Monitoring protein conformational dynamics with smFRET

Detecting protein conformational dynamics remains challenging. Therefore, we used single-molecule Förster resonance energy transfer (smFRET) to investigate how the N— and C—termini of Dbp2 regulate its core domain and enable full helicase activity. Biochemical assays revealed that both the full-length Saccharomyces cerevisiae Dbp2 protein and truncated variants containing the N— or C—terminal tails retain RNA-binding activity, with high affinity for G-quadruplex (G4) RNA. In contrast, the core domain alone had no binding capacity. Previous smFRET studies demonstrated that the Dbp2 core domain exhibits conformational dynamics in solution. Without ligands, its RecA1 and RecA2 domains adopt an open conformation that shifts to a closed state upon binding to ATP nalogs and RNA substrates. Notably, locking the core domain in a closed state through an engineered intramolecular disulfide bond abolishes its helicase activity, indicating that the enzymatic function is strictly dependent on ATP- and RNA-regulated conformational dynamics.

Bioinformatic analysis was conducted to identify cysteine residues with high intrinsic reactivity (i.e., low natural abundance and high sulfhydryl reactivity) in wild-type Dbp2. Using an AlphaFold-predicted full-length Dbp2 structural model, we measured distances between the C_β atoms of candidate cysteine residues in the core domain and those in the intrinsically disordered N— and C—terminal regions. Cysteine pairs with FRET distances of 3–8 nm were selected for subsequent mutagenesis. Site-directed mutagenesis was then performed to introduce cysteine residues at specific terminal sites while replacing native cysteines that were either unsuitable for labeling or functionally non-essential with alanine. This strategy produced two monocysteine mutants, Dbp2^T2C/C141 and Dbp2^C373/Y546C, with a single cysteine pair between the two domains of interest. Mutant proteins were purified to high homogeneity using a nickel-affinity chromatography protocol identical to that for the wild-type protein. An electrophoretic mobility shift assay confirmed that both mutants retained helicase activity comparable to that of the wild-type protein in unwinding DNA/RNA hybrid strands. Under reducing conditions, mutants were site-specifically labeled with Cy3 and Cy5 maleimide fluorophores. For smFRET imaging, labeled proteins were immobilized on the glass surface of a microfluidic channel via a biotin-PEG-streptavidin-biotin-anti-His antibody bridge, and experiments were performed in the presence of RNA G4 substrates.

In the absence of RNA G4, the FRET histograms of Dbp2^T2C/C141 and Dbp2^C373/Y546C presented a wide and flat peak-shaped distribution, and the main FRET efficiency (E_FRET) peaks were 0.65 and 0.59, respectively. Upon the addition of RNA G4, the FRET distribution peak shape narrowed, and the main E_FRET peaks increased to ~0.78 and 0.76, respectively. These results suggest that RNA G4 brings the N— and C—terminal domains of Dbp2 close to the core domains of RecA1 and RecA2, respectively. This proximity stabilizes the protein’s conformation and fully activates its helicase activity.

Combining site-directed mutagenesis, fluorescently labeled proteins, and smFRET, we systematically investigated how the intrinsically disordered N— and C—terminal domains of Dbp2 regulate its core domain to achieve full helicase activity upon binding to nucleic-acid ligands. This experimental strategy offers a new paradigm for studying structure-function relationships in other proteins, particularly those containing intrinsically disordered regions.