Computational Modelling of Pyridopyrimidine Scaffolds; Unlocking New Successors as Tyrosine Kinase Inhibitors

Sreeja S., Vani V., A. B. Abin, S. R. Anchu, V. S. Athira, L. Venkitachalam, K. Nandhini,

doi:10.5281/zenodo.17492359

Research Paper | Open Access
Volume 03 | Issue 10 | Article Id IJPS/250310405

Computational Modelling of Pyridopyrimidine Scaffolds; Unlocking New Successors as Tyrosine Kinase Inhibitors
Sreeja S.* Vani V. A. B. Abin S. R. Anchu V. S. Athira L. Venkitachalam K. Nandhini
Pharmaceutical Chemistry, Mar Dioscorus College of Pharmacy.

Abstract

Background: The main aim of the research is to carry out Insilico design of pyridopyrimidine scaffolds and assess their binding affinity along with its toxicity evaluation. Methods: Insilico design was performed using various computational tools like Chem sketch, Molinspiration and PASS. Binding affinity against the receptor tyrosine kinase was determined by Auto Dock Vina and their toxicity evaluation by Swiss ADME/ ADME Tlab 2.0. Result: We designed seventy-five ligands having pyridopyrimidine nucleus that shows binding affinity against the tyrosine kinase receptor (PDB id: 1QCF). Among the designed derivatives, ten derivatives showed more binding affinity and less toxicity as compared to standard drug piritrexim. Conclusion: Among the ten designed derivatives, pyridopyrimidine with benzopyran moiety can be proposed as promising lead for tyrosine kinase inhibitor.

Keywords

Pyridopyrimidine, Tyrosine kinase, Anticancer, Molecular docking, Piritrexim

Introduction

Cancer has continued to occupy the top position as the leading cause of mortality and morbidity worldwide. Cancer is characterized by uncontrolled proliferation [1], invasion, and metastasis of cells. There are many advances in the medical treatment of cancer; however, hepatotoxicity, liver damage, skin sensitization, drug resistance, carcinogenicities are some adverse effects. Depending on genetics and lifestyle, the rate at which cancers can occur holds great relevance. Tyrosine Kinase is an enzyme that act by phosphorylation of tyrosine residue on their substrate proteins. This mismanagement in the tyrosine kinase pathway results in cancer development and progression. Errors involved in such processes are gene mutation, gene amplification, chromosomal translocation, and autocrine signalling. Tyrosine kinase inhibitors prevent the tyrosine kinase receptors from signalling. A few examples of tyrosine kinase inhibitors are Piritrexim, Erlotinib, Trastuzumab, and Sunitinib.

Among the heterocyclic compounds, pyridopyrimidine show versatile activity [2,3] in cancer treatment due to its structural analogue similar to purines enabling them to interact with the tyrosine kinase receptors. In particular, their role as kinase inhibitors has positioned them as potential therapeutic agents for various malignancies [4]. By examining the structural activity relationship pyridopyrimidine have reactive sites in the 2^nd& 4^th position which help in its biological action. Pyridopyrimidine derivatives show therapeutic activity such as [2,3] anti-cancer, anti-microbial, anti-viral and analgesic. Pyridopyrimidine was synthesized by reaction of aldehydes with cyanoacetyl-pyrimidinone in DMF under heating and nano catalysis conditions. Pyridopyrimidines is selected as lead nucleus and developed into suitable tyrosine kinase inhibitors scaffolds by computational molecular modelling and the potent compounds which are most suitable for binding to the receptors were selected by means of docking in terms of docking energy. The selected compounds ADMET evaluation was performed and compared with the standard drug piritrexim.

2.MATERIALS AND METHODS

A number of computational chemistry software was used to conduct an in silico screening [5] of all the proposed structures of novel pyridopyrimidine derivatives such as ACD LAB/Chem Sketch 12.0, Molinspiration, PASS and AutoDockVina.

Table2.1: Software used in In silico design

Software Used	Usage
ACD/Chemsketch	Drawing
Novopro	Converting SDF from PDBQT
AutoDock	Docking
PyMol	Visualising
SwissADME/ ADMETlab2.0	Pharmacokinetic & Toxicity evaluation

2.1ACD/CHEMSKETCH

Drawing chemicals structures, including organics, organometallics, polymers, and Markush structures, is possible with ACD/ChemSketch. Molecular weight, density, molar refractivity, 2D and 3D structure viewing, functionality for naming structures, finding of log P and certain other molecular properties calculation features are included in this software.

2.2 MOLINSPIRATION

Molinspiration is a research organization that is independent and focuses on developing and implementing modern computational techniques, particularly in relation to the internet. It offers a comprehensive set of cheminformatics software applications that aid in molecule manipulation and processing, which encompasses SMILES and SD file conversion, molecule normalization, tautomer generation, molecular fragmentation, calculation of various molecular properties needed QSAR, molecular modelling and drug design, high quality molecule depiction, molecular database tools supporting substructure search or similarity and pharmacophore similarity search. All the designed compounds were then subjected to Lipinski rule is also known as Pfizer rule of five/ Lipinski’s rule of 5.

The Lipinski rule of five states that an orally active drug should obey:

Not more than 5 hydrogen bond donors
Not more than 10 hydrogen bond acceptors
An octanol-water partition coefficient log P not greater than 5
Molecular weight not more than 500 Dalton
Not more than 5 rotatable bonds.

2.3 PASS

A computer program called PASS can be used to estimate sample size or determine the power of a statistical test or confidence interval. NCSS LLC is responsible for the production of PASS. PASS encompasses more than 920 documented sample size and power procedures. Clinical trials, pharmaceutical, medical, and many other research areas rely on PASS as the leading sample size software. An estimated biological activity profile can be obtained by using the structural formula of a drug-like substance as an input. The likelihood of being active is called 'Pa' while the likelihood of being inactive is called 'Pi'. Pa and Pi values are not connected and can range between zero and one.

2.4 NOVOPRO

Conversion of SMILES to 3D structures is possible using online tools provided by NovoPro Biosciences. You have the option to download the 3D structure in any desired format, such as.pdb. Mole, Sdf. The SMILES are a method to convert various derivatives and compute their 3D structures.

2.5 RCSB PROTEIN DATA BANK

A database that stores 3D structure data is known as the protein databank. The primary purpose of it is to acquire information about large biological molecules such as proteins and nucleic acids.

2.6 AUTODOCK

AutoDock [6] is a program designed to simulate molecular modeling. Protein-ligand docking can be achieved effectively with it. The GNU General Public License grants access to AutoDock 4. One of the most popular methods for docking ligands to set grids that describe target proteins is this. Auto Grid has the ability to pre-calculate these grids. Preparation of both proteins and ligands is necessary before docking.

2.7 PYMOL

PyMol [6] is both an open source and a proprietary system for analysing molecular data. By using PyMol, it is possible to obtain high-quality 3D images of small and biological molecules. PyMol used to visualize the images of the docked complex.

2.8 SWISSADME

Swiss ADME is a software to compute pharmacokinetic properties, physiochemical properties and toxicity of the micromolecules. Swiss ADME web tool gives free access to Absorption, Distribution, Metabolism and Elimination. Boiled egg is obtained from this software and gives idea about where drug is absorbed from.

2.9 ADMETLAB2.0

ADMET is a software tool used for predicting the toxicity of the compounds. It provides predictions according to severity denoted by plus (+) if present and negative (-) if absent.

3.EXPERIMENTAL

From the literature reviews, we came across the lead nucleus pyridopyrimidine. To the lead, we attached suitable linkers and made over 77 derivatives using Insilico computational tools. The drug likeness property was verified through molinspiration. Two ligands that showed violation were casted out and remaining were selected to predict the activity. Target for our study was identified from PDB and utilized for docking. The target protein taken was tyrosine kinase (1QCF) which had a resolution of 2.00 A°, R-value free of 0.257, R-value work of 0.215. This 1QCF is a Homo sapiens protein. Docked complex are formed from target and ligands which are evaluated for binding affinity. Among the 75 derivatives, ten derivatives were selected based on best docking score and compared with the standard. Toxicity evaluation was carried out for these ten derivatives by using SwissADME and ADMET2.0. Finally, the suitable ligand is selected on the basis of docking score and ADMET evaluation.

4.RESULTS

Table 4.1: Drug likeness property calculated using Molinspiration

SL NO.	Mol.Wt<500	nHAcc<5	nHDon<10	Log P <5	Nrotb<5	nViolations
1	305.30	7	1	1.64	3	0
2	305.30	7	1	1.49	3	0
3	272.31	7	2	-0.67	3	0
4	255.24	7	1	0.14	3	0
5	255.24	7	1	0.14	3	0
6	303.32	6	2	1.88	3	0
7	254.25	7	1	-0.32	3	0
8	305.30	7	1	1.49	3	0
9	305.30	7	1	1.77	3	0
10	272.31	7	2	-0.54	3	0
11	255.24	7	1	0.27	3	0
12	255.24	7	1	0.27	3	0
13	254.25	7	1	-0.18	3	0
14	253.26	6	2	0.77	3	0
15	253.26	6	2	0.57	3	0
16	271.31	6	1	0.78	3	0
17	271.31	6	1	0.78	3	0
18	271.31	6	1	0.78	3	0
19	254.25	6	1	0.87	3	0
20	254.25	6	1	0.67	3	0
21	270.32	5	1	1.31	3	0
22	270.32	5	1	1.51	3	0
23	255.24	8	2	-0.30	3	0
24	255.24	8	2	-0.50	3	0
25	255.24	8	2	-0.50	3	0
26	255.26	6	2	0.90	3	0
27	253.26	6	2	0.70	3	0
28	271.31	6	1	0.91	3	0
29	271.31	6	1	0.91	3	0
30	271.31	6	1	0.91	3	0
31	254.25	6	1	1	3	0
32	254.25	6	1	0.81	3	0
33	270.32	5	1	1.45	3	0
34	270.32	5	1	1.64	3	0
35	255.24	8	2	-0.17	3	0
36	255.24	8	2	-0.37	3	0
37	255.24	8	2	-0.37	3	0
38	256.27	7	2	-0.24	3	0
39	256.27	7	2	-0.16	3	0
40	254.25	7	2	0.04	3	0
41	255.24	7	1	0.33	3	0
42	255.24	7	1	0.33	3	0
43	304.31	7	2	1.54	3	0
44	320.38	5	1	2.62	3	0
45	320.38	5	1	2.62	3	0
46	356.27	7	2	-0.11	3	0
47	256.27	7	2	-0.02	3	0
48	254.25	7	2	0.17	3	0
49	255.24	7	1	0.47	3	0
50	255.24	7	1	0.47	3	0
51	304.31	7	2	1.67	3	0
52	320.38	5	1	2.75	3	0
53	320.38	5	1	2.80	3	0
54	304.31	6	1	2.17	3	0
55	315.34	6	1	2.18	3	0
56	322.32	7	1	1.73	3	0
57	265.28	6	1	0.83	3	0
58	265.28	6	1	0.44	3	0
59	266.26	7	1	0.12	3	0
60	304.31	6	1	2.31	3	0
61	315.31	6	1	2.21	3	0
62	332.32	7	1	1.86	3	0
63	265.28	6	1	0.96	3	0
64	265.28	6	1	0.57	3	0
65	266.26	7	1	0.26	3	0
66	271.32	6	2	0.70	3	0
67	332.32	7	1	1.73	3	0
68	268.28	6	1	0.68	3	0
69	284.34	5	1	1.22	3	0
70	266.26	7	1	-0.08	3	0
71	271.32	6	2	0.83	3	0
72	332.32	7	1	1.86	3	0
73	268.28	6	1	0.76	3	0
74	284.34	5	1	1.22	3	0
75	266.26	7	1	0.06	3	0

Table 4.2: Prediction of biological activity using Pa & Pi

Compounds	Pass Values
Compounds	Pa	Pa
1	0,213	0,213
2	0,306	0,306
3	0,356	0,356
4	0,269	0,269
5	0,179	0,179
6	0,180	0,180
7	0,213	0,213
8	0,457	0,457
9	0,185	0,185
10	0,157	0,157
11	0,316	0,316
12	0,265	0,265
13	0,322	0,322
14	0,208	0,208
15	0,153	0,153
16	0,138	0,138
17	0,476	0,476
18	0,222	0,222
19	0,191	0,191
20	0,177	0,177
21	0,128	0,128
22	0,230	0,230
23	0,176	0,176
24	0,179	0,179
25	0,130	0,130
26	0,288	0,288
27	0,123	0,123
28	0,218	0,218
29	0,535	0,535
30	0,144	0,144
31	0,147	0,147
32	0,215	0,215
33	0,188	0,188
34	0,123	0,123
35	0,123	0,123
36	0,113	0,113
37	0,146	0,146
38	0,165	0,165
39	0,193	0,193
40	0,113	0,113
41	0,204	0,204
42	0,187	0,187
43	0,227	0,227
44	0,135	0,135
45	0,133	0,133
46	0,135	0,135
47	0,127	0,127
48	0,163	0,163
49	0,137	0,137
50	0,260	0,260
51	0,196	0,196
52	0,116	0,116
53	0,188	0,188
54	0,116	0,116
55	0,205	0,205
56	0,219	0,219
57	0,190	0,190
58	0,159	0,159
59	0,164	0,164
60	0,168	0,168
61	0,312	0,312
62	0,115	0,115
63	0,243	0,243
64	0,242	0,242
65	0,240	0,240
66	0,287	0,287
67	0,505	0,505
68	0,147	0,147
69	0,154	0,154
70	0,420	0,420
71	0,093	0,093
72	0,545	0,545
73	0,120	0,120
74	0,165	0,165
75	0,365	0,365

Table 4.3: Docking score of designed ligand with respective protein

Sl. No	Ligands	Receptor	Docking Score
1	1	1QCF	-9.1
2	2	1QCF	-8.7
3	3	1QCF	-7.7
4	4	1QCF	-7.7
5	5	1QCF	-7.6
6	6	1QCF	-9.1
7	7	1QCF	-7.5
8	8	1QCF	-7.6
9	9	1QCF	-8.6
10	10	1QCF	-8.8
11	11	1QCF	-7.3
12	12	1QCF	-8.3
13	13	1QCF	-7.5
14	14	1QCF	-8.1
15	15	1QCF	-7.6
16	16	1QCF	-7.6
17	17	1QCF	-7.6
18	18	1QCF	-7.6
19	19	1QCF	-7.7
20	20	1QCF	-7.5
21	21	1QCF	-7.4
22	22	1QCF	-7.4
23	23	1QCF	-7.7
24	24	1QCF	-7.9
25	25	1QCF	-7.9
26	26	1QCF	-7.7
27	27	1QCF	-7.5
28	28	1QCF	-7.5
29	29	1QCF	-7.6
30	30	1QCF	-7.9
31	31	1QCF	-7.7
32	32	1QCF	-7.5
33	33	1QCF	-7.3
34	34	1QCF	-7.1
35	35	1QCF	-8.1
36	36	1QCF	-8.2
37	37	1QCF	-8.2
38	38	1QCF	-7.8
39	39	1QCF	-8.1
40	40	1QCF	-7.7
41	41	1QCF	-7.6
42	42	1QCF	-7.8
43	43	1QCF	-9.0
44	44	1QCF	-8.3
45	45	1QCF	-8.5
46	46	1QCF	-7.8
47	47	1QCF	-7.8
48	48	1QCF	-8.1
49	49	1QCF	-8.2
50	50	1QCF	-7.7
51	51	1QCF	-8.5
52	52	1QCF	-8.8
53	53	1QCF	-8.1
54	54	1QCF	-8.9
55	55	1QCF	-9.1
56	56	1QCF	-9.3
57	57	1QCF	-8.2
58	58	1QCF	-8.1
59	59	1QCF	-8.2
60	60	1QCF	-8.5
61	61	1QCF	-8.7
62	62	1QCF	-8.7
63	63	1QCF	-7.9
64	64	1QCF	-7.8
65	65	1QCF	-7.8
66	66	1QCF	-8.1
67	67	1QCF	-9.3
68	68	1QCF	-8.3
69	69	1QCF	-7.8
70	70	1QCF	-8.0
71	71	1QCF	-7.7
72	72	1QCF	-9.4
73	73	1QCF	-7.7
74	74	1QCF	-7.8
75	75	1QCF	-7.9
	Standard [piritrixim]	1QCF	-7.7

Table 4.4: docked images of the selected 10 compounds

Compound

Structure

Docking Score

Images Of Docked Complex

-9.1

-9.1

-8.8

-9.0

-8.8

-8.9

-9.1

-9.3

-9.3

-9.4

Piritrexim

-7.7

Compound 10

Fig 4.1: Absorption of compound 10 is from GIT

Compound 67

Fig 4.2: Absorption of compound 67 is from GIT

Table 4.5: Toxicity evaluation of derived 10 compounds

Compound No	Hepatotoxicity	Drug Induced Liver Injury	Respirative Toxicity	Carcinogenicity
1	++	+++	+	+
6	++	+++	+++	---
10	+	+++	++	-
43	+	+++	+++	--
52	++	+++	+++	--
54	+	+++	++	+
55	++	+++	+++	-
56	++	+++	++	-
67	-	+++	++	+
72	++	+++	+++	--
piritrexim	+++	+++	++	++

5. DISCUSSION

The above table 4.1 shows molecular weight, no. of hydrogen bond donors, no. of hydrogen bond acceptors, Log P, no. of rotational bonds and no. of violations the of the designed 75 ligands that was determined using Molinspiration. All these molecules obey Lipinski rule of five. The smiles of designed ligand were given in the pass to predict the biological activity. PASS software predicted that these derivatives have anti-cancer activity [6,7]. Pa and Pi values were denoted in table 4.2. Docking was carried out in software AutodockVina [6,8] to obtain the docking score. Docking score predicts the binding affinity of the ligand with the receptor, the docking score are used to evaluate the strength of interaction between ligand and receptor. Scores of 75 compounds are given above and from that 45 compounds show better docking score, 9 compounds showed similar docking score and 21 compounds showed poor docking score than standard in table 4.3. Docked complex is the predicted structure of protein-ligand complex where the ligand is bound to the protein in specific orientation and conformation. From the designed 75 compounds top 10 compounds which showed better docking score than standard was selected and its docked image complex was visualized using PyMol [8] as shown in table 4.4. In SwissADME the boiled egg denotes the weather drug absorption is from GIT [6,9] or any other route. The yellow region in boiled egg indicated absorption is possible through blood brain barrier and white region indicated absorption through gastro intestinal tract (Fig 4.1 & 4.2) .In table 4.5 toxicity evaluation, all the selected ligands exhibited DILI ( Drug Induced Liver Injury [9]) similar to that of standard. Compound 67 shows least human hepatotoxicity among the selected ligands than standard. The taken standard drug posses high carcinogenicity.Majority of the selected ligand’s probability of having carcinogenicity was predicted to be less than that of standard. Compound 10 shows less hepatotoxicity and no carcinogenicity as compared to other selected ligand as well as standard. Overall from analysing the data and discussion , two compounds 67 and 10 fits the desired hypothesis.

6. CONCLUSION

The data were collected and analysed in order to interpret the result. From the collected data, we can conclude that designed compound 67 pyridopyrimidine having benzopyran moiety and compound 10 pyridopyrimidine with piperazine shows more binding affinity and less toxicity as compared to standard piritrexim as tyrosine kinase inhibitor thus; we propose that these compounds can be a promising tyrosine kinase inhibitor candidate for anticancer therapy.

7. ACKNOWLEDGEMENTS

We, 8^th semester B Pharm students of Mar Dioscorus College of Pharmacy, take privilege to acknowledge to all who have helped us in the completion of the project work. First of all, we are grateful to God Almighty, the author of knowledge and wisdom for bestowing us with all the blessing. It is a great pleasure to acknowledge our deepest thanks and gratitude to our esteemed guide, Dr SREEJA S.

REFERENCES

Hornberg JJ, Bruggeman FJ, Westerhoff HV, Lankelma J. Cancer: A systems biology disease. Biosystems 2006; 83:81–90.
Kumar A, Bhagat KK, Singh AK, et al. Medicinal chemistry perspective of pyrido[2,3-d]pyrimidines as anticancer agents. RSC Adv. 2023; 13:6872.
Yadav P, Shah K. Pyridopyrimidines as anticancer agents. ECS Trans. 2022;107(1):11577.
Palopa JA, Plano D, Moreno E, Sanmartín C. Novel quinazoline and pyrido[2,3-d]pyrimidine derivatives and their hydroselenite salts as antitumoral agents. Arkivoc 2014; (ii):187–206. http://dx.doi.org/10.3998/ark.5550190.p008.244
Wadood A, Ahmed N, Shah L, Ahmad A, Hassan H, Shams S. In-silico drug design: An approach which revolutionarised the drug discovery process. OA Drug Des Deliv. 2013; 1(1):3.
Fan J, Fu A, Zhang L. Progress in molecular docking. Quant Biol. 2019; 7(2):83–89. https://doi.org/10.1007/s40484-019-0172-y
Elzahabi HSA, Nossier ES, Khalifa NM, Alasfoury RA, El-Manawaty MA. Anticancer evaluation and molecular modelling of multi-targeted kinase inhibitors based on pyrido[2,3-d]pyrimidine scaffold. J Enzyme Inhib Med Chem. 2018; 33(1):546–57.
Khalifa NM, Al-Omar MA, Alkahtani HM, Bakheit AH. Kinase inhibitors of novel pyridopyrimidinone candidates: Synthesis and in vitro anticancer properties. J Chem. 2019; 2019:2962. https://doi.org/10.1155/2019/2635219
Bandral SK, Genne S, Golla S, Nayana RJ, et al. Identification of potent lead to inhibit DHFR for the treatment of cancer: A computational approach. Bull Environ Pharmacol Life Sci. 2023; 12(6):200–8.