Olaf Rolinski
Senior Lecturer

Research

 Time-resolved fluorescence spectroscopy,  modelling fluorescence kinetics, single molecule detection, lifetime sensors, yoctomolar sensors, proteins activities and misfolding

Fluorescence lifetime sensors based on FRET

We have successfully used FRET for developing transition metal ion sensors based on the dye→metal ion and dye→ligand:metal ion FRET, achieving ppb sensitivity. The experimental data obeyed the Förster-type equations for the random distributions of acceptors in bulk solutions, but some deviations were observed for the donor-acceptor systems placed in the polymer, and in our work on glucose sensing. The glucose recognition occurs in our sensor by competitive binding of glucose and acceptor-labelled polysugar to the donor-labelled protein. Thus, the change in glucose level does not affect the bulk concentration of acceptor, but changes its distribution, contrary to the Förster-type theories. Indeed, these models poorly fitted the experimental data. Although γ vs. [glucose] calibration curve can be used for sensing purposes, it was clear that a new approach, dealing with the complex distributions in biological media, was required.

In our approach, instead of using the assumed analytical formula for the distribution function ρ_DA(r), the solution of the inverse problem is supported by generic properties of the ρ_DA(r) as a distribution function and known parameters of the donor and acceptor. The method enables determination of the ρ_DA(r) in a range of distances, offering richer than Förster approach structural information. The glucose studies based on this approach, revealed morphological details on the protein used for sensing. Also, we revisited the dye/metal-ion systems in polymers, phospholipid bilayers, and sol-gels and found that our data did not fit to the Förster models, as the actual donor-acceptor distributions significantly different from random. More recently we have demonstrated that a time-resolved FRET experiment is able to reveal ρ_DA(r) in the limited region of distances ~0.4R₀<r<~1.6R₀ and we are currently using this approach to study sol-gel pore morphology.

Fluorescence kinetics in complex media

The decay of an ideal unperturbed fluorescence sensor is expected to be single exponential, and any deviation from this simple behaviour is attributed to the detected process. However, in a complex biological sytem a single exponential decay is rare. For example three intrinsic fluorophores, aminoacids phenylalanine (Phe), tyrosine (Tyr) and tryptophan (Trp) present decays that have to be fitted by at least three-exponential model decay. Our recent extensive kinetics studies of fluorescent aminoacids demonstrated that intrin6 May, 2011etics resulting in a gamma decay function, which becomes exponential for the additional parameter q→1.

We demonstrated that gamma function is much more realistic model of fluorescence decay in biological media and finds much stronger background in statistical physics than multiexponential model. The importance of parameter q for sensing is still to be explored.

Highlights