New 1H-Benzo[f]indazole-4,9-diones Conjugated with C-Protected Amino Acids and Other Derivatives: Synthesis and in Vitro Antiproliferative Evaluation

1H-Benzo[f]indazole-4,9-dione derivatives conjugated with C-protected amino acids (glycine, l-alanine, l-phenylalanine and l-glutamic acid) 6a–l were prepared by chemically modifying the prenyl substituent of 3-methyl-7-(4-methylpent-3-enyl)-1H-benzo[f]indazole-4,9-dione 2 through epoxidation, degradative oxidation, oxidation and N-acyl condensation reactions. The chemical structures of the synthesized compounds were elucidated by analyzing their IR, 1H-NMR and 13C-NMR spectral data together with elemental analysis for carbon, hydrogen and nitrogen. The preliminary in vitro antiproliferative activity of the synthesized derivatives was evaluated on KATO-III and MCF-7 cell lines using a cell proliferation assay. The majority of the derivatives exhibited significant antiproliferative activity with IC50 values ranging from 25.5 to 432.5 μM. These results suggest that 1H-benzo[f]indazole-4,9-dione derivatives are promising molecules to be researched for developing new anticancer agents.

To rationalize the possible products, we performed full geometry optimizations of the structures using preliminary density functional theory (DFT) calculations (see the Experimental Section). With the aim of exploring the reactants and possible products of the first step of this reaction from a thermodynamics perspective, we subtracted the total bonding energies of the products from the total bonding energies of the reactants. The comparison between these energy changes indicated that the formation of 1' is 4-fold more favorable than the formation of 1T' (Scheme 1). Moreover, the HOMO-LUMO gaps of compounds 1' and 1T' were 2.28 eV and 1.39 eV, respectively, which indicates that the most stable derivative is compound 1' (Figure 1). These results suggest that cyclization must initially occur by the conjugate addition to afford the 3-substituted 1,4-hydroquinone compound 1', but a complete transition state (TS) search and intrinsic reaction coordinate (IRC) study are needed to confirm this hypothesis. For this reason, our research group is currently working to obtain a full reaction pathway for all reaction steps.
To rationalize the possible products, we performed full geometry optimizations of the structures using preliminary density functional theory (DFT) calculations (see the Experimental Section). With the aim of exploring the reactants and possible products of the first step of this reaction from a thermodynamics perspective, we subtracted the total bonding energies of the products from the total bonding energies of the reactants. The comparison between these energy changes indicated that the formation of 1' is 4-fold more favorable than the formation of 1T' (Scheme 1). Moreover, the HOMO-LUMO gaps of compounds 1' and 1T' were 2.28 eV and 1.39 eV, respectively, which indicates that the most stable derivative is compound 1' (Figure 1). These results suggest that cyclization must initially occur by the conjugate addition to afford the 3-substituted 1,4-hydroquinone compound 1', but a complete transition state (TS) search and intrinsic reaction coordinate (IRC) study are needed to confirm this hypothesis. For this reason, our research group is currently working to obtain a full reaction pathway for all reaction steps.  The second step of the reaction is the cyclization of the 3-substituted 1,4-hydroquinone compound 1' by the nucleophilic addition/elimination reaction between the amino and carbonyl groups of 1' to afford the fused pyrazolo-1,4-naphthohydroquinone compound 1''. Finally, this compound is oxidized to the 1,4-naphthoquinone 2 by the initial 2-acetyl-1,4-naphthoquinone 1 (Scheme 2), a step that is supported by the isolation of 2-acetyl-6-(4-methylpent-3-enyl)-1,4-naphthohydroquinone as a by-product [10,28]. To synthesize the new 1H-benzo[f]indazole-4,9-diones conjugated with C-protected amino acids 6a-l, we followed the synthetic pathway shown in Scheme 3. The epoxidation of the double bond in the 7-(4-methylpent-3-enyl) group of 2a-c to afford oxiranyl compounds 3a-c was accomplished with m-chloroperoxybenzoic acid (mCPBA), and treatment of these compounds with periodic acid afforded the aldehydes 4a-c [29]. Oxidation of these aldehydes to the carboxylic acids 5a-c was performed with sodium chlorite in the presence of a catalytic amount of 2-methyl-2-butene. The reactivity of the The second step of the reaction is the cyclization of the 3-substituted 1,4-hydroquinone compound 1' by the nucleophilic addition/elimination reaction between the amino and carbonyl groups of 1' to afford the fused pyrazolo-1,4-naphthohydroquinone compound 1". Finally, this compound is oxidized to the 1,4-naphthoquinone 2 by the initial 2-acetyl-1,4-naphthoquinone 1 (Scheme 2), a step that is supported by the isolation of 2-acetyl-6-(4-methylpent-3-enyl)-1,4-naphthohydroquinone as a by-product [10,28]. The second step of the reaction is the cyclization of the 3-substituted 1,4-hydroquinone compound 1' by the nucleophilic addition/elimination reaction between the amino and carbonyl groups of 1' to afford the fused pyrazolo-1,4-naphthohydroquinone compound 1''. Finally, this compound is oxidized to the 1,4-naphthoquinone 2 by the initial 2-acetyl-1,4-naphthoquinone 1 (Scheme 2), a step that is supported by the isolation of 2-acetyl-6-(4-methylpent-3-enyl)-1,4-naphthohydroquinone as a by-product [10,28]. To synthesize the new 1H-benzo[f]indazole-4,9-diones conjugated with C-protected amino acids 6a-l, we followed the synthetic pathway shown in Scheme 3. The epoxidation of the double bond in the 7-(4-methylpent-3-enyl) group of 2a-c to afford oxiranyl compounds 3a-c was accomplished with m-chloroperoxybenzoic acid (mCPBA), and treatment of these compounds with periodic acid afforded the aldehydes 4a-c [29]. Oxidation of these aldehydes to the carboxylic acids 5a-c was performed with sodium chlorite in the presence of a catalytic amount of 2-methyl-2-butene. The reactivity of the To synthesize the new 1H-benzo[f ]indazole-4,9-diones conjugated with C-protected amino acids 6a-l, we followed the synthetic pathway shown in Scheme 3. The epoxidation of the double bond in the 7-(4-methylpent-3-enyl) group of 2a-c to afford oxiranyl compounds 3a-c was accomplished with m-chloroperoxybenzoic acid (mCPBA), and treatment of these compounds with periodic acid afforded the aldehydes 4a-c [29]. Oxidation of these aldehydes to the carboxylic acids 5a-c was performed with sodium chlorite in the presence of a catalytic amount of 2-methyl-2-butene. The reactivity of the carboxylic group was then enhanced through the in situ formation of the mixed anhydride with ethyl chloroformate followed by the addition of the corresponding methyl ester of glycine, L-alanine, L-phenylalanine and L-glutamic acid [25]. In all of these synthesized compounds, the L-configuration must be retained in the amino acid unit. The physical and analytical data of the compounds are presented in the experimental section along with the IR, 1 H and 13 C spectroscopic data; chemical shifts are reported according to the carbon numbering of compounds 2 in Scheme 3.
Molecules 2015, 20, page-page 4 carboxylic group was then enhanced through the in situ formation of the mixed anhydride with ethyl chloroformate followed by the addition of the corresponding methyl ester of glycine, L-alanine, L-phenylalanine and L-glutamic acid [25]. In all of these synthesized compounds, the L-configuration must be retained in the amino acid unit. The physical and analytical data of the compounds are presented in the experimental section along with the IR, 1 H and 13 C spectroscopic data; chemical shifts are reported according to the carbon numbering of compounds 2 in Scheme 3. The common features from the spectral data of compounds 6a-l are closely related to those previously reported for the starting compounds 2a-c [23], and they are as follows: - In some cases, their IR spectra show two carbonyl-quinone absorptions at approximately 1680 and 1670 cm −1 , but the latter absorption is primarily observed. - In the 1 H spectra, the singlet of the C-10 methyl group appears at approximately 2.60 to 2.80 ppm, the coupled methylene groups of C-11 and C-12 carbons show triplets or multiplets between 2.50 and 3.00 ppm (J = 7.3-8.0 Hz), and the coupled aromatic hydrogen of carbon C-5, C-6 and C-8 are observed as doublets of doublets and two doublets at 7.70 to 8.10 ppm (J = 7.6 and 1.6 Hz). - The 13 C-NMR spectra contain signals for carbonyl-quinone C-4 and C-9 carbon atoms at 170 to 180 ppm.

Biological Assay
The antiproliferative activity of the synthesized compounds was assessed on KATO-III and MCF-7 cell lines using a CellTiter 96 ® AQueous One Solution Proliferation Assay (MTS) from Promega The common features from the spectral data of compounds 6a-l are closely related to those previously reported for the starting compounds 2a-c [23], and they are as follows: - In some cases, their IR spectra show two carbonyl-quinone absorptions at approximately 1680 and 1670 cm´1, but the latter absorption is primarily observed. - In the 1 H spectra, the singlet of the C-10 methyl group appears at approximately 2.60 to 2.80 ppm, the coupled methylene groups of C-11 and C-12 carbons show triplets or multiplets between 2.50 and 3.00 ppm (J = 7.3-8.0 Hz), and the coupled aromatic hydrogen of carbon C-5, C-6 and C-8 are observed as doublets of doublets and two doublets at 7.70 to 8.10 ppm (J = 7.6 and 1.6 Hz). - The 13 C-NMR spectra contain signals for carbonyl-quinone C-4 and C-9 carbon atoms at 170 to 180 ppm.

Biological Assay
The antiproliferative activity of the synthesized compounds was assessed on KATO-III and MCF-7 cell lines using a CellTiter 96 AQueous One Solution Proliferation Assay (MTS) from Promega (Madison, WI, USA) with doxorubicin as a control. The results were expressed as the concentration determining 50% inhibition of cell proliferation (IC 50 ).
Tables 1 and 2 show the IC 50 values for the antiproliferative activity obtained for each derivative tested in KATO-III and MCF-7 cell lines, respectively. Each column in Tables 1 and 2 contains the IC 50 values for derivatives belonging to Series-I, -II and -III, respectively.
By comparing Tables 1 and 2 is possible to observe the similarity between the patterns generated from the IC 50 values obtained from Series-I, -II and -III derivatives. Moreover, these patterns were very similar in both cell models.    A closer analysis using a two-way ANOVA test followed by a Dunnett's multiple comparison post-test showed that the most promising derivatives were compounds 2 (i.e., 2a, 2b and 2c), compounds 4 (i.e., 4a, 4b and 4c) and derivatives conjugated with L-alanine (i.e., 6b, 6f and 6j), L-phenylalanine (i.e., 6c, 6g and 6k) and L-glutamic acid (i. e., 6d, 6h and 6l). Additionally, the statistical analysis shows that the compounds of Series-II and -III have better IC 50 values compared to compounds of Series-I.

General
All reactions were performed using reagents and solvents purchased from commercial sources and purified by standard procedures as necessary. Starting N-substituted 1H-benzo[f ]indazole-4,9-diones 2a-c were synthesized according to a previously described procedure [23]. IR spectra were recorded on a Perkin Elmer FT IR 1600 spectrophotometer (Norwalk, CN, USA) as a film over NaCl discs. NMR spectra were recorded on a Bruker Avance 400 Digital NMR spectrometer (Bruker/Analytic, Karlsruhe, Germany) operating at 400.13 MHz for 1 H and 100.62 MHz for 13 C in CDCl 3 , acetone-d 6 or DMSO-d 6 with internal TMS as a reference. Chemical shifts (δ) were expressed in ppm, followed by multiplicity and coupling constant (J) in Hz. Elemental analyses of C, H and N were performed using a Perkin Elmer 2400 Series II CHN Elemental Analyzer (Perkin Elmer Inc., Waltham, MA 02451, USA). The reaction progress was monitored by thin layer chromatography with Silica gel 60 F 254 (0.25 mm thick, Merck, Darmstadt, Germany) aluminum sheets, whereas column chromatographies were performed on Silica gel 60 (230-400 mesh, Merck) using solvent mixtures with variable proportions as eluents. Melting points were determined on a Stuart SMP 10 apparatus (Stone, Staffordshire, UK), and they were not corrected.

General Procedure for the Preparation of [3-(3-Methyl-4,9-dioxo-4,9-dhydro-1H-benzo[f ]indazol-7-yl)propanoylamino]-methyl Ester 6a-l
A solution containing 0.37 mmol of carboxylic acids 5a-c, 0.041 g (0.407 mmol, 56 µL) of triethylamine and 0.044 g (0.407 mmol, 38 µL) of ethyl chloroformate in 12 mL of dry THF was stirred for 20 min at 0˝C. After the addition of 0.407 mmol of the protected L-amino acid (Gly, Ala, Phe, and Glu), the mixture was stirred 16 h at r.t. After filtration over Celite-545 and evaporation of the solvent, the residue was dissolved in 70 mL of ethyl acetate. The organic solution was extracted with 40 mL of a 5% NaHCO 3 solution and water (40 mL). After drying with Na 2 SO 4 , the solvent was removed under reduced pressure, and the crude product was purified by column chromatography with chloroform/acetone 7:3 as the eluent.

Computational Details
DFT calculations [30][31][32][33] were conducted using the Amsterdam Density Functional (ADF) program [34]. The Vosko-Wilk-Nusair parametrization [35] was used to treat electron correlation within the local density approximation (LDA). The numerical integration procedure applied for the calculation was developed by teVelde [33]. The standard ADF TZ2P basis set was used for all atoms. The frozen core approximation was used to treat core electrons at the following levels: C, 1s; N, 1s; and O, 1s [33]. Full geometry optimizations were performed on each complex using the analytical gradient method implemented by Verluis and Ziegler [36]. The geometries for all the model compounds discussed in the text were fully optimized and checked via analytical frequency calculations as either true minima (no imaginary values).

Antiproliferative Assay
KATO-III (human gastric cancer cell line) and MCF-7 (human breast adenocarcinoma cell line) cells were obtained from the American Type Culture Collection (ATCC). KATO-III and MCF-7 cells (2ˆ10 3 ) were grown in DMEM supplemented with 10% FBS and 1% penicillin/streptomycin. Cells were subcultured into fresh medium (100-mm-diameter plate dish) until a density of approximately 80% was obtained. Briefly, 2ˆ10 3 cells were seeded in 96-well culture plates. After 24 h of incubation at 37˝C in a humidified 5% CO 2 atmosphere, different concentrations (10´9 to 10´3 M) of 1H-benzo[f ]indazole-4,9-dione-based derivatives were added. After 72 h of incubation, 20 µL of MTS (Promega, Madison, WI, USA) was added, and the wells were incubated for an additional 2 h at 37˝C. The absorbance at 490 nm was recorded using a Varioskan Flash Multimode Reader (Thermo Scientific, Waltham, MA, USA). Each variant of the experiment was performed in triplicate. To obtain IC 50 values for each compound, dose-response curves were constructed in both KATO-III and MCF-7 cell lines. Doxorubicin was included in all evaluation to provide a reference of antiproliferative activity.

Statistical Analysis
Data are expressed as means˘CI 95% (95% confidence intervals) for three independent experiments. The concentration inducing a 50% decrease of cell proliferation (IC 50 ) was performed using the four-parameters logistic fit-also known as "4PL"-supported by GraphPad Prism 6 (GraphPad Software, San Diego, CA, USA). Statistical differences among means were assessed using a two-way ANOVA test followed by a Dunnett's multiple comparison post-test. A p < 0.05 was taken as statistically significant.

Conclusions
In this study, we have synthesized three series of new 1H-benzo[f ]indazole-4,9-dione-based derivatives containing oxiranyl, formyl, carboxylic and L-and C-protected N-aminoacidyl substituents attached to the side chain of the 1,4-naphthoquinone group in moderate to good yields.