Z. Naturforsch. 2016; 71(11-12)c: 423–427 Dorimar Stiz, Adriana Campos, Ana Lúcia Tasca Gois Ruiz, João Ernesto de Carvalho, Rogério Corrêa and Valdir Cechinel-Filho* Antiproliferative effect of synthetic cyclic imides (methylphtalimides, carboxylic acid phtalimides and itaconimides) against human cancer cell lines DOI 10.1515/znc-2016-0067 Received April 1, 2016; revised September 6, 2016; accepted September 20, 2016 Abstract: This work describes the antiproliferative potential of 14 cyclic imides (methylphtalimides, carboxylic acid phtalimides and itaconimides) against several human cancer cell lines. The antiproliferative effect was evaluated using the sulforhodamine B assay. Although some compounds from methylphtalimide and carboxylic acid phtalimide classes exhibited a selective antiproliferative activity, the itaconimides (11–14) exhibited the best results, especially compound 14, which presented a TGI (concentration that produces total growth inhibition) value of 0.0043 µM against glioma (U251), being inactive against the non-tumor cell line (HaCat). Absorption, distribution, metabolism and excretion in silico evaluations suggest that these compounds are promising candidates. Keywords: ADME; antiproliferative effect; cyclic imides; itaconimides. 1 Introduction Cancer represents a serious and worrying global public health problem, with expressively high mortality rates. According to the World Health Organization, it is expected to rise to 22 million per year over the next two decades. *Corresponding author: Valdir Cechinel-Filho, Programa de PósGraduação em Ciências Farmacêuticas and Núcleo de Investigações Químico-Farmacêuticas (NIQFAR), Universidade do Vale do Itajaí – UNIVALI, Caixa postal 360, CEP 88302-202, Itajaí, Santa Catarina, Brazil, Tel.: +55-473-341-7557, E-mail: cechinel@univali.br Dorimar Stiz, Adriana Campos and Rogério Corrêa: Programa de Pós-Graduação em Ciências Farmacêuticas and Núcleo de Investigações Químico-Farmacêuticas (NIQFAR), Universidade do Vale do Itajaí – UNIVALI, Itajaí, Santa Catarina, Brazil Ana Lúcia Tasca Gois Ruiz and João Ernesto de Carvalho: Centro Pluridisciplinar de Pesquisas Químicas, Biológicas e Agrícolas (CPQBA) – Universidade Estadual de Campinas (UNICAMP), Campinas, São Paulo, Brazil The scientific community has made great efforts to discover new and effective molecules from nature and synthetic routes with anticancer potential. In relation to cancer, nature has inspired the design of a great variety of molecules (derivatives and analogs), several of which are available in the clinic to treat different kinds of cancer [1, 2]. Among the various possibilities, cyclic imides are good examples of promising molecules with reported anticancer activities [3–5]. Our research group has studied some sub-classes of cyclic imides considering the discovery of phyllanthimide, an alkaloid isolated from Phyllanthus sellowianus [6]. Several cyclic imides, such as N-phenylmaleimides and glutarimides, have demonstrated pronounced anticancer properties in different experimental models [7–9]. The present study describes the evaluation of three sub-families of cyclic imides – methylphtalimides (A), carboxyl acid phtalimides (B) and itaconimides (C) – as antiproliferative agents against several human cancer cell lines in vitro. Absorption, distribution, metabolism and excretion (ADME) in silico evaluations were also performed to predict variables such as absorption capacity, distribution, metabolism and excretion for the studied molecules. Variables such as Lipinki’s Rule of Five, the Bioavailability Score, the Egan Violations Count and the Veber Violations Count were used. All the analyzed substances were previously described in the literature [10]. 2 Materials and methods 2.1 Chemistry The synthetic cyclic imides were obtained by the reaction of the respective anhydrides with appropriate amines in ether, or directly refluxed with acetic acid as previously described [10]. Fourteen compounds were obtained and were divided into three series as indicated in Figures 1–3. 424 Stiz et al.: Antiproliferative effect of synthetic cyclic imides 2.2.3 Antiproliferative assay X = 4-CH3 (1); 4-OCH3 (2); H (3); 3,4-Cl2 (4); 4-Cl (5). Figure 1: Molecular structure of cyclic imides derived from 4-methylphthalic anhydride (methylphtalimides). X = 4-CH3 (6); 4-OCH3 (7); -H (8); 4-Cl (9); 3,4-Cl2 (10). Figure 2: Molecular structure of cyclic imides derived from 4-carboxylphthalic (carboxyl acid phtalimides). Cells in 96-well plates (100 µL cells/well) were exposed to various sample concentrations (approx. 0.001–1 µM) in DMSO/RPMI1640 at 37 °C, 5% of CO2 in air for 48 h. Doxorubicin (DOXO) was used as standard (46 × 10−6, 46 × 10−5, 46 × 10−4 and 46 × 10−3 µM). The final DMSO concentration did not affect cell viability (0.25%). Cells, before (T0) and after 48 h of exposure (T1) were then fixed with 50% trichloroacetic acid, and cell proliferation was determined by spectrophotometric quantification of the cellular protein content at 540 nm, using the sulforhodamine B assay [11]. The TGI (concentration that produces total growth inhibition) was determined through nonlinear regression analysis using the concentration-response curve for each cell line in the software ORIGIN 8.0® (Origin-Lab Corporation) [12]. 3 Results X = 4-CH3 (11); 4-OCH3 (12); H (13); 4-Cl (14). Figure 3: Molecular structure of cyclic imides derived from itaconic anhydride (itaconimides). 2.2 In vitro antiproliferative assay 2.2.1 Cell lines Human tumor cell lines U251 (glioma), MCF7 (breast), NCI/ADR-RES (ovary expressing multi-drug resistance phenotype), 786-0 (kidney), NCI-H460 (lung, nonsmall cells), HT-29 (colon), PC-3 (prostate), OVCAR-3 (ovary) and K562 (leukemia) were kindly provided by the National Cancer Institute (Frederick, MA, USA). The non-tumor cell line HaCat (human keratinocytes) was donated by Professor Dr. Ricardo Della Coletta, FOP/ UNICAMP. 2.2.2 Cell culture Stock cultures were grown in medium RPMI 1640 (GIBCO BRL) supplemented with 5% fetal bovine serum (GIBCO) and 10 U/mL penicillin, 10 µg/mL streptomycin at 37 °C in 5% CO2. Fourteen previously synthesized cyclic imides (methylphtalimides, carboxylic acid phtalimides and itaconimides) [10] were evaluated for the first time to evidence their possible anticancer potential by the analysis of the antiproliferative effects. Table 1 shows the antiproliferative effect of methylphtalimides (1–5) in different cancer cell lines. As can be observed, compound 5 showed an antiproliferative effect against prostate (PC-3), ovary (OVCAR-3) and leukemia (K562) cells, with TGI values 0.202, 0.375 and 0.198 µM, respectively. Compounds 1 and 2 showed a pronounced and selective activity only against the ovary cancer cell (OVCAR-3), with TGI values of 0.0276 and 0.0112 µM, respectively. The most active compounds are those containing the substituent –CH3 (1) and –OCH3 (2) groups with electron donating σ+, moving the electron density of the molecule to the imide portion. Additional studies should be conducted to enhance the antiproliferative activity and determine their possible mechanisms of action. The results described in Table 2 demonstrate the antiproliferative effect of phtalimides. All the compounds presented an antiproliferative activity against the ovary (OVCAR-3) cancer cell line, with TGI values of between 0.116 and 0.0036 µM. Phtalimide 10 showed the best antiproliferative effect against this cell line and almost all the other cell lines tested. Its molecular structure presents 425 Stiz et al.: Antiproliferative effect of synthetic cyclic imides Table 1: Antiproliferative activity of doxorubicin (DOXO – positive control) and cyclic imides derived from 4-methylphthalic anhydride against human cancer cell lines.a TGI (µM).b DOXO 1 2 3 4 5 2 m a 7 4 p o h k Q 46 × 10 >1 > 0.939 > 1.059 > 0.819 > 0.924 < 46 × 10 >1 > 0.939 > 1.059 > 0.819 > 0.924 23 × 10 >1 > 0.939 > 1.059 > 0.819 > 0.924 51 × 10 >1 > 0.939 > 1.059 > 0.819 > 0.924 75 × 10 >1 > 0.939 > 1.059 > 0.819 > 0.924 < 46 × 10 >1 > 0.939 > 1.059 > 0.819 0.202 16 × 10 0.0276 0.0112 > 1.059 > 0.819 0.375 42 × 10 >1 > 0.939 > 1.059 > 0.819 > 0.924 18 × 10 >1 > 0.939 > 1.059 > 0.819 0.198 53 × 10−6 >1 > 0.939 > 1.059 > 0.819 > 0.924 −6 −6 −5 −6 −6 −6 −5 −5 −5 a Human tumor cell lines: 2 = U251 (glioma); m = MCF7 (breast); a = NCI/ADR-RES (ovary expressing multi-drug resistance phenotype); 7 = 786-0 (kidney); 4 = NCI-H460 (lung, non-small cells); p = PC-3 (prostate); o = OVCAR-3 (ovary); h = HT-29 (colon); k = K562 (leukemia). Nontumor cell line: Q = HaCat (keratinocyte). Assessed by the SRB assay. b TGI values represent the necessary concentration (µM) for total inhibition of cancer cell proliferation. Values were determined through nonlinear regression analysis using the ORIGIN 8.0® (OriginLab Corporation). Table 2: Antiproliferative activity of doxorubicin (DOXO-positive control) and cyclic imides derived from 4-carboxylphthalic against human cancer cell linesa. TGI (µM)b. DOXO 6 7 8 9 10 2 m a 7 4 p o h k Q 46 × 10−6 > 0.889 > 0.841 > 0.936 > 0.829 0.347 < 46 × 10−6 > 0.889 > 0.841 > 0.936 > 0.829 0.306 23 × 10−5 > 0.889 0.730 > 0.936 > 0.829 0.570 51 × 10−6 > 0.889 > 0.841 > 0.936 > 0.829 > 0.744 75 × 10−6 > 0.889 > 0.841 > 0.936 > 0.829 0.144 < 46 × 10−6 0.532 0.812 > 0.936 0.409 0.147 16 × 10−5 0.117 0.116 0.082 0.0036 0.080 42 × 10−5 > 0.889 > 0.841 > 0.936 > 0.829 0.349 18 × 10−5 > 0.889 > 0.841 > 0.936 > 0.829 > 0.744 53 × 10−6 > 0.889 > 0.841 > 0.936 > 0.829 0.744 Human tumor cell lines: 2 = U251 (glioma); m = MCF7 (breast); a = NCI/ADR-RES (ovary expressing multi-drug resistance phenotype); 7 = 786-0 (kidney); 4 = NCI-H460 (lung, non-small cells); p = PC-3 (prostate); o = OVCAR-3 (ovary); h = HT-29 (colon); k = K562 (leukemia). Nontumor cell line: Q = HaCat (keratinocyte). Assessed by the SRB assay. b TGI values represent the necessary concentration (µM) for total inhibition of cancer cell proliferation. Values were determined through nonlinear regression analysis using the ORIGIN 8.0® (OriginLab Corporation). a two chlorine atoms in positions 3 and 4, suggesting that steric or conformational parameters are involved in the observed effect. The cyclic imides derived from itaconic anhydride (itaconimides) were the most potent and promising series evaluated (Table 3). All the synthesized compounds were active against almost all cancer cell lines, especially glioma (U251), ovary expressing multi-drug resistance phenotype (NCI/ADR-RES), kidney (786-0), lung (NCIH460), prostate (PC-3), ovary (OVCAR-3) and leukemia (K562), demonstrating TGI values that were often better than the positive control, DOXO. Table 3: Antiproliferative activity of doxorubicin (DOXO – positive control) and cyclic imides derived from itaconic anhydride against human cancer cell lines.a TGI (µM).b DOXO 11 12 13 14 2 m a 7 4 p o h k Q 77 × 10 0.0065 0.0050 0.0059 0.0043 > 45 × 10 > 1.243 0.049 0.082 0.014 > 45 × 10 > 1.243 0.015 0.019 0.014 17 × 10 0.011 0.0073 0.0075 0.0059 39 × 10 0.068 0.0202 0.028 0.010 35 × 10 0.010 0.010 0.008 0.007 68 × 10 0.048 0.015 0.030 0.0059 > 45 × 10 > 1.243 > 1.150 0.604 > 1.120 31 × 10 0.016 0.0078 0.009 0.0094 23 × 10−3 1.243 0.066 0.0084 1.120 −6 −3 −3 −4 −3 −4 −4 −3 −4 a Human tumor cell lines: 2 = U251 (glioma); m = MCF7 (breast); a = NCI/ADR-RES (ovary expressing multi-drug resistance phenotype); 7 = 786-0 (kidney); 4 = NCI-H460 (lung, non-small cells); p = PC-3 (prostate); o = OVCAR-3 (ovary); h = HT-29 (colon); k = K562 (leukemia). Nontumor cell line: Q = HaCat (keratinocyte). Assessed by the SRB assay. b TGI values represent the necessary concentration (µM) for total inhibition of cancer cell proliferation. Values were determined through nonlinear regression analysis using the ORIGIN 8.0® (OriginLab Corporation). 426 Stiz et al.: Antiproliferative effect of synthetic cyclic imides It is noteworthy that the substance 14, with a chlorine atom in position 4, similar to compound 10, presented the best antiproliferative effect, especially against glioma (U251), with a TGI value of 0.0043 µM. This compound also showed no cytotoxicity against the non-tumor cell line (HaCat). 4 Discussion Several studies have demonstrated that N-substituted cyclic imides have cytotoxic activity, which can be attributed to the intrinsic nature of the imidic ring and its electrically neutral potential and hydrophobicity, which facilitates the penetration of substances through the cell membrane. Other studies report that the cytotoxic effects may be related to the characteristics and size of the substituent groups of the imide ring, modifying the electronic and steric properties of the substances, and altering the cytotoxic activity [3, 7, 13–15]. In general, the cyclic imides are active being the imidic ring considered a pharmacophoric group. In this paper, the double bond in the imidic ring seems to enhance the antiproliferative effects in comparison with the other compounds. For the antiproliferative evaluation of these samples, we followed the methodology described by Developmental Therapeutics Program NCI/NIH (https://dtp.cancer. gov/). As described by Monks et al. [11] and DPT/NCI/NIH (https://dtp.cancer.gov/discovery_development/nci-60/ methodology.htm), in this methodology, there is one more measurement of cell population density at time zero (the time at which drugs are added); using these three measurements [optical density at time zero (T0), control (C) and test (T) optical densities after 48 h], cellular responses can be calculated by [(T − T0) / (C − T0)] × 100 for concentrations for which T ≥ T0 [(T − T0) / T0] × 100 for concentrations for which T < T0. Thus, three dose response parameters could be calculated for each sample, namely growth inhibition of 50% (GI50), “which is the sample concentration resulting in a 50% reduction in the net protein increase (as measured by SRB staining) in control cells during the drug incubation”; the sample concentration resulting in total growth inhibition (TGI), “where the amount of protein at the end of drug incubation is equal to the amount at the beginning”; and the LC50, “ which is the concentration of drug resulting in a 50% reduction in the measured protein at the end of the drug treatment as compared with that at the beginning”. As described by Sebaugh [16], the definition of IC50 (inhibitory concentration) “…contain some builtin assumptions: that there is a monotonic relationship between the dose of the compound…”. In the methodology used for the evaluation of the antiproliferative activity of our substances, two responses can be assessed, cytostatic and cytocidal effects. Thus, IC50 is not a recommended parameter to express our results. Since it is not yet known conclusively that the mechanism of cell death is induced by the compounds under study, a direct comparison is not possible with DOXO, which acts by inhibiting topoisomerases and DNA intercalation. On the other hand, the most active compounds (11–14) have the opposite behavior to a non-tumor cell (HaCaT line) in relation to the tumor lines. In general, they interfere with the least proliferation of HaCaT (at least a concentration 10× higher than the average of the most promising concentrations) suggesting the minimum possible action in normal tissue cells. Regarding the ADME approach, it was possible to evidence the molecular characteristics required for promising drug candidates. The results of the bioavailability score for all substances tested were 0.55, suggesting the suitability of the studied compounds for in vitro and in vivo tests. No violation was found for the Egan violations Table 4: Descriptors involved in theoretical ADME evaluation (compounds 11–14). Compound MW, amu HBD HBA RBC logP TPSA, Å2 ABS EVC VVC LRF 11 12 13 14 201.2212 217.2206 187.1946 221.6397 0 0 0 0 3 4 3 3 1 2 1 1 1.3811 1.0813 1.0727 1.7261 37.380 46.610 37.380 37.380 0.5500 0.5500 0.5500 0.5500 0 0 0 0 0 0 0 0 0 0 0 0 ABS, bioavailability score; EVC, Egan violations count; HBA, no of H-bond acceptors; HBD, no of H-bond donors; logP, partition coefficient octanol/water; LRF, Lipinski rule of five violations count; MW, molecular weight; RBC, no of rotatable bonds; TPSA, total polar surface area; VVC, Veber violations count. Stiz et al.: Antiproliferative effect of synthetic cyclic imides count, Veber violations count and Lipinski rule of five violations count filters, indicating that they have good conditions for passive intestinal absorption and likelihood of oral bioavailable. Table 4 shows these parameters for the most active compounds (11–14). In conclusion, these results demonstrate, for the first time, that most of the synthesized imides exhibit antiproliferative properties, especially against the ovary cancer cell line (OVCAR-3). Itaconimides represented the most promising series, particularly compound 14, which is currently being studied in other experimental models to confirm the anticancer profile and elucidate its possible mechanism of action. Finally, these compounds also demonstrated excellent drugability profiles according to the ADME. Acknowledgments: The authors are grateful to the National Council for Scientific and Technological Development (Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPq), Fundo de Apoio à Manutenção e ao Desenvolvimento da Educação Superior – FUMDES (State of Santa Catarina) and RIBECANCER (RT 0464)/ CYTED/CNPq Network, for their financial support. 5. 6. 7. 8. 9. 10. 11. 12. References 1. Newman DJ, Cragg GM. Natural products as sources of new drugs over the 30 years from 1981 to 2010. J Nat Prod 2012;75:311–35. 2. Cragg GM, Newman DJ. Natural products as sources of new anticancer agents. In: San Feliciano A, Cechinel Filho V, editors. Descoberta, Desenho e Desenvolvimento de Novos Agentes Anticâncer no Âmbito do Programa Iberoamericano CYTED. Itajaí: UNIVALI, 2014:67–114. 3. Cechinel Filho V, Buzzi FC, Corrêa R, Yunes RA, Nunes RJ. Aspectos químicos e potencial terapêutico de imidas cíclicas: uma revisão da literatura. Quim Nova 2003;26:230–41. 4. Machado KE, Oliveira KN, Santos-Bubniak L, Licínio MA, Nunes RJ, Santos-Silva MC. Evaluation of apoptotic effect of cyclic imide 13. 14. 15. 16. 427 derivatives on murine B16F10 melanoma cells. Bioorg Med Chem 2011;19:6285–91. Kumar A, Kumar N, Roy P, Sondhi SM, Sharma A. Microwaveassisted synthesis of benzenesulfonohydrazide and benzenesulfonamide cyclic imide hybrid molecules and their evaluation for anticancer activity. Med Chem Res 2015;24:3760–71. Tempesta MS, Corley DG, Beutler JA, Metral CJ, Yunes RA, Giacomozzi CA, et al. Phyllanthimide, a new alkaloid from Phyllanthus sellowianus. J Nat Prod 1988;51:617–8. Prado S, Cechinel Filho V, Campos Buzzi F, Corrêa R, Cadena S, Oliveira M. Biologics evaluation of some selected cyclic imides: mitochondrial effects and in vitro cytotoxicity. Z Naturforsch C 2004;59:663–72. Yunes JA, Cardoso AA, Yunes RA, Corrêa R, Campos-Buzzi F, Cechinel Filho V. Antiproliferative effects of a series of cyclic imides on primary endothelial cells and a leukemia cell line. Z Naturforsch C 2008;63:675–80. Noldin VF, Locatelli C, Cordova CA, Noldin AT, Vanzin F, Fae JD, et al. Cytotoxicity of N-phenylmaleimide derivatives and inhibition of melanoma growth in a preclinical mouse melanoma model. Res Rev J Pharm Pharm Sci 2015;4:1–2. Stiz D, Corrêa R, D’Auria FD, Simonetti G, Cechinel-Filho V. Synthesis of cyclic imides (methylphtalimides, carboxylic acid phtalimides and itaconimides) and evaluation of their antifungal potential. Med Chem 2016;12:647–54. Monks A, Scudiero D, Skehan P, Shoemaker R, Paull K, Vistica D, et al. Feasibility of a high-flux anticancer drug screen using a diverse panel of cultured human tumor cell lines. J Natl Cancer Inst 1991;83:757–66. Shoemaker RH. The NCI60 human tumor cell line anticancer drug screen. Nat Rev Cancer 2006;6:813–23. Loh W, Cosby LA, Sartorelli AC. Synthesis and antineoplastic activity of phenyl-substituted benzenesulfonylhydrazones of 2-pyridinecarboxyaldehyde 1-oxide. J Med Chem 1980;23:631–4. Fournel M, Trachy-Bourget MC, Yan PT, Kalita A, Bonfils C, Beaulieu C, et al. Sulfonamide anilides, a novel class of histone deacetylase inhibitors, are antiproliferative against human tumors. Cancer Res 2002;62:4325–30. Kendall JD, Rewcastle GW, Frederick R, Mawson R, Denny WA, Marshall ES, et al. Synthesis, biological evaluation and molecular modelling of sulfonohydrazides as selective PI3K p110alpha inhibitors. Bioorg Med Chem 2007;15:7677–87. Sebaugh JL. Guidelines for accurate EC50/IC50 estimation. Pharm Stat 2011;10:128–34.