INTRODUCTION:

Cancer has become the major cause of death in both developing and developed countries (Sun et al. 2013). Breast cancer is one of the leading cancers worldwide in women; India has highest breast cancer related mortality.The human genome encodes approximately 500 predicted protein kinases (PKs), many of them are participating in signal transduction pathways that regulate cell growth and survival, suggesting potential roles in cancer initiation and progression (Ibrahim and Ismail 2011) so inhibition of the CDK4 can results in potential anticancer activity (Ghosal et al. 2010). PKs, cyclin-dependent kinases, a family of serine/ threonine PKs, attract special attention owing to the close association of their over expression or alteration with carcinomas (Li, et. al. 2013).Although mutations of CDK4 have been described in some tumors more frequent is inactivation of endogenous CDK inhibitors (Doleckova et al. 2013).Substitution of the deficit of endogenous CDKIs by pharmacological counterparts provides a very promising therapeutic option to curb proliferative disorders of malignant cells when they escape from the proper control of the cell cycle in cancers (Li et al. 2013).

Design and synthesis of compounds with interactions with specific target proteins and no interactions with other protein targets is the need of time.However, the new drugs are still far from desired objectives and hence development of newer agents with properties closer to that of an ideal anticancer remains a challenge. QSAR is more helpful to correlate physicochemical properties of molecules to their biological activities (Choudhari et al 2011). 3D-QSAR uses probe-based sampling within a molecular lattice to determine three-dimensional properties of molecules mainly steric and electrostatic values and can then correlate these 3D descriptors with biological activity. Pharmacophore is the three dimensional geometry of interaction features that a molecule must have in order to bind in proteins active site(Young 2009). The various types of interaction centers are involved in that mainly hydrogen bond acceptors and donors, positive charge centres, hydrophobic centres and aromatic ring centers. Here, we report the 3D QSAR and pharmacophore modeling analysis of some novel CDK4 inhibitors which would give helpful structural requirements for development of new series of CDK4 inhibitors.

MATERIALS AND METHODS:

Experimental protocols

Selection of data set

Synthesized and screened data set for present study as per literature reported by Breault et al. (2002).

QSAR analysis

Ligand preparation

The structure of N-phenylpyrimidin-2-aminewas used as the template to build the molecules in the builder module on VlifeMDS 4.3 software. The ligand geometries were optimized by energy minimization using MMFF94 force field and Gasteiger-Marsili charges for the atoms, till a gradient of 0.001 kcal/mol/A° was reached, maintaining the template structure rigid during the minimization.

Molecular alignment

The template based technique used for alignment of the dataset molecules. Template for alignment was selected on basis of the most active molecule and the alignment of all the molecules is shown in Fig. 1.

Fig. 1 Alignment of molecules using the most active compound template

Descriptor calculation

The suitable alignment of the given dataset molecules was performed using the Vlife MDS4.3 software for 3D QSAR. The hydrophillic, steric, and electrostatic interaction energies were computed at the lattice points of the grid using a methyl probe of charge +1. These interaction energy values were utilized as values of the 3D-descriptors for generation of correlation with biological activity. The term descriptor is utilized in the following discussion to indicate field values at the lattice points.

Data set

The dataset was assorted into a training set (16 Molecules, Table 1) and a test set (7 molecules, Table 1) using random selection on the basis of chemical and biological diversity. For the present 3D-QSAR study used biological activity as inhibitory concentration (IC₅₀) values for CDK4 inhibition.

Full search multiple linear regression method

By using regression analysis determined relationship between independent and dependent variables (3D fields and biological activities, respectively). Describe the fitness of data by using a regression model, r². Those models which having correlation coefficient, q² above 0.8 were used to check the external predictability while the significance of the model was assessed by F value. Those models showing q²below 0.7 were discarded. For CDK4 inhibitory activities selected models are shown in Table 2.

Table 1 Table showing molecules which are utilized in present study

Sr No.	R	X	Y	Sr No.	R	X	Y
1	H	All H	H	13	H	2-Cl, 5-CH₃	H
2	H	2-Cl	H	14	H	2-F, 5-CF₃	H
3	H	2-F	H	15	CH₂CN	2-Cl, 5-Cl	H
4	H	2-Br	H	16	CH₂CCN	2-Cl, 5-Cl	H
5	H	2-CN	H	17	CH₂CH₂CN	2-Cl, 5-Cl	H
6	H	2-OCH₃	H	18	CH₂Ph	2-Cl, 5-Cl	H
7	H	3-Cl	H	19	(CH₂)₃CF₃	2-Cl, 5-Cl	H
8	H	4-Cl	H	20	H	All H	CH₃
9	H	4-OCH₃	H	21	H	All H	F
10	H	4-CH₃	H	22	H	All H	Cl
11	H	2-Cl, 5-Cl	H	23	H	All H	Br
12	H	2-F, 5-CH₃	H

Table 2 Table showing the selected QSAR equation along with statistical parameters employed for model selection

Model	QSAR model	N	r²	q²	F value	Pred r²
A	pIC₅₀= 0.5948(±0.0211) S_1119+0.2674(±0.0979) E_1106-109.4730(±55.7753) S_938-0.0077	23	0.9853	0.9504	267.4822	0.7967
B	pIC50= 5.4473(±0.1800) S_1011+ 2.9675(±0.3704) H_888 -1.0176(±0.3971) E_722 + 0.0212	23	0.9875	0.8231	316.2329	0.6739

Table 3 Table showing the observed activity and predicted activity of molecules

Sr. No.	Observed activity	Predicted activity	Residuals	Sr. No.	Observed activity	Predicted activity	Residuals
1	2	1.2	0.7	13	2	1.5	0.5
2	1	1.1	-0.1	14	0.6	1.2	-0.6
3	1	1.5	-0.5	15	0.1	1.2	-1.1
4	1	1.1	-0.1	16	0.2	1.2	0
5	2	1.0	0.9	17	0.3	0.2	0.1
6	19	19	0	18	1	0.8	0.2
7	2	1.6	0.4	19	0.2	0.5	-0.3
8	1	1	0	20	0.8	1.7	-0.9
9	2	1.5	0.5	21	0.3	0.8	-0.5
10	2	1.7	0.3	22	0.1	0.1	0
11	0.8	1	-0.2	23	0.1	0.4	-0.3
12	0.7	0.1	0.6

Activity prediction

A reliable validation is required for QSAR model. Evaluation of the predictive results for the given dataset is done by QSAR model. Select those models which having r²above 0.8 and checked for their external predictive ability measured as predicted r². The observed and the predicted values for CDK4 inhibition are shown in Table 3.

Pharmacophore modeling

First align the most active molecule as template from series of CDK4 inhibitors. The minimum number of pharmacophore features generated for an alignment was taken as 4 and the tolerance for the distance separating two features was kept 10 A⁰.

RESULTS:

In the present study, 16 molecules were used in the training set (Table 1) and 7 molecules in the test set(Table 1) to derive 3D QSAR models with the number of field grid points being not more than seven per model. The electrostatic and steric fields were calculated using the Tripos force field and Gasteiger-Marsili charges. To evaluate the predictive ability of generated 3D QSAR models with regularly distributed biological activities. By the successful runs of MLR (Multiple liner regression), different sets of equations were generated and these equations were analyzed statistically to select the best QSAR model. After screening various combinations of different descriptors, two models were selected which shown in Table 2. The models were selected on the basis of r², q², pred r², F and P values (Table 2).

DISCUSSION:

Interpretation of QSAR model:

The QSAR model was selected as best model among all developed models to represent the optimum structural requirements of pyrimidine analogs to act as CDK 4 Inhibitors (Fig. 2and3 ).Optimization of steric and electrostatic properties in selected QSAR model will lead to an increase or decrease in binding affinities and selectivity’s. The contributing descriptor in the QSAR model A are the steric interactions at lattice point S_1119, S_938 and electrostatic interaction at lattice point E_1106. The steric interaction at lattice point S_1119 is positively contributing to activity, so substitution of bulky R-substituent’s at the aromatic ring could show increased activity. The steric interactions along the lattice point S938contribute negatively substation of aliphatic group will lead to increase in activity. Electrostatic interactions at the lattice point E1106is contributing positively hence substitution of electron donating substituent-R at the ortho position of aromatic ring could increase activity. Correlation between the observed biological activity and activity predicted by the 3D QSAR model (Table 3) indicate that each of the selected 3D descriptors has appropriate weight age in the selected 3D QSAR equation representing the correlation of biological activity with these descriptors.

Fig. 2 Field points of selected QSAR model A

Fig. 3: Correlation graph for selected QSAR model A

Pharmacophore identication Studies:

A pharmacophore model is the three-dimensional geometry of interaction features that are necessary for bioactive ligands. Logically, an align molecules based on features that are responsible for bioactivity of the various CDK4 inhibitors. A pharmacophore model for coagulation CDK4 inhibitor was generated using the mole sign module of Vlife 4.3. The hypothesis contains characteristic features like hydrogen bond donor (magenta color), hydrogen bond acceptor (buff color), hydrophobic, and aliphatic regions of the structure (orange color) (Fig. 4).

Fig. 4: Pharmacophore model for molecules under study

CONCLUSION:

Computational approach involving 3D QSAR was employed to identify molecular structural features which required for effective inhibition of CDK4, in an aim to discover new leads for treatment and prevention of cancers. For the prediction of highly pharmacophore model used16 training set compounds, which consists of hydrogen bond donor and acceptor, aromatic groups and hydrophilic groups. The utility of our pharmacophore model is to predict CDK4 inhibitory activity. Thus, the design and development of lead molecules on the basis of data obtained from this 3D QSAR is likely to yield potent compounds.

ACKNOWLEDGEMENT:

Authors are sincerely thankful to Dr. H. N. More, Principal, Bharati Vidyapeeth College of Pharmacy, Kolhapur for creating driving force in us to carry out something helpful for new research.

REFERENCES:

1. Breault GA, Ellston RP, Green S, Pease JE, Tucker JA. Cyclin-dependent kinase 4 inhibitors as a treatment for cancer. Part 2: identification and oprtimization of substituted 2,4-bis aniline pyrimidines. Bio organic Medicinal Chemistry Letter. 13; 2002: 2961-2966.

2. Choudhari PB, Bhatia MS. 3D QSAR, docking studies and pharmacophore modeling of selected factor Xa inhibitors. Medicinal Chemistry Research. Doi: 10.1007/s00044-011-9663-8; 2011.

3. Doleckova I, Cesnek M, Dracinsky M, Brynda J, Voller J, Janeba Z. Synthesis and biological evaluation of guanidino analogues of roscovitine. European Journal of Medicinal Chemistry. 62; 2013: 443-452.

4. Ghoshal N, Mascarenhas MN, Bhattacharyya D. Why pyridine containing pyrido[2,3-d]pyrimidin-7-ones selectively inhibit CDK4 than CDK2: Insights from molecular dynamics simulation. Journal of Molecular Graphics and Modelling. 28; 2010: 695–706.

5. Ibrahim D, Ismail N. Design, synthesis and biological study of novel pyrido[2,3-d]pyrimidine as anti-proliferative CDK2 inhibitors. European Journal of Medicinal Chemistry. 46; 2011: 5825-5832.

6. Li G, Jung K, Lee H, Son M, Seo J, Hong S, Jeong Y. A novel imidazopyridine derivative, HS-106, induces apoptosis of breast cancer cells and represses angiogenesis by targeting the PI3K/mTOR pathway. Cancer Letter. 329; 2013: 59-67.

7. Sun j, Lv X, Qiu H, Wang Y, Du Q, Li D, Yang Y, Zhu H. Synthesis, biological evaluation and molecular docking studies of pyrazole derivatives coupling with a thiourea moiety as novel CDKs inhibitors. European Journal of Medicinal Chemistry. 68; 2013: 1-9.

8. Young DC. Computational drug design- A guide for computational and medicinal chemist. John Wiley and Sons, Inc., Hoboken, New Jersey. 2009.

Received on 16.03.2015 Modified on 04.04.2015

Asian J. Research Chem 8(4): April 2015; Page 231-235

DOI: 10.5958/0974-4150.2015.00040.1