The photophysical mechanisms which determine the spectral properties and decay rates of 4',6-diamidine-2-phenylindole (DAPI) in solution and in association with nucleic acids have not yet been fully elucidated. We have performed steady-state and time-resolved fluorescence experiments on DAPI in a wide pH range to investigate the hypothesis that different ground-state conformations are responsible for the photophysical properties of the probe. Several excited-state mechanisms are investigated and it is concluded that among the proposed models, the hypothesis of ground-state heterogeneity with rapid interconversion among conformations is the only one consistent with the experiments in the entire pH range investigated.

Steady-state and dynamic fluorescence measurements have been performed on DAPI in solution and in complexes formed with a number of synthetic and natural polydeoxynucleotides. The decay of DAPI in buffer at pH 7 was decomposed using two exponentials having lifetime values of approximately 2.8 ns and 0.2 ns. The double exponential character of the decay was maintained over a large pH range from 3 to 9. At pH 1 the short component dominated, whereas at pH 12, only the long component was detectable. Two distinct spectra were associated with the two lifetime components; the short component was shifted to the red. The short lifetime component occurs in the presence of water. In water the excitation spectra depended on the emission wavelength and there was no viscosity dependence of the two forms. To explain these results we propose that there is a ground state conformer in which preferential solvation of the indole ring allows proton transfer in the excited state. DAPI complexed with polydeoxynucleotides retained most of the features of the decay of DAPI in solution. However, the complexes with fully AT-containing polymers stabilized the longer lifetime form of DAPI because the stronger binding enhanced solvent shielding. A gradual increase of the short lifetime component, which monitors dye solvent exposure, was obtained as the AT content was decreased. For polyd(GC) the decay was similar to that of free DAPI.

Fluorescence depolarization of synthetic polydeoxynucleotide/4'-6-diamidino-2-phenylindole dihydrochloride complexes has been investigated as a function of dye/polymer coverage. At low coverage, fluorescence depolarization is due to local torsional motions of the DNA segment where the dye resides. At relatively high coverage, fluorescence depolarization is dominated by energy transfer to other dye molecules along the DNA. The extent of the observed depolarization due to torsional motion depends on the angle the dye molecule forms with the DNA helical axis. A large torsional motion and a small angle produce the same depolarization as a small torsional motion and a large projection angle. Furthermore, the extent of transfer critically depends on the relative orientation of dye molecules along the DNA. The effect of multiple transfer is examined using a Monte Carlo approach. The measurement of depolarization with transfer, at high coverage, allows determination of the dye orientation about the DNA helical axis. The value of the torsional spring constant is then determined, at very low coverage, for few selected polydeoxynucleotides.

Fluorescence decay studies, obtained by multifrequency phase-modulation fluorometry, have been performed on DAPI in solution and complexed with natural and synthetic polydeoxynucleotides. DAPI decay at pH 7 was decomposed using two exponential components of 2.8 and 0.2 ns of lifetime values, respectively. The double exponential character of the decay was maintained over a large pH range. Phase- and modulation-resolved spectra, collected between 420 and 550 nm, have indicated at least two spectral components associated with the two lifetime values. This, plus the observation of the dependence of the emission spectrum on the excitation wavelength, suggests a lifetime heterogeneity originating from ground-state molecular conformers, partially affected by pH changes. DAPI complexed with natural polydeoxynucleotides retained most of the features of DAPI decay in solution, except for the value of the long lifetime component that was longer (approximately 4 ns) and the relative fractional fluorescence intensities of the two components that were inverted. AT polymers/DAPI complexes show single exponential decay. Solvent shielding when DAPI is bound to DNA changes the indole ring solvation and stabilizes the longer lifetime decay component. For poly(GC)/DAPI complex, the decay was similar to that of free DAPI in solution, proving the dependence on the polydeoxynucleotides sequence the different types of binding and the reliability of the fluorescence method to solve them.

The binding of 4',6-diamidino-2-phenylindole (DAPI) to double-stranded GC polymers either in the alternating or in homopolymer sequence was investigated using fluorescence techniques. We employed fluctuation correlation spectroscopy, which measures the diffusion coefficient of fluorescent particles, to demonstrate that the fluorescence was originating from relatively slowly diffusing entities. These entities display a very large heterogeneity of diffusing coefficients, indicating that molecular aggregation is extensive in our samples. We used frequency domain fluorometry to characterize the fluorescence lifetime of the species, while varying the GC polymer-dye coverage systematically. At very low coverage we observed a relatively bright fluorescent component with a lifetime value of approximately 4 ns. The stoichiometry of binding of this bright species was such that it can only arise from rare molecular structures, either unusual loops or large molecular aggregates. The amount and characteristics of this bright fluorescent component were different between the homo and the alternating polymer, indicating that the difference in sequence of the two polymers is responsible for the different aggregates which are then detected in the fluorescence experiment. At large GC polymer coverage we observed a relatively wide distribution of fluorescent species with short lifetime values, in the range between 0.12 and 0.2 ns. Given the stoichiometry of binding of this fluorescent component, we concluded that it could arise either from intercalative and/or non-specific binding to the DNA double-stranded molecules. We comment on the origin of the rare but brightly fluorescent binding sites and discuss the potential to detect such unusual DNA structures.