Fluorescence Correlation Spectroscopy as a Versatile Method to Define Aptamer-Protein Interactions with Single-Molecule Sensitivity

Abstract

Aptamers are folded oligonucleotides that selectively recognize and bind a target and are consequently regarded as an emerging alternative to antibodies for sensing and therapeutic applications. The rational development of functional aptamers is strictly related to the accurate definition of molecular binding properties. Nevertheless, most of the methodologies employed to define binding affinities use bulk measurements. Here, we describe the use of fluorescence correlation spectroscopy (FCS) as a method with single-molecule sensitivity that quantitatively defines aptamer-protein binding. First, FCS was used to measure the equilibrium affinity between the CLN3 aptamer, conjugated with a dye, and its target, the c-Met protein. Equilibrium affinity was also determined for other functional aptamers targeting nucleolin and platelet-derived growth factors. Then, association and dissociation rates of CLN3 to/from the target protein were measured using FCS by monitoring the equilibration kinetics of the binding reaction in solution. Finally, FCS was exploited to investigate the behavior of CLN3 exposed to physiological concentrations of the most abundant serum proteins. Under these conditions, the aptamer showed negligible interactions with nontarget serum proteins while preserving its affinity for the c-Met. The presented results introduce FCS as an alternative or complementary analytical tool in aptamer research, particularly well-suited for the characterization of protein-targeting aptamers.

Figure 1

(a) Schematic representation of the principle of FCS binding measurements between the dye-conjugated aptamer and the target protein. (b) FCS autocorrelation curves measured for 1 nM CLN3-atto633, alone (magenta) and exposed to 200 nM of c-Met (green). The black lines represent the results of the fitting with a model comprising a single diffusing species (eq 1). The graphical representation of the diffusion time (τ_D) obtained by the fit is highlighted for the magenta curve.

Figure 2

(a) FCS autocorrelation curves, normalized to 1, measured on solutions containing 100 pM of CLN3-atto633 and an increasing concentration of c-Met, from 0 to 100 nM. The change of the curves with c-Met concentration is highlighted by the arrow. (b) Corresponding data showing the calculated fraction of CLN3-atto633 bound to c-Met at increasing protein concentration (black). The red line shows the result of the fitting with a binding model (eq 5). The calculated fraction of protein-bound aptamer is shown for the control aptamer G5mut-atto633 exposed to c-Met (magenta) and for CLN3-atto633 exposed to IgG (green). Error bars: mean ± st. dev., 3 repetitions.

Figure 3

(a) FCS autocorrelation curves, normalized to 1, measured on a solution containing 100 pM of CLN3-atto633 and 5 nM of c-Met. The curves are measured at increasing times, from 0 to 120 min, starting from the initial mixing of the solution. The change of the curves in time is highlighted by the arrow. (b) Observed Δτ_D of the FCS autocorrelation curves measured at increasing times from the initial mixing of a solution containing 100 pM of CLN3-atto633 and 7 nM of c-Met (black). The red line shows the result of the fitting with an exponential model, yielding the k_equil parameter (eq 6). (c) k_equil values, obtained by the same procedure shown in panel (b), measured on solutions containing 100 pM of CLN3-atto633 and an increasing concentration of c-Met between 2 and 9 nM (black). The red line shows the result of the fitting with a linear model (eq 7).

Figure 4

Observed τ_D values obtained on solutions containing 1 nM of CLN3-atto633 and an increasing concentration of HSA. The highlighted region indicates the range of HSA concentrations found in human serum.

Figure 5

(a) FCS autocorrelation curves, normalized to 1, measured on heterogeneous solutions containing 100 pM of CLN3-atto633 and an increasing concentration of c-Met, from 0 to 100 nM, in the presence of HSA ∼210 mg/mL, IgG ∼48 mg/mL, Tf ∼12 mg/mL, and Fibr ∼12 mg/mL. The change of the curves with c-Met concentration is highlighted by the arrow. (b) Corresponding data showing the observed average τ_D at increasing c-Met concentration in the presence of serum proteins (black). The red line shows the result of the fitting with a binding model (eq 5). The results obtained from a control experiment with the G5mut-atto633 are shown (green). Error bars: mean ± st. dev., 3 repetitions.