Swaranjali Pawar, Shalini Shinde, Pooja Chavan, S. Rathod*, P. Choudhari, Snehal Arvindkar and S. Patil1

Department of Pharmaceutical Chemistry, Bharati Vidyapeeth College of Pharmacy, Kolhapur, Maharashtra 416013, 1Department of Pharmacology, Dr. Shivajirao Kadam College of Pharmacy, Sangli, Maharashtra 416305, India

*Corresponding Author:
S. Rathod
Department of Pharmaceutical Chemistry, Bharati Vidyapeeth College of Pharmacy, Kolhapur, Maharashtra 416013, Sangli, Maharashtra 416305, India
E-mail: sanket.rathod-copk@bvp.edu.in
Date of Received 14 July 2022
Date of Revision 16 March 2023
Date of Acceptance 12 February 2024
Indian J Pharm Sci 2024;86(1):322-329  

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as the author is credited and the new creations are licensed under the identical terms

DOI: 10.36468/pharmaceutical-sciences.1281

Abstract

Cancer is one of the leading causes of mortality worldwide and researchers are working to find new ways to cure it. Cancer is a disease characterized by abnormal cell development with the ability to invade and spread to other sections of the body. Breast cancers, colon cancers, lung cancers, liver cancers, rectum cancers, and stomach cancers are the most common cancers over the globe. Cyclin-dependent kinases are one of the causes of cancer and they are protein kinases. Cyclin-dependent kinases play important roles in controlling cell division in cancer cells. This study aims to find potential cyclin-dependent kinases 1 and cyclin-dependent kinases 2 inhibitors. Several phytochemicals were subjected to molecular docking against cyclin-dependent kinases 1 and cyclin-dependent kinases 2 and phytochemicals having good binding affinity were subjected to molecular dynamic simulation to determine the stability of protein-ligand complexes.

Keywords

Cancer, cyclin-dependent kinases 1, cyclin-dependent kinases 2, computational, molecular docking, molecular dynamics

Cancer is the second leading cause of mortality in the world, in 2019 with approximately 10.1 million deaths (17.83 % of total deaths)[1]. In 2020, 19.3 million new cases of cancer were estimated worldwide, according to predictions from the Global Cancer Observatory (GLOBOCAN). After China and United States of America, India came in third. According to GLOBOCAN, 2.08 million more cases of cancer would be diagnosed in India between 2020 and 2040, an increase of 57.5 %[2]. In 2020, the most prominent malignancies diagnosed globally were female breast cancer (2.26 million cases), lung (2.21) and prostate cancer (1.41) and the most common causes of cancer death were lung (1.79 million fatalities), liver (830000) and stomach cancer (769000)[3]. Worldwide, 6.9 million new cases of cancer will be detected among persons 80 y of age or older in 2050 (20.5 % of all cancer cases)[4-6].

Cyclin-Dependent Kinase (CDK) activation abnormalities, which are extremely common in human cancers, offered a rationale for developing synthetic CDK inhibitors as anticancer drugs[7].

Pharmaceutical companies have behaved focused on the proteins that regulate cellular proliferation, differentiation, transformation and metastasis. In the human genome, there are approximately 518 distinct protein kinases[8]. At present, 13 members of the CDK family have been reported, which are named from CDK1 to CDK13. Among these members, CDK1, CDK2, CDK4 and CDK6 are directly involved in cell cycle regulation, while CDK7 controls the cycle by activating other CDKs. In addition, CDK8, CDK9 and CDK7 are transcription regulatory factors. Moreover, CDK2, CDK4, and CDK6 mediate the G1 phase of the cell cycle, whereas CDK2 and CDK1 regulate S and G2 phases, and the M phase, respectively. Among them, CDK is a key class of cell cycle proteins. CDK activity requires the binding of the regulatory subunits known as cyclins. Cyclins are generated and eliminated at precise intervals during the cell cycle, allowing for timely control of kinase activity. Each stage in the cell cycle is regulated by CDKs belonging to a family of threonine/serine protein kinases. The CDKs are master regulators of the cell division cycle in association with the cyclin regulatory subunits[9]. The CDK1/cyclin B complex is responsible for completing the mitotic process. Despite the highly regulated nature of the process, CDK1 is the only CDK required for cell cycle modulation[10]. CDK1 controls the start of the cell cycle and its progression through mitosis. Recent research has linked loss of CDK1 function or abnormal CDK1 expression to G2 phase arrest and a variety of tumor types, validating CDK1 as a therapeutic target[11]. CDK2 is the family of serine/ threonine protein kinases that have developed as a potent, selective, and low-toxic cancer therapy target[12]. CDK2 binds to cyclins A, B, and E and plays an important role in the G1/S transition, the start of the synthesis of DNA, and the regulation of the exit from the S phase. Furthermore, altered expression of cyclins, whose interaction with CDKs is required for their catalytic activity, can promote cancer proliferation[13]. Highly selective CDK2 inhibitors will aid in the continuous discovery of cancer subtypes that are responsive to CDK2 inhibition by defining the additional unique role of CDK2 in cancer development[14].

Phytochemicals and their derivatives have the potential to improve cancer therapy efficiency while reducing side effects[15]. Several of these phytochemicals are physiologically active substances found in nature with high anticancer potential. Phytochemicals often work as anticancer agents by regulating molecular pathways that have been related to cancer growth and progression[15]. Because of their structural variety, natural products have remained one of the most important frameworks in future anticancer agent development[16]. More than 100 novel compounds derived from natural products such as Etoposide, Paclitaxel and Vinblastine have been introduced into the market[17].

Materials and Methods

Protein and ligand preparation:

The previously reported Three Dimensional (3D) crystal structure of CDK1 (PDB: 6GU7)[18-21] having 2.75 Å resolution and CDK2 (PDB: 6GUE) [22-24] having 1.99 Å resolution was retrieved from the RCSB Protein data bank[7]. The downloaded structures of the protein were prepared for docking as per the previously described protocol[25]. Thereafter prepared structures were evaluated for stereochemical adaptability and the quality of the protein structure using the PROCHECK and ProSAweb server[26,27]. The structures of phytochemicals were downloaded in mol2 file format from Indian Medicinal Plants, Phytochemistry, and Therapeutics (IMPPAT) (https://cb.imsc.res.in/imppat/home)[28-32]. IMPPAT is an Indian medicinal phytochemical compound database and it helps to find the accessible compounds of the required plant. This database has various filters and can help to find the compounds based on biological activity[33-35]. All downloaded mol2 files of phytochemicals were imported into BIOVIA Discovery Studio 2020 to optimize by adding hydrogen atoms and convert into a PDB file format. Further, all ligand groups were energy minimized using the OpenBabel module of the PyRx with MMFF94 force field and steepest descent algorithm and used for further molecular docking studies[36,37]. Dinaciclib was used as a standard to compare the docking results[38].

Molecular docking studies:

Molecular docking was carried out to investigate the binding affinity and interactions between the subjected phytochemicals and target protein (PDB: 6GU7 and PDB: 6GUE). AutoDock Vina of PyRx 0.8 was used to perform molecular docking[37-40]. The prepared structures of proteins and phytochemicals were imported in PyRx 0.8. A maximized grid box with center X: 22.0372, Y: 12.6379, Z: 4.7538 and dimensions X: 60.73 Å, Y: 41.37 Å, Z: 66.32 Å for PDB: 6GU7 and with center X: -10.889, Y: -7.5226, Z: 9.3097 and dimensions X: 94.9893Å, Y: 108.3853Å, Z: 98.3694Å for PDB: 6GUE was selected in the Vina workspace to cover the entire protein structure. The exhaustiveness was kept to default at eight and nine different conformations was predicted for each phytochemical with the selected target protein (PDB: 6GUE and 6GU7) [25,37]. Further, the best pose with the highest negative binding affinity was selected for each molecule[41]. Docking interaction, visualization, and analysis of saved conformations were carried out with the help of BIOVIA Discovery Studio.

Molecular dynamic simulation:

Protein-ligand complexes of phytochemicals having good binding affinity against both targeted proteins are subjected to Molecular Dynamic (MD) simulation. The selected 6GUE-107876 and 6GU7-4871complex files were subjected to MD simulation studies with the help of the GROMACS simulation package through the WebGRO and the GROMOS96 43a1 force field was chosen to perform this simulation[42]. Topology files of phytochemicals were built with the help of the PRODRG 2.5 server[43]. The entire protein-ligand complex system was solvated using the Simple Point Charge (SPC) water model with a triclinic box[44]. The 5000 steps of steepest descent were performed for Energy Minimization (EM) of the subjected complex[41]. The complex system was neutralized and the simulation was performed in the presence of 0.15 M NaCl[45]. Equilibration of the simulated systems was done using canonical (NVT) and isothermal-isobaric (NPT) ensembles after each step of EM[46-48]. The temperature on the system kept maintained using the Berendsen thermostat approach and pressure was kept maintained at 1.0 bar using the Parrinello-Rahman barostat approach to control the simulated proteinligand complex systems[49-51]. The simulation was performed under the 300 K temperature for 100 ns, respectively. The obtained trajectories of MD simulation of 100 ns were subjected to determine the conformational changes and stability of the protein-ligand complex. Obtained MD trajectories were subjected to analysis of the Root Mean Square Deviation (RMSD), Root Mean Square Fluctuation (RMSF), Radius of gyration (Rg) and Hydrogen Bonds (HBs), of the subjected complex system of hit molecule[52,53].

Results and Discussion

Virtual screening techniques help to increase the efficiency of the drug discovery process. Target proteins were selected on the basis of previously reported literatures. Selected phytochemicals were prepared for molecular docking using the free version of BIOVA Discovery Studio 2020. The selected protein structure was subjected to quality evaluation and binding site analysis by using PROCHECK and ProSA-web servers. The Ramachandran plot revealed that 92.7 % of PDB: 6GUE and 87.9 % of 6GU7 residues are present in favored regions as shown in fig. 1. The z-score indicates overall model quality for both the prepared protein structures was found to be -7.93 for PDB: 6GUE and -6.42 for PDB: 6GU7. Fig. 2 represents a plot that contains the z-scores of all experimentally determined protein chains in both PDB: 6GU7 and 6GUE.

pharmaceutical-sciences-region

Fig. 1: Ramachandran plot of the (a) CDK2 (PDB ID: 6GUE) showing 92.2 % residues in the favored region and (b) CDK1 (PDB: 6GU7) showing 87.9 % residues in the favored region

pharmaceutical-sciences-chain

Fig. 2: Z-Score plot for experimentally determined protein chains in both (a) CDK2 (PDB: 6GU7) and (b) CDK1 (PDB: 6GU7) Note: (Image): X-ray and (Image): NMR

Over the past few decades molecular docking technique established as an outstanding tool to determine the binding energies and best binding confirmations in the interaction of target protein and potential small molecules and nowadays, it is a widely applied computational tool. In the present work, molecular docking was done by using the AutoDock Vina package of PyRx 0.8 software[41]. Downloaded phytochemicals were subjected to molecular docking against CDK1 and CDK2 (PDB ID: 6GU7 and 6GUE). The binding affinity of phytochemicals against CDK1 (PDB: 6GU7) ranged from -4.1 to -10.4 kcal/mol and for phytochemicals against CDK2 (PDB: 6GUE) ranged from -4.5 to -9.9 kcal/mol. Among all the phytochemicals, PubChem CID 4871 showed a good binding affinity of -10.4 kcal/mol with CDK1 (PDB: 6GU7) and for CDK2 (PDB: 6GUE), PubChem CID 107876 showed good binding affinity of -9.9 kcal/mol. The binding affinity of the docked control i.e. Dinaciclib showed a binding affinity of -8.6kcal/mol with PDB:6GU and -9.2kcal/mol with PDB: 6GUE. Both the selected hits i.e. PubChem CID 4871 against CDK1 (PDB: 6GU7) and PubChem CID 107876 against CDK2 (PDB: 6GUE) showed better binding affinity than compared standard (Dinaciclib). The BIOVA Discovery Studio Visualizer was used to determination of detailed Two Dimensional (2D) and 3D interactions of best conformation. The interaction of PubChem CID 4871 with residues of CDK1 (PDB: 6GU7) and PubChem CID 107876 with CDK2 (PDB: 6GUE) showed the highest affinity and interaction. GLN132, ASP146, ALA31, LEU135, PHE82, ILE10, LEU83, GLU130, THR14, GLY13, LYS130, PHE80, VAL64, ALA145, VAL18 are the interacting amino acid residues between PubChem CID 4871 and CDK1 (PDB: 6GU7) as shown in fig. 3a and fig. 3b and they formed van der Waals, conventional hydrogen bond, carbon hydrogen bond, π-Cation, π-Sigma, Alkyl, π-Alkyl interactions. While in the case of PubChem CID 107876 and CDK2 (PDB: 6GUE), LYS178, PRO155, ARG157, VAL154, GLY153, THR316, GLN313, GLU230, GLN228, ASN229, TYR418, MET334, TYR413, SER331, HIS419, LYS417, GLU330, PHE319, TYR179, VAL156 are interacting amino acid residues as shown in fig. 3c and fig. 3d. van der Waals interaction was not indicated in the 2D and 3D visitation of interaction as it does not have a direct interaction between the protein-ligand complex. Both the phytochemicals were further subjected to a molecular dynamics simulation study to determine the stability of the complex over 100 ns of simulation time.

pharmaceutical-sciences-pub

Fig. 3: 2D and 3D interaction of (a-b) PubChem CID 4871 with 6GU7 and (c-d) PubChem CID 107876 with 6GUE

MD simulation studies were performed on proteinligand complexes having good binding affinity. Through the MD simulation technique, the dynamic behavior and configurational changes between 6GUE-107876 and 6GU7-4871 complexes were analyzed over 100 ns. Analysis of parameters like RMSD, RMSF, Rg and HBs were done over 100 ns using generated MD trajectories from WebGro. The conformational stability of the protein-ligand complex system was analyzed with the help of RMSD. RMSD of the protein-ligand complex was calculated to obtain the equilibrium time of the simulated complexes (6GUE-107876 and 6GU7- 4871). Low RMSD indicates more stability of the protein-ligand complex system. Fig. 4a represents the RMSD plot for the 6GUE-107876 and 6GU7- 4871. Through the results, it is observed that the RMSD of complexes showed minimum deviation over the simulation time of 100 ns and it indicates that both complexes have good conformational stability. The number of hydrogen bonds present in the complexes with its consistency throughout the 100 ns simulation was determined at 300 K, as shown in fig. 4b. There were not many considerable changes observed in the hydrogen bond interaction within the protein-ligand complexes. RMSF was plotted at 300 K as given in fig. 4c and fig. 4d. Through the RMSF values, it is observed that fluctuations occur in 6GUE-107876 complex and LEU25 (RMSF value: 2.3530 Å), THR97 (RMSF value: 1.9860 Å), GLU138 (RMSF value: 2.0070 Å), ARG157 (RMSF value: 1.9930 Å), TYR159 (RMSF value: 2.9340 Å), SER171 (RMSF value: 3.7090 Å), PRO272 (RMSF value: 1.9310 Å) amino acid residues showed fluctuations. And in case of 6GU7-4871, GLU40 (RMSF value: 3.8220 Å), GLU41 (RMSF value: 3.9000 Å), PRO95 (RMSF value: 2.4580 Å), PRO96 (RMSF value: 3.0990 Å), GLY97 (RMSF value: 3.0930 Å), ASN 224 (RMSF value: 4.4950 Å), ASN225 (RMSF value: 5.4240 Å), GLU226 (RMSF value: 3.4240 Å) amino acid residues showed fluctuations, in which ASN225 showed highest RMSD value. However, the RMSF values of both complexes remain in the acceptable range indicating the stability of particular amino acid residues. Rg values for both 6GUE-107876 and 6GU7-4871complexes have been studied over 100 ns at 300 K and it is observed that the Rg values for the 6GUE-107876 ranged between 2.46-2.58 nm and for 6GU7-4871 Rg value ranged between 1.85 to 1.95 nm with showing conformational stability of the complex with minimum fluctuation in the plot, as shown in fig. 4e and fig. 4f.

pharmaceutical-sciences-radius

Fig. 4: (a) RMSD; (b) Hydrogen bonds; (c-f) Radius of gyration, for the complex of PubChem CID 4871-CDK1 (PDB: 6GU7) and PubChem CID 107876-CDK2 (PDB: 6GUE) at 300 K for 100 ns Note: (Image):6GUE-107876 and (Image): 6GU7-4871

Phytochemicals having anticancer activity were virtually screened against CDK1 (PDB: 6GU7) and CDK2 (PDB: 6GUE) to determine the binding affinity and stability. Through the results of molecular docking, it is observed that out of all, PubChem CID 4871 and PubChem CID 107876 showed a good binding affinity with the targeted proteins and the visualization of interaction formed between both the complex reviled that VDW, Conventional H Bond, Carbon H Bond, π-Cation, π-Sigma, Alkyl, π-Alkyl played a major role in the binding of the ligand with targeted protein. MD simulation of protein-ligand complexes having good binding affinity was performed over 100 ns using the RMSD, RMSF, Rg and hydrogen bonds was analyzed. Through the results of MD simulation, it is observed that subjected 6GUE- 107876 and 6GU7-4871 complexes were stable throughout the performed 100 ns simulation time, although the CDK1/CDK2 inhibitory activity of subjected phytochemicals requires an in vivo or in vitro experimental confirmation.

Conflict of interest:

The authors confirm that this article's content has no conflict of interest.

References