Determining Kd With an Alpha Assay

Overview

The Kd of a protein-protein interaction may be determined with Alpha assays using one of two approaches: saturation curves or competition assays. Of these, competition assays are applicable to a wider range of assay conditions.

Saturation curves for determination of Kd

In limited situations, saturation curve assays can be performed to determine the affinity constant (Kd) for the protein-protein interaction in Alpha format. In a saturation assay, one binding protein is titrated across a range of concentrations to generate a saturation curve. The Kd can be derived from the data as the protein concentration at which half maximal signal is reached. Saturation curves can be used to determine Kd only if the assay meets the following criteria:

1. The Kd for the protein-protein interaction is far enough below the binding capacity of each bead so that all protein concentrations used to derive the Kd (usually up to 5X Kd) fall below the bead binding capacities. (This typically means that the Kd needs to be in the sub-nanomolar range, though some beads have a binding capacity up to ~300 nM – refer to Table below.)
2. There is an excess of the binding partner that will be varied in comparison to the concentration of the fixed binding partner. For example, if you are varying the concentration of Protein Y in the saturation curve, Protein Y should be in at least 10X excess (molar concentration) compared to Protein X. Otherwise, there is a ligand-depletion (protein-depletion), meaning that the percentage of “bound” protein can be above 10%. In this case, the amount of free ligand (protein) cannot be estimated as being equal to total ligand (protein) for the curve fitting equation.

If the expected Kd is not below the binding capacity of the beads used in the assay, a competition binding assay must be used to determine the Kd for the protein-protein interaction. The reason for this is that the bead system will become saturated when the binding capacity is reached, and two competing equilibria will be occurring: excess protein will be dissociating the interaction between the two beads, decreasing signal (the hooking effect), while the excess protein may simultaneously be driving the interaction between the two proteins, increasing signal. The curve generated will be a composite of these two equilibria, once the binding capacity has been reached. This composite curve may appear to reach saturation, may appear to begin to hook (decrease) slightly, or may continue increasing without reaching saturation (depending on the actual Kd, the concentrations of proteins being tested, and the binding capacities of the beads used in the assay).

Symptoms that a saturation curve cannot be used to accurately predict the Kd:

1. Underestimated Kd in Alpha assay. If the Alpha data is giving a Kd that is below what is expected for the protein-protein interaction, it is possible that bead saturation is skewing the data. A competition binding curve should be tested for Kd determination.
2. Kd is calculated to be ~5 nM when using streptavidin beads. Because the streptavidin-biotin interaction is a relatively tight interaction, the binding capacity of these beads is fairly predictable. If the calculated Kd is between 5 and 10 nM and streptavidin beads are being used, it is likely the binding capacity of the beads has been exceeded. A competition binding curve should be tested for determination of Kd.

Example data: 
An excellent example of the use of saturation curves to determine Kd in an Alpha assay format is in a paper by Cassel et al. (2010). In this work, the Kd of the interaction between TAR DNA binding protein 43 (TDP-43) and biotinylated single-stranded TAR DNA and 6 TG repeats was determined to be 0.75 nM and 0.63 nM, respectively. Association and dissociation rates were also calculated and used to derive Kd values. In both cases, the Kd derived was consistent with previous reports using other methods.

In the example below, saturation curves were used to determine the Kd for the interaction between biotin-murine EGF and Fc-fusion human EGFR (Figure 1). Streptavidin Donor beads and Protein A AlphaLISA Acceptor beads were used in the assay. The expected Kd is ~2.8 nM as determined by a radioligand binding assay (Lax, I. et al., 1988). In the protocol for the data below, a 3-step assay was performed as follows. 1) biotinylated EGF and EGFR-Fc addition, 60 minute incubation. 2) Protein A AlphaLISA Acceptor bead addition, 60 minute incubation. 3) Streptavidin Donor bead addition, 30 minute incubation followed by reading the plate. The streptavidin Donor beads were added last (after the second incubation) to prevent the multi-biotinylated EGF from bridging two streptavidin Donor beads before other interactions could occur.