Fluorescence lifetime studies of ligand-protein binding using the impulse response technique with refined data analysis methods

An instrument for the measurement of fluorescence decay lifetimes was constructed and used for studies of ligand-protein binding. The light source was a nitrogen-pumped dye laser with a subnanosecond pulse width. Detection was by the impulse response method using a boxcar averager to poll and digitize the output of a photomultiplier tube mounted on a laboratory-made base. Digitized signal information was sent from the boxcar to a dedicated microcomputer from which it was uploaded over a network to a workstation.;Data acquired in this way contained significant instrumental distortions. In order to resolve biexponential decay laws from such data, an analysis method was developed which took advantage of the computing power of the workstation. Lifetime parameters were roughly determined by the simulated annealing method and these estimates were used to provide starting guesses and upper and lower bounds for each parameter for analysis by simplex searching. The combination of these two techniques enabled accurate estimation of decay law parameters in the presence of instrumental distortion and noise.;One of the systems examined was an aqueous solution of quinine. The fluorescence from this system followed a biexponential decay law whose parameters were dependent upon emission wavelength.;The ultimate goal of this project was to estimate binding parameters of fluorescent ligands to proteins. The system chosen for this study was the binding of dansylated amino acids to bovine serum albumin. In this case, the bound amino acid derivative has a much longer fluorescence lifetime than the free form. Hence, it was possible to follow, for example, the increase in fluorescence from the longer lifetime component (the bound form) as the ratio of protein to ligand was increased. It was also possible to obtain evidence of the displacement of the fluorescent dansylated amino acid ligand from its binding site on albumin by nonfluorescent competitor ligands. Unfortunately, because the quantity proportional to concentration calculable in these experiments is dependent upon the spectral quantum yield of the fluorescent species, which is different for the free and the bound forms of the ligands, it was not possible to quantitatively estimate binding parameters from these experiments.