Abstract
The bacterial host factor Hfq is an RNA-binding protein that facilitates the interaction of mRNAs with small regulatory RNAs (sRNAs) Hfq self-assembles as hexameric toroid and functions by simultaneously binding U-rich regions of sRNA on one face and A-rich regions of mRNA on the other face. More recently a third site on the lateral rim of Hfq has been implicated in binding RNA, though little is known about the molecular details of RNA-binding at this site. Due to Hfq’s broad functional role in binding RNAs, it has been demonstrated to be required for numerous physiological pathways, including stress response, quorum sensing, and biofilm formation. Through structural similarities, Hfq has been identified as the bacterial branch of the Sm superfamily of proteins. While bacteria typically have one Hfq homolog, eukaryotes encode multiple Sm paralogs, which oligomerize as heteroheptamers through a complex chaperoned assembly pathway. The eukaryotic Sm and Sm-like (LSm) proteins are involved in splicing and various other mRNA processing pathways.
Several species of bacteria have been identified as having two putative Hfq paralogs. Such proteins could provide valuable insight into the evolutionary transition from bacterial Hfq to eukaryotic Sm. Currently though, little is known about the physiological role of additional Hfq paralogs and no structural information is available. This work presents a new phylogenetic analysis of Hfq sequences, demonstrating the presence of 2 (or more) Hfq paralogs in bacteria of diverse lineages including the deep-branching extremophilic Aquificae. Here, the structures of two such paralogs, from Aquifex aeolicus, have been determined to atomic resolution, offering the first instance of two genuine Hfq proteins from a single species. Atomic-level details of the conserved lateral rim site have also been revealed through co-crystallization of Hfq1 with U6 RNA. Intriguingly this lateral rim site is not conserved in Hfq2, which was also found to bind RNAs with pH dependence and co-purify with endogenous DNA. These results suggest that Hfq1 and Hfq2 function independent, likely in different cellular pathways.
A single hfq open reading frame with two putative linked Sm domains (HfqN- HfqC) has also been identified in the α-proteobacterium Novosphingobium aromaticivorans. Here N. aromaticivorans (Nar) Hfq has been expressed, purified and biochemically characterized. The oligomeric state of Nar Hfq in solution is a trimer, likely assembling as a ring of alternating HfqN and HfqC subunits. This represents the first known instance of “pseudo-heteromeric” Hfq, which poses intriguing consequences for differences in RNA-binding affinities. Initial efforts to crystallize Nar Hfq were hindered by the presence of an ~40 residue proline-rich N-terminal tail which appears to increase the conformational heterogeneity of the Nar Hfq crystals. Ongoing efforts are currently aimed at crystallization of ΔN-Nar Hfq construct.
