Thirty novel derivatives of aza-cyclopenta[b]fluorene-1,9-dione were synthesized, and their cytotoxic activities were tested against HeLa, LS180, MCF-7, and Raji cancer cell lines by MTT assay. Two derivatives containing nitrofuryl moiety, including 10-(5-nitro-furan-2-yl)-2,3-dihydro-4-aza-cyclopenta[b]fluorene-1,9-dione (IC50 range: 5.7–13.0 μm) and 10-(5-Nitro-furan-2-yl)-2,3,4,10-tetrahydro-4-aza-cyclopenta[b]fluorene-1,9-dione (IC50 range: 3.6–20.2 μm), as well as 10-(2-Nitro-phenyl)-2,3,4,10-tetrahydro-4-aza-cyclopenta[b]fluorene-1,9-dione (IC50 range: 3.1–27.1 μm) with nitrophenyl moiety on C10 position, were the most effective compounds. Furthermore, the effect of physiochemical descriptors on the cytotoxicity was evaluated by quantitative structure–activity relationship analysis. The quantitative structure–activity relationship results showed that molecular dipole moment, molar refractivity, fragment-based parameters, and some topological indices were influential on the cytotoxic effect. Finally, the good correlation that was found among cytotoxic data obtained from different cell lines may be an implication of a common cytotoxic mechanism in these cell lines. These findings provide useful structural information for the rational design and synthesis of efficient chemotherapeutic agents for treatment for cancer.
The condensed polycyclic heteroaromatic compounds are naturally occurring compounds with widespread biological activities (1). Different polycyclic compounds have demonstrated promising cytotoxic activities (2,3). One of the common properties of these compounds is the presence of planar polycyclic chromospheres, which can intercalate into the DNA macromolecule and accounts for the cytotoxic properties of these compounds. They also have at least one basic side chain, which can improve DNA binding affinity and in some cases enhance solubility under physiological conditions (4). Different studies have demonstrated that the planar portion of the molecule could be tricyclic, tetracyclic, or even pentacyclic. Although some results have shown a positive relationship between DNA binding and cytotoxicity, it is generally accepted that DNA intercalation is necessary but not enough for the antitumoral activity (4,5). Therefore, the condensed heterocyclic system has been receiving much attention for further modification as a cytotoxic scaffold and different heterocyclic systems such as acridin and acridinone (6–8), diindenopyridine derivatives (5) and others have been evaluated for this purpose.
In continuation of our work on condensed heteroaromatic systems (5,8–10), we have synthesized some novel polycyclic derivatives of azacyclopenta[b]fluorene-1,9-dione and then evaluated their cytotoxic properties on four different human cancer cell lines.
Materials and Methods
The synthesis of the tetrahydro-4-aza-cyclopenta[b]fluorene-1,9-dione or dihydro-4-aza-cyclopenta[b]fluorene-1,9-dione derivatives was achieved following the steps outlined in Scheme 1. Reaction of ammonium acetate 1 with 1,3-cyclopentanedione 2 afforded 3-imino cyclopentanone 3 (67% yield). The tetrahydro analogues 6a–o were synthesized by molecular condensation of equivalent amount of 3-imino cyclopentanone 3, corresponding aldehyde 4a–o and 1,3-indandion 5. Then, the dihydro-4-aza-cyclopenta[b]fluorene-1,9-dione 7a–o derivatives were achieved by oxidizing the corresponding tetrhydro form using MnO2. The yield and melting point of final compounds are summarized in Table 1.
Table 1. Physical characterization and cytotoxic activity data of synthetic compounds assessed by the MTT reduction assay
Values represent the mean ± SD of at least three different experiments. Compounds were tested at the maximum final concentration of 100 μm.
RPMI 1640, fetal bovine serum (FBS), trypsin, and phosphate-buffered saline (PBS) were purchased from Biosera (Ringmer, UK). 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was obtained from Sigma (Saint Louis, MO, USA) and penicillin/streptomycin was purchased from Invitrogen (San Diego, CA, USA). Doxorubicin and dimethyl sulphoxide were obtained from EBEWE Pharma (Unterach, Austria) and Merck (Darmstadt, Germany), respectively.
Cell lines and cell culture
HeLa (human cervical adenocarcinoma), LS180 (human colon adenocarcinoma), MCF-7 (human breast adenocarcinoma), and Raji (human B lymphoma) cells were obtained from the National Cell Bank of Iran, Pasteur Institute, Tehran, Iran. All cell lines were maintained in RPMI 1640 supplemented with 10% FBS, 100 units/mL penicillin-G, and 100 μg/mL streptomycin. Cells were grown in monolayer cultures, except for Raji cells, which were grown in suspension, at 37 °C in humidified air containing 5% CO2.
Cell viability following exposure to synthetic compounds was estimated using the MTT reduction assay (11–12). MCF-7 and Raji cells were plated in 96-well microplates at a density of 5 × 104 cells/mL (100 μL/well). LS180 and HeLa cells were plated at densities of 1 × 105 and 2.5 × 104 cells/mL, respectively. Control wells contained no drugs and blank wells contained only growth medium for background correction. After overnight incubation at 37 °C, half of the growth medium was removed and 50 μL of medium supplemented with different concentrations of synthetic compounds dissolved in DMSO were added in triplicate. Plates with Raji cells were centrifuged before this procedure. Maximum concentration of DMSO in the wells was 0.5%. Cells were further incubated for 72 h, except for HeLa cells, which were incubated for 96 h. At the end of the incubation time, the medium was removed and MTT was added to each well at a final concentration of 0.5 mg/mL, and plates were incubated for another 4 h at 37 °C. Then, formazan crystals were solubilized in 200 μL DMSO. The optical density was measured at 570 nm with background correction at 655 nm using a Bio-Rad microplate reader (Model 680; Bio-Rad, Hercules, CA, USA). The percentage of inhibition of viability compared to control wells was calculated for each concentration of the compound, and IC16 and IC50 values (13) were calculated with the software CurveExpert version 1.34 for Windows (Hyams Development, OH, USA). Each experiment was repeated four times. Data are presented as mean ± SD.
The chemical structure of each molecule was drawn with Hyperchem software (Version 7; Hypercube Inc., http://www.hyper.com, USA). The Gaussian98 programa was employed to optimize the molecular structure. The structures were optimized by the ab initio method at the RHF/3-21G level. Some quantum chemical descriptors including the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) energies, the difference between HOMO and LUMO (HLG), and the molecular dipole moment (MDP) were calculated in Gaussian98. A pool of different descriptors including physicochemical properties, constitutional descriptors, geometrical descriptors, and topological indices were calculated using dragon (http://www.talete.mi.it/dragon_exp.htm) software. The multiple linear regression (MLR) was performed by the SPSS software (SPSS Inc., Chicago, IL, USA) using the stepwise selection and elimination method for variable selection. The cross-validation process was performed by home-made subroutines in matlab environment (Natick, MA, USA).
Results and Discussion
In this project, 30 dihydro and tetrahydro analogous of aza-cyclopenta[b]fluorene-1,9-dione was synthesized (Figure 1). The structures presented for 6a–o and 7a–o have been confirmed by spectroscopic data using NMR and MS instruments.
The cytotoxic activities of synthesized compounds were evaluated in four human cancer cell lines, and IC16 and IC50 values were calculated for each derivative (Table 1). On the basis of IC16 and IC50 values, the most and least potent compounds in each cell line were identified. In HeLa cell line, the most potent compounds in decreasing order of efficiency were 6l > 6g > 7g > 6i, with IC50 values of 3.1, 3.6, 5.7, and 8.0 μm, respectively. The weakest compounds were 6f, 7f, and 7l which had IC16 and IC50 values higher than 100 μm. In LS180 cell line, the most potent compounds were 6l > 6e > 6i > 7g > 6g, which had IC50 values of 6.9, 8.0, 11.5, 13.0, and 14.3 μm, respectively. Compounds 6f, 7f, 7b, 6j, 7j, 6h, and 7h with IC16 and IC50 values of higher than 100 μm were the weakest compounds. In MCF-7 cells, the order of potency was as follows; 7g > 6g > 6e > 6l (IC50 values: 10.2, 20.2, 26.5, and 27.1, respectively). Compound 7f, whose IC16 and IC50 values were both greater than 100μm, was the weakest compound followed by 6f, 7m, and 7e, which had IC16 values of 28.7, 25.5, and 24.4 μm, respectively. In Raji cells, the most potent compounds were 6g > 7g > 6l > 6i, with IC50 values of 4.4, 9.7, 13.5, and 22.4 μm, respectively. Compounds 7a, 7c, 7i, 6j, 7l, 7m, 6o, and 7o had IC16 and IC50 values of higher than 100 μm.
According to the above data, it can be concluded that the most potent compounds in all cell lines were 6g, 7g, 6l, and 6i. Although the potency of these compounds is much lower than doxorubicin, they are comparable and in some instances more potent than cisplatin. On the other hand, the weakest compounds were evidently 6f and 7f, whose their IC50 values were greater than 100 μm in all cell lines and had only one IC16 value of lower than 100 μm (7f, in Raji cell line) and two IC16 values lower than 100 μm (6f, in MCF-7 and Raji cell lines).
In general, it seems that the derivatives of aza-cyclopenta[b]fluorene-1,9-dione are less potent on Raji and LS180 cell lines and more potent on the HeLa cell line.
An interesting finding was obtained by the analysis of the correlation between cytotoxicity in various cell lines demonstrating significant associations of cytotoxic properties of these compounds in various cell lines (Table 2). In fact, in all cases, the correlation between two cell lines is more than 60% which means that IC50 values of compounds in different cell lines have more than 60% similarity in their trends. This may indicate that these compounds probably have similar cytotoxic mechanisms on different cancer cell lines that we used in this study.
Table 2. Correlation coefficients (R2) between IC50 values of synthesized compounds obtained by the MTT assay in various cell lines
Some important points were observed in respect for the structure–activity relationship of these compounds. The dihydro and tetrahydro forms showed similar cytotoxic activity, which indicates that the oxidation of pyridine ring does not seem to significantly alter the cytotoxic potency of these derivatives. Furthermore, compounds 6g, 7g, and 6l are among the most potent derivatives and all of them bear a nitro group on C10 aryl position. This indicates the important role of this moiety in the cytotoxicity.
To evaluate the effects of the structural parameters of the investigated derivatives on their cytotoxic activities, quantitative structure–activity relationship (QSAR) analysis with different types of molecular descriptors was performed. The activity data were converted to logarithmic scale (i.e., LogIC16 for cytotoxicity in different cell lines). Because IC50 values could not be calculated for some of the compounds attributable to their low activity, we used IC16 values for QSAR analysis to have a numerical value for almost all compounds.
For each set of descriptors, the best linear regression equations were obtained by the stepwise selection-based MLR subroutine of spss software. The correlation coefficient (R2), standard error (SE) of regression, correlation coefficient for leave-one-out cross-validation (QLOO2), and correlation coefficient for leave-ten-out cross-validation (QLTO2) were employed to judge the validity of regression equation. Because colinearity degrades the performances of the MLR-based QSAR equation; first, correlation analysis was performed to detect the colinear descriptors (R > 0.9).
In the first step, it was tried to find an appropriate model for cytotoxic activity in HeLa cell line. The obtained equation is shown by eqn 1:
In this equation, the values in the parenthesis represent the SD of the coefficients. N, R, RMSE, and F are number of components, correlation coefficient, root mean square error of regression, and Fisher’s F-ratio, respectively. As seen in the eqn 1, this equation has moderate statistical quality, which can explain and predict 59% and 52% of variances in the cytotoxic activity data in HeLa cell line, respectively. This model contains the molecular dipole moment at z co-ordinate (MDPz), a fragment-based parameter (C-027) defining number of R--CH—X group in the molecule and molar refractivity.
The next equation was obtained for cytotoxicity of LS180 cell line:
It is clear that this equation has a comparable quality with the former model explaining the 57% of variances in the cytotoxic activity data in LS180 cell line. The important point, in this equation, is the presence of two topological parameters, JGI1 as a Galvez topological charge index and J the Balaban Index, which indicates the importance of this type of descriptors. Moreover, the existence of another fragment-based parameter (C-026) identifying the number of R--CH--R group is a sign of similarity between these models.
The next equation was obtained for cytotoxicity of MCF-7 cell line:
The correlation coefficient of this equation shows its better quality compared to the former equations, although it is an average model. The interesting point is its high similarity with eqn 1 in which both models include a molecular dipole moment at z or x co-ordinate (MDPz for first equation and MDPx for latter one), a fragment-based parameter (C-027), and molar refractivity. On the other hand, these activities have a good correlation (intercorrelation between these cytotoxicity is 0.687) even not so high. These factors may imply that similarity of trend and mechanism of compounds in these two cell lines.
Finally, an equation for cytotoxicity in Raji cell line was obtained as described below:
This four-parametric QSAR model showed the best statistical quality among all four cytotoxicity data. Again, in this model, a molecular dipole moment and a fragment-based parameter (C-029) are presented indicating the heavy impact of these parameters in overall cytotoxicity of these compounds.
Thirty dihydro and tetrahydro analogous of aza-cyclopenta[b]fluorene-1,9-dione synthesized in this study showed weak-to-relatively-good cytotoxic activities in four human cancer cell lines, which in some instances were superior to cisplatin cytotoxicity. QSAR analysis found proper relationships between structural characteristics and the cytotoxic effect. Antitumoral activity in the HeLa, MCF-7, and Raji cell lines showed that molecular dipole moment and fragment-based parameters (C-027, C-027, and C-029, respectively) were important for the activity of these derivatives. The presence of dipole moment may indicate that some transient electrostatic interactions are involved in the mechanism of action. In LS180 cell line, in addition to a fragment-based parameter (C-026), two topological parameters also influenced the cytotoxic activity. Similarity of the cytotoxic activities of different compounds in various cell lines indicated the presence of a similar cytotoxic mechanism in all cell lines. These findings provide helpful information to guide the rational design and synthesis of more potent cytotoxic molecules that could be useful chemotherapeutic agents for treatment for cancer. According to the condensed and planar structure of these novel compounds, we speculate that DNA intercalation is a possible mechanism of cytotoxicity.
Frisch M.J., Trucks G.W., Schlegel H.B., Scuseria G.E., Robb M.A., Cheeseman J.R., Zakrzewski V.G., Montgomery Jr. J.A., Stratmann R.E., Burant J.C., Dapprich S., Millam J.M., Daniels A.D., Kudin K.N., Strain M.C., Farkas O., Tomasi J., Barone V., Cossi M., Cammi R., Mennucci B., Pomelli C., Adamo C., Clifford S., Ochterski J., Petersson G.A., Ayala P.Y., Cui Q., Morokuma K., Salvador P., Dannenberg J.J., Malick D.K., Rabuck A.D., Raghavachari K., Foresman J.B., Cioslowski J., Ortiz J.V., Baboul A.G., Stefanov B.B., Liu G., Liashenko A., Piskorz P., Komaromi I., Gomperts R., Martin R.L., Fox D.J., Keith T., Al-Laham M.A., Peng C.Y., Nanayakkara A., Challacombe M., Gill P.M.W., Johnson B., Chen W., Wong M.W., Andres J.L., Gonzalez C., Head-Gordon M., Replogle E.S., Pople J.A. (2001) GAUSSIAN 98 Rev. A11, Pittsburgh, PA, USA: Gaussian Inc.
This work was supported by Research Contract No. 425/1401 from Tehran University of Medical Sciences, Tehran, Iran. Also, financial support of the Shiraz University of Medical Sciences, vice-chancellor of research is acknowledged. The cytotoxicity section of this project is a part of Pharm. D thesis of P. Peymani (Thesis Number: 390).