YANG Qi ZHANG Shu-Ping ZHAO Shi-Peng

?

3-QSAR Studies on a Series of Indoleamide Derivatives as Antiplasmodial Drugs①

YANG QiaZHANG Shu-Pinga②ZHAO Shi-Pengb

a(200093)b(200093)

In this study, three-dimensional quantitative structure-activity relationship (3- QSAR) was studied for the antiplasmodial activity of a series of novel indoleamide derivatives by comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA). 3-QSAR model was established by a training set of 20 compounds and was externally validated by a test set of 4 compounds. The best prediction (2= 0.593 and 0.527,2= 0.990 and 0.953,2pred=0.967 and 0.962for CoMFA and CoMSIA) was obtained according to CoMFA and CoMSIA. Those parameters indicated the model was reliable and predictable. We designed several molecules with high activities according to the contour maps produced by the CoMFA and CoMSIA models.

3-QSAR, CoMFA, CoMSIA, antiplasmodial, indoleamide;

1 INTRODUCTION

It’s known to us that malaria is caused mainly by P. falciparum and every year a lot of people die because of it, especially children and pregnant women[1]. Although the progress has been made in reduction of malaria cases, parasite resistance of traditional antiplasmodial drugs is growing, such as artemisinins and chloroquine[2]. Thus, the problem of research and development of new antiplasmodial drugs must be solved urgently. According to the recent report, anti- plasmodial indole alkaloids have been discovered[3, 4]. In addition, Garcia group have also synthesized indole derivatives with antiplasmodial activity, and discovered that the presence of carboxamide group at the 3rdposition of indole was crucial for activity[5]. Recently, a series of novel indoleamide derivatives with antiplasmodial activity have been discovered by N. Devender group[1].

Three-dimensional quantitative structure-activity relationship (3-QSAR) was used to explore the relationship between the structures of molecules and their biological activities using theoretical calculation and statistical analysis tools[6]. In 3-QSAR, comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) were common models. CoMFA and CoMSIA methods were mainly based on the assump- tions in which the biological activity of the com- pound was mainly dependent on their structures[7, 8]. CoMFA model mainly considered the factors of steric and electrostatic fields. The steric and electro-static contour maps of CoMFA could be obtained by the CoMFA model. Based on this, we could know the relationship between the activity and stereostructure or electrostatic distribution[9, 10]. Comparing with CoMFA, the CoMSIA model uses a smoother gauss function which improves the sensitivity of molecular superimposed methods and spatial orientation, not only considering the factors of steric and electrostatic fields, but also the factors of hydrophobic, hydrogen bond donor field and receptor field[11, 12]. Due to the advancement of computational chemistry, many groups have employed the 3-QSAR method to predict a correlation between biological activity and chemical structure by the CoMFA and CoMSIA[13-15]. Therefore, we attempted todesign some new indo-leamide derivativeswith high predicted activity by taking advantage of useful guidelines of the 3-QSAR

2 MATERIALS AND METHODS

2.1 Preparation of the data set

A series of indoleamide derivatives, reported by Devender[1], were taken for the study. The database of 24 compounds consisted of indoleamide derivatives with antiplasmodial agents. We divided randomly the data into two subsets, a training set of 20 compounds used to build the CoMFA and CoMSIA models and a test set of 4 compounds used to further evaluate their predictive ability. The structures of all the compounds and biological data are listed in Table 1.

Table 1. Indoleamide Derivatives and Their Experimental and Predicted Activities

aTraining set molecules,bTest set molecules

2.2 Minimization and alignment

The 3structures of indoleamide derivatives were established using standard geometric parameters of molecular modeling software package SYBYL-X 2.0[16].Partial atomic charges were calculated by the Gasteiger Huckel method and energy minimizations were performed using the Tripos force field with a distance-dependent dielectric constant and Powell conjugate gradient method with a convergence crite- rion of 0.005 kcal/mol[17]. The maximum iteration was set to 10000. In 3-QSAR study, a proper molecular alignment played a vital role in main-taining bioactive conformation for the final result[18]. Compound 15 was selected as a template, because it was the most activecompound. Red bold atom in Fig. 1 was used as the common structure. Then, the alignment of all compounds of the training set is shown in Fig. 2.

Fig. 1. Alignments of all molecules. (a) Common substructure for alignment (shown in red bold atoms), (b) Alignments of all molecules

Fig. 2. Plot of the experimental activity against the predicted activity by CoMFA model (a) and CoMSIA model (b)

2. 3 CoMFA and CoMSIA studies

The training set containing 20 compounds was used to build the CoMFA and CoMSIA models, and then the predictability of models was further vali- dated bythe test set of 4 molecules. The CoMFA modelemployed Lennard Jones and Coulomb poten-tials to calculate the steric and electrostaticproperties, and used the3carbon having a +1charge as the probe atom to determine the magnitude ofthe field values[19].The distance between each two adjacentlattices of the spaced grid was 2.0 ?[20].The defaultvalue of 30 kcal/mol was set as a maximum steric and electrostatic energy cutoff[21]. Then we used partial least-squares () for analyzing the model[22]. In analysis, leave-one-out () method was employed tocross-validate in order to acquire cross-validated correlation coefficient (2)and the optimal number of components (), then non-cross-validation was used to calculate non-cross-validated correlation coefficient (2), test value () and standard error of the estimate (). PLS analysis process of CoMSIA was basically the same as CoMFA.In CoMSIA, the effects of hydrophobic, hydrogen-bond donor and hydrogen-bond acceptor on the activity of compounds were also considered[23].

3 RESULTS AND DISCUSSION

3.1 CoMFA and CoMSIA

The CoMFA and CoMSIA models wereestablishedby using compounds in the training set. Experimental andpredicted50 values of all compounds areshown in Table 1. The differences between the pre- dicted and experimental50 values are also given in Table 1. Table 2 shows the PLS results of CoMFA and CoMSIA models. As shown inTable 2, cross-vali- dated correlation coefficient (2=0.593 and 0.527) and non-cross-validated correlation coefficient (2= 0.990 and 0.953) indicated that the models were predictive and credible. External validation correla- tion coefficient (2pred=0.967 and 0.962)showed that the 3-QSAR model had high external predictable ability. The relationship of experimental and pre- dicted50 valves is shown inFig. 2. All points were almostlocated on the diagonal line,indicating the models had reasonable predictability.

3.2 3D-QSAR contour maps

The field distribution of the models could be seen from the 3-QSAR contour maps. We could modify moleculesto increase or decrease theactivity accor- ding to the 3-QSAR contour maps. Fig. 3 shows the electrostatic and stericcontour maps of CoMFA while Fig. 4 showsthe electrostatic, steric, hydrophobic and hydrogen bond receptor contour maps of CoMSIA. We selected compound 15 as the reference structure. All the contours represented the default 80% and20% level contributions for favored and disfavored regions, respectively.

Table 2. Statistical Results of the CoMFA and CoMSIA Models

2: Cross-validated correlation coefficient.: Optimum number of components.2: Non-cross-validated correlation coefficient.: Standard error of the estimate.: F -test value.2pred:External validation correlation coefficient.

Fig. 3. Contour maps of the CoMFA models. (a) Blue and red represent favorable andunfavorable regions for electrostatic field. Compound 15 was used as the template. (b) Green and yellow represent favorable and unfavorable regions for the steric field

Fig. 4. CoMSIA contour maps. (a) Electropositive (blue) and electronegative (red) fields. (b) Favorable (green) and unfavorable (yellow) steric fields. (c) Favorable (yellow) andunfavorable (red) hydrophobic fields. (d) Favorable (magenta) and unfavorable (purple) hydrogen bond donor fields. Compound 15 was used as the template

3.3 CoMFA contour maps

As Fig. 3a shows, the electrostatic contour map of CoMFA has two colors, blue and red, which repre- sents electrostatic interactions of the CoMFA model. The red contour near benzene position indicates that the addition of electron-donatinggroups in R1position will decrease the activity.The blue contour near the tertiary butyl position suggests that addingelectron-withdrawing groups in this region will de- crease the activity of the molecule.

As Fig. 3b shows, the steric contour map of CoMFA has two colors, green and yellow, which representssteric interactions of the CoMFA model. The greencontour near the benzene position indi- cates that increasing the volume of substituents in this region can enhance the activity of small molecules. The yellow contour near the tertiary butylpositionshows that decreasing the volume of substituents in this region can enhance the activity of small mole- cules.

3.4 CoMSIA contour maps

The electrostatic and steric field contour maps of CoMFA are shown in Fig. 4a and 4b. The CoMSIAelectrostaticandstericcontour maps are similar toCoMFA. Thus, the way of modifying the molecule is the same asCoMFA.

As Fig. 4c shows, thehydrophobic contour map of CoMSIA has two colors, yellow and red. The yellow contour near benzene position indicates hydrophobic groups are favored in R1, R2, R3andR4positions. The red contours indicate hydrophilic groups are favored in the region and it is nearly invisible. As Fig. 4d shows, the hydrogen bond receptorcontour map of CoMSIA has two colors, magenta and purple. The magenta and purple contours depict favorable and unfavorable positions for hydrogen bond acceptors, respectively. The magenta contour near benzene positionindicates the addition of hydrogen bond receptor in R1, R2, R3andR4positions would de- crease the activity of the molecule. The purple contour is nearly invisible.

4 DESIGN OF NEWLY ANTIPLASMODIAL DRUGS

We modified compound 15 and obtained a series of newly antiplasmodial compounds in Table 3 according to the 3-QSAR model. On the basis of 3-QSAR model, the addition of electron-donatinggroups, bulky groups and hydrophobic groups in R1position would decrease the activity, such as phenyl and methyl. Thus, compounds N1 and N2 were designed. Compound N1 shows higher activity (50 = 1.39 and 0.96 for CoMFA and CoMSIA models respectively) than compound 15. The addition of hydrogen bond receptor or electron-donatinggroups in R1, R2, R3and R4positions would decrease the activity, such as hydroxyl, phenyl and methyl. Thus, compound N3-8 was designed. The structures of newly antiplasmodial drugs and predicted50 values are shown in Table 3. Compound N1-8 showed lower50 value than compound 15, so these compounds have higher activity than com- pound 15.

Table 3. Structures and Predicted IC50 Values of Newly Designed Antiplasmodial Drugs

5 CONCLUSION

In this study, 3-QSAR was applied to explore the structure-activity relationship of novel indoleamide derivatives as antiplasmodial agents. The CoMFA and CoMSIA models exhibited good internal con- sistency in terms of non-cross-validated correlation coefficient (2= 0.990 and 0.953), Standard error of the estimate (= 0.341 and 0.678) and high pre- dictive ability (2= 0.593 and 0.527). The excellent external predictable ability of the model wasfurtherproved in terms of external validation correlation coefficient (2pred= 0.967 and 0.962). Furthermore,we designed a series of newly antiplasmodial mole- culesaccording to the CoMFA and CoMSIA contour maps. The lower50 values of these newlydesigned moleculesalso indicated that they have better activity and deserve further study.

Fig. 5. Structuresof compounds N1-8

(1) Devender, N.; Gunjan, S.; Tripathi, R.; Tripathi, R. P. Synthesis and antiplasmodial activity of novel indoleamide derivatives bearing sulfonamide and triazole pharmacophores.2017, 131, 171-184.

(2) Rottmann, M.; McNamara, C.; Yeung, B. K. S.; Lee, M. C. S.; Zou, B.; Russell, B.; Seitz, P.; Plouffe, D. M.; Dharia, N. V.; Tan, J.; Cohen, S. B.; Spencer, K. R.; Gonzalez-Paez, G. E.; Lakshminarayana, S. B.; Goh, A.; Suwanarusk, R.; Jegla, T.; Schmitt, E. K.; Beck, H. P.; Brun, R.; Nosten, F.; Renia, L.; Dartois, V.; Keller, T. H.; Fidock, D. A.; Winzeler, E. A.; Diagana, T. T. Spiroindolones, a potent compound class for the treatment of malaria.2010, 329, 1175-1180.

(3) dos Santos Torres, Z. E.; Silveira, E. R.; Rocha e Silva, L. F.; Lima, E. S.; de Vasconcellos, M. C.; de Andrade Uchoa, D. E.; Braz Filho, R.; Pohlit, A. M. Chemical composition of aspidosperma ulei markgr and antiplasmodial activity of selected indole alkaloids.2013, 18, 6281-6297.

(4) Nugroho, A. E.; Hirasawa, Y.; Piow, W. C.; Kaneda, T.; Hadi, A. H. A.; Shirota, O.; Ekasari, W.; Widyawaruyanti, A.; Morita, H. Antiplasmodial indole alkaloids from leuconotis griffithii.2012, 66, 350-353.

(5) Schuck, D. C.; Jordao, A. K.; Nakabashi, M.; Cunha, A. C.; Ferreira, V. F.; Garcia, C. R. S. Synthetic indole and melatonin derivatives exhibit antimalarial activity on the cell cycle of the human malaria parasite plasmodium falciparum.2014, 78, 375-382.

(6) Bouaziz-Terrachet, S.; Terrachet, R.; Tairi-Kellou, S. Receptor and ligand-based 3D-QSAR study on a series of nonsteroidal anti-inflammatory drugs.2013, 22, 1529-1537.

(7) Cramer, R. D.; Patterson, D. E.; Bunce, J. D. Comparative molecular-field analysis (CoMFA). 1. Effect of shape on binding of steroids to carrier proteins.1988, 110, 5959-5967.

(8) Klebe, G.; Abraham, U.; Mietzner, T. Molecular similarity indices in a comparative analysis (CoMSIA) of drug molecules to correlate and predict their biological activity.1994,37, 4130-4146.

(9) Rodrigues, C. R.; Flaherty, T. M.; Springer, C.; McKerrow, J. H.; Cohen, F. E. CoMFA and HQSAR of acylhydrazide cruzain inhibitors.2002, 12, 1537-1541.

(10) Kulkarni, S. S.; Patel, M. R.; Talele, T. T. CoMFA and HQSAR studies on 6,7-dimethoxy-4-pyrrolidylquinazoline derivatives as phosphodiesterase10A inhibitors.2008, 16, 3675-3686.

(11) Perez-Villanueva, J.; Medina-Franco, J. L.; Caulfield, T. R.; Hernandez-Campos, A.; Hernandez-Luis, F.; Yepez-Mulia, L.; Castillo, R. Comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) of some benzimidazole derivatives with trichomonicidal activity.2011, 46, 3499-3508.

(12) Phosrithong, N.; Ungwitayatorn, J. Ligand-based CoMFA and CoMSIA studies on chromone derivatives as radical scavengers.2013, 49, 9-15.

(13) Pourbasheer, E.; Aalizadeh, R.; Ebadi, A.; Ganjali, M. R. 3D-QSAR analysis of MCD inhibitors by CoMFA and CoMSIA.2015, 18, 751-766.

(14) Guariento, S.; Franchini, S.; Tonelli, M.; Fossa, P.; Sorbi, C.; Cichero, E.; Brasili, L. Exhaustive CoMFA and CoMSIA analyses around different chemical entities: a ligand-based study exploring the affinity and selectivity profiles of 5-HT1A ligands.2017, 32, 214-230.

(15) Sharma, R.; Dhingra, N.; Patil, S. CoMFA, CoMSIA, HQSAR and molecular docking analysis of ionone-based chalcone derivatives as antiprostate cancer activity.2016, 78, 54-64.

(16) Xu, H. R.; Fu, L.; Zhan, P.; Liu, X. Y. 3D-QSAR analysis of a series of S-DABO derivatives as anti-HIV agents by CoMFA and CoMSIA.2016, 27, 999-1014.

(17) Potshangbam, A. M.; Tanneeru, K.; Reddy, B. M.; Guruprasad, L. 3D-QSAR and molecular docking studies of 2-pyrimidinecarbonitrile derivatives as inhibitors against falcipain-3.2011, 21, 7219-7223.

(18) Burke, T. R.; Fesen, M. R.; Mazumder, A.; Wang, J.; Carothers, A. M.; Grunberger, D.; Driscoll, J.; Kohn, K.; Pommier, Y. Hydroxylated aromatic inhibitors of HIV-1 integrase.1995, 38, 4171-4178.

(19) Zhi, H.; Zheng, J.; Chang, Y.; Li, Q.; Liao, G.; Wang, Q.; Sun, P. QSAR studies on triazole derivatives as sglt inhibitors via CoMFA and CoMSIA.2015, 1098, 199-205.

(20) Ding, L.; Wang, Z. Z.; Sun, X. D.; Yang, J.; Ma, C. Y.; Li, W.; Liu, H. M. 3D-QSAR (CoMFA, CoMSIA), molecular docking and molecular dynamics simulations study of 6-aryl-5-cyano-pyrimidine derivatives to explore the structure requirements of LSD1 inhibitors.2017, 27, 3521-3528.

(21) Srivastava, V.; Kumar, A.; Mishra, B. N.; Siddiqi, M. I. CoMFA and CoMSIA 3D-QSAR analysis of DMDP derivatives as anti-cancer agents.2008, 2, 384-391.

(22) Lv, Y.; Yin, C.; Liu, H.; Yi, Z.; Wang, Y. 3D-QSAR study on atmospheric half-lives of POPs using CoMFA and CoMSIA.2008, 20, 1433-1438.

(23) Shah, P.; Kumar, S.; Tiwari, S.; Siddiqi, M. I. 3D-QSAR studies of triazolopyrimidine derivatives of plasmodium falciparum dihydroorotate dehydrogenase inhibitors using a combination of molecular dynamics, docking, and genetic algorithm-based methods.2012, 5, 91-103.

13 October 2017;

15 March 2018

① This project was supported by the Shanghai leading talent projects, pharmacology laboratory in China state institute of pharmaceutical industry

. E-mail: 2510848682@qq.com

10.14102/j.cnki.0254-5861.2011-1855