Combination of β2-AR Blocker and COX-2 Inhibitor Suppresses Development of Oral Squamous Cell Carcinoma

Background To study the effect of inhibiting progression of oral squamous cell carcinoma (OSCC) through blocking β2-adrenergic receptor (β2-AR) and inhibiting Cyclooxygenase-2 (COX-2). Methods Transwell invasion assay and wound-healing assay were used to investigate whether invasion and metastasis of OSCC cells (SCC9, Cal27 cell lines and primary OSCC cells) would be inhibited after treated by β2-AR blocker (ICI118,551), COX-2 inhibitor (Etodolac) or both. Real time quantitative polymerase chain reaction (RT-qPCR), Western blot and enzyme linked immunosorbent assay (ELISA) were used to detect expression change of genes related to invasion and metastasis in OSCC cells after drug treatment. In vivo, 6-8 weeks old female Balb/c nude mice were randomly divided into control group, ICI118,551 treatment group, etodolac treatment group and combined treatment group. Cal27 cells were used to establish OSCC xenograft models. The survival days, tumor sizes and lymph nodes metastasis were compared among these groups. Immunohistochemistry, Western blot and ELISA were carried out to detect expression change of invasion and metastasis relative genes. Results ICI118,551 or etodolac could inhibit invasion and metastasis of OSCC cells, and the inhibition of combined treatment was more signicant. Moreover, RT-qPCR and Western blot showed that epidermal growth factor receptor (EGFR), transforming growth factor-β1 (TGF-β1), interleukin-1β (IL-1β), matrix metalloproteinase 2 (MMP2) were downregulated after treatment, especially in combined treatment group. In vivo, the combined treatment group could signicantly prolong survival days and inhibit tumor size. Immunohistochemistry, Western blot and ELISA showed the expression of above-mentioned genes were also downregulated.


Background
Head and neck cancer is the seventh most common cancer worldwide, accounting for 3% of all cancers [1]. Oral squamous cell carcinoma (OSCC) is the most common cancer among head and neck cancer, accounting for about 90% of cases [2]. Therefore, the diagnosis and treatment of OSCC are very crucial.
At present, the main treatment methods of OSCC include surgery, radiotherapy and chemotherapy.
Surgery combined with radiotherapy and chemotherapy is still the standard combined and sequential treatment of OSCC [3][4][5]. However, despite the combination of multiple treatments, the overall 5-year survival rate of OSCC patients has not been signi cantly improved in the past few decades, OSCC still belongs to the malignant tumor with low survival rate, just about 50% [6,7]. So, it is important to improve the overall survival rate of OSCC patients.
In past few years, different studies pointed out that β2-adrenoceptor (β2-AR) was overexpressed and played an important role in the occurrence and development of cancers [8][9][10]. It is closely related to the growth, angiogenesis, invasion and metastasis of cancers, including OSCC [11][12][13][14]. Meanwhile, our previous study and other studies found that blocking β2-AR could inhibit the development of OSCC, propranolol combined with cisplatin had a synergistic effect and could weaken the resistance of OSCC cells after radiotherapy and chemotherapy treatment [15][16][17].
Recently, some studies of colon and breast cancer found that the combination of β2-AR blocker and COX-2 inhibitor could signi cantly inhibit the occurrence and development of malignant tumors [30,31]. They found that perioperative use of β-AR blockers and COX-2 inhibitors might improve immune competence and inhibit tumor metastasis [32]. However, there is not any study reports whether combination of β2-AR blocker and COX-2 inhibitor is effective on treatment of OSCC or not. We put forward the hypothesis that combined treatment of β2-AR blocker and COX-2 inhibitor could inhibit the development of OSCC. So, in this study, we explored the effect of combination of β2-AR blocker and COX-2 inhibitor on OSCC in vitro and vivo, as well as the related molecular mechanisms Methods All methods were performed in accordance with the relevant guidelines and regulations.
The primary OSCC cells for the study were from tongue squamous cell carcinoma tissue of one patient.
The patient, a 29-year-old man (T4N2M0); had left tongue low-differentiated squamous cell carcinoma with bilateral cervical lymph node metastases. Fresh specimens from tongue carcinoma were dissected promptly after operations carried out at the Department of Oral and Maxillofacial Surgery, Guangxi Medical University School and Hospital of Stomatology. Cells were cultured in DMEM supplemented with 14% FBS and penicillin (50 units/mL)/streptomycin (50 μg/mL) in a humidified atmosphere (5% CO 2 ) at 37 °C.

ICI118,551
To block β2-AR activation, we used the selective β2-AR blocker ICI118,551 (MedChemExpress, USA). The drug was dissolved in dimethyl sulfoxide (DMSO) and added to a mixture of culture medium to 5μmol/L concentration for vitro assay. In vivo assay, it was dissolved in water and diluted to 0.5 mg/ml concentration.

Etodolac
The semi-selective COX-2 inhibitor Etodolac (MACKLIN, China) was used to inhibit COX-2 activation. It was dissolved in DMSO and added to a mixture of culture medium to 5μmol/L concentration for vitro assay. In vivo assay, it was dissolved in corn oil and diluted to 12.5 mg/ml concentration.

Animals and counterbalancing
Balb/c nude mice, purchased from the Animal Research Laboratory of Guangxi Medical University, were distributed randomly into a control group, ICI118,551 treatment group, etodolac treatment group and combined treatment group, 6 mice in each group. Cal27 cells (1 × 10 7 /mL) in 50 μL phosphate buffer saline (PBS) were subcutaneously injected into in the ank of the tongue at day 0. Tumor growth was monitored in an assessor-blind trial, and tumor size was measured every 2 days. Tumor volume was estimated using the following equation: tumor volume = ((A × B 2 )/2) mm 3 , where A is the largest dimension and B is the largest dimension perpendicular to A. ICI118,551 (5 mg/kg in a 0.5 mg/ml concentration) and Etodolac (50 mg/kg in a 12.5 mg/ml concentration) were given orally to the mice daily. PBS was given orally to the control group daily. Mice were euthanized 24 h after completion of treatment. The tumor and submandibular lymphatic node histology were analyzed. The Kaplan-Meier method was used to evaluate survival. All experimental procedures received approval by the Laboratory Animal Care and Use Committees of Guangxi Medical University.
Wound healing assay OSCC cells were grown in 6-well plates to until reaching 100% con uence. A cell scratch-wound was established by a 200µl pipette tip. After this, the culture medium was changed to different drugs, which were diluted in DMEM or DMEM/F12. After 24h, the wound width was photographed, and the percentage of wound closure was determined based on the wound width at 0 h and the areas were measured using ImageJ software.

Transwell invasion assay
Cell invasion assays were performed using a modi ed 24-well chamber with a membrane which was precoated with Matrigel (Corning, USA). 100 µl of cells (1 × 10 5 /ml) in FBS-free medium were seeded into the upper chamber, while different drugs were added 100μL into the upper chamber. In each lower chamber, 500μL medium containing 10% or 14% FBS was placed. After 24h, invasive cells on the undersurface of the membrane were stained with crystal violet (Thermo Fisher Scienti c, USA) staining solution for 20 minutes. Quanti cation was measured by counting the invasive cells in ve randomly selected elds.
Real time quantitative polymerase chain reaction Total ribonucleic acid (RNA) was isolated using RNAiso Plus (Takara Bio, Japan) and according to the and reverse transcribed using PrimeScript™RT reagent Kit (Takara Bio, Japan) according to the instructions. We used SYBR ® Premix Ex Taq™II Kit (Takara Bio, Japan) to detect expression of messenger ribonucleic acid (mRNA). The polymerase chain reaction (PCR) primer sequences used for each gene were showed in Table1. PCR reaction was carried out with Step One Plus TM Real-Time PCR System (Takara Bio, Japan): the rst step is initial denaturation followed by 95°C for 30s, the second step is denaturation followed by 40 cycles of 95°C for 5s and 60°C for 30s. All the values were standardized with 2 -ΔΔCT method.

Western blot analysis
Total protein was extracted from cells or tissues using RIPA buffer in the presence of protease inhibitor (Fdbio science, China) and phosphatase inhibitor cocktail (Fdbio science, China). Proteins at 60μg per lane. Target proteins were separated based on their molecular weight on 10% SDS-PAGE gels (Fdbio science, China) and then transferred to poly vinylidene uoride membranes (Millipore, USA). The membranes were incubated with the appropriate primary antibody overnight at 4°C. The primary antibody used for each gene were as follows: Anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (FD0063, Fdbio science, China) at 1/5000 dilution, Anti-epidermal growth factor receptor (EGFR) (ab52894, abcam, USA) at 1/1000 dilution, Anti-transforming growth factor-β1 (TGF-β1) (ab179695, abcam, USA) at 1/1000 dilution, Anti-matrix metalloproteinase 2 (MMP2) (ab92536, abcam, USA) at 1/1000 dilution, Anti-VEGF-A (ab52917, abcam, USA) at 1/5000 dilution. Then membranes were incubated in corresponding species secondary antibodies for 1h. GAPDH was used as an internal control. Protein expression levels were detected by ChemiDocM universal imager system (Bio-Rad, USA) and gray value was analyzed by Image lab software. The relative expression of target protein (%) = gray value of target protein / gray value of GAPDH.

Enzyme-linked immunosorbent assay
Cytokine interleukin-1β (IL-1β) from culture supernatants of treated cells and serum of xenograft mice were determined by an enzyme-linked immunosorbent assay (ELISA) method using an ELISA kit (Solarbio, China) according to manufactures' introduction.

Statistical analysis
Statistical analysis was carried out using SPSS software (version 22.0, IBM, USA) and GraphPad Prism (version 7.00, GraphPad Software, Inc., La Jolla, USA). The data are presented as mean ± standard deviation. Analysis of Variance (ANOVA) was used to compare the means of multiple groups. P-values less than 0.05 were considered statistically signi cant.

Combined application of ICI118,551 and etodolac inhibited the migration and invasion ability of OSCC cells
The results of wound-healing assay showed that ICI118,551 (5μmol/L) or etodolac (5μmol/L) alone could inhibit the migration ability of OSCC cells, while the migration ability of the cells treated with the combination of two drugs was more signi cantly inhibited than that treated alone. The statistical results of three OSCC cells (Cal27, SCC9 and the primary OSCC cells) showed that compared with the control group, the other three groups could signi cantly inhibit the migration ability of OSCC cells. Compared with ICI118,551 group (ICI) and etodolac group (Eto), the combination group (ICI+Eto) could substantially inhibit the migration ability of OSCC cells (Fig.1a-f). Same as the wound-healing assay, the results of transwell invasion assay showed that the invasion rate is the lowest in ICI+Eto group of each OSCC cell (Fig.1g-j).  (Fig.2a-b). And the average survival time was 13.33 ±1.022 days in control group, 20.833±1.327 days in the ICI group, 15±0.365 days in Eto group and 23.5 ±1.962 days in combination group. The results showed that combined application of ICI118,551 and etodolac could inhibit the growth of tumor and prolong the overall survival time of Balb/c nude mice with OSCC (Fig.2c).
After mice were dead, the submandibular lymph nodes were resected to detect whether the combination of drugs could inhibit the metastasis of submandibular lymph nodes. H&E staining was combined with immunohistochemical staining, which was carried out to detect the expression of cytokeratin (CK), to analyze the metastasis of submandibular lymph nodes. The results showed that submandibular lymph node metastasis was found in 3 mice in control group, 2 in ICI group, 2 in Eto group, and only 1 in ICI+Eto group (Fig.2d-g).

Combined application of ICI118,551 and etodolac down-regulated genes relate to OSCC invasion and metastasis
After combined treatment of ICI118,551 and etodolac for 48 hours, the expression of some genes relate to invasion and metastasis in OSCC cells were detected by real time quantitative polymerase chain reaction (RT-qPCR), Western blot and ELISA. Results of RT-qPCR showed that EGFR, TGF-β1, IL-1β and MMP2 were substantially down-regulated after combined treatment. Among these, EGFR was up-regulated after treated with ICI118,551 in Cal27, but it was down-regulated after combined treatment and the relative expression was also lower than etodolac treatment. Moreover, the change of relative expression of MMP2 was not obvious after treated with ICI118,551 in two OSCC cell lines, but it was down-regulated in three OSCC cells after combined application. However, VEGF-A was not down-regulated in two OSCC cell lines after combined treatment, it was down-regulated after combined treatment in primary OSCC cell instead (Fig.3). Results of Western blot showed that EGFR, MMP2, VEGF-A were down-regulated after combined treatment in three OSCC cells, TGF-β1 was down-regulated in Cal27 and SCC9 cell lines (Fig.4a-o). ELISA showed that IL-1β was down-regulated after combined treatment, the concentration of IL-1β in combination group was the lowest. (Fig.4p) Genes related to invasion and metastasis of OSCC in xenograft animal models were also downregulated after combined treatment.
In vivo assay, another four groups were set as above. On day 12, mice were sacri ced, tongue tumors and blood samples were collected for the analysis of genes relate to invasion and metastasis. Results of IHC showed that the mean optical density of EGFR, TGF-β1, MMP2 and VEGF-A in combination group were lower than other three groups, as same as results of Western blot (Fig.5). Interestingly, Western blot showed that the expression of VEGF-A in ICI or Eto group was either higher than control group. However, it was down-regulated in combination group (Fig.6a-e). In ELISA, the mean concentration of IL-1β in ICI or Eto group was either lower than control group. Same as vitro assay, in combination group, the concentration of IL-1β was the lowest (Fig.6f).

Discussion
OSCC is the most common cancer in oral and maxillofacial region. Surgery combined with chemotherapy and radiotherapy is still the standard treatment of OSCC, but the overall 5-year survival rate of OSCC patients has not been signi cantly improved in the past few decades. Recent years, β2-AR and COX-2 were found that overexpressed in OSCC, different studies also found that β2-AR blocker or COX-2 inhibitor could inhibit the progression of OSCC in vitro and vivo. However, a few clinical trials showed that postoperative use of β-AR blocker couldn't increase the survival rate of OSCC patients [33]. For COX-2 inhibitor, some clinical trials reported that celecoxib, a selective COX-2 inhibitor, could improve the prognosis of OSCC patients [34,35]. Recent years, some studies of colon and breast cancer found that the combination of β2-AR blocker and COX-2 inhibitor could inhibit the occurrence and progression of malignant tumors and improve the prognosis of patients [30,31]. So, we wanted to know whether the combined treatment could also inhibit the progression of OSCC.
To our knowledge, it is the rst time to study the effect of combined use of β2-AR blocker and COX-2 inhibitor in OSCC. In vitro, we found that after combined treatment, the migration and invasion ability of OSCC cells were signi cantly inhibited. In vivo, the combined application of ICI118,551 and etodolac could inhibit the growth of tumor and prolong the overall survival time of Balb/c nude mice with OSCC. Moreover, to explore whether combined use of these two drugs could inhibit the progression of OSCC by regulating the expression of some genes related to invasion and metastasis further. We detected the expression of EGFR, TGF-β1, IL-1β, MMP2 and VEGF-A after combined treatment. The results of RT-qPCR showed that EGFR, TGF-β1, IL-1β and MMP2 were down-regulated in OSCC cells after combined treatment, except VEGF-A. The protein expression level of these above four genes were also the lowest in the four groups in vitro and vivo. EGFR, TGF-β1, IL-1β and MMP2 were all related to invasion and metastasis of malignant tumors, including OSCC [36][37][38][39][40][41]. So, we have veri ed that combined use of ICI118,551 and etodolac can inhibit the development of OSCC and the expression of EGFR, TGF-β1, IL-1β and MMP2 indeed.
Among these, we found that the expression change of EGFR was signi cant. EGFR, also known as ErbB1, belongs to EGF receptor family and is closely related to the growth and development of epidermal cells [42]. It has been found that EGFR is overexpressed in more than 80% invasive head and neck squamous cell carcinoma (HNSCC), which is closely related to the poor prognosis of the patients [43,44]. EGFR can promote tumor cell proliferation, metastasis, anti-apoptosis and angiogenesis by activating its downstream signaling pathways Ras/Raf/MEK/ERK and phosphatidylinositide 3-kinases/protein kinase B (PI3K/Akt) [45,46]. Some studies reported that EGFR also could promote the tumor immune escape mechanism [47,48]. In recent years, EGFR-targeted therapy has been used to clinical treatment of OSCC [49][50][51][52]. It was also mentioned in the clinical guideline of national comprehensive cancer network (NCCN) in 2019 that cetuximab, a kind of EGFR-targeted drugs, could combine with radiotherapy and chemotherapy in the systematic treatment of OSCC [53]. However, many families with poor nancial condition cannot afford the expensive price of cetuximab, and some studies reported that cetuximab had effective therapeutic effect on only 10-20% cancer patients [52,54]. In our study, we found that the combination of common and inexpensive β2-AR blocker and COX-2 inhibitor could signi cantly downregulate the expression of EGFR and inhibit the development of OSCC in vitro and vivo. Nasry et al.
reported that cetuximab could suppress the proliferation of cancer cells by inhibiting COX-2/PGE2/EGFR pathway [23]. But the molecular mechanism of β2-AR regulates the expression of EGFR has not been determined. Further studies need to explore the molecular mechanism of β2-AR combines with COX-2 to regulate EGFR. If we can clarify the mechanism clearly further, the combined use of β2-AR blocker and COX-2 inhibitor will be another treatment which can inhibit the expression of EGFR. And there will be one more choice of pharmaco-adjuvant therapy for the OSCC patients who cannot afford the price of cetuximab.
In this study, we also found that TGF-β1, IL-1β and MMP2 were down-regulated after combined treatment. TGF-β1 is a crucial inducer of epithelial-mesenchymal transition (EMT) [55]. EMT refers to the biological process in which cells transition from epithelial state to mesenchymal state. It is closely related to embryonic development, wound healing, and cancer metastasis especially [37]. In HNSCC, the expression of TGF-β1 is higher than normal tissues. Chen, et al. reported that the over-expression of TGF-β1 in OSCC was related to lymph node metastasis, recurrence and poor survival rate [56]. Moreover, IL-1β plays an important role in tumor microenvironment. IL-1β can increase the proliferation of dysplasia oral cells, stimulate oncogenic cytokines, and promote aggressiveness of OSCC [39,57]. Similar to the abovementioned genes, MMP2 was found to be overexpressed in HNSCC. Among MMPs, MMP2 is one of the most commonly identi ed in HNSCC and is associated with invasion and metastasis [58,59]. Meanwhile, MMP2 was found that could be used for evaluating lymph node metastasis and tumor progression in OSCC [60]. So, the molecular mechanism still needs to be clari ed. And, the clinical trials also need to proceed to determine the effect of this combined treatment in the future.

Conclusions
In conclusion, this study suggests that combined use of ICI118,551 and etodolac can inhibit the development of OSCC in vitro and vivo, especially the ability of invasion and metastasis. Meanwhile, the combination of these two drugs is more effective than ICI118,551 or etodolac treatment alone, the combined use of β2-AR blocker and COX-2 inhibitor will be a potential treatment for OSCC via downregulating expression of EGFR, TGF-β1, IL-1β and MMP2.  Table 1 The RT-qPCR primer sequences used for each gene we detected.

Genes
The