Evaluation of Vimentin as a Potential Poor Prognostic Indicator and Salivary Biomarker for Oral Cancers and Pre-Cancers by Mass Spectrometry Based Proteomics

Background : Oral tongue squamous cell carcinoma (OTSCC) is an aggressive cancer with high morbidity and mortality rates, despite multimodality management. There are currently no clinically relevant molecular markers to help identify patients at a higher risk of recurrence and failure. Methods: 2D-DIGE coupled with tandem mass spectrometry was performed on tissues obtained from early staged OTSCC along with its paired apparently adjacent normal tissue samples (n=10). Top upregulated protein was validated using another independent set of tissue samples by Immunohistochemistry (n=346), comprising of retrospective early stage OTSCC (n=150) and prospective series of oral pre-cancers, normal and oral cancers (n=195). For further validation of protein expression, saliva samples collected from Oral Cancer and pre-cancer samples were analysed by ELISA (n=80). Results: We found vimentin, the mesenchymal protein to be the most upregulated protein in tongue tumour tissues compared to adjacent apparent normal tissues. Vimentin was found to be signicantly overexpressed in oral pre-cancers along with cancers compared to normal tissues. Conclusion: Vimentin detection in saliva can be useful diagnostic test to detect oral precancers that may have malignant potential needing closer follow up. Salivary ELISA for vimentin can additionally be useful for disease monitoring in oral cancers. of the cell, which is a promising approach for the identication of novel protein biomarkers. These proteins can be used as key targets for therapeutic intervention and also as promising candidates for early detection of cancers[4, 5]. We have some recent studies showing the preliminary application of proteomics for the identication of biomarkers for OSCC [6, 7]. Comparison of protein expression proles between OSCC and normal cell lines or tissues has revealed replicable and signicant changes in the expression levels of number of proteins, including some metabolic enzymes, modulators of signal transduction pathways, and oncoproteins[8]. In the current study, we have performed 2D DIGE based proteomic proling coupled with mass spectrometry approach and have validated the expression of top upregulated protein Vimentin, eventually to explore the prognostic stratication of early staged OTSCC. We have additionally studied the vimentin expression in Oral precancers and cancers comparing with normal tissues. Salivary ELISA for vimentin has been attempted to evaluate the secretion of vimentin in samples from normal healthy volunteers comparing with saliva from patients presenting with oral leukoplakia, OSMF and Oral squamous cell carcinoma. OTSCC tumour pooled adjacent uninvolved proteomic proling. labelling) tissue labelled pooled tissue proteins samples The nal volume for all preparations was adjusted to a total of 340 uL with rehydration buffer (7 M urea, 2 M thiourea, 1% IPG buffer, 50 mM DTT, 4% CHAPS, and a trace amount of bromophenol A reciprocal labelling experiment was also performed. the percentage grade and intensity score with the IR score ranged from 0 to 12. The immunoreactivity was divided into three groups on the basis of the nal score: negative immunoreactivity was dened as a total score of 0, low immunoreactivity was dened as a total score of 1–4, and high immunoreactivity was dened as a total score >4. The immunostaining of the tumor invasive front was evaluated using the same method as mentioned for tumor areas. early a poor prognostic role of vimentin in oral precancers and in vimentin useful oral determining prognosis early precancer Vimentin be good biomarker because none of the normal buccal mucosa tissues vimentin as of vimentin is indicative of mesenchymal transition. current study showed aberrant vimentin expression in oral discovery in the current and have shown that proteins involved in cytoskeletal remodelling that are involved in the process of tumorigenesis. Proteomics approach can give rise to several markers, as shown in the current study but these biomarkers need to be chosen as per their clinical relevance based on validation studies Our studies has shown additional 9 top differentially regulated proteins that can be subsequently validated in oral cancer tissue and saliva samples. Along with vimentin, we also obtained laminin A/C transcript, myoglobin as signicant markers by proteomics approach and they are well known stromal components playing important role in ECM modulation. Vimentin has been evaluated as a useful marker for aggressive pathology and poor prognosis in tongue cancers, its utility can be explored in patients to assess patients who are more likely to fail treatment, despite being early staged. T1 T2 are adverse factors, there still an unanswered need objective molecular useful to patients further attention.

Histologically apparently normal adjacent tongue tissues along with paired early staged OTSCC tumour tissues (n=10) were obtained from patients presenting with OTSCC and undergoing surgery. Formalin xed para n embedded samples from buccal leukoplakia (n=50), oral cancers (n=71) were obtained from Sree Balaji Dental College and hospital and formalin xed tissues from oral submucous brosis samples (n=32) and normal buccal mucosa tissues (n=42) were obtained during third molar extraction were collected from Sree Ramachandra Dental College and Hospital. This was an independent cohort (n=195) for validation studies. Saliva samples (n=80) were collected from oral cancer patients (n=45), patients with oral potentially malignant lesions (n=15) and normal (n=20) obtained from Saveetha Dental College and Hospital and Sree Ramachandra Dental College and Hospital. Additionally, formalin xed para n embedded sections from retrospective series of exclusively early staged tongue cancer patients [T1 and T2] (n=150) were obtained from Cancer Institute WIA who had been treated between 1995 to 200 for validation studies of the ndings with the complete treatment follow up. All the FFPE sections (n=345) used for the study were histologically examined by oral pathologist VS and PR.
Patient Saliva samples ELISA was done for saliva samples (n=80) collected from Oral cancer patients, patients presenting with oral potentially malignant lesions and absolute normal volunteers. The study participants were requested to refrain from drinking, eating, chewing tobacco or smoking 1 hour prior to the collection of saliva. After obtaining the informed consent of the patient, 0.5 to 1 ml of whole unstimulated saliva was collected by passive expectoration and patients were asked to spit into a 50-mL sterile tube containing 10µL of proteinase inhibitor (Proteinase inhibitor cocktail P2714 Sigma Aldrich). The saliva samples were transferred to 1.5-mL sterile microtubes and centrifuged for 3 minutes at 13,000 rpm. Supernatants, separated from the cellular phase, were immediately aliquoted and stored at -80 o C within 60 minutes after saliva collection.

Proteomics -Proteins Labelling with CyDyes
Pooled OTSCC tumour and pooled adjacent uninvolved tissues were used for the proteomic pro ling. Lysine labelling protocol, (minimal labelling) used in this study is described before [9]. The processed tissue proteins were labelled individually with dyes Cy3 and Cy5 while the pooled tissue proteins prepared by mixing equal aliquot of protein from all samples in an experimental set up were labelled with Cy2. The nal volume for all preparations was adjusted to a total of 340 uL with rehydration buffer (7 M urea, 2 M thiourea, 1% IPG buffer, 50 mM DTT, 4% CHAPS, and a trace amount of bromophenol blue). A reciprocal labelling experiment was also performed.
Proteomics -2D gel electrophoresis Two-dimensional gel electrophoresis of CyDye labeled proteins was done as described before [10,11] with the following modi cations. Eighteen cm IPG strips of pH 4-7 (GE Healthcare, Uppsala, Sweden) was employed in the rst dimension. Labelled proteins were focused for a total of 80,000 Vhs at a constant temperature (20 o C) under linear voltage ramp after an active IPG rehydration at 30 V in a IPGPhor III (GE Healthcare, Uppsala, Sweden) apparatus. Following IEF, each IPG strip was placed in the equilibration buffer containing 2% DTT rst followed by incubation in another buffer in which the DTT was replaced by 2.5% iodoacetamide. The second dimension PAGE (12.5%) was carried out in an EttanDaltSix systems (GE Healthcare, Uppsala, Sweden) at 1 W/gel for 1 hr and 13 W/ gel for 5 hr. All experimental procedures were performed in dim light or in the dark.

Protein Visualisation and DeCyder Image analysis
The protocols for the protein visualization and image analysis using DeCyder has been mentioned previously [12]. Brie y, after second dimension electrophoresis, the gels were scanned with Typhoon FLA 9500 Variable Mode Imager (GE Healthcare, Uppsala, Sweden). Cy2, Cy3 & Cy5 images were captured using the settings recommended by the manufacturer. A DeCyder differential in-gel analysis (DIA) module was used for image analysis between samples within the same gel while a DeCyder biological variation analysis (BVA) module was performed for pairwise image analysis among multiple gels. Student's t-test and ANOVA were used to compare the average spot volume and differences of protein abundance for all detectable spots between the tumor and normal groups. Reciprocal dye labelling was performed to normalize bias in labelling.

Protein Identi cation and Mass Spectrometry (MS)
Pooled tongue tissue proteins (250 µg) were separated on 18 cm IPG strips of pH 4-7 in the rst dimension. First and second dimension electrophoresis were done as given under 2D DIGE method. The second-dimension gels were stained with colloidal coomassie blue G-250 and gel spots from this preparative gel were excised manually for in-gel trypsin digestion and LC-MS/MS was performed. Extracted peptides were dried under vacuum for 90 min and stored at 4 °C. Zip tip puri ed peptides were analysed using nano-RPLC (Thermo Scienti c, USA) coupled with an Orbitrap Elite Mass spectrometer (Thermo Scienti c, USA). Peptides were ionized by positive mode electrospray with an ion spray voltage of 1.9 kV. The MS data were acquired in positive ion mode over mass range m/z 350-4000 Da using Xcalibur software (version 2.2.SP1.48) (Thermo Scienti c USA). MS data were analysed using Proteome Discoverer software v.1.4 (Thermo Scienti c) using Sequest algorithm with database downloaded from Uniprot as described earlier [12]. The combined list of o cial gene symbols corresponding to the identi ed proteins was used for input. We used STRING (www.string.db.org) [13] network construction.

Immunohistochemistry (IHC)
The IHC detection methods was as mentioned previously [14]. Brie y, IHC for vimentin was performed on 5 μm sections of FFPE tissues. The sections were depara nized in xylene and rehydrated in absolute ethanol. Antigen retrieval was done with 0.05M citrate Buffer (pH 9) in pressure cooker for 20 minutes. Endogenous peroxidase activity was blocked by incubation in 0.03% hydrogen peroxide in distilled water for 10 minutes and then washed with phosphate buffered saline (PBS). Sections were counterstained with hematoxylin, dehydrated, and mounted in DPX. Positive controls and negative controls were included appropriately where primary anti-body was replaced with 2% BSA in negative control. Immunostaining of the sections was reviewed with the corresponding haematoxylin and eosin-stained sections.

IHC Scoring
Immunohistochemical scoring for the target was done as described earlier [15]. Brie y, the percentage grade of stained tumor cells was scored as 0, negative; 1, <10%; 2, 11-50%; 3, 51-80%; or 4, >80% positive cells and the intensity of stain was scored as 0, negative; 1, weak; 2, moderate; or 3, strong. The immunoreactive (IR) score was calculated as a product of the percentage grade and intensity score with the IR score ranged from 0 to 12. The immunoreactivity was divided into three groups on the basis of the nal score: negative immunoreactivity was de ned as a total score of 0, low immunoreactivity was de ned as a total score of 1-4, and high immunoreactivity was de ned as a total score >4. The immunostaining of the tumor invasive front was evaluated using the same method as mentioned for tumor areas.

ELISA in Saliva Samples
The RayBio® Human Vimentin ELISA (Enzyme-Linked Immunosorbent Assay) kit is used to quantify the expression of vimentin in saliva samples (n=80), of which saliva samples from patients with oral cancer (n=45), patients with oral potential pre-malignant lesions (n=20) and healthy volunteers (n=15) were used. This is an in vitro enzyme-linked immunosorbent assay for the quantitative measurement of human Vimentin in saliva samples. This assay employs an antibody speci c for human Vimentin coated on a 96-well plate. Standards and samples are pipetted into the wells and Vimentin present in a sample is bound to the wells by the immobilized antibody. The wells are washed and biotinylated anti-human Vimentin antibody is added. After washing away unbound biotinylated antibody, HRPconjugated streptavidin is pipetted to the wells. The wells are again washed, a TMB substrate solution is added to the wells and color develops in proportion to the amount of Vimentin bound. The Stop Solution changes the color from blue to yellow, and the intensity of the color is measured at 450 nm. The standard graph was plotted with the vimentin standard protein provided in the kit. Using the standard graph the protein concentration were extrapolated for the unknown OD values obtained from saliva samples obtained from patients and healthy volunteers.

Statistical analysis
The relative levels of stained protein spots compared with the internal standard spots were analyzed by DeCyder Difference In-gel Analysis (DIA) and DeCyder Biological Variation Analysis (BVA) software modules (GE Healthcare). Student's t-test was used to calculate statistically signi cant differences between 2 groups in relative abundance of individual protein spots among the groups in 2D-DIGE. P<0.05 was considered statistically signi cant. Other statistical analysis was done using SPSS (IBM Corporation version 16).

Results
Quantitative proteomics using 2D DIGE and Mass Spectrometry in OTSCC shows vimentin as the most upregulated differentially expressed protein in early staged tongue cancer Comparative proteomic analysis of pooled tongue tissue samples obtained from OTSCC patients compared to pooled adjacent apparent normal samples are shown in Fig. 1. More than 95 protein pairs were obtained in the image analysis platform, among which 45 were upregulated in tumour samples compared with adjacent normal protein samples. Out of the differentially expressed proteins, top 10 differentially expressed spots were taken for mass spectrometry. Table 1 describes the protein identi cation details for the differentially regulated proteins with the accession number matched in the database, mass spec probability score, percentage of sequence coverage match. Table 2 describes the average fold ratio with ANOVA value for differentially regulated proteins. The top 10 differentially regulated proteins have been listed in Table 2, in which eight proteins were signi cantly upregulated and two proteins were signi cantly downregulated in tumour samples. Quantitative 2D-DIGE proteomic approach coupled with tandem mass spectrometry identi ed, Vimentin as the topmost upregulated protein in OTSCC (Fig. 1a).  The average ratio of upregulation of vimentin was 4.89 folds higher compared to the normal counterpart. The one-way ANOVA p value was found to be 0.00064 (Fig. 1d). The log standardized abundance was signi cantly high compared to the normal tissue protein (Fig. 1e).
Functional classi cation of the identi ed Vimentin protein and biological network analysis reveals its signi cance in tumorigenesis The STRING [13] cluster analysis revealed that Vimentin forms a strong protein interaction with other partners, comprising of three major networks; rst cluster is with most of the Small Nuclear Ribonucleo Proteins (SNRP) proteins; second cluster with different members of tropomyosin and third cluster with many caspases (Fig. 2). All these protein networks are known to have a key role in different tumorigenesis pathways; thus, this interaction analysis describes the signi cance of the identi ed Vimentin protein's regulation in tumorigenesis process.

Validation of Vimentin Expression to evaluate the role in early staged OTSCC
To validate the expression of Vimentin and its role in OTSCC, we undertook a retrospective cohort of exclusively early staged (T1 and T2) OTSCC patients (n = 150) treated as per the decision of the multispeciality board between 1995 to 2007. The clinicopathological features of early staged OTSCCs analysed based on vimentin protein expression is shown in Table 3. Median age of the cohort was 55 years and median OS was 74 months and DFS was 22 months. The pattern of vimentin expression was predominantly cytoplasmic.
Positive immunoreactivity for Vimentin was identi ed in 58 (38.6%) patients and Vimentin was negative in 92 (61.31%) patients. Vimentin expression at Invasive front of the tumours indicated failure of treatment and locoregional recurrence in early staged OTSCC Vimentin at ITF was positive in 47.6% (50/105) of patients whose tumours showed an invasive tumour front. Among the patients who showed locoregional recurrence with ITF in the tumours, 68% (17/25) showed a positive vimentin at ITF compared to 32% (8/25) patients whose tumours ITF had negative immune-expression as shown in Table 4. Patients undergoing modi ed neck dissection for neck management had the best survival among the early staged OTSCC Patients whose tumours had positive vimentin expression had an decreased DFS compared to the patients whose tumours had negative vimentin expression and this association was statistically signi cant (Log Rank = 4.068 ; p = 0.044) (Fig. 3a). The patients in this cohort were subjected to upfront neck management, namely neck node observation (n = 73), modi ed radical neck dissection [MRND] (n = 33) and radiation to neck (n = 44). Kaplan Meier survival curves of DFS based on upfront neck management showed that patients who underwent MRND having a better survival (p = 0.006; log rank = 10.094) compared to patients given radiation to neck. (Fig. 3b)

Vimentin expression correlates with severity of dysplasia in oral pre-cancers and indicates malignant potential
Based on the ndings of the retrospective study, we undertook another cohort of patients (n = 196) with oral leukoplakia (n = 50), OSMF(n = 32) and invasive cancers involving buccal cavity (to ascertain if Vimentin can useful in oral cancer inclusive of both sites oral tongue and buccal cavity. This cohort had 82 oral precancers comprising of leukoplakia (n = 50) and OSMF (n = 32) along with normals (n = 42) and Cancers of buccal cavity (n = 72). Table 5 shows the demographic details of subjects and patients who presented with oral pre-cancers at different stages and oral cancers presenting to the dental clinic. The median age of this group was 45 years.   Table 5 describing the positive expression of vimentin having a signi cant association (p = 0.000 : χ 2 = 68.524) with oral precancers and cancers compared to apparent normals. Interestingly, vimentin was signi cantly correlated to the dedifferentiated state of the oral precancers with oral cancers (p = 0.000 ; χ 2 = 77.037). Vimentin is a signi cant biomarker for oral precancers that may have an aggressive potential to turn into malignancy. Figure 4 shows the IHC based patterns of vimentin expression in Oral cancers.
Vimentin secretion is signi cantly elevated in saliva samples obtained from Oral Cancer and Oral pre-cancer patients compared to healthy volunteers Since tissue availability in oral cancers and pre-cancers involve invasive procedures as a biopsy, we wanted to evaluate if vimentin could be detected in saliva, as saliva can serve as a non-invasive medium for early detection and also for disease monitoring. Vimentin secretion in saliva could be detected as signi cantly high in Oral cancer and pre-cancer patients. The concentration obtained in salivary ELISA for Vimentin detection range from 2.3-6.4 ng/ml for healthy volunteer samples, for precancer samples, the concentration ranges from 4.5-16.8 ng/ml and for cancer samples, the concentration ranges from 4.8-127 ng/ml. The fold increase between healthy volunteer sample and precancer samples were statistically signi cant. The differential expression was analysed using SPSS software and observed to be statistically signi cant and the data was represented in Fig. 5a. The diagnostic potential of the underlying pathological implication could be detected by ROC curve analysis showing AUC = 0.8 which had a high statistical signi cance (Fig. 5b &  5c).

Discussion
Vimentin is a well-known mesenchymal protein acting as a scaffold for signalling proteins that are important for cancer cell invasion [16] wound healing, tissue repair [17] tissue ageing and apoptosis [18,19]. This study describes the role of vimentin that emerged as the most upregulated protein in OTSCC by quantitative proteomics. We have validated the expression of vimentin in early staged OTSCC inferring its role as a poor prognostic indicator. The role of vimentin was further evaluated in oral precancers and in saliva indicating that vimentin can be a very useful marker in oral cancer for determining prognosis and can be used for early detection of the disease in saliva and Oral precancer tissues. Vimentin can be good biomarker because none of the normal buccal mucosa tissues expressed vimentin as expression of vimentin is indicative of mesenchymal transition. The current study showed aberrant vimentin expression in oral premalignant lesions and oral cancers.
Previous study done as a meta-analysis of differentially expressed genes in OTSCCs has shown the comprehensive expression pro ling of genes identifying the role of extracellular matrix with EMT based deregulation in OTSCC showing the role of tumour microenvironment in OTSCC with a number of extracellular matrix (ECM) components playing a crucial role in patient prognosis [14].
As the rst step, we evaluated the global proteomic pro les in early staged oral tongue cancer samples by 2D DIGE followed by mass spectrometry analysis. 2D-DIGE technique enables direct comparison of protein pro le between tumor and normal samples on the same single 2D gel, thus reducing technical variability which could affect the expression pattern of proteins. Recent studies have used the conventional 2D electrophoresis [7] and identi ed a panel of 12 proteins in tongue cancer, but absence of validations in normal tissues in the same gels can possibly lead to biased conclusions owing gel-to gel variations. To overcome this, we attempted 2D DIGE based proteomic discovery in the current study and have shown that proteins involved in cytoskeletal remodelling that are involved in the process of tumorigenesis. Proteomics approach can give rise to several markers, as shown in the current study but these biomarkers need to be chosen as per their clinical relevance based on validation studies Our studies has shown additional 9 top differentially regulated proteins that can be subsequently validated in oral cancer tissue and saliva samples. Along with vimentin, we also obtained laminin A/C transcript, myoglobin as signi cant markers by proteomics approach and they are well known stromal components playing important role in ECM modulation. Vimentin has been evaluated as a useful marker for aggressive pathology and poor prognosis in tongue cancers, its utility can be explored in patients to assess patients who are more likely to fail treatment, despite being early staged.
To validate this nding, as a second step, we evaluated the vimentin expression in retrospective series of exclusively early staged OTSCC. Early staged OTSCC need a biomarker to identify the patients who are more likely to fail despite being in T1 or T2 stage. The aggressive nature of OTSCC is re ected by the increased rates of local recurrence, occult node and distant metastasis. Though several histological features like extracapsular spread, perineural invasion, and presence of lymphovascular emboli are adverse factors, there is still an unanswered need of objective molecular markers that can be useful to identify patients needing further attention.
This study showed that upfront neck management was an important factor to predict event free survival. Patients who underwent modi ed neck dissection had the best overall survival among the early staged OTSCC showing the importance of neck dissection in OTSCC. This result was in agreement to the previous randomised controlled trial showing elective neck dissection showed higher rates of overall and disease-free survival in early staged OTSCC [20]. We did not nd a signi cant correlation with overall survival and vimentin expression unlike the disease-free survival which is similar to previous reports [21]. Positive vimentin expression was found to be associated with increased stage, increased size of the tumour, increased treatment failures, increased locoregional recurrence and poorer disease-free survival. Earlier studies have identi ed vimentin over-expression to be a poor prognostic indicator in OTSCC by univariate analysis [21][22][23][24]. Our study con rms the earlier ndings. It has been suggested that cancer cells present in the invasive tumour front (ITF) are more aggressive in terms of their metastatic potential [25]. Our study emphasises that expression of vimentin assessed at ITF can indicate the EMT switch which is known to be associated with increased motility and invasiveness. We found a signi cant association with locoregional recurrence and vimentin expression at ITF. Vimentin at ITF and tumour sites has been shown to be strongly correlated to aggressive phenotype contributing to poor prognosis [26]. Aberrant expression of vimentin has been incriminated in various epithelial cancers [27][28][29][30][31][32] including OSCC [33]. Vimentin has been shown as a predictive biomarker for tumor growth and metastasis, although its understanding is limited in OTSCC prognosis [34,35].
As a third step, we wanted to evaluate the signi cance of vimentin in oral pre-malignant lesions that would comprise of leukoplakia, OSMF as well as buccal cancers. We wanted to evaluate if Vimentin can be a biomarker for oral cancers, comprising of both the major sites buccal as well as tongue. We found a signi cant association of vimentin expression in oral precancers and cancers. Increased age was correlated to the vimentin expression mainly because of higher incidence of cancers in subjects with increasing age. Most of the potentially malignant disorders are asymptomatic and treatment can be of three types namely close observation, surgical excision/ablation and medical treatment. There is a lack of standardised diagnostic criteria in visual inspection of oral cavity to identify potentially malignant lesions that may eventually progress. Previous studies have shown Vimentin expression in lesions of leukoplakia and submucous brosis could be an early event in tobacco and areca nut associated tumorigenesis process.
As the fourth step, we have evaluated vimentin expression in saliva as a non-invasive means and have shown that vimentin can be a good marker for both early detection and disease monitoring in oral cancers. Our current study con rms this nding with a higher vimentin expression in saliva samples using ELISA method with a signi cant diagnostic potential for identifying patients with poor prognosis. This can be validated in higher number of samples of oral pre-cancers and cancers for early detection and disease monitoring. To our knowledge this is the rst study evaluating vimentin in saliva from our country. Vimentin has been shown to be secreted by distinct population of vascular endothelial cells and activated macrophages and can accumulate in the blood previously [36,37].
Vimentin secretion has been shown to be induced by pro-in ammatory cytokines TNF-α, and LPS suggesting that vimentin secretion is an in ammatory response [38]. We found that vimentin secretion was indeed very elevated in certain oral precancer samples that they could be an in ammatory response and more prone for malignant transformation in future.
A recent report reviewing all the promising biomarkers identi ed in tongue cancers have shown vimentin as one of the strongest biomarker with signi cant relevance as a marker with clinical utility [39] proving that it is an important marker of OTSCC which innately shows a higher propensity to metastasize con rming our reports. As per our ndings Vimentin in pre-cancers can help identify patients most likely to progress to malignancy as is an useful early detection marker. In recent days, the non-invasive nature of saliva and its signi cant relationship with plasma levels made saliva an very attractive diagnostic tool [40].Up to our knowledge, this is the rst report to describe the status of vimentin expression in saliva samples obtained from precancer, cancer patients with oral squamous cell carcinoma. Oral cancer is a very common cancer in Indian population and there were very few studies aimed to identify the biomarkers with validation study. We have identi ed the candidate proteins altered in our Indian population and validated them with both IHC and ELISA analysis.

Conclusion
In conclusion, 2D DIGE coupled with tandem mass spectrometry was found useful to identify differentially expressed proteins in OTSCC tissues. All the quantitative tissue proteomics-based markers identi ed in current study needs validation in OTSCC tissues as a prospective study with larger numbers of samples. The current study has been pursued for vimentin, a well characterised EMT marker. It was found clinically relevant to prognosticate early OTSCC patients most likely to fail treatment, requiring speci c tailored treatment.
Vimentin was also useful as an early detection biomarker of precancers in oral cavity. In addition, the vimentin protein expression has been validated in saliva samples obtained from precancer and cancer patients and found to be signi cantly upregulated when compared to normal samples, proving its role as a useful biomarker for early detection and disease monitoring.

Declarations Ethics approval and consent to participate
For all the samples used in this study, the ethics approval had been obtained and the details were enclosed in the methodology section.
The consent to participate in the study was obtained from all the patients.

Consent for publication
Not Applicable Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Con ict of Interest Statement Representative 2DE gels of Normal and Tumor sample, 3D view, graphical representation and fold ratio analysis for differentially regulated proteins Figure 2 Cluster Analysis for the differentially regulated protein Vimentin using STRING database