Benzydamine, a CDK2 Kinase Inhibitor, Suppresses The Growth of Esophageal Squamous Cell Carcinoma in Vitro and in Vivo.

Background: Esophageal squamous cell carcinoma (ESCC) is among one of the leading causes of cancer death worldwide owing to late detection and low survival rate. The clinical outcome of ESCC remains dismal. To date, the disease lacks available targeted therapies. Recently, drugs approved by the Food and Drug Administration have been reported to have potential as cancer chemoprevention agents. Methods: Benzydamine, available as a hydrochloride salt, a locally acting non-steroidal anti-inammatory drug, was screened out among FDA-approved drugs owing to its effective cytotoxic effect on KYSE450 cells, which remained unexplored. Mass spectrometry, kinase prediction and Swiss Target Prediction were used to verify the potential target(s) of benzydamine. Patient-derived oesophageal xenograft mouse model were used to investigate the effect of benzydamine on tumor growth in vivo. Results: We found that benzydamine inhibited anchorage-dependent and -independent growth of ESCC cells. Kyoto Encyclopedia of Genes and Genomes pathway enrichment revealed that benzydamine attenuated ve signaling pathways, including the DNA replication pathway. We further found that benzydamine could bind to CDK2 in its ATP-binding site. Inhibition of the activity of CDK2 suppressed the growth of ESCC cells and led to a G1/S cell cycle arrest. Additionally, knocking-down CDK2 decreased the sensitivity of ESCC cells to benzydamine hydrochloride. Notably, benzydamine suppressed tumour growth in a patient-derived oesophageal xenograft mouse model of ESCC in vivo. Conclusions: We have identied CDK2 as a potential target of benzydamine for the treatment and prevention of ESCC. Benzydamine suppressed the growth of ESCC cells by inhibiting the activity of CDK2. Pull-down assay using CNBr-Sepharose 4B beads Benzydamine-Sepharose 4B beads and dimethyl sulfoxide (DMSO)-Sepharose 4B beads were provided by GE Healthcare Bio-Science (Uppsala, Sweden) and prepared according to the manufacturer’s instructions. Cell lysates (500 μg), active CDK2 (200 ng), and 293F cell lysate (500 μg) were incubated with benzydamine-Sepharose 4B beads and DMSO-Sepharose 4B beads in 1× reaction buffer (50 mM Tris pH 7.5, 5 mM EDTA, 150 mM NaCl, 1 mM DTT, 0.01 % NP-40, 2 μg/mL bovine serum albumin) at 4 °C with gentle rotation overnight. Beads were washed thrice with washing buffer (50 mM Tris-HCl pH 7.5, 5 mM EDTA, 150 mM NaCl, 1 mM DTT, 0.2 mM PMSF, and 0.01 % NP-40) after incubation. CDK2 bands were analyzed by western blotting. Dei SA, Postigo AA, Dean DC. Cdk phosphorylation triggers sequential intramolecular interactions that progressively block Rb functions as cells move through G1. Cell. conjugated Sepharose 4B beads or with Sepharose 4B beads alone. Pull-down proteins were analyzed by western blotting. (e) The specicity of the binding of benzydamine to active CDK2 in the presence of ATP was evaluated. (f, g) Active CDK2 (500 ng) and various doses of benzydamine were incubated with MCM2 (1 μg) as a substrate at 30 °C for 30 min. The phosphorylation of MCM2 (Serine, Ser 41) was detected by western blotting. Our results revealed that benzydamine inhibited the activity of the CDK2 kinase in vitro. CDK2 and MCM2 proteins were detected by western blotting and Coomassie blue staining (The bottom picture).

Cyclin-dependent kinase 2 (CDK2) is considered to be a vital kinase in cell cycle regulation, participating in a series of biological processes. In many cancers, CDK2 plays an important role in tumor proliferation and correlated with cancer patients' survival [11,12]. Additionally, emerging evidence has demonstrated that inhibition of CDK2 elicits an anti-tumor activity in a subset of tumors [13]. Therefore, CDK2-selective inhibitors might present a therapeutic opportunity for CDK2-highly expressed cancers [14]. In this study, we identi ed benzydamine, a FDA-approved drug, as a CDK2 inhibitor. We elucidated the anti-tumor effect of benzydamine in ESCC in vitro and in vivo, which was shown to be achieved through the attenuation of the CDK2/ minichromosome maintenance protein 2 (MCM2) signaling pathways. In sum up, our results suggested that targeting CDK2 is a potentially effective therapeutic strategy for improving the treatment of ESCC. Anchorage-independent cell growth assay KYSE150 and KYSE450 cells (8 × 10 3 cells/well) were suspended in RPMI 1640 and DMEM containing 0.3 % agar and 10 % FBS at various concentrations of benzydamine (0, 2.5, 5, 10, and 20 μM). Cells were cultured at 37 °C and 5 % CO 2 for 10 days. Colonies were measured and analyzed using the IN Cell Analyzer 6000 software.
Cell sample preparation and phosphoproteomics analysis KYSE150 cells (4.5 × 10 6 ) were seeded in 15 cm dishes. After 16-18 h of culture followed by 20 µM benzydamine treatment for 24 h, cells were lysed in lysis buffer (RIPA lysate, Solarbio, Beijing, China, #R0020). After centrifuging the samples to remove debris, the supernatant was collected. The samples were digested with trypsin and the tryptic peptides were fractionated by high pH reverse-phase HPLC using a Thermo Betasil C18 column (5 μm particles, 10 mm ID, 250 mm length). In brief, peptides were separated with a gradient of 8-32 % acetonitrile (pH 9.0) for approximately 60 min, resulting in 60 fractions. Subsequently, peptides were combined into six fractions and dried by vacuum centrifugation. Peptides were rst subjected to a nanospray ionization source and then tandem mass spectrometry (MS/MS) in a Q ExactiveTM Plus (Thermo Fisher Scienti c, Waltham, MA, USA) coupled online to the UPLC. Data were obtained by searching through the human UniProt database for identi ed peptides assembled as proteins. The resulting MS/MS data were processed by the Maxquant search engine (v.1.5.2.8) and analyzed.

Cell cycle analysis
Cells (3 × 10 5 cells) were seeded in 60-mm plates and synchronized by serum starvation for 24 h. Cells were then treated with benzydamine (0, 2.5, 5, 10, or 20 μM) for 24 h or 48 h in 10 % serum-supplemented medium. For cell cycle analysis, cells were harvested and washed with phosphate buffered saline twice, xed in 70 % ethanol (Tianjin Zhiyuan Chemical Reagent Co., Ltd, China), and stored at -20 °C for 24 h. Cells were stained with propidium iodide (Beyotime, Shanghai, China) for cell cycle assessment, followed by analysis using a BD FACSCalibur Flow Cytometer (BD Biosciences, San Jose, CA).

Computational modelling of CDK2 with benzydamine
To explore the binding and interaction of CDK2 with benzydamine, we performed in silico docking using the autodock software programs. First, we downloaded the CDK2 crystal structure from the PDB (ID: 1AQ1) and prepared it using the standard procedures of the Protein Preparation Wizard (autodock). Hydrogen atoms were added at a pH of 7, and all water molecules were removed. The drug benzydamine was prepared for docking by using the default parameters in the LigPrep program. Subsequently, the docking of benzydamine to CDK2 was achieved using the default parameters in the extra precision (XP) mode in the Glide program.
Protein puri cation PET-28a-CDK2 and MCM2 plasmids (purchased from YouBio Biotechnology Company, Hunan, China) were transformed in E. coli and ampli ed. Following ampli cation, the cells were lysed using ultrasounds, and the proteins obtained were puri ed by nickel column adsorption and elution. CDK2 and MCM2 proteins were identi ed by western blotting and Coomassie blue staining.

In vitro kinase assay
In vitro kinase assays were performed according to the manufacturer's instructions. Human recombined MCM2 protein (1 μg), used as a substrate for CDK2, was mixed with active CDK2 (500 ng) and varying doses of benzydamine in a 25 μL reaction mixture, which was supplemented with 20 μM ATP and 1× kinase buffer (Cell Signaling Technology, Danvers, MA, USA), followed by incubation at 30 °C for 30 min.
Reactions were then blocked by the addition of 5 μL 6× loading buffer, and proteins were analyzed by western blotting.
Cells were then re-incubated in fresh medium for 24 h. Subsequently, 1 μg/mL (KYSE450) or 2 μg/mL (KYSE150) puromycin was used for cell selection. The transduction e ciency was analyzed by western blotting. The cell proliferation and colony formation ability of knockdown cells were examined in comparison with mock-transfected cells. KYSE450 and KYSE150 cells were transfected with short hairpin CDK2 (shCDK2) in a similar manner.

Patient-derived xenograft (PDX) mouse model
All research protocols used in this study were approved by the Research Ethics Committee of Zhengzhou University. ESCC tissues were obtained from the Linzhou Tumour HospitalHenan. The protocols for establishing a PDX mouse model have been previously described [15,16]. For these experiments, 6-to 8week-old severe combined immunode cient (SCID) female mice were used. Tumor tissues from patients were subcutaneously implanted into the back of SCID mice. When the tumor mass reached an average volume of 100 mm 3 , mice were randomly divided into three treatment groups (n = 8-10/group) as follows: (1) vehicle group (n = 10); (2) 5 mg/kg benzydamine (n = 10); (3) 50 mg/kg benzydamine hydrochloride (n = 10). Benzydamine hydrochloride was administered using a gavage once a day for 30 days. Body weight was monitored three times per week. The tumor volume of each mouse was measured twice a week. Tumor volume was calculated using the following formula: V = LD × (SD) 2 /2, where V is the tumor volume. When the average tumor volume reached 1.0 cm 3 , mice were euthanized under anesthesia and tumors were extracted.
Immunohistochemical staining assay Formalin-xed tumor tissue sections were depara nized, hydrated, and cut into 4-µm sections for immunohistochemistry. Samples were baked in a constant-temperature oven at 65 °C, and citrate acid was used for antigen retrieval. All tumor tissue sections were blocked with 3 % H 2 O 2 for 10 min in the dark. Slides were then hybridized using speci c antibodies (Ki-67, 1:50,Abcam, Cambridge, England; p-MCM2 S41 , 1:200, Abcam, Cambridge, England) for 16 h at 4 °C and then incubated with an HRPconjugated goat anti-rabbit or mouse IgG antibody (ZSGB-BIO, Beijing, China) for 30 min. After DAB staining for 2 min, sectioned tissues were counterstained with hematoxylin (Baso, Zhuhai, Guangdong, China) for 1 min, dehydrated in a graded series of alcohol to xylene, and covered with glass coverslips. All slides were observed under a microscope and scanned using the Tissue Faxes (version 4.2). The Image-Pro Plus software (v. 6) was used for measuring positive cells.

Statistical analysis
One-way analysis of variance or the Student's t-test was used to compare signi cant differences; p < 0.05 was considered statistically signi cant. All quantitative results were expressed as the mean ± standard deviation or ± standard error, as indicated.

Results
Benzydamine suppressed anchorage-dependent and -independent growth of ESCC To identify a novel drug against ESCC, we screened FDA-approved drugs by performing a cytotoxic assay on KYSE450 cells. Benzydamine, a NSAID, exhibited signi cant cytotoxic effects in KYSE450 cells (Fig.  1a, b, Supplementary table1) (Fig. 1c). Subsequently, we used different doses of benzydamine to examine its effects on the anchorage-dependent growth of KYSE150, KYSE450, and SHEE cells. We found that benzydamine inhibited the anchorage-dependent growth of KYSE150 and KYSE450 cells (Fig. 1d). We then veri ed the effect of benzydamine on the anchorage-independent growth of KYSE150 and KYSE450 cells using a soft agar assay. Our results demonstrated that benzydamine suppressed the anchorageindependent growth of ESCC cells in a dose-dependent manner (Fig. 1e). Plate clone formation assays of KYSE150 and KYSE450 cells also indicated that benzydamine-treated groups had fewer colony numbers than the control group (Fig. 1f).

Phosphorylation pro les revealed the anti-tumor mechanism of benzydamine
To explore the inhibitory mechanism of benzydamine on ESCC, we conducted phosphoproteomic analysis to comprehensively analyze the changes in the phosphorylation level of proteins after treatment with benzydamine. Phosphorylation changes upon treatment with benzydamine were analyzed following a precise standard (t-test p-value < 0.05, 1.5-fold change from baseline as the threshold). We identi ed a total of 3 496 proteins, of which 2 982 proteins were quanti ed. Among the differentially expressed proteins, 159 proteins were upregulated, whereas 363 proteins were downregulated. We also identi ed a total of 14069 phosphorylation sites, among which 8509 sites were quanti ed. Among these quanti ed phosphosites, 191 sites were upregulated, whereas 500 sites were downregulated in KYSE150 cells after treated with 20 µM benzydamine ( Fig. 2a-b). We then mapped the quanti ed phosphosites to Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment. We observed that the top ve down regulated signaling pathways included RNA transport, DNA replication, spliceosome, ferroptosis, and protein processing in the endoplasmic reticulum. The phosphoproteomic data suggested that multiple phosphorylation sites that have been essentially related with "cancer driver" genes were down regulated in the DNA replication signaling pathway (Fig. 2c). For instance, we found that the MCM2 S41 phosphosite was obviously down regulated in both the DNA replication and cell cycle signaling pathways (Fig. 2d). This phosphorylation site is known to participate in the initiation of DNA synthesis and is reportedly required for entry into the S phase and for cell division. Our results suggested that the phosphorylation level of MCM2 S41 was evidently inhibited after benzydamine treatment (Fig. 2e). To explore the underlying mechanism of the anti-tumor effect of benzydamine, we performed kinase prediction and SwissTargetPrediction to identify the upstream kinases. We observed that CDK2 was the most promising molecular target of benzydamine. Additionally, Spearman's correlation analysis of the expression of the CDK2 and MCM2 genes representing the CDK2/MCM2 signaling pathway also showed signi cant differences (p = 2.58e-14) (Fig. 2f).
Benzydamine-induced G1/S phase arrest and inhibited the DNA replication pathway We found that treatment with benzydamine affected the cell cycle distribution in the KYSE150 and KYSE450 ESCC cell lines (Fig. 3a, b). More speci cally, benzydamine caused a signi cant G1/S phase cell cycle arrest in KYSE150 and KYSE450 cells (p < 0.05). Additionally, immuno uorescent experiments suggested the colocalization between p-MCM2 (Ser41) and CDK2 and benzydamine decreaed the phosphorylation of MCM2 S41 in KYSE150 and KYSE450 cells. Based on the results obtained, we detected the phosphorylation and expression levels of proteins involved in the transition from G1 to S phase during the cell cycle (Fig. 3c, d). We found that compared with the DMSO vehicle control, benzydamine reduced the phosphorylation of MCM2 S41 , c-Myc S62 and Rb T826 in ESCC cells in a dose-dependent manner (Fig. 3e, f).
Benzydamine bound directly to CDK2 and inhibited CDK2 activity Based on the SwissTarget and kinase prediction analysis, we hypothesized that CDK2 might be the upstream regulator of benzydamine in ESCC. To investigate the possible mechanism involved in the benzydamine-CDK2 interaction, we downloaded the CDK2 kinase domain (PDB: 1AQ1) and docked it with benzydamine following the protocols in the autodock software programs [17]. According to the docking model, we found that benzydamine formed a hydrogen bond with the CDK2 aspartic acid at site 145, which is also known to be an ATP-binding site affecting kinase activity. This suggested that benzydamine might be competing with ATP for binding at the ATP-binding site of CDK2 (Fig. 4a). To further verify this result, we performed a pull-down assay in vitro and in vivo, by conjugating benzydamine with Sepharose 4B beads. We found that the recombinant CDK2 protein kinase bound with benzydamine-conjugated Sepharose 4B beads but not with Sepharose 4B beads alone (Fig. 4b). We then constructed mutant CDK2 (F80A, D145A, F146A, D145A+146A) and ectopically expressed these mutants in 293F cells. Pull-down assays using wild-type or mutant CDK2 proteins and conjugated Sepharose 4B beads revealed that the D145A CDK2 mutant exhibited the greatest reduction in the binding a nity to benzydamine (Fig. 4c), indicating that Asp145 site is essential for binding (Fig. 4c). We further performed an in vivo pull-down assay using KYSE150 and KYSE450 cell lysates, and our results suggested that benzydamine could also bind to CDK2 in vivo (Fig. 4d). Interestingly, we also determined that the binding of benzydamine to CDK2 was ATP-competitive. We observed that benzydamine competed with ATP for binding at the ATP-binding site of CDK2, Asp145 (Fig. 4e). And, sequence alignment of CDK2 phosphorylation consensus show highly evolutionarily conserved among multiple species suggesting that Asp145 was a functional phosphorylation site (Fig. S3). Next, we performed an in vitro kinase assay using an active recombinant CDK2 protein and MCM2 as a substrate, to verify whether CDK2 was the direct target of benzydamine. Our results suggested that the activity of CDK2 was strongly inhibited by treatment with benzydamine in a dose-dependent manner (Fig. 4f, g) and con rmed that benzydamine directly suppressed the activity of CDK2, resulting in inhibition of CDK2 phosphorylation of MCM2 on serine, especially on serine 41.
Knockdown of CDK2 decreased the sensitivity of ESCC cells to benzydamine TCGA analysis of CDK2 suggested that CDK2 is highly expressed in EC (Fig. 5a). To further explore the function of CDK2 in ESCC tumor growth, we performed experiments after knocking-down CDK2. After transfecting KYSE150 and KYSE450 cells with sh-CDK2, the total protein level of CDK2 was measured to verify the e ciency of transfection and CDK2 knockdown also decreased the phosphorylation of related downstream biomarkers (Fig. 5b). We also found that the CDK2 knockdown suppressed the proliferation of KYSE150 and KYSE450 cells (Fig. 5c). Moreover, we noticed that the colony formation ability of these cells was also inhibited after CDK2 was knocked-down (Fig. 5d, e). effect of the CDK2 knockdown was also detected in cell cycle distribution (Fig. 5f, g). Our results suggested that knocking-down CDK2 induced a G1/S phase arrest in KYSE150 and KYSE450 cells. Benzydamine was shown to speci cally target CDK2; therefore, we further investigated whether knockdown of CDK2 could affect the sensitivity of ESCC cells to benzydamine. We observed that following CDK2 knockdown, treatment with benzydamine had no obvious effect on the anchorage-dependent growth compared with that in mock-transfected cells. Moreover, the benzydamine-induced inhibition of both the proliferation of ESCC cells was weakened by the CDK2 knockdown (Fig. 5h).

Benzydamine suppressed patient-derived esophageal xenograft tumor growth in vivo
We used the PDX model to evaluate the anti-tumor activity of benzydamine in vivo. Tumor tissue from patients was implanted into the backs of SCID mice. Mice were administered benzydamine hydrochloride (5 and 50 mg/kg) or vehicle. We found that benzydamine suppressed the volume of tumor growth relative to the vehicle-treated group, while, the average body weight was not obviously different between different groups (Fig. 6a, b, Fig. S4). The mice were euthanized, and the weight of tumor tissues was measured; we observed that an increase in tumor weight was inhibited in mice treated with benzydamine (Fig. 6c).
In addition, we also observed that benzydamine-treated mice did not exhibit an obvious loss of body weight compared with the vehicle-treated group. We then veri ed the ndings of the phosphoproteomic pro le at the tissue level using immunohistochemistry. Our results demonstrated that benzydamine decreased the expression of Ki67 and the phosphorylation level of MCM2 S41 (Fig. 6d).

Discussion
The incidence and mortality of EC was reportedly ranked in the seventh and sixth place, respectively, among all cancers in 2020 [18]. Thus, there is an urgent need for the identi cation of new inhibitors for effective targeted therapies. Recently, the application of FDA-approved drugs has become a new approach for chemoprevention [8]. These agents have the potential to reduce the onset of cancer and its relapse after treatment by inhibiting signaling molecules of great importance in carcinogenic events [19]. In this study, we showed that the NSAID benzydamine signi cantly inhibited the proliferation and colony formation of ESCC cells in vitro, as well as the tumor growth of patient-derived esophageal xenografts of ESCC in vivo ( Fig. 1 and 6). We identi ed a novel drug with chemopreventive potential, especially against CDK2-dependent cancers.
We performed mass spectrometry to explore the systemic pharmacology of the inhibitory effect of benzydamine on ESCC cells. Analysis of the phosphoproteome and KEGG enrichment suggested that benzydamine signi cantly downregulated the DNA replication signaling pathway, which is essential for cell cycle and genome stability [20]. Additionally, analysis of differentially-phosphorylated sites indicated that the phosphorylation levels of MCM2 S41 , Rb T826 , and c-Myc S62 were decreased. Combined with kinase prediction, Swiss target prediction, and in vitro assays, CDK2 was identi ed as the potential target of benzydamine (Fig. 2, 4).
CDK2 belongs to the family of cyclin-dependent kinase complexes and plays a key role in the cell division cycle [21]. In particular, CDK2 is known to be activated by binding to cyclin E1 or E2, and the activity of both cyclin A2 and CDK2 is critically related to tumor growth in many cancer types [14]. The cyclin E-CDK2 complex acts as a master regulator of the progression from the late G1 phase to the S phase [13,22], which is essential for eukaryotic cell proliferation and its mis-regulation has been shown to enhance oncogenesis [23]. Inhibition of cyclin E-CDK2 leads to a G1/S arrest, blocking cell cycle progression [13,24]. Therefore, we also investigated the potential inhibitory mechanisms of benzydamine on the growth of EC. We found that benzydamine induced a G1/S arrest in ESCC cells in a dose-dependent manner by inhibiting the activity of CDK2 (Fig. 3).
Moreover, CDK2 has been shown to speci cally phosphorylate MCM2, Rb, and c-Myc. During the G1 to S transition, cyclin E-CDK2 was found to phosphorylate MCM2 at Ser 41 to initiate DNA synthesis [25,26]. Meanwhile, this complex has also been reported to catalyze the hyper-phosphorylation of Rb at Thr 826, resulting in the promotion of tumor growth [27]. Additionally, cyclin E/CDK2 is known to phosphorylate c-Myc at Ser 62 and the phosphorylation can increase the biological activity of Myc, which regulating cell cycle and tumor initiation [28,29]. Our results indicated that benzydamine decreased the phosphorylation of MCM2 S41 , RB T826 , and c-Myc S62 by binding to CDK2 and inhibiting its activity, which has been critically related to tumor growth in many cancer types. To our knowledge, this is the rst study elucidating the anti-tumor effect of benzydamine. In addition, we explored whether the inhibitory effect of benzydamine on ESCC cells depended on the expression of CDK2. To this end, we established cells expressing sh-CDK2. We found that knocking-down CDK2 restrained the growth and colony formation of ESCC cells. Furthermore, ESCC cells expressing sh-CDK2 were less sensitive to benzydamine than cells expressing shcontrol (Fig. 5). Therefore, we assumed that CDK2 is the major target protein of benzydamine in EC cells. A computational modelling system could identify the critical interacting residues between target proteins and natural compounds [30,31]. By utilizing computational modelling, we simulated the binding model of benzydamine to CDK2. We computationally predicted that benzydamine possesses a series of binding orientations in the ATP-binding site of CDK2 (Fig. 4a). In addition, an in vitro and in vivo binding assay suggested that the Asp145 residue of CDK2 is critically essential for binding, affecting the kinase activity (Fig. 4b-e).
PDX models provide a series of advantages over human cell line xenograft models for the evaluation of preclinical therapies and prediction of responsiveness to anti-cancer treatments because they retain more genetic characteristics of the tumor specimens of patients [32][33][34]. In this study, we constructed PDX models in immunode cient mice by subcutaneously inoculating them with tumor tissues from an ESCC patient [35]. We found that benzydamine suppressed tumor growth of ESCC and PDX tumors in mice by attenuating the phosphorylation of MCM2 at serine 41. Additionally, the expression of Ki67, a proliferation marker [36], was also shown to be decreased in tumor tissues after treatment with benzydamine (Fig. 6). These ndings indicated that benzydamine could suppress the tumor growth of ESCC in vivo by targeting the CDK2/MCM2 signaling pathway.
Benzydamine has been used to prevent postoperative sore throat [37], to treat oral mucositis, and for pain relief in patients with cancer receiving chemotherapy, radiotherapy, or both [38]. However, beyond its original use, this is the rst study to establish the anti-tumor activity of benzydamine in EC through the targeting of CDK2. Based on the similarity of the ATP-binding pocket and interaction between the CDK proteins, we assumed that a number of speci c cancers could bene t from the combined inhibition of CDK2 with other CDKs [39]. Thus, the inhibition of CDK2 could be combined with various modalities of treatment, such as chemotherapy, radiotherapy, and targeted inhibitors, to reduce the proliferation of tumor cells [13,40,41].
Overall, our study suggested that benzydamine could inhibit the progression of ESCC tumors, both in vitro and in vivo, by suppressing the DNA replication signaling pathway by targeting CDK2 (Fig. 7). Thus, this study could provide useful information for the clinical application of benzydamine in ESCC chemoprevention and combination therapy.

Conclusions
In summary, our results revealed a novel drug against ESCC. Benzydamine directly binds with CDK2 and inhibits its kinase activity, down-regulating the phosphorylation levels of MCM2, c-Myc and Rb, induces cell cycle arrest, and exerts the inhibitory effect on ESCC proliferation in vivo and in vitro. This study provides a new strategy for cancer chemoprevention of esophageal squamous cell carcinoma.

Declarations
Ethics approval and consent to participate This study was approved by the Research Ethics Committee of Zhengzhou University. Written informed consent was provided by all patients for the use of the tissue samples.

Consent for publication
All authors have given their consent for the publication of this article.

Availability of data and materials
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation, to any quali ed researcher.

Competing interests
All authors declare that they have no competing interests.  The data in (d-f) were derived from three independent experiments and are presented as mean ± SD.
Asterisks (*, **, ***) indicate a signi cant decrease (p < 0.05, p < 0.01, p < 0.001, respectively) by one-way ANOVA followed by multiple-comparison tests. Similar results were obtained in three independent experiments.   and KYSE450 cells with or without benzydamine treatment for 24 h. Cells were stained with DAPI (blue in), anti-CDK2 (red) and anti-MCM2S41 (green). Double immuno uorescence indicated colocalization between MCM2S41 and CDK2. Scale bars, 10 μm. (e, f) Effect of benzydamine on the expression of biomarkers associated with the G1 to S phase transition during the cell cycle. KYSE150 and KYSE450 cells were treated with 0, 2.5, 5, 10, or 20 μM benzydamine and incubated for 24 h before detection. For the data in (a-d), asterisks (**p < 0.01, ***p < 0.001) indicate a signi cant difference between untreated control and benzydamine-treated KYSE150 and KYSE450 cells. Data are expressed as mean ± SD by one way ANOVA followed by multiple comparisons.   Benzydamine suppresses patient-derived esophageal xenograft tumor growth in vivo (a, b, c) Effect of benzydamine on tumor growth. Vehicle or benzydamine hydrochloride (5 or 50 mg/kg) was administered by gavage. Benzydamine signi cantly suppressed the tumor growth of cases EG20, LEG34, and LEG110.
Plotted tumor volume and tumor weight. (d) Immunohistochemistry analysis of Ki67, p-MCM2 S41 stained with DAB in cases EG20, LEG34, and LEG110 after administration of benzydamine. The positive cell rate was evaluated by Histo Quest-shortcut software. All data are shown as the mean ± SD. Asterisks (*, **, ***) indicate a signi cant decrease (p < 0.05, p < 0.01, p < 0.001, respectively) by one-way ANOVA followed by multiple-comparison tests.