Colorimetric and Fluorimetric DNA Detection with a Hydroxystyryl–Quinolizinium Photoacid and Its Application for Cell Imaging

Abstract The combination of styryl dye properties with the acidity and strong photoacidity of the 2,2′‐[(1′′‐hydroxy‐4′′‐methyl‐(E)‐2′′,6′′‐phenylene)]‐bisquinolizinium enables the detection of DNA by distinct absorption and emission color changes and the fluorimetric detection of DNA in cells with epifluorescence and confocal fluorescence microscopy.


Acid-base titrations
i) Acid-base titrations were performed in Britton-Robinson buffer solution, as obtained from phosphoric acid, boric acid, and sodium acetate (0.04 M each) in water, adjusted to a particular pH value by addition of aq. solution of NaOH (2 M) or aq. HCl (2 M). 2 Following each addition step, the pH and the absorption spectra were recorded. The absorption at a particular wavelength was plotted versus the pH of the solution, and the resulting isotherm was used to determine acidity constant pK a by numerical fitting to the Henderson-Hasselbalch equation. 3 The pKa value and the corresponding error of 3a were obtained by numerical fit of the titration isotherms to the "dose resp" function of equation 1.
A 1 = Absorption at initial pH A 2 = Absorption at final pH LOGx 0 = pKa p = Hill slope ii) Emission spectra were recorded in aq. HClO 4 (11.8 M) solutions and in aq. NaOH (pH 8.5) solutions to assign the bands of 3a and the conjugate base 3a cB ( Figure S2). The energy of the 0-0 transitions were estimated by means of average between the wavenumbers corresponding to the maxima of absorption and emission (Eq. 2).
The excited-state acidity constant pK a * was estimated according to the Förster cycle (Eq. 3).   iii) The assignment of the emission bands of 3a were further confirmed by experiments in 2,2,2trifluoroethanol (TFE), i.e. a protic solvent with no proton-acceptor properties. 4 Aliquots of a stock solution of 3a in MeOH (c = 1.0 mM) were evaporated under a stream of nitrogen and redissolved in the solvent mixture of trifluoroethanol/ethanol with increasing EtOH content from 20% to 100%. Fluorescence spectra were recorded with excitation and emission slits of 5 nm and an excitation wavelength λ ex = 440 nm. In TFE, 3a exhibits a strong emission (Φ fl = 0.35) with maximum at 530 nm. On addition of EtOH, which is a proton acceptor, this emission band decreased and an additional emission band developed at 790 nm corresponding to the emission of the conjugate base 3a cB .

Determination of fluorescence quantum yield
Aliquots of a stock solution of derivatives 3a and 3b in MeOH (c = 1.00 mM) were evaporated under a stream of nitrogen and redissolved in the respective solvent for fluorimetric analysis. The relative fluorescence quantum yields, Φ fl , of 3a and 3b were determined according to Eq. 4 5 under identical conditions (detection wavelength, excitation wavelength, detector voltage, slit bandwidths, collection rate).
Φ fl,x = Φ fl,s ( F x / F s )( A s / A x )(n x 2 / n s 2 ) (Eq. 4) The indices x and s indicate the analyte (x) and standard (s) solution. F = area under the emission curve, A = absorbance at the excitation wave length, n = index of refraction of the solvent.

Photometric and fluorimetric DNA titrations
For photometric and fluorimetric titration of 3a and 3b with ct DNA, aliquots of a stock solution in MeOH (c = 1.00 mM) were evaporated under a stream of nitrogen, redissolved in DMSO (5% v/v) and BPE buffer to obtain a ligand concentration of c = 20 µM. The DNA solutions contained also the ligand at the same concentration in order to avoid dilution effects. For the detection of emission spectra the excitation and emission slits were adjusted to 5 nm. The spectra were smoothed with the implemented moving average function by a factor of 5. Samples of the ligand solutions were placed into quartz cells and absorption or emission spectra were recorded. The titrations were performed up to saturation. All spectrometric titrations were performed at least two times to ensure reproducibility. The binding constants were determined from plots of the absorption at a given wavelength versus relative DNA concentration, c DNA / c ligand , and fitting of the experimental binding isotherm ( Figure S7) to the theoretical model (Eq. 4). 6 (Eq. 5) Q = I / I 0 = the minimal absorbance in the presence of excess ligand n = number of independent binding sites per DNA Standard deviations (SD) of K b values were calculated from equation 6.

SD (K b ) = {(SD of A / A) / A} / c lig (Eq. 6)
The difference of SD values for Kb values of 3a and 3b originates in in different contributions of the A value for 3b that is bit of higher as compared to the A value for 3a. Moreover, the A parameter is a kind of intersection.

CD-and LD-spectroscopic experiments
CD and LD spectra were recorded in BPE buffer solution at different ligand-DNA ratios (LDR) at fixed DNA concentration (c DNA = 50.0 µM). The CD and LD measurements were performed at LDR = 0, 0.3, 0.5, 0.6, 0.8, 1.0 with band width of 1 nm, recording speed of 1 nm/s and time per data point of 0.5 seconds. Flow-LD spectra were recorded on CD spectrometer equipped a High Shear Couette Cell Accessory (Applied Photophysics). The LD samples were recorded in a rotating couette with a shear gradient of 1200 s -1 .

NIH 3T3 cell culture and fluorescence staining
The NIH 3T3 mouse fibroblasts 7 were cultured at standard conditions (37°C, 5% CO 2 ) in Dulbecco's modified Eagle medium (DMEM high glucose; Gibco, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (Gibco, Thermo Fisher Scientific), 2 mM L-glutamine (Gibco, Thermo Fisher Scientific), 100 U mL -1 penicillin (Gibco, Thermo Fisher Scientific) and 100 µg mL -1 streptomycin (Gibco, Thermo Fisher Scientific). Subcultivation of cells was performed by treatment with 0.25% Trypsin/EDTA of cell culture layers washed with PBS (Lonza, Belgium). Released cells were collected and centrifuged (270 x g for 4 min) in a conical tube, counted with a Neubauer improved cell counting chamber (Brand, Wertheim, Germany) and seeded at a density of 10.000 cells cm -2 on glass cover slips (Menzel Gläser 20 x 20 mm, thickness 0.17 mm) in 6-well plates (Standard F, Sarstedt AG. & Co.). Cells were cultured to a 60% confluency before fluorescent labeling. If cells were fixed before the fluorescent labeling with 3a, they were washed gently three times with prewarmed (37 °C) PBS (Lonza, Belgium), fixed with 4% paraformaldehyde in PBS for 20 min at 21 °C and washed again three times with PBS at 21 °C.
If the fluorescent labeling with 3a was performed with living cells, the culture medium was removed only once.
For the fluorescent labeling, cells were incubated with 2.5 µM 3a as a final concentration (3a stock concentration: 1.0 mM in DMSO) in either pre-warmed cell culture media for staining of living cells or PBS for staining of fixed cells for 1 h at 37 °C and 5% CO 2 (living cells) or at room temperature (fixed cells). As controls, cells have been in parallel incubated with cell culture media or PBS only or with the same final concentration of DMSO [0.25 % (v/v)].
After staining of living cells they were washed two times with pre-warmed (37°C) PBS and the fixation protocol was performed as described above.
For counter staining with Hoechst 33258 (Carl Roth) as a gold standard, cells were optionally incubated with 1 µg/mL Hoechst 33258 in PBS for 15 min at room temperature. Before final sample embedding with Mowiol® 4-88 (Carl Roth) samples were washed three times with PBS and once with Milli-Q water (from a Millipore Direct Q8 system, advantage A10 system, Schwalbach, with Millimark Express 40 filter, Merck, Germany).
The camera exposition time for filterset 1) was 4000 ms, for 2) 10000 ms and for 3) 20000 ms. Brightness and contrast were optimized by the "auto" adjustment of the Zen2.3 lite software without information loss.  Figure 4C). Labeling 3a stands for both the ligand 3a and its conjugate base 3a cB compound. Scale bar: 20 µm. Figure S17. Epifluorescence microscopy images of NIH 3T3 mouse fibroblasts after fixation. The cell nuclei were stained with a standard Hoechst 33258 stain solution. The background fluorescence in the red and green channel are negligibly low. Corresponding exposition times as well as automatic brightness and contrast adjustment were applied. The images show pseudo coloring of the fluorescence emission resulting from three different filter sets: green channel: ʎ ex = 450-490 nm, ʎ em >515 nm; B) red channel: ʎ ex = 540-552 nm, ʎ em > 590 nm; C) Blue channel: ʎ ex = 320-390 nm, ʎ em = 420-470 nm (Hoechst 33258) ; D) Overlay of blue channel with the corresponding bright field image. Figure S18. Epifluorescence microscopy images of NIH 3T3 mouse fibroblasts, which were stained with 3a (2.5 µM in PBS) for 1 h in cell medium prior to fixation with 4 % PFA. The images show pseudo coloring of the fluorescence emission resulting from three different filter sets: A) green channel: ʎ ex = 450-490 nm, ʎ em >515 nm; B) red channel: ʎ ex = 540-552 nm, ʎ em > 590 nm; C) Overlay of red and green channel; D) Overlay of red and green channel with the corresponding bright field image. Labeling 3a stands for both the ligand 3a and its conjugate base 3a cB compound. Figure S19. Confocal fluorescence microscopy images of NIH 3T3 mouse fibroblasts after fixation and incubation with 3a (2.5 µM) for 1 h. Excitation wavelength ʎ ex = 485 nm, emission wavelength λ em = 515-652 nm (A1) and 652-732 nm (B1). Semi-logarithmic plot of the TCSPC triple-exponential fluorescence decays in corresponding images A1 and B1 at λ em > 515 nm (A2) and 652-732 nm (B2) (black dots) together with the corresponding instrument response function (gray line) and calculated fits (red line) at the time resolution of 16 ps. In the A2 and B2 panels, the weighted deviations (σ) and the autocorrelation function (A-C) plots as well as the decay times (τ), their amplitudes (a) and χ 2 fit quality parameter are shown. Excitation wavelength ʎ ex = 485 nm, emission wavelength λ em = 515-652 nm (G) and 652-732 nm (R). A close to zero negative intensity of background is shown in blue.