Redundancy of primary RNA-binding functions of the bacterial transcription terminator Rho.

The bacterial transcription terminator, Rho, terminates transcription at half of the operons. According to the classical model derived from in vitro assays on a few terminators, Rho is recruited to the transcription elongation complex (EC) by recognizing specific sites (rut) on the nascent RNA. Here, we explored the mode of in vivo recruitment process of Rho. We show that sequence specific recognition of the rut site, in majority of the Rho-dependent terminators, can be compromised to a great extent without seriously affecting the genome-wide termination function as well as the viability of Escherichia coli. These terminators function optimally only through a NusG-assisted recruitment and activation of Rho. Our data also indicate that at these terminators, Rho-EC-bound NusG interaction facilitates the isomerization of Rho into a translocase-competent form by stabilizing the interactions of mRNA with the secondary RNA binding site, thereby overcoming the defects of the primary RNA binding functions.

Suppression of in vivo Rho-dependent transcription termination defects: evidence for kinetically controlled steps.

The conventional model of Rho-dependent transcription termination in bacteria requires RNA-dependent translocase activity of the termination factor Rho as well as many kinetically controlled steps to execute efficient RNA release from the transcription elongation complex (EC). The involvement of the kinetically controlled steps, such as RNA binding, translocation and RNA release from the EC, means that this termination process must be kinetically coupled to the transcription elongation process. The existence of these steps in vivo has not previously been delineated in detail. Moreover, the requirement for translocase activity in Rho-dependent termination has recently been questioned by a radical view, wherein Rho binds to the elongating RNA polymerase (RNAP) prior to loading onto the mRNA. Using growth assays, microarray analyses and reporter-based transcription termination assays in vivo, we showed that slowing of the transcription elongation rate by using RNAP mutants (rpoB8 and rpoB3445) and growth of the strains in minimal medium suppressed the termination defects of five Rho mutants, three NusG mutants defective for Rho binding and the defects caused by two Rho inhibitors, Psu and bicyclomycin. These results established the existence of kinetically controlled steps in the in vivo Rho-dependent termination process and further reinforced the importance of 'kinetic coupling' between the two molecular motors, Rho and RNAP, and also argue strongly that the Rho translocation model is an accurate representation of the in vivo situation. Finally, these results indicated that one of the major roles of NusG in in vivo Rho-dependent termination is to enhance the speed of RNA release from the EC.