Active RNA cycling on Hfq

Post-transcriptional regulation of gene expression in bacteria involves numerous so-called sRNAs, a heterogeneous class of RNAs that predominantly acts by an antisense mechanism. This mode of action can result in up- or down-regulation of gene expression. In many bacteria, the homohexameric Hfq protein (a bacterial homolog of eukaryotic Lsm proteins, carrying an Sm motif) is a key player required for sRNA-mediated control. E.g., the majority of enterobacterial sRNAs requires Hfq for regulatory efficiency, and mutations/ deletions in Hfq have strong pleiotropic phenotypes. Theses are often associated with compromised stress responses and loss of virulence in pathogens.

The Hfq requirement may be explained by several established effects: RNA-RNA binding rate enhancement, RNA chaperone activity, protection of sRNAs from degradation - or any combination thereof. In all these mechanisms, tight binding to RNAs is needed. Indeed, many studies have supported binding to sRNAs and mRNAs at low Kd-values (sub- to mid-nanomolar). However, in any changing internal/ external environment, RNA pools change rapidly, and newly induced sRNAs must be able to quickly gain access to their share of the Hfq pool to carry out there functions adequately. Hence, rapid exchange of RNAs on Hfq needs to take place, on a time scale known for sRNA regulation in vivo (1-2 minutes).

I will address how rapid cycling of RNAs on Hfq works. Previous work implicitly assumed that resident RNAs have to dissociate from Hfq before a second RNA can bind (passive cycling). In contrast, we show that a second RNA can bind to an RNA-Hfq complex, followed by rearrangent steps of the two RNAs on the six binding surfaces, until dissociation of one of the RNAs occurs. This active cycling mechanism implies that 1) exchange rates are driven by RNA concentration and 2) the time scale of exchange shrinks to the minute range – in line with regulation in vivo, and 3) RNAs therefore can rapidly "equilibrate" on the entire Hfq pool. Passive cycling cannot achieve this since it is inherently limited by the very low RNA-Hfq dissociation rate constants which gives half-lives of complexes exceeding one hour. We suggest that RNA cycling on Lsm proteins follows the same model.