Synthesis, cytotoxicity, pharmacokinetic profile, binding with DNA and BSA of new imidazo[1,2-a]pyrazine-benzo[d]imidazol-5-yl hybrids

Novel derivatives possessing imidazo[1,2-a]pyrazine and 1H-benzo[d]imidazole scaffolds were synthesized using Suzuki-Miyaura cross-coupling reactions. In vitro anticancer activities against NCI-60 cancer cell panels were tested at 10 µM concentration. The best results were obtained from substitution of two 1-cyclohexyl-1H-benzo[d]imidazole groups present at C-6 and C-8 positions of imidazo[1,2-a]pyrazine (31). Compound 31 was found to be cytotoxic against 51 cell lines and cytostatic against 8 cell lines with broad range of growth inhibitions (−98.48 to 98.86%). GI50 value of compound 31 was found in the range of 0.80–2.87 µM for 59 human cancer cell lines at five-dose concentration levels. DNA binding study of potent compound 31 was suggested that this compound was intercalated into DNA base pairs with binding constant of 1.25 × 104 M−1. Compound 31 showed effective binding with bovine serum albumin (BSA) and presented binding constant value of 3.79 ×104 M-1. Pharmacokinetic studies revealed that all compounds are following Lipinski’s rule of five and expected to be orally active.

. Over the past years, benzimidazole derivatives have been widely studied for their antimalarial 34 , anticancer 35 , antiprotozoal 36 , anti-inflammatory, and analgesic 37 activities. On the other hand, imidazo [1,2-a]pyrazine is also reported as cyclic nucleotide phosphodiesterase inhibitor 38 , anticancer 39 , anti-inflammatory 40 , antioxidant 41 , antimicrobial 42 , antiviral 43 , and antimalarial agents 44 . Thus, the designing of benzimidazole-imidazo[1,2-a] pyrazine hybrid ring system has been taken that provides new compounds related to various biological activities of benzimidazole and imidazopyrazine, in the hope that new anti-tumor agents might be discovered. Alicyclic substitution at NH of benzimidazole moiety can enhance the pharmacokinetic properties of the scaffold. By taking these perspectives into consideration, the lead molecule was designed. To explore the effect of substitution on cytotoxicity, a library of compounds was synthesized by modification of different substitution on phenyl ring and benzimidazole moiety (Fig. 1).

Results and Discussion
Chemistry. Suzuki-Miyaura cross-coupling approach has been employed for the synthesis of imidazo [1,2-a] pyrazine-benzimidazole hybrids (9-23, 30 and 31) and outlined in Figs. 2 and 3. Nitration of 1,4-dibromobenzene 1 afforded intermediate 2 in 93% yield followed by nucleophilic substitution with cyclohexylamine in the presence of K 2 CO 3 in DMF at 100 °C for 18 h to obtain 4-bromo-N-cyclohexyl-2-nitroaniline 3 in 75% yield. Formation of intermediate 3 was supported by NMR spectral analysis, the characteristic multiplet at δ 3.52-3.43 ppm for one proton and signals ranging from 2.14 to 1.25 for ten protons corresponding to cyclohexyl ring. Boronation of intermediate 3 with bis(pinacolato)diboron in 1,4-dioxane in the presence of Pd(PPh 3 ) 2 Cl 2 and KOAc afforded product 4 in 82% yield. Compound 4 was further characterized by NMR with characteristic signals of four methyl groups of boronate having singlet of 12 protons at δ 1.32 ppm.
The synthesis of bisbenzimidazole (benzimidazole at C-6 and C-8 positions of imidazo[1,2-a]pyrazine) has been shown in Fig. 3. First, nitration of 1,3-dibromobenzene with nitric acid afforded derivative 25 in 96% yield followed by treatment with cyclohexylamine in the presence of K 2 CO 3 in DMF at 100 °C yielded aminated precursor 26 in quantitative yield. Derivatives 3 and 26 were then reduced with sodium dithionite in THF and water to afford intermediate 27a-b. Derivatives 27a-b were cyclized using triethyl orthoformate in the presence of acetic acid to get benzimidazole 28a-b. Boronation was carried out with bis(pinacolato)diboron using Pd(PPh 3 ) 2 Cl 2 and KOAc in dioxane afforded intermediate 29a-b. Suzuki-Miyaura cross-coupling of intermediates 29a and 29b with 8 were accomplished using Pd(PPh 3 ) 4 and K 2 CO 3 to obtain 30 and 31 in 75% and 81% yields, respectively. Compounds 30 and 31 were distinguished by difference in the signal of CH corresponding to cyclohexyl group in 1 H NMR. All the final compounds were confirmed by NMR and mass spectrometry (Figs S1-S54).
Cytotoxicity. Cytotoxicity at one dose concentration (10 µM). Amongst all newly synthesized derivatives, nine compounds were selected by NCI for their in-vitro cytotoxicity at single dose concentration (10 µM) towards sixty subcancer cell lines of nine main panels. Analyzed all compounds exhibited diverse activity for different cancer cells (Table S1). The bisbenzimidazole derivatives 30 and 31 with benzimidazole rings at C6 and C8 positions of imidazo[1,2-a]pyrazine showed promising cytotoxicity against human cancer cell lines. Compound 31 exhibited a broad spectrum of activity and showed sensitivity against all 59 tested cancer cell lines. Derivative 31 has been observed as most potent derivative in each group of cancer. Derivative 31 was found to be cytotoxic effect against 51 cell lines and cytostatic effect towards 8 cell lines with a broad range of growth inhibition (−98.48 to 98.86%). SK-MEL-5 (melanoma) and HCC-2998 (colon cancer) cell lines have been found to be the most sensitive cell lines for compound 31 (Fig. 4). Compound 30 showed maximum inhibitory effect against RPMI-8226 (leukemia) amongst all cancer cell lines with growth inhibition of 52.18%. About 50% growth inhibition for HL-60 (TB) and MOLT-4 (leukemia) cancer cell lines was observed for compound 12 having naphthalene substitution at the C6 position of imidazo [1,2-a]pyrazine. Similarly, 3-thiophene substitution (compound 14) displayed more than 50% growth inhibition against T-47D (breast cancer) cancer cell lines.
Interestingly, among all evaluated compounds, compound 31 was found to be most active in the preliminary test at single dose concentration and thus preceded to five dose concentration assays.
Cytotoxicity at five dose concentrations (0.01-100 µM). In vitro screening of compound 31 against full panel of 60 cancer cell lines at five different concentrations has been performed and results are shown in Table 1. Compound 31 exhibited broad spectrum of growth inhibition for nine panels of cell lines with GI 50 values in the range of 0.80-2.87 µM and full panel mean graph mid-point (MG-MID) to be 2.12 µM. Compound 31 was observed to be sensitive towards most of the given cell lines and showed excellent activity with RPMI-8226 of leukemia having GI 50 value of 806 nM.
Cytotoxicity at normal cell line. To check the safety profile, compounds 31 was evaluated against Human Embryonic Kidney cells (HEK293) through MTT assay. The results indicated that compound has not shown any significant toxicity against HEK293 cells at 10 −4 , 10 −5 , 10 −6 , 10 −7 and 10 −8 M concentrations, suggesting great potential for their in-vivo use as antitumor agents (Fig. 5). It was found that on treating the cells with compound 31, the % survivals of HEK293 cells were well above 79%, 83%, 84%, 89% and 90% at 10 −4 , 10 −5 , 10 −6 , 10 −7 and 10 −8 M concentrations, respectively. This clearly indicated that these compounds were well within the toxicity limits and thereby, exhibiting good safety profile and have a high prospective for in vivo use as antitumor agents.
DNA binding studies. DNA is a key drug target and numerous derivatives show their antitumor activity by binding to DNA and interfering with DNA replication and preventing the growth of cancer cells, which is the foundation of designing novel and potent anticancer agents. The efficiency of DNA targeting drug depends upon its binding mode and affinity 9 . Thus, DNA binding study of small molecule is essential in the development of novel therapeutic substance 10 . Therefore, the interaction of compound 31 with DNA was analyzed by a number of techniques, such as UV-Vis absorption, fluorescence, and circular dichroism spectroscopy.
Absorption spectral studies. The electronic absorption spectrum of compound 31 (20 µM) consists of band in the range of 270-400 nm in phosphate buffer (pH 7.4). Compound 31 showed the high energy absorption band in the spectrum at 290 nm. Upon increasing concentrations of CT-DNA (0-15 µM), the band at 290 nm showed hypochromism (Fig. 6a). Titration was extended until the saturation point. These results suggested that the compound used in this study showed binding to DNA in an intercalative mode. To find the binding efficiency of derivative 31, binding constant (K b ) with CT-DNA has been obtained by applying the Benesi-Hildebrand equation (Eq. 1) 11 and calculated to be 1.25 × 10 4 M −1 ( Table 2) ( Figure S55). The observed value of binding constant (K b ) revealed that compound 31 was effectively bound with DNA.
Fluorescence spectral studies. To investigate the effect of adding DNA on compound 31, luminescence study has been used to get the information about binding constant and binding mode. The fluorescence spectrum of compound 31, exhibiting an emission band at 445 nm (λ ex = 295 nm), was monitored at a fixed concentration of 5 µM in phosphate buffer having pH 7.4 at 298 K. On addition of CT-DNA (0-50 µM) into solution of compound 31, gradual quenching of fluorescence intensity by 75% was detected (Fig. 6b). The result proposes that the quenching of fluorescence intensity of compound 31 might be due to the interaction between CT-DNA and compound. Stern-Volmer equation (equation-2) was applied to calculate the value of K sv (Stern-Volmer constant) and K q (apparent bimolecular quenching constant) 12 . The value of K sv (Stern-Volmer constant) has been obtained from the slope and was found to be 93 × 10 2 M −1 ( Figure S56a). Linear Stern-Volmer plot showed a single quenching process, either static or dynamic. The value of K q has been calculated using τ o (lifetime of the fluorophore) 13 = 10 −8 s and was found to be 93 × 10 10 M −1 s −1 ( Table 2). The value of K q is greater than maximum dynamic quenching constant (~1 × 10 10 M −1 s −1 ) 14 , represents that interaction of DNA with compound 31 is probably involved the static quenching mechanism.
The binding constant (K bin ) and the number of binding sites (n) for static quenching interaction of DNA with compound 31 were determined by Scatchard equation (Eq. 3). 15 The calculated value of K bin = 9.58 × 10 4 M −1 suggested that derivative 31 has strong binding affinity towards DNA. The number of binding sites (n) was obtained to be 1.21, indicating 1:1 stoichiometry between compound 31 and DNA ( Figure S56b).
Competitive binding between compound 31 and ethidium bromide for CT-DNA. Competitive binding experiment with compound 31 was conducted to get more evidence for the binding of the compound to DNA. Ethidium bromide (EtBr) shows no emission band in the buffer 16 . EtBr displayed enhanced emission band at 606 nm upon intercalating with CT-DNA, when excited at 520 nm. The EtBr-CT-DNA (3 µM: 30 µM) complex showed significant quenching after addition of compound 31 (0-80 µM) due to the displacement of EtBr from DNA, suggesting strong binding of compound 31 to DNA (Fig. 6c). The detected linearity in the plot of F o /F versus concentration of compound 31 is in good agreement with linear Stern-Volmer equation (Eq. 2) 12 . The Stern-Volmer plot ( Figure S57) has been used to determine the quenching constant and was calculated to be 1. 16   Bovine serum albumin (BSA) binding interaction. Serum albumins are the major plasma proteins and impart crucial role in nutrients and exogenous drug transport to the cells and tissues, and their metabolism 19 . Serum albumins are important blood proteins that have their ability to transport multitude of ligands. The binding ability of drug-albumin in the bloodstream has a significant impact on distribution, free concentration, metabolism and toxicity of drug 20 . Optimal interactions of any bioactive compound with serum albumin may increase drug efficiency. Bovine serum albumin (BSA) is the most extensively studied serum albumin owing to its structurally similar to human serum albumin.
Absorption spectroscopic studies. To find the interaction between compound 31 and BSA, UV-Vis spectra were recorded. The absorption spectrum of a fixed amount of BSA (10 µM) in phosphate buffer of pH 7.4 has been recorded with increasing concentration of compound 31 (0-8 µM). The absorption spectrum of BSA showed a band at 280 nm as the results of aromatic amino acids present in structure of BSA. An increase in intensity of the absorption band at 280 nm has been achieved with incremental addition of derivative 31 without affecting position of band (Fig. 7a). These changes were obtained due to the variations in the conformation of BSA along with changes in the microenvironment polarity of aromatic residues, indicating interaction between compound 31 and BSA protein. To check the binding affinity of compound 31 with BSA, binding constant (K b ) was determined by the Benesi-Hildebrand equation (Eq. 1) 11 and calculated to be 3.79 ×10 4 M -1 ( Figure S58).
Fluorescence quenching studies with BSA. Bovine serum albumin binding affinity with compound 31 was examined by emission quenching of tryptophan. BSA shows an emission band near 350 nm, due to the presence of trp-212 and trp-134 residues in its structure, on excitation at 295 nm. A solution of fixed amount of BSA  01-100 µM). NT = not tested; MG-MID = average sensitivity of derivative against all cancer cell (µM); GI 50 = concentration of the derivative required for 50% of maximal growth inhibition; TGI = concentration of the derivative required for total growth inhibition; LC 50 = concentration of the derivative required to kill 50% of population. www.nature.com/scientificreports www.nature.com/scientificreports/ (10 µM) when titrated with incremental addition of compound 31 (0-16 µM), displayed quenching of emission band at 350 nm up to 70% of the initial fluorescence intensity of BSA (Fig. 7b). A hypsochromic shift (5 nm) was appeared as a result of formation of complex between compound 31 and BSA. The Stern-Volmer quenching constant (K sv ) and apparent bimolecular quenching constant (K q ) of compound 31 have been calculated with Stern-Volmer equation (Eq. 2) 12 and were found to be 1.04 ×10 5 M −1 and 1.04 × 10 13 M −1 s −1 , respectively (Table 2) ( Figure S59a). The calculated value of K sv for compound 31 is demonstrating promising binding affinity of compound for serum albumin. The linearity of the Stern-Volmer plot denotes single quenching phenomenon, either static (complex formation by quencher and fluorophore) or dynamic (collision process) 14 . A higher value of K q indicating the existence of static quenching phenomena and the formation of the complex between compound 31 and BSA.
The binding constant (K bin ) and the number of binding sites (n) for interaction of compound 31 with BSA have been obtained from the Scatchard equation (Eq. 3) 15 and were calculated to be 4.70 × 10 4 M −1 and 0.92, respectively (Table 2) ( Figure S59b). The value of binding constant (K bin ) suggested that BSA has a good affinity to compound 31, as the known K bin value of non-covalent interaction between BSA and drug is generally in the range of 10 4 -10 6 M −1 . The binding constant for interaction of compound 31 and BSA recommended that compound can easily be transported by the protein.  www.nature.com/scientificreports www.nature.com/scientificreports/ In order to estimate the distribution of compound 31, complex of compound 31 (10 µM) and BSA (10 µM) was titrated with increasing concentration of ibuprofen (0-15 µM). On excitation at 280 nm, the emission band of complex at 350 nm was effectively quenched by the addition of ibuprofen ( Figure S60a). Results showed that ibuprofen efficiently displayed compound 31 from the complex. The dissociation constant for compound 31 was calculated using the Stern-Volmer equation (Eq. 2) and found to be 1.07 × 10 5 M −1 ( Figure S60b). The calculated dissociation constant was found equal to quenching constant for compound 31 that concludes the potential dissociation of compound 31 from BSA.

Förster resonance energy transfer (FRET) studies.
To determine the spatial distances between the donor (BSA) and the acceptor (compound 31), Förster resonance energy transfer (FRET) phenomenon was used 21 . According to the fundamental Förster mechanism, the energy transfer occurs when emission spectrum of the donor (BSA) overlap with the absorption spectrum of the acceptor (compound 31). The fluorescence energy transfer method can be used to determine the Förster distance (r) between BSA and compound 31. Figure 7c showed the overlap pattern of UV-visible absorption spectrum of compound 31 with emission spectrum of BSA. According to Eqs. 5 to 7, the parameters were calculated as the overlap integral value (J) between the emission spectrum of donor (BSA) and the absorption spectrum of acceptor (compound 31) = 5.35 × 10 −15 cm 3 L mol −1 , critical distance (R o ) = 2.29 nm, energy transfer efficiency (E) = 64% (0.64), and distance between BSA and the compound 31 (r) = 2.08 nm. The distance between BSA and compound 31 (r) is less than 10 nm which is full agreement with the rule 0.5 R o < r < 1.5 R o . Thus, the transfer of energy from BSA to derivative 31 must occur with high probability 22 . Physicochemical properties evaluation. Significant pharmacokinetic properties of any drug candidate can overcome the problem of failure of clinical trials during development. The rule-of-five (RO5) deals with orally active candidates and describe various ranges of four physicochemical factors (log P ≤ 5, Mol. Wt ≤ 500, no. of H-bond acceptors ≤ 10 and no. of H-bond donors ≤ 5) for suitable aqueous solubility and intestinal permeability 23 . We have studied all the synthesized compounds (9-23, 30 and 31) for these pharmacokinetic parameters. These compounds were found to have experimental log P values less than five and (Eq. 8) have molecular weight less than 500 except compounds 30 and 31. All compounds have found number of hydrogen bond acceptors less than 10 and the number of hydrogen bond donors less than 5, according to the rule-of-five. It was also observed that these compounds showed % absorbance (ABS) in the range of 86.54-92.43% (Table 3) (Eq. 9), indicating good bioavailability. These observations projected that these compounds are expected to be orally active because they are following the parameters of Lipinski's rule of five.  (Table S2).  www.nature.com/scientificreports www.nature.com/scientificreports/ The docking studies of derivative 31 with DNA exhibited hydrogen bonding (d = 2.60 Å) interaction between nitrogen of benzimidazole ring and hydrogen (H-21) atom linked with nitrogen (N-2) of guanine (DG-10) of chain A. Similarly, compound 31 also showed hydrogen bonding (d = 2.60 Å) interaction of nitrogen of benzimidazole ring with H-3 of nitrogen (N-3) of guanine base (DG-16) present in chain B (Fig. 8).

Conclusion
A new series of 17 compounds (9-23, 30 and 31) possessing imidazo[1,2-a]pyrazine and 1-cyclohexyl-1H-benzo[d]imidazole scaffold was synthesized using the Suzuki-Miyaura cross-coupling approach. Nine compounds (8-10, 12-14, 22, 30 and 31) were tested at single dose of 10 µM at NCI over 60 human cell line panel, amongst which compound 31 was subsequently tested at five-dose concentration levels. Compound 31 demonstrated potent and broad-spectrum anticancer activity over all the tested nine cancer types. Compound 31 was found to be cytotoxic against 51 cell lines and cytostatic against 8 cell lines with a broad range of growth inhibition (−98.48 to 98.86%). Compound 31 exhibited a broad spectrum of growth inhibition for nine panels of cell lines with GI 50 values in the range of 0.80-2.87 µM. Compound 31 showed strong interaction with CT-DNA by intercalation binding mode with a binding constant value of 1.25 × 10 4 M −1 . Compound 31 also exhibited strong affinity towards bovine serum albumin (BSA) with binding constant of 3.79 ×10 4 M -1 . From Forster's non-radiative energy transfer equations, it has been found that the distance of compound from BSA is 2.08 nm which predicts the possibility of energy transfer. All the compounds showed good pharmacokinetic properties and are expected to be orally active. Docking studies showed exceptional binding energy of DNA (PDB: 1BNA) with derivative 31 that has been found to be −11.1 Kcal/mol.

Experimental protocol
Cytotoxicity. Cytotoxicity was performed at National Cancer Institute (NCI), Bethesda, USA as per their protocols 26 .
MTT assay. HEK-293 cells (human embryonic kidney cells) were grown in Dulbecco's modified eagle's medium (DMEM) supplemented with 100 U/ml penicillin, 10% FBS, 100 mg/ml streptomycin and 50 mM glutamine. Furthermore, cells were harvested using trypsin and seeded in a 96-well cell culture plate for the MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay. The test compound was solubilized in cell culture grade DMSO. Once the cells in 96-well plates were 70% confluent, the compound was injected to the cells at five concentrations (0.01, 0.1, 1, 10, 100 µM) at 37 °C for 48 h (made in DMEM supplemented with 1% antibiotic and 10% FBS). After 48 h, cells were washed with PBS and added with 100 μL of fresh media in each well along with 10 μL of MTT reagent (5 mg/mL) for another 4 h. After 4 h, the media was again removed and added 100 μL cell culture grade DMSO to dissolve the formazan crystals formed by the reduction of MTT by live cells. The quantity of formed formazan crystal was determined as alterations in absorbance at 570 nm wavelength using ELISA plate reader (Bio-Tek). All tests were accomplished three times. The cell toxicity (%) was calculated using the following formula: All spectra were recorded in the range of 200-800 nm using reference and sample cuvettes of 1 cm path length. Phosphate buffer was used for the corrections of baseline. Titration procedures were repeated until not any further change in spectrum was observed, demonstrating that saturation in the binding process has been achieved. Absorption data were then fit to the Benesi-Hildebrand equation (Eq. 1) to get binding constant (K b ). Binding parameters K SV and K q were calculated using the Stern-Volmer equation (Eq. 2): Competitive binding fluorescence measurements. Ethidium bromide (EtBr) displacement experiments were performed by incremental addition of compound 31 to complex of EtBr-DNA. The EtBr (3 µM) and CT-DNA (30 µM) were titrated with incremental addition of compound 31 (0-85 µM) in phosphate buffer (pH 7.4). The emission spectra of the EtBr-CT-DNA complex were noted in the range of 200 nm-800 nm using 520 nm as excitation wavelength. The quenching constant (K q ) was calculated using the Stern-Volmer equation (Eq. 2). The apparent binding constant (K app ) was obtained from known binding constant for ethidium bromide (K EtBr = 1 × 10 7 M −1 ) using the following equation (Eq. 4).
app EtBr   F and F o are the BSA fluorescence intensities in the presence and absence of the derivative 31, r is the distance among the acceptor and the donor, whereas R o is the critical distance when the efficiency of energy transfer is 50%. R o can be calculated using Eq. 6 28 : where k 2 is the orientation factor of the dipole, ɳ is the refracted index of the medium, Ф the fluorescence quantum yield of the donor, and J is the overlap integral of the fluorescence emission spectrum of the donor with the absorption spectrum of the acceptor. The value of J can be calculated using Eq. 7 29 : Molecular docking. The AutoDock software package (vina) was used to execute the docking study of compound 31 with DNA (PdB: 1BNA). AutoDockTool (1.5.6rc3) was used to set up each ligand DNA interaction. Polar hydrogen atoms were added and water molecules were deleted. Gasteiger charges were calculated and nonpolar hydrogen atoms were merged to carbon atoms. To optimize the 3D structure of compound 31, Gaussian 09 W program was used and saved it in pdf format. The ADT package (version 1.5.6rc3) was used to modify the partial charges of the pdf file of compound 31 and saved the resulting file into Pdbqt format. The size of a grid box 44 Å, 78 Å, 106 Å, indicating x, y and z directions was retained throughout the docking. The spacing of the grid was used to be 0.375 Å. All default settings were used to perform the docking.

Statistical analysis.
The experiments were performed in triplicates and the results were displayed as mean ± SD (Standard Deviations). The Microsoft Office Professional Plus 13 (excel) was used to plot the graph, and to calculate SD and other binding parameters.