Varlitinib induced differential Protein Expression analysis on oral carcinoma cell line: A therapeutic approach

Receptor-ligand complex mediated signaling signicantly contributesin cellular activities such as growth, proliferation, differentiation, and survival. However, augmented expression of signal transducing receptors and ligands is the most frequent molecular event and major hallmark of oral carcinogenesis. Among these receptors, Epidermal Growth Factor Receptor (EGFR) with intracellular tyrosine kinase activity is the most frequently overexpressed molecule by Squamous epithelial cells of oral cavity. Aberrated EGFR mediated signaling has laid the foundation of targeted therapy thus providing rationale for the conducted study. We have selected EGFR pathway as targeted intracellular signaling cascade inOral squamous cell carcinoma (OSCC). Deactivating EGFR by blocking the binding sites is likely to result in prevention of intracellular downstream signaling. In this context, Tyrosine Kinase Inhibitors (TKIs) have come into play. Quinazolines (aromatic heterocyclic compounds) and their derivatives have shown promising clinical outcomes. Present study focused to investigate anti-EGFR potential of quinazoline derivative, varlitinib-a pan-EGFR inhibitor on oral squamous epithelial cell lines. We performed proteomic analyses to identify differential expression pattern of proteins in SCC-25 cells in response to varlitinib treatment. Identied proteins include Binding Immunoglobulin Protein (BiP), Heat Shock Protein 7C (HSP7C), Protein Disulde Isomerase 1 A (PDIA1), Vimentin (VIME), Keratin type I Cytoskeletal 14 (K1C14), and β-Actin (ACTB). Among these, ve proteinswere found to be downregulated upon varlitinib treatment whereas only Keratin type I Cytoskeletal 14 was upregulated. Differential expression of proteins and possible role of varlitinib as potential antitumor drug in oral carcinoma is discussed.


Introduction:
Oral squamous cell carcinoma (OSCC) constitutes a major proportionwith more than 100,000cancer related fatalitiesworld-wide in 2020 [1]. The pathophysiology of OSCC in part involves Epidermal Growth Factor Receptor (EGFR) and associated signaling pathway [2,3]. EGFR is activated upon ATP mediated phosphorylation and subsequent receptor dimerization resulting in stimulation of pleiotropic downstream signaling cascade. These include Ras/Raf/MAPKand, PI3K/Akt, that control cell proliferation, differentiation, migration, and survival [4]. Dysregulation of EGFR and receptor mediated aberrated signaling cascade has been signi cantly involved in the development and progression of OSCC, resulting in invasion, metastasis, and angiogenesis [2][3][4].High expression of EGFR has been often associated with chemotherapeutic resistance towards the drugs being currently used in clinical settings such as methotrexate, 5 uorouracil, cyclophosphamide etc. [5,6]. Other therapeutic strategies including radiotherapy have not gained substantial clinical outcomes for health improvement and overall survival [7]. These clinical ndings draw attention towards pressing need to develop new, affective, and improved therapies for oral epithelial malignancy.
Studies based on signaling mechanisms led scientists to follow targeted therapeutic approach.Identi cation of EGFR as an oncogene has made EGFR pivotal drug target for the treatment of OSCC [8,9]. Compelling investigations on EGFR therapeutic ventureshave shown two main approaches targeting EGFR i.e., monoclonal antibodies (mAbs) and Tyrosine Kinase Inhibitors (TKIs). Each of this approach has distinct mechanism; mAbs (anti-EGFR antibodies) block the extracellular ligand binding domain whereas, TKIs target intracellular tyrosine kinase domain thus, preventing RTK activity [8].
Amongst TKIs, quinazolines have been signi cantly important.Quinazoline based drugs such as ge tinib, erlotinib, lapatinib, etc. are currently being clinically and pre-clinically tested for several epithelial malignancies including OSCC [8,10] Varlitinib is an oral, pan, quinazoline based TK inhibitor of Human Epidermal Receptor (HER) with maximum e cacy in sub molar potency and minimal side effects [11]. Earlier, we reported anticancer effect of varlitinib on oral cancer cell line SCC-25 and identi ed that varlitinib mediates its action via MAPK/EGFR pathway [12]. Focus of current study was to investigate protein expression pattern in SCC-25 squamous epithelial cell line in response to varlitinib treatment using proteomic approach. heat-inactivated fetal bovine serum, 400 ng/ml hydrocortisone, 20 mmol/L glutamine, 100 U/mL penicillin and 100 µg/mL streptomycin. Cells were grown in humidi ed environment at 37°C and 5% CO2.

Cell Proliferation assessment:
After varlitinib treatment,SCC-25 cell viability was determined using MTS assay (Cell Proliferation assay kit, Abcam) as per manufacturer's instructions. Brie y, MTS reagent (20 µl) was added in each well of 96well microtiter plate, followed by 2 hours incubation at 37°C. The optical density of each well was then measured with a microplate reader (Beckman Coulter, California, USA) at 520 nm. The percentage cytotoxicity was calculated as a growth percentage of cells relative to untreated controls. All assays were performed in triplicate.

Two-Dimensional Gel Electrophoresis (2D-GE):
Total protein was extracted from varlitinib treated and untreated SCC-25 cells using lysis buffer (Urea, Glycerol, NP-40, Ampholyte buffer 3-10, 0.5M Tris-HCL pH 6.8, 0.5M DTT, 0.25M EDTA). Total protein concentration in cell lysate was estimated using Pierce BCA Protein Assay Kit (Thermo Fisher Scienti c, Waltham, MA, USA) as per manufacturer's protocol. 85µg protein from50µM varlitinib treated cells for 24 hours and untreated cellswas dissolved in rehydration buffer containing 6M Urea, 2M Thiourea, 4%CHAPS, carrier ampholyte (Bio-Rad, Hercules CA, USA) and applied on immobilized pH-gradient (IPG) gel strips (7 cm. pH 3-10; Bio-Rad) at room temperature. The strips were rehydrated overnight. Subsequently, rst dimension IEF was carried out using Multiphor II system (GE Health-care, England, UK) at 20°C.The total voltage applied was 10,000 V/hwith constant current of 2 mA.Following the IEF, focused strips were equilibrated in equilibration buffer A (6M Urea, 50 mM Tris pH 8.8, 30% Glycerol, 2% SDS with 10mg/ml DTT) and buffer B (0.5M Tris-HCl pH 6.8, 12M Urea, 10% SDS, 60% glycerol, 25mg/ml Iodoacetamide) for 30 minutes each. IPG strips were then loaded on to 12% SDS-Polyacrylamide gels. SDS-PAGE was run using Protean Cell system Bio Rad at constant voltage of 60 V. Protein spots were stained using Coomassie Brilliant Blue R-250. a) Image Analysis: Differential expression pattern of proteins from 50 µMVarlitinib treated,and control gels were analyzed via PDQuest Gel Analysis Software Version 8.0.1 (Bio-Rad). We identi ed differentially expressed protein spots among control and treated gels. Spot intensities were calculated using PDQuest software followed by excision of differentially expressed spots via EXQuest spot cutter (BIO-RAD) for subsequent identi cation.

MS/MS Analysis for protein identi cation:
Differentially expressed protein spots were digestedusing trypsin. Brie y, protein spots were de-stained with 50mM Ammonium Bicarbonate, and 50 % Acetonitrile (ACN) thrice for 15 minutes followed by dryingvia speed-vac concentrator (Eppendorf, Hamburg, Germany). Desiccated samples were rehydrated using10 mM DTT and incubated for 30 min at 37°C to reduce disul de linkages mediated by cysteine. Reduced cysteine residues were exposed with50 mM iodoacetamide in dark for 30 minutes at room temperature. Protein spots were enzymatically digested with 2ng/µL of sequence grade trypsin (Promega). Resulting peptide fragments were obtained with 25mM Ammonium bicarbonate, 10% (v/v) formic acid and acetonitrile (1:1) used in equal proportions. Subsequently, the peptide fragments were vacuum dried (Speed-Vacuum concentrator Eppendorf) and were re-suspended in 0.1% formic acid.For MS/MS analysis, peptides were subjected to Impact II UHR QqTOF, Bruker Mass Spectrometer. Protein identi cation was carried out using MGF les through Matrix Science search engine. Peptide fragmentation nger printing was opted for MASCOT search and the selected parameters were Carbamidomethyl (C) for xed modi cation, Oxidation (M) for variable modi cation, maximum 2 missed cleavages were selected, peptide mass tolerance of 20 ppm and fragment mass tolerance 0.5 Da with p < 0.05 was selected. Protein identi cation of digested peptides was carried by using Swiss-Prot data base with MASCOT search engine.

Bioinformatics:
We analyzed PPI (protein-protein interaction) ofproteins identi ed in the current study using STRING (Search Tool for the Retrieval of Interacting Genes/Proteins) database with median con dence (scores greater than 0.4).

Cytotoxic potential of varlitinib on SCC-25 cells:
Approximately 5000 cells were cultured in each well of 96 well plate. After 24 hours cells were incubated with different concentrations of varlitinib (µm Conc.) for 24, 48 and 72 hours. Cytotoxic activity of varlitinib on human oral epithelial SCC-25 cellswas identi ed by MTS cell-proliferation assay.SCC-25 cells attained 50% cell growth inhibition (median dosage affecting 50% cell population) when exposed with 50 µM varlitinib for 24 hourswith respect to their control (Figure 1), (p < 0.05). We observed linear pattern of cytotoxicity in terms of dose and incubation time ( Figure 2). We chose 50µM as cytotoxic dose for further proteomic analysis.
3.2. Analysis of differential expression of proteins in human SCC-25 cell line: Comparison of protein expression levels in SCC-25 cell line treated with 50 µM varlitinib for 24hours was examined by treating cells along with respective control. Proteins extracted from treated and untreatedcells were subjected to 2D gel electrophoresis (2D-GE)to compare differentially expressed proteins. Maximum proteins were found in the pI range(s) of 4-7 and 7-8.5with molecular weight more than 25 kDa. Respective gel images of untreated control and 50 µM varlitinib treated SCC-25 cells are shown in Figure 3.
In this study, we identi edsix differentially expressed spots via PDQuest softwarein a proteomic map of OSCC cells. The differential expression of these proteins was statistically signi cant ( Figure 4) among control versus varlitinib treated gels.We found ve proteins that happen to be downregulated in varlitinib treated SCC-25 cellswhereas oneprotein was found to be up regulated.

Identi cation of differentially expressed proteins:
Differentially expressed spots were submitted to Mass spectrometric analysis and the resulting data was analyzed using MASCOT protein identi cation search engine. A list of identi ed proteins with their molecular weight, pI, score, percent coverage and expression levels are given in Table 1. Relative spot intensities from spot 1 to spot 6 are given in Figure 4.

Correlation of identi ed proteins with oral carcinogenesis:
Interaction analyses of differentially expressed proteins identi ed via Mass spectrometric analyses were conducted using STRING network analysis tool.Based on protein interaction and pathway(s) data, obtained through STRING database and PubMed literature searches, we constructed a putative model of protein interaction networksinvolved in cellular regulating activities. Through experiments, databases, text mining and interaction with 2nd node methods in STRING, we found correlation of all differentially expressed proteins ACTB, K1C14, HSPs and VIME( Figure 5).

Discussion:
OSCC remains one of the deadliest malignancies in the past couple of decades possibly due to paucity of appropriate diagnostic and treatment modalities.Proteomic analyses with advances in mass spectrometry has paved the way by exploring novel biomarkers that could be used to timely and accurately diagnose OSCC. The aim of this study is to identify altered expression of proteins (that could be clinically useful) in OSCC upon varlitinib treatment. Cytotoxicity assessment of varlitinib showed dose and time dependent inhibitory action on cell proliferation. Growth inhibitory potential (in terms of reduced cell survival) was attained at 24 hours of treatment.
To determine the effects of varlitinib on protein expression levels in SCC-25 cells, we performed 2D-GE (two-dimensional gel electrophoresis) followed by Mass spectrometric analysis to identify differentially expressed proteins. The study exhibited 6 different proteinspots in total from varlitinib treated as well as control group of SCC-25 cells.We observed signi cant fold changes in protein expression levels for 6 protein spotsafter 24 hours of treatment. The catalogue of these identi ed proteins using massspectrometry after treatment is shown in Table 1. Each protein is brie y described in the subsequent paragraphs.
The rst group includes proteins that are correlated to evasion of apoptosis such asBiP, HSP7C and PDIA1. Binding immunoglobulin protein (BiP, also known as Glucose Regulated Protein-78; GRP-78) is a member of HSP 70 family hence encoded by HSPA5 gene [13]. BiP is an Endoplasmic Reticulum (ER) residing molecular chaperone protein. It is responsible for mediating folding, translocation of proteins, initiating unfolded protein response (UPR) and ER-associated degradation (ERAD).BiP is of utmost importance for regulating homeostasis [13]. Under stress condition(s), BiP undergoes several changes in terms of function, expression and activation which blocks translocation of proteins to ER. Subsequently causing impaired degradation resulting in pathological conditions such as cancers, cardiovascular as well as neurodegenerative disorders [13]. BiP is an established master regulator of cancer and has been found to be over-expressed in wide variety of cancers including lung cancer, liver cancer, breast cancer etc. [13,14]. Increased expression of BiP inactivatespro-apoptotic markers such as BIK and BAX andinhibits apoptosis [13,15]. BiP, through signal transducer, activates intracellular kinases thus activating cellular pathways that promote cell proliferation and survival [16]. Moreover, BiP plays a critical role in OSCC progression by inducing enhanced proliferation, chemoresistance and metastasis [17].BiP implicitly activates MAPKand PI3Ksignaling cascades facilitating cancer cell progression. [18]. BiP pathogenesis is evidenced through different studies [16][17][18][19]. It has been suggested that downregulation of BiP can be potentially linked with impeded tumor formation and growth resulting in apoptotic induction and improved survival. Our study shows down-regulation of BiP in response to varlitinib treatment suggesting therapeutic potential of the drug.
Protein disul de isomerase 1 (PDIA1) -a dithiol-disul de oxidoreductase is also an ER residing molecular chaperone [20,21]. PDIA1 primarily catalyzes disul de bond formation via its oxidizing, reducing and isomerizing capabilities thusmediating proteinfolding, translocation to ER and degradation thus contributing to maintain cellular homeostasis [20,21]. Deregulated PDIA1 expression and/or enzyme activity is often associated with different human diseases such as CVDs, neurodegenerative disorders, and different cancer types [21]. Higher expressions of PDIA1 have been demonstrated in brain cancer [22], lymphoma [23], colorectal cancer [24], breast cancer [25], often conferring metastasis, invasiveness and chemoresistance. The mechanisms causing increased expression of PDIA1 and associated pathological outcomes are poorly understood [24]. However, PDIA1 is considered as an upstream regulator for balancing Reactive Oxygen Species (ROS) as well as controlling the activity of a metalloprotease ADAM17 which in turn acts as an intermediate in EGFR signaling [25]. ADAM17 activation is based on change in redox balance -an effect which is important for growth factor dependent signaling in cancer cells. It is reported that gene silencing or knocking out PDIA1 results in induction of apoptosis and decreased cell proliferation [24]). Ourproteomics data revealed upregulation of PDIA1 in control SCC-25 cellswhile varlitinib treated group showed reduction in PDIA1 expression levels. The results remainedconsistent with previously conducted studies.
Another interesting protein found to be decreased in varlitinib treated cells was HSP7C. HSP7C is a 71 KDaheat shock related proteinwhich is constitutively expressed in cells and regulates tra cking and folding of nascent proteins. Ithas been frequently observed in human breast cancer [26] autoimmune retinopathy [27] etc.Increased expression of HSP7C due to heat, oxidative stress and hypoxia promotes cell survival, apoptotic evasion from proteolytic stress related rapid and abnormal cell growth [28]. Our proteomics data showed increased expression of HSP7C among SCC-25 cells which is in consensus with the already performed studies [28,29]. The expression was found to be decreased following 24 hours varlitinib treatment of SCC-25 cells con rming its oncogenic role.
The second group of proteins belong to the family of intermediate lament proteins including Vimentin and cytoskeletal Keratin Type 14. Vimentin plays key role in cellular, physiological,structural, mechanical, biological, and developmental processes. Vimentin is known to provide phosphorylation site(s) for kinases thereby regulating signal transduction [30]. Vimentin has been found to be present in wide range of cells including mesenchymal cells and is typically associated with Epithelial to mesenchymal transition -a prominent feature associated with tumor progression [31]. Increased vimentin expression is often associated with breast cancers [31], gastrointestinal [31] and prostate cancer [32].Knock down clones of these proteins have shown decline in carcinoma cell proliferation and decreased vimentin expression could likely lead to anti-neoplastic behavior of cells by driving them less aggressive and controlled proliferation [33][34][35]. We also observed reduced expression of vimentin in varlitinib treated SCC-25 cell line in comparison to control cells. Second protein, Keratin is also a member of the family of intermediate lament and is reported as protein marker of epithelial differentiation generating polarity which plays a signi cantly importantrole in providing mechanical strength and integrity to epithelial cells [36,37]Besides mechanical support and stability, keratins contribute in intracellular signal transduction [36,37]. We identi ed Type I Keratin cytoskeletal 14 which is typically expressed in basal layer cells of non-keratinized strati ed squamous epithelial lining [36]. K1C14 has been associated with carcinogenesis and it is usually overexpressed in different cancers [37,38]. However, we found up-regulated expression of K1C14 among varlitinib treated group of SCC-25 cellswith respect to their counter control. Our results are contradictory to previous reports and needs to be explored further to identify the exact pattern of K1C14 in response to TKI mediated treatment.   Data sets were analysed for statistically signi cant differences by one way ANOVA. * represents statistically signi cant difference (P < 0.05) between treated and untreated cells. Spot intensities of differentially expressed identi ed proteins from untreated and varlitinib treated SCC-25 cells. Error bars indicate ± SD whereas * indicates statistical signi cance (p<0.05). Figure 5