EphA2/SHP2/SOX8 Axis: A Novel Target for Regulation of Migration and Cetuximab Treatment Sensitivity in Oral Squamous Cell Carcinoma Authors: Dayong Yan, Lele Guo, Fei Pei, Ketao Zhong Published: 2026 ABSTRACT ABSTRACTThe development of resistance to anticancer therapies in cancer cell is a major challenge in the treatment of oral squamous cell carcinoma (OSCC), a common malignant tumor. Erythropoietin‐producing hepatocellular A2 (EphA2) affects several cancers, and this study examined its role in enhancing OSCC. Following screening, EphA2 and SRY‐Box transcription factor 8 (SOX8) knocked down and overexpression cell lines were constructed, followed by treatment with Cetuximab. Quantitative real‐time polymerase chain reaction, western blot, cell counting kit‐8, wound healing, and transwell assay were used to detect the relevant indicators. Co‐immunoprecipitation was used to detect the interaction between EphA2 and SH2 domain‐containing protein‐tyrosine phosphatase‐2 (SHP2). Double luciferase reporter gene experiment and chromatin immunoprecipitation experiment were performed to verify the regulatory mechanism of EphA2 and SOX8. The mouse tumor model was established, and the tumor development was observed after plasmid transfection and Cetuximab treatment. EphA2 was highly expressed in OSCC cells, and overexpression of EphA2 up‐regulated SOX8. EphA2 regulated SOX8 and associated protein expression to increase OSCC cell migration and invasion. EphA2 knockdown made OSCC cells more sensitive to Cetuximab, whereas SOX8 overexpression reduced this sensitivity. We found SHP2 bound to SOX8. Down‐regulation of EphA2 decreased tumor growth in mice, increased Cetuximab sensitivity, and overexpression of SOX8/SHP2 decreased Cetuximab sensitivity. OSCC had elevated EphA2 levels, which enhanced cell migration and invasion via SHP2/SOX8 and impaired Cetuximab sensitivity. Down‐regulation of EphA2 decreased tumor growth and enhanced Cetuximab sensitivity, suggesting new OSCC targets and potential treatments. Abbreviations epidermal growth factor receptor epithelial‐mesenchymal transformation erythropoietin‐producing hepatocellular A2 hepatocyte growth factor head and neck squamous cell carcinoma mesenchymal to epithelial transition factor oral squamous cell carcinoma receptor tyrosine kinase SH2 domain‐containing protein‐tyrosine phosphatase‐2 SRY‐Box transcription factor tongue squamous cell carcinoma Introduction Oral squamous cell carcinoma (OSCC) constitutes more than 90% of oral and maxillofacial malignant tumors and is an important part of maxillofacial squamous cell carcinoma [1]. OSCC is more likely to occur on the tongue, cheek, lip, and bottom of the mouth, which seriously endangers the health and life safety of humans [2]. Surgical treatment is the first choice for early‐stage OSCC, while surgery for advanced OSCC is difficult to completely resect due to its infiltrating and invasive characteristics [3]. Cetuximab is a monoclonal antibody that acts on epidermal growth factor receptor (EGFR), which specifically binds to EGFR and plays an anti‐tumor effect by inhibiting intracellular signal transduction, as well as opsonizing immune cell‐mediated phagocytosis and cytotoxicity [4, 5, 6]. Cetuximab was first used to treat metastatic colorectal cancer and non‐small cell lung cancer, and was approved by the US Food and Drug Administration in 2006 for the treatment of head and neck squamous cell carcinoma (HNSCC) [7]. However, some patients may develop drug resistance after Cetuximab treatment. Therefore, it is of great significance to find new related targets that regulate the invasion and metastasis of OSCC and control the drug resistance of OSCC. As a transmembrane receptor tyrosine kinase (RTK), Erythropoietin‐producing hepatocellular A2 (EphA2) plays a key role in the development of various tumors [8]. In OSCC, overexpression of EphA2 is also associated with tumor malignancy, highlighting its potential importance in the development of OSCC [9]. It is worth noting that inhibition of EphA2 has worked well in the treatment of other tumors. Studies have shown that inhibition of EphA2 can significantly reduce the resistance of colorectal cancer cells to Cetuximab and effectively inhibit cell growth, suggesting that the sensitivity of anticancer drugs can be enhanced by regulating EphA2 [10]. In the intracellular signaling system, EphA2 does not function independently. In nasopharyngeal carcinoma cells, EphA2 binds to SH2 domain‐containing protein‐tyrosine phosphatase‐2 (SHP2) protein, which may trigger a series of signal transmission and regulation [11]. Tyrosine phosphatase SHP2 is a key signal regulatory molecule, which can affect the phosphorylation of SRY‐Box Transcription Factor 9 (SOX9) and promote its degradation by the proteasome, thus regulating the stability and transcriptional activity of SOX9 [12]. Therefore, EphA2 may indirectly regulate the transcriptional activity and stability of SOX9 through its interaction with SHP2. Considering that SOX8 is a paralogue of SOX9, based on the similarity of molecular structure and function, it is speculated that SHP2 can also regulate the stability and transcriptional activity of SOX8, and JASPAR predicted the discovery of related promoter binding sequences, which provided theoretical support for this speculation. In addition, SOX8 plays a key role in drug resistance and malignant progression of tumors. For example, it can activate the hepatocyte growth factor (HGF)/mesenchymal to epithelial transition factor (MET) bypass pathway, causing Caco‐2 cells to develop resistance to Cetuximab [13]. In cisplatin‐resistant tongue squamous cell carcinoma (TSCC) cells, SOX8 expression is up‐regulated, and exhibits cancer stem cell‐like properties and epithelial‐mesenchymal transformation (EMT) phenomenon [14]. In the study of OSCC, SOX8 may be a risk factor affecting prognosis according to the REASON risk score [15]. However, the mechanism of action of EphA2/SHP2/SOX8 in OSCC is unclear, and their effects on Cetuximab sensitivity in OSCC are rarely reported. Based on this, we hypothesized that EphA2 activates SHP2 to regulate SOX8 transcriptional activity, inhibit the migration and invasion of OSCC cells, and increase their sensitivity to Cetuximab. This study aims to explore this mechanism and provide new targets and a theoretical basis for the treatment of OSCC. Materials and Methods Animal Ethics Forty‐two male BALB/c nu/nu mice, weighing 23–28 g and aged 7 weeks, were provided by Hangzhou Medical College. The experiment was conducted after the individual had been fed an adaptive diet for a week. The feeding conditions included a specified pathogen‐free (SPF) environment, a 12‐h light and 12‐h dark cycle, a general meal, and drinking water. The laboratory temperature was maintained at 23 ± 1°C. The Ethics Committee of Zhejiang Baiyue Biotechnology Co. Ltd. (Approval No. ZJBYLA‐IACUC‐20240927) has approved the animal research in compliance with ethical guidelines. Cell Culture and Grouping Human oral keratinocytes, hNOK (CSC‐7745W), and human OSCC cell line, HSC‐3 (CSC‐C6335J) were derived from Creative Bioarray (https://www.creative-bioarray.com/). SCC‐25 (CRL‐1628), CAL‐27 (CRL‐2095), and SCC‐9 (CRL‐1629) were derived from ATCC (USA). The cell short tandem repeat was identified correctly. Cells were maintained in Dulbecco's Modified Eagle Medium/Nutrient Mixture F‐12 (DMEM/F‐12) medium (12634010, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, 16000‐044, Invitrogen, USA) and 1% penicillin‐streptomycin (15140122, Gibco, USA) in a cell incubator (3111, THERMO, MA, USA). To research the effect of EphA2 on OSCC cells, cells were transfected with negative control (NC), short hairpin negative control (shNC), EphA2 overexpression plasmid or short hairpin RNA targeting EphA2 (shEphA2). In the second stage, to explore the regulation of OSCC by SOX8, we divided the HSC‐3 cells into the Control, shNC + NC, shEphA2 + NC (EphA2 knockdown + NC), shNC + SOX8 (shNC + SOX8 overexpression), and shEphA2 + SOX8 (EphA2 knockdown + SOX8 overexpression) groups. And the SCC‐25 cells were divided into the Control, NC + shNC, EphA2 + shNC (EphA2 overexpression + shNC), NC + shSOX8 (NC + SOX8 knockdown), and EphA2 + shSOX8 (EphA2 overexpression + SOX8 knockdown) groups. Subsequently, to check the impacts of EphA2 and Cetuximab (205923‐56‐4, Kelong, China) on OSCC cells, cells were distributed to the Control, Cetuximab (treated with 400 µg/mL Cetuximab for 48 h) [16], Cetuximab + NC, Cetuximab + shEphA2, and Cetuximab+SOX8 groups (OSCC cells transfected with shEphA2 or SOX8 overexpression plasmid before exposing to Cetuximab). Animal Modeling and Grouping HSC‐3 cells (including conventional cells, cells stably transfected with NC/shEphA2 /SOX8 overexpressed/SHP2 overexpressed plasmids) were injected subcutaneously into the flanks of mice. About 10 days after incubation, the tumor‐bearing mice were injected with phosphate‐buffered saline (PBS, C0221A, Beyotime, China) (control group) or Cetuximab (20 mg/kg/day) [17] in the abdominal cavity, once every 3 days for 28 days. Each group had seven mice. Tumor volume and weight were measured once a week. The tumor volume was measured with a caliper along two diameter axes, calculated as V = π/6{(D + d)/2}3, where D is the larger diameter and d is the smaller diameter [18]. Cell Transfection The pcDNA3.1 vector (VT9221, YouBio, China) was used to generate the plasmids for EphA2/SOX8/SHP2 overexpression, with the empty vector as the negative control. The pSilencer 3.1‐H1 neo vector (VT1394) provided from YouBio was taken to synthesize shEphA2/shSOX8. The shNC was constructed using the same vector. The above plasmids were transfected into OSCC cells via Lipofectamine 3000 (L3000‐015, Invitrogen, USA). After 48 h, quantitative real‐time polymerase chain reaction (qRT‐PCR) and Western blot were performed to test the transfection efficiency. QRT‐PCR First of all, Trizol reagent (R0016, Beyotime, China) was utilized to extract RNA from OSCC cells. After that, cDNA was obtained by BeyoRT II cDNA First Strand Synthesis Kit (D7168S, Beyotime, China). Thereafter, the cDNA template was amplified by a C1000 Touch PCR system (E1138, Bio‐rad, USA) with SYBR Green qPCR Mix (D7265, Beyotime, China). Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) was applied to the endogenous reference. The results were normalized by 2‐ΔΔCt method for relative expression [19]. The primers were exhibited in Table 1. Table 1Gene sequence primers.NameForward primer (5′–3′)Reverse primer (5′–3′)EphA2GATCGGACCGAGAGCGAGAAGATCAGGTCCCACTTCCTTGCCSOX8CAAAACGGAAGCTGGGAAGCCCAAACTGGTACTGGGTGCTSHP2TCACTTTCTGCAAACGACGCTTGCAGCTCGATTTCCCTTTTTGAPDHGACAAGATGGTGAAGGTCGGTCTTTGGCATCGTGGAAGGGCTCJohn Wiley & Sons, Ltd. Gene sequence primers. Western Blot According to the literature [19], the proteins were lysed from OSCC cells after transfection using a radio immunoprecipitation assay (RIPA) buffer (PC901, Biomiga, USA) and quantified by bicinchoninic acid assay (BCA) kit (KGP903, KeyGEN, China). The proteins were added to the loading buffer and boiled to denaturation and then segregated by electrophoresis. After that, the polyvinylidene fluoride (PVDF) membranes (2215, Millipore, USA) were wet‐transferred and blocked in 5% non‐fat milk (37°C, 120 min). Then, the blocked membranes were exposed to primary antibodies (Table 2) overnight at 4°C followed by incubation with goat anti‐rabbit or goat anti‐mouse antibodies (Table 2) at room temperature. The bands were visualized with ECL reagent (P0018FS, Beyotime, China) using iBright CL750 (Thermo Fisher, USA). GAPDH was employed to the endogenous control. Table 2Antibodies used in this study.NameCatalogMolecular weightDilutionManufacturerEphA2#6997125 kDa1/1000CST,SOX820627‐1‐AP47 kDa1/1000Proteintech, USAE‐Cadherinab4077297 kDa1/1000Abcam, UKN‐Cadherinab18203130 kDa1 µg/mLAbcam, UKSnailab21634729 kDa1/1000Abcam, UKVimentinab9254754 kDa1/1000Abcam, UKSHP2#375272 kDa1/1000CST, UKGAPDHab824536 kDa1/10000Abcam, UKGoat anti‐rabbitab205718—1/2000Abcam, UKGoat anti‐mouseab205719—1/2000Abcam, UKJohn Wiley & Sons, Ltd. Antibodies used in this study. Cell Counting Kit (CCK)−8 Cells (5 × 104 per well) with or without transfection were dripped into 96‐well plates. After being reached cell monolayer, cells were reacted with 10 μL CCK‐8 solution (C0037, Beyotime, China) for 2 h. Finally, the absorbance (450 nm) of cells in each group was determined utilizing a microplate reader (PHERAstar, BMG LABTECH, Germany) at 0, 24, 48, and 72 h. Wound Healing Experiment Cells (1 × 104 per well) were placed in the 37°C incubator for 1 day. Thereafter, a 200 μL pipette tip was taken to scratch on the horizontal lines, and floating cells were removed by PBS. Then, the migrated cells were inspected with the help of an inverted microscope at magnification ×100 (Axiovert 135, ZEISS, Germany) at 0 and 48 h. Cell Invasion Experiment Matrigel (356234; Corning Life Sciences, USA) was diluted into working solution and added into the upper chamber of the Transwell insert (3422; 8.0 µm pore; Corning Life Sciences, USA), which was then placed in the 37°C incubator for 30 min. Then, cells suspended in serum‐free medium was seeded into the upper chamber of transwell culture plate. Complete medium with 10% FBS was added into the lower chamber, and then cultured at 37°C in an incubator. After 48 h of transfection, cells that had invaded into the lower chamber were fixed by 4% paraformaldehyde (YM‐MY803J, Yuanmu, Shanghai, China) for 20 min and stained with a 0.1% crystal violet (C0775, Sigma‐Aldrich, USA) at 37°C for 15 min. Finally, the invaded cells were photographed by the inverted microscope at magnification ×200. Bioinformatics Analysis JASPAR database (http://jaspar.genereg.net/) was used to predict the binding sequence of the SHP2 and SOX8. Dual‐Luciferase Reporter Assay The wild‐type (WT) and mutated (MUT) fragments of SOX8 were amplified into pmirGLO vector (E1330, Promega, USA) to get SOX8‐WT and SOX8‐MUT, which were then co‐transfected with SHP2 overexpression plasmid or NC into HSC‐3 cells. After 48 h, the role of SHP2 on the luciferase activity was measured by a luciferase assay kit (D0010, Solarbio, China). Chromatin Immunoprecipitation (ChIP) The experimental procedures make reference to earlier research. The ChIP assay confirmed the interaction between the promoter regions of SHP2 and SOX8. After fixing HSC‐3 cells with 1% formaldehyde, the buffer was lysed and cross‐linked before being subjected to sonication. After that, SHP2 antibody (#3752, 1:1000, CST, USA) was added, and the mixture was incubated overnight. To obtain the protein‐DNA complex, protein A/G beads were added. In order to extract DNA molecules, 5 mmol/L NaCl (01‐157, Solarbio, China) was then added for de‐crosslinking. SOX8 promoter enrichment was detecetd by qRT‐PCR. Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling (TUNEL) Staining The apoptosis of tumor tissue was stained according to the TUNEL staining kit (C1091, Beyotime, China), and the number of positive cells was observed under an optical microscope (×400). Tunel‐positive nuclei showed brown‐yellow granules, which were apoptotic cells. Each slice was randomly selected with six fields of view. Statistical Analysis GraphPad Prism 8.0 (GraphPad Software, San Diego, CA, USA) was utilized for statistical assay and exhibited as mean ± standard deviation. Independent sample t‐test was used for comparison between two groups; two‐way analysis of variance (ANOVA) was used for comparison between multiple groups in Figures 3A,B and 5A; two‐way repeated measures ANOVA was used for comparison between multiple groups in Figure 7B; one‐way ANOVA was used for comparison between other multiple groups; Tukey's post hoc test was used for multiple comparison between groups. p < 0.05 was deemed to statistically significant. Results EphA2 Was Highly Expressed in OSCC Cells, and Overexpression of EphA2 Up‐Regulated the Level of SOX8 in OSCC Cells It was manifested that the EphA2 level was prominently elevated in OSCC cells (Figure 1A, p < 0.01). Among 4 OSCC cell lines, EphA2 expression was the highest in HSC‐3 cells and lowest in SCC‐25 cells, so these two cells were subsequently selected for experiments (Figure 1A, p < 0.01). QRT‐PCR results showed that EphA2 knockdown and overexpressed cell lines were successfully established (Figure 1B–G, p < 0.01). It can be seen that the level of EphA2 and SOX8 were apparently augmented in SCC‐25 cells after transfection with EphA2, while shEphA2 presented a completely opposite effect in HSC‐3 cells (Figure 1B–G, p < 0.01). Erythropoietin‐producing hepatocellular A2 (EphA2) was overexpressed in oral squamous cell carcinoma (OSCC) cells and overexpression of EphA2 promoted the expression of EphA2 and SRY‐Box transcription factor 8 (SOX8). (A) Quantitative real‐time polymerase chain reaction (qRT‐PCR) was used to detect the expressions of EphA2 in OSCC cells and normal cells. Glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) was applied to the endogenous control. (B–G) The expressions of EphA2 and SOX8 were detected by qRT‐PCR and western blot. GAPDH was employed to the internal reference. The experiment was repeated three times. ## p < 0.01, ### p < 0.001 versus hNOK group; ++ p < 0.01, +++ p < 0.001 versus short hairpin RNA negative control (shNC); ***p < 0.001 versus NC. EphA2 Promoted OSCC Cell Migration and Invasion Through SOX8 We examined the expression of EphA2 or SOX8 after transfection. In HSC‐3 cells, EphA2 silencing down‐regulated the levels of EphA2 and SOX8, while SOX8 overexpression showed the opposite regulatory effect (Figure 2A–E, p < 0.01). However, the regulation of EphA2 silencing and SOX8 overexpression could reverse each other (Figure 2A–E, p < 0.001). In SCC‐25 cells, EphA2 overexpression up‐regulated the levels of EphA2 and SOX8, while SOX8 silencing showed opposite regulatory effects, while the regulation of EphA2 overexpression and SOX8 silencing could reverse each other (Figure 2F–J, p < 0.05). In addition, we also detected that SOX8 silencing inhibited the cell viability of OSCC and reversed the effect of EphA2 overexpression, while SOX8 overexpression promoted cell viability and reversed EphA2 silencing (Figure 3A,B, p < 0.01). In cell migration and invasion experiments, SOX8 silencing significantly inhibited the migration and invasion of OSCC cells, reversing the effect of EphA2 overexpression, SOX8 overexpression promoted cell migration and invasion, reversing the regulation of EphA2 silencing on cells (Figures 3C–F, and 4A–D, p < 0.05). Later, the results of the Western blot experiment showed that EphA2 overexpression promoted the expression of N‐Cadherin (N‐cad), Snail Family Transcriptional Repressor 1 (Snail), and Vimentin, inhibited the expression of E‐cad, and SOX8 silencing reversed this effect (Figure 4F,K–N, p < 0.05). On the contrary, EphA2 silence inhibits N‐cad, Snail, and Vimentin, promotes E‐cad, and overexpression of SOX8 reversed this effect (Figure 4E,G–J, p < 0.05). SOX8 silenced/overexpressed and EphA2 overexpressed/silenced cell lines were established successfully. (A–J) The transfection efficiency of SOX8 and EphA2 was detected by qRT‐PCR and western blot. GAPDH was employed to the internal reference. The experiment was repeated three times. ^ p < 0.05, ^^ p < 0.01, ^^^ p < 0.001 versus shNC+NC; $$$ p < 0.001 versus shEphA2 + NC; & & & p < 0.001 versus shNC + SOX8; ΔΔΔ p < 0.001 versus EphA2 + shNC; θ p < 0.05, θθθ p < 0.001 versus NC + shSOX8. SOX8 silencing/overexpression reversed the regulation of EphA2 overexpression/silencing on cell viability and migration. (A, B) The effect of SOX8/EphA2 on cell viability was detected by Cell Counting Kit (CCK)−8 assay. (C–F) The effect of SOX8/EphA2 on cell migration was detected by wound healing assay (magnification ×100, scale bars = 100 μm). The experiment was repeated three times. ^ p < 0.05, ^^ p < 0.01, ^^^ p < 0.001 versus shNC + NC; $$ p < 0.01 versus shEphA2 + NC; & p < 0.05, & & & p < 0.001 versus shNC + SOX8; ΔΔ p < 0.01, ΔΔΔ p < 0.001 versus EphA2 + shNC; θθθ p < 0.001 versus NC + shSOX8. SOX8 silencing/overexpression reversed the regulation of EphA2 overexpression/silencing on cell invasion and epithelial‐mesenchymal transition (EMT) protein. (A−D) The effect of SOX8/EphA2 on the cell invasion was detected by transwell assay (magnification ×200, scale bars = 100 μm). (E–N) The effect of SOX8/EphA2 on the EMT markers (E‐cadherin, N‐cadherin, Snail and vimentin) was detected by western blot. GAPDH was utilized as the internal reference. The experiment was repeated three times. ^ p < 0.05, ^^ p < 0.01, ^^^ p < 0.001 versus shNC + NC; $ p < 0.05, $$$ p < 0.001 versus shEphA2 + NC; & & & p < 0.001 versus shNC + SOX8; Δ p < 0.05, ΔΔ p < 0.01, ΔΔΔ p < 0.001 versus EphA2 + shNC; θ p < 0.05, θθθ p < 0.001 versus NC + shSOX8. ShEphA2 Enhanced the Sensitivity of OSCC Cells to Cetuximab, Inhibited Cell Migration and Invasion, and SOX8 Overexpression Reversed the Effect of Cetuximab In cell biology experiments, we observed that Cetuximab decreased cell viability, migration, and invasion, with shEphA2 amplifying this impact of Cetuximab, whereas SOX8 overexpression counteracted the regulatory effects of Cetuximab on cells (Figure 5A−E, p < 0.05). Cetuximab inhibited the expression of N‐cadherin, Snail, and Vimentin, while promoted E‐cadherin expression; shEphA2 augmented this impact of Cetuximab, while SOX8 overexpression negated Cetuximab's regulatory influence on these proteins (Figure 5F,G, p < 0.05). Cetuximab inhibited cell biological function and regulated EMT protein, which was enhanced by EphA2 silencing and reversed by SOX8 overexpression. (A) The cell viability was detected by CCK‐8 assay. (B, D) The cell migration was detected by wound healing assay (magnification ×100, scale bars = 100 μm). (C, E) The cell invasion was detected by transwell assay (magnification ×200, scale bars = 100 μm). (F, G) The EMT markers (E‐cadherin, N‐cadherin, Snail, and vimentin) were detected by western blot. GAPDH was utilized as the internal reference. The experiment was repeated three times. *p < 0.05, **p < 0.01, ***p < 0.001 versus Control; + p < 0.05, ++ p < 0.01, +++ p < 0.001 versus Cetuximab + NC. SHP2 Bound to SOX8 Next, we conducted a series of experiments to verify the regulatory mechanism between SHP2 and SOX8. Protein detection results showed that EphA2 silencing inhibited the expression of SHP2 protein in cells (Figure 6A,B, p < 001). Combining the JASPAR database prediction results, we predicted the combination sequence of SHP2 and SOX8 (Figure 6C). Subsequently, we successfully transfected the SHP2 overexpression plasmid in HSC‐3 cells (Figure 6D, p < 0.001). The results of double luciferase reporter gene experiment confirmed that SHP2 overexpression could significantly enhance luciferase activity in SOX8‐WT group (p < 0.001), but did not affect luciferase activity in SOX8‐MUT group (Figure 6E). The ChIP experiment also confirmed the successful binding of SHP2 and SOX8, and the results showed that the relative enrichment of SOX8 in the anti‐SHP2 group was significantly increased (Figure 6F, p < 0.001). The final western blot experiment also showed that the overexpression of SHP2 promoted the cell SOX8 protein (Figure 6G,H, p < 0.05). SHP2 bound to SOX8. (A, B) The effect of EphA2 silencing on SHP2 was detected by Western blot. GAPDH was utilized as the internal reference. (C) JASPAR predicted SHP2 and SOX8 binding sequences. (D) The transfection efficiency of SHP2 overexpression was detected by qRT‐PCR. GAPDH was utilized as the internal reference. (E) A dual luciferase reporter gene was used to detect the binding of SHP2 and SOX8. (F) Chromatin immunoprecipitation (ChIP) detected the binding of SHP2 and SOX8. (G, H) Western blot analysis was performed to determine the regulation of SOX8 protein by SHP2 overexpression. GAPDH was utilized as the internal reference. The experiment was repeated three times. & & & p < 0.001 versus shNC; ^ p < 0.05, ^^^ p < 0.001 versus NC; ### p < 0.001 versus Anti‐Immunoglobulin G (IgG). Down‐Regulation of EphA2 Inhibited Tumor Development in Mice and Increased Sensitivity to Cetuximab To further verify the mechanism by which EphA2 regulates OSCC, we established a mouse tumor model and treated it with plasmid transfection and Cetuximab. Cetuximab was found to inhibit tumor growth, which was enhanced by shEphA2, and reversed by overexpression of SOX8 or SHP2 (Figure 7A,B, p < 0.001). In addition, Cetuximab inhibited N‐cad, snail, and Vimentin proteins in tumor tissue, and promoted the expression of E‐cad protein, which was enhanced by shEphA2, while overexpression of SOX8 or SHP2 reversed the Cetuximab effect (Figure 7C−G, p < 0.05). Finally, in the detection of apoptosis of tumor tissue, Cetuximab was found to promote apoptosis of tumor tissue cells, and shEphA2 enhanced this effect, while overexpression of SOX8 or SHP2 reversed the Cetuximab effect (Figure 7H). Down‐regulation of EphA2 inhibited tumor development and increased Cetuximab sensitivity in mice. (A, B) Detection of tumor volume in mice (n = 7). (C−G) Western blot was used to detect EMT‐related proteins (E‐cadherin, N‐cadherin, Snail, and vimentin) in tumor tissues (n =7). GAPDH was utilized as the internal reference. (H) The apoptosis of tumor tissue was detected by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining (n = 7). Tunel‐positive nuclei showed brown‐yellow granules, which were apoptotic cells (magnification ×400, scale bars = 50 μm). *p < 0.05, ***p < 0.001 versus Control; + p < 0.05, ++ p < 0.01, +++ p < 0.001 versus Cetuximab + NC. Discussion As an important member of the Eph receptor family, EphA2 is highly expressed in a variety of cancers, such as prostate cancer, colorectal cancer, cervical cancer, and skin cancer [20]. In tumors, EphA2 overexpression is associated with non‐classical pathway activation, tumor progression, and poor prognosis, which emphasizes its importance as a marker of malignant tumors [21]. Studies have shown that Senescence‐associated Epithelial Membrane Protein 1 (CLDN1) promotes cancer dryness and chemotherapy resistance of colorectal cancer through direct interaction with EphA2 [22]. However, the specific action mechanism of EphA2 in OSCC remains to be further explored. OSCC is a common malignant tumor of the head and neck, and its invasion and metastasis characteristics seriously affect the prognosis of patients [23]. Previous studies have found that EphA2 is overexpressed in OSCC tissues and is associated with poor prognosis. It can also promote Kruppel‐like factor 4 (KLF4) transcription and thus enhance the dryness of OSCC cells [24]. This study found that EphA2 was highly expressed in OSCC cells, which was consistent with previous studies. In addition, the establishment of EphA2 knockdown and overexpression cell lines and the selection of HSC‐3 and SCC‐25 cells with significant differences in expression levels provided a good model for the study of their functions. SHP2, as a protein tyrosine phosphatase, is involved in a variety of cell signal transduction pathways and plays an important role in cell proliferation, differentiation, and survival, and is also a well‐known carcinogenic phosphatase in cancer [25]. SOX8 belongs to the SOX transcription factor family and is involved in the regulation of cell development and differentiation [26]. Studies have proved that overexpression of SOX8 is associated with poor prognosis of colorectal cancer [27]. However, the relationship between the two and the specific mechanism of action in OSCC remains unclear. This study confirmed that EphA2 can affect the migration and invasion of OSCC cells through the SHP2/SOX8 pathway. Previous study has shown that the occurrence of EMT was correlated with the invasion and metastasis of OSCC [28]. OSCC cells lose the functions of epithelial cells through EMT, thereby facilitating the metastasis and invasion of OSCC [29]. E‐cadherin, N‐cadherin, Snail, and vimentin are EMT‐related proteins. Highly aggressive OSCC usually has high expressions of N‐cadherin, Snail, and vimentin and low expression of E‐cadherin [30]. E‐cadherin is an important protein that maintains the properties of epithelial cells, and its decreased expression is often associated with increased invasion of tumor cells [31]. Snail is a zinc finger protein that can suppress the level of E‐cadherin and elevate the level of interstitial cell molecules, and Snail is strengthened in OSCC and is closely related to the poor prognosis of patients [32]. This study found that EphA2 overexpression promoted the expression of N‐cadherin, Snail and Vimentin, and inhibited the expression of E‐cadherin, while SOX8 silencing could reverse this effect. These results indicated that EphA2 affected the migration and invasion of OSCC cells through the SHP2/SOX8 pathway by regulating the expression of these proteins. The accumulated evidence shows that the EMT‐related signaling pathway is related to drug resistance of cancer cells [33]. In addition, cells that have experienced EMT show stem cell‐like characteristics and have the advantage of selective growth in the presence of drugs [34]. Hiroko Kitahara et al. reported that the sensitivity of OSCC to Cetuximab can be enhanced by triggering MET [35]. Zhenlv Lin et al pointed out that inhibition of EMT can promote the sensitivity of colorectal cancer cells to Cetuximab [36]. In this study, we found that down‐regulation of EphA2 can enhance the sensitivity of OSCC cells to Cetuximab and inhibit cell migration and invasion, while overexpression of SOX8 can reverse the effect of Cetuximab. These results provide new targets and ideas for the treatment of OSCC, suggesting that EphA2 and SHP2/SOX8 pathways may be potential intervention targets to improve the therapeutic effect of Cetuximab. Conclusion In summary, this study revealed that EphA2 promotes the migration and invasion of OSCC cells through the SHP2/SOX8 pathway and reduces the sensitivity of cancer cells to Cetuximab. Down‐regulation of EphA2 can inhibit tumor development and increase sensitivity to Cetuximab, providing a new target and idea for the treatment of OSCC. However, there are still some limitations in this study. For example, the specific molecular mechanism of EphA2 regulating OSCC cell biological behavior through the SHP2/SOX8 pathway needs to be further studied, and the efficacy and safety of EPHA2 in clinical application need to be verified by more clinical trials. Future studies could focus on these aspects to further refine the understanding of EphA2's role in OSCC and provide more effective strategies for the treatment of OSCC. Author Contributions Dayong Yan: conceptualization, methodology, validation, writing – original draft, writing – review and editing. Lele Guo: conceptualization, investigation, formal analysis, writing – original draft, writing – review and editing. Fei Pei: data curation, formal analysis, methodology, visualization, writing – original draft, writing – review and editing. Ketao Zhong: data curation, formal analysis, investigation, supervision, validation, writing – original draft, writing – review and editing. Conflicts of Interest The authors declare no conflicts of interest.