Move over and bind

In bacteria, all RNA synthesis is carried out by a multisubunit enzyme composed of two small α subunits, large β and β' subunits, and the σ subunit, of which there are several varieties. It exists in two forms — the core enzyme (α₂ββ') and the holoenzyme (α₂ββ'σ). While the core polymerase can elongate RNA, only the holoenzyme can initiate transcription at specific promoter sites. The promoter consists of two highly conserved sequences (-10 and -35 with respect to the transcription start site +1), each six nucleotides long and separated from each other by ~17 nucleotides. Thus, the σ subunit recognizes the promoter but only when it is bound to the core enzyme. The σ⁷⁰ subunit is composed of a conserved autoinhibitory N-terminal domain (region 1) and two DNA binding domains (regions 2.4 and 4.2) that are involved in recognition of the -10 and -35 promoter elements. These and other findings have led to a model in which region 1 sterically blocks access of the DNA binding domains of σ⁷⁰ to promoter DNA in the absence of core polymerase. Binding of the core enzyme to σ⁷⁰ was proposed to induce movement of region 1 to unmask the DNA binding domains of σ⁷⁰. A schematic diagram of this conformational change is depicted, but no structural details are known.

To test whether core enzyme does induce conformational changes in σ⁷⁰, Heyduk and coworkers (Callaci, S., Heyduk, E., & Heyduk, T. Mol. Cell 3, 229–238; 1999) used a combination of site-directed mutagenesis and luminescence resonance energy transfer measurements to introduce luminescence donors and acceptors in different conserved regions of the protein and monitor the distances between them in the free σ⁷⁰ and in the holoenzyme. Their studies have provided direct evidence for a major movement of region 1 induced by core binding. Region 1 moves by ~20 Å away from the DNA binding domains. In addition, their experiments have suggested another way in which core binding could regulate promoter recognition by σ⁷⁰.

In the absence of core, regions 2.4 and 4.2 of σ⁷⁰ are too close together to interact simultaneously with the -10 and -35 regions on the DNA. The distance between regions 2.4 and 4.2 is increased by ~15 Å upon core binding which is more compatible with binding to the promoter. Thus core binding to σ⁷⁰ has at least two effects with regard to promoter binding: it displaces the autoinhibitory region 1 and it increases the spacing between the two DNA binding domains. Both of these changes increase the affinity of the holoenzyme for the promoter DNA.

One interesting question is whether the movements observed within σ⁷⁰ upon core binding are interdependent. For example, does the modulation of spacing between DNA binding regions occur in the absence of region 1? If not, can this movement be restored if domain 1 is added back in trans? Answers to these questions would address the role of region 1 in promoter recognition by regions 2.4 and 4.2. In one scenario, region 1 holds the DNA binding domains together in the absence of core and when core binds they spring apart. Alternatively, region 1 could push the DNA binding regions apart upon core binding. Clearly, even in this 'simple' system, many important dynamic properties have yet to be revealed.