The RNA-binding protein KSRP aggravates malignant progression of clear cell renal cell carcinoma through transcriptional inhibition and post-transcriptional destabilization of the NEDD4L ubiquitin ligase

Abstract

Background: KH-type splicing regulatory protein (KHSRP, also called KSRP), a versatile RNA-binding protein, plays a critical role in various physiological and pathological conditions through modulating gene expressions at multiple levels. However, the role of KSRP in clear cell renal cell carcinoma (ccRCC) remains poorly understood.

Methods: KSRP expression was detected by a ccRCC tissue microarray and evaluated by an in silico analysis. Cell loss-of-function and gain-of-function, colony-formation, anoikis, and transwell assays, and an orthotopic bioluminescent xenograft model were conducted to determine the functional role of KRSP in ccRCC progression. Micro (mi)RNA and complementary (c)DNA microarrays were used to identify downstream targets of KSRP. Western blotting, quantitative real-time polymerase chain reaction, and promoter- and 3-untranslated region (3'UTR)-luciferase reporter assays were employed to validate the underlying mechanisms of KSRP which aggravate progression of ccRCC.

Results: Our results showed that dysregulated high levels of KSRP were correlated with advanced clinical stages, larger tumor sizes, recurrence, and poor prognoses of ccRCC. Neural precursor cell-expressed developmentally downregulated 4 like (NEDD4L) was identified as a novel target of KSRP, which can reverse the protumorigenic and prometastatic characteristics as well as epithelial-mesenchymal transition (EMT) promotion by KSRP in vitro and in vivo. Molecular studies revealed that KSRP can decrease NEDD4L messenger (m)RNA stability via inducing mir-629-5p upregulation and directly targeting the AU-rich elements (AREs) of the 3'UTR. Moreover, KSRP was shown to transcriptionally suppress NEDD4L via inducing the transcriptional repressor, Wilm's tumor 1 (WT1). In the clinic, ccRCC samples revealed a positive correlation between KSRP and mesenchymal-related genes, and patients expressing high KSRP and low NEDD4L had the worst prognoses.

Conclusion: The current findings unveil novel mechanisms of KSRP which promote malignant progression of ccRCC through transcriptional inhibition and post-transcriptional destabilization of NEDD4L transcripts. Targeting KSRP and its pathways may be a novel pharmaceutical intervention for ccRCC.

Keywords: EMT; KSRP; NEDD4L; Progression; ccRCC.

Fig. 1

KH-type splicing regulatory protein (KSRP) expression is correlated with a poor prognosis of clear cell renal cell carcinoma (ccRCC) patients and induces clonogenicity, invasion, and anoikis resistance of ccRCC cells. A Upper panel, KSRP expression was analyzed in 72 matched ccRCC tissues and their corresponding normal tissues using data from TCGA. Lower panel, Plot depicting expression levels of KSRP in ccRCC specimens (n = 69) and normal tissue samples (n = 23). The plot was made using GSE15641 from GEO. B Upper panel, IHC staining of KSRP in ccRCC specimens and normal tissue samples. Lower panel, IHC summed scores were applied to determine KSRP expression. Scale bars of black color and red color are 100 and 30 μM, respectively. C Kaplan–Meier analysis of overall survival (OS) and disease-specific survival (DSS) rates in patients with ccRCC presenting with high or low KSRP expression using data from TCGA or E-MTAB-1980. D Knockdown efficiencies of two KSRP shRNAs were determined by Western blotting in Caki-1 and 786-O cells. E–G The effects of KSRP-knockdown on the growth, invasion, and anoikis of Caki-1 and 786-O cells were respectively examined by colony formation (E), Matrigel invasion (F), and anoikis resistance (G) assays. Values are presented as the mean ± SD of three independent experiments. *p < 0.05; **p < 0.01; ***p < 0.001, compared to the control group

Fig. 2

Neural precursor cell-expressed developmentally downregulated 4 like (NEDD4L) is negatively regulated by KH-type splicing regulatory protein (KSRP) and associated with a favorable prognosis in clear cell renal cell carcinoma (ccRCC). A Flowchart for integrating microRNA and cDNA microarrays regulated by KSRP and selection of NEDD4L as a potential KSRP-targeted gene. B, C Western blotting (B) and real-time qPCR (C) analyses of NEDD4L levels in ccRCC cells transfected with a KSRP shRNA, pEGFP-KSRP, or corresponding control vector. Values from (C) are presented as the mean ± SD of three independent experiments. *p < 0.05, compared to the control group. D Correlation analysis of the KIRC database (TCGA, PanCancer Atlas) using cBioPortal revealed negative correlations between mRNA levels of KSRP and NEDD4L. E NEDD4L expression was analyzed in 72 matched ccRCC tissues and their corresponding normal tissues using data from TCGA-KIRC. F NEDD4L expression levels obtained from TCGA-KIRC were compared according to the tumor size (T stages), clinical stages, and metastasis (M stages). Statistical significance was analyzed by a Wilcoxon signed-rank test. G Kaplan–Meier analysis of overall survival (OS) and disease-specific survival (DSS) rates in patients with ccRCC presenting with high or low expression of NEDD4L using data from TCGA and E-MTAB-1980

Fig. 3

KH-type splicing regulatory protein (KSRP) negatively regulates neural precursor cell-expressed developmentally downregulated 4 like (NEDD4L) to trigger the epithelial-mesenchymal transition (EMT) and promotes invasion of clear cell renal cell carcinoma (ccRCC) cells. A NEDD4L was overexpressed (left panel) and knocked-down (right panel) in Caki-1 cells as determined by Western blotting (WB). B Invasive abilities of Caki-1 cells overexpressing (left panel) or knocked-down (right panel) NEDD4L were determined by a Matrigel invasion assay. C, D Expression levels of NEDD4L and KSRP, and the invasive abilities were respectively determined by WB (C) and Matrigel invasion assays (D) in Caki-1 cells expressing shKSRP with or without co-expressing shNEDD4L as indicated. Data from D are shown as a percent of control cells. Values are presented as the mean ± SD of three independent experiments. ***p < 0.001, compared to the control group. ^###p < 0.001, compared to KSRP shRNA-infected cells. E Gene set enrichment analysis of TCGA-KIRC patients with high (top 5%) versus low (bottom 5%) expression of KSRP or NEDD4L. A gene set of the EMT derived from HALLMARK was used. Normalized enrichment scores (NESs) and false discovery rates (FDRs) are shown in the enrichment plot. F Correlations of gene expressions of KSRP or NEDD4L with EMT markers are demonstrated in a correlation plot. RNA sequencing data of TCGA-KIRC patients (n = 510) were utilized to perform this analysis. Correlation coefficients and p values were evaluated by a Pearson correlation analysis. G Caki-1 cells expressing either a KSRP shRNA, NEDD4L-expressing, or respective control vector and subjected to WB to determine expressions of EMT-related regulators. H Dot plots indicating NEDD4L substrates-enriched signaling. I NEDD4L-overexpressing and control Caki-1 cells were treated with or without 20 μM MG-132 for 4 h, and then the Akt, p-Akt, Snail, and GAPDH proteins were analyzed by WB. J All ccRCC patients from TCGA or E-MTAB-1980 were separated into a negative correlation with KSRP and NEDD4L expressions (low KSRP with high NEDD4L and high KSRP with low NEDD4L) and others. Data showed that patients in the KSRP^high/NEDD4L^low group had the worst prognosis

Fig. 4

Oncogenic role of KH-type splicing regulatory protein (KSRP) in vivo. A Dissemination of human clear cell renal cell carcinoma (ccRCC) cells in zebrafish embryos. Caki-1 cells were implanted into 48 h post-fertilization zebrafish embryos. Tumor cell dissemination was detected at day 4 post-injection. White arrows indicate the primary tumor and blue arrowheads indicate disseminated tumor foci. Frequencies of fish showing trunk and end-tail dissemination were monitored and counted every day (right panel). shCtrl: n = 12; shKSRP: n = 14. 4hpi, 4 h post-injection; 4dpi, 4 days post-injection. B Male NOD/SCID mice were orthotopically injected with luciferase-tagged and KSRP-knockdown (shKSRP), neural precursor cell-expressed developmentally downregulated 4 like (NEDD4L)-knockdown (shNEDD4L), or double-knockdown (shKSRP/shNEDD4L) Caki-1 cells. Whole-body bioluminescence imaging was conducted at the indicated time points after injecting cells into mice. C Quantitative analysis of Xenogen imaging signal intensity (photons/s/cm²/sr) every week. *p < 0.05, **p < 0.01, compared to the control group. ^###p < 0.001 compared to the KSRP-knockdown only group. D Gross appearance of orthotopic tumors (indicated by the yellow dashed line) in a kidney. E Upper panel, Representative ex vivo bioluminescence imaging of metastatic sites (pancreas, liver, and lungs) at the end of this in vivo study. Lower panel, Signal intensities of metastatic organs were imaged with bioluminescence at the end of the study, with the mean signal for each group indicated. *p < 0.05, **p < 0.01, ***p < 0.001. F Caki-1 xenografts infected with the indicated shRNAs were isolated to detect expressions of vimentin and E-cadherin by IHC staining. Original magnification, 400 ×

Fig. 5

KH-type splicing regulatory protein (KSRP) negatively regulates neural precursor cell-expressed developmentally downregulated 4 like (NEDD4L) mRNA stability via inducing miR-629-5p upregulation and targeting AU-rich elements (AREs) of the 3' untranslated region (3’UTR). A Decay rate of NEDD4L mRNA after treatment with 5 μg/mL actinomycin D for the indicated times in Caki-1 cells with or without KSRP-knockdown. B Flowchart for identifying miRNA induced by KSRP. The left part shows that TCGA miRNA sequencing data were retrieved from UCSC Xena. TCGA KIRC patients with low expression of miRNA (> 10% patients with reads per million mapped reads (RPM) of < 1) were excluded. The remaining 401 miRNAs were used to perform a differentially expressed (DE)-miRNA analysis between tumor (n = 241) and normal tissues (n = 70). Forty miRNAs were identified as upregulated miRNAs (with fold change (FC) of > 2 and a false discovery rate (FDR) of < 0.01) in KIRC tumor tissues. Right part shows 29 miRNAs with an FC of < 0.5 as identified in KSRP-depleted Caki-1 cells. By intersecting miRNAs upregulated in TCGA KIRC tissues with miRNAs induced by KSRP in ccRCC cells, miR-629-5p was identified as a critical candidate. C Real-time qPCR analysis of miR-629-5p in Caki-1 cells transfected with a KSRP shRNA, pEGFP-KSRP, or their corresponding control vector. D, E Caki-1 cells were transfected with an miR-629-5p mimic (D) or miR-629-5p inhibitor (E) for 24 h. NEDD4L protein levels and invasive ability of cells were respectively determined by Western blot (WB) (left panel) and Matrigel invasion (right panel) assays. F Expressions of KSRP and NEDD4L (left panel) and invasive ability of cells (right panel) were determined in Caki-1 cells transfected with pEGFP-KSRP with or without an miR-629-5p inhibitor. G Correlation between miR-629-5p and NEDD4L expression in TCGA KIRC patients as demonstrated in dot plots. Correlation coefficients and p values were evaluated by a Pearson correlation analysis. H GSEA of TCGA KIRC patients with high (top 5%) versus low (bottom 5%) expressions of miR-629-5p. A gene set of the epithelial-mesenchymal transition (EMT) derived from HALLMARK was used. The normalized enrichment score (NES) and FDR are shown in the enrichment plot. I Correlations of miR-629-5p expression with EMT markers were demonstrated in a correlation plot using miRNA sequencing data of TCGA KIRC patients. Correlation coefficients and p values were evaluated by a Pearson correlation analysis. J WB analysis of E-cadherin, N-cadherin, and vimentin expressions in Caki-1 and 786-O cells transfected with an miR-629-5p mimic. K Left panel, Schematic of the human NEDD4L-3' untranslated region (3'UTR) inserted in the luciferase reporter vector. Positions of AU-rich elements (AREs) are indicated by two gray oblongs, and the corresponding sequences are shown underneath. Right panel, Relative luciferase activities of Caki-1 cells transfected with the NEDD4L luciferase 3’UTR reporter vector containing wild-type or mutant ARE domains with or without KSRP shRNA. Values are presented as the mean ± SD of three independent experiments. **p < 0.01, compared to the respective control groups

Fig. 6

KH-type splicing regulatory protein (KSRP) transcriptionally suppresses neural precursor cell-expressed developmentally downregulated 4 like (NEDD4L) via inducing upregulation of Wilm's tumor 1 (WT1). A Upper panel, schematic representation of the promoter region (1506 bp) of NEDD4L. Lower panel, relative luciferase activities of Caki-1 cells transfected with the NADD4L promoter reporter with or without KSRP shRNA. *** p < 0.001, compared to the control group. B Schematic illustration of truncated NEDD4L promoter constructs (F1–F3). The position of the yellow box represents the fragment location relative to the full-length NEDD4L. C Luciferase reporter activity driven by different NEDD4L promoter segments in Caki-1 cells infected with shKSRP or shCtrl. D Flowchart for identifying possible transcription factors (TFs) involved in KSRP-regulated NEDD4L promoter activity. TF binding to F2 or F3 fragments was predicted by the PROMO and LASAGNA websites. Common TF-binding regions on F2 and F3 were collected, and WT1 was ultimately identified as a candidate TF. E–F WB analysis of KSRP, WT1, and NEDD4L expression levels in whole-cell lysates (E) or cytoplasmic and nuclear fractions (F) from Caki-1 and 786-O cells infected with shKSRP or shCtrl. G-H Luciferase reporter (G) and WB (H) assays were respectively used to detect promoter activities and expression levels of NEDD4L in Caki-1 and 786-O cells infected with shWT1 or shCtrl. I Caki-1 and 786-O cells were infected with shCtrl, shWT1 or shWT1 + shNEDD4L, and the invasive ability of cells were respectively determined by Matrigel invasion assays. J Correlations among KSRP, NEDD4L, and WT1 were analyzed in clear cell renal cell carcinoma (ccRCC) patients from TCGA KIRC and E-MTAB-6692 datasets. Correlation coefficients and p values were evaluated by a Pearson correlation analysis. K Kaplan–Meier analysis of overall survival (OS) rates in patients with ccRCC presenting with high or low expression of WT1 using data from TCGA KIRC dataset

Fig. 7

Schematic presentation depicting the mechanisms of KH-type splicing regulatory protein (KSRP) in promoting the progression of clear cell renal cell carcinoma (ccRCC) via targeting neural precursor cell-expressed developmentally downregulated 4 like (NEDD4L) (Created with BioRender.com)