On-line kinetic model discrimination for optimized surface plasmon resonance experiments.

In order to improve the throughput of surface plasmon resonance-based biosensors, an on-line iterative optimization algorithm has been presented aiming at reducing experimental time and material consumption without any loss of confidence on kinetic parameters [De Crescenzo (2008) J. Mol Recognit., 21, 256-66.]. This algorithm was based on a simple Langmuirian model to compute the confidence and predict optimal injections. However, this kinetic model is not suitable for all interactions, as it does not include mass transfer limitation that may occur for fast interaction kinetics. If a simple model was to be used when this phenomenon influenced the interactions, kinetic parameters would be biased. On the other hand, we show in this paper that data analysis with a kinetic model including a mass transfer limitation step would lead to longer experiments and poorer confidence if the interactions were simple. So, in this manuscript, we present an on-line model discrimination and optimization approach to increase the throughput of surface plasmon resonance biosensors.