Hi Xianchi, First of all: a very slow dissociation rate can also be an artifact: the analyte can be simply precipitating on the surface. You do need to rule this out by proper controls.
But there is no rule saying that 10e-6 s-1 off rate is not realistic. Even with protein-small molecule binding, one can get extremely slow dissociation if the interaction is very strong. For example, the dissociation of biotin from streptavidin has a rate constant of the 10e-6 s(-1) order. A very slow dissociation rate is often (if not always) correlated with very tight binding (for example KD in picomolar range or even smaller). What is your calculated KD? The binding phase of the BIAcore curves should also reflect the fact that the off rate is low (the Kon-obs has an off rate term). We have measured some pico molar bindings with BIAcore in the past. It is doable, but difficult. You may also want to consider in-solution methods (so that you do not need to worry about artifacts caused by the surface), such as ITC (by competition method), or fluorescence-based methods. For BIAcore, when working with very strong bindings (KD <100pM), there are a few things to consider: 1) You may find great difficulties regenerating the chip - probably the biggest concern with BIAcore (considering the ridiculous price of chips). 2) How much RU should you conjugate to the chip? With strong interactions, as low as possible amount of your ligand should be labeled on the chip, for 4 purposes: a) to make regeneration easier; b) to reduce rebinding effect; c) to reduce mass transfer effect; d) to make sure you do not take away significant amount of analyte from the solution (discussed in 3)). 3) If you plan to span the KD range with the analyte, then if the KD is in picomolar molar range, you are supplying the surface with extremely dilute analyte solutions. In such case, you need to calculate if your flow rate is high enough, to compensate the loss of solute due to binding to surface, otherwise the real concentration of the analyte in the mobile phase will be much lower than the assumed analyte concentration. 4) The association phase of the BIAcore experiment is also affected by the dissociation rate. The observed binding rate Kon-obs contains a Koff term. When you are loading the analyte at near KD concentrations, the binding will take similar amount of time as the dissociation phase to reach plateau. The Kon-obs also contains a concentration terml, so the time required for reaching plateau will be shorter and shorter when you load with higher concentrations of the analytes. 5) The only proper way of labeling chip for slow dissociations is by covalent means. HTH, Zhijie From: xianchi dong Sent: Wednesday, February 20, 2013 12:03 PM To: CCP4BB@JISCMAIL.AC.UK Subject: [ccp4bb] off topic question BIAcore problem Dear all, Recently I have measure a set of kinetic data of a receptor ligand interaction using BIAcore 3000. To my surprise, the dissociation rate is very low (~ 10e-6). During the measurement, I use a long dissociation time (2 hours) .I repeat several time which give very similar results. So I am wondering if the BIAcore can measure such a low off-rate kinetic. What is the limitation of BIAcore? Any review about that? Thanks in advance. Xianchi Dong Research Fellow Children's Hospital Boston Harvard Medical School
