Synthesis

The phenanthridine derivative of alanine amino acid (Phen-AA) has been prepared according to the procedure described earlier [13].

General procedure for the synthesis of the compounds Phen-Py-1 and Phen-Py-2

To the solution of Boc-protected amino acid Phen-AA [19] in dichloromethane (4 mL) was added a TFA/H2O mixture (9:1, v/v; 2 mL) and the reaction was stirred at room temperature for 20 hours. The trifluoroacetate salt of the deprotected amino acid was obtained as yellow oil after evaporation of the solvent. The deprotected compound was then dissolved in anhydrous acetonitrile (3 mL) and appropriate pyrenecarboxylic acid, HBTU, HOBt and Et3N were added. The reaction was stirred at room temperature for 20 hours. The products Phen-Py-1 and Phen-Py-2 were isolated by preparative thin-layer chromatography in dichloromethane/methanol 9:1.

Phen-Py-1: Phen-AA (12.3 mg, 0.03 mmol), 1-pyrenebutyric acid (11.2 mg, 0.04 mmol), HBTU (11.4 mg, 0.03 mmol, 98%), HOBt (4.2 mg, 0.03 mmol, 97%) and Et3N (16.8 µL, 0.12 mmol) were used according to the general procedure. Phen-Py-1 was obtained as a white solid (9.4 mg, 56%). mp = 131–132 °C; Rf = 0,8 (CH2Cl2/MeOH 9:1); IR (KBr) νmax/cm−1: 3418 (s), 3294 (s), 3038 (m), 2947 (m), 2858 (m), 1738 (s), 1643 (s), 1582 (m), 1535 (m), 1435 (m), 1377 (m), 1209 (m), 843 (s), 760 (s), 723 (m); 1H NMR (CDCl3) δ 8.42 (d, J = 8.5 Hz, 1H, Phen-10), 8.31 (d, J = 8.1 Hz, 1H, Phen-1), 8.21–8.10 (m, 3H, Py), 8.07–7.93 (m, 6H, Phen, 5Py), 7.88 (d, J = 1.5 Hz, 1H, Phen-4), 7.73 (d, J = 7.8 Hz, 1H, Py), 7.66–7.45 (m, 3H, Phen), 5.98 (d, J = 7.7 Hz, 1H, NH), 5.12–5.03 (dd, J = 13.7, 6.0 Hz, 1H, CH), 3.75 (s, 3H, OCH3), 3.48–3.39 (m, 1H, CH2), 3.36–3.18 (m, 3H, CH2) 2.92 (s, 3H, Phen-CH3), 2.38–2.24 (m, 2H, CH2), 2.22–2.09 (m, 2H, CH2) ppm; 13C NMR (CDCl3) δ 172.4 (Cq), 172.1 (Cq), 158.4 (Cq), 143.7 (Cq), 135.6 (Cq), 135.3 (Cq), 131.8 (CH-Ar), 131. 7 (Cq), 131.5 (Cq), 131.0 (Cq), 130.1 (Cq), 129.4 (CH-Ar), 128.8 (Cq), 128.7 (CH-Ar), 127.6 (CH-Ar), 127.5 (CH-Ar), 127.3 (CH-Ar), 126.9 (CH-Ar), 126.8 (CH-Ar), 126.5 (CH-Ar), 126.0 (CH-Ar), 125.0 (CH-Ar), 124.9 (CH-Ar), 124.9 (CH-Ar), 123.6 (Cq), 123.3 (CH-Ar), 122.9 (CH-Ar), 121.9 (CH-Ar), 53.2 (CH-Ala), 52.7 (OCH3), 38.4 (CH2), 36.1 (CH2), 32.8 (CH2), 27.3 (CH2), 23.5 (CH3) ppm; HRMS (m/z): [M + H]+ calcd. for C38H32N2O3+, 565.2485; found, 565.2464.

Phen-Py-2: Phen-AA (12.0 mg, 0.03 mmol), 1-pyrenecarboxylic acid (9.2 mg, 0.04 mmol), HBTU (11.6 mg, 0.03 mmol, 98%), HOBt (4.2 mg, 0.03 mmol, 97%) and Et3N (16.8 µL, 0.12 mmol) were used according to the general procedure. Phen-Py-2 was obtained as a white solid (15.9 mg, 84%). mp = 230–231 °C; Rf = 0.8 (CH2Cl2:MeOH 9:1); IR (KBr) νmax/cm−1: 3435 (s), 3261 (s), 1740 (m), 1634 (s), 1531 (m), 849 (m), 760 (m); 1H NMR (CDCl3) δ 8.59 (d, J = 8.5 Hz, 1H, Phen-10), 8.51 (d, J = 7.3 Hz, 1H, Phen-1), 8.36 (d, J = 9.3 Hz, 1H, Py), 8.21 (d, J = 7.4 Hz, 1H, Py), 8.18–7.98 (m, 8H, 2Phen, 6Py), 7.91 (d, J = 9.3 Hz, 1H, Py), 7.78–7.67 (m, 2H, Phen), 7.66–7.58 (m, 1H, Phen), 6.66 (d, J = 7.6 Hz, 1H, NH), 5.51–5.40 (m, 1H, CH-Ala), 3.89 (s, 3H, OCH3), 3.81–3.72 (m, 1H, CH2-Ala), 3.57–3.47 (m, 1H, CH2-Ala), 2.90 (s, 3H, Phen-CH3) ppm; 13C NMR (CDCl3) δ 172.1 (Cq), 169.5 (Cq), 158.7 (Cq), 135.6 (Cq), 133.1 (Cq), 132.1 (CH-Ar), 131.3 (Cq), 130.7 (Cq), 129.9 (CH-Ar), 129.4 (CH-Ar), 129.1 (CH-Ar), 129.1 (CH-Ar), 128.9 (CH-Ar), 127.4 (CH-Ar), 127.2 (CH-Ar), 126.7 (CH-Ar), 126.6 (CH-Ar), 126.1 (CH-Ar), 126.1 (CH-Ar), 124.6 (CH-Ar), 124.5 (CH-Ar), 124.2 (CH-Ar), 123.7 (Cq), 123.1 (CH-Ar), 122.1 (CH-Ar), 54.0 (CH-Ala), 52.9 (OCH3), 38.5 (CH2), 23.4 (CH3) ppm; HRMS (m/z): [M + H]+ calcd. for C35H26N2O3+, 523.2022; found, 523.2025.

Study of DNA/RNA and enzyme interactions

General procedures: Solvents were distilled from appropriate drying agents shortly before use. TLC was carried out on DC-plastikfolien Kieselgel 60 F254 and preparative thick-layer (2 mm) chromatography was done on Merck 60 F254. 1H and 13C NMR spectra were recorded in DMSO-d6 or CDCl3 on Bruker AV 300 and 600 MHz spectrometers using TMS as the internal standard. The assignment of C-atoms and protons were confirmed on the basis of 2D NMR HETCOR, COSY, and NOESY. Chemical shifts (δ) are expressed in ppm, and J values in Hz. Signal multiplicities are denoted as s (singlet), d (doublet), t (triplet), q (quartet) and m (multiplet). High-resolution mass spectra (HRMS) were obtained using a MALDI–TOF/TOF mass spectrometer 4800 Plus MALDI TOF/TOF analyzer (Applied Biosystems Inc., Foster City, CA, USA). The electronic absorption spectra of newly prepared compounds, UV–vis titration and thermal melting experiments were measured on a Varian Cary 100 Bio spectrometer. Fluorescence spectra were recorded on a Varian Cary Eclipse fluorimeter. CD spectra were recorded on JASCO J815 spectrophotometer. Absolute quantum yields (Φf) were determined using software implemented with the instrument by the Integrating sphere SC-30 of the Edinburgh FS5 spectrometer. Quantum yields were measured for argon-purged solutions in sodium cacodylate buffer, pH 7.0, I = 0.05 mol dm−3, or pH 7.0, Ic = 0.05 mol dm−3 (λexc= 280 nm) at room temperature (25 °C) in a quartz cuvette of 10 mm path length; to avoid the scattering of incident light at the liquid–air interface, testing solutions with a 2 mL volume were used. Fluorescence and CD spectra were recorded using appropriate 1 cm path quartz cuvettes; UV–vis spectra were recorded in 1 cm path quartz cuvettes or using an immersion probe with 5 cm light path length. Isothermal titration calorimetry (ITC) titrations were performed on a Malvern PEAQ-ITC microcalorimeter (MicroCal, Inc.,Northampton, MA, USA). MicroCal PEAQ-ITC analysis software, supplied by the manufacturer, was used for data analysis. Polynucleotides were purchased as noted: calf thymus (ct)-DNA, poly dAdT–poly dAdT, poly dGdC–poly dGdC and poly rA–poly rU (Sigma) and dissolved in sodium cacodylate buffer, Ic = 0.05 mol dm−3, pH 7.0. The calf thymus (ct) DNA was additionally sonicated and filtered through a 0.45 µm filter [45]. The polynucleotide concentration was determined spectroscopically and expressed as the concentration of phosphates [45,46]. Recombinant human DPP III was obtained by heterologous expression in Escherichia coli and purification according to Špoljarić et al. [47,48]. Stock solutions of Phen-Py-1 and Phen-Py-2 were prepared by dissolving the compounds in DMSO; the total DMSO content was below 1% in UV–vis and below 0.1% in fluorimetric measurements. All measurements were performed in sodium cacodylate buffer, Ic = 0.05 mol dm−3.

UV–vis, CD, and fluorescence titrations: UV–vis and fluorimetric titrations were performed by adding portions of polynucleotide solution into the solution of the studied compound. After mixing polynucleotides/protein with studied compounds it was observed that the equilibrium was reached in less than 120 seconds. Compounds Phen-Py-1 and Phen-Py-2 showed a decrease of their excimer fluorescence emission intensity upon time. Therefore, buffer solutions of compounds were prepared 24 hours before titration with polynucleotides to ensure stable spectra of compounds. In fluorimetric titrations, the concentrations of studied Phen-Py-1 and Phen-Py-2 were 2 × 10−6 mol dm−3. An excitation wavelength of λexc = 352 nm was used for titrations to avoid absorption of excitation light caused by increasing absorbance of the polynucleotide or protein. The emission was measured in the range of λem = 350–650 nm. Fluorescence spectra were collected at r < 0.3 (r = [compound]/[polynucleotide]) to assure one dominant binding mode. Titration data were processed by means of Scatchard equation [38] and Global Fit procedure [40]. Calculations mostly gave values of ratio n = 0.2 ± 0.05, but for easier comparison all Ka values were re-calculated for fixed n = 0.2. Values for Ka have satisfactory correlation coefficients (>0.98). In Scatchard equation values of association constant (Ka) and ratio (n = [bound compound]/[polynucleotide]) are highly mutually dependent and similar quality of fitting calculated to experimental data is obtained for ±20% variation for Ks and n; this variation can be considered as an estimation of the errors for the given binding constants. CD experiments were performed by adding portions of Phen-Py-1 and Phen-Py-2 compound stock solution into the solution of polynucleotide (c ≈ 1–2 × 10−5 mol dm–3). The examined Phen-Py-1 and Phen-Py-2 compounds were chiral and therefore possessed intrinsic CD spectra. CD spectra were recorded with a scanning speed of 200 nm/min. Buffer background was subtracted from each spectrum, thus each spectrum was a result of two accumulations.

Thermal melting experiments: Thermal melting curves for ds-DNA, ds-RNA and their complexes with studied compounds were determined by following the absorption change at 260 nm as a function of temperature. The absorbance scale was normalized. Tm values were 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 here reported was the average of at least two measurements. The error in ΔTm is ±0.5 °C.

Isothermal titration calorimetry (ITC) experiments: A non-covalent interaction study of Phen-Py-1 with the human enzyme mutant DPP III E451A was performed on a MicroCal PEAQ-ITC microcalorimeter (Malvern, UK). Measurements were made in 20 mM Tris-HCl buffer, pH 7.4 at 25.0 °C with the addition of 5% DMSO. All experiments were performed under the same conditions; temperature 25.0 °C, reference power 30.0 μW, high feedback, stirring speed 500 rpm and initial delay 60 s. The enzyme solution (190–240 μM) was in the syringe (40 μL) and the compound Phen-Py-1 (20 μM) was in the reaction cell (200 μL). The reaction was started with a 0.4 μL injection of enzyme followed by 18 injections 2.0 μL each, with 150 seconds spacing to allow for equilibration. Blank experiments were carried out to determine the heats of dilution of the ligand and the enzyme. The resulting data were analyzed by using MicroCal PEAQ-ITC analysis software, supplied by the manufacturer, according to the model based on a single set of identical binding sites to estimate the binding constants (Ka) and the enthalpy of binding (∆rH). The reaction Gibbs energies (∆rG) were calculated by using the following equation: ∆rG = −RTln(Ka). The entropic contribution to the binding Gibbs energy was calculated by the equation: T∆rS = ∆rH − ∆rG.

Confocal microscopy: HeLa cells were cultured and maintained in complete high glucose (4.5 g/L) Dulbecco's Modified Eagle's Medium (DMEM, Sigma Aldrich) with the addition of 10% fetal bovine serum (FBS), 1% non-essential amino acids and 1% antibiotic/antimycotic solution (all chemicals were purchased by Capricorn Scientific GmbH). The cells were kept at 37 °C and 5% CO2 in a Heracell 150 humidified incubator (Heraeus, Germany). Before confocal microscopy experiments, HeLa cells were counted on LUNA-II Automated Cell Counter (Logos Biosystems) and transferred to 4-chamber 35 mm glass-bottom dishes (IBL, Austria) at a concentration of 15.000 cells per chamber and grown overnight. The dye Phen-Py-1 was added to the cells at a final concentration of 1 × 10−6 M, an hour before confocal imaging.

Computational details: In order to sample the conformational flexibility of investigated systems and probe their intrinsic dynamics in the aqueous solution, classical molecular dynamics (MD) simulations were performed employing standard generalized AMBER force fields (ff14SB [49] and GAFF [50]) as implemented within the AMBER16 program package [51]. All structures were subsequently solvated in a truncated octahedral box of TIP3P water molecules spanning a 10 Å thick buffer of solvent molecules around each system, and submitted to periodic simulations where the excess positive charge was neutralized with an equivalent number of chloride anions in monoprotonated systems corresponding to pH 5. Upon gradual heating from 0 K, MD simulations were performed at 300 K for a period of 300 ns, maintaining the temperature constant using the Langevin thermostat with a collision frequency of 1 ps−1. The obtained structures in the corresponding trajectories were clustered based on DBSCAN density-based algorithm according to recommended procedures. The idea behind this computational strategy was to investigate whether intrinsic dynamical features of studied conjugates both affect and can explain their tendency to undergo mutual association and form stacking interactions. The mentioned approach recently turned out as very useful in interpreting the affinities of several nucleobase – guanidiniocarbonyl–pyrrole conjugates towards single stranded RNA systems [19,52].

To confirm that the described clustering analysis elucidated the most representative structures of each conjugate at both experimental pH values, we proceeded by calculating energies of the excited states responsible for the experimental UV–vis spectra corresponding to isolated conjugates in the aqueous solution. For that purpose, we used the most abundant structure of each system in Figure 2 and performed the geometry optimization by the M06-2X DFT approach [53] together with the 6–31+G(d) basis set [54] in the Gaussian 16 program package [55], with the water solvent effects modeled through the implicit SMD solvation [56]. The choice of such computational setup was prompted by its success in reproducing various features of different organic [57,58], organometallic [59], and protein systems [60], being particularly accurate for relative trends among similar systems, which is the focus here. This was followed by the TD-DFT computations at the same level of theory considering 32 lowest singlet electronic excitations. The choice of this setup was prompted by its recent success in modeling UV–vis spectra of organic and inorganic systems in various solvents [61-63].