Integrated analysis identifies RAC3 as an immune‐related prognostic biomarker associated with chemotherapy sensitivity in endometrial cancer

Abstract Endometrial cancer (EC) is one of the most common gynaecological malignant tumours with a high incidence, leading to urgent demands for exploring novel carcinogenic mechanisms and developing rational therapeutic strategies. The rac family of small GTPase 3 (RAC3) functions as an oncogene in various human malignant tumours and plays an important role in tumour development. However, the critical roles of RAC3 in the progression of EC need further investigation. Based on TCGA, single‐cell RNA‐Seq, CCLE and clinical specimens, we revealed that the RAC3 was specifically distributed in EC tumour cells compared to normal tissues and functioned as an independent diagnostic marker with a high area under curve (AUC) score. Meanwhile, the RAC3 expression in EC tissues was also correlated with a poor prognosis. In detail, the high levels of RAC3 in EC tissues were reversely associated with CD8+T cell infiltration and orchestrated an immunosuppressive microenvironment. Furthermore, RAC3 accelerated tumour cell proliferation and inhibited its apoptosis, without impacting cell cycle stages. Importantly, silencing RAC3 improved the sensitivity of EC cells to chemotherapeutic drugs. In this paper, we revealed that RAC3 was predominantly expressed in EC and significantly correlated with the progression of EC via inducing immunosuppression and regulating tumour cell viability, providing a novel diagnostic biomarker and a promising strategy for sensitizing chemotherapy to EC.


| INTRODUC TI ON
As the most common gynaecological tumour originating from the endometrial epithelium, endometrial cancer (EC) was diagnosed in more than 415,000 women in 2020, with the majority of cases occurring in postmenopausal women. 1,2 More worrisome is that EC-related mortality increased by an average of 1.9% per year. 3 However, the five-year relative survival rate of patients with highgrade or advanced-stage EC has still not experienced any substantial improvement for decades. 4 That earlier diagnosis produces a better prognosis is true for EC patients, while the recent diagnostic biomarkers depending on carbohydrate antigen 125 (CA125) and human epididymis protein 4 (HE4) have poor sensitivity and specificity for early EC detection. 5,6 Chemotherapy and hormone therapy remains the first-line treatment options for advanced and recurrent EC, yet they can potentially induce drug resistance, which in turn poses an inevitable clinical challenge. 7 Thus, overcoming the difficulty of early diagnosis and potentiating the sensitivity of chemotherapeutic drugs for EC have been of great urgency. Accumulating evidence suggests that oncogenes play critical roles in orchestrating tumour progression and drug resistance. 8 However, reports on the role of novel oncogenes in EC are still scarce.
RAC3, a member of the Rho GTPases family, participates in cytoskeleton formation, cellular and developmental biology and pathological processes. 9 Meanwhile, evidence illustrated that the RAC3 was highly expressed in a variety of human cancers, such as breast cancer, 10 lung cancer 11 and bladder cancer. 11 Furthermore, RAC3 inhibited apoptosis and promoted tumour invasion, high expression of which indicated a poor prognosis for breast cancer. 12,13 However, few studies have investigated its prognostic implication and clinical significance in EC. 14,15 In the present research, we revealed that RAC3 was accumulated in EC and served as a reliable diagnostic marker with a high AUC value. Subsequently, RAC3 was associated with poor prognosis in EC via its immunosuppressive phenotype and the regulation of tumour cell viability. Finally, we figured out that RAC3 overexpression enhanced the chemo-resistance of EC cells. In conclusion, our study highlighted RAC3 as a potential diagnostic and therapeutic target for EC patients.

| Tissue collection
Normal endometriums were obtained from eight women with uterine curettage. The tumour tissues were recruited from 61 patients with EC who underwent tumorectomy ( Table 1). The diagnosis was confirmed by two experienced pathologists. This project was approved by the Ethics Committee of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology.

| Single-cell RNA-Seq analysis
Single-cell raw data were loaded into R software (v4.0.5) and generated by the 'Seurat' package (v4. and tumour clusters, respectively. RAC3 was also visualized by feature plots according to mean expression.

| Kaplan-Meier plotter analysis and ROC analysis
The prognostic value of RAC3 was analysed by a Kaplan-Meier plotter (http://kmplot.com), with the hazard ratio (HR), 95% confidence interval (CI), and log-rank p-value. The receiver operating characteristic (ROC) curves were applied to determine the diagnostic accuracy using the 'survivalROC' package. The AUC value is the area under the ROC curve and represents the diagnostic ability of a single factor. AUC values of more than 0.9 represented high accuracy and 0.7 ≤ AUC ≤0.9 reflected moderate accuracy.

| Gene set enrichment analysis (GSEA)
GSEA was downloaded for the analysis of different pathways related to target genes (http://softw are.broad insti tue.org/gsea/). Related pathways were obtained from the molecular signatures database (MSigDB).

| Apoptosis assay
The apoptosis assay was performed using FITC Annexin V Apoptosis Detection Kit I (BD Biosciences) with the staining method. The apoptosis rate (sum of Annexin V − /PI + and Annexin V + /PI + cell proportion) was analysed by flow cytometry (Beckman Counter) within 1 h.

| Cell cycle assay
The cell cycle assay was performed using PI/RNase staining buffer (BD Biosciences) according to the instruction. The distribution of the cell cycle was analysed by flow cytometry (Beckman Counter).

| Western blot analysis
EC cells were harvested and lysed in RIPA lysis buffer (Servicebio).
After extraction, the total protein was measured by the Bicinchoninic

| Statistical analysis
All data were analysed using SPSS version 19.0 (IBM SPSS). Unpaired t-tests were used to compare differences between the groups.

| RE SULTS
3.1 | The highly expressed RAC3 in EC functions as a potential diagnostic marker and an independent prognostic marker RAC3 was originally identified in a screening for RAC family members involved in EC analysis ( Figure S1). Among the 33 tumour subtypes archived in TCGA cohorts, 15 kinds had statistically significant RAC3 expression differences between tumour and normal tissues, including EC ( Figure S1). In the TCGA-UCEC cohort, we found that RAC3 is highly expressed in EC tissues by comparison with paired or unpaired normal endometrium ( Figure 1A,B). To further validate these findings, we collected clinical specimens of EC (n = 61, Table 1) and normal endometrium (n = 8,  Figure 1D,E). In addition, the results also  showed the RAC3 expression levels were not related to the clinical stage ( Figure 1F,G). Furthermore, RAC3 had good diagnostic accuracy in EC, indicated by the area under the curve (AUC) value of 0.938 ( Figure 1H). Kaplan-Meier survival curve elucidated that EC patients with low-RAC3 expression had significantly longer survival time than those with high-RAC3 ( Figure 1I,J). These data showed that RAC3 was a potential diagnostic marker and prognostic factor for EC patients.

| RAC3 is mainly distributed in tumour epithelial cells
The previous subsections covered the RAC3 expression at the tissue level. To explore the principal distribution of RAC3, we employed the single-cell RNA-Seq dataset of two EC patients (Patient1#-2#).
We yielded two distinct cell populations on a UMAP plot including immune cells and non-immune cells (Figure 2A,B). Additionally, we subdivided cell populations into three subpopulations. The results indicated that RAC3 was mainly co-localized with EPCAM, a considerable marker highly expressed in epithelial tumour cells ( Figure 2C,D). To analyse RAC3 location intuitively, we detected its expression in collected specimens with IHC staining. The results showed that RAC3-positive cells were mainly enriched in tumour epithelial regions than in stromal regions. (Figure 2E,F). To sum up, the above results suggested that RAC3 was predominantly abundant in tumour epithelial cells.

| The highly expressed RAC3 in EC is correlated with an immunosuppressive tumour microenvironment
Next, we focused on how RAC3 reshaped the immune-  Figure 3G). The above results were also confirmed by a single gene correlation analysis in the TCGA cohort ( Figure 3H). Thus, the highlevel expression of RAC3 in EC orchestrated an immunosuppressive microenvironment and promoted tumour progression.

| RAC3 accelerates the proliferation and strengthens the chemo-resistance of endometrial cancer cells
To further investigate how RAC3 modulated tumour cell growth, the CCLE data involving 25 EC cell lines with different RAC3 expression levels were analysed by GSEA. The results indicated that the RAC3-low group was related to the negative regulation of the epithelial cell proliferation gene set ( Figure 4A). Moreover, a RAC3-high cell line (HEC-1B) and a RAC3-low (HEC-1A) were chosen according to CCLE and verified by Q-PCR and western blotting ( Figure 4B-D). We found that HEC-1B harboured a higher intrinsic proliferative rate than HEC-1A by CCK8 assay ( Figure 4E). Furthermore, the transcription level of RAC3 was modified by silencing RNA or overexpression plasmid to verify the cell viability modified by RAC3 ( Figure 4F . CCK8 assay was employed to detect the proliferation with mock, cisplatin, paclitaxel, and doxorubicin for 24-72 h. p value was denoted as *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

| RAC3 inhibits intrinsic cell apoptosis and drug-induced cell death
According to apoptosis-related GSEA analysis, RAC3 also modulated the apoptosis process of EC tumour cells ( Figure S2). Next, we detected the intrinsic apoptotic rate of EC cell lines and the results indicated that HEC-1B harboured a higher intrinsic apoptotic rate than HEC-1A by apoptotic rate analysis ( Figure 5A). To further verify the apoptosis process affected by RAC3, the transcription level of RAC3 was modified with silencing RNA or overexpression plasmid. The results showed that silencing RAC3 promoted the apoptosis of HEC-1B while overexpressing RAC3 constrained the apoptosis of HEC-1A ( Figure 5C,D,F,G). Meanwhile, RAC3-silenced cells displayed increased sensitivity to the above-mentioned chemotherapy drugs in varying degrees ( Figure 5C,E). Conversely, RAC3-overexpressed cells exhibited enhanced resistance to chemotherapy drugs ( Figure 5F,H). Overall, RAC3 inhibited EC cell apoptosis and restricted cell death caused by chemotherapy drugs. Consequently, developing drugs that directly target RAC3 needs urgent attention for EC patients.
There were several limitations in this study. First, there was no external data to predict the prognosis of RAC3 in EC patients.
Second, more experiments were needed to explain the molecular mechanisms and pathways of RAC3 in EC.
In conclusion, this study systematically revealed the expression patterns of RAC3 in EC. Our work also provided a practical biomarker for predicting the prognosis of EC and suggested that targeting RAC3 might provide new strategies for EC treatment. Dun-an Zang for the technical guidance and thank available online databases.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors have no conflict of interest to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
Publicly available datasets can be accessed by the websites provided in the methods and the data supporting this study are available from the corresponding author upon reasonable request.