DUSP9 is down regulated in CRC tissues and associated with tumor progression and poor prognosis
To assess the role of DUSP9 in human CRC, we first evaluated the protein expression levels of DUSP9 in 18 paired CRC tissues by western blot. Our results demonstrated that DUSP9 was significantly down regulated in CRC tissues compared with peritumor tissues (Fig. 1A). The results of protein quantification were shown in Fig.1B. In addition, quantitative real-time reverse transcription (qRT-PCR) analysis showed that DUSP9 was down regulated in CRC tissues at mRNA level (Fig. 1C). As shown in Fig. 1D, immunohistochemical staining analysis in CRC tissues from another cohort of 245 patients further confirmed the downregulation of DUSP9 at the protein level in CRC tissues compared with paired peritumor tissues. Finally, we explored the clinical relevance of DUSP9 expression. Kaplan–Meier survival analysis showed that the overall survival (OS) and recurrence-free survival (RFS) of patients with low expression of DUSP9 were significantly shorter than that of patients with high expression of DUSP9 (Fig. 1E, F). Moreover, low DUSP9 expression in CRC was closely associated with tumor size, depth of invasion and advanced TNM stage, indicating that DUSP9 may be involved in the progression of CRC (Table 1). In addition, DUSP9 expression, depth of invasion and TNM stage were found to be associated with overall survival (OS) and disease-free survival (DFS) of these patients in univariate survival analysis. Multivariate analysis indicated that DUSP9 expression could be a prognostic factor for OS and DFS of patients with CRC after adjusting for gender, age at diagnosis, depth of invasion and TNM stage, indicating that DUSP9 may be an independent prognostic factor for CRC (Supplementary Table 1 and 2).
RNA-seq revealed the pathways that DUSP9 inhibited proliferation and metastasis in CRC
To determine the biological function of DUSP9 in CRC, we performed RNA-seq analysis of significantly differently expressed genes (DEGs) between DUSP9 stable knockdown-SW480 cells and shControl cells. Of the 4096 dysregulated genes between two groups, 2113 genes were up-regulated (fold change ≥2) and 1983 genes were down-regulated (fold change ≥2) (Fig. 2A). Part of representative dysregulated genes between SW480 cells with DUSP9 stable knockdown and SW480-shControl cells are listed in Supplemental Table 3. In order to explore the role of DURP9 in tumor progression, we performed KEGG pathway analysis of these DEGs between two groups, and the results showed that many tumor growth and metastasis-related pathways, such as JNK、Erk、Wnt、Akt/mTOR and ErbB signaling pathways were significantly enriched in DUSP9 knockdown group. This suggests that DUSP9 knockdown can activate tumor growth and metastasis-related pathways (Fig. 2B). A hierarchical cluster of DEGs is partially shown in Fig. 2C. As shown in the heatmap, mesenchymal markers (N-cadherin and vimentin) were significantly increase and epithelial markers (E-cadherin and zonula occludens-1) were remarkablely decrease in shDUSP9 SW480 cells than control cells, indicating that DUSP9 knockdown significantly promotes the progression of CRC. Moreover, GSEA (Gene Set Enrichment Analysis) also showed that downregulation of DUSP9 led to the activation of Erk and Akt/mTOR signaling pathway (Fig. 2D).
DUSP9 suppresses tumor growth in vivo
In order to further verify the tumorigenicity of DUSP9 in vivo, we constructed a tumor model of nude mice using SW480 and 293T CRC cell lines with stable overexpression or knockdown of DUSP9. These cells were then transplanted into nude mice to test their tumorigenicity in vivo. Subcutaneous tumor growth was monitored every 3 days, and these mice were euthanized after 30 days. Tumors derived from SW480 cells with stable DUSP9
knockdown showed an increased growth rate and less net weight at the fourth week when compared with control mice (Fig. 3A). In contrast, the tumor growth rate was slower, and the average tumor weight was significantly reduced in mice inoculated with SW480 cells with stable DUSP9 overexpression at the fourth week when compared with control mice (Fig. 3B).
Moreover, when compared with controls, the xenografts developed from SW480 cells with DUSP9 stable knockdown exhibited a significant increase of positive PCNA staining. However, overexpression of DUSP9 exhibited a considerable decrease of positive PCNA staining in xenografts developed from 293T cells (Fig. 3C, D). In conclusion, the above results further confirmed the antitumor effect of DUSP9 on the progression of CRC.
Enforced expression of DUSP9 inhibits proliferation of CRC cells
Next, a variety of in vitro assays were carried out with loss-of-function or gain-of-function of DUSP9 to evaluate the potential role of DUSP9 on CRC cell functions. The results of MTS cell viability assay showed that SW480 cells with DUSP9 stable knockdown had a faster growth rate than control cells, whereas 293T cells with DUSP9 stable overexpression had a slower growth rate than control cells (Fig. 4A). Colony forming assays revealed that down-regulation of DUSP9 significantly promoted cell proliferation of SW480 cells, whereas up-regulation of DUSP9 significantly inhibited cell proliferation of 293T cells (Fig. 4B). Similarly, the number of 5-ethynyl-2′-deoxyuridine (EdU) incorporation was significantly increased in SW480 cells with DUSP9 knockdown, but were markedly decreased in 293T cells with DUSP9 overexpression compared with controls (Fig. 4C). These results demonstrated that DUSP9 functioned as a key regulatory factor in cell proliferation of CRC.
DUSP9 inhibits tumor migration, invasion and metastasis in vitro
We next investigated the effects of DUSP9 on the migration, invasion and metastasis of CRC cells. The scratch wound healing assays showed that knockdown of DUSP9 significantly promoted the migratory ability of SW480 cells. In contrast, DUSP9 overexpression remarkablely inhibited the migratory ability of 293T cells (Fig. 5A). Accordingly, the transwell invasion assay showed that DUSP9 overexpression significantly impaired the invasion of 293T cells. However, knockdown of DUSP9 in SW480 cells acted the opposite way (Fig. 5B). It is well established that the epithelial–mesenchymal transition (EMT) plays a key role in tumor metastasis by increasing cell mobility and reducing cell-cell contact. We further investigated whether EMT is involved in DUSP9 mediated invasion and metastasis of CRC cells. Western blot analysis and quantitative RT-PCR showed that mesenchymal markers (N-cadherin and vimentin) were significantly increase and epithelial markers (E-cadherin and zonula occludens-1) were remarkablely decreased when DUSP9 were knockdown in SW480 cells. However, the opposite effect was observed following DUSP9 overexpression in 293T cells (Fig. 5C, D). These findings indicated that DUSP9 inhibited tumor migration, invasion and metastasis in CRC.
miR-1246 promotes the proliferation and invasion of CRC cells by inhibiting DUSP9
Accumulating evidence has suggested that MicroRNA is critical to the regulation of gene-expression network and is frequently dysregulated in many types of cancers. In this study, MicroRNA Data Integration Portal (mirDIP)-based target prediction programs were used to identify the potential microRNAs involved in the downregulation of DUSP9 in CRC. The top ten predicted miRNAs targeting DUSP9 were listed in Fig. 6A. Real-time PCR and Western blot showed that miR-1246 remarkablely reduced DUSP9 expression in SW480 and 293T cells (Fig. 6B, C). On the contrary, micro-1246 inhibition can elevate the expression of DUSP9 both in mRNA and protein levels (Fig. 6D, E). Moreover, a significant negative correlation (r = −0.642, p < 0.01) between miR-1246 and DUSP9 was found in tumor tissues from 30 CRC patients (Fig. 6F). In addition, miR-1246 mimics attenuated the ability of DUSP9 to inhibit the proliferation of CRC cells, whereas miR-1246 inhibition decreased the proliferation by DUSP9 knockdown in CRC cells (Fig. 6G). Moreover, transwell invasion experiment revealed that miR-1246 mimics attenuated the ability of DUSP9 to inhibit the invasion of CRC cells (Fig. 6H). Altogether, these results suggest that miR-1246 promotes the proliferation and invasion of CRC cells by inhibiting DUSP9.
DUSP9 expression is silenced via promoter hypermethylation in CRC
In order to further explore the reason for the decrease of DUSP9 in CRC, we used MethPrimer software to predict the methylation status in DUSP9 gene promoter. The results showed that DUSP9 was hypermethylated in a variety of cancers, such as colon adenocarcinoma (COAD), bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), lung adenocarcinoma (LUAD) and pancreatic adenocarcinoma etc (Fig. 7A). Based on the criteria and algorithm described by Dahiya, the DUSP9 promoter contains a large CpG island near the transcription start site (Fig. 7B). Of note, 11 CpG sites in DUSP9 promoter were involved in this study for bisulfite sequencing analysis (Fig. 7C). The results showed that in normal intestinal mucosa (n=12), DUSP9 promoter showed hypomethylation status (average methylation level was 15.4%), while in colorectal cancer (n=12), DUSP9 promoter showed hypermethylation status (average methylation level was 87.4%, P < 0.01) (Fig. 7D). To further determine the relationship between the methylation level and expression level of DUSP9 in CRC, we treated SW480 cells with 5-aza-2'-deoxycytidine (5-aza-dC) and examined DUSP9 promoter methylation and protein expression changes. The results showed that with the increase of 5-aza-dC concentration, the methylation level of DUSP9 decreased (Fig. 7E), while the protein expression level increased gradually (Fig. 7F). This suggests that promoter hypermethylation is one of the reasons for low expression of DUSP9 in CRC.