Biomedical Research Journal

: 2019  |  Volume : 6  |  Issue : 1  |  Page : 25--33

Transforming growth factor beta receptor 2 single-nucleotide polymorphism association with oral cancer and In silico identification of small drug-like molecules as inhibitors to transforming growth factor Beta-2 receptor

Shaleen Multani1, Hetal Damani Shah1, Dhananjaya Saranath2,  
1 Department of Biological Sciences, Sunandan Divatia School of Science, SVKM's NMIMS (Deemed-to-be) University, Mumbai, Maharashtra, India
2 Dr. Kantilal J. Sheth Memorial Building, 84-A, RG Thadani Marg, Worli, Mumbai, Maharashtra, India

Correspondence Address:
Dr. Dhananjaya Saranath
Dr. Kantilal J. Sheth Memorial Building, 84-A, RG Thadani Marg, Worli, Mumbai - 400 018, Maharashtra


Objective: Oral cancer, in India, constitutes 26% of global oral cancer burden. The major risk factors include tobacco, areca nut, alcohol, and human papillomavirus 16/18; however, only 5%–10% of the high-risk individuals develop oral cancer, indicating the role of genomic variants in susceptibility to oral cancer. Conventional treatment options in oral cancer have resulted in relatively poor prognosis and an unmet need of treatment. In silico analysis, therefore, was performed to identify small drug-like molecules as potential inhibitors of transforming growth factor beta-2 receptor (TGFβRII). Materials and Methods: Seven single-nucleotide polymorphisms (SNPs) were analyzed in 500 histopathologically confirmed oral cancer samples and 500 long-term tobacco users (LTTUs) as controls using allelic discrimination real-time polymerase chain reaction or high-resolution melting analysis. The differential frequencies in oral cancer and LTTUs were calculated using SPSS software (version 19), and odds ratio (OR) to indicate risk to oral cancer using structure-based virtual screening of drug-like molecules was performed to identify lead inhibitor molecules to TGFβRII using Schrödinger Suite 2015-4. Results: Heterozygous GC genotype of TGFBR2 rs9843143 demonstrated increased risk ([P = 0.011; OR 1.61 [1.25–2.1]) while CC genotype showed decreased risk (P = 0.005; OR 0.61 [0.44–0.83]) to oral cancer. Increased/decreased risk to oral cancer was not observed for the other SNPs. In silico analysis identified six molecules as inhibitors of TGFβRII kinase domain from 17,723 conformers from Maybridge HitFinder library and 2685 conformers from MEGx AnalytiCon natural product library. Conclusion: SNP rs9843143 (TGFBR2) demonstrated a significant association (P < 0.05) with oral cancer and six potential inhibitors of TGFβRII kinase were identified using in silico analysis.

How to cite this article:
Multani S, Shah HD, Saranath D. Transforming growth factor beta receptor 2 single-nucleotide polymorphism association with oral cancer and In silico identification of small drug-like molecules as inhibitors to transforming growth factor Beta-2 receptor.Biomed Res J 2019;6:25-33

How to cite this URL:
Multani S, Shah HD, Saranath D. Transforming growth factor beta receptor 2 single-nucleotide polymorphism association with oral cancer and In silico identification of small drug-like molecules as inhibitors to transforming growth factor Beta-2 receptor. Biomed Res J [serial online] 2019 [cited 2022 Jan 20 ];6:25-33
Available from:

Full Text


Oral cancer is a high incidence cancer in Asian countries including India with an annual incidence of 77,003 and is the most common cancer in males (53,842) and fifth most common cancer in females (23,161) in India.[1] Despite easy accessibility to the oral cavity, a majority of oral cancers are diagnosed in advanced stages III/IV, resulting in poor prognosis with a 5-year survival rate of 40%.[2] Tobacco, areca nut, alcohol, and high-risk oncogenic human papillomavirus 16/18 constitute major risk factors to oral cancer.[3],[4] The attributable risk of tobacco in the development of oral cancer is 80%; however, only a small proportion of tobacco users develop persistent premalignant lesions, transforming into oral cancer (3%–8%).[5] Thus, genomic variants represented as single-nucleotide polymorphisms (SNPs) could play a crucial role in the development of oral cancer.

About 90% of the genomic variants are SNPs, present as single base changes in intronic or exonic regions of the gene with a frequency of >1% in the population.[6] SNPs may be synonymous with no change in amino acid or nonsynonymous resulting in amino acid substitution causing alteration in the protein molecule. SNPs in intronic regions may alter conformation and stability of the DNA. Further, intronic SNPs may impact DNA polymerase processivity, transcription factor binding, and nucleosome assembly.[7],[8] The ancestral allele is considered wild-type (WT) and the altered nucleotide is the SNP. The genotypes are either homozygous WT, homozygous SNP, or heterozygous. The differential distribution of the genotypic and allelotypic frequencies in oral cancer patients and control group may indicate association of SNPs with risk to oral cancer identifying potential biomarkers.

Whole genome-wide association studies using high-throughput microarray have reported association of SNPs in several cancers including colorectum, breast, lung, and oral cancer.[9],[10] Validation of SNPs identified in microarray studies using larger sample sizes and alternative technology such as real-time polymerase chain reaction (PCR) and sequencing is mandatory. SNPs have been examined in context of DNA methylation in breast, colon, and lung cancers demonstrating a direct relationship of the SNPs in the noncoding sequence in the regulation of methylation with functional roles in cancer.[11] Our study focused on SNPs in several signal transduction genes in proliferation and apoptosis pathway including Endothelin Receptor Type A (EDNRA), interleukin-1 receptor-associated kinase 3 (IRAK3), protein tyrosine phosphatase receptor G (PTPRG), spleen tyrosine kinase (SYK), breakpoint cluster region (BCR), and transforming growth factor beta-2 receptor (TGFβR2). EDNRA is associated with angiogenesis, cell proliferation and cell survival,[12] and endothelial-to-mesenchymal transition.[13] IRAK3 is a negative regulator of Toll-like receptor and prevents inflammation against tumor cells by deactivating the action of IRAK2/4.[14] PTPRG regulates cell growth, differentiation, mitotic cycle and transformation,[15] and as a tumor suppressor.[16] SYK is associated with the development of immune cells, cell proliferation, and differentiation and regulates cell migration and invasion in head and neck cancers.[17] BCR is involved in the formation of Philadelphia chromosome, pathogenic in chronic myeloid leukemia. The BCR-ABL fusion protein activates RAS, JAK/STAT, and PI-3 kinase, enhancing proliferation, differentiation, and decreases apoptosis.[18] TGFRβII binds TGF-β 1, 2, 3 ligands and has dual function in cell proliferation and apoptosis. The ligand subsequently phosphorylates TGFβRI, leading to heterodimerization and activation of SMAD proteins[19] [Figure 1]. The deregulation of TGFβRII is observed in several cancers including breast,[20] colorectal, and hepatocellular cancer.[21],[22] Targeting the kinase domain of TGFβRII with inhibitors may block its function and inhibit tumor progression. Hence, small drug-like lead molecules were identified to block the kinase domain of TGFβRII by virtual screening of the chemical compounds in the Maybridge HitFinder and natural products in the MEGx AnalytiCon library.{Figure 1}

 Materials and Methods

Study subjects

The study subjects comprised 500 histopathologically confirmed oral cancer patients admitted to Prince Aly Khan Hospital, Mumbai, India, and unrelated 500 long-term tobacco users (LTTUs) with minimum 10 years of tobacco habit as normal controls. The control LTTUs were obtained from cancer screening camps conducted by the Cancer Patients Aid Association, Mumbai, India. Demographic data including age, gender, tobacco habits, and clinicopathological profile including tumor size, site, nodal infiltration, differentiation, and stage of cancer were recorded. The study was approved by the Institute Ethics Committees of NMIMS (deemed-to-be) University, Mumbai, Prince Aly Khan Hospital, Mumbai, and Cancer Patients Aid Association, Mumbai. Informed consent for voluntary participation was obtained from the participants.

DNA extraction

DNA was extracted from peripheral blood samples, using PureLink DNA extraction kit as per the manufacturer's instructions (Invitrogen, California, USA). Qualitative and quantitative analysis was performed using NanoDrop Spectrophotometer 2000 (ThermoScientific, Walthman, USA), and samples with OD 260/280 nm ratio between 1.8 and 2.0 were used in the study.

Single-nucleotide polymorphisms analysis

Allele-specific primer sequences were designed using Allele ID software and primers were purchased from Eurofins (Luxembourg City, Luxembourg). SNPs were determined using Allelic Discrimination Real-time PCR assay using SYBR green dye[23] and high-resolution melting (HRM) analysis with Syto 9 dye. The HRM protocol included an initial holding step at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 15 s and annealing at 57°C for 1 min. Melt curve analysis was performed at 95°C for 10 s, with gradual increase at 0.3°C/min from 60°C to 95°C, and gradual decrease at 1°C/min from 95°C to 60°C. Representative samples were subjected to nucleotide sequencing. The sequencing primers were designed to obtain fragment sizes from 461 to 620 bp and are given in [Supplementary Table S1]. Nucleotide sequencing for confirmation of the alleles were performed at SciGenome, Adyar, Chennai, India.[INLINE:1]

Statistical analysis

Hardy-Weinberg equilibrium analysis for the genotypes was performed using SNPstats software. Genotypic and allelotypic frequencies and association with oral cancer were analyzed by Fisher's exact test using SPSS software (version 19, IBM, New York, New York city, United States), and P < 0.05 was considered statistically significant. Odds ratio (OR) and 95% confidence intervals were determined using software. Association of genotypes with clinicopathological characteristics in oral cancer was analyzed using GraphPad Prism (v5.0, GraphPad, San Diego, California, United States).

Structure-based virtual screening


The molecular modeling studies were performed using Schrödinger Small-Molecule Drug Discovery suite 2015-4 and modules-Protein Preparation Wizard, Prime v4.2, Ligprep v4.6, Site Glide v5.9, and Qikprop v4.6.

Protein structure and preparation

The crystal structure of WT TGFBR2 protein was obtained from the protein data bank (PDB).[24] The apo TGFβRII-6M crystal structure was used for the docking experiments (PDB 5E8V)[25] using Glide v5.9.[26] The mutations induced in the protein to develop the crystal structure were reversed to obtain configuration of the native protein, missing side chains were added, and water molecules were deleted using Protein Preparation Wizard with default settings as in Schrödinger suite 2015-4.[27] Hydrogen bonds (H-bonds) were optimized and the protein was subjected to energy minimization using Prime (default settings).

Preparation of ligands

LigPrep v4.6 was used to prepare and optimize 14,400 small drug-like molecules imported from Maybridge HitFinder™ library.[27] The library includes a diverse collection of small molecules following Lipinski's rule of five for drug-likeliness with purity >90%. MEGx AnalytiCon library consisting 268 natural products extracted and purified from plants, following Lipinski's rule of five, with appropriate absorption, distribution, metabolism, excretion, and toxicity (ADMET) profile and available for screening was prepared using default settings as implemented in Schrödinger suite 2015-4.[28]

Receptor grid generation and Ligand docking

Prediction of optimal binding site is a cornerstone of drug design. SiteMap was employed for identification of putative binding sites for TGFBR2-ligand interaction. Five active sites were generated and the largest site with encompassing the residues of importance was chosen. The residues critical for TGFβRII kinase activity include Lysine277 (Lys277), at the ATP binding site, and a unique site consisting of DFG motif of Aspartic acid397 (Asp397), Phenylalanine398 (Phe398), and Glycine399 (Gly399).[29] The mapped site was used to define the area for ligand docking using Receptor Grid Generation Module v5.9.

The virtual screening protocol was performed using Glide v5.9, by screening initially in HTVS mode, followed by standard precision (SP) and extra precision (XP) modes. The binding modes of the conformers were analyzed and the hits were used for the Prime-MMGBSA calculations.[30] The number of heavy atoms was calculated, and the ratio of the MMGBSA score with number of heavy atoms provided the ligand efficiency of the molecules.[31]

In silico analysis for absorption, distribution, metabolism, excretion, and toxicity

The in silico analysis of the ADMET properties was performed for top 30% hits obtained from XP mode using Qikprop v4.6 as per Schrödinger suite 2015-4. The parameters included molecular weight, partition coefficient, solvent-accessible molecular surface, globularity, polarizability, aqueous solubility, blood/brain barrier partition coefficient, IC50 value for blockage of HERG K+ channels, Caco-2 cell membrane, MDCK and skin permeabilities, likely metabolic reactions and percent human oral absorption. The properties provide an overview of the physiochemical, pharmacokinetic, and toxic properties of small drug-like molecules.


Demographic and clinicopathological data

The 500 oral cancer patients comprised 94.4% males and 7.6% females; the control group constituted 82.8% males and 17.2% females. The age range was 22–80 years with mean age of 48.7 years in the oral cancer cases and 46.8 years in the control group. The subjects were tobacco habitués with an average duration of 14.2 ± 7.6 usage in cases and 18.1 ± 9 years usage in controls. The site most affected in the patients was the buccal mucosa (57.4%), followed by tongue (27.8%), and alveolus-gingival complex (9.2%), with other sites as retromolar trigone, palate and labial mucosa constituting 5.6%. Histopathologically, the oral cancers were squamous cell carcinomas, with the majority of cancers moderately differentiated (89.2%), 7.2% well differentiated and 3.6% poorly differentiated. The size of the tumors were T1 (22.2%), T2 (47.6%), T3 (14.4%), and T4 (15.6%) with the absence of nodal infiltration in 60.8% and 39.2% demonstrated nodal infiltration. Patients were diagnosed in early stages I/II (44.4%) and late stages III/IV (55.6%).

Single-nucleotide polymorphisms genotyping

The genotypes of the samples were determined by melt curve analysis, with differential temperatures indicating the presence of homozygous WT, heterozygous WT/SNP, and homozygous SNP genotypes. The melt curve temperatures of the WT and SNP alleles were as follows: rs9843143 (TGFBR2) WT 82.5°C and SNP 78.9°C; rs6827096 (EDNRA) WT 79.2°C and SNP 81.3°C; rs1821777 (IRAK3) WT 79.0°C and SNP 75.6°C; rs1911746 (PTPRG) WT 80.4°C and SNP 77.6°C; rs290974 (SYK) WT 76.5°C and SNP 80.8°C; rs2156921 (BCR) WT 78.0°C; and SNP 78.6°C and rs144761391 (GAPDH) WT 83.34°C. The heterozygous WT/SNP demonstrated specific peaks of WT and SNP melt temperatures. A representative melt curve analysis indicating WT, WT/SNP, and SNP peaks for rs9843143 (TGFBR2) is given in [Figure 2]a,[Figure 2]b,[Figure 2]c and the respective nucleotide sequencing indicated in [Figure 2]d,[Figure 2]e,[Figure 2]f. The HRM analysis of SNP rs2156921 (BCR) indicated the WT genotype as a median line, heterozygous WT/SNP genotype showed 2 peaks either below or above the standard, and homozygous SNP genotype as a single peak above the standard [Figure 2]g and HRM genotype nucleotide sequencing in [Figure 2]h,[Figure 2]i,[Figure 2]j.{Figure 2}

The SNPs rs9843143 (TGFBR2), rs6827096 (EDNRA), rs1821777 (IRAK3), rs1911746 (PTPRG), and rs2156921 (BCR) followed the Hardy-Weinberg equilibrium (HWE) in the cumulative Indian cohort (n = 1000) and individually oral cancer and control groups whereas SNP rs290974 (SYK) showed a deviation from HWE (P = 0.01). A significant difference between the genotypic frequencies of oral cancer cases and control group was observed for SNP rs9843143 (TGFBR2) with the heterozygous GC genotype indicating increased oral cancer risk (P = 0.011; OR 1.61 [95% CI 1.25–2.10]) and the homozygous SNP CC genotype showed decreased oral cancer risk (P < 0.005; OR 0.61 [95% CI 0.44–0.83]) [Table 1]. A significant difference in the genotypes in cases and controls was not observed for SNPs rs6827096 (EDRNA), rs1821777 (IRAK3), rs1911746 (PTPRG), rs290974 (SYK), rs215692 (BCR), and rs144761391 (GAPDH). A significant association was not observed between the SNPs and clinicopathological parameters (P > 0.05) of oral cancer. Whereas WT A allele of SNP rs2156921 (BCR) showed a decreased risk (OR 0.82 [95% CI 0.68–0.98]), and G allele showed an increased risk (OR 1.21 [95% CI 1.01–1.45]) to oral cancer. A significant difference was not observed between allelic frequencies of the other SNPs and oral cancer (P > 0.05).{Table 1}

Identification of inhibitors of TGFBR2 kinase domain

Virtual screening for the prepared TGFβRII protein (PDB ID 5E8V) and the ligand conformers from the Maybridge HitFinder™ library (17,723) and MEGx AnalytiCon library (2658) was analyzed. The top 30% hits selected on the basis of the Glide GScore enumerated 5100 conformers from the Maybridge HitFinder™ library and 797 conformers from MEGx AnalytiCon library, further screened in the SP mode. The top 30% hits from the SP mode indicated 1529 and 239 conformers, respectively, from the libraries, further screened in the XP mode, finally providing 435 and 71 conformers. The MMGBSA binding free energy was calculated for the top hits, and the ligands were arranged in ascending order of energy. Five molecules from Maybridge HitFinder™ library and one from MEGx AnalytiCon library were selected and subjected to further analysis. The ligands showed bond formation with the DFG motif residue, Asp397 and ATP binding site Lys277 [Figure 3] and [Table 2], and these did not bind to the other residues of the DFG motif (Phe398 and Gly399). The number of heavy atoms and ligand efficiency was also considered for the stability profile of the ligands with the natural product ligand NP018768 showing the best ligand efficiency of 2.72 [Table 2].{Figure 3}{Table 2}

Drug likeliness properties: In silico absorption, distribution, metabolism, excretion, and toxicity analysis

Fourteen pharmacokinetic properties were assessed using Qikprop module [Table 3]. The compounds exhibited predicted properties within 95% of the acceptable range of oral drugs. The molecular weights of the ligands were in the range of 328.4–390.5, while all the chemical ligands had hydrophobic nature with logP ranging from 1.16 to 3.84 and the natural product had hydrophilic nature with logP-1.35. The solvent-accessible molecular surface, volume, polarizability, globularity, aqueous solubility, and permeability descriptors for the compounds were within the normal range [Table 3]. The compounds were predicted to be involved in 2–7 metabolic reactions. All compounds showed a high level (>69%) with respect to human oral absorption.{Table 3}


Despite the lifestyle habits resulting in exposure to high-risk factors for oral cancer, a small proportion of the tobacco habitués develop oral cancer, indicating a role of genomic variants in development of oral cancer. Hence, we investigated 500 oral cancer patients and 500 LTTUs as controls for seven SNPs in genes associated with cell proliferation and apoptosis and a control SNP in GAPDH gene using Allelic Discrimination Real-Time PCR assay with SYBR green dye and HRM with Syto 9 dye. The results indicated that five of the SNPs rs9843143 (TGFBR2), rs6827096 (EDNRA), rs1821777 (IRAK3), rs2156921 (BCR) and rs1911746 (PTPRG) followed the HWE whereas SNP rs290974 (SYK) showed deviation from HWE. The deviation may be due to sample size, since an 'infinite population' is favored for a locus to exhibit HWE.[32]

The control SNP rs144761391 (GAPDH) is also absent in several populations as also in the Indian cohort and hence HWE analysis was not feasible. The prevalence of the SNPs in the Indian population (n = 1000) was calculated and compared with the genotypic and allelotypic distribution in the HapMap databases, including Gujrati Indians in Houston (GIH), Han Chinese (HCB), Japanese (JPT), African tribals (YRI), and Central Europeans (CEU) ( The comparative analysis indicated a concordance with SNP rs9843143 (TGFBR2) with the HCB, JPT, and CEU populations; SNP rs6827096 (EDNRA) with GIH, JPT, and CEU populations; SNP rs1821777 (IRAK3) with GIH, HCB, and JPT populations; SNP rs1911746 (PTPRG) with GIH, HCB, JPT, and YRI populations; SNP rs290974 (SYK) with GIH and YRI populations; and SNP rs2156921 (BCR) with GIH, JPT, and CEU populations. The control SNP rs144761391 (GAPDH) matched the varied ethnic groups documented in the HapMap database. SNPs vary in different ethnic populations and the preference of an allele is based on natural selection. The available HapMap data sample sizes varied and were relatively small which may lead to discordance in distribution of alleles in specific populations.

The WT and SNP alleles of the rs2156921 (BCR) showed significant association with oral cancer, with the WT allele demonstrated decreased risk OR 0.82 (0.68–0.98) and SNP allele increased risk OR 1.21 (1.01–1.45). The genotypic analysis in the present study demonstrated a role of TGFBR2 SNP in oral cancer, as the heterozygous genotype was associated with an increased risk to oral cancer with OR 1.61 (1.25–2.10), while the homozygous SNP genotype showed an increased frequency in controls and OR 0.61 (0.44–0.83), indicating a decreased risk and consequent protection to oral cancer. The other SNPs did not show an association with susceptibility to oral cancer. SNPs are low penetrant gene variants and are less deterministic and more probabilistic and hence a high-risk panel may indicate better accuracy and sensibility of the correlation with risk.[33] Earlier studies from our laboratory had demonstrated three SNPs rs2124437 (RASGRP3), rs1335022 (GRIK2), and rs4512367 (PREX2) with risk association in oral cancer. The genotype AA in SNP rs2124437 (RASGRP3) (OR 1.34), genotype TT in SNP rs1335022 (GRIK2) (OR 1.58), and genotypes CC (OR 1.56) and TT (OR 2.77) in SNP rs4512367 (PREX2) demonstrated increased risk to oral cancer.[23] We examined coinheritance of the four high-risk associated genotypes and three low-risk associated genotypes in oral cancer [Supplementary Table S2]. An additional increased risk to oral cancer (P < 0.000; OR 4.89 [1.65–14.50]) was observed on coinheritance of the four high-risk genotypes. Coinheritance of low-risk panel of genotypes CT (GRIK2) (OR 0.68), CT (PREX2) (OR 0.49), and CC (TGFBR2) (OR 0.61) further decreased risk to oral cancer as indicated by OR 0.37 (0.21–0.64).[INLINE:2]

TGF-β signaling pathway has a dual role in cancer, as a tumor suppressor in the early stages and an oncogene in late stages of cancer.[34] TGFβRII is oncogenic in ER/PR-negative, node-positive breast cancers with increased expression in the stroma of breast cancer cells, leading to increased metastasis by remodeling the extracellular matrix.[20] Besides, TGFβRII enhances angiogenesis in the late stages of breast cancer by activating SMAD 1, 5, 8 leading to transcription of tumor progressing genes.[34] However, in hepatocellular cancer, TGFβRII is downregulated in metastatic cells with low expression associated with large tumor size, poor differentiation, invasion, intra-hepatic metastasis, and shorter recurrence-free survival.[22] Several mutations inhibiting the activity of TGFβRII are observed in several cancers including colon cancer, gastric cancer, head and neck squamous cell carcinoma, and breast cancer.

The monoclonal antibody for EGFR, cetuximab, is the most common targeted therapy in oral cancer. With the aim of developing additional targeted therapy for oral cancer, we used in silico drug designing to inhibit the kinase activity of TGFβRII. The crystallized structure of the TGFβRII kinase domain has been recently published[25] and has 41% homology with TGFβRI.[29] Therefore, the unique DFG motif, consisting Asp397, Phe398, and Gly399 along with the ATP binding site, Lys277 in the TGFβRII, was used as a target site to inhibit the oncogenic activity. The Maybridge HitFinder™ and MEGx AnalytiCon libraries were investigated by virtual screening for TGFβRII ligands. We identified six molecules from the 14,668 molecules which followed the Lipinski guidelines for “drug-likeliness” with ADMET properties in the optimal range of 95% of known oral drugs. The identified ligands demonstrated low MMGBSA scores from a group of 506 ligands obtained after virtual screening for interaction with the unique site in TGFβRII. Ligand NP018768 demonstrated appropriate ligand efficiency and BTB14371 ligand was the smallest lead molecule with 22 heavy atoms.


Our data demonstrate that SNPs are critical 'predictive biomarkers' indicating increased /decreased risk to oral cancer. Our study defines the percent frequency of the SNPs in the Indian population and association of rs9843143 (TGFBR2) genotype and rs2156921 (BCR) allele with oral cancer. Coinheritance of a panel of high-risk SNPs further increased risk to oral cancer with an OR of 4.89 (1.65–14.5); and coinheritance of a low-risk panel further decreased the risk with OR 0.37 (0.21–0.64). Thus, the mentioned panel of SNPs may act as a “predictive biomarker” in LTTUs to indicate the risk of oral cancer. In addition, in silico analysis for identification of lead molecules to inhibit TGFβRII kinase activity identified six molecules. The identified ligands provide possible candidate molecules and need to be validated in vitro and in vivo to indicate its potential use as targeted therapy in oral cancer.


The authors gratefully acknowledge Sunandan Divatia School of Science, NMIMS (Deemed-to-be) University for its facilities. We also would like to acknowledge Dr. Sultan Pradhan from Prince Aly Khan Hospital, Mumbai, for his support in helping us collect patient samples.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, et al. Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 2015;136:E359-86.
2Wang B, Zhang S, Yue K, Wang XD. The recurrence and survival of oral squamous cell carcinoma: A report of 275 cases. Chin J Cancer 2013;32:614-8.
3Petti S. Lifestyle risk factors for oral cancer. Oral Oncol 2009;45:340-50.
4Saranath D. Contemporary Issues in Oral Cancer. Mumbai, India: Oxford University Press; 2000.
5Scheifele C, Reichart PA. Is there a natural limit of the transformation rate of oral leukoplakia? Oral Oncol 2003;39:470-5.
6Chien MH, Yang JS, Chu YH, Lin CH, Wei LH, Yang SF, et al. Impacts of CA9 gene polymorphisms and environmental factors on oral-cancer susceptibility and clinicopathologic characteristics in Taiwan. PLoS One 2012;7:e51051.
7Ignatova Z, Martínez-Pérez I, Zimmermann KH. DNA Computing Models. Springer Science and Business Media; 2008. Available from: &pgis=1. [Last Accessed 2018 Dec 01].
8Langley SA, Karpen GH, Langley CH. Nucleosomes shape DNA polymorphism and divergence. PLoS Genet 2014;10:e1004457.
9Bhatnagar R, Dabholkar J, Saranath D. Genome-wide disease association study in chewing tobacco associated oral cancers. Oral Oncol 2012;48:831-5.
10Easton DF, Eeles RA. Genome-wide association studies in cancer. Hum Mol Genet 2008;17:109-15.
11Heyn H, Sayols S, Moutinho C, Vidal E, Sanchez-Mut JV, Stefansson OA, et al. Linkage of DNA methylation quantitative trait loci to human cancer risk. Cell Rep 2014;7:331-8.
12Lange K, Kammerer M, Hegi ME, Grotegut S, Dittmann A, Huang W, et al. Endothelin receptor type B counteracts tenascin-C-induced endothelin receptor type A-dependent focal adhesion and actin stress fiber disorganization. Cancer Res 2007;67:6163-73.
13Nanni S, Priolo C, Grasselli A, D'Eletto M, Merola R, Moretti F, et al. Epithelial-restricted gene profile of primary cultures from human prostate tumors: A molecular approach to predict clinical behavior of prostate cancer. Mol Cancer Res 2006;4:79-92.
14Rhyasen GW, Starczynowski DT. IRAK signalling in cancer. Br J Cancer 2015;112:232-7.
15Cheung AK, Lung HL, Hung SC, Law EW, Cheng Y, Yau WL, et al. Functional analysis of a cell cycle-associated, tumor-suppressive gene, protein tyrosine phosphatase receptor type G, in nasopharyngeal carcinoma. Cancer Res 2008;68:8137-45.
16Shu ST, Sugimoto Y, Liu S, Chang HL, Ye W, Wang LS, et al. Function and regulatory mechanisms of the candidate tumor suppressor receptor protein tyrosine phosphatase gamma (PTPRG) in breast cancer cells. Anticancer Res 2010;30:1937-46.
17Luangdilok S, Box C, Patterson L, Court W, Harrington K, Pitkin L, et al. Syk tyrosine kinase is linked to cell motility and progression in squamous cell carcinomas of the head and neck. Cancer Res 2007;67:7907-16.
18Salesse S, Verfaillie CM. BCR/ABL: From molecular mechanisms of leukemia induction to treatment of chronic myelogenous leukemia. Oncogene 2002;21:8547-59.
19Derynck R, Zhang YE. Smad-dependent and smad-independent pathways in TGF-beta family signalling. Nature 2003;425:577-84.
20Barlow J, Yandell D, Weaver D, Casey T, Plaut K. Higher stromal expression of transforming growth factor-beta type II receptors is associated with poorer prognosis breast tumors. Breast Cancer Res Treat 2003;79:149-59.
21Kubiczkova L, Sedlarikova L, Hajek R, Sevcikova S. TGF-β – An excellent servant but a bad master. J Transl Med 2012;10:183.
22Mamiya T, Yamazaki K, Masugi Y, Mori T, Effendi K, Du W, et al. Reduced transforming growth factor-beta receptor II expression in hepatocellular carcinoma correlates with intrahepatic metastasis. Lab Invest 2010;90:1339-45.
23Multani S, Pradhan S, Saranath D. Gene polymorphisms and oral cancer risk in tobacco habitués. Tumour Biol 2016;37:6169-76.
24Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, et al. The protein data bank. Nucleic Acids Res 2000;28:235-42.
25Tebben AJ, Ruzanov M, Gao M, Xie D, Kiefer SE, Yan C, et al. Crystal structures of apo and inhibitor-bound TGFβR2 kinase domain: Insights into TGFβR isoform selectivity. Acta Crystallogr D Struct Biol 2016;72:658-74.
26Friesner RA, Murphy RB, Repasky MP, Frye LL, Greenwood JR, Halgren TA, et al. Extra precision glide: Docking and scoring incorporating a model of hydrophobic enclosure for protein-ligand complexes. J Med Chem 2006;49:6177-96.
27Sastry GM, Adzhigirey M, Day T, Annabhimoju R, Sherman W. Protein and ligand preparation: Parameters, protocols, and influence on virtual screening enrichments. J Comput Aided Mol Des 2013;27:221-34.
28Clavio NA, Billones JB. Virtual screening against Mycobacterium tuberculosis isocitrate lyase and in silico ADME-tox evaluation of top hits. J Chem Pharm Res 2014;6:727-38.
29Huse M, Chen YG, Massagué J, Kuriyan J. Crystal structure of the cytoplasmic domain of the type I TGF beta receptor in complex with FKBP12. Cell 1999;96:425-36.
30Vijayakumar B, Parasuraman S, Raveendran R, Velmurugan D. Identification of natural inhibitors against angiotensin I converting enzyme for cardiac safety using induced fit docking and MM-GBSA studies. Pharmacogn Mag 2014;10:S639-44.
31Hopkins AL, Keserü GM, Leeson PD, Rees DC, Reynolds CH. The role of ligand efficiency metrics in drug discovery. Nat Rev Drug Discov 2014;13:105-21.
32Salanti G, Amountza G, Ntzani EE, Ioannidis JP. Hardy-weinberg equilibrium in genetic association studies: An empirical evaluation of reporting, deviations, and power. Eur J Hum Genet 2005;13:840-8.
33Hirschhorn JN, Lohmueller K, Byrne E, Hirschhorn K. A comprehensive review of genetic association studies. Genet Med 2002;4:45-61.
34Principe DR, Doll JA, Bauer J, Jung B, Munshi HG, Bartholin L, et al. TGF-β: Duality of function between tumor prevention and carcinogenesis. J Natl Cancer Inst 2014;106:djt369.