European Journal of Medicinal Chemistry 45 (2010) 5493e5497 Contents lists available at ScienceDirect European Journal of Medicinal Chemistry journal homepage: http://www.elsevier.com/locate/ejmech Short communication Development of 3-aryl-1-isoquinolinamines as potent antitumor agents based on CoMFA Su Hui Yang a, Hue Thi My Van a, Thanh Nguyen Le a, Daulat Bikram Khadka a, Suk Hee Cho a, Kyung-Tae Lee b, Eung-Seok Lee c, Young Bok Lee d, Chang-Ho Ahn d, Won-Jea Cho a, * a College of Pharmacy and Research Institute of Drug Development, Chonnam National University, Buk-gu, Gwangju 500-757, Korea College of Pharmacy, Kyung-Hee University, Seoul 130-701, Korea College of Pharmacy, Yeungnam University, Kyongsan 712-749, Korea d Rexahn Pharmaceuticals, Inc., Rockville, MD 20850, USA b c a r t i c l e i n f o a b s t r a c t Article history: Received 10 June 2010 Received in revised form August 2010 Accepted 17 August 2010 Available online 24 August 2010 Various substituted 3-aryl-1-isoquinolinamines were designed and synthesized based on the previously constructed CoMFA model Most of the synthesized compounds showed excellent potency in eight different human tumor cell lines as expected In order to find the exact cytotoxic mechanism of these 3aryl-1-isoquinolinamines, we analyzed the cell cycle dynamics by flow cytometry and found that 3-aryl1-isoquinolinamine 6k-treated HeLa cells were arrested in G2/M phase, which is related to apoptosis Ó 2010 Elsevier Masson SAS All rights reserved Keywords: CoMFA Synthesis Antitumor agents Cell cycle arrest Apoptosis Introduction The major purpose of a QSAR study is the prediction of the activity of new compounds on the basis of previously synthesized compounds Among the modern 3D QSAR techniques, CoMFA has generally been used in drug design because it affords fast generation of a QSAR equation with a contour map for analysis of newly synthesized molecules [1] Moreover, CoMFA models have also been used for virtual screening to find lead compounds, although it has limitations regarding reasonable activity prediction of novel compounds that contain different 3D geometries [2] In our continuous effort to develop potent antitumor agents, we investigated the synthesis, biological evaluation and QSAR study of 3-arylisoquinoline compounds such as indeno[1,2-c]isoquinolines [3], isoindolo[2,1-b]isoquinolinones [4], 12-oxobenzo[c]phenanthridinones and benz[b]oxepines [5] as the rigid forms of the 3-aryl rings via 3-arylisoquinolinone as a key intermediate (Fig 1) However, we did not develop any compounds that surpassed the activity of the original hit compound 5, which was found during the natural alkaloid synthesis [6] Our research then focused * Corresponding author Tel.: ỵ82 62 530 2933; fax: þ82 62 530 2911 E-mail address: wjcho@jnu.ac.kr (W.-J Cho) 0223-5234/$ e see front matter Ó 2010 Elsevier Masson SAS All rights reserved doi:10.1016/j.ejmech.2010.08.042 on developing potent water-soluble 3-arylisoquinolines by applying a previously constructed CoMFA model; our former CoMFA model provided significantly high cross-validated r2 (q2) value of 0.715 as well as a high conventional r2 value of 0.927 [7] To date, most of the CoMFA studies on diverse molecules have provided reasonable statistical values that could explain the biological activities, but application of CoMFA for finding novel bioactive compound has been rare [2] In this work, we synthesized various 3-aryl-1-isoquinolinamines based on the validation of their antitumor activities through the CoMFA model Moreover, cell cycle dynamics analyses with apoptotic induction studies were performed because the mode of action of 3-aryl-1-isoquinolinamines remained unclear; some molecules were suggested to be partial inhibitors of topoisomerase I The CoMFA QSAR model was constructed from 58 diverse substituted 3-arylisoquinolines with varied cytotoxicities as shown in Table [7] The steric and electrostatic contributions to the equation were 40.1 and 59.9%, respectively, and the reasonable F value (96.804) with low standard deviation value (0.251) suggested its potential for application Among the synthesized compounds, 3-aryl-1-isoquinolinamine 6a exhibited the most potent cytotoxicities against human tumor 5494 S.H Yang et al / European Journal of Medicinal Chemistry 45 (2010) 5493e5497 O R2 R2 R1 N R O 12-Oxobenzo[c]phenanthridinone (3) R2 N O R2 N O R O Isoindolo[2,1-b]isoquinolines (2) Benz[b]oxepines(4) MOMO O R2 R1 R2 R2 NH R1 O R1 R3 3-Arylisoquinolinone NMe N 3-Arylisoquinoline (6) MeO O NMe MeO O O Indeno[1,2-c]isoquinolines (1) Me R1 OMOM NEt2 R2 CN O Fig Structure of constrained structures of 3-arylisoquinolines; indeno[1,2-b]isoquinoline 1, isoindolo[2,1-b]isoquinoline 2, 12-oxobenzo[c]phenanthridine and benz[b]oxepine Structural modification of 3-arylisoquinolinone to 3-arylisoquinolines cell lines Therefore, we designed synthetically possible derivatives, and their cytotoxic activities were virtually calculated using our previously constructed CoMFA model as shown in Table As a result, most of the designed compounds exhibited sub-micromolar activities except dimethylamino-substituted compounds 6nep on the C-5 (R1) position position to afford the desired 3-aryl-1-isoquinolinamines in moderate yield as shown in Scheme The free amines were then treated with concentrated HCl in acetone to give the HCl salt forms of the amines [9] Results and discussion Chemistry 3.1 Chemistry In order to synthesize 3-aryl-1-isoquinoliamines 6, the previously reported lithiated toluamideebenzonitrile cycloaddition method was used [8] N-diethyl-o-toluamides were reacted with n-BuLi to give the anions, which were then treated with benzonitrile to provide the 3-arylisoquinolines in moderate yield Treatment of with POCl3 afforded the corresponding imine chloride 10, which was then treated with p-methoxybenzylamine in the presence of K2CO3 in DMF to give the PMB-protected isoquinolinamine 11 in good yield In our previous study, benzylamine at the C-1 position was introduced and catalytic hydrogenation reaction in the presence of PdeC was used to obtain the amine in moderate yield However, the above reaction condition could not be applied to several substituted analogs Finally, we found that trifluoroacetic acid was useful to remove the PMB group on the C-1 3.1.1 General procedure for synthesis of 6aem A solution of p-methoxybenzylamine compound 11 in CH2Cl2 in the presence of trifluoroacetic acid was stirred at room temperature for overnight To this solution sat aq NaHCO3 was added and reaction mixture was extracted with CH2Cl2 The combined organic extracts were washed with water, dried, and concentrated to give the residue which was then purified by column chromatography on silica gel to afford 3-aryl-1-isoquinolinamines (n-hexane: ethyl acetate ¼ 1:1 was used) To a solution of 3-aryl-1-isoquinolinamine in acetone was added concentrated HCl to afford the precipitate which was collected and washed with acetone to give the desired hydrochloride salt Table CoMFA correlation statistics for 3-arylisoquinolines Type ma Nb Sc F r2 Cross-Validd SEP Conformer Defined Region a b c d e f g q RCe Stericf Elecg 58 0.251 96.804 0.924 0.485 0.715 40.1 Number of components Number of compounds Standard deviation Obtained from the leave-one-out cross-validation Relative contribution Steric field descriptors Electrostatic field descriptors 59.9 3.1.2 6-Methyl-3-o-tolylisoquinolin-1-amine hydrochloride salt (6a) H NMR (DMSO-d6) d: 13.84 (s, 1H), 9.40 (s, 1H), 8.79e7.57 (m, 7H), 7.38 (s, 1H), 2.75 (s, 3H), 2.61 (s, 3H) mp: 253e258 C (dec.) LCMS: m/z (%) 248 (Mỵ, 100) 3.1.3 3-(3,4-Dimethoxyphenyl)-6-methylisoquinolin-1-amine (6b) H NMR (300 MHz, CDCl3) d: 7.67e7.64 (m, 2H), 7.54 (m, 1H), 7.45 (s, 1H), 7.27 (s, 1H), 7.18 (m, 1H), 6.91 (d, J ¼ 8.5 Hz, 1H), 5.76 (s, 2H), 3.96 (s, 3H), 3.89 (s, 3H), 2.50 (s, 3H) mp: 270e273 C (dec.) LCMS: m/z (%) 294 (Mỵ, 100) 3.1.4 3-(Benzo[d][1,3]dioxol-6-yl)-6-methylisoquinolin-1-amine (6c) H NMR (300 MHz, CDCl3) d: 7.63 (d, J ¼ 8.5 Hz, 1H), 7.54 (m, 2H), 7.45 (s, 1H), 7.22 (m, 2H), 6.88 (d, J ¼ 8.5 Hz, 1H), 5.98 5495 S.H Yang et al / European Journal of Medicinal Chemistry 45 (2010) 5493e5497 Table Observed and calculated cytotoxicity (-log IC50 value) values against A-549 of the 3-aryl-1-isoquinolinamines based on the CoMFA model No Compound R1 R2 R3 R4 Yield(%) Obsda Calcdb Diffc 10 11 12 13 14 15 16 6a 6b 6c 6d 6e 6f 6g 6h 6i 6j 6k 6l 6m 6n 6o 6p H H H H H H H H H H H H H NMe2 NMe2 NMe2 eCH3 eCH3 eCH3 eCH3 eCH3 eCH3 eCH3 H H eOCH3 eOCH3 eOCH3 eOCH3 H H H H H H H H H H eCH3 eCH3 eOCH3 eOCH3 eOCH3 eOCH3 H H H 2-methyl 3,4-dimethoxy 3,4-methylenedioxide 4-methoxy-2-methyl 2,6-dimethyl 3-methoxy 3-methyl 3-methoxy 2-methyl H 3-methoxy 3,5-dimethoxy 2-methyl H 2-methyl 3,4-dimethoxy 67 79 71 67 68 89 75 92 91 71 60 78 89 73 60 40 6.79 6.25 6.85 6.77 6.42 6.97 6.74 6.35 7.27 6.89 6.12 6.55 6.36 10 0.24 1.75 0.22 0.24 0.21 1.28 0.20 0.47 0.20 0.67 0.19 0.70 0.83 >10 >10 >10 0.19 0.85 0.16 0.17 0.17 0.59 0.18 0.21 0.12 0.52 0.16 0.77 0.78 >10 >10 >10 0.24 2.69 0.19 0.25 0.24 1.19 0.22 0.42 0.26 0.58 0.18 0.72 1.25 >10 >10 >10 0.15 0.57 0.16 0.17 0.16 0.47 0.15 0.24 0.10 0.27 0.074 0.19 0.38 >10 >10 >10 0.35 1.41 0.22 0.29 0.23 0.72 0.25 0.41 0.18 0.75 0.23 0.69 0.91 >10 >10 >10 0.12 0.51 0.11 0.12 0.12 0.69 0.16 0.14 0.14 0.59 0.17 0.54 0.67 >10 >10 >10 0.26 2.15 0.23 0.28 0.24 0.75 0.21 0.48 0.16 0.71 0.22 0.40 0.56 >10 >10 >10 This compound was reported and used as the reference molecule for cytotoxicity comparison with other compounds (s, 3H), 3.86 (s, 6H) mp: 246e254 C (dec.) LCMS: m/z (%) 340 (Mỵ, 100) 3.1.14 6,7-Dimethoxy-3-o-tolylisoquinolin-1-ylamine (6m) H NMR (300 MHz, CDCl3) d: 7.45e7.03 (m, 7H), 5.01 (s, 2H), 4.02 (s, 3H), 4.02 (s, 3H), 2.40 (s, 3H) mp: 259e268 C (dec.) LCMS: m/z (%) 294 (Mỵ, 100) 3.1.15 N5,N5-Dimethyl-3-phenylisoquinolin-1,5-diamine (6n) H NMR (300 MHz, CDCl3) d: 8.08 (m, 2H), 7.82 (s, 1H), 7.48e7.33 (m, 5H), 7.18 (m, 2H), 5.24 (s, 2H), 2.89 (s, 6H) mp: 235e250 C (dec.) LCMS: m/z (%) 263 (Mỵ, 95) 3.1.16 N5,N5-Dimethyl-3-o-tolylisoquinolin-1,5-diamine (6o) H NMR (300 MHz, DMSO-d6) d: 13.87 (s, 1H), 8.42 (d, 1H), 7.82e7.71 (m, 2H), 7.50e7.33 (m, 5H), 2.91 (s, 6H), 2.42 (s, 3H) mp: 212e219 C (dec.) LCMS: m/z (%) 277 (Mỵ, 100) 3.1.17 3-(3,4-Dimethoxyphenyl)-N5,N5-dimethylisoquinolin-1,5diamine (6p) H NMR (300 MHz, DMSO-d6) d: 9.69 (s, 2H), 8.40 (d, 1H), 7.93e7.91 (m, 2H), 7.75e7.67 (m, 3H), 7.13 (d, 1H), 3.94 (s, 3H), 3.82 (s, 3H), 3.03 (s, 6H) 242e248 C (dec.) LCMS: m/z (%) 323 (Mỵ, 100) 3.2 Biological evaluation The in vitro cytotoxicity experiments were carried out with the synthesized 3-aryl-1-isoquinolinamines against several tumor cell lines including Human MDA-MB-231 (breast), PC3 (prostate), HCT116 (colon), OVCAR-3 (ovary), Caki-1 (kidney), PANC-1 (pancreas), SNB-19 (glioblastoma) and SK-MEL-28 (melanoma) cells from the American Type Culture Collection (ATCC) (Manassas, VA) A-549 (lung) cells for CoMFA study and HeLa (cervical) cell lines for flow cytometry were used The growth inhibition assays of the 3-aryl-1-isoquinolinamine analogs against eight different human tumor cell lines were carried out using the Sulforhodamine B method [10] As expected, most of the 3-aryl-1-isoquinolinamine derivatives exhibited potent cytotoxicities, as shown in Table The IC50 values of the compounds were less than mM or even 0.1 mM against seven tumor cell lines, suggesting that the potency of these compounds was not greatly dependent on the substitution pattern of the modified phenyl ring Among these compounds, 6i showed equal or better growth inhibitory activity against MDA-MB-231 breast cancer cells compared to 6a When we prepared 5-dimethylaminosubstituted 3-aryl-1-isoquinolinamines 6nep, no compounds exceeded the activity of 6a against several tumor cell lines and this Fig Effects of compound 6k on cell cycle progression in HeLa cells Cells were treated with or without compound 6k (50e500 nM for 24 h), washed and then collected The cells were stained with propidium iodide and then DNA contents were analyzed by flow cytometry S.H Yang et al / European Journal of Medicinal Chemistry 45 (2010) 5493e5497 5497 Fig Effects of compound 6k on apoptotic induction in HeLa cells Cells were treated with or without compound 6k (50e500 nM for 24 h) and then co-stained with PI and FITCconjugated annexin V The externalization of phosphatidylserine (PS) was detected by flow cytometry after 6k treatment result was a good match with the CoMFA model prediction as shown in Table The average activity difference between the observed values with calculated ones was only 0.26 and it means our constructed CoMFA model is quite predictive To assess whether compound 6k-induced cytotoxicity of cells was mediated via alterations in the cell cycle [11], the effects of 6k on cell cycle progression were evaluated As shown in Fig 2, compound 6k-induced a significant dose-dependent increase in the cell population in the G2/M phase of the cell cycle The cell cycle progression of G2/M phase was 35.80, 48.59, 73.06, 83.14 and 86.98% with 50, 100, 200, 300 and 500 nM compound 6k, respectively This increase in the G2/M phase cell population was accompanied by a concomitant decrease in the G1-phase cell population Apoptosis is an essential physiological process for the normal development and maintenance of tissue homeostasis Most anticancer agents eradicate cancer cells by inducing apoptosis To see whether the cytotoxic effect of compound 6k derived from the induction of apoptosis, HeLa cells were treated with 6k (50e500 nM for 24 h), and then the externalization of phosphatidylserine (PS) was analyzed using Annexin V and PI double staining [12] As shown in Fig 3, cell numbers increased up to 70.12% in the lower-right quadrant, representing the increased early apoptotic cells (Annexin V-positive) The other 12.52% of cells belonged to the late apoptotic stage This finding indicates that compound 6kinduced cell death in HeLa cells is apoptotic rather than necrotic cell death Conclusion In conclusion, we synthesized various substituted 3-aryl-1-isoquinolinamines based on the previously constructed CoMFA model Most of the synthesized compounds showed excellent potency in eight different human tumor cell lines In particular, dimethylsubstituted 3-aryl-1-isoquinolinamine 6i displayed stronger potency than the lead compound 6a Interestingly, 6i was also predicted to be most active by the CoMFA model The potency of these compounds does not seem to be significantly dependent on the substitution pattern of the 3-arylisoquinoline aromatic rings We then undertook more intensive studies of these compounds to determine their mode of action Although some 3-arylisoquinoline derivatives showed moderate topoisomerase I inhibitory activities, no confirmed cytotoxic mechanism was clarified In order to find the exact cytotoxic mechanism of these 3-aryl-1-isoquinolinamines, we analyzed the cell cycle dynamics by flow cytometry and found that 3-aryl-1-isoquinolinamine 6k-treated HeLa cells were arrested in G2/M phase, which is related to apoptosis Finding water-soluble potent cytotoxic compounds will provide valuable information to researchers developing antitumor chemotherapeutic agents Acknowledgement This study was financially supported by Special Research Program of Chonnam National University, 2009 References [1] Q.Y Zhang, J Wan, X Xu, G.F Yang, Y.L Ren, J.J Liu, H Wang, Y Guo, J Comb Chem (2007) 131 [2] R.D Cramer, P Cruz, G Stahl, W.C Curtiss, B Camphell, B.B Masek, F Soltanshahi, J Chem Inf Model 48 (2008) 2180 [3] a) W.J Cho, Q.M Le, H.T.M Van, K.Y Lee, B.Y Kang, E.S Lee, S.K Lee, Y Kwon, Bioorg Med Chem Lett 17 (2007) 3531; b) H.T.M Van, Q.M Le, K.Y Lee, E.S Lee, Y Kwon, T.S Kim, T.N Le, S.H Lee, W.J Cho, Bioorg Med Chem Lett 17 (2007) 5763 [4] H.T.M Van, W.J Cho, Bioorg Med Chem Lett 19 (2009) 2551 [5] S.H Lee, H.T.M Van, S.H Yang, K.T Lee, Y Kwon, W.J Cho, Bioorg Med Chem Lett 19 (2009) 2444 [6] W.J Cho, S.J Yoo, M.J Park, B.H Chung, C.O Lee, Arch Pharm Res 20 (1997) 264 [7] W.J Cho, E.K Kim, I.Y Park, E.Y Jeong, T.S Kim, T.N Le, D.D Kim, E.S Lee, Bioorg Med Chem 10 (2002) 2953 [8] H.T.M Van, S.H Yang, D.B Khadka, Y.C Kim, W.J Cho, Tetrahedron 65 (2009) 10142 [9] S.H Yang, H.T.M Van, T.N Le, D.B Khadka, S.H Cho, K.-T Lee, H.-J Chung, S.K Lee, C.-H Ahn, Y.B Lee, W.-J Cho, Bioorg Med Chem Lett 20 (2010) 5277 [10] L.V Rubinstein, R.H Shoemaker, K.D Paull, R.M Simon, S Tosini, P Skehan, D.A Scudiero, A Monks, M.R Boyd, J Natl Cancer Inst 82 (1990) 1113 [11] a) A Skladanowski, P Bozko, M Sabisz, Chem Rev 109 (2009) 2951; b) S Maddika, S.R Ande, E Wiechec, L.L Hansen, S Wesselborg, M Los, J Cell Sci 121 (2008) 979 [12] a) A Moore, C.J Donahue, K.D Bauer, J.P Mather, Methods Cell Biol 57 (1998) 265; b) C.G Ferreira, M Epping, F.A Kruyt, G Giaccone, Clin Cancer Res (2002) 2024 ... than necrotic cell death Conclusion In conclusion, we synthesized various substituted 3-aryl-1-isoquinolinamines based on the previously constructed CoMFA model Most of the synthesized compounds... with calculated ones was only 0.26 and it means our constructed CoMFA model is quite predictive To assess whether compound 6k-induced cytotoxicity of cells was mediated via alterations in the cell... progression of G2/M phase was 35.80, 48.59, 73.06, 83.14 and 86.98% with 50, 100, 200, 300 and 500 nM compound 6k, respectively This increase in the G2/M phase cell population was accompanied by a concomitant