Interactions between proteins are at the heart of almost all cellular mechanisms. SKi Pro has been used by researchers around the world to study intracellular protein binding events in order to elucidate the mechanisms of specific transcription factors, apoptotic pathways and single stranded DNA binding proteins, to name but a few examples. Below, a screening experiment performed on the SKi Pro X10 using refractive index detection is described.
Here researchers were interested in engineering the variable region of a soluble proteins to bind to a certain interleukin with the goal of adding that protein into an antibody molecule like scaffold. The goal in the SKi Pro experiments is to quantify the binding properties to a therapeutically important interleukin.
They had developed a phage display approach to producing potential binders and used e. coli to express variable regions of the protein (marked as 'V') tagged with maltose binding protein (marked as 'MBP'). They used SKi Pro to screen for good binders from the bacterial supernatant solutions.
Biotinylated interleukin was linked to a streptavidin SKi sensor prior to the binding experiments here. The plate of 96 bacterial supernatant solutions was spun down and placed in the AutoPrep and serially introduced into the interleukin surface (A in the figure). Binding was monitored in real time with ΔOPD values between 0.1—0.3 nm (see OPD explained).
As the binding is weak, with essentially complete dissociation in 10 minutes, there was no need to regenerate the surface, as is often performed with e.g. mild acid.
The kinetic traces of the sort shown above are fit with a two-state model. The amplitude of the association is reported as ΔOPD and indicates how much binding takes place. However as the concentrations of each individual analyte in the supernatant is unknown, this is not sufficient to indicate a good binder. Therefore, the dissociation rate (koff, or simply off rate) is fit with each trace. Almost always a slower off rate indicates a better binder.
The offrates span several order of magnitude. Therefore, pkoff (the -log10(koff)) for each well position is shown. Larger numbers mean slower off-rates which almost always means better binding.
The error bars shown on the right are the fit uncertainties that account for the noise inherent in the data and the suitability of the model.