Knockdown of DIAPH3 Inhibits the Proliferation of Cervical Cancer Cells through Inactivating mTOR Signaling Pathway

Cervical cancer (CC) ranks fourth for both incidence and mortality among females in worldwide. Therefore, it is urgent to explore new therapeutic and diagnostic targets for cervical cancer. Diaphanous-related formin 3 (DIAPH3) has been identified to play crucial roles in many malignant tumors. But its function and potential mechanism in CC remain largely unknown. In our study, DIAPH3 was frequently upregulated in CC tissue samples and increased expression of DIAPH3 was associated with poor overall survival according to several databases. Through in vitro and in vivo experiments, we found that decreased expression levels of DIAPH3 significantly inhibited the progression of CC. The GSEA analysis and western blot assay indicated that DIAPH3 was associated with the mTOR signaling pathway. The univariate and multivariate Cox analysis indicated that DIAPH3 was an independent prognosis risk factor in TCGA-CESC. And we confirmed that DIAPH3 expression was clearly related to tumor immune infiltrating cells (TIICs) by the analysis of CIBERSORT and TIMER databases. Taken together, we revealed that DIAPH3 plays as an oncogene through mTOR signaling pathway and DIAPH3 might be a potential prognostic biomarker in CC.


Introduction
Cervical cancer (CC) is a kind of common malignancies worldwide. And both the incidence and the mortality of CC remain fourth among females [1].
ere is a substantial increase in cervical cancer incidence in China [2]. Most cases are often diagnosed at an advanced stage. Almost all cases are due to human papillomavirus (HPV) infection [3]. e treatments of CC include surgery or a concurrent chemoradiotherapy program composed of cisplatin-based chemotherapy with brachytherapy and external beam radiotherapy [4]. Despite the advancement of diagnosis and treatment technology, the five-year survival rates of advanced CC are about 30%-50% [5]. Hence, it is urgent to explore new therapeutic and diagnostic targets for cervical cancer.
e diaphanous-related formin 3 (DIAPH3) is considered a core modulator of the cytoskeleton [6] and the host gene locates on human chromosome 13q21.2 [7]. DIAPH3 is investigated to reshape the cytoskeleton [8] and regulates the expression of GSK3β to preserve microtubule stabilization [9]. Moreover, previous studies reported that DIAPH3 plays many crucial roles in many malignancies. Di Vizio D et al. confirmed that DIAPH3 facilitates the metastasis process of prostate cancer [10]. Jiang et al. verified that the overexpression of DIAPH3 suppresses the migration of triplenegative breast cancer [11]. Dong et al. reported that DIAPH3 facilitates the malignant biological behavior of HCC cells through activating β-catenin/TCF signaling pathway [12]. To our knowledge, the functions of DIAPH3 and its molecular mechanism in cervical cancer were largely unknown. In the present study, we revealed that DIAPH3 plays as an oncogene through mTOR signaling pathway. We revealed that DIAPH3 promoted proliferation through mTOR signaling pathway in cervical cancer.
In recent years, tumor immunity has gradually attracted the attention of many researchers. e regulatory T cells (T regs) could affect the role of cytotoxic T cells to facilitate tumor cells to get rid of the immune system [13]. A large number of papers have deeply explored the complex tumor immune regulation process and immune checkpoints to influence the process of antitumor immune response. For example, the advancement of antibody drugs against PD1 and PD-L1 for the therapy of prostate cancer and renal cell carcinoma has shown great efficacy in clinical practice [14][15][16][17]. erefore, in this study, we further evaluated the prognostic value of DIAPH3 and assessed the association between the expression of DIAPH3 and tumor immune infiltration level in CC.

CC Tissues and Cells.
e CC tissues were collected from the Department of Obstetrics and Gynecology, e Affiliated Jiangyin Hospital of Southeast University Medical College. All patients have signed informed consent. Human cervical cancer cell lines C33 A, Caski, Siha, and HeLa were purchased from the Cell Center of Shanghai Institutes for Biological Sciences. HaCaTcells were used as normal controls, which were obtained from Nanjing Kaiji Biotechnology Company. Cells were cultured in Dulbecco's modified Eagle's medium (Gibco, USA) supplemented with 10% fetal bovine serum (Gibco, USA) and 1% penicillin/streptomycin (Gibco, USA) at 37°C with an atmosphere of 5% CO 2 in a humidified cell chamber.

CCK8 Assay.
We utilized the Cell Counting Kit-8 (Dojindo, Kumamoto, Japan) to assess the proliferation ability of CC cells. Cells in different treatment groups were seeded into 96-well plates (1,000 cells/well) and cultured with complete medium for 5 days. en the cells were incubated with CCK-8 solution for 2 h. Next, the proliferation ability was evaluated by OD value at 450 nm. Each sample was performed in triplicate.

Colony Formation
Assay. Cells in different treatment groups were seeded in 6-well plates (500 cells/well) and placed in the incubator for 2 weeks. After that, the cells were stained with crystal violet for 30 minutes. e numbers of colony formations were counted. e experiment was performed in triplicate.

EdU Assay.
To further assess the proliferation ability, we also utilized the EdU kit (Beyotime Biotechnology, Shanghai, China). e treated cells were firstly plated into 96-well plates (3 × 10 4 /well) and cultured for 24 hours. en we added the EdU reagent (50 μmol/L). Next, cells were permeabilized with 0.5% TritonX-100 for 10 minutes after fixing in 4% formaldehyde for 2 hours. 1×Apollo reaction solution (400 μL) reacted with EdU (Beyotime Biotechnology, Shanghai, China) for 30 minutes, and DAPI (400 μL) stained the nucleus after washing with PBS for three times. Finally, the microscope was used to capture the images of cells. e experiment was performed in triplicate.

Animal Experiment.
To establish the tumor growth models, we purchased four-week-old female BALB/c nude mice. e transfected cells (2 × 10 6 cells/100 μL) with PBS suspension were subcutaneously injected. After 5 weeks, we sacrificed the mice and observed the tumor progression.

Analysis of Tumor Immune Infiltrated Cells (TIICs).
We conducted the association analysis between the expression of DIAPH3 and the fractions of tumor infiltrated immune cells in CESC by using the data from the CIBERSORT and TIMER databases. e correlation was examined by Spearman test.

Statistical Analysis.
e data were presented as mean ± SD. We statistically analyzed the data through Student's t-test using GraphPad Prism 7. Survival analysis was calculated by Kaplan-Meier plots. p values <0.05 represented a statistically significant difference. e association of DIAPH3 with clinicopathological characteristics was analyzed by the χ 2 test.

DIAPH3 Was Overexpressed in Cervical Cancer and
Showed a Poor Prognosis. We first retrieved data of cervical cancer from the TCGA database, and we selected DAIPH3 as a potential oncogene in CC according to fold change >2 and P value < 0.01 (Figure 1(a)). To further confirm it in a large cohort of patient samples, we analyzed the expression of 2 Journal of Oncology Journal of Oncology DIAPH3 in CC in TCGA database, consisting of 305 tumor tissues and 3 normal tissues. It showed that DIAPH3 was significantly overexpressed in CC compared with normal samples (Figure 1(b)). Moreover, we found that high expression of DIAPH3 indicated a poor overall survival in CC according to TCGA database ( Figure 1(c)). Simultaneously, we verified the above results through GEPIA database and it showed the similar outcome (Figures 1(d) and 1(e)). en we added e Human Protein Atlas and we found that patients with high expression of DIAPH3 exhibited a lower overall survival rate ( Figure 1(f )). Moreover, we performed the analysis of DIAPH3 expression with survival in CC by using the TCGA-CESC cohort. Combined with the univariate cox (Table 1) and multivariate cox analysis (Table 2), it indicated that the expression of DIAPH3 was an independent prognostic risk index of cervical cancer. And we found that DIAPH3 expression was significantly upregulated in 42 pairs of CC samples compared to normal cases ( Figure 1(g)). To determine the functions of DIAPH3 in CC in vitro, we examined DIAPH3 expression in four CC cells (HeLa, Siha, C33 A, and Caski) and HaCaT cells (normal control). It was shown that DIAPH3 was increased in CC cells, especially in HeLa and Siha cells (Figure 1(h)), so we chose these two cell types for further investigation. Additionally, we transfected si-NC and si-DIAPH3 into HeLa and Siha cells, respectively, and we validated the knockdown efficiency of DIAPH3 using qRT-PCR and western blot assay. We observed that the expression of DIAPH3 was obviously decreased in the si-DIAPH3 group (Figures 1(i)-1(k)). Additionally, we evaluated the association between the expression of DIAPH3 and patients' pathological characteristics. In the group of tumor size larger than 3 cm, the expression level of DIAPH3 was significantly upregulated (Table 3).

Knockdown of DIAPH3 Inhibited the Progression of CC In Vitro and In Vivo.
To determine the biological functions of DIAPH3 in CC, we transfected si-NC and si-DIAPH3 into HeLa and Siha cells, respectively. After transfection, we utilized different means to detect cell proliferation in vitro. In CCK8 assay, it was shown that the ability of proliferation in si-DAIPH3 group was notably declined compared to that in si-NC group in HeLa and Siha cells (Figures 2(a) and 2(b)). In the colony formation assay, the results showed that knockdown of DIAPH3 weakened colony formation ability in HeLa cells (Figures 2(c) and 2(e)), and it showed similar results in Siha cells (Figures 2(d) and 2(f )). en EdU incorporation assay was utilized for further assessment of the influence of DIAPH3 on CC cell proliferation. We found that the rate of EdU in the si-DIAPH3 group was obviously declined compared to that in si-NC group in HeLa cells (Figures 2(g) and 2(h)), and the tendency was consistent with that in Siha cells (Figures 2(i) and 2(j)). demonstrated that high expression of DIAPH3 was positively correlated with the mTOR pathway (P value < 0.01). Next, we used western blot assay for further investigation. In HeLa cells, it showed that downregulated level of DIAPH3 decreased the expression of related proteins p-AKT, mTOR, and p-p70s6k and increased the expression of PTEN in mTOR signaling pathway and there is a similar trend in Siha cells (Figure 4(c)). Accordingly, these data indicated that knockdown of DIAPH3 inhibited the proliferation of cervical cancer cells through inactivating the mTOR signaling pathway.

Correlation between the Expression of DIAPH3 and the Fractions of TIICs in CC.
To further probe the association between DIAPH3 expression and TIICs in CC, we analyzed the subgroups of tumor-infiltrating immune cells by analysis of the data from CIBERSORT. en we constructed 22 kinds   Journal of Oncology 5 of immune cell profiles ( Figure 5(a)). As shown in Figure 5(b) and Table 4   Journal of Oncology Meanwhile, we utilized the TIMER database to further analyze the correlation between the expression of DIAPH3 and diverse kinds of TIICs. Figure 7(a) shows that the expression of DIAPH3 was positively associated with the infiltrating levels of T cell CD8+ naive (cor � 0.13, p � 3.04e-02) and T cell CD4+ 2 (cor � 0.309, p � 1.48e-07) and negatively related to the infiltrating levels of B cells (cor � -0.161, p � 7.3e-03), monocytes (cor � -0.166, p � 5.76e-03), macrophages (cor � -0.166, p � 5.52e-03), and myeloid dendritic cells (cor � -0.198, p � 9.48e-04). We found that DIAPH3 expression is associated with part of the markers of different immune cells (Table 5), especially negatively related to the markers of B cell and M1 macrophages (Figure 8).

Correlation between DIAPH3 Expression and Immune
Checkpoint. Firstly, we divided the expression of DIAPH3 in TCGA-CESC into high and low groups according to the median value. e results of the correlation analysis between immune cell subgroups and functions revealed that T_cell_costimulation, T_cell_coinhibition, checkpoint, APC_coinhibition, APC_costimulation, type I INF response, CCR, and parainflammation were significantly different between the lowand high-risk groups (Figure 9(a)). Simultaneously, in view of the increasing importance of immunotherapy based on checkpoint inhibitors in antitumor therapy, we probed the difference in the expression of immune checkpoints between the low-risk and high-risk groups. We found a significant difference in the expression of BTLA, CD48, PDCD-1, etc. between the low-risk and high-risk groups of patients (Figure 9(b)).

Discussion
Many researchers verified that DIAPH3 could regulate growth and migration by affecting cytoskeleton formation in different types of cancer [11,12,18]. And Hager MH et al. identified that DIAPH3 could accelerate the cellular transition to amoeboid tumor phenotype [19]. However, its role and molecular mechanism remain unknown in cervical cancer. We firstly reported that DIAPH3 was overexpressed in cervical cancer and enhanced the ability of CC cell proliferation. Moreover, we found that high expression of DIAPH3 indicated a poor prognosis. e mTOR kinase is a core downstream molecule of PI3K/Akt to modulate cellular growth, metabolism, and migration, and its signaling pathway is frequently dysregulated in a variety of cancers [20][21][22]. Activation of the       mTOR signaling pathway has been observed in cervical cancer [23,24]. e inhibition of mTOR by rapamycin and its analogs seems to be effective in cancer treatment [25]. Importantly, therapy targeting the mTOR signaling pathway indicates clinical benefits in cervical cancer [26]. In our study, we firstly explored the notion that knockdown of DIAPH3 inhibits the proliferation of cervical cancer cells through inactivating the mTOR signaling pathway.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
L. Wan and J. Zhu are contributed equally to this work.