Molecular Docking of Known Carcinogen 4- (Methyl-nitrosamino)-1-(3-pyridyl)-1-butanone (NNK) with Cyclin Dependent Kinases towards Its Potential Role in Cell Cycle Perturbation

Cell cycle is maintained almost all the times and is controlled by various regulatory proteins and their complexes (Cdk+Cyclin) in different phases of interphase (G1, S and G2) and mitosis of cell cycle. A number of mechanisms have been proposed for the initiation and progression of carcinogenesis by abruption in cell cycle process. One of the important features of cancer/carcinogenesis is functional loss of these cell cycle regulatory proteins particularly in CDKs and cyclins. We hypothesize that there is a direct involvement of these cell cycle regulatory proteins not only at the genetic level but also proteins level, during the initiation of carcinogenesis. Therefore, it becomes significant to determine inconsistency in the functioning of regulatory proteins due to interaction with carcinogen 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK). Hence, we investigated the interaction efficiency of NNK, against cell cycle regulatory proteins. We found a different value of ΔG (free energy of binding) among the studied proteins ranging between -3.29 to -7.25 kcal/mol was observed. To validate the results, we considered Human Oxy-Hemoglobin at 1.25 Å Resolution, [PDB_ID:1HHO] as a +ve control, (binding energy -6.06 kcal/mol). Finally, the CDK8 (PDB_ID:3RGF) and CDK2 (PDB_ID:3DDP) regulatory proteins showing significantly strong molecular interaction with NNK -7.25 kcal/mol, -6.19 kcal/mol respectively were analyzed in details. In this study we predicted that CDK8 protein fails to form functional complex with its complementary partner cyclin C in presence of NNK. Consequently, inconsistency of functioning in regulatory proteins might lead to the abruption in cell cycle progression; contribute to the loss of cell cycle control and subsequently increasing the possibility of carcinogenesis.

induce downstream processes by phosphorylating the regulatory proteins in cell cycle [2]. However, different types of cyclins D (cyclin D1, cyclin D2 and cyclin D3) are essential for the regulation of cell cycle. Further, cyclins bind to CDKs and form CDK-cyclin complex. This complex/s is/are very important for phase progression in cell cycle [3]. The functional activity of that CDK-cyclin complex is induced conformational changes due to phosphorylation in conserved threonine and tyrosine residues of CDKs part. Thus the complex (Cdk+Cyclin) enhances or suppresses the binding efficiency of their specific cyclin partner [4]. In case of cancer, mutations have been observed in genes encoding CDKs, Cyclins (D1, D2, D3, A, H, T & C), CDK-activating enzymes (CAK), CKI and CDK-substrates [5]. Consequently, changes in functioning of CDKs activity to formed Cdk+Cyclin complex, an important part of many cancers, as well as other disease states, generally through elevated and/or inappropriate activation [6]. One of the probable ways of loss of cell cycle control could be disruption of these regulatory proteins CDKs (check points) by a direct interaction with chemical/carcinogen. In the present study, we investigated the possible molecular interactions between CDKs, cyclins with potent cigarette smoke carcinogen 4-(Methylnitrosamino)-1-(3-Pyridyl)-1-butanone (NNK).
NNK is a yellow crystalline compound with a molecular formula C10H13N3O2, (Mol Weight = 207.2316) derived from tobacco alkaloids (nitrosamine) as a potent carcinogen [7]. The concentrations of NNK in tobacco substances such as 1-20 μg/g, 20-310 ng/cigarette and ≤26 ng/m 3 in snuff, cigarette mainstream smoke, and in indoor air respectively [8]. The existences of substantial amount of NNK in tobacco products play a very significant role as a main cause of cancer in population [9]. Ultimately, NNK is considered as a major contributor as well as risk factor to lung carcinogenesis [10][11]. NNK and its metabolite NNAL, has already been reported to molecular interaction with DNA repair proteins primarily by our group [12].
We analyzed the molecular interaction, based on binding efficiency of potent carcinogen NNK against CDKs (check points) and cyclins involved in the cell cycle process. In case of cancer, cells develop an autonomous set of instructions against normal rules, leading to uncontrolled, undifferentiated growth and proliferation called an abnormal condition of a cell known as carcinogenicity. The carcinogen NNK and its metabolite NNAL directly bind with DNA repair proteins to make DNA adduct. Therefore, possibility exists that NNK may directly interact with these regulatory protein (CDKs, cyclin) and affect the functional activity of CDK-cyclin complex in cell cycle regulation. The carcinogen NNK induced cell cycle abruption may in turn result in hastened DNA replication, with a compromised proofreading by DNA pol/RNA pol. This process in a whole may give rise to daughter cell with loads of mutations in their DNA. Therefore we designed this study to investigate and determine whether the carcinogen NNK, apart from directly causing damage to regulatory proteins, is also capable of affecting their functionalities in term of binding efficiency of CDKs to their specific cyclins partner (A, C, D, D1, D2, D3, E, and H) for proper signal to execute cell cycle phases normally.

Methodology: Preparation of receptor-protein structures
The 3D structures of check point proteins (CDKs & Cyclins) involved in cell cycle regulation were obtained from PDB (Protein Data Bank) and some other proteins retrieved from MODBASE server Table 1 (see supplementary material). MODBASE is a queryable database of annotated protein structures models, theoretically calculated models, which may contain significant errors, not experimentally determined structures [13]. Published protein structures were edited, to remove HETATM by using Discovery Studio Visualizer (Version 2.5.5). And Chimera was used for energy minimization, removal of steric collision (forces) with the steepest descent steps 1000, steepest descent size 0.02 Å, conjugated gradient steps 1000 and gradient step size 0.02 Å for the conjugate gradient minimization [14 -15]. Protein structure visualization and image generation were performed using PyMOL software (DeLano Scientific, Palo Alto, CA).

In silico Studies
In silico studies were performed by Autodock version 4.0 suit with Cygwin interface tool [18][19]. We selected molecular docking methods for CDKs and NNK interaction, followed by retrieving the best conformations of check point (CDKs) as regulatory proteins and carcinogen (NNK), on the basis of binding energy value (kcal/mol). First of all, we marked all water molecules (H2O) in proteins then removed from targeted protein structure, before apply docking performance. Then only hydrogen atoms were added to all target proteins. After that Kollman united charges and salvation parameters were applied to selected regulatory proteins (CDKs and cyclins). Gasteiger charge also was charged to ligand (NNK). Then defined Grid box was set to cover the maximum part (including target site) of selected protein for ligand interaction. The value was set to standard 60×60×60 Å in X, Y and Z coordinate of grid point with default value of grid points spacing 0.375 Å. Lamarckian Genetic Algorithm (LGA) was applied for receptor protein and ligand for flexible docking calculations [20]. The LGA parameters like population size (ga_pop_size), energy evaluations (ga_num_generation), mutation rate, crossover rate and step size were set to 150, 2500000, 27000, 0.02, 0.8 and 0.2 Å, respectively. The LGA runs were set to as standard 50 runs. We observed all 50 conformations of selected proteins with ligand complex were analyzed for the interaction orientations including binding energy of the docked structure using Discovery Studio Visualizer version 2.5.5.

Protein-Protein Interaction analysis
The interacting regulatory proteins (CDKs) and cyclins were found using STRING 9.0 database that predict, interacting proteins against your query. We found interface residues in CDKs & Cyclins using PDBe PISA, an interactive tool for the exploration of macromolecular (protein, DNA/RNA and ligand) interfaces [21]. The Discovery studio 2.5 was used for Zdock (Dock Proteins) and Zdock score obtained from Protein-Protein interactions (Cdk+Cyclin) as well as from Protein-Complex (NNK+CDK & Cyclin) interaction.

Z dock calculations
Zdock is one of the successful suites that have shown great prediction abilities in Critical Assessment of Predicted Interactions (CAPRI) [22]. Zdock uses a fast Fourier transform to search all possible binding modes for the proteins, evaluating based on shape, desolvation energy, and electrostatics. The top 2000 predictions from Zdock where they are minimized by CHARMM with create fixed atom constraint in backbone of protein and again create Harmonic restraint in selected protein for improving the energies and eliminate clashes. ZDOCK is an initial stage rigid body molecular docking algorithm that uses a fast Fourier transform (FFT) algorithm to improve performance for searching in translational [23]. All of the available structures from NMR were used to calculate docking poses and the structures obtained were subjected to energy minimization using, smart minimize algorithm (Max steps 200, RMS gradient 0.01) in the program D.S. 2.5. The resulting highest values of score were used as appropriate conformational pose with Zdock score value Table 3 (see supplementary material).

Results:
The ligand structure NNK (4-(Methylnitrosamino)-1-(3-Pyridyl)-1-butanone) as a potent cigarette smoke carcinogen described in this study was retrieved from latest version of Chem Spider Chemical Database in (dot).mol format (  material). The residues ARG, ASP, LEU, LYS, TYR, VAL and ALA from all regulatory proteins (CDKs) commonly interacted with carcinogen NNK through hydrogen bonds. The functional integrity with interaction efficiency of CDKs once bound to NNK was also evaluated by using Zdock method. We obtained Zdock-scores of CDK8 with its corresponding partner cyclin C was 30.12, and of CDK2 with its corresponding partner cyclin A was 21.62. Whereas, the complex CDK8+NNK when docked with its partner cyclin C showed Zdock score 19.12, and CDK2+NNK when docked with its partner cyclin A showed Zdock score 21.34. Reduction in the Zdock score of CDK8 after it forms complex with NNK may represent a loss in its capacity to bind with partner cyclin C, once bound with NNK. Whereas, regulatory protein CDK2 doesn't show significant change in Zdock value, after binding with NNK as shown in (Table 3).

Discussion:
An insilico approaches were applied to evaluate protein-ligand and protein-protein interaction (PPI) for identifying possible targets of carcinogen (NNK) amongst cyclin dependent-kinases (CDKs) as well as with their respective partner cyclins. In order to gather in-depth knowledge on, an important issue that how the cigarette smoke carcinogen NNK interfere the mechanism of signals through CDK complex (Cdks+Cyclins) in proliferation, cell division and in abruption of cell cycle. In this study, we characterized and identified the molecular interaction of NNK with all regulatory proteins (CDK2, CDK4, CDK6, CDK7, CDK8, and CDK9) and with their respective partner cyclins (Cyclin A, Cyclin E, Cyclin D, D1, D2, D3, Cyclin C, Cyclin H, and Cyclin T) by using Autodock and Zdock (protein-protein interaction) methods. To validate our in silico study, we considered Human oxy-hemoglobin at 1.25 Å resolution, [PDB ID-1HHO] as a +ve control, which showed binding energy of -6.06 kcal/mol, as the binding of Human oxyhemoglobin with NNK has previously been quantified in tobacco users and is considered as a biochemical marker for uptake of tobacco specific nitrosamines [24]. The docking outputs indicated that potent cigarette smoke carcinogen NNK shows the binding efficiency (∆G) against cell cycle regulatory enzymes/proteins (CDKs) ranging, -3.95 to -7 (Figure 1 B & C) respectively. During the proteinligand interaction, the amino acid residues of CDK8 involved in interaction with ligand (NNK) were identified as ARG150,  ASP151, LEU152, LYS153, THR196, PHE197, TRP198, TYR199,  ARG200, ALA201, LEU204, TYR211, ILE215, and ALA219. While, in case of protein CDK2 the amino acid residues, namely ILE10, VAL18, ALA31, LYS33, VAL64, PHE80, GLU81, PHE82, LEU83, HIS84, GLN85, ASP86, LEU134, and ALA144 were found to be involved efficiently in the interaction with ligand NNK. In fact, amino acid residues, particularly ARG, ASP, LEU, LYS, TYR, VAL and ALA of both CDK8 and CDK2 were found to be essential for the interaction of carcinogen NNK (Table 5). This molecular interaction between CDKs and NNK is validated on the basis of their binding energy (∆G) obtained from best docked conformations. It has been reported that Cyclin Dependent Kinase-8 is actively involved in the regulation of mRNA transcription and considered as a potent oncogene in colon carcinogenesis. In addition, mutated or amplified CDK8 with increased expression is a common observation during a variety of human cancers [25]. However, we further explored the interaction impact of NNK on CDK8 (3RGF) in term of binding efficiency towards its respective partner cyclin C. For this purpose, we applied an in silico Zdock method for calculating Zdock score of protein-protein interaction (PPI) between CDK8 Vs Cyclin C complex and further compared it with protein complex interaction (CDK8+NNK Vs Cyclin C) by using Discovery studio 2.5. Results clearly depicted that significant loss of binding energy of CDK8 Vs Cyclin C from 30.12 (PPI) to 19.12 for NNK bound with CDK 8 complex Vs Cyclin C (PCI) at coordinates X -17.585, Y 11.939 & Z 17.689. These results predicted that binding of NNK at the active site of CDK8 strongly interferes with the natural binding of cyclin C to the active site of CDK8, rendering it unable to form functional complex (CDK8+cyclin C). It has been reported that CDK8 positively regulates transcription, by directly phosphorylating p53 and histone H3, or by facilitating assembly of Pol II elongation complex [26]. The CDK8-cyclin C complex is the part of RNA Polymerase II, which regulate the transcription of general transcription initiation factor IIH (TFIIH), controlling the basal transcription machinery. Failing the formation of CDK8-cyclin C complex may eventually will result in failure of the transcriptional regulation of a member of RNA-Pol-II dependent genes [27]. CDK8 plays significant role in regulating cell cycle progression [28] And CDK 8-cyclin C complex abnormality has many times reported to result in tumorogenesis [29]. When we performed same Zdock study with other protein CDK2, it was observed that NNK binding with the CDK2 didn't significantly interfere natural binding of its partner cyclin A Table 3. Previously QMS Jamal, et al., 2012 presented Zdock based analysis to determine the loss/incapability in formation of functional complex of regulatory proteins in DNA repair pathways after its binding with a concerned chemical/carcinogen [12]. The binding of NNK with CDK8 therefore, may be an important event in carcinogenesis caused by cigarette smoke carcinogen and should be studied in depth.

Conclusion:
In silico study explores the interaction of CDK8 (cyclin dependent kinase-8) with NNK, a widely inhaled potent cigarette smoke carcinogen among the young generation of population. CDK8, cyclin C and its complex (CDK8+cyclin C) are the key mediator of cell cycle progression which play a vital role in cell cycle perturbation due to potential interaction of cigarette smoke carcinogen NNK. The hydrogen bonds and certain amino acid residues ARG, ASP, LEU, LYS, TYR, VAL and ALA play a key role in the correct positioning of CDKs within the active site of NNK to permit docking interaction. The effect of molecular interaction of NNK on the binding with CDKs is elucidated. Study indicated the loss of functional complex of these enzymes/regulatory proteins (Cdk+Cyclin), which probably could be a reason for perturbation in cell cycle process resulting in occupied active site of CDKs by NNK. Furthermore, our study suggests that carcinogens (NNK) positively alter the mechanisms of cell cycle progression pathways and enzymes functioning could be affected by carcinogens. Computer based structural analysis of bio macromolecules and their molecular interactions (ligand and protein) could play an important role in assessment of risk to a number of diseases including cancer. At last but not least, a deep analysis is needed to elucidate the perturbation of cell cycle mechanisms with best suitable techniques and tools. In vivo and In vitro validation is needed to authenticate insilico results obtained from this study.

Conflict of Interest:
We have no conflict of interest with anybody working in the area and among the authors in the manuscript.