Precision and accuracy of single-molecule FRET measurements-a multi-laboratory benchmark study

Abstract

Single-molecule Förster resonance energy transfer (smFRET) is increasingly being used to determine distances, structures, and dynamics of biomolecules in vitro and in vivo. However, generalized protocols and FRET standards to ensure the reproducibility and accuracy of measurements of FRET efficiencies are currently lacking. Here we report the results of a comparative blind study in which 20 labs determined the FRET efficiencies (E) of several dye-labeled DNA duplexes. Using a unified, straightforward method, we obtained FRET efficiencies with s.d. between ±0.02 and ±0.05. We suggest experimental and computational procedures for converting FRET efficiencies into accurate distances, and discuss potential uncertainties in the experiment and the modeling. Our quantitative assessment of the reproducibility of intensity-based smFRET measurements and a unified correction procedure represents an important step toward the validation of distance networks, with the ultimate aim of achieving reliable structural models of biomolecular systems by smFRET-based hybrid methods.

Fig. 1. Schematic of the FRET standard molecules.

Double-stranded DNA was labeled with a FRET pair at 15-bp or 23-bp separation for the “lo” and “mid” samples, respectively (sequences are provided in the Methods). The accessible volumes (AVs) of the dyes (donor, blue; acceptor, red) are illustrated as semi-transparent surfaces and were calculated with freely available software. The mean dye positions are indicated by darker spheres (assuming homogeneously distributed dye positions; Supplementary Note 3). The distance between the mean dye positions is defined as R_MP,model. Calculated values for R_MP,model and the errors obtained by varying parameters of the AV model are shown (Supplementary Note 3). The B-DNA model was generated with Nucleic Acid Builder version 04/17/2017 for Amber.

Fig. 2. Stepwise data correction for 1-lo and 1-mid samples.

a–d, Workflow for correction of the confocal data for background (a → b); leakage (factor α); and direct excitation (δ) (b → c), excitation, and detection factors (β, γ) (c → d). e–h, Workflow for correction of TIRF data for background and photobleaching by selection of the prebleached range (e → f); leakage; and direct excitation (f → g), detection, and excitation factors (g → h). The efficiency histograms show a projection of the data with a stoichiometry between 0.3 and 0.7. The general terms “stoichiometry” and “FRET efficiency” are used in place of the corresponding specific terms for each correction step. Donor (D)-only, FRET, and acceptor (A)-only populations are specified.

Fig. 3. Summary of the results of the intensity-based methods.

a, Confocal measurements. b, TIRF measurements. Note that some laboratories performed measurements with both methods. The mean ± s.d. is depicted in the upper portion of each plot. Dashed lines indicate mean values (summarized in Supplementary Table 4). Example correction factors are given in Supplementary Table 3.

Fig. 4. Mean interdye distances determined from 19 〈E〉 values measured in 16 different labs.

a,b, R_〈E〉 for samples 1 (a) and 2 (b). c,d, R_MP for samples 1 (c) and 2 (d). Data are shown as individual values (colored symbols) and as the mean (black dots) and s.d., assuming R₀ = 62.6 Å and R₀ = 68.0 Å for samples 1 and 2, respectively. The black bars at the top of each plot indicate the static model values and their error (determined by variation of model parameters); see Supplementary Table 4 for values. The depicted errors include only the statistical variations of the FRET efficiencies, and do not include the error in the Förster radii; thus these errors represent the precision of the measurement, but not the accuracy. Exp., experimental.

Fig. 5. Error propagation of experimental uncertainty.

a, R_DA uncertainty contributions from the experimental correction factors: ∆R_γ (gamma factor), ΔR_bgD and ΔR_bgA (background), ∆R_α (leakage), ∆R_δ (direct excitation), and total uncertainty with known R₀; crosses indicate the uncertainty of experimental values of R_〈E〉 across the labs. b, Uncertainty in R_DA (black line) based on the efficiency-related uncertainty (gray line) and the uncertainty for determining R₀ (blue line). Here we used the following uncertainties, which were determined for the confocal-based measurements on sample 1: ΔR₀/R₀ = 7%, Δγ/γ = 10%, ΔI^(BG)/I = 2%, Δα/α = 10%, and Δδ/δ = 10%. Absolute values are presented in Supplementary Table 3.