N-Phenacyldibromobenzimidazoles—Synthesis Optimization and Evaluation of Their Cytotoxic Activity

Antifungal N-phenacyl derivatives of 4,6- and 5,6-dibromobenzimidazoles are interesting substrates in the synthesis of new antimycotics. Unfortunately, their application is limited by the low synthesis yields and time-consuming separation procedure. In this paper, we present the optimization of the synthesis conditions and purification methods of N-phenacyldibromobenzimidazoles. The reactions were carried out in various base solvent-systems including K2CO3, NaH, KOH, t-BuOK, MeONa, NaHCO3, Et3N, Cs2CO3, DBU, DIPEA, or DABCO as a base, and MeCN, DMF, THF, DMSO, or dioxane as a solvent. The progress of the reaction was monitored using HPLC analysis. The best results were reached when the reactions were carried out in an NaHCO3–MeCN system at reflux for 24 h. Additionally, the cytotoxic activity of the synthesized compounds against MCF-7 (breast adenocarcinoma), A-549 (lung adenocarcinoma), CCRF-CEM (acute lymphoblastic leukemia), and MRC-5 (normal lung fibroblasts) was evaluated. We observed that the studied cell lines differed in sensitivity to the tested compounds with MCF-7 cells being the most sensitive, while A-549 cells were the least sensitive. Moreover, the cytotoxicity of the tested derivatives towards CCRF-CEM cells increased with the number of chlorine or fluorine substituents. Furthermore, some of the active compounds, i.e., 2-(5,6-dibromo-1H-benzimidazol-1-yl)-1-(3,4-dichlorophenyl)ethanone (4f), 2-(4,6-dibromo-1H-benzimidazol-1-yl)-1-(2,4,6-trichlorophenyl)ethanone (5g), and 2-(4,6-dibromo-1H-benzimidazol-1-yl)-1-(2,4,6-trifluorophenyl)ethanone (5j) demonstrated pro-apoptotic properties against leukemic cells with derivative 5g being the most effective.

Recently we described the synthesis and antifungal activity against Candida albicans and Candida neoformans of N-phenacyl-4,6/5,6-dibromobenzimidazoles [29]. These compounds were obtained by N-alkylation of appropriate dibromobenzimidazoles with that would allow the simplification of the procedure of their isolation and obtain the title compounds in satisfactory yields. In addition, for the first time we evaluated the cytotoxic effect of the synthesized compounds [68] on three tumor cell lines, i.e., MCF-7 (breast adenocarcinoma), A-549 (lung adenocarcinoma), CCRF-CEM (acute lymphoblastic leukemia), and MRC-5 (normal lung fibroblasts).

Synthesis of Title Compounds
N-Phenacyldibromobenzimidazoles 4-5a-j were obtained by N-alkylation of 5,6-or 4,6-dibromobenzimidazole (1 or 2) with various phenacyl halides 3a-j. The reaction conditions were optimized using a model reaction, N-alkylation of 5,6-dibromobenzimidazole 1 with 2,4-dichlorophenacyl chloride 3e (Scheme 1). The optimized parameters were: the type of solvent and base, the molar ratio of substrates, the temperature, and the time of the reaction. The conversion of each reaction was controlled by HPLC analysis. We started our investigations by evaluating the reaction of 1 with 3e in the K2CO3-MeCN system, using various molar ratios of the substrates ( Table 1, Entry 1-6). The best result was observed for a 3-fold excess of benzimidazole 1 over the alkylating agent 3e (Table 1, Entry 1). The yield of the reaction reached only 28%. Subsequent experiments were carried out using other base-solvent systems at rt for 24 h (  [20][21][22]. When the reaction was conducted in the presence of NaHCO3 in MeCN (Table 1, Entry 18) the yield was very low (<1%) and alkylating agent 3e remained unreacted. In other systems, despite low product 4e yields, high chloroketone conversion was observed. To increase the yield of the reaction, the following experiments (Table 1, Entry [24][25][26][27] were carried out at reflux for 3 h. Depending on the conditions, the product was obtained in yields ranging from 28% to 44%. The highest values were reached using the NaHCO3-MeCN system. Considering the results of further experiments (Table 1, Entry 28-31), the NaHCO3-MeCN system was chosen for optimization of the molar ratio of substrate 1: alkylating agent 3e: base. The highest yield (54%) was observed using the NaHCO3-MeCN system, with 0.3 mol excess of α-chloroketone 3e at reflux for 24 h (Table 1, Entry 34).  Legend: a all reactions were carried out with 0.1 mmol of 5,6-dibromobenzimidazole 1 and 5 mL of a solvent (5 mL) (see Supporting Information), b determined by HPLC (see Supporting Information S1).
Further optimization of the reaction conditions was carried out on a preparative scale ( Table 2) and yields of the reactions were calculated after the purification of the product. Legend: a at the beginning of the reaction 1.3 equiv. of 3e was added to the reaction flask; b at the beginning of the reaction 1 equiv. of 3e was added to the reaction flask, 0.5 equiv. of 3e was added after 17 h and 20 h; c crude compound 4e containing some by-products; d small amount of contaminated product 4e.
The reaction of equimolar amounts of compounds 1 and 3e with 20 equiv. of NaHCO 3 in 50 mL of MeCN at reflux afforded product 4e with the yield of 58% (after column chromatography), while at 60 • C the yield was reduced by 20% ( Table 2, Entry 1-2). When the molar ratio of compound 3e:1 increased to 1.3:1 or the volume of the solvent was reduced to 30 mL, the yield of 4e was not improved (Table 2, Entry 3,4). Further optimization was performed with a lower excess of NaHCO 3 (5 equiv.). In all cases (Table 2, Entry 5-9) the reactions were initially carried out with equimolar amounts of substrates. The additional portions of chloroketone 3e were added gradually, after 17 h and 20 h of reaction time. This method allowed us to receive product 4e in a yield of 67% (after column chromatography, Table 2, Entry 5). Further investigation of the product separation method revealed that good results were obtained by treating the crude product with a small volume of ethyl acetate and filtering the precipitated solid. This procedure allowed us to obtain product 4e in a 63-66% yield, using 50 and 40 mL of MeCN, respectively (Table 2, Entry 6-7). A further decrease in solvent volume to 30 and 20 mL did not improve the results. In the first case, product 4e contained some impurities ( Table 2, Entry 8), while in the case of the lowest volume of solvent only traces of the product were precipitated when ethyl acetate was added (Table 2, Entry 9).
Additionally, one more base-solvent system was investigated. Carrying out the reaction of equimolar amounts of substrates 1 and 3e in the presence of K 2 CO 3 in MeCN at reflux for 0.5 h, product 4e was afforded in a yield of 61%. This value was obtained when the product was purified by column chromatography. In this case, the precipitation method failed ( Table 2, Entry 10-11).
With the optimal conditions in hand, we carried out the reactions of dibromobenzimidazoles 1 and 2 with phenacyl chlorides 3e-j in the NaHCO 3 -MeCN system, at reflux, with benzimidazole/alkylating agent ratio 1/1 at the beginning. Further portions of chlorides 3 were added during the reaction. In most cases full conversion was observed after 24 h, using 1.5-2 equiv. of chloroketone 3. Usually, N-alkylation of 4,6-dibromobenzimidazole required smaller excesses of alkylating agent 3. Most of the crude products were precipitated from oils obtained after evaporation of the filtered reaction mixtures. In the case of 2,4,6-trifuoroderivatives 4j,5j and 2,4-difluoroderivative 4h, products were isolated by column chromatography.
The only exceptions were the N-alkylation of benzimidazoles 1,2 with 2,4,6-trichlorophenacyl chloride 2g. In the NaHCO 3 -MeCN system, no product was formed. The reactions were carried out in the presence of K 2 CO 3 in MeCN at reflux. Products 4g and 5g were isolated by column chromatography due to the formation of large amounts of by-products.
To compare the effect of base-solvent systems, the respective reactions of benzimidazoles 1,2 with phenacyl bromide 3a and their monosubstituted derivatives 3b-d were performed. In the case of 4,6-dibromobenzimidazole 2, a nearly full conversion after 16-20 h, using 1.05-1.25 equivalent of alkylating agent 3a-d was observed. The yields of compounds 5a-d exceeded 80%. Meanwhile, the rate of 5,6-dibromobenzimidazole 1 alkylation with 3a-d was significantly slower (after carrying out the reaction for 24 h with two equivalents of 3a-d added in portions, the full conversion of substrate 1 was often not observed). As a result, the yields of compounds 4a-d were lower than those obtained in the K 2 CO 3 -MeCN system at rt [29]. A double decrease in acetonitrile volume mainly resulted in the formation of more by-products, but it did not allow the reduction in the reaction time (Schemes 2 and 3, Table 3).  i, 2,5-F2C6H3 65 59 10 j, 2,4,6-F2C6H2 54 57 Legend: a conditions: 1.05 to 2 equiv. of phenacyl halide 3 (added gradually), NaHCO3 (5 equiv.), 40 mL MeCN/1 mmol of 1 or 2, reflux, 24 h; b reaction in K2CO3-MeCN system, 2.5 equiv. of phenacyl halide 3 (added gradually), reflux, 8 h. In the case of the alkylation of 4,6-dibromobenzimidazole, the formation of two isomeric products, having bromine atoms at 4,6 or 5,7 positions, is possible. The formation of the 4,6-isomer was unambiguously confirmed by the X-ray crystallography of compound 5d (Figure 1). In the case of the alkylation of 4,6-dibromobenzimidazole, the formation of two isomeric products, having bromine atoms at 4,6 or 5,7 positions, is possible. The formation of the 4,6-isomer was unambiguously confirmed by the X-ray crystallography of compound 5d (Figure 1).
To explain the low synthesis yields of compounds 4 and 5 in the K 2 CO 3 -MeCN system, we carried out the reaction of the model substrate, 2,4-dichlorophenacyl chloride 3e, under these conditions, for 24 h at rt. We observed nearly a full conversion of chloroketone 3e and the formation of a complex mixture of products with similar R f values. Three of these products were isolated. Based on NMR spectra, HRMS analysis, and the literature  [69][70][71][72][73][74], one of the compounds was assigned the structure of chloromethyl oxirane 6e. For the other two separated products, the structures were not unambiguously assigned. For one of these products, the structure of diepoxide 7e can be proposed (Scheme 4). Similar structures were reported earlier [73,75,76]. In the case of the alkylation of 4,6-dibromobenzimidazole, the formation of two isomeric products, having bromine atoms at 4,6 or 5,7 positions, is possible. The formation of the 4,6-isomer was unambiguously confirmed by the X-ray crystallography of compound 5d (Figure 1). To explain the low synthesis yields of compounds 4 and 5 in the K2CO3-MeCN system, we carried out the reaction of the model substrate, 2,4-dichlorophenacyl chloride 3e, Molecules 2022, 27, x FOR PEER REVIEW under these conditions, for 24 h at rt. We observed nearly a full conversion of chlorok 3e and the formation of a complex mixture of products with similar Rf values. Thr these products were isolated. Based on NMR spectra, HRMS analysis, and the liter data [69][70][71][72][73][74], one of the compounds was assigned the structure of chloromethyl ox 6e. For the other two separated products, the structures were not unambiguousl signed. For one of these products, the structure of diepoxide 7e can be proposed (Sc 4). Similar structures were reported earlier [73,75,76].
In the case of the last product, containing nine aromatic and three aliphatic pro the formation of trans-1,2,3-tribenzoylcyclopropane 8e was suggested. The analysis omatic proton signals indicated the presence of two identical benzene rings. On the hand, for such compounds, the presence of a doublet and triplet in the proton spec in the aliphatic range is characteristic. [71,77,78]. However, a doublet and a doub doublets are present in the spectrum of the isolated compound. Previous literature indicate that analogous compounds, with unsubstituted benzene rings, can be form reactions of phenacyl chloride under alkaline conditions [73,74,79,80].

Evaluation of Biological Activity
To evaluate the cytotoxic activity of compounds 4a-j and 5a-5j, we performe MTT test for three tumor cell lines, i.e.,: MCF-7 (breast adenocarcinoma), A-549 (lun enocarcinoma), CCRF-CEM (acute lymphoblastic leukemia), and one normal cell MRC-5 (normal lung fibroblasts). The EC50 values, describing the half maximal effe concentration of each tested compound, were calculated and are summarized in Ta The representative sigmoidal dose-response curves are shown in Figure S2 (see Sup ing Information S2). In the case of the last product, containing nine aromatic and three aliphatic protons, the formation of trans-1,2,3-tribenzoylcyclopropane 8e was suggested. The analysis of aromatic proton signals indicated the presence of two identical benzene rings. On the other hand, for such compounds, the presence of a doublet and triplet in the proton spectrum in the aliphatic range is characteristic. [71,77,78]. However, a doublet and a doublet of doublets are present in the spectrum of the isolated compound. Previous literature data indicate that analogous compounds, with unsubstituted benzene rings, can be formed in reactions of phenacyl chloride under alkaline conditions [73,74,79,80].

Evaluation of Biological Activity
To evaluate the cytotoxic activity of compounds 4a-j and 5a-5j, we performed an MTT test for three tumor cell lines, i.e.,: MCF-7 (breast adenocarcinoma), A-549 (lung adenocarcinoma), CCRF-CEM (acute lymphoblastic leukemia), and one normal cell line, MRC-5 (normal lung fibroblasts). The EC 50 values, describing the half maximal effective concentration of each tested compound, were calculated and are summarized in Table 4. The representative sigmoidal dose-response curves are shown in Figure S2 (see Supporting Information S2).  The study showed that most of the tested derivatives exhibited biological activity, with the exception of derivative 4d (4-Br), which was sparingly soluble in DMSO. Interestingly, derivative 5d (4-Br) demonstrated moderate activity against all the tested cell lines, with the lowest value of EC 50 , 27.75 µM for MCF-7. The sensitivity of the studied cell lines was different, and the MCF-7 line shows the greatest sensitivity to the tested compounds, with the lowest EC 50 for derivative 5g (2,4,6-Cl 3 ) (23.98 µM). This compound also showed the best activity against the CCRF-CEM line with an EC 50 of 26.64 µM; however, it was also cytotoxic to normal cells (MRC-5) with an EC 50 of 26.9 µM. Among the studied cell lines, A-549 cell line appeared to be the least sensitive to the tested compounds, with compound 4i (2,5-F 2 ) showing the most activity with an EC 50 of 37.87 µM. Interestingly, we observed that as the number of chlorine or fluorine substituents in 4,6-dibromobenzimidazole derivatives increased, their cytotoxicity was better towards CCRF-CEM cells, i.e., from the least cytotoxic 5b (4-F) and 5c  to the most active 5j (2,4,6-F 3 ) and 5g (2,4,6-Cl 3 ), respectively. However, among the tested compounds, the highest selectivity index was obtained for 5a (2.04) for MCF-7, respectively.

Molecules 2022, 27, x FOR PEER REVIEW
10 of 17 apoptosis effectively in CCRF-CEM (Figure 2) with the highest percent of apoptotic cells, i.e., 75.6% obtained after treatment with 5g (Figure 2b). This compound also induced apoptosis at a 30 µM concentration, giving 32% of cells in late apoptosis (Figure 2b). Graphs represent mean values ± s.e.m. *** p < 0.001 relative to control; ns-not significant. Statistical analysis for three to six independent replicates was performed using a one-way ANOVA analysis (GraphPad Software Inc., San Diego, CA, USA).

Materials and Methods
Commercially available reagents from Sigma Aldrich (Darmstadt, Germany) and Avantor (Gliwice, Poland) were used as supplied. Solvents: DMF, THF, DMSO, and MeCN (for reaction with NaH) were dried with standard methods. Thin-layer chromatography was carried out on TLC aluminum plates with silica gel Kieselgel 60 F254 (Merck, Darmstadt, Germany) (0.2 mm thickness film). The column chromatography was performed using Silica gel 60 (Merck, Darmstadt, Germany) of 40-63 µm.
Dimethyl sulphoxide (DMSO), Molecular Biology grade, used as a solvent for all stocks of the chemical agents, was obtained from Roth (Karlsruhe, Germany). All reagents used in flow cytometry analysis were purchased from BD Biosciences Pharmingen (San Diego, CA, USA).
The melting points were measured with a commercial apparatus Thomas-Hoover "UNI-MELT" on samples placed in glass capillary tubes and were not corrected. The 1 H and 13 C NMR spectra were measured with a Varian 500 spectrometer operating at 500 MHz for 1 H and 125 MHz for 13 C nuclei. Chemical shifts (δ) are given in parts per million (ppm) relative to the residual solvent signal (CDCl3, δH of residual CHCl3 7.26 ppm); signal multiplicity assignment: s, singlet; br s, broad singlet; d, doublet; dd, doublet of doublets; m, multiplet; coupling constants (J) are given in hertz (Hz). All these measurements were made at Warsaw University of Technology. High-resolution mass spectrometry (HRMS) was carried out on Q Exactive Hybrid Quadrupole-Orbitrap Mass Spectrometer (Bremen, Germany), ESI (electrospray) with spray voltage 4.00 kV at Institute of Biochemistry and . Cells were stained with annexin V-FITC and PI (propidium iodide). Flow cytometry analyses were run on FACSCanto II flow cytometer (BD Biosciences, San Diego, CA, USA) and analysed using BD FACSDiva software. Graphs represent mean values ± s.e.m. *** p < 0.001 relative to control; ns-not significant. Statistical analysis for three to six independent replicates was performed using a one-way ANOVA analysis (GraphPad Software Inc., San Diego, CA, USA).

Materials and Methods
Commercially available reagents from Sigma Aldrich (Darmstadt, Germany) and Avantor (Gliwice, Poland) were used as supplied. Solvents: DMF, THF, DMSO, and MeCN (for reaction with NaH) were dried with standard methods. Thin-layer chromatography was carried out on TLC aluminum plates with silica gel Kieselgel 60 F 254 (Merck, Darmstadt, Germany) (0.2 mm thickness film). The column chromatography was performed using Silica gel 60 (Merck, Darmstadt, Germany) of 40-63 µm.
Dimethyl sulphoxide (DMSO), Molecular Biology grade, used as a solvent for all stocks of the chemical agents, was obtained from Roth (Karlsruhe, Germany). All reagents used in flow cytometry analysis were purchased from BD Biosciences Pharmingen (San Diego, CA, USA).
The melting points were measured with a commercial apparatus Thomas-Hoover "UNI-MELT" on samples placed in glass capillary tubes and were not corrected. The 1 H and 13 C NMR spectra were measured with a Varian 500 spectrometer operating at 500 MHz for 1 H and 125 MHz for 13 C nuclei. Chemical shifts (δ) are given in parts per million (ppm) relative to the residual solvent signal (CDCl 3 , δ H of residual CHCl 3 7.26 ppm); signal multiplicity assignment: s, singlet; br s, broad singlet; d, doublet; dd, doublet of doublets; m, multiplet; coupling constants (J) are given in hertz (Hz). All these measurements were made at Warsaw University of Technology. High-resolution mass spectrometry (HRMS) was carried out on Q Exactive Hybrid Quadrupole-Orbitrap Mass Spectrometer (Bremen, Germany), ESI (electrospray) with spray voltage 4.00 kV at Institute of Biochemistry and Biophysics Polish Academy of Science (IBB PAS, Warsaw, Poland. The most intensive signals are reported.

Synthesis of N-Phenacylbenzimidazoles 4,5a-f, 4,5h-j
The mixture of benzimidazole 1 or 2 (4 mmol), MeCN (160 mL), phenacyl halide 3 (4 mmol), and NaHCO 3 (20 mmol, 1.68 g) was stirred magnetically at reflux. The reaction was monitored by TLC (CHCl 3 /MeOH 95/5 v/v) typically after 16, 20, and 24 h. If necessary, additional portions of phenacyl halide (1-4 mmol) were added after 16.5 and 20.5 h. In most cases, full conversion of benzimidazole 1 or 2 was observed after ca 24 h. The mixture was cooled, filtered through a short pad of celite, washed with MeCN (4 × 10 mL), evaporated to dryness. The solid residue was transferred to a Schott funnel and washed with the respective solvent (5-7 × 2-10 mL, see Supporting Information S2). The oily residue was purified by column chromatography (silica gel, CHCl 3 as eluent). If full conversion of benzimidazole 1 or 2 was not observed after 24 h, further portions of substrate 3 were added, the reaction was continued and worked-up as described above.

Cell Culture and Agents Treatment
CCRF-CEM (ECACC 85112105) human Caucasian acute lymphoblastic leukaemia and MRC-5 pd30 Human fetal lung fibroblasts (ECACC 05090501) were purchased from European Collection of Authenticated Cell Cultures, whereas A-549 (ATCC CCL 185) human lung carcinoma and MCF-7 (ATCC HTB-22) human Caucasian breast adenocarcinoma cell line were purchased from American Type Culture Collection. A-549 and MCF-7 cell lines were cultured in high glucose DMEM (Biowest) supplemented with 10% fetal bovine serum (Biowest), 2 mM L-glutamine and antibiotics (100 U/mL penicillin, 100 µg/mL streptomycin). CCRF-CEM were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (EuroClone), 2 mM L-glutamine and antibiotics (100 U/mL penicillin, 100 µg/mL streptomycin). MRC-5 pd30 human fibroblasts (ECACC) were cultured in MEME, Minimum Essential Medium Eagle (Merck) supplemented with 10% fetal bovine serum (Merck), 2 mM l-glutamine, antibiotics (100 U mL −1 penicillin, 100 µg mL −1 streptomycin, Merck) and 1% non-essential amino-acids (Merck). Cells were grown in 75 cm 2 cell culture flasks (Sarstedt, Nümbrecht, Germany), in a humidified atmosphere of CO 2 /air (5/95%) at 37 • C. All the experiments were performed in exponentially growing cultures. Stock solution of tested compounds were prepared in DMSO and stored in −20 • C for maximum one month. For the cytotoxicity studies, stock solutions were diluted 200-fold with the proper culture medium to obtain the final concentrations. For cytotoxicity studies 2-fold serial dilutions of the tested compounds were prepared in the proper medium in the range from 3.125 µM to 200 µM.

Detection of Apoptosis by Flow Cytometry
CCRF-CEM cells were seeded in 24-well plates at 2 × 10 5 cells/well and treated with the tested compounds used in 15 µM, 30 µM, and 45 µM concentrations. After exposure to the examined compounds, the cells were collected, centrifugated at 200× g at 4 • C for 5 min, washed twice with cold phosphate-buffered saline (PBS), and subsequently suspended in binding buffer. Subsequently, 100-µL aliquots of the cell suspension were labelled according to the kit manufacturer's instructions. Briefly, annexin V-fluorescein isothiocyanate, and propidium iodide (BD Biosciences, Pharmingen, San Diego, CA, USA) were added to the cell suspension, and the mixture was vortexed and then incubated for 15 min at RT in the dark. A cold binding buffer (400 µL) was then added and the cells were vortexed again and kept on ice. Flow cytometric measurements were performed within 1 h after labeling. Viable, necrotic, early, and late apoptotic cells were detected by FACSCanto II flow cytometer (BD Biosciences, San Diego, CA, USA) and analysed using BD FACSDiva software.

Conclusions
In summary, the above results indicate that N-phenacyldibromobenzimidazoles can be efficiently synthesized in different base-solvent systems depending on the substitution pattern in the phenacyl moiety. All compounds, substituted with two chlorine 4,5e,f or fluorine atoms 4,5h,i, as well as trifluoroderivatives 4,5j, are easily obtained in the NaHCO 3 -MeCN system at reflux, while synthesis of the most sterically crowded trichloroderivatives 4,5g required K 2 CO 3 instead of NaHCO 3 . On the other hand, unsubstituted 5a or monosubstituted 5b-d derivatives of 4,6-dibromobenzimidazole are efficiently synthesized in both the K 2 CO 3 -MeCN system at rt and the NaHCO 3 -MeCN system at reflux, while for isomeric 5,6-dibromoderivatives 4a-d, the first system is the best choice.
Taking into account the results of cytotoxicity studies, we conclude that the introduction of chlorine or fluorine substituents into 4,6-dibromobenzimidazole derivatives increases their cytotoxicity towards leukemic cells. On the one hand, the derivatives demonstrate some pro-apoptotic properties, which is an important feature of potential anticancer drugs; however, they are also cytotoxic to normal cells.
Further modifications and synthetic applications of the title compounds as well as evaluation of their biological activity are under investigation.
Funding: This research was funded by the Warsaw University of Technology (NChem2).

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.