Computational analysis of benzofuran-2-carboxlic acids as potent Pim-1 kinase inhibitors

Abstract

Context: The three Pim serine/threonine kinases (Pim-1, Pim-2, and Pim-3) belong to a small family of kinases that regulate numerous signaling pathways fundamental to the development of tumors. Pim kinases’ overexpression has been reported in numerous solid and hematological tumors and, in particular, prostate cancer (Pim-1).

Objectives: This study investigated the binding modes of benzofuran-2-carboxlic acids against Pim-1 kinase, hence providing useful information for the active inhibition of it.

Materials and methods: In present study, molecular docking approach via MOE-Dock program was applied to predict the binding interactions of some known Pim-1 kinase inhibitors. First validation of the docking protocol was carried out by calculating RMSD for the co-crystallized and docked ligands. Using the same protocol, all the compounds were docked into the active site of Pim-1 kinase.

Results: All the compounds showed significant interactions and good correlation with the experimental data. The results illustrate that compounds with optimum basicity and relevant distance between the acidic and basic groups showed optimum interactions with the active site residues of Pim-1 kinase.

Conclusion: We hope that this study will be helpful in designing new, structurally diverse and more potent compounds for the active treatment of prostate cancer and other related diseases caused by deregulation of Pim-1 kinase.

• Cancer and inhibitor
• molecular docking
• Pim kinase

Introduction

Pim kinases of the serine/threonine family regulate cell survival (Amaravadi & Thompson, 2005). This family has three members, i.e., Pim-1, Pim-2, and Pim-3. All three Pim kinases isoforms are very much homologous at the amino acid level but still differ moderately in their tissue distribution (Allen et al., 1997; Eichmann et al., 2000). The proto-oncogene of Pim-1 was first recognized as a locus often activated by incorporation of provirus in Moloney murine leukemia virus-induced mouse T-cell lymphoma. Pim-2 was recognized as a gene commonly activated in a secondary transplant of virus-induced lymphomas. The identification of Pim-3 kinase is based on the relationship with Pim-1 and Pim-2. The observation of transgenic mice provided a confirmation that overexpressed Pim-1 and Pim-2 kinases in lymphoid system and developed lymphomas; these kinases are oncogenic in nature (Nawijn et al., 2011) and the involvement of these kinases in the development of tumor is also well studied. Pim-1 and Pim-2 are found to be responsible for prostate cancer development in some cases (Aho et al., 2004; Cibull et al., 2006). Recent studies revealed that prostate cancer is increasing rapidly (Dai et al., 2005). As this cancer is less sensitive to chemotherapy, there is a need to develop a new and potent drug for its treatment (Siu et al., 2011). The over expression of Pim-1 in head and neck squamous cell carcinoma and bladder cancer is also evidence for causing cancer. In colorectal, pancreatic and hepatocellular carcinoma, Pim-3 is over expressed (Li et al., 2006; Popivanova et al., 2007; Wu et al., 2010).

The mechanism by which Pim kinase control the proliferation of tumor cells may include phosphorylation, activation of molecules that positively regulate cell cycle progression or inactivation of cell cycle inhibitors p27kips or p2cipi (Bachmann & Moroy, 2005; Geromichalos, 2007; Morishita et al., 2008). Pim kinase may control cell capability by phosphorylating apoptotic protein BAD and ASK1 (Morishita et al., 2011) and are concerned with the capability of drug resistance.

In the current study, molecular docking of some known inhibitors of Pim-1 kinase (Xiang et al., 2011) was carried out to understand their binding modes. Molecular docking was carried out by MOE-Dock implemented in molecular operating environment (MOE) (www.chemcomp.com). The results of this study might be used to design new and more potent Pim-1 kinase inhibitors.

Materials and methods

In this study, an effort was made to carry out the docking of some known ligands into the binding pocket of Pim-1 kinase with the following communications. Intel (R) xenon (R) CPU [email protected] GHz system (Hewlett-Packard Development Company, Houston, TX) having 3.8GB RAM with the open 11.4 (X 86_64) operating platform was used. The software package MOE was used for docking. MOE is a software system designed by the Chemical Computing Group to support cheminformatics, molecular modeling, bioinformatics, virtual screening, and structure-based-design and can be used to build new applications based on Scientific Vector Language (SVL). To imagine the interaction between Pim-1 protein and ligands, ligPlot implementation in MOE was used.

Retrieval of ligands

The inhibitors for Pim-1 kinase incorporated in our study were all collected from the previous literature (Xiang et al., 2011). The structures of these inhibitors were constructed using MOE-Builder tool (Emerson Electric Co, St. Louis, MO). The correlated 3D structures were modeled and partial charges were calculated using MOE. The energies of the recognized molecules were minimized using the energy minimization algorithm of MOE tool. The following parameters were used for energy minimization; gradient: 0.05, Force Field: MMFF94X, Chiral Constraint: Current Geometry. All the minimized molecules were saved in the (mdb) file format. In the next step, the prepared ligands were used as input files for MOE-Dock.

Protein preparation

The protein molecule (Pim-1 kinase) included in our study was obtained from Protein Data Bank. Water molecules were removed and the 3D protonation of the protein molecule was performed via MOE. The energy of the protein molecule was minimized using the energy minimization algorithm of MOE tool. The following parameters were used for energy minimization; gradient: 0.05, force field: MMFF94X +dolvation, chiral constraint: current geometry. Energy minimization was terminated when the root mean square gradient falls below 0.05. The minimized structure was used as the template for docking.

Molecular docking

The binding mode of the ligands into the binding pocket of protein molecule was predicted by MOE-Dock implemented in MOE. Similar to our previous study (Wadood et al., 2013), molecular docking was carried out with most of the default parameters. After the completion of docking, we analyze the best poses for hydrogen bonding/π–π interactions and root mean-square deviation (RMSD) calculation using MOE applications.

Results and discussion

Validation of docking protocol

In order to assess the accuracy of MOE-Dock, the co-crystallized ligand (PDB ID; 3R02) was extracted from the active site and re-docked into the binding site of Pim-1 kinase. The RMSD between the co-crystallized ligand and the top-ranked docked conformation was observed to be 1.2434 Å, suggesting the high docking reliability of MOE-Dock protocol. The MOE-Dock and set protocol could be extended to explore the Pim-1 kinase binding conformations for other inhibitors accordingly.

Docking analysis of 5-substituted benzofuran-2-carboxylic acids

A total of 14 compounds of this class were docked into the binding cavity of Pim-1 kinase. The docking results showed a good correlation with the experimental data. The experimentally top active compounds were also ranked as top on the basis of docking score. The results produce a good correlation coefficient (r²= 0.6903) between the docking scores and IC₅₀ values of the ligands (Figure 1A).

Figure 1. (A) Correlation graph between experimental values and docking scores of 5-substituted benzofuran-2-carboxylic acids. (B) Correlation graph between the experimental values and the docking scores of 5-bromo-7-substituted benzofuran-2-carboxylic acids.

On the basis of IC₅₀ values and docking scores, the top 6 among the 14 docked compounds were discussed in this report. Furthermore, these top six compounds were classified as the most active (potent), moderately active and least active on the basis of their IC₅₀ values.

Binding interactions between ligand and Pim-1 kinase

Most active ligands

According to IC₅₀ values, compounds 10 (IC₅₀ 0.12 µM) and 13 (IC₅₀ 0.053 µM) were the most active in the series. Interestingly, these two compounds also showed the best docking scores as shown in Table 1.

Table 1. Experimental values and docking scores of compounds (5-substituted benzofuran-2-carboxylic acids).

Compounds	Structures	Pim-1 IC50 µM	S scores
1		5.3	−5.7845
2		6.1	−5.3042
3		7.8	−5.4838
4		7.4	−5.3949
5		5.4	−5.5014
6		0.38	−6.1006
7		6.0	−6.0374
8		1.9	−6.2159
9		0.46	−6.0340
10		0.12	−6.2304
11		0.20	−6.1287
12		0.21	−6.1294
13		0.053	−6.2317
14		0.45	−6.0407

Compound 13, having the highest biological activity, established five different contacts with the active site residues of Pim-1 kinase (Figure 2A). The hydrophobic residue Ile185 interacts with benzofuran core through the arene–hydrogen bond. Acidic residue Glu89, basic Lys67, and hydrophobic Phe187 are linked to the carbonyl oxygen of the carboxylic group through a water molecule. The acidic residue Asp186 interacted directly with the same oxygen. In the case of compound 10, a total of three interactions were observed. The basic Lys67 showed interaction with the nitrogen of aniline moiety. Hydrophobic residue Leu174 was involved in the arene–hydrogen interaction with the pyrazine ring. Polar Gln127 interacted through the hydrogen bond with the carbonyl oxygen of the carboxylic group.

Figure 2. (A) The docking conformation of the most active compound 13 in the active of Pim-1 kinase. (B) The docking conformations of compound 11. (C) Interaction of compound 12 with the active site of Pim-1 kinase. (D) The docking conformation compound 6. (E) Interaction of compound 14 with the active site.

The structural difference between compounds 13 and 10 is the methylene (CH₂) group bonded to the terminal amine group in compound 13. This methylene group is suggested here to be the cause of somewhat superior enzymatic potency and binding interactions of this compound as compared with compound 10 because it enhances the distance between the basic (amine) and the acidic (benzofuran) group.

Moderately active compounds

Compounds 11 (IC₅₀ 0.20 µM) and 12 (IC₅₀ 0.21 µM) were grouped as moderately active compounds according to their IC₅₀ values in the series. These two compounds are equipotent and also showed similar binding modes. Both the compounds showed bonding with the residues Lys67, Glu89, Asp186, and phe187. Compound 12 also showed two additional interactions with Leu44 and Gln127 (Figure 2B and C).

These moderately active compounds showed virtually similar behavior to the most active compounds 10 and 13, but a slightly low enzymatic potency and docking scores of these compounds might be due to some structural features which are the different spatial positions of cyclohexyl amine group in compound 11 and the lack of terminal amine group in compound 12, as shown in Table 1.

Least active compounds

Compounds 6 (IC₅₀ 0.38 µM) and 14 (IC₅₀ 0.45 µM) are almost equipotent and ranked as the least active compounds based on their IC₅₀ values and docking scores. In the case of compound 6, three interactions were observed with the active site residue as shown in Figure 2D. Acidic Asp186 was found to establish a hydrogen bond with the carbonyl oxygen of the carboxylic group, Lys67 made polar interaction with the oxygen atom of benzofuran core whereas Leu174 was involved in arene–hydrogen interaction with the pyrazine ring of the compound. The least active compound 14 interacted with the residues Lys67 and Leu44 (Figure 2E).

The structural difference between compounds 6 and 14 is the methyl piperidine group in compound 6 and the azitidine in compound 14. The docking study suggested that the methyl piperidine group is somewhat more active than azitidine. Both these compounds lack the terminal amine group as present in the most active compounds 10 and 13 which might be one of the reasons of slightly inferior activities and docking scores of these compounds.

Docking analysis of 5-bromo-7-substituted benzofuran-2-carboxylic acids

A total of 16 compounds from this class were docked into the binding pocket of the Pim-1 kinase. The docking score for each compound showed a significant correlation with the experimental data, as shown in Table 2. A correlation coefficient (r²= 0.859) was obtained from the docking score and IC₅₀ values of the ligands (Figure 1B).

Table 2. Experimental values and docking scores of the compounds (5-bromo-7-substituted benzofuran-2-carboxylic acid) against Pim-1 kinase.

Compounds	Structures	Pim-1 IC50 µM	S scores
15		0.060	−6.6492
16		4.9	−6.2419
17		0.071	−6.6895
18		0.020	−6.9834
19		0.051	−6.6734
20		4.5	−6.1614
21		0.040	−6.6990
22		0.15	−6.4726
23		0.055	−6.6100
24		0.22	−6.4228
25		0.98	−6.3989
26		2.7	−6.3236
27		0.001	−6.9951
28		0.019	−6.9928
29		2.3	−6.2385
30		2.4	−6.2412

Binding interactions between ligands and Pim-1 kinase

All the docked compounds were observed to fit well in the binding pocket of Pim-1 kinase. Some of the most active compounds of this class were selected here for a discussion on the basis of their experimental enzymatic potency and docking scores.

The most active compounds 27 (IC₅₀ 0.001 µM) and 28 (IC₅₀ 0.019 µM) having almost similar biological activities were observed to show similar interactions and docking scores. These compounds interacted with seven important active site residues Glu89, Glu171, Asp128, Asp186, Asn172, Lys67, and Phe187. These compounds (27 and 28) are stereoisomers having almost same inhibitory activities, and therefore showed almost similar interactions and docking scores (Figure 3A).

Figure 3. (A) The docking conformation of a potent compound 27 in the active site of Pim-1 kinase. (B) The docking conformation of compound 19. (C) The docking conformation of compound 21. (D) Interaction analysis of compound 23. (E) The docking conformation of compound 15 in the active site of Pim-1 kinase.

In the case of compound 18, four amino acids, Asp186, Lys67, Glu89, and Phe187, contributed in the interactions. Acidic residue Asp186 interacted directly with the carbonyl oxygen of the carboxylic group. Basic residue Lys67 formed hydrogen bond with the oxygen of the carboxylic group directly and interacted through water molecule as well. The acidic residue Glu89 and hydrophobic residue Phe187 were linked through a solvent with the carbonyl oxygen atom of the carboxylic group.

Compounds 19, 21, and 23 are equipotent (IC₅₀ 0.041–0.055 µM) and also showed similar binding modes (Figure 3B–D). Compound 21 showed interactions with Asp128, Glu171, and Asn172. Lys67 interacted directly and Glu89 through water with the carboxylic group of the compound. Compound 23 established similar interactions like compound 21. Compound 19 showed interactions with Asp128 and Lys67, Glu89 and Phe187. In addition, Val52 showed arene–hydrogen interaction with the benzofuran core of the compound.

In the case of compound 15, four interactions were observed with active site residues (Figure 3E). The hydrophobic residues Val52 made arene–hydrogen bond with the benzofuran core of the compound and Phe187 interacted with the carboxylic group through water-bridge. Acidic residue (Glu89) and basic residue (Lys67) showed similar interaction as Phe187 with an additional interaction of Lys67 with the carbonyl oxygen of carboxylic acid of the compound.

The docking study of these compounds suggests that the basic nature of terminal amine group plays a significant role in the inhibitory activity of this class of compounds. The regular increase in the alkyl chain, associated with the piperidine ring of compounds 15 and 19 and replacing the oxygen atom by the amine group in compound 21 (Table 2) enhances the basicity of the terminal piperidine in these compounds (19 and 21) gradually. Therefore, with a consistent improvement in the docking scores, binding interactions were observed that well correlated with the enzymatic potencies of these compounds. Similarly, a decrease in the distance between the piperidine ring and the associated amine bridge in compound 18 as compared with compound 21 showed the same result.

Conclusion

In the present study, 14 compounds of 5-substituted benzofuran-2-carboxylic acid and 16 compounds from 5-bromo-7-substituted benzofuran-2-carboxylic acid class were docked into the binding pocket of the Pim-1 kinase. The docking results showed a good correlation with the experimental data. From the docking study, it was observed that in the case of 5-substituted benzofuran-2-carboxylic acids, methyl piperidine might play an important role whereas in the case of 5-bromo-7-substituted benzofuran-2-carboxylic acids, the basic nature of terminal group of these compounds played a significant role in their activities. Thus, we hope that this study would be helpful in designing new and more potent compounds for the treatment of prostate cancer and other related diseases caused by deregulation of Pim-1 kinase.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.