Structure-Based Studies of 2-Bezylidine-Benzofuran-3-One Class of Compounds as the Cyclin Dependent Kinases (CDKs) Inhibitor.

 

Sunil H. Ganatra* and. Shilpa A. Gurjar

Department of Chemistry, Institute of Science, R. T. Road, Civil Lines, Nagpur-440001, M.S., India.

ABSTRACT:

Structured based molecule design is utilized to identify potent inhibitors of a Kinase family that collectively functions as key regulators of cell cycle. Our approach uses virtual molecular screening methods to indentify and refine chemical entities that act as inhibitors and increase the success rate of discovering potent and selective molecules that can be future drugs to incurable diseases like Cancer. The active pockets of the CDK’s are unique in terms of size, shape and amino acid composition. Benzofuranone based series of molecules designed virtually and docked to CDK4 mimic CDK2 crystal structure having PDB reference number 1GII.

 

Using flexible ligand – rigid enzyme docking techniques, the binding energy values (ΔG kcal/mol) of various conformations of 2-Benzylidene-benzofuran-3-one based molecules are evaluated. The best-selected conformations and their binding energies values along with cluster results are explained in detail.

 

From these initial studies and through iterative rational drug design process, more potent, selective, and most importantly, chemically unique Substituted 2-Benzylidines-Benzofuran-3-one based compounds have been identified as an effective CDK inhibitors. 

 

 

 

INTRODUCTION:

New Classes of Compounds must be introduced into the Cancer drug development pipeline in an effort to develop new chemotherapy options to fight against cancer. Diseases like cancer are among top global diseases of the world. Anti-Cancer therapies which are given to patients are ineffective today. This is because of the side effects and high toxicity in these drugs. There is now a scramble to find new therapeutic approaches.

 

The discovery of new medicines is a very long, expensive and risky process. However recent advances in computation, chemistry and biological sciences are now yielding new paradigms that allow the researchers to model and understand drug targets and to discover drugs that are cheaper and safer than the existing ones within a short time span. The availability of three-dimensional coordinates for target protein enables the use of structure-based drug design techniques. These technologies include virtual screening, pharmacophore development and structure based optimization [1].

 

The procedure that is normally followed in this process is the building of in-silico set of small molecules (electronic chemical structures) using computer programs. Then the energy minimizations and conformational analysis is performed using advanced calculating paradigms. Finally the virtual sets of compounds are optimally positioned into the binding site of the receptor enzyme and to explore the energy of interaction with respect to individual docking poses of the molecule.

 

Rational drug design (RDD) is a process used in the biopharmaceutical industry to discover and develop new drug compounds. RDD uses a variety of computational methods to identify novel compounds, design compounds for selectivity, efficacy and safety, and develop compounds into clinical trial candidates. These methods fall into several natural categories – structure-based drug design, ligand-based drug design, de novo design and homology modelling – depending on how much information is available about drug targets and potential drug compounds. [2]

 

Structure-based drug design (SBDD) is one of several methods in the rational drug design toolbox. Drug targets are typically key molecules involved in a specific metabolic or cell signalling pathway that is known, or believed, to be related to a particular disease state. Drug targets are most often proteins and enzymes in these pathways. Drug compounds are designed to inhibit, restore or otherwise modify the structure and behaviour of disease-related proteins and enzymes [2,3].

 

SBDD uses crystal structure of proteins to assist in the development of new drug compounds. The 3D structure of protein targets is most often derived from x-ray crystallography or nuclear magnetic resonance (NMR) techniques. X-ray and NMR methods can resolve the structure of proteins to a resolution of a few angstroms. At this level of resolution, researchers can precisely examine the interactions between atoms in protein targets and atoms in potential drug compounds that bind to the proteins. This ability to work at high resolution with both proteins and drug compounds makes SBDD one of the most powerful methods in drug design. [4, 5, 6]. 

 

Targeting the Cyclin Dependent Kinases:

Cell cycle progression is regulated by a series of sequential events that include the activation and subsequent inactivation of cyclin dependent kinases (CDKs) and cyclins. CDKs are a group of Serine/Threonine kinases that form active hetero-dimeric complexes by binding to their regulatory subunits, Cyclins. Several CDKs, mainly CDK2, CDK4, and CDK6, work cooperatively to drive cells from G₁ phase into S phase. CDK4 and CDK6 are involved in early G₁ phase, whereas CDK2 is required to complete G₁ phase and initiate S phase. Both CDK4 and CDK6 form active complexes with the D type of cyclins (Cyclins D1, D2, and D3).

 

The enzymatic activity of a CDK is regulated at three levels: cyclin association, subunit phosphorylation, and association with CDK inhibitors. When cyclins initially bind to CDKs, the resulting complex is inactive. The phosphorylation of CDKs by CDK activating kinases leads to their activation [7].

 

From a therapeutic standpoint CDKs are considered promising targets in cancer chemotherapy. The most promising strategies involve designing inhibitors that either block CDK activity or prevent their interaction with cyclins. Most of the currently available molecules target the ATP-binding site of these enzymes. Such an approach might create serious problems as catalytic residues are well conserved across eukaryotic protein kinases.[7]

 

CDK’s are currently pursued as drug targets in numerous diseases that include cardiovascular, neurological disorders and cancer. There is a growing need in designing drugs that show novelty in chemical structure and mechanism for action.  2-Bezylidine-Benzofuran-3-one substituted sets of drugs are one such promising target for Kinase inhibitor type of drugs. Present study is to understand and screen the best candidates as an anti-CDKs agent. The selected class of compounds are Substituted 2-Benzylidines-Benzofuran-3-one and their interactions with CDK4 mimic CDK2 enzyme are studied using in silico techniques.

 

The aim of the present study is to understand the CDKs inhibition possibilities and strength of substituted 2-Benzylidines-Benzofuran-3-one class of compounds using molecular modelling techniques. To utilize the structure based drug design techniques, evaluate the possible confirmations of small molecules and their binding energies with CDK4 mimic CDK2 crystal structure having PDF reference number 1GII. Also to rank the best five molecules by evaluating them on the basis of higher cluster size, possibilities of hydrogen bonding and binding energies of complex.

 

MATERIAL AND METHODS:

Preparation of Receptor Enzyme:

The selected target enzyme is the crystal structure of  CDK4 mimic CDK2.  Crystal structure downloaded from online resources [8]. Its  PDB reference number is 1GII. The structure is having uniqueness by having active site of CDK4 into the structure of CDK2.   Generally all CDKs are homologues in their structures and hence an inhibitor interacts with one CDK, also interacts  withanother with slight change in its activities. But there is the need to prepare unique inhibitor, which inhibit only CDK4 and avoid others. 1GII  provides the facilities of CDK4  active site within CDK2 structure. The pocket size of CDK4 is larger and can accommodate larger molecules.[9]

The structure is checked for any missing atoms. The hetro atoms are removed which are not involved in docking process from selected enzyme. Solvation process performed followed by preparation of enzyme grid for docking process using AutoDock 4.0 [10]

 

Selection of Binding Site:

The selection of binding site is the site of Natural inhibitor. 1PU: 1-(5-Oxo-2,3,5,9b-Tetrahydro-1h-Pyrrolo[2,1- A]Isoindol-9-Yl)-3-Pyridin-2-Yl-Urea is available with 1GII and its position is selected as the active site for docking process.

Figure 1. Crystal structure of Cyclin Dependent Kinase (CDK4 mimic CDK2) PDB Ref.: 1GII [9].

Small molecule preparation and verification (Ligand preparation):

Virtual (in-silico) set of compounds or small molecular structures were created on a high processing computer using various molecular modelling environments. Computational schemes like molecular mechanics and energy minimizations are employed.  The search restricted for the low key conformations that are presumed to be biologically relevant. With the growing recognition that the pharmokinetics of drugs are of great importance, these virtual set of chemical structures were screened on the basis of the physio-chemical factors related to drug-likeness. [11,12]

 

Library of compounds were design using ChemOffice software [13] from the basic structure of lead compound, Substituted 2-Benzylidines-Benzofuran-3-one. Initially 2-D structures designed followed by 3-D design. The general structure of lead compound is depicted in figure 2. The pharmacophores were attached to the molecule at positions R1, R2, R3, R4 to prepare a series of new molecules.

 

The designed molecules were sterically modified to confirm the global minimum of the same. The global minimum energy is the main parameter to know the best stable conformation of molecule. 3-D geometry optimisations were done using QM/MM techniques [14,15]. This is semi-empirical techniques gives fairly good geometry optimisation. It was not possible to use ab initio calculation due to non-availability of programme and facilities at our laboratory. All molecules were tested for global minimum energies but only those having negative global minimum energy values were selected for further processing. Those molecules having higher values are rejected and not taken for further study [16,17]. 

 

The list of compounds designed along with various substitutions is reported in Table 1.

Figure 2. General Formula for Substituted 2-Benzylidines-Benzofuran-3-one.

The designed compounds are called as small molecules or ligand in the process of docking. The docking process is the formation of stable complex of ligand and receptor enzyme. The stability of complex is verified by calculating the docking or binding energy in the form of difference of Gibb’s free energy. (ΔG in Kcal. Mol^-1 .).

 

Docking of inhibitor molecules into the Receptor site

Binding energy calculation:

Binding affinity can be estimated by kinetic experiments that measure the inhibition of enzyme or protein in presence of both inhibitor and the substrate and is reported as inhibition constant K. The dissociation constant K_d is the ratio of the concentration of the reactants (protein (P) or Enzyme and ligand (L)) to the products complex (PL).

 

The dissociation of PL is related to the free energy of binding.

 

Where DG is the free energy change for the reaction, R is the gas constant , and T is the temperature, DG⁰  is the free energy of the reaction at standard conditions where all the concentrations are 1M, temperature is 298 K and pressure is 1atm [17].

 

Docking algorithm makes use of force field equations and parameters to calculate the binding energy. The binding free energy is the sum of intermolecular interactions between ligand and enzyme. The interactions include van der Waals, H-bond, electrostatic and steric energy of the ligand-enzyme complex. It can be represented by the equation 3.  [16,17,18].

 

Docking procedure and selected parameters:

The selected ligand and enzyme were subjected to Genetic Algorithm (GA) docking procedures. Prior to docking the selected ligand molecules were optimised by UFF methods and templates of binding sites were optimised by adding hydrogen atoms and solvation parameters. For GA algorithm, the selected parameters are listed in table 2.

 

The conformations were sorted out showing clustering’s ranked to get lowest energy conformations. All results with GA docking methods showed a similarity in RMSD value indicating that the docking methods were viable and valid.

 

 

Table 1: List of Substituted 2-Benzylidines-Benzofuran-3-one based designed molecules along with substitution groups at R1, R2, R3 and R4 substitution position along with obtained binding energy values in kcal-mol-1.
No..	R1	R2	R3	R4	Binding energy (kcal/mol)
1	-Cl	-H	-H	-H
2	-Cl	-NH2	-H	-H
3	-Cl	-N(CH3)2	-H	-H	-10.21
4	-Cl	-NH (COCH3)	-H	-H	-10.83
5	-Cl	-OH	-H	-H	-10.17
6	-Cl	-COCH3	-H	-H	-11.06
7	-Cl	-CH2CH3	-H	-H	-10.85
8	-Cl	-OCH3	-H	-H	-10.10
9	-Cl	-OH	-H	-H	-10.17
10	-Cl	OH	-H	-H	-9.82
11	-Cl	-OH	-OCH3	-H	-9.55
12	-NH2	-Cl	-H	-H	-9.62
13	-NH2	-Br	-H	-H	-9.34
14	-Cl	-(CH3)2N	-H	-H	-9.70
15	-CL	-OH	-H	-H	-10.33
16	-Cl	-C2H5	-H	-H	-10.42
17	-Cl	-OCH3	-H	-H	-9.682
18	-Cl	-OH	-H	-H	-10.15
19	-Br	-OH	-H	-H	-10.21
20	-F	-OH	-H	-H	-9.89
21	-NHC6H5	-Cl	-H	-H	-11.39
22	-NH(C6H4)Cl	-Cl	-H	-H	-11.32
23	-NH(C6H4)OH	-Cl	-H	-H	10.51
24	-NH(C5H3)OCH3	-Cl	-H	-H	-10.58
25	-NH(C6H4)C2H5	-Cl	-H	-H	-10.58
26	-NHC5H5	-Cl	-H	-OH	-10.65
27	-NHC6H5	-Cl	-H	-OCH3	-10.30
28	-NHC6H5	-Cl	-H	-C2H5	-10.76
29	-NHC6H5	-Cl	-H	-CH3	-10.42
30	-NHC6H5	-Cl	-H	-C3H7	-11.31

 

Table 2: Selected parameters for Genetic Algorithm Docking Process.
Parameter Name	Setting Value
Population Size	100
Maximum Generation	5000
Local Maximum Iteration	20
Mutation Rate	0.2
RMSD Calculation	2 A with defined rotatable bonds and active torsions set to fewest atoms.
Grid Setting	40,40,40 in X,Y,Z dimensions

 

RESULTS:

The docking results of twenty seven ligands with selected enzyme are depicted in Table 1. Though more number of molecules designed and docked with enzyme, but only those showing negative values are selected and finalized for further studies.

 

From docking results, it is observed that molecule number 21, 22 and 30 shows better docking values along with number of hydrogen bonds. These three molecules report binding energy less than -10 Kcal mol^-1. Molecule 21 shows the possibilities of one hydrogen bond with 86-ASP amino acid of enzyme, where as molecule number 22 and 30 show possibilities of 2 hydrogen bonds. 86-ASP amino acid of enzyme participates in forming the hydrogen bonding with best three molecules. 85-GLN and 84-HIS and 74-HIS amino acids also participate in making hydrogen bonds.  Figure 3 shows the docking pictures of molecule number 21. Table 3 shows the docking results of molecule number 21, 22 and 30 along with the hydrogen bonding details.

 

 

 

 

 

LogP is one criterion used in medicinal chemistry to assess the druglikeness of a given molecule, and used to calculate lipophilic efficiency The logarithm of the ratio of the concentrations of the un-ionized solute in the solvents is called logP: The logP value is also known as a measure of lipophilicity[19,20].

 

The logP parameter, which is the key parameter for drug molecules, also calculated and reported in table 4. Computationally it is possible to calculate this parameter using Crippen’s fragmentation and Viswanathan’s fragmentation methods [20].  It is reported that this parameter show values lower than 5 for the selected molecules (mol no. 21,22 and 30).

 

Table 4. LogP values of selected molecules
Mol. No.	Chemical formula	LogP by Crippen's fragmentation	LogP by Viswanathans fragmentation
21	C21H14ClNO2	4.59	4.21
22	C21H13Cl2N2	5.15	4.73
30	C24H20ClNO2	5.91	5.47

 

DISCUSSION:

The introduction of amino substituted group at position R2 seen to increase the binding energy but, introduction of -OH and -OCH₃ does not seem to allow the benzofuranone moiety to be favourably bound to the protein. On the Other hand, introduction of chlorine atom makes favourable hydrophobic contact with the protein.

 

In particular, the carbonyl group (C=O) at the 4^th position of benzofuranone is found to interact with the binding site of cyclin dependent kinase with amino acids 86 ASP and 84 HIS of the ATP binding site. The hydroxy-phenyl at the 5’ and 2’ positions actively participate in binding and establish hydrogen bonds with the backbone of CDK protein. They bind to the amino acids not previously in binding with the original inhibitor and thereby increase the binding energy of the CDK protein. Similarly the chlorine atom at the 7’ position and the hydroxy-phenyl analogues seem to add to the hydrophobic effect and increase the interactions with the selected protein.

 

As justified by the inhibition data of compounds, targeting ASP 86, HIS 84 and VAL 83 for hydrogen bond interactions witha introduction of an acceptor group substituent was a good strategy to increase inhibition activity. The increase in activity by changing the substituent from halogens (-Cl, -Br) to secondary amines (-NHR) with bulky groups seem to favour the stability of the molecule and hence the higher binding values are obtained. Finally, Compounds derived from the general structure of 2-Bezylidine-Benzofuran-3-ones as kinase inhibitors showed significant inhibition of Cyclin dependent kinase 2 complexes with CDK4 inhibitor. Following structure based approach, it was possible to improve the inhibitory activity of the virtual set of compounds against CDKs and hence obtain selectivity against CDK4.

 

The reverse isn’t possible since the nature of CDK4 ATP binding site that offers fewer opportunities than those of CDK2.

 

CONCLUSION:

The docking studies of substituted 2-Benzylidine-Benzofuran-3-one class of compounds give negative binding energies. The results show that they must be excellent CDKs inhibitors, particularly CDK4 due to its larger cavities. It is already known that those molecules which inhibit one of the CDK enzymes must also inhibit another one due to structure homology among various CDKs [9]. CDK4 have larger cavities then other members of this class of proteins. The selected class of ligand are also having larger structure and fits into the CDK4 mimic CDK2 active site. i.e. into the crystal structure of 1GII.  The molecule number 6,21,22 and 30 show better binding energies and also show the possibilities of hydrogen bonding give stabilities to their complexes with crystal structure of CDK4 mimic CDK2 enzyme.

 

This study supports the further processing of these classes of molecules as anti-cancer agents.

 

ACKNOWLEDGEMENTS:

We sincerely acknowledge the fund support by U.G.C, New Delhi. (Research Award Scheme Xth Plan.)  to Dr. Sunil H. Ganatra

 

We would like to thank Dr. B. J Ghiya, Emeritus Scientist and Ex-Reader, Department of Chemistry, Institute of Science, Nagpur for fruitful scientific discussions.

 

REFERENCES:

1.        Susan M Keeman , Jeanne A.Geyer, William J. Welsh, Sean T. Prigge,  and Norman C. Waters., Combinatorial Chem. and high throughput Screening, 2005, 8, 27-38.

2.        W. Graham Richards., Computer aided drug desing, Pure and Appl Chem., 1994 (66)8, pp. 1589-1596,.

3.        Joseph Scoepfer, Heinz Fretz, Bhabatosh Chaudhuri et. al.,, Structure-Based Design and Synthesis of 2-Benzylidene-benzofuran-3-ones as Flavopiridol Mimics”, Journal of Medicinal Chemistry,2001,pp 1741–1747. DOI: 10.1021/jm0108348

4.        Reviews of QM/MM are J.Rao, “Reviews in Computational Chemistry Volume 7” K.B. Lipkowitz, D.B. Boyd, Eds., 119, VCH, New York (1992).

5.        A.Warshel, “Computer Modelling of Chemical Reactions in Enzymes and Solutions” Wiley, New York (1991)

6.        E.Clementi, “Computational Aspects for Large Chemical Systems” Springer, New York (1980).

7.        Loyer P, Trembley J, Katona R, Kidd V, Lahti J., Role of CDK/cyclin complexes in  transcription and RNA splicing. Cell Signal, 2005, 17 (9): 1033-51.

8.        H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig, I.N. Shindyalov, P.E. Bourne (2000) The Protein Data Bank,Nucleic Acids Research, 28: 235-242. (http://www.rcsb.org)

9.        Ikuta, M.,  Kamata, K. et.al., Crystallographic approach to identification of cyclin-dependent kinase 4 (CDK4)-specific inhibitors by using CDK4 mimic CDK2 protein, J.Biol.Chem. (2001)276: 27548-27554, DOI:10.2210/pdb1gii/pdb, [PubMed: 11335721] 

10.     Morris, G. M., Goodsell, D. S., Halliday, R.S., Huey, R., Hart, W. E., Belew, R. K. and Olson, A. J. (1998), Automated Docking Using a Lamarckian Genetic Algorithm and and Empirical Binding Free Energy Function J. Computational Chemistry, 19: 1639-1662.

11.     G.H. Grant, W.G. Richards “Computational Chemistry” Oxford (1995).

12.     A.R.Leach “Molecular Modelling Principals and Applications” Addison Wesley Longman (1996)

13.     Chemdraw Software, Cambridge Corporation, Cambridge M.A. 02140 U.S.A.

14.     D.A.MacQuarrie “Quantum chemistry” University Science Books(1983)

15.     C.Cohen-Tannoudji, B.Diu, F.Laloe “Quantum Mechanics Volumes I and II” Wliey-Interscience (1997).

16.     F.Jensen “Introduction to Computational Chemistry” John Wiley andSons (1999).

17.     Annual Reports in Computational Chemistry, American Chemical Society (2005), Volume 1.

18.     Kaushik Raha and Kenneth.M.Merz Jr., Calculating Binding Free Energy in Protein-Ligand Interactions, . Department of Chemistry, The Pennsylvania State University, University Park, PA 16802 USA.

19.     Edwards MP, Price DA (2010). "Role of Physicochemical Properties and Ligand Lipophilicity Efficiency in Addressing Drug Safety Risks". Annual Reports in Medicinal Chemistry. Annual Reports in Medicinal Chemistry 45: 381–391. doi:10.1016/S0065-7743(10)45023-X. ISBN 9780123809025.

20.      Leeson PD, Springthorpe B (November 2007). "The influence of drug-like concepts on decision-making in medicinal chemistry". Nat Rev Drug Discov 6 (11): 881–90. doi:10.1038/nrd2445. [PMID 17971784]

 

 

 

 

Received on 05.05.2012        Modified on 20.05.2012

Accepted on 24.05.2012        © AJRC All right reserved