¹Centre for Drug Discovery and Development, Sathyabama Institute of Science & -Technology, Chennai – 600119, India

² Department of Bioinformatics, Sathyabama Institute of Science & -Technology,  Chennai – 600119,India

³ Sathyabama Institute of Science & -Technology, Chennai – 600119

⁴ California University of Science & Medicine, School of Medicine, CA, USA

⁵ Musculoskeletal Disease Research Laboratory US Department of Veteran Affairs, Loma Linda, CA, USA

Corresponding author email: jerrine.jj@gmail.com

Article Publishing History

INTRODUCTION

Cancer is the leading causes of death globally where 1 in 6 die. Low economic prosperity and lack of awareness of prevention strategies are noted to be vital factors contributing to the burden and incidence for cancer ERBB is abbreviated from erythroblastic oncogene B, a gene isolated from avian genome. Heterodimerization of this receptor with other members of the EGFR family, typically owing to HER2 overexpression, results in the autophosphorylation of tyrosine residues within the cytoplasmic domain of the heterodimer and initiates a variety of signalling pathways leading to cellular proliferation and tumorigenesis (Yarden,  et al.,2001) ).

Amplification or over-expression of this oncogene has been shown to play an important role in the development and progression of certain aggressive types of breast cancer. According to TCGA (The cancer Genome Atlas) data portal HER2 aberration studied in various solid tumours associated with cervical, bladder and were have limited therapeutic options (Cancer Genome Atlas,2013).  Approximately, 70% of death due to cancer occurs in low- and middle-income countries. Receptor tyrosine-protein kinase erbB-2, otherwise known as HER2, a significant member in the EGFR family of receptor tyrosine kinases family, (Pegram  et al.,2020 Siegel  et al.,2020).

So, outspreading the HER2-based therapeutic option beyond breast cancer to other solid tumours with HER2 overexpression will be beneficial to the affected individuals. In recent years, the protein has become an important biomarker and target of therapy for approximately 30% of breast cancer patients. In this study, the natural compounds from plant source is considered based on their pharmacological properties. The compounds have been characterized in detail for breast cancer.  The natural source would overcome the existing synthetic drugs in mode of action and also reduce the side effects caused by the commercial compounds (Christy and Swetha 2019).

The present study focuses on the comparative in-silico based investigation of the plant  natural compounds namely Cubenin, Curcumin, Delta Cadiene, Eugenol,  Terpene and Quercetin binding efficacy towards HER2 and their intermolecular interactions were compared with the commercial drugs:Cyclophosphamide, Doxorubin, Letrozole, Methotrexate and Tamoxifen. The in-silico approach enables one to screen for ADMET properties of vast number of molecules within short span thus reducing the time and is a non- expensive and non-tedious process with great accuracy, which is not possible in standard experimental methods.

MATERIAL AND  METHODS

HER2/ERBB2 expression datamining: The cancer genome data repository was used to assess the HER2 positive cancer types and the Figure 1 revealed that HER2 expression was seen in other solid tumors associated with bladder, uterine, cervical regions also (TCGA GDC portal).Human pathology atlas module of human protein atlas also revealed the HER2 varied expression with breast, colorectal, cervical, renal and urothelial cancers and dataset related information’s listed in  Figure 2. (Uhlen  et al.,2017)

Figure 1: HER2 aberrations in various cancer types

Figure 2: HER2 profile based on human pathology atlas

HER2 protein structure quality assessment:

The target is retrieved from the Protein Data Bank, RCSB-PDB (H.M Berman et al.,2000). Upon searching the protein structure for the HER2 protein, the structure 3RCD has been chosen the fitting protein model (Ishikawa  et al.,2011). Stereochemical properties based assessment revealed that nearly 90% of residues are residing in the allowed region( ).Before proceeding  with the docking the protein structure has been energy minimized to  find an arrangement in space of  a collection of atoms where, the net inter-atomic force on each atom is acceptably close to zero and the position on the potential energy surface (PES) is a stationary point . the energy value of the protein molecule after Energy minimization is-44093.590. The energy minimisation was done in Swiss PDB viewer, it is a downloadable software (Swiss PDB viewer). Then after this the protein structure was uploaded in the Discovery Studio, where the force field CHARMM was applied and then receptor binding pockets where determined so as the ligand fits in.

Preparation of Ligand: The commercial compounds Cyclophosphamide, Doxorubin, Letrozole, Methotrexate and Tamoxifen and the natural compounds Cubenin, Curcumin, Delta Cadiene, Eugenol,  Terpene and Quercetin three dimensional structures were retrieved from Pubchem database  (Kim et al.,2016). Then the ligands were analysed if they are druggable or not using the online tool, which is based on the Lipinski’s rule of 5 ( Jayaraman et al.,2012).

ADMET Profiling:The ADME profiling was done using the Accelrys Discovery Studio 2.5 software. The lead compounds from natural resources fail to enter the market due to the poor pharmacokinetic properties. So, designing ligands satisfying the Adsorption, Distribution, Metabolism Elimination and Toxicity (ADMET) properties will go through the market as a good drug. The drugs should be orally absorbed and distributed to the site of action and eliminated from the body without leaving any traces, which produces adverse effects. Hence, the tools and computer-aided methods, nowadays, have become popular in identifying good drug candidate molecule.

Molecular Docking: Molecular docking is a key tool in structural molecular biology and computer-assisted drug design (Venkatachalam et al.,2003). The goal of ligand-protein docking is to predict the predominant binding mode(s) of a ligand with a protein of known three-dimensional structure. The receptor-ligand interaction uses Ligand Fit protocol, which docks the ligands into the binding site of receptors using shape-based searching and Monte Carlo sampling of ligands (Gangopadhyay  et al,2017).

The parameters used are PLP1 algorithm for energy grid and conjugate gradient for energy minimization. The scores for docked poses are obtained by LigScore1, LigScore  which predicts the binding affinities, -PLP1, -PLP2 known as Piecewise Linear Potential scoring function calculates both the shape and hydrogen bond complementarity of poses to the active site and Jain scoring function which scores the non-covalent protein-ligand interactions( Krammer  et al.,2005;Jain 1996).

RESULTS AND  DISCUSSION

The protein structure retrieved from RCSB-PDB: There were eight druggable cavities populated in the chains of HER2 three-dimensional structure and were ranked based on the cavity score and drug score. In general The ligandability score used to assess the possible small ligands  affinity to the specified cavity, whereas druggability scores reveal and rank the good target for HER2 binding .Specified caities of HER2 were predisposed based on cavity volume, pocket lip size, hydrophobic volume, cavity surface area, and hydrogen-bond-forming surface area of the chosen HER2 cavity.

Figure 1a: Three-dimensional structure of HER2 and druggable cavities

The Ligand Structures retrieved from Pubchem: PubChem an open source data repository for chemical substances and their related biological activities. This structure based search module compiled with variety of chemical structure format like SMILES, SMARTS, InChI, CID, molecular formula and SDF format. Since our ligandfit module accepts the SDF format we have downloaded all the Plant based natural compounds and synthetic compounds in SDF format.  HER2 specific synthetic drugs as well natural compounds were listed in the Figure 2

Figure 2a: Natural compounds and synthetic compounds three dimensional structures

ADME Studies: ADME profiling of the natural compounds aids in proving the compounds are less toxic and it passes through various barriers.

Table 1. ADME Profiling of Natural Compounds

Compound	ADMET-BBB	Solubility	Hepatotoxicity Probability	CYP2D6  Probability	AlogP98	ADMET_ PSA_2D
Cubenin	-0.203	-4.448	0.708	0.227	3.194	65.466
Curcumin	-0.544	-3.537	0.887	0.376	3.554	94.092
Delta Cadien	1.373	-5.713	0.258	0.178	4.939	0
Eugenol	0.172	-2.416	0.35	0.029	2.579	29.745
Terpene	0.242	-2.285	0.37	0.029	2.346		20.815

Table 2. ADME Profiling of Synthetic Compounds

Compound	ADMET-BBB	Solubility	Hepatotoxicity Probability	CYP2D6  Probability	AlogP98	ADMET_ PSA_2D
Cyclophosphamide	-0.723	-1.296	0.589	0.059	0.33	42.393
Doxorubicin		-4.798	0.9	0.455	-0.044	209.31
Letrozole	-0.471	-3.828	0.841	0.663	2.749	73.74
Methotrexate		-4.018	0.741	0.178	0.376	207.82
Tamoxifen	1.605	-6.711	0.894	0.702	6.319		12.282

Lipinski Rule : Lipinski rule of 5 helps in distinguishing between drug like and nondrug like molecules. It predicts high probability of success or failure due to drug likeness for molecules complying with 2 or more of the following rules:

Table 3. Lipinski Rule of Druggability for Synthetic Compounds

	MASS	HBD	HD A	Log P	MR
Cyclophosphamide	356	1	6	2.5102	91.47678
Doxorubicin	368	2	6	3.369898	102.0166
Letrozole	204	0	0	4.725199	66.74298
Methotrexate	164	1	2	2.1293	48.55979
Tamoxifen	152	1	1	2.2797	47.30179

Table 4. Lipinski Rule of Druggability for Natural Compounds

	MASS	HBD	HD A	Log P	MR
Cubenin	356	1	6	2.5102	91.47678
Curcumin	368	2	6	3.369898	102.0166
Delta Cadeine	204	0	0	4.725199	66.74298
Eugenol	164	1	2	2.1293	48.55979
Terpene	152	1	1	2.2797	47.30179

Molecular docking studies using ligandfit module : The docking studies where done to find the Receptor-Ligand Interactions LigandFit assess the small molecules affiinity into the HER2 protein active site by applyinng its shape complementary screening. In general docking process list the top conformations of ligand after the energy minimization method based on steepest descent method and conjugate gradient method. Ligandfit uses the consensus scoring to minimize the false positive. Proposed consensus scoring method was depends on LigScore1, LigScore2, piecewise linear potential 1 (PLP1) , piecewise linear potential 2, potential mean force (PMF) , Jain score.

The above are the docking 2D & 3D images of the receptor with the ligand, with their docked positions and the respective amino acids. The dock scores are evaluated based on the parameters of lig score, PLP1, PLP2,JAIN score and PMF score.

Table 5. Dock scores of Synthetic compounds.

Compounds	Lig score 1		Lig score 2		PLP1		PLP 2		JAIN		PMF
Cyclophosphamide	-0.168	1.652		30.288		32.14		1.035		13.01
Doxorubin	2.961	4.52		45.541		41.176		1.873		41.452
Letrozole	1.375	3.749		50.645		46.388		-0.093		19.141
Methotrexate	2.883	4.105		38.079		30.442		-2.452		34.579
Tamoxifen	1.878	3.238		34.473		32.632		-1.709		41.811

Table 6. Dock scores of natural compounds

Compounds	Lig score 1	Lig score 2	PLP1	PLP 2	JAIN	PMF
Cubenin	3.563	4.065	47.011	48.088	0.271	37.173
Curcumin	3.235	4.174	45.657	44.984	-1.815	50.533
Delta Cadiene	6.009	2.93	42.196	41.498	1.496	39.294
Eugenol	1.321	2.519	19.224	20.169	1.367	30.905
Terpene	1.368	2.549	26.964	29.576	-0.044	33.962

CONCLUSION

Thus, the results of the docking studies carried out on HER2 corroborate to the findings that the most suitable drug like properties are possessed by the compound. In comparison with other compounds natural compounds are better and it is druggable. This provides evidence of how a natural compounds from plants can be a source of potential anti- cancer agent. The preclinical studies will pave way for a potential anti-cancer compound. The study aimed in finding the compatible natural lead molecules from plants and it shows that the compounds are druggable based on the Lipinski’s rule and the Insilico toxicity study studies are also positive, adding advantage for the compounds to be druggable. With the docking studies , the compounds that dock with the target has been found. The dock scores are also supporting for the further study. The work will be continued on breast cancer cell line study invitro  and then in vivo studies.

REFERENCES

Aditi Gangopadhyay A  Hirak Jyoti Chakraborty  Abhijit Datta  (2017) Targeting the Dengue β-OG With Serotype-Specific Alkaloid Virtual Leads Journal of molecular graphics & Modeling 73: 129. [PMID: 28279821]

Berman, HM 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.

Cancer Genome Atlas Research Network, Weinstein, J. N., Collisson, E. A., Mills, G. B., Shaw, K. R., Ozenberger, B. A., Ellrott, K., Shmulevich, I., Sander, C., & Stuart, J. M. (2013). Nature genetics, 45(10), 1113–1120. https://doi.org/10.1038/ng.2764

Christy J & Swetha V (2019). Differential expression analysis of kruppel like factors 6 and antagonistic effect study of cinnamic acid – an in silico approach. Asian Journal of Pharmaceutical and Clinical Research, 12(7), 107-113. https://doi.org/10.22159 /ajpcr. 2019.v12i7.33406

Ishikawa T, Seto M, Banno H, et al.(2011)  Design and synthesis of novel human epidermal growth factor receptor 2 (HER2)/epidermal growth factor receptor (EGFR) dual inhibitors bearing a pyrrolo[3,2-d]pyrimidine scaffold. J Med Chem. 54(23):8030‐8050. doi:10.1021/jm2008634

Jain A.N.(1996)  Scoring Noncovalent Protein-Ligand Interactions: A Continuous Differentiable Function Tuned to Compute Binding Affinities Journal of Computer-Aided Molecular Design,10,427–440

Jayaram, B Tanya Singh, Goutam Mukherjee, Abhinav Mathur, Shashank Shekhar, and Vandana Shekhar, Sanjeevini (2012) a freely accessible web-server for target directed lead molecule discovery BMC Bioinformatics 13, S7.

Kim, S., Thiessen, P. A., Bolton, E. E., Chen, J., Fu, G., Gindulyte, A., Han, L., He, J., He, S., Shoemaker, B. A., Wang, J., Yu, B., Zhang, J., & Bryant, S. H. (2016). PubChem Substance and Compound databases. Nucleic acids research, 44(D1), D1202–D1213. https://doi.org/10.1093/nar/gkv951

Krammer A, Kirchhoff PD, Jiang X, Venkatachalam CM, Waldman M.(2005) The Journal of Molecular Graphics and Modelling, 23,395-407

Nat. Rev. Mol. Cell Biol. 2,127–137

Pegram MD, Miles D, Tsui CK, Zong Y.(2020)  Clin Cancer Res. 26(4):775‐786.    doi:10.1158/1078-0432.CCR-18-1976

Siegel RL, Miller KD, Jemal A. CA (2020) Cancer J Clin. 70(1):7‐30. doi:10.3322/caac.21590

SWISS-MODEL and the Swiss-Pdb Viewer: An environment for comparative protein modeling. Electrophoresis 18, 2714-2723.

Uhlen  M. Cheng Zhang Sunjae Lee  et al (2017). A pathology atlas of the human cancer transcriptome. Science. PubMed: 28818916 DOI: 10.1126/science.aan2507

Venkatachalam C.M., Jiang X., Oldfield T., Waldman M, (2003)The Journal of Molecular Graphics and Modelling,21, 289-07

Yarden, Y. & Sliwkowski,M X (2001) Kinase Inhibitor Drugs