Bis(phenylethynyl)arene Linkers in Tetracationic Bis‐triarylborane Chromophores Control Fluorimetric and Raman Sensing of Various DNAs and RNAs

Abstract We report four new luminescent tetracationic bis‐triarylborane DNA and RNA sensors that show high binding affinities, in several cases even in the nanomolar range. Three of the compounds contain substituted, highly emissive and structurally flexible bis(2,6‐dimethylphenyl‐4‐ethynyl)arene linkers (3: arene=5,5′‐2,2′‐bithiophene; 4: arene=1,4‐benzene; 5: arene=9,10‐anthracene) between the two boryl moieties and serve as efficient dual Raman and fluorescence chromophores. The shorter analogue 6 employs 9,10‐anthracene as the linker and demonstrates the importance of an adequate linker length with a certain level of flexibility by exhibiting generally lower binding affinities than 3–5. Pronounced aggregation–deaggregation processes are observed in fluorimetric titration experiments with DNA for compounds 3 and 5. Molecular modelling of complexes of 5 with AT‐DNA, suggest the minor groove as the dominant binding site for monomeric 5, but demonstrate that dimers of 5 can also be accommodated. Strong SERS responses for 3–5 versus a very weak response for 6, particularly the strong signals from anthracene itself observed for 5 but not for 6, demonstrate the importance of triple bonds for strong Raman activity in molecules of this compound class. The energy of the characteristic stretching vibration of the C≡C bonds is significantly dependent on the aromatic moiety between the triple bonds. The insertion of aromatic moieties between two C≡C bonds thus offers an alternative design for dual Raman and fluorescence chromophores, applicable in multiplex biological Raman imaging.


S4
were collected on a RIGAKU OXFORD DIFFRACTION XtaLAB Synergy diffractometer with a semiconductor HPA-detector (HyPix-6000) and multi-layer mirror monochromated Cu-K radiation. The crystals were cooled using an Oxford Cryostream or Bruker Kryoflex lowtemperature device. Data were collected at 100 K (3N, 4N) or 120 K (C). The images were processed and corrected for Lorentz-polarization effects and absorption as implemented in the Bruker software packages (3N, 4N) or in the CrysAlis Pro software (C), respectively. The structures were solved using the intrinsic phasing method (SHELXT) [7] and Fourier expansion technique. All non-hydrogen atoms were refined in anisotropic approximation, with hydrogen atoms 'riding' in idealized positions, by full-matrix least squares against F 2 of all data, using SHELXL [8] software and the SHELXLE graphical user interface. [9] For both, 3N and 4N, each asymmetric unit contains one ethylacetate molecule disordered over three positions, while for C it contains two hexane molecules disordered over two positions each in addition to the main molecules. Diamond [10] software was used for graphical representation. Crystal data and experimental details are listed in Table S3; full structural information has been deposited with Cambridge Crystallographic Data Centre. CCDC-1997113 (3N), 1997114 (4N), and 1997115 (C).
Linear Optical Properties. All measurements were performed in standard quartz cuvettes (1 cm x 1 cm cross-section) under ambient conditions. UV-visible absorption spectra were recorded using an Agilent 8453 diode array UV-visible spectrophotometer. The molar extinction coefficients were calculated from three independently prepared samples in hexane (3N-5N) and MeCN and H2O (3)(4)(5) solution. The emission spectra were recorded using an Edinburgh Instruments FLSP920 spectrometer equipped with a double monochromator for both excitation and emission, operating in right-angle geometry mode, and all spectra were fully corrected for the spectral response of the instrument. All solutions used in photophysical measurements had a concentration lower than 5 × 10 -6 M to minimize inner filter effects during fluorescence measurements. The fluorescence quantum yields were measured using a calibrated integrating sphere (inner diameter: 150 mm) from Edinburgh Instruments combined with the FLSP920 spectrometer described above. For solution-state measurements, the longest-wavelength absorption maximum of the compound in the respective solvent was chosen as the excitation wavelength, unless stated otherwise. Fluorescence lifetimes were recorded using the time-correlated single-photon counting (TCSPC) method using an Edinburgh Instruments FLS980 spectrometer equipped with a high speed photomultiplier tube positioned after a single emission monochromator. Measurements were made in right-angle geometry mode, and the emission was collected through a polarizer set to the magic angle.
Solutions were excited with a pulsed diode laser at a wavelength of 376.6 nm (3N, 4N, 3, 4) and 472.6 nm (5N, 5) at repetition rates of 10 or 20 MHz, as appropriate. The full-width-at-halfmaximum (FWHM) of the pulse from the diode laser was ca. 80 ps with an instrument response S5 function (IRF) of ca. 230 ps FWHM and ca. 200 ps with an instrument response function (IRF) of ca. 1120 ps FWHM, respectively. The IRFs were measured from the scatter of an aqueous suspension of Ludox at the excitation wavelength. Decays were recorded to 10 000 counts in the peak channel with a record length of 8 192 channels. The band pass of the emission monochromator and a variable neutral density filter on the excitation side were adjusted to give a signal count rate of <60 kHz. Iterative reconvolution of the IRF with one decay function and non-linear least-squares analysis were used to analyse the data. The quality of all decay fits was judged to be satisfactory, based on the calculated values of the reduced χ 2 and Durbin-Watson parameters and visual inspection of the weighted residuals.
Optical Properties in Sodium Cacodylate. UV-visible absorption spectra were recorded on a Varian Cary 100 Bio spectrometer; excitation and emission spectra were recorded on a Varian Cary Eclipse fluorimeter Study of Interactions with DNA and RNA. Polynucleotides were purchased as noted: poly dGdCpoly dGdC, poly dAdTpoly dAdT, poly Apoly U, poly A, poly G, poly C, poly U (Sigma), calf thymus (ct)-DNA (Aldrich) and dissolved in sodium cacodylate buffer, I = 0.05 M, pH=7.0. The ct-DNA was additionally sonicated and filtered through a 0.45 mm filter to obtain mostly short (ca. 100 base pairs) rod-like B-helical DNA fragments. [11] The polynucleotide concentration was determined spectroscopically [12] as the concentration of phosphates (corresponds to c(nucleobase)). Thermal melting experiments were performed on a Varian Cary 100 Bio spectrometer in quartz cuvettes (1 cm). The measurements were carried out in aqueous buffer solution at pH 7.0 (sodium cacodylate buffer I = 0.05 M). Thermal melting curves for ds-DNA, ds-RNA and their complexes with 3-6 were determined by following the absorption change at 260 nm as a function of temperature. [13] Tm values are the midpoints of the transition curves determined from the maximum of the first derivative and checked graphically by the tangent method. The ΔTm values were calculated subtracting Tm of the free nucleic acid from Tm of the complex. Every ΔTm value reported here was the average of at least two measurements. The error in ΔTm is  0.5 °C.  [14,15]  software, at the B3LYP/6-31G(d) level of theory, and the parametrization procedure was performed using the Antechamber module within the AMBER16 program suite wherein the Mullikan charges were used as the partial atomic charges. Parametrization, energy minimization, and molecular dynamics (MD) simulations of the complexes between compound 5 and DNA were performed using the AMBER16 suite of programs. [16] The solutes were prepared using the AMBER16 utility program tLeap wherein ligand and DNA were parametrized within general AMBER force field gaff [17] and ff99bsc0 [18] force fields, respectively. For details of the parametrization procedure see our previous work. [19] Initial conformations were prepared in PyMOL (The PyMOL Molecular Graphics System, Version 1.7 SchroÈdinger, LLC), wherein compound 5 was docked into the DNA minor groove using DNA-1 complex [19] as a template. The systems were solvated in the truncated octahedron box filled with TIP3P water molecules [20,21] whereas the sodium ions were added to achieve electroneutrality. The complexes were minimized, equilibrated and simulated for 300 ns by the programs sander.MPI and pmemd.MPI. The simulations were performed using periodic boundary conditions (PBC). The particle mesh Ewald (PME) method was used for calculation of the long-range electrostatic interactions, and in the direct space the pairwise interactions were calculated within the cut-off distance of 8 Å. The solvated complexes were geometry optimized by using steepest descent and conjugate gradient methods (altogether 7000 steps), and equilibrated for 100 ps with time step of 1 fs. During the first stage of equilibration (30 ps), the temperature was linearly increased from 0 to 300 K and the volume was held constant. In the second stage (NPT ensemble with T and P about 300 K and 1 atm, respectively) the solution density was optimized. The equilibrated complexes were subjected to productive molecular dynamics simulation using NPT conditions and a time step of 2 fs. The temperature was held constant using a Langevin thermostat [22] with a collision frequency of 0.2 ps, and the pressure was regulated by a Berendsen barostat. [23] Raman and SERS measurements. Raman and SERS spectra were measured on a Bruker Equinox 55 interferometer equipped with a FRA 106/S Raman module using Nd-YAG laser S7 excitation at 1064 nm of 500 mW laser power. The spectra were acquired in the 3500-100 cm 1 spectral range at 4 cm 1 resolution. A total of 512 and 128 scans were averaged for a Raman and SERS spectrum, respectively. Quartz cuvettes were used for handling samples.
Solutions of the bis-triarylborane compounds were prepared by dissolution of the solid substance in water, the concentration of which was determined spectroscopically using the respective molar absorption coefficient. The stock solutions were further on diluted in water to obtain solutions of 1  10 4 M (3, 4 and 5) and 2  10 3 M (6), used for Raman measurements.
For the SERS measurement a silver colloidal suspension was used as the SERS active substrate, prepared by reduction of silver nitrate with trisodium citrate according to the procedure described in our previous work. [2] The resulting colloidal suspension was gray colored, characterized by a maximum at 416 nm in the UV/Vis spectrum, pointing to the typical silver plasmon resonance frequency. The pH value of the silver colloid was 7.5. Working samples for the SERS measurements were prepared by dissolution of bis-triarylborane compound solution in an appropriate volume of water, followed by addition of 400 L of the silver colloid. The total sample volume was 500 L. For the concentration dependent measurements the final concentrations of 3-6 in the Ag colloid were 1  10 7 , 5  10 7 , 1  10 6 and 5  10 6 M. To measure the SERS spectra of the complexes of 3-5 with ct-DNA, samples were prepared in the buffered solution by mixing the appropriate volume of the bis-triarylborane compound with ct-DNA in molar ratios r[compound]/[ct-DNA] = 1, 0.2 and 0.1, followed by addition of the of 400 L silver colloid. The total volume of the working sample was 500 L and the final concentration of the bis-triarylborane compound was 1  10 -6 M.
Theoretical Studies. All calculations (DFT and TD-DFT) were carried out with the Gaussian 16 (16.A.03) [24] program package and were performed on a parallel cluster system. GaussView (6.0.16), Avogadro (1.2.0) [25] and multiwfn [26] were used to visualize the results, to measure calculated structural parameters, and to plot orbital surfaces (isovalue: ± 0.030 [e a0 -3 ] 1/2 ). The ground-state geometries were optimized using the B3LYP functional [27] in combination with the 6-31G(d) basis set. [28,29] The ultrafine integration grid and no symmetry constraints were used for all molecules. Frequency calculations were performed on the optimized structures to confirm them to be local minima showing no negative (imaginary) frequencies. Based on these optimized structures, the lowest-energy vertical transitions (gas-phase) were calculated (singlets, 25 states) by TD-DFT, using the CAM-B3LYP functional in combination with the 6-31G(d) basis set. [28,29] For calculations of the Raman spectra the ground state geometries were optimized at the B3LYP/6-31+G(d, p) level of theory. The resulting frequencies were multiplied by a scaling factor of 0.964 (as suggested by the Computational Chemistry Comparison and Benchmark database (https://cccbdb.nist.gov/vibscalejustx.asp)).

S34
Studies of Interactions with DNA and RNA Table S4. Groove widths and depths for selected nucleic acid conformations. [30,31] Structure poly dGdCpoly dGdC c 13.