Abstract
A "σ cycle" in which the initiation factor σ associates with RNA polymerase (RNAP) core enzyme to permit transcription initiation and dissociates from RNAP core enzyme to permit transcription elongation, has been proposed to occur and to be an essential step for σ-exchange, with all principal σ factors from all bacteria. These proposals were based on studies of the principal σ factor of Escherichia coli, σ(70), which generally, albeit not obligatorily, is released from RNAP upon the transition from transcription initiation to elongation. Here, we show that, in contrast to E. coli σ(70)(,) the Bacillus subtilis principal σ factor, σ(A), is not released and is retained on RNAP core throughout transcription elongation. We further show that a mutant E. coli σ(70) derivative lacking σ region 1.1 (σ R1.1) is not released and is retained on RNAP core throughout transcription elongation. We also observe that B. subtilis σ(A) and the mutant E. coli σ(70) derivative lacking σ R1.1 interact much more stably with RNAP than full-length E. coli σ(70). Our results indicate that the σ cycle is not a universal phenomenon in bacteria.