Fluorescence anisotropy binding assays

Fluorescence anisotropy was used to measure the binding of synthetic VPg peptides (labelled at the N terminus with fluorescein isothiocyanate (FITC)) to HEAT-1 domains from eIF4G1, eIF4GII and DAP5. Measurements were performed at 29°C using a Spectramax i3 Spectrometer with excitation/emission wavelengths of 485 and 535 nm respectively, slit widths set to 20 nm (excitation) and 25 nm (emission) and an integration time of 400 ms. Samples were loaded into non-binding surface black half area plates (Corning) using a MES binding buffer composed of 25 mM MES pH 6.5, 150 mM NaCl, 3 mM DDT, 0.05% Tween-20.

N-terminally FITC labelled peptides MNV VPg 104–124 (VGPSWADDDRQVDYGEKINFE-COOH) and MNV VPg 108–124 were purchased at >95% purity from ChinaPeptides Co., Ltd. (Shanghai, China). In both cases the FITC group was attached with via a 6-aminohexanoic acid group to the N terminus of the peptide. Peptides were dissolved in binding buffer and titrated to pH 6.5 using a small volume of concentrated NaOH. The final peptide concentration was kept at a constant 10 nM across all binding experiments. The proteins tested for binding–untagged eIF4GI HEAT-1 truncated (748–993), untagged eIF4GII HEAT-1 (745–1003), MBP tagged eIF4GII HEAT-1 (745–1003) and DAP5 HEAT-1 (61–323)– were all dissolved in MES binding buffer. For each experiment a two-fold dilution series of each HEAT-1 protein was performed (typically starting at 20 or 40 μM) using MES binding buffer as diluent. For each dilution series, a no-protein control was performed in which binding buffer replaced protein. Each measurement typically involved two independent dilution series performed in parallel. The average fluorescence anisotropy values for the no-protein controls were subtracted from the fluorescence anisotropy values for each point in the titration giving the difference (ΔFA) for each data point. The ΔFA values for each point in the titration were used to calculate the dissociation constant (K_D) for each interaction tested. This was done by fitting all of the titration data to a single-site binding model (ΔFA = FA_max*[protein]/(K_D + [protein]) where FA_max is the maximal change in fluorescence anisotropy) in Prism6 (GraphPad Software).

Titration of up to 40 μM eIF4GI into an irrelevant peptide ([FITC]-GSSHRYFLERGLESATSL; [64]) did not result in significant increases in fluorescence anisotropy (ΔFA < 0.004), confirming the specificity of the interaction with the MNV VPg peptides.