Abstract
Oral cancer etiology is complex and controlled by multi-factorial events including genetic events. Candidate gene studies, genome-wide association studies, and next-generation sequencing identified various chromosomal loci to be associated with oral cancer. There is no available review that could give us the comprehensive picture of genetic loci identified to be associated with oral cancer by candidate gene studies–based, genome-wide association studies–based, and next-generation sequencing–based approaches. A systematic literature search was performed in the PubMed database to identify the loci associated with oral cancer by exclusive candidate gene studies–based, genome-wide association studies–based, and next-generation sequencing–based study approaches. The information of loci associated with oral cancer is made online through the resource “ORNATE.” Next, screening of the loci validated by candidate gene studies and next-generation sequencing approach or by two independent studies within candidate gene studies or next-generation sequencing approaches were performed. A total of 264 loci were identified to be associated with oral cancer by candidate gene studies, genome-wide association studies, and next-generation sequencing approaches. In total, 28 loci, that is, 14q32.33 (AKT1), 5q22.2 (APC), 11q22.3 (ATM), 2q33.1 (CASP8), 11q13.3 (CCND1), 16q22.1 (CDH1), 9p21.3 (CDKN2A), 1q31.1 (COX-2), 7p11.2 (EGFR), 22q13.2 (EP300), 4q35.2 (FAT1), 4q31.3 (FBXW7), 4p16.3 (FGFR3), 1p13.3 (GSTM1-GSTT1), 11q13.2 (GSTP1), 11p15.5 (H-RAS), 3p25.3 (hOGG1), 1q32.1 (IL-10), 4q13.3 (IL-8), 12p12.1 (KRAS), 12q15 (MDM2), 12q13.12 (MLL2), 9q34.3 (NOTCH1), 17p13.1 (p53), 3q26.32 (PIK3CA), 10q23.31 (PTEN), 13q14.2 (RB1), and 5q14.2 (XRCC4), were validated to be associated with oral cancer. “ORNATE” gives a snapshot of genetic loci associated with oral cancer. All 28 loci were validated to be linked to oral cancer for which further fine-mapping followed by gene-by-gene and gene–environment interaction studies is needed to confirm their involvement in modifying oral cancer.
Keywords
Introduction
Every year, around 274,300 new cases of oral cancer (OC) are diagnosed in Asia associated with a high mortality rate due to probable delayed diagnosis of the disease. 1 The risk of OC development depends on multiple factors, such as exposure to environmental factors, human papillomavirus (HPV), alcohol, tobacco, areca nut, and genetic factors, such as an alteration in DNA.2,3
Several decades of intensive research have expanded our knowledge of cancer pathogenesis at the molecular level, thus providing new targets for early diagnosis, disease characterization, and therapy selection. Genetic profiling of individual patient is being expected to play a key role in efficient disease diagnosis, adverse drug responses, and drug treatment. Moreover, mutations within genetic loci are used for very early cancer diagnosis, prognosis, and predicting therapeutic responses.4–6 Advances in methods and technology in the field of genomics are now expected to enable construction of a comprehensive panel of genes that may be targeted for defining treatment strategies for OC.
Genetics of OC is studied at the population level following three approaches, that is, (a) candidate gene studies (CGS) which target associations between genetic variations within pre-specified genes of interest,7,8 (b) genome-wide association studies (GWAS) that, unlikely to CGS, scan the whole genome through the single-nucleotide polymorphism (SNP) array and target sequence variants with small effects,9,10 and (c) next-generation sequencing (NGS)-based studies where entire genome is deep sequenced and variants with large effects are studied.11,12 Each of these approaches identified various genetic loci to be associated with OC. However, there is no available review of the literature (Supplement S1) that could give us the comprehensive knowledge of genetic loci identified to be associated with OC by CGS, GWAS, and NGS. Absence of any relevant literature inspired us to perform a comprehensive review with the following aims: (a) to give an overview of genetic loci reported to be associated with OC by CGS-, GWAS-, and NGS-based studies and (b) to screen/identify loci validated by at least two independent CGS or NGS studies or validated by both CGS and NGS study approaches.
Materials and methods
Comprehensive search was performed (based on reported odds ratios) to know the genetic association studies of polymorphism/s associated with OC using exclusive CGS, GWAS, and NGS approaches (Figure 1 and Supplement S2; flow charts 1–4). Since the polymorphism/s in tight linkage disequilibrium (LD; r2 ≥ 0.8) with the reported OC-associated polymorphism/s could be the causal one (if two or more than two polymorphisms are in tight LD, it is difficult to dissect which among them is causal), we considered and discussed the whole genetic locus rather than individual polymorphism. (a) Studies screening by CGS approach: only those studies in which the same genetic locus was identified by at least 2 independent studies and having >500 studied subjects were picked up (single study having >500 studied subjects was not excluded); (b) studies screening by GWAS approach: studies in which same locus was identified by at least two independent studies, one from GWAS and other one/s validated by at least one CGS- or NGS-based approach were selected; (c) studies screening by NGS approach: first criterion was to screen the studies in which same locus was identified by at least two independent studies. Second criterion was to screen studies where same locus was identified by at least two independent studies, one from NGS and other one/s validated by at least one CGS or GWAS approaches. Due to limited GWAS studies on OC,10,13,14 the validation of OC-associated genetic loci was performed for CGS and NGS studies.

Flow chart describing the included/excluded literature.
The studies were screened using the following criteria.
Inclusion criteria
Original studies on human OC, squamous cell carcinoma of the oral cavity, lip, and buccal mucosa, showing genetic associations were considered. The genetic locus was screened if polymorphism/s within it was/were found to be associated with OC.
Exclusion criteria
The studies in which trait (cancer types) was not clearly mentioned.
The studies published in languages other than English.
Studies on (a) copy number changes, (b) methylation, (c) meta-analysis, (d) oral premalignant lesions, including oral sub-mucous fibrosis, and actinic cheilitis, (e) oro-pharyngeal carcinoma, and (f) oral mucosal melanoma.
Open source link development and genetic loci mapping on chromosome
For an overview of the loci screened in this study, a resource “ORNATE” was developed, which comprises the information of loci (encompassing genes) associated with OC and their underlying references (http://bmi.icmr.org.in/ornate/). ORNATE was designed using front-end technologies HTML, CSS, jQuery (JavaScript library) and back end with PHP (version 5.6) and MySql (version 5.4).
To locate the OC-associated loci on human chromosomes, mapping was done using a Genome Decoration tool (http://www.ncbi.nlm.nih.gov/genome/tools/gdp/).
Results
A total of 264 loci were found to be linked to OC by CGS (155 loci), GWAS (20 loci), and NGS (89 loci) approaches (Table S1 and “ORNATE”). Of these 264 loci, 28 loci, that is, 14q32.33 (AKT1), 5q22.2 (APC), 11q22.3 (ATM), 2q33.1 (CASP8), 11q13.3 (CCND1), 16q22.1 (CDH1), 9p21.3 (CDKN2A), 1q31.1 (COX-2), 7p11.2 (EGFR), 22q13.2 (EP300), 4q35.2 (FAT1), 4q31.3 (FBXW7), 4p16.3 (FGFR3), 1p13.3 (GSTM1-GSTT1), 11q13.2 (GSTP1), 11p15.5 (H-RAS), 3p25.3 (hOGG1), 1q32.1 (IL-10), 4q13.3 (IL-8), 12p12.1 (KRAS), 12q15 (MDM2), 12q13.12 (MLL2), 9q34.3 (NOTCH1), 17p13.1 (p53), 3q26.32 (PIK3CA), 10q23.31 (PTEN), 13q14.2 (RB1), and 5q14.2 (XRCC4), were validated to be associated with OC by CGS and NGS approaches (Figure 2 and Table 1). For better understanding of loci distribution, the validated loci were mapped on human chromosomes (Figure 3).

Validated 28 oral cancer–associated genetic loci. On x-axis is the associated genetic loci and their representative gene, whereas on y-axis is the number of identified associated studies.
List of chromosome locus and their corresponding oral cancer–associated and validated gene.
CGS: candidate gene studies; NGS: next-generation sequencing.
The studies represent the total number of articles screened where the corresponding genetic locus (encompassing gene) was found to be associated with oral cancer. Subjects are dealt with the sum of the number of participants from the screened studies.

Validated 28 oral cancer–associated genetic loci mapped on the human chromosomes. Information of genetic loci and their encompassing gene mapped on the figure can be found in Table 1. References underlying the loci are available on “ORNATE.”
A total of 172 OC-associated loci were identified by CGS approach (Table S1 and Supplement S2; flow chart 2); of which 14q32.33 (AKT1), 5q22.2 (APC), 11q22.3 (ATM), 2q33.1 (CASP8), 16q22.1 (CDH1), 7p11.2 (EGFR), 4p16.3 (FGFR3), 11p15.5 (H-RAS), 12p12.1 (KRAS), 9q34.3 (NOTCH1), 17p13.1 (p53), and 13q14.2 (RB1) were validated by NGS (Table 1) and 18q21.2 (DCC), 3p14.2 (FHIT), 6q16.3 (GRIK2), 16q12.2 (MMP-2), and 2p22.3 (RASGRP3) by GWAS (Table S1). Hence, overall, 155 loci were identified only in CGS.
A total of 20 loci were found to be linked to OC via GWAS approach (Table S1); of which 18q21.2 (DCC), 3p14.2 (FHIT), 6q16.3 (GRIK2), 16q12.2 (MMP-2), and 2p22.3 (RASGRP3) were further validated in studies by CGS approach.
Through NGS-based approach, 89 OC-associated loci were identified (Table 1); of which 9p21.3 (CDKN2A), 22q13.2 (EP300), 4q35.2 (FAT1), 4q31.3 (FBXW7), 12q13.12 (MLL2), 3q26.32 (PIK3CA), and 10q23.31 (PTEN) were found in at least two independent studies and 14q32.33 (AKT), 5q22.2 (APC), 11q22.3 (ATM), 2q33.1 (CASP8), 16q22.1 (CDH1), 7p11.2 (EGFR), 4p16.3 (FGFR3), 11p15.5 (H-RAS), 12p12.1 (KRAS), 9q34.3 (NOTCH1), 917p13.1 (p53), and 13q14.2 (RB1) were also confirmed in studies following CGS approach.
Conclusion
The “ORNATE” gives the status of studies from literature in which genetic variants (within loci) were found to be associated with OC. With the non-availability of any resource giving systematic information about genetic loci associated with OC, “ORNATE” may be a useful tool for comparing/deciding experiments on OC genetic association studies.
In this study, 28 loci were validated to be linked to OC via CGS and NGS approach. Genetic alterations are considered to be important in the development and progression of OC, leading to dysregulation of essential cellular signaling pathways such as the Ras-MAPK-ERK (Ras–mitogen-activated protein kinase–extracellular signal–regulated protein kinase), the PI3K/Akt/mTOR (phosphoinositide 3-kinase/Akt/mechanistic target of rapamycin), the JAK/STAT3 (Janus kinase/signal transducer and activator of transcription 3), and the PLCγ/PKC (phospholipase C-γ/protein kinase C).67,68 In addition to mutation correlates of OC, several studies identifying differential gene expression patterns in OC patient samples are available.69,70 Integrating mutation and gene expression data set to identify patterns of molecular events relevant to initiation and progression of tumor may be of importance for OC diagnosis, prognosis, and therapy. 71
The tumor suppressor gene p53 is commonly mutated in OC.72,73 Alterations in the p53 gene make it functionally inactive in oral tumors. Moreover, functional restoration of the p53 gene in OC cell lines and in animal models of OC has shown to relapse oral carcinogenesis. 74 It is suggested that p53 polymorphism together with HPV infection increases the susceptibility of OC. 75 Another important tumor suppressor gene NOTCH1 is reported to be mutated in 54% for oral squamous cell carcinoma (OSCC) and 60% for preneoplastic lesions in Chinese patients. 57 NOTCH1 mutations and altered expression are proposed to be the drivers of OSCC progression.57,76 The MDM2 gene, a proto-oncogene promoting tumor formation by targeting tumor suppressor proteins, such as p53, is amplified in 25%–40% of all human cancers. 77 The co-expression of p53/MDM2 proteins is suggested to be an indicator of aggressive tumor behavior in OSCC. 78 The proto-oncogene Akt1 promotes cell proliferation, survival, and metastasis of cancer cells and its overexpression relates to the progression of OC. 79
In the RAS gene family, involved in the molecular pathogenesis of OSCC, mutations in the HRAS gene are reported to be the highest in OC. Importantly, HRAS mutations are common in Asian patients who chew betel quid. 80
Expression of the oncogene PIK3CA increases in OSCC. 81 However, the association of SNPs with the expression level has not been observed. 82 This may be due to the occurrence of somatic mutations of the PIK3CA late in the development of OSCC and tumor progression occurring through the PI3K-AKT signaling pathway. 83 The frequency of PIK3CA and HRAS mutation in Asian OC patients is similar to the Caucasian OC patients, despite differences in their risk habit exposure as betel quid chewing versus smoking and alcohol drinking, respectively. 84
The epidermal growth factor receptor (EGFR) is involved in the regulation of cell proliferation, metastasis, migration, invasion, angiogenesis, and inhibition of apoptosis. Overexpression of EGFR is associated with aggressive phenotype and poor prognosis in OC. 85 Cetuximab, a monoclonal antibody targeting EGFR, is a Food and Drug Administration (FDA)-approved drug in OSCC, with a significant effect in combination therapy. 86
Overexpression of cyclooxygenase-2 (COX-2), enzyme involved in the production of prostanoids from arachidonic acid, results in enhanced synthesis of prostaglandins promoting angiogenesis, apoptosis, and cancer progression. 87 Polymorphisms and expression of the COX-2 gene are associated with the pathogenesis of OSCC and may be of prognostic significance.87–89
Deregulation of CASP8, a key controller of apoptosis, is crucially involved in carcinogenesis. Importantly, polymorphisms in the CASP8 are suggested to play a protective role in the development of tumor in OSCC patients. 24
Interleukin (IL)-8 and IL-10 cytokines play an essential role in OC. IL-10 is an immunosuppressive cytokine, while IL-8 is a pro-inflammatory cytokine both involved in the process of angiogenesis. These cytokines, studied at both mutation and expression level, may serve as biomarkers of prognostic value.90–93 The heterogeneity in cancer cells between and/or within patients makes it a complex disease to cure. 94 Identifying germ line mutations which may serve as biomarkers for predicting OC-susceptible individuals and/or who may respond differently to a particular cancer therapy is very important. 4 In addition to this, studying genetic mutations and aberrations in tumor cells, all together provide a window of opportunity for defining treatment strategies for OC. Recent data demonstrate that OC is a mutationally heterogeneous class of cancer. Efforts to improve disease prognosis by combining genetic information from NGS with traditional clinicopathological prognostic parameters are ongoing. 5 Precision oncology goals to define treatment strategies based on patient-specific somatic mutations and abnormal molecular pathways. However, discrepancies in data sets may be accepted due to differences in disease stage, exposed carcinogen, ethnicity, and genetic background of individual patient. 5 Recent research focuses on OC subtype stratification, and molecular characterization may help in prognosis and defining personalized treatment strategies. In addition, retrospective study design with formalin-fixed paraffin-embedded samples paved the path to study the prognostic significance of common genetic alterations. Such study has aided in identifying a prognostic gene signature, HRAS, BRAF, FGFR3, SMAD4, KIT, PTEN, NOTCH1, AKT1, CTNNB1, and PTPN11, that predicts both disease-free survival and overall survival. 16
On 28 validated hotspots, further fine-mapping on larger cohorts, based on the 1000 Genomes Project data sets, 95 is needed to confirm already known associations and find novel genetic variants associated with OC. Moreover, gene-by-gene interaction analysis 96 using complex statistical models 97 may also help to comprehend the genomic network regulating oral carcinogenesis. Polygenic disease, such as OC, is also influenced by environmental factors, including alcohol, betel quid, and smoking, which along with genetic variations plays a modulating role in determining an individual’s susceptibility to disease. 3 Statistical methods using multifaceted gene–environment analysis are also required to unravel the complex biological mechanism/s modulating OC. 96
The holistic approach used for screening studies on OC-associated loci could be utilized for studying other cancer subtypes, but for now, well-defined studies on larger cohorts of different ethnicity needs to be performed. In order to improve patient care with targeted therapies leading to more accuracy and specificity, the combined analysis of the so far available genetic data sets is required. This review is a step toward this direction with an aim to filter down the genetic loci from the available data set in the public domain which may be of clinical relevance in defining strategies pertaining to OC treatment.
Footnotes
Acknowledgements
The authors are thankful to Dr Shashi Sharma and Dr Diksha Bhartiya from National Institute of Cancer Prevention and Research, NOIDA, India for their critical suggestions on the manuscript. V.S. wrote the first draft of the review and along with H.S. conceived the study. A.N. performed bioinformatics, reviewed, and edited the manuscript. A.K.S. created the “ORNATE.” R.M., D.N.S., and M.B. lead, managed, and wrote the final version of the manuscript. The article was scanned for plagiarism using online plagiarism software “https://www.plagiarismsoftware.net/” and the grammar was checked by “
.”
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by the Indian Council of Medical Research by grant no. BIC/11(25)/2015.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
