Identification of Thioflavin T Binding Modes to DNA: A Structure-Specific Molecular Probe for Lasing Applications

The binding mechanism of thioflavin T (ThT) to DNA was studied using polarized light spectroscopy and fluorescence-based techniques in solutions and in solid films. Linear dichroism measurements showed that ThT binds to DNA duplex by intercalation. Time-resolved fluorescence studies revealed a second binding mode which is the external binding to the DNA phosphate groups. Both binding modes represent the nonspecific type of interactions. The studies were complemented with the analysis of short oligonucleotides having DNA cavities. The results indicate that the interplay between three binding modes—intercalation, external binding, and binding inside DNA cavities—determines the effective fluorescence quantum yield of the dye in the DNA structures. External binding was found to be responsible for fluorescence quenching because of energy transfer between intercalated and externally bound molecules. Finally, amplified spontaneous emission (ASE) was successfully generated in the ThT-stained films and used for detecting different DNA structures. ASE measurements show that ThT-stained DNA structures can be used for designing bioderived microlasers.

The concentrations of G4 DNA solutions were assessed on the basis of absorbance at 258 nm with the molar extinction coefficient of 262000 M -1 cm -1 and the sequence molecular weight (M w ) of 8935 g/mol. The stock solution concentration was equal to 110 μM (1mg/ml). The G-qudruplex sequence was kept at 95 o C for 3 minutes and then slowly cooled down to room temperature in order to form the G4 structure.
ThT-stained solutions were prepared by mixing [dye]/[DNA base] in ratio from 0.00067 to 0.04. In order to obtain film samples the solutions were drop casted on glass slides in volume 100 μl and allowed to dry in ambient conditions. UV-Vis Spectroscopy: Absorption spectra were recorded on a CARY-5000 spectrophotometer. Linear dichroism (LD) spectra were measured using Chirascan CD spectrophotometer. LD is defined as the difference in absorbance of light linearly polarized parallel (A ∥ ) and perpendicular (A ⊥ ) to the orientation axis of the DNA. Orientation was achieved using unit flow Couette cell. The sample was shear aligned in 0.5 mm space between two concentric quartz cylinders in a custom made outer-rotating Couette cell under a shear gradient of 3100 s -1 . The total path length of the Couette cell was 1 mm. All spectra were corrected for background contributions by subtracting corresponding spectra recorded without shear/rotation.
The reduced LD (LD r ) is obtained by dividing the LD by the absorbance of the corresponding isotropic sample (A iso ), which for an anisotropic sample can be obtained as A iso = (A ∥ +2A ⊥ )/3, so that For molecules with a uniaxial orientation distribution, as can be expected for DNA in a Coutte cell, LD r is a product of an orientation factor S and an optical factor O.
The optical factor O is related to the angle α between the helix axis and the light absorbing transition moment of the actual chromophore, which could be a DNA base or a DNA-bound ligand. The orientation factor S = 0.5 (3<cos 2 θ> -1) corresponds to the average orientation of the DNA helix in the flow solution, where θ is the angle between the macroscopic orientation direction and the local helix axis of a particular molecule and where the average runs over all DNA molecules in the sample. The degree of helix orientation S was calculated from the LD r values at 260 nm using the average angle α DNA = 86 o for the DNA bases. The binding angle α dye for a dye with respect to the DNA helix axis can then be calculated from the LD r value in the visible region of its absorption spectra.
Steady-state fluorescence spectroscopy: For recording fluorescence spectra samples were excited with the light of a xenon arc lamp passing through a monochromator with the central wavelength and transmission bandwidth set to 400 nm and 1 nm, respectively, and a short-pass absorption filter (absorption edge at 450 nm). The emitted light was analysed with a SpectraPro 150 Czerny-Turner imaging monochromator equipped with a CCD camera (Andor DU420A-BU2). The fluorescence light was collected at the right angle to the direction of the excitation light. The cuvette with the solution was oriented at 90 o and the films were oriented at approx. 45 o angle with respect to the excitation and detection directions. A long-pass filter (edge at 430 nm) in front of the monochromator was used to eliminate the scattered excitation light. The recorded spectra were corrected for the transmission curve of this filter. The overall S3 spectral responsivity curve of the apparatus was determined using secondary emission standards according to the procedure proposed by Gardecki and Maroncelli 2 .
Time-resolved fluorescence. Fluorescence decays were recorded using a custom-built setup based on the PicoQuant HydraHarp 400 Multichannel Picosecond Event Timer and the picosecond diode laser PicoQuant LDH-P-C-405B working at 405 nm. Repetition frequency of the laser was set to 10 MHz and the average excitation power was in the range 0.5 μW to 2.5 μW. Liquid samples were oriented at 90 o and film samples were oriented at approx. 45 o with respect to the excitation beam, with the fluorescence light collected at the approximately right angle with respect to the direction of the excitation beam. The collected light passed through a long-pass cut-off filter in order to remove the scattered excitation light. As the setup was used for measurements of liquid samples, the fluorescence light passed through a polarizer, whose transmission axis was oriented at the magic angle (54.7 o ) with respect to the polarization of the excitation beam, in order to eliminate artifacts due to the rotational diffusion. The fluorescence light was focused on the entrance slit of a Czerny-Turner monochromator (Acton Research SpectraPro 150) used to select a given wavelength from the fluorescence spectrum. After the monochromator the light was detected with the Hamamatsu R3809U-50 photomultiplier sensitive in the 160-850 nm range. The output signal from the photomultiplier was amplified with the Ortec Model 9327 amplifier prior to routing it to the HydraHarp module. The overall instrumental response function (IRF), measured by scattering the excitation light in a suspension of titanium dioxide, was narrower than 100 ps FWHM.
Recorded fluorescence kinetics were analyzed by fitting with multiexponential decay functions convoluted with the IRF, using the Horiba FelixGX software. The average fluorescence lifetime was calculated as the amplitude-weighted mean of the decay times of each component.

Amplified spontaneous emission (ASE).
Spectra of amplified spontaneous emission of ThT mixed with DNA structures in films were recorded in the experimental setup shown in Fig. S1.

S4
Studied films were excited at 400 nm with femtosecond pulses obtained by frequency doubling the output beam of a Ti:Sapphire femtosecond amplifier system (Legend Elite Duo) working at 5 kHz repetition rate. A beta-barium borate (BBO) crystal 1 mm thick was used to generate the second harmonic (SH) light of the amplifier output beam. Energy of the resulting SH pulses was adjusted by rotating the half-wave plate mounted in front of the BBO crystal. The SH light was separated from the remaining fundamental light by a pair of dichroic mirrors. The maximum energy of an SH pulse achievable in this configuration was approximately 300 μJ. After the dichroic mirrors the beam was first expanded with a telescope formed by a pair of spherical lenses with focal lengths of -50 and 125 mm. Then it was focused with a cylindrical lens in order to form a narrow stripe of light at the position of the film sample. The profile of the beam in the focal point was determined with the Thorlabs LC1-USB line CCD camera. It was approximately Gaussian in both directions with the full width at half maximum (FWHM) of (7.7 ± 0.3) mm in the horizontal and (17 ± 1) m in the vertical direction, respectively. The area of the focal point at the FWHM intensity level, used for the calculation of average intensities of the excitation light reported in the current work, was taken as (0.0013 ± 0.0001) cm 2 .
The studied films were oriented perpendicularly to the direction of the excitation beam. The amplified spontaneous emission was detected in the plane of the films, along the direction of the elongated linear focus of the cylindrical lens. The ASE light was collected with a fiber and delivered to the Ocean Optics USB 2000 spectrometer, whose spectral resolution was around 4 nm. For the determination of ASE thresholds the intensity of the excitation light was gradually increased by rotating the half-wave plate and the spectrum of the emitted light was simultaneously monitored. Plots of the dependence of the light intensity at the wavelength corresponding to the maximum of the ASE spectrum on the excitation intensity were used to determine the ASE generation thresholds.