Generality of the branched pathway in transcription initiation by Escherichia coli RNA polymerase.

Transcription initiation has been assumed to be a multi-step sequential process, although additional steps could exist. Initiation from the T7A1 promoter, in particular, apparently behaves in vitro in a manner that can be fully explained by the sequential pathway. However, initiation from the lambda P(R)AL promoter has been shown to follow a branched pathway from which a part of the enzyme-promoter complex is arrested at the promoter raising the question as to which mechanism is general. We found that a moribund complex, characteristic of the arrested branch, is formed at the T7A1 promoter, especially in low salt condition indicating that the initiation mechanism for this promoter is also branched. The results of DNA footprinting suggested that holoenzyme in the moribund complex is dislocated on DNA from the position of productive complex. However, only a small fraction of the binary complex becomes arrested at this promoter, and the interconversion between subspecies of binary complex is apparently more reversible than at the lambda P(R)AL promoter, which explains why the reaction pathway appears to be sequential. These findings suggest a generality of the branched pathway mechanism, which would resolve contradictory observations that have been reported for various promoters.