microRNA-210-3p depletion by CRISPR/Cas9 promoted tumorigenesis through revival of TWIST1 in renal cell carcinoma

Abstract

Previous studies showed that five miRNAs (miR-885-5p, miR-1274, miR-210-3p, miR-224 and miR-1290) were upregulated the most in clear cell renal cell carcinoma (ccRCC). Our focus was to understand from a clinical standpoint the functional consequences of upregulating miR-210-3p. Towards this, we utilized the CRISPR/Cas9 gene editing system to deplete miR-210-3p in RCC cell lines (786-o, A498 and Caki2) and characterized the outcomes. We observed that miR-210-3p depletion dramatically increased tumorigenesis, including altering the morphology of A498 and Caki2 cells in a manner characteristic of epithelial-mesenchymal transition (EMT). These results were corroborated by in vivo xenograft studies, which showed enhanced growth of tumors from miR-210-3p-depleted A498 cells. We identified Twist-related protein 1 (TWIST1) as a key target of miR-210-3p. Analysis of the ccRCC patient data in The Cancer Genome Atlas database showed a negative correlation between miR-210-3p and TWIST1 expression. High TWIST1 and low miR-210-3p expression associated with poorer overall and disease-free survival as compared to low TWIST1 and high miR-210-3p expression. These findings suggest that renal cell carcinoma progression is promoted by TWIST1 suppression mediated by miR-210-3p.

Keywords: CRISPR/Cas9; TWIST1; miR-210-3p; microRNA; renal cell carcinoma.

Figure 1. The expression levels of the top 5 upregulated miRNAs in RCC

A. qRT-PCR data showing higher expression levels of miR-885-5p, miR-1274a, miR-210-3p, and miR-224 in RCC tissues compared to adjacent noncancerous tissues (p = 0.0022, p = 0.0048, p = 0.0022 and p = 0.0022, respectively). B. qRT-PCR data showing high expression levels of miR-210-3p in 786-o, A498, and Caki2 RCC cell lines. C. qRT-PCR data showing high expression levels of miR-210-3p in RCC clinical tissues than in normal kidney samples (* P < 0.0001).

Figure 2. Downregulation of miR-210-3p in RCC cell lines using the CRISPR/Cas9 system

A. Design of sgRNAs for miR-210-3p. Two sgRNAs were designed for miR-210-3p depletion by using CRISPR DESIGN (http://crispr.mit.edu/). Sequencing data showed that these sgRNAs were appropriately inserted into the lentiCRISPR v2 vector. B. qRT-PCR data showing that miR-210-3p was significantly depleted in the three RCC cell lines (786-o, A498, Caki2).

Figure 3. Morphological and functional characteristics of miR-210-3p-depleted ccRCC cells

A. Representative images of miR-210-3p-depleted ccRCC cells showing morphological changes in the miR-210-3p-depleted A498 and Caki2 cells compared to the control. B. Matrigel invasion assay performed in A498 and Caki2 cell lines showing accelerated cell invasion in all the miR-210-3p-depleted cells (* P < 0.01). C. Representative images of colonies of A498 and Caki2 control cells and miR-210-3p-depleted cells with (right) or without (left) sunitinib malate.

Figure 4. Characterization of tumor suppressive properties of miR-210-3p

A. Representative images of the mice carrying tumors and the harvested tumors. Tumor volumes were calculated in nude mice 20 days after subcutaneous injection of A498 cells treated with miR-210-3p-depleted cells or with the empty vector (P = 0.0039). B. Patients with low miR-210-3p expression show reduced overall survival in comparison to those with high miR-210-3p expression (P =0.0618). C. Matrigel cell invasion assay performed with the ACHN cell line 72 h after miR-210-3p transfection (* P = 0.0003). D. Representative image of colony formation between miR-control and miR-210-3p transfectants in ACHN and Caki1 cells.

Figure 5. Characterization of TWIST1 as a candidate target of miR-210-3p

A. TWIST1 mRNA and protein expression in miR-210-3p-depleted A498 and Caki2 cell lines. B. Correlation between miR-210-3p and TWIST1 mRNA expression in clinical RCC patients. Correlation between miR-210-3p and TWIST1 mRNA expression in RCC patients with metastasis. Gene expression values were estimated using RSEM.

Figure 6. miR-210-3p directly regulates TWIST1

A. Luciferase reporter assays using two vectors encoding either putative miR-210-3p binding site in the 3′-UTR of TWIST1 (nucleotides 3058-3064) or after deleting the miR-210-3p binding site. Luminescence intensity was measured in the miR-210-3p transfectants in comparison with the miR-control transfectant. Renilla luciferase values were normalized to firefly luciferase values (* P = 0.0495). B. Luciferase reporter assays using the vector encoding TWIST1 3′-UTR in miR-210-3p-depleted A498 and Caki2 cell lines (* P < 0.0001). C. qRT-PCR data showing that the expression levels of RAD52 and EFNA3, known targets of miR-210-3p, were significantly higher in miR-210-3p-depleted A498 and Caki2 cell lines (* P < 0.0001).

Figure 7. Kaplan-Meier survival plots for high and low TWIST1 expression groups in a TCGA cohort

A. Overall survival and B. Disease-free survival periods (left) were significantly reduced in the patients with high TWIST1 expression in comparison with the patients with low expression (P =0.00054 and P = 0.00347 respectively). (A) Cox proportional analysis for the prediction of overall survival or (B, left) disease-free survival. C. GSEA indicating six significantly enriched pathways, of which EMT was the most significant.