Dissecting RNA chaperone activity

Abstract

Many RNA-binding proteins help RNAs to fold via their RNA chaperone activity. This term has been used widely without accounting for the diversity of the observed reactions, which include complex events like restructuring of misfolded catalytic RNAs, promoting the assembly of RNA-protein complexes, and mediating RNA-RNA interactions. Proteins display very diverse activities depending on the assays used to measure RNA chaperone activity. To classify proteins with this activity, we compared three exemplary proteins from E. coli, host factor Hfq, ribosomal protein S1, and the histone-like protein StpA for their abilities to promote two simple reactions, RNA annealing and strand displacement. The results of a FRET-based assay show that S1 promotes only RNA strand displacement while Hfq solely enhances RNA annealing. StpA, in contrast, is active in both reactions. To test whether the two activities can be assigned to different domains of the bipartite-structured StpA, we assayed the purified N- and C- terminal domains separately. While both domains are unable to promote RNA annealing, we can attribute the RNA strand displacement activity of StpA to the C-terminal domain. Correlating with their RNA annealing activities, only Hfq and full-length StpA display simultaneous binding of two RNAs, suggesting a matchmaker-like model for this activity. For StpA, this "RNA crowding" requires protein-protein interactions, since a dimerization-deficient StpA mutant lost the ability to bind and anneal two RNAs. These results underline the difference between the two reaction types, making it necessary to distinguish and classify proteins according to their specific RNA chaperone activities.

FIGURE 1.

RNA chaperone activities in RNA annealing and strand displacement can be analyzed in a fluorescence-based assay. (A) Annealing of two fluorophore-labeled RNA 21mers yields a FRET-signal that is reduced upon RNA chaperone-facilitated strand displacement with a competitor RNA. (B) In phase I, 5 nM each of two fully complementary RNAs (Cy5–21R+, Cy3–21R−) were annealed in a microplate reader in the absence or presence of 1 μM of protein. The donor (Cy3) and acceptor (Cy5) fluorescence emissions were quantified every second; the FRET index was calculated as F_Cy5/F_Cy3 and normalized at t_180s. Hfq and StpA accelerated this reaction. Phase II was initiated by the injection of an excess of nonlabeled competitor RNA, and either RNA annealing continued (L7/L12 and Hfq) or the tested protein induced strand displacement (S1 and StpA). Representative curves are shown. (C) Comparison of the observed reaction constants for RNA annealing in phase I and II (k _ann,1 and k _ann,2) and strand displacement (k _SD).

FIGURE 2.

Only dimerization-competent StpA enhances RNA annealing, whereas the ability to facilitate strand displacement resides in the C-terminal domain. (A) The N- and C-terminal domains, as well as the position of the L30P mutation, are indicated in the protein sequence of E. coli StpA. (B) Wild-type StpA (wt), StpA L30P, the N-terminal, and the C-terminal fragment were assayed for their RNA annealing and strand displacement activity as described in the Materials and Methods. (C) Kinetic evaluation of the reactions. Only StpA wt accelerates annealing; StpA wt, StpA L30P, and StpA C-term facilitate strand displacement in phase II. (D) Coomassie blue–stained SDS-polyacrylamide gels showing the analysis of protein–protein cross-link reactions. X-link indicates the incubation with cross-link agents EDC/NHS. The calculated molecular weights of the monomers are as follows: StpA, 15 kDa; StpA L30P, 15 kDa; N-terminal domain, 9 kDa; and C-terminal domain, 5 kDa.

FIGURE 3.

Two noncomplementary short RNAs bind simultaneously to Hfq or dimerization-competent wild-type StpA. (A) Monitoring “RNA crowding.” Two noncomplementary fluorophore-labeled RNAs give a FRET signal when they are in close proximity because of simultaneous binding to a protein. In this model, dual RNA binding is mediated by two StpA proteins that dimerize via their N-terminal domain. (B) In a microplate reader, the noncomplementary RNAs Cy5–21R+ and Cy3-Duplex− were injected into buffer containing the respective proteins. The FRET index was measured and calculated as described in the Materials and Methods but not normalized. Only incubation with StpA and Hfq yielded quantifiable reaction curves; the observed rate constants for dual binding k _db are 0.028±0.002 sec⁻¹ and 0.027±0.002 sec⁻¹.