MicroRNA-200b is downregulated and suppresses metastasis by targeting LAMA4 in renal cell carcinoma

Abstract

Background: Metastasis is the primary cause of tumor death in renal cell carcinoma (RCC). Improved diagnostic markers of metastasis are critically needed for RCC. MicoRNAs are demonstrated to be stable and significant biomarkers for several malignancies. In this study, we aimed to explore the metastasis related microRNAs and its mechanism in RCC.

Methods: The relationship between microRNAs expression and prognosis and metastasis of RCC patients were explored by data mining through expression profiles from The Cancer Genome Atlas (TCGA). A total of 80 RCC tissues and adjacent normal kidney tissues were obtained from Department of Urology, Peking University First Hospital. Expression of microRNA-200b (miR-200b) in RCC tissues and cell lines were determined by bioinformatic data mining and quantitative real-time PCR (qRT-PCR). The effects of miR-200b on cell proliferation, migration and invasion were determined by cell counting kit-8 and colony formation assay, wound healing assay and Boyden chamber assay. Mouse cell-derived xenograft and patient-derived xenograft model were also performed to evaluate the effects of miR-200b on tumor growth and metastasis in vivo. The molecular mechanism of miR-200b function was investigated using bioinformatic target predication and high-throughput cDNA sequencing (RNA-seq) and validated by luciferase reporter assay, qRT-PCR, Western blot and immunostaining in vitro and in vivo.

Findings: Our findings indicates that miR-200b is frequently downregulated and have potential utility as a biomarker of metastasis and prognosis in RCC. Interestingly, ectopic expression of miR-200b in the Caki-1 and OSRC-2 cell lines suppresses cell migration and invasion in vitro as well as tumor metastases in vivo. However, miR-200b has no effect on cell proliferation in vitro and tumor growth in vivo. In addition, bioinformatics target predication and RNA-seq results reveals that Laminin subunit alpha 4 (LAMA4) is one target of miR-200b and significantly inhibited by miR-200b in vitro and in vivo.

Interpretation: These results demonstrate a previously undescribed role of miR-200b as a suppressor of tumor metastasis in RCC by directly destabilizing LAMA4 mRNA.

Keywords: LAMA4; Metastasis; MicroRNA-200b; Renal cell carcinoma.

Fig. 1

Downregulation of miR-200b was associated with metastasis and poor prognosis in RCC. (a) Study flow for seeking microRNAs that related with overall survival (OS) and metastasis of RCC patients by in silico analysis using LinkedOmics website. (b) Relative miR-200b expression between M0 and M1 stage from TCGA KIRC dataset. The boxes span the interquartile range. Horizontal lines in the boxes represent median value of miR-200b expression. Whiskers extend to the highest and lowest values. (c) Relative miR-200b expression in normal kidney tissues and ccRCC tissues from TCGA KIRC dataset. RPM, reads per million. (d)Kaplan-Meier curves of recurrence free survival (RFS) time between high and low miR-200b group using Tarone-Ware test in 88 ccRCC patients. Cutoff value was the mean values of miR-200b expression in carcinoma tissues. (e) Relative miR-200b expression in 80 RCC tissues and matched adjacent normal tissues. (f, g) Relative expression data of miR-200b in 80 cases were further analyzed. The expression levels of miR-200b were negatively correlated with T stages (f) and G grades (g). Horizontal lines in (c) and (e–g) represent the mean value of miR-200b expression in TCGA KIRC dataset (c), and mean values of miR-200b expression in 80 RCC patients from PUFH (e-g).

Fig. 2

Ectopic expression of miR-200b impedes cell migration and invasion but not growth inRCC cells in vitro. (a) qRT-PCR analysis of relative miR-200b expression in renal proximal tubule epithelial cells and RCC cell lines. (b) Wound-healing assay. Left panel showed the representative images of wound of Caki-1 and OSRC-2 cells transfected with Control and miR-200b at 0 h and 24 h. Scale bar, 100 μm. Relative cell migratory area (%) of Caki-1 and OSRC-2 cells were statistically analyzed (right panel). (c) Boyden chamber assays were performed with matrigel showed the cell migration and invasion abilities through matrigel of miR-200b over-expressed cells compared with the control cells after 16–18 h' culture. Scale bar, 100 μm. (d) Migration and invasion abilities of 786-O cells were analyzed by Boyden chamber invasive assays between miR-200b TUD cells and control cells. Scale bar, 100 μm. (e) CCK-8 assays were performed to evaluate proliferation ability of miR-200b over-expressed and TUD cells comparing with control cells. (f) Illustration of flow cytometric analysis of Ki-67 staining to evaluate cell proliferation abilitiy between miR-200b over-expressed group and control group. (g) Statistical analysis of positive percentage of Ki-67 staining in miR-200b over-expressed and TUD cells comparing with control cells. (h) Apoptosis detection assay with Annexin V-FITC and PI dual staining. Left panel showed the representative results of apoptosis rate in miR-200b over-expressed and TUD cells comparing with control cells. Apoptosis rate of Caki-1, OSRC-2 and 786-O cells were statistically analyzed (right panel). Data were presented as mean ± SD from at least three independent experiments. *, p < .05. n.s, not significant.

Fig. 3

MiR-200b targets LAMA4 directly. (a) Venn diagram describing the process of predicting target genes of miR-200b that related with metastasis. (b) Sequence alignment of human miR-200b seed sequence with the 3′-UTR of LAMA4. The mutated sequence in the matched binding sites for the gene that was used for creating firefly luciferase reporter constructs is shown in the bottom of the gene set. (c) Luciferase reporter assay demonstrates that miR-200b inhibited the transcription of wild-type, but not the mutant 3′-UTRs of LAMA4. (d, e) The expression of endogenous LAMA4 was inhibited in the pool of miR-200b infected Caki-1 and OSRC-2 cells, compared with the control, at mRNA level as detected by qRT-PCR. The mRNA expression of LAMA4 was normalized to GAPDH mRNA. (f) The expression of LAMA4 but not PCNA in miR-200b infected renal cancer cell lines was inhibited compared with control counterparts at protein level as detected by Western blot. Data were presented as mean ± SD from at least three independent experiments. *, p < .05.

Fig. 4

Expression of LAMA4 increases in RCC patients and miR-200b expression inversely correlates with LAMA4 mRNA levels in RCC tissues. (a) Relative LAMA4 mRNA expression in 80 RCC tissues and matched adjacent normal tissues by qRT-PCR. (b, c) Relative expression data of LAMA4 in 80 cases were further analyzed. The relationship between LAMA4 expression and T stages (b) and G grades (c). (d) Western blot results showed protein expressions of LAMA4 in 12 RCC specimens. (e) The linear regression and correlation between miR-200b-3p and LAMA4 mRNA levels in all 80 RCC tissues. Expression status is shown as the T (tumor)/N (normal) ratio in a log scale. (f) The linear regression and correlation between miR-200b-3p and LAMA4 mRNA levels in TCGA 241 ccRCC tissues. Expression status is shown as log2 (RPM + 1) from RNA-seq results, RPM, reads per million. Horizontal lines in (a–c) represent the mean values of LAMA4 expression relative to endogenous reference GAPDH for each series of samples.

Fig. 5

LAMA4 is a critical role in cell metastasis which induced by silencing miR-200b in cell-derived xenografts (CDXs). (a) Growth curves of mice subcutaneous xenografts derived from Caki-1 Control and Caki-1 miR-200b cells. (b) Image of 10 xenografts formed by Caki-1 Control and Caki-1 miR-200b cells (Left panel). Tumor weights were statistical analyzed (Right panel). (c) OSRC-2 subcutaneous CDXs growth curves. (d) Left panel shown the OSRC-2 subcutaneous CDXs of control and miR-200b group. Statistically analysis of tumor weights of OSRC-2 subcutaneous CDXs (Right panel). (e) Growth curves of 786-O subcutaneous CDXs of control, miR-200b TUD and miR-200b TUD + shLAMA4 group. (f) The image of 786-O subcutaneous CDXs (Left panel). Statistically analysis of tumor weights of 786-O subcutaneous CDXs (Right panel). (g) Left panel shown the mice lung metastatic lesions derived from three groups: Control cells, miR-200b TUD cells and miR-200b TUD + shLAMA4 cells by tail vein injection. Black circles present the position of metastatic nodules. The middle panel is to enlarge the metastatic nodules clearly (arrows). The right panel shown the morphology of lung metastasis by HE staining. Scale bar, 100 μm. (h) QRT-PCR determined relative Alu-sequence expression of mice lung metastases from 786-O CDXs formed by 786-O control, 786-O miR-200b TUD and 786-O miR-200b TUD + shLAMA4 cells, respectively. *, p < .05. n.s, not significant.

Fig. 6

Extrinsic injection of miR-200b mimics in RCC PDXs decreases the expression of LAMA4 and prolongs animal survival. (a) qRT-PCR analysis of miR-200b expression in Caki-1 CDX tumor tissues one week after intratumoral injection of miR-200b mimics or control RNA. (b) Growth curves of Caki-1 CDXs formed by Caki-1 cells injected with miR-200b mimics (n = 6) or control RNA (n = 6). (c) Representative picture of IHC staining of PCNA positive cells of tissues from Caki-1 CDXs. Scale bar, 200 μm. (d) Relative Alu-sequence expression of mice lung metastases from Caki-1 CDXs formed by Caki-1 cells injected with miR-200b mimics or control RNA by qRT-PCR. (e) Representative micrographs and quantitative data of metastatic colonies in mice lung. DiI-positive lung metastatic colonies were photographed and counted under a laser confocal microscope. (f) The hematoxylin and eosin (HE) staining of PDX tissues in nude mice was similar to that of the tumor tissues of the patient with RCC. (g) qRT-PCR analysis of miR-200b expression in PDX tumor tissues one week after intratumoral injection of control RNA, PBS and miR-200b mimics. (h) Growth curves of PDX injected with control RNA (n = 6), PBS (n = 6) and miR-200b-3p mimics (n = 6). (i) Relative Alu-sequence expression of mice lung metastases from PDXs injected with control RNA, PBS and miR-200b mimics by qRT-PCR. (j) Kaplan-Meier curves of overall survival time between control RNA, PBS and miR-200b group using Tarone-Ware test. (k) Representative pictures of IHC staining of PCNA and LAMA4 positive cells of tissues from PDXs. Scale bar, 100 μm. Data were presented as mean ± SD from at least three independent experiments. *, p < .05.

Fig. 7

Effects of LAMA4 on ECM related genes expression and the mechanism of miR-200b-LAMA4 axis on metastasis in RCC. (a) Analysis of the correlation between LAMA4 and numbers of integrin family genes expression in TCGA ccRCC dataset. (b) The linear regression and correlation between LAMA4 and ITGB1, ITGA5, ITGA1 or ITGA9 mRNA levels in TCGA ccRCC dataset. (c, d) Changes of ITGB1, ITGA5, ITGA1 and ITGA9 expression affected by rLAMA4 in Caki-1 miR-200b cells using qRT-PCR (c) and flow cytometry (d). (e) Using Boyden chamber invasive assays, the migration and invasion abilities of Caki-1 miR-200b cells treated with rLAMA and ATN-161 were shown in Left panel. And the effect of LAMA4 knockdown and/or ATN-161 on migration and invasion abilities of 786-O miR-200b TUD cells measured by Boyden chamber invasive assays (right panel). (f) CCK-8 assays was performed to evaluate the role of rLAMA and ATN-161 in cell proliferation of Caki-1 miR-200b cells (left panel). The right panel shown the cell proliferation abilities of 786-O miR-200b TUD cells treated with LAMA4 knockdown and/or ATN-161. (g) Western blot shown the expression of ITGA5, ITGB1, ILK, FAK and ERK after overexpression of miR-200b and treatment with rLAMA4, ATN-161, Cpd-22 and SCH772984 in Caki-1 cells. (h) Diagram of miR-200b-LAMA4 axis on metastasis in RCC. Step 1. Oncology signals induced downregulation of pri-miR-200b; Step 2. Pre-miR-200b was exported and matured in cytoplasm; Step 3. Loss of miR-200b in RCC contributes to upregulation of LAMA4; Step 4. LAMA4 was secreted into extracellular matrix; Step 5. LAMA4 stimulates expression of integrin α5β1; Step 6. ILK/FAK/ERK pathway was activated. Step 7. ERK induces expression of the oncogenes. Data were presented as mean ± SD from at least three independent experiments. *, p < .05.

Fig. S1

Analyzing the TCGA database of RCC using LinkedOmics website. (a) Volcano plot of association results between microRNAs (n = 713) and overall survival in TCGA RCC dataset. (b) Heat plot of 215 microRNAs which are negatively correlated with overall survival in TCGA RCC dataset. (c) Comparison analysis of relative expression of 6 microRNAs (miR-202, miR-101-2, miR-10a, miR-676, miR-10b and miR-139) in primary carcinoma tissues between non-metastatic RCC (M0) and metastatic RCC (M1) in the TCGA database of RCC. All the boxes span the interquartile range. Horizontal lines in the boxes represent median value of miRNAs' expression. Whiskers extend to the highest and lowest values.

Fig. S2

Analyzing the TCGA database of RCC using UCSC Xena website. Comparison analysis of relative miR-202 (a), miR-101-2 (b), miR-10a (c), miR-676 (d), miR-10b (e) and miR-139 (f) expression between tumor tissues (T) and normal kidney tissues (N) in the TCGA database of RCC. RPM, reads per million.

Fig. S3

Kaplan-Meier curves of 6 microRNAs in TCGA ccRCC dataset. Kaplan-Meier curves of recurrence free survival (RFS) time between high and low miR-202 (a), miR-101-2 (b), miR-10a (c), miR-676 (d), miR-10b (e) and miR-139 (f) group using Tarone-Ware test in TCGA ccRCC dataset. Cutoff value was the mean values of the corresponding microRNA expression in carcinoma tissues.

Fig. S4

Transfection efficiency validation and cell function assays of miR-200b. (a) qRT-PCR validated the over-expression of miR-200b mimics infected Caki-1 and OSRC-2 cells, and inhibition of miR-200b expression in stable 786-O miR-200b TUD cells. (b) Colony formation ability of Caki-1, OSRC-2 and 786-O cells were evaluated by colony formation assays between miR-200b up or down-regulated group and control group. (c) qRT-PCR validated the knock down of miR-200b in OSRC-2 cells by miR-200b inhibitor. (d) Boyden chamber assays with matrigel evaluated cell migration and invasion abilities between miR-200b inhibition group and control group in OSRC-2-miR-200b after 16–18 h' culture. Scale bar, 100 μm. (e) Proliferation ability of renal cancer cells were evaluated by CCK-8 assays between miR-200b inhibitor group and control group in OSRC-2-miR-200b cells. Data were presented as mean ± SD from at least three independent experiments. *, p < .05. n.s, Not significant.

Fig. S5

Relative LAMA4 mRNA expression and clinical significance in TCGA KIRC dataset. (a) Total different genes were found by RNA sequencing assay in Caki-1-miR-200b cells compared with the control cells. (b) Top 10 significant genes of 102 genes related to extracellular matrix (ECM) were shown. (c) Relative LAMA4 mRNA expression in TCGA KIRC dataset. (d) Kaplan-Meier curves of overall survival time between high and low LAMA4 group using Log-Rank test in TCGA KIRC dataset. Cutoff value was the median value of LAMA4 expression in carcinoma tissues. (e) Relative LAMA4 mRNA expression in 16 human cancer types from STRING profiles. All the boxes in (c, e) span the interquartile range. Horizontal lines in the boxes represent median value of LAMA4 expression. Whiskers extend to the highest and lowest values.

Fig. S6

Relative expression of top 10 genes (except of LAMA4) which down regulated significantly by miR-200b in TCGA KIRC dataset. Relative CALM1 (a), EXO1 (b), WFS1 (c), ZEB2 (d), SEPP1 (e), NR4A2 (f), LSM1 (g), LDB2 (h) and ACOX1 (i) mRNA expression in TCGA KIRC dataset. All the boxes span the interquartile range. Horizontal lines in the boxes represent median value of LAMA4 expression. Whiskers extend to the highest and lowest values.

Fig. S7

Kaplan-Meier curves of top 10 genes (except of LAMA4) which down regulated significantly by miR-200b in TCGA KIRC dataset. Kaplan-Meier curves of overall survival time between high and low CALM1 (a), EXO1 (b), WFS1 (c), ZEB2 (d), SEPP1 (e), NR4A2 (f), LSM1 (g), LDB2 (h) and ACOX1 (i) mRNA expression group using Log-Rank test in TCGA KIRC dataset. Cutoff value was the mean value of genes expression in carcinoma tissues.

Fig. S8

Correlation between LAMA4 and other miR-200b family (miR-200a, miR-200c, miR141). The linear regression and correlation between LAMA4 and other miR-200b family (miR-200a (a), miR-200c (b), miR141 (c)) mRNA levels in all 80 RCC tissues. The linear regression and correlation between LAMA4 and other miR-200b family (miR-200a (d), miR-200c (e), miR141 (f)) mRNA levels in TCGA KIRC dataset.

Fig. S9

Validation of LAMA4 knock down stably in 786-O miR-200b TUD cells. qRT-PCR (a) and Western blot (b) assays shown the downregualtion of LAMA4 expression by shLAMA4.