Knockdown of RFC4 inhibits the cell proliferation of nasopharyngeal carcinoma in vitro and in vivo

Abstract

Nasopharyngeal carcinoma (NPC) is a malignant tumor that mainly occurs in East and Southeast Asia. Although patients benefit from the main NPC treatments (e.g., radiotherapy and concurrent chemotherapy), persistent and recurrent diseases still occur in some NPC patients. Therefore, investigating the pathogenesis of NPC is of great clinical significance. In the present study, replication factor c subunit 4 (RFC4) is a key potential target involved in NPC progression via bioinformatics analysis. Furthermore, the expression and mechanism of RFC4 in NPC were investigated in vitro and in vivo. Our results revealed that RFC4 was more elevated in NPC tumor tissues than in normal tissues. RFC4 knockdown induced G2/M cell cycle arrest and inhibited NPC cell proliferation in vitro and in vivo. Interestingly, HOXA10 was confirmed as a downstream target of RFC4, and the overexpression of HOXA10 attenuated the silencing of RFC4-induced cell proliferation, colony formation inhibition, and cell cycle arrest. For the first time, this study reveals that RFC4 is required for NPC cell proliferation and may play a pivotal role in NPC tumorigenesis.

Keywords

nasopharyngeal carcinoma / WGCNA / RFC4 / proliferation

1 Introduction

Nasopharyngeal carcinoma (NPC), which is closely associated with Epstein–Barr virus infection, is a kind of epithelial carcinoma that arises from the nasopharyngeal mucosal lining [1,2]. According to the International Agency for Research on Cancer (IARC), a total of 133,000 new NPC cases were diagnosed in 2020, with more than 70% of the patients from East and Southeast Asia [3]. Currently, radiation therapy (mainly for primary NPC), concurrent chemoradiotherapy (mainly for locally advanced NPC), and immune therapy (usually as an adjuvant therapy to concurrent chemoradiotherapy for advanced or recurrent NPC) are the main treatments for NPC [1,4,5]. Although the five-year overall survival rate of NPC has been improved with concurrent chemotherapy and radiotherapy, persistent and recurrent diseases still occur [6]. Thus, understanding the pathogenesis of NPC and discovering potential targets are of great importance for the diagnosis and therapy of NPC.

Replication factor C (RFC) family proteins, which can help load proliferating cell nuclear antigen (PCNA) onto DNA, are involved in DNA replication and repair [7,8]. RFC4, which is an essential component of the RFC machine, acts as a pivot point for PCNA loading [9]. Recently, the expression of RFC4 in some cancers has been detected, revealing that RFC4 is overexpressed in liver cancer [10], colorectal cancer [11], and oral tongue squamous cell carcinoma [12]. In addition, the downregulation of RFC4 inhibited cell proliferation and enhanced chemosensitivity in hepatocellular carcinoma [13]. Wang et al. found that RFC4 could interact with ku70/80, promote non-homologous end joining-mediated DNA repair, and enhance the radioresistance of colorectal cancer [14]. However, the expression and mechanism of RFC4 in NPC remain unknown.

In the present study, RFC4 was identified as a potential target of NPC via integrated bioinformatics analysis, which include weighted gene co-expression network analysis (WGCNA), differential expression analysis, gene enrichment analysis, protein–protein interaction (PPI), survival analysis, and gene alteration analysis. Furthermore, the expression and mechanism of RFC4 in NPC cell proliferation were investigated in vitro and in vivo.

2 Materials and methods

2.1 Data collection and differentially expressed gene (DEG) analysis

Three gene expression data sets of NPC (e.g., GSE12452, GSE63634, and GSE34573) were downloaded from the Gene Expression Omnibus (GEO) database. The LIMMA package in R was used to analyze DEGs with criteria of |fold change| > 2 and P < 0.05. Gene alteration analysis was performed using cBioPortal, which contains 530 cases of head and neck squamous cell carcinoma (HNSC).

2.2 Weighted gene co-expression network analysis

GSE12452 and GSE64634 were merged after removing the batch effects through the COMBAT function provided by SangerBox. Next, the genes with the top 25% variance were selected for WGCNA through the “WGCNA” package of R software (version 4.0.0). The soft thresholding power (β = 12) was selected according to the criterion of the approximate scale-free topology that followed the construction of the co-expression network via “one-step network construction and module detection”.

2.3 KEGG and GO enrichment analysis

Genes in the turquoise module were submitted to the Database for Annotation, Visualization, and Integrated Discovery, and Gene Ontology and KEGG analysis were conducted with a cutoff criterion of P < 0.05.

2.4 Kaplan–Meier plotter analysis

The relationships between gene expression and patient survival were analyzed by Kaplan–Meier plotter with P < 0.05 indicating statistical significance.

2.5 Immunohistochemical (IHC) assay

The use of NPC samples was approved by the Institutional Review Board of Guangxi University (GXU2021-076). All tumor and normal tissues were obtained with written informed consent from patients. Briefly, specimens from human NPC tissue and xenograft mouse models were embedded in paraffin after their fixation with formalin. Subsequently, the paraffin-embedded tissues were made into histologic sections (5 µm) and deparaffinized and rehydrated by xylene and graded alcohol, respectively. Heat-induced epitope retrieval and citrate buffer were used to retrieve the antigen, followed by blocking with 5% BSA and incubating with the indicated antibodies overnight at 4 °C. After 24 h, the primary antibodies were washed away, and the secondary antibodies were added to the sections and incubated for 90 min at 37 °C followed by a reaction with 3, 3′-diaminobenzidine (DAB, Zhongshan Golden Bridge Biotechnology Co., Ltd., Beijing, China) and staining with hematoxylin. After staining, the sections were dehydrated, cleared, mounted, and scored. The IHC assay score was calculated according to the formula, IRS (0–12) = RP (0–4) × SI (0–3), in which RP and SI refer to the percentage of staining-positive cells and staining intensity, respectively.

2.6 Cell culture and transfection assay

CNE2 cells were purchased from Shanghai Aulu Biological Technology Co., Ltd. (Shanghai, China) and cultured with RPMI 1640 medium. HK-1 cells were purchased from Shanghai Shunran Biological Technology Co., Ltd. and cultured with DMEM. Both cell lines were cultured with their respective media supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin at 37 °C with 5% CO₂ in a humidified incubator. Cells within 10 passages after thawing were transfected with siRNAs or plasmids using Lipofectamine 3000 Reagent (Thermo Fisher Scientific, USA) according to its protocols.

2.7 Plate cloning formation assay

Cells transfected with siRNAs or plasmids were digested and then seeded into 6-well plates at 500–800 cells per well. The medium was changed every 2 or 3 days until the cells grow into visible clones. Then, the clones were fixed with methanol and stained with 0.05% crystal violet.

2.8 Cell cycle detection

48 h after their transfection with siRNAs or plasmids, cells were harvested, washed with phosphate-buffered saline, and fixed with 70% cold ethanol at 4 °C for no less than 4 h. Then, the fixed cells were treated with 100 µL of ribonuclease at 37 °C for 30 min and 400 µL of propidium iodide (PI) at 4 °C for 30 min, followed by flow cytometry measurement.

2.9 RNA extraction and real-time qPCR (RT-qPCR)

TRIzol Reagent (Life Technologies, Scotland, UK), chloroform, and isopropanol were used to extract the RNA according to a previously published protocol. Next, RNA (1 µg) was reverse-transcribed by HiScript II Q Select RT SuperMix (Vazyme, Nanjing, China). RT-qPCR was performed using SYBR Select Master Mix (Applied Biosystems, catalog no. 4472908) with the ABI7300 system (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions. The primer sequences are listed in Table S1.

2.10 RNA sequencing (RNA-seq)

RNA was isolated from CNE2 cells transfected with siRFC4 or NC for 48 h. Then, the RNA samples were sent to Biomarker Technologies (Qingdao, China) for RNA-seq and subsequent DEG analysis.

2.11 Western blot

Cells were lysed in RIPA buffer that contained protease inhibitor cocktail (Sigma–Aldrich). Proteins (20 µg) were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis followed by transfering to a polyvinylidene difluoride membrane (Sigma–Aldrich), which was blocked with 5% milk in TBS for 1 h. The blocked membrane was incubated with specific primary antibodies overnight at 4 °C. Subsequently, the membranes were washed with TBST and then incubated with secondary antibodies for 2 h at room temperature. Finally, the membranes were detected by a Luminescent Image Analyzer LSA 4000 (GE, Fairfield, CO, USA). The primary antibodies included those that fight against RFC4 (ABclonal, A5485), CCNB1 (ABclonal, A2056), HOXA10 (Proteintech, 26497-1-AP), CDK1 (Santa Cruz Biotechnology, sc-8395), and β-actin (Santa Cruz Biotechnology, sc-47778).

2.12 Nude mouse xenograft model

The animal studies were performed according to protocols approved by the Animal Ethics Committee of Guangxi University (GXU-2021-077). Briefly, six-week old female BALB/c nude mice were purchased from Charles River (Beijing, China), which were inoculated with 3 × 10⁶ cells through subcutaneous injection. Three weeks later, the mice were euthanized, and their tumors were excised, weighed, and photographed. Tumor growth was monitored every other day when the tumor grew to the proper size, and the tumor volume was measured and calculated according to the following formula: volume = width²× length × 0.5.

2.13 Statistics analysis

All statistical analyses were carried out using GraphPad Prism 7.0. The correlation coefficients (r) between RFC4 and two transcription factors (TFs), namely, HOXA10 and RELB, were measured by the Spearman correlation method. Two-tailed Student’s t-test was applied to compare two independent groups, and one-way ANOVA was used to compare multiple independent samples. P values less than 0.05 were recognized as statistically significant.

3 Results

3.1 RFC4 was identified as a potential key gene in the tumorigenesis of NPC via bioinformatic analysis

The NPC microarray data sets (e.g., GSE12452 and GSE64634) were downloaded from the GEO database and merged after removing the batch effects. The merged data set contained 14 normal and 43 tumor samples with TN and WHO histological diagnosis information retrieved from previously published works (Table S2). Subsequently, genes with the top 25% variance were used to construct the co-expression modules via WGCNA (Fig.1). To construct a scale-free network, 12 was chosen as the soft power (β = 12, scale free R² = 0.85) (Fig.1). Subsequently, 8 modules were identified through one-step construction and module detection (Fig.1). The results of module–trait relationship analysis showed that the turquoise module was most positively correlated with NPC (NPC, r = 0.85, P = 8^e−17; WHO histological diagnosis, r = 0.79, P = 2^e−13) (Fig.1). Therefore, genes in the turquoise module were selected for further investigation. First, gene enrichment analysis was conducted. The results showed that the top 3 enriched items include DNA replication, mitotic nuclear division, and DNA-dependent DNA repair for GO_Biological_process (GO_BP), chromosomal region, condensed chromosome, and chromosome, centromeric region for GO_cell_component (GO_CC), catalytic activity that acts on DNA, ribonucleoprotein complex binding, and single-stranded DNA binding for GO_ molecular_function (GO_MF), and cell cycle, DNA replication, and RNA transport for KEGG (Fig. S1).

To further identify the hub genes in the turquoise module, protein–protein interactions were assessed and visualized by Cytoscape via applying the MCODE plug-in. As a result, four clusters were identified (Fig. S2), among which Cluster 1, which included 39 nodes and 623 edges, was the largest and most dense cluster (Fig.1). A total of 10 genes in the inner smaller circle of Cluster 1 were recognized as key targets for NPC because these genes were the top 10 DEGs with the highest significance (P value). Next, the prognosis of these 10 key genes was analyzed in HNSC from TCGA via KM-plotter. The results displayed that the expression of 7 genes was significantly associated with patient survival. In summary, a high expression of KIF23, RFC4, and CSE1L was remarkably correlated with poor survival, whereas the high expression of ATAD2, BRIP1, FANCI, and GMNN was markedly associated with better survival (Fig.1). Moreover, the gene alterations in these seven genes were tested via cBioPortal, which revealed that RFC4 was the most variable gene with a 20% of gene alteration rate in HNSC patients (Fig.1). Considering the prognostic and gene alteration results, RFC4 was selected as a potential target for the following investigation.

3.2 RFC4 was overexpressed in NPC tumor tissue

Subsequently, the expression of RFC4 in normal and tumor tissues of NPC was assessed by analyzing the data from the GEO database. The results exhibited that the expression of RFC4 in NPC tumor tissues was higher than in normal tissues (Fig.2). To validate the expression of RFC4, seven local NPC specimens, including tumor and paired normal tissues, were obtained, and the RFC4 expression was tested by performing an IHC assay. In agreement with the results of GEO analysis, the IHC assay confirmed that RFC4 was elevated in NPC tumor compared with normal tissue (Fig.2).

3.3 Knockdown of RFC4 inhibited NPC cell proliferation in vitro

The expression and prognostic results of RFC4 implied that RFC4 might be a driver gene in NPC progression. To examine the role of RFC4 in NPC, RNA interference was carried out in CNE2 and HK-1 cell lines, and cell proliferation and clone formation were tested. As shown in Fig.3 and 3B, the downregulation of RFC4 led to the inhibition of cell proliferation and clone formation (Fig.3 and 3B). Moreover, upon RFC4 siRNA treatment, the cell cycle was arrested at G2/M (Fig.3). The expression of key regulators in G2/M, CCNB1, and CDK1 was also suppressed in cells transfected with RFC4 siRNA (Fig.3). To further confirm the oncogenic effects of RFC4 in NPC, we overexpressed RFC4 in HK-1 cells, which have lower RFC4 expression than the CNE2 cell line. Then, cell viability and clone formation were tested. The results displayed that the upregulation of RFC4 promoted cell proliferation and clone formation in HK-1 cells (Fig.3 and 3F). In addition, the elevated expression of RFC4 increased the expression of CCNB1 and CDK1 (Fig.3). These results indicated that RFC4 could promote cell proliferation by regulating the cell cycle in NPC cells.

3.4 RFC4 regulated the expression of HOXA10 in NPC

To investigate the underlying mechanism of RFC4-induced NPC cell proliferation, we extracted RNA from CNE2 cells transfected with RFC4 siRNA or control siRNA (NC) to perform RNA-seq. A total of 557 genes, including 331 upregulated and 226 downregulated genes, were identified as DEGs (Fig.4). The signaling pathway enrichment analysis of these 557 DEGs revealed the significant enrichment of transcriptional misregulation in cancer (Fig.4). Given that the aberrant expression of TFs causes tumorigenesis [15], 13 differentially expressed TFs involved in transcriptional misregulation in cancer were selected for further investigation (Fig.4). Among these TFs, EGR1 [16,17], ETV7 [18–20], HOXA10 [21,22], MYC [23,24], and RELB [25–27] have been involved in NPC development. Thus, these five TFs were chosen for validation via qPCR. As shown in Fig.4, only HOXA10 and RELB changed consistently in CNE2 and HK-1 cells, thereby verifying the results in the RNA-seq (Fig.4). Additionally, the correlation analysis revealed that the mRNA expression of HOXA10 was significantly correlated with that of RFC4 in NPC tumor tissues in the GEO and Gene Expression Profiling Interactive Analysis (GEPIA) data sets (Fig.4). Therefore, HOXA10 was recognized as a potential target of RFC4. To further verify the correlation between RFC4 and HOXA10, western blot assays, which showed that the expression of HOXA10 was restrained in CNE2 and HK-1 cells upon RFC4 siRNA transfection, were conducted (Fig.4). Additionally, we discovered that overexpression of RFC4 elevated the expression of HOXA10 in HK-1 cells (Fig.4). These results demonstrated that RFC4 could regulate the expression of HOXA10 in NPC cells.

3.5 Overexpression of HOXA10 attenuated RFC4 silencing-induced inhibition of NPC cell proliferation

HOXA10 promotes cell proliferation in NPC. In the present study, RFC4 could positively regulate the expression of HOXA10 and promote cell proliferation. Accordingly, we hypothesized that RFC4 promoted cell proliferation by upregulating HOXA10. To test this idea, we overexpressed HOXA10 in CNE2 cells transfected with shRFC4 or control plasmid and performed cell proliferation, clone formation, and cell cycle assays. The results showed that the overexpression of HOXA10 alleviated the inhibition of cell proliferation and clone formation, G2/M cell cycle arrest, and CDK1 and CCNB1 downregulation caused by RFC4 silencing (Fig.5). These results suggested that RFC4 partially regulated NPC proliferation via HOXA10.

3.6 Knockdown of RFC4 restrained NPC tumor growth in vivo

To examine the pro-proliferative effect of RFC4 in vivo, a cell line with stable expression at low levels of RFC4 (CNE2-shRFC4) and its control cell line (CNE2-shNC) were constructed and transplanted into BALB/c nude mice. Obviously, tumor volume and weight in the shRFC4 group were significantly reduced compared with those in the control group (Fig.6–6C). We also performed IHC assays to test the expression of Ki67 (a marker of cell proliferation) in tumors, which showed that Ki67 was downregulated remarkably in the shRFC4 group, indicating that tumor growth was inhibited in the shRFC4 group (Fig.6). Moreover, the expression levels of CDK1, CCNB1, and HOXA10 obviously decreased in the shRFC4 group, which verified the results in vitro (Fig.6). These results implied that the downregulation of RFC4 restrained NPC tumor growth in vivo.

4 Discussion

WGCNA, a tool for revealing the relationship between gene expression and cell phenotype, is frequently used in the discovery of tumor targets, including gastric cancer [28], lung cancer [29], and cervical cancer [30]. In the present study, the turquoise module was identified as the most significantly relevant module for NPC progression via WGCNA (Fig.1–1D). To further identify the key genes correlated with NPC progression in the turquoise module, the MCODE plug-in of Cytoscape was used, thereby leading to the discovery of 10 hub genes in Cluster 1 (Fig.1). Among these 10 hub genes, DTL promoted cancer development via the degradation of the tumor suppressor gene, PDCD4 [31]. Luo et al. reported that miRNA451a inhibited NPC cell proliferation, migration, and invasion through the downregulation of CSE1L [32]. Wang et al. discovered that MAD2L1 was upregulated in NPC and was involved in NPC progression [33]. Our results verified previous reports, which indicated that the method used in this study was feasible and effective.

RFC4, one of the 10 hub genes identified in this study, is upregulated in some types of cancer [10,34–36], whereas its effects and mechanism in tumor progression have rarely been reported. In the present study, the role of RFC4 in NPC progression was investigated for the first time. Findings also suggested that RFC4 was overexpressed in NPC and that up to 20% of HNSC patients harbored RFC4 amplification, indicating that RFC4 is a potential target for NPC (Fig.1 and 2). Interestingly, we discovered that loss of RFC4 resulted in cell proliferation inhibition by inducing cell cycle arrest at G2/M (Fig.3) instead of S phase arrest, as reported previously [11]. These results imply that RFC4 may exert an oncogenic role in NPC, and that its pro-proliferative mechanism in NPC might be different from those in other types of cancer.

HOXA10, a homeobox gene, is a TF engaged in cell differentiation and morphogenesis. The deregulated expression of HOXA10 is identified in various cancers, including NPC [37–40]. HOXA10 overexpression deteriorates tumor progression by promoting cell growth and metastasis in NPC [21,40]. Currently, limited information is known about the regulation of HOXA10 expression in NPC. In this study, HOXA10 was revealed as a potential downstream target of RFC4 via RNA-seq (Fig.4–4D). The expression correlation analysis revealed a significant positive correlation between HOXA10 and RFC4 expression in NPC (Fig.4). Moreover, overexpressing or knocking down RFC4 caused the up- or down-regulation of HOXA10, respectively, in NPC cells in vitro and in vivo (Fig.4, 4G and 6E). These results demonstrated that RFC4 could regulate the expression of HOXA10. Interestingly, we further found that the proliferative inhibitory effects of RFC4 silencing could be attenuated partially via the overexpression of HOXA10 (Fig.5). Our results imply that RFC4 promotes cell proliferation via the induction of HOXA10 in NPC. However, the mechanism by which RFC4 regulates HOXA10 expression was not elucidated in this study and needs to be clarified in future investigations.

Taken together, RFC4 was overexpressed in NPC. Silencing RFC4 suppressed cell proliferation in vitro and in vivo. RFC4 promoted cell proliferation via the induction of HOXA10. Our results indicate that RFC4 is a potential target of NPC.

References

[1] Chen YP, Chan ATC, Le QT, Blanchard P, Sun Y, Ma J. Nasopharyngeal carcinoma. Lancet 2019; 394(10192): 64–80

[2] Wang M, Gu B, Chen X, Wang Y, Li P, Wang K. The function and therapeutic potential of Epstein-Barr virus-encoded microRNAs in cancer. Mol Ther Nucleic Acids 2019; 17: 657–668

[3] Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2021; 71(3): 209–249

[4] Chen YP, Ismaila N, Chua MLK, Colevas AD, Haddad R, Huang SH, Wee JTS, Whitley AC, Yi JL, Yom SS, Chan ATC, Hu CS, Lang JY, Le QT, Lee AWM, Lee N, Lin JC, Ma B, Morgan TJ, Shah J, Sun Y, Ma J. Chemotherapy in combination with radiotherapy for definitive-intent treatment of stage II-IVA nasopharyngeal carcinoma: CSCO and ASCO Guideline. J Clin Oncol 2021; 39(7): 840–859

[5] Guan S, Wei J, Huang L, Wu L. Chemotherapy and chemo-resistance in nasopharyngeal carcinoma. Eur J Med Chem 2020; 207: 112758

[6] Wang S, Claret FX, Wu W. MicroRNAs as therapeutic targets in nasopharyngeal carcinoma. Front Oncol 2019; 9: 756

[7] Bowman GD, O’Donnell M, Kuriyan J. Structural analysis of a eukaryotic sliding DNA clamp-clamp loader complex. Nature 2004; 429(6993): 724–730

[8] Okumura K, Nogami M, Taguchi H, Dean FB, Chen M, Pan ZQ, Hurwitz J, Shiratori A, Murakami Y, Ozawa K, Eki T. Assignment of the 36.5-kDa (RFC5), 37-kDa (RFC4), 38-kDa (RFC3), and 40-kDa (RFC2) subunit genes of human replication factor C to chromosome bands 12q24.2-q24.3, 3q27, 13q12.3-q13, and 7q11.23. Genomics 1995; 25(1): 274–278

[9] Gaubitz C, Liu X, Magrino J, Stone NP, Landeck J, Hedglin M, Kelch BA. Structure of the human clamp loader reveals an autoinhibited conformation of a substrate-bound AAA+ switch. Proc Natl Acad Sci USA 2020; 117(38): 23571–23580

[10] Zhu W, Zhang Q, Liu M, Yan M, Chu X, Li Y. Identification of DNA repair-related genes predicting pathogenesis and prognosis for liver cancer. Cancer Cell Int 2021; 21(1): 81

[11] Xiang J, Fang L, Luo Y, Yang Z, Liao Y, Cui J, Huang M, Yang Z, Huang Y, Fan X, Wang H, Wang L, Peng J, Wang J. Levels of human replication factor C4, a clamp loader, correlate with tumor progression and predict the prognosis for colorectal cancer. J Transl Med 2014; 12(1): 320

[12] Zhang J, Wang L, Xie X. RFC4 promotes the progression and growth of oral tongue squamous cell carcinoma in vivo and vitro. J Clin Lab Anal 2021; 35(5): e23761

[13] Arai M, Kondoh N, Imazeki N, Hada A, Hatsuse K, Matsubara O, Yamamoto M. The knockdown of endogenous replication factor C4 decreases the growth and enhances the chemosensitivity of hepatocellular carcinoma cells. Liver Int 2009; 29(1): 55–62

[14] Wang XC, Yue X, Zhang RX, Liu TY, Pan ZZ, Yang MJ, Lu ZH, Wang ZY, Peng JH, Le LY, Wang GY, Peng QH, Meng Y, Huang W, Liu RY. Genome-wide RNAi screening identifies RFC4 as a factor that mediates radioresistance in colorectal cancer by facilitating nonhomologous end joining repair. Clin Cancer Res 2019; 25(14): 4567–4579

[15] Bushweller JH. Targeting transcription factors in cancer—from undruggable to reality. Nat Rev Cancer 2019; 19(11): 611–624

[16] Xie L, Xu L, He Z, Zhou W, Wang L, Zhang L, Lan K, Ren C, Liu W, Yao K. Identification of differentially expressed genes in nasopharyngeal carcinoma by means of the Atlas human cancer cDNA expression array. J Cancer Res Clin Oncol 2000; 126(7): 400–406

[17] Zhu Q, Zhang Q, Gu M, Zhang K, Xia T, Zhang S, Chen W, Yin H, Yao H, Fan Y, Pan S, Xie H, Liu H, Cheng T, Zhang P, Zhang T, You B, You Y. MIR106A-5p upregulation suppresses autophagy and accelerates malignant phenotype in nasopharyngeal carcinoma. Autophagy 2021; 17(7): 1667–1683

[18] Sang Y, Cheng C, Zeng YX, Kang T. Snail promotes metastasis of nasopharyngeal carcinoma partly by down-regulating TEL2. Cancer Commun (Lond) 2018; 38(1): 58

[19] Shikata M, Ishikawa F, Kanoh J. Tel2 is required for activation of the Mrc1-mediated replication checkpoint. J Biol Chem 2007; 282(8): 5346–5355

[20] Sang Y, Chen MY, Luo D, Zhang RH, Wang L, Li M, Luo R, Qian CN, Shao JY, Zeng YX, Kang T. TEL2 suppresses metastasis by down-regulating SERPINE1 in nasopharyngeal carcinoma. Oncotarget 2015; 6(30): 29240–29253

[21] Li H, You J, Xue H, Tan X, Chao C. CircCTDP1 promotes nasopharyngeal carcinoma progression via a microRNA-320b/HOXA10/TGFβ2 pathway. Int J Mol Med 2020; 45(3): 836–846

[22] Shen ZH, Zhao KM, Du T. HOXA10 promotes nasopharyngeal carcinoma cell proliferation and invasion via inducing the expression of ZIC2. Eur Rev Med Pharmacol Sci 2017; 21(5): 945–952

[23] Huang W, Song W, Jiang Y, Chen L, Lu H. c-Myc-induced circ-NOTCH1 promotes aggressive phenotypes of nasopharyngeal carcinoma cells by regulating the miR-34c-5p/c-Myc axis. Cell Biol Int 2021; 45(7): 1436–1447

[24] Liu J, Wen Y, Liu Z, Liu S, Xu P, Xu Y, Deng S, Hu S, Luo R, Jiang J, Yu G. VPS33B modulates c-Myc/p53/miR-192-3p to target CCNB1 suppressing the growth of non-small cell lung cancer. Mol Ther Nucleic Acids 2021; 23: 324–335

[25] Chung GT, Lou WP, Chow C, To KF, Choy KW, Leung AW, Tong CY, Yuen JW, Ko CW, Yip TT, Busson P, Lo KW. Constitutive activation of distinct NF-κB signals in EBV-associated nasopharyngeal carcinoma. J Pathol 2013; 231(3): 311–322

[26] Ma F, Zhi C, Wang M, Li T, Khan SA, Ma Z, Jing Z, Bo C, Zhou Q, Xia S, Huang S, Huang S, Zhang Z, Jia H, Cui X, Yao M, Ji T. Dysregulated NF-κB signal promotes the hub gene PCLAF expression to facilitate nasopharyngeal carcinoma proliferation and metastasis. Biomed Pharmacother 2020; 125: 109905

[27] Huang YX, Chen XT, Guo KY, Li YH, Wu BY, Song CY, He YJ. Sunitinib induces NK-κB-dependent NKG2D ligand expression in nasopharyngeal carcinoma and hepatoma cells. J Immunother 2017; 40(5): 164–174

[28] Ren ZH, Shang GP, Wu K, Hu CY, Ji T. WGCNA co-expression network analysis reveals ILF3-AS1 functions as a ceRNA to regulate PTBP1 expression by sponging miR-29a in gastric cancer. Front Genet 2020; 11: e.g.39

[29] Tian F, Zhao J, Fan X, Kang Z. Weighted gene co-expression network analysis in identification of metastasis-related genes of lung squamous cell carcinoma based on the Cancer Genome Atlas database. J Thorac Dis 2017; 9(1): 42–53

[30] Liu J, Li S, Lin L, Jiang Y, Wan Y, Zhou S, Cheng W. Co-expression network analysis identified atypical chemokine receptor 1 (ACKR1) association with lymph node metastasis and prognosis in cervical cancer. Cancer Biomark 2020; 27(2): 213–223

[31] Cui H, Wang Q, Lei Z, Feng M, Zhao Z, Wang Y, Wei G. DTL promotes cancer progression by PDCD4 ubiquitin-dependent degradation. J Exp Clin Cancer Res 2019; 38(1): 350

[32] Luo Y, Qu X, Kan D, Cai B. The microRNA-451a/chromosome segregation 1-like axis suppresses cell proliferation, migration, and invasion and induces apoptosis in nasopharyngeal carcinoma. Bioengineered 2021; 12(1): 6967–6980

[33] Wang B, Wang W, Wang H, Liu W. Microarray analysis of novel genes involved in nasopharyngeal carcinoma. Bull Exp Biol Med 2021; 170(5): 658–664

[34] Ding X, Duan H, Luo H. Identification of core gene expression signature and key pathways in colorectal cancer. Front Genet 2020; 11: 45

[35] Chen W, Zhu S, Zhang Y, Xiao J, Tian D. Identification of key candidate tumor biomarkers in non-small-cell lung cancer by in silico analysis. Oncol Lett 2020; 19(1): 1008–1016

[36] Qiu HZ, Huang J, Xiang CC, Li R, Zuo ED, Zhang Y, Shan L, Cheng X. Screening and discovery of new potential biomarkers and small molecule drugs for cervical cancer: a bioinformatics analysis. Technol Cancer Res Treat 2020; 19: 1533033820980112

[37] Yoshida H, Broaddus R, Cheng W, Xie S, Naora H. Deregulation of the HOXA10 homeobox gene in endometrial carcinoma: role in epithelial-mesenchymal transition. Cancer Res 2006; 66(2): 889–897

[38] Hatanaka Y, de Velasco MA, Oki T, Shimizu N, Nozawa M, Yoshimura K, Yoshikawa K, Nishio K, Uemura H. HOXA10 expression profiling in prostate cancer. Prostate 2019; 79(5): 554–563

[39] Han Y, Lu S, Wen YG, Yu FD, Zhu XW, Qiu GQ, Tang HM, Peng ZH, Zhou CZ. Overexpression of HOXA10 promotes gastric cancer cells proliferation and HOXA10⁺/CD44⁺ is potential prognostic biomarker for gastric cancer. Eur J Cell Biol 2015; 94(12): 642–652

[40] Gong D, Zhu H, Zeng L, Hu R, Hu J, Ding J. Overexpression of HOXA10 promotes the growth and metastasis of nasopharyngeal carcinoma. Exp Biol Med (Maywood) 2021; 246(23): 2454–2462

RIGHTS & PERMISSIONS

Higher Education Press