2.1. Data Mining

The gene expression of FOXM1 in both tumor tissues and matching normal tissues was obtained in the GEPIA2 based on TCGA and GTEx databases (http://gepia2.cancer-pku.cn/#analysis). The gene expression of FOXM1 for EC was also obtained from the Gene Expression Omnibus database (GSE17025). The association between the FOXM1 expression and the infiltrated immune cell types in EC was analyzed using the “Immune-Gene” module in the online Sanger box (http://sangerbox.com/Index) and the Tumor Immune Estimation Resource (TIMER) 2.0 database (http://timer.cistrome.org/). The genetic alterations and the interactive network of FOXM1 in EC were obtained from cBioPortal (https://www.cbioportal.org). The differentially expressed genes (DEGs) and the gene set enrichment analysis (GSEA) related to FOXM1 in EC were identified using the LinkedOmics database based on the TCGA database (http://www.linkedomics.org/login.php). A protein–protein interaction network analysis was applied to analyze the expression of FOXM1 and the potential interaction partners using STRING (https://string-db.org/).

2.2. Cell Culture and Regents

From American Type Culture Collection (ATCC), four human EC cell lines (ISHIKAWA, HEC-1A, ECC1, HECE-1B) were purchased. All of these cell lines were incubated at 37°C in a humidified incubator with 5% CO₂ and cultured using the ATCC-indicated medium, which was supplemented with 1% antibiotics (100 units/ml penicillin and 100 μg/ml streptomycin) as well as with fetal bovine serum (FBS, 10% v/v); negative for mycoplasma contamination of all of above cell lines were verified before using them for further assays.

2.3. Clinical Samples

In this study, the human EC tissue microarray, including 120 EC samples and 17 noncancerous endometrial samples, was used for FOXM1 staining [21]. To investigate the correlation between FOXM1 and CDF66b, another human EC tissue microarray, including 13 EC samples and five noncancerous endometrial samples, was purchased from Servicebio, Wuhan, China. All procedures were carried out according to the rules of the China Ethical Review Committee.

2.4. Construction of shRNA-FOXM1 Lentivirus Constructs

1 × 10⁶ recombinant lentivirus-transducing units containing either shRNA-FOXM1 or shRNA-control (purchased from GeneCopoeia) mixture with 5 μg/ml polybrene were infected into EC cells. After 24 hr, the selected EC cells were maintained with 2 μg/ml puromycin. After the section, the efficiency of FOXM1 knockdown was confirmed using qRT-PCR and WB assays. Then the cells with higher knockdown efficiency were cultured for the following experiments for the cell function.

2.5. Isolation of RNA and Quantitative Real-Time PCR (qPCR)

2.6. Western Blotting (WB)

RIPA buffer with 1 mM PMSF and protease inhibitor cocktail was used to lyse EC cells. Then proteins contained in EC cell lysis solution were divided using accurate SDS-PAGE gel. After transferring onto nitrocellulose membrane (Shaanxi, China), the protein was blocked by 1% BSA within PBS/Tween at room temperature (RT) for 1 hr. Next, the membrane was further cultured overnight at 4°C with anti-FOXM1 (1 : 1,000, ab207298) or anti-GAPDH (1 : 1,000 dilution, Santa Cruz). After washing, the membranes were further incubated using corresponding secondary antibodies, and the Odyssey imaging system (LI-COR Biosciences, Lincoln, NE) was used for detecting targeted protein signals.

2.7. Viability of Cell and Formation of Colony Assay

For the viability of cell assay, 1,500 EC cells per well were added into a 96-well plate and incubated using 100 μl complete culture medium for 24, 48, 72, 96, and 120 hr. Each well medium was replaced at the indicated time points with solution of 10% (v/v) cell counting kit-8 (CCK-8; Share-Bio, China). After 1 hr of incubation, colorimetric dye that was produced by metabolized CCK-8 was scaled at 450 nm by a Power Wave XS microplate reader (BIO-TEK). For the colony formation assay, 1,000 EC cells per well were cultured within a six-well plate for 14 days in the cell culture medium. After staining with 0.1% crystal violet and counting, cell colonies were statistically analyzed.

2.8. Migration and Invasion of Cell Assay

For the cell migration assay, 2 × 10⁴ cells contained in 200 μl medium serum-free medium were put into the upper chambers of a 24-cell plate with 8 μm pore size cell culture inserts (Millipore, USA). After migrating for 24 hr at 37°C, 4% paraformaldehyde (PFA) and 0.1% crystal violet were used to fix and stain the migrated cells respectively. The migrated cells in six random fields under a light microscope were counted and statistically analyzed.

For the cell invasion assay, the Matrigel was put into the upper chamber of the inserts mixed with the serum-free medium, and then 5 × 10⁴ EC cells were put onto the Matrigel and migrated for 48 hr. After invading through the membrane and fixing with 4% PFA, the cells were stained using 0.1% crystal violet and analyzed.

2.9. Immunohistochemistry (IHC) Staining

As described previously in literature, immunohistochemistry (IHC) staining was performed [22]. In brief, the paraffin-embedded EC and normal tissue slides were deparaffinized and rehydrated, and then blocked using 10% BSA in PBS for 30 min. Next, the sections were incubated using different primary antibodies (FOXM1, Abcam ab207298; CD66b, Biolegend 305102) overnight at 4°C, followed by incubation using the matching secondary antibodies for 1 hr at RT. Then DAB substrate kit (Thermo Scientific) was used for treating the sections and hematoxylin staining for nuclear counterstaining. Lastly, a microscope (Carl Zeiss) was used for photography and all the sections were analyzed.

2.10. Statistical Method

Statistical analyses were performed using GraphPad Prism 9 (San Diego, CA) in this study. Means ± standard deviation (SD) was the presented form of data. Comparisons between the two groups were carried out by Student’s t-test. P < 0.05 was considered statistically significant.

3.1. FOXM1 Expression Is Highly Increased in EC and Closely Associated with Poor Prognosis of EC Patients

Based on the CCLE database, we found that mRNA of FOXM1 was diffusely detected in diverse tissues (Figure S1a). next, the mRNA expression level of FOXM1 in pan-cancer through TCGA and GTEx database was analyzed for checking the expression of FOXM1 in different types of cancer. The result displayed that FOXM1 mRNA was clearly upregulated in most kinds of cancers, indicating the vital functions of FOXM1 on the occurrence and progression of tumors (Figure S1b). Furthermore, we explored that FOXM1 high expression level predicted bleak prognosis in most types of cancer (Figure S1c).

In EC, the mRNA expressed level of FOXM1 was highly increased in EC tissues in comparison with corresponding health tissues (Figure 1(a)). Additionally, the gene expression data from GEO database GSE17025 displayed that FOXM1 mRNA expression level was significantly increased in EC and also was positively associated with EC pathological grades (Figures 1(b) and 1(c)). Highly expressed FOXM1 was indicated by Kaplan–Meier analysis significantly correlated with bleak prognosis in EC patients (Figure 1(d)). Furthermore, immunohistochemical staining of EC tissues indicated that the protein level of FOXM1 was clearly increased in EC tissues in comparison with the matching health tissues (Figure 1(e)).

3.2. Proliferation, Migration, as well as Invasion of EC Cells Was Promoted by FOXM1 In Vitro

To evaluate the role of FOXM1 within EC development, the mRNA and protein expressed levels of FOXM1 were first tested in four cell lines of EC including ISHIKAWA, HEC-1A, ECC1, and HECE-1B by qPCR and WB. Among them, ISHIKAWA and HEC-1A cells showed relatively higher mRNA and protein expression of FOXM1, which were chosen for the subsequent experiments (Figure 2(a)). In order to verify the oncogenic functions of FOXM1 in EC, we established stable EC cell lines with knockdown of FOXM1 by short hairpin RNA (shRNA) in ISHIKAWA as well as HEC-1A cells, which were verified by qPCR and WB analyses (Figure 2(b)). The FOXM1 knockdown significantly inhibited the proliferation of the two EC cell lines above as shown by CCK8 as well colony formation assay (Figures 2(c) and 2(d)). What’s more, the migration as well as invasion were also suppressed by FOXM1 knockdown in both ISHIKAWA and HEC-1A cells (Figures 2(e) and 2(f)).

3.3. The Genetic Alteration and the Interaction Networks of FOXM1 in EC Patients

The genetic alterations, correlations, and interaction networks of FOXM1 were analyzed in EC based on the TCGA database using the cBioPortal online tool. Two or more alterations were detected in EC using the TCGA database (Figure 3(a)). FOXM1 was altered in 4% of patients with EC (Figure 3(b)). In addition, the protein–protein interaction networks were analyzed for exploring possibility of proteins with FOXM1 using the STRING approach. The interaction network showed that FOXM1 was closely correlated with those proteins who are involved in the cell cycle, such as cyclin-dependent kinase 1 (CDK1), CDK2, Cyclin B1 (CCNB1), as well as CCNB2, indicating FOXM1 may act as an important regulator of the tumor cell cycle (Figure 3(c)). Those might be closely associated with FOXM1-related molecular functions.

3.4. The Coexpression Analysis of FOXM1 in EC

In order to gain deeper understanding of the exact function as well as the potential mechanism of FOXM1 in EC, the coexpression network of FOXM1was identified in EC using the LinkedOmics database (LinkFinder module) and found that 4,001 genes (red color dots) were positively associated with FOXM1, and 3,791 genes (green color dots) were negatively associated with FOXM1 (Figure 4(a)). The top positive and negative 50 genes associated with FOXM1 were shown in the heatmaps, respectively (Figures 4(b) and 4(c)). These highly coexpressed genes of FOXM1, studied by GO biological and KEGG pathways analysis, were largely gathered to different progress linking to cell cycle, including DNA replication, cell cycle G2/M phase transition, cell cycle checkpoint, and mitotic cell cycle phase transition (Figures 4(d) and 4(e)). Besides, coexpressed genes of FOXM1 also participated in leukocyte migration, response to chemokines, cell adhesion molecules, acute inflammatory response, etc. (Figures 4(d) and 4(e)). Taken together, those analyses indicated that FOXM1 has influence on EC cell cycle and the infiltration of immune cells in EC.

3.5. FOXM1 Enhances the CD276 Expression in the Tumor Cells and the Neutrophil Infiltration in the Tumor Microenvironment of EC

To check whether FOXM1 was involved in the immune microenvironment of tumor or not, the “Immune-Gene” analysis was performed using the Sanger box and GEPIA2 database, which showed that FOXM1 has a strong positive correlation with CD276 expression (Figure 5(a), Figure S2a). Also, mRNA of CD276 was notably upregulated in EC tissues in comparison with corresponding health tissues (Figure 5(b)). Furthermore, FOXM1 knockdown remarkedly suppressed CD276 expression not only in ISHIKAWA but also in HEC-1A cell lines (Figure 5(c)). Next, we further explored a significantly strong association between FOXM1 and neutrophils (Figure 5(d), Figure S2b). Therefore, we checked the neutrophil infiltration in EC tissue microarray. IHC assay displayed that a number of infiltrated neutrophils in the tumor microenvironment were positively associated with the expression of FOXM1 in EC (Figure 5(e)). Taken together, FOXM1 may contribute to regulating the tumor immune microenvironment through neutrophil infiltration in EC.