Impaired microRNA processing by DICER1 downregulation endows thyroid cancer with increased aggressiveness

Abstract

The global downregulation of microRNAs (miRNAs) is emerging as a common hallmark of cancer. However, the mechanisms underlying this phenomenon are not well known. We identified that the oncogenic miR-146b-5p attenuates miRNA biosynthesis by targeting DICER1 and reducing its expression. DICER1 overexpression inhibited all the miR-146b-induced aggressive phenotypes in thyroid cells. Systemic injection of an anti-miR-146b in mice with orthotopic thyroid tumors suppressed tumor growth and recovered DICER1 levels. Notably, DICER1 downregulation promoted proliferation, migration, invasion, and epithelial-mesenchymal transition through miRNA downregulation. Our analysis of The Cancer Genome Atlas revealed a general decrease in DICER1 expression in thyroid cancer that was associated with a worse clinical outcome. Administration of the small-molecule enoxacin to promote DICER1 complex activity reduced tumor aggressiveness both in vitro and in vivo. Overall, our data confirm DICER1 as a tumor suppressor and show that oncogenic miR-146b contributes to its downregulation. Moreover, our results highlight a potential therapeutic application of RNA-based therapies including miRNA inhibitors and restoration of the biogenesis machinery, which may provide treatments for thyroid and other cancers.

Fig. 1

miR-146b directly targets DICER1, which in turn blocks miR-146b-induced proliferation, migration and invasion. a Table shows the main up- and downregulated miRNAs in thyroid cancer [25] and their predicted binding sites in the DICER 3′UTR (position and the mirSVR score for the miRs predicted by miRanda). Also shown is the fold change (FC) of normal vs PTC and the correlations between DICER1 and miRNAs using Cancer Regulome analysis in TCGA database. b, c Stable cell lines were generated from Nthy-ori cells transfected with a pEGP-Null vector (Null cells) or a pEGP-miR-146b vector (146b cells). b Left: relative DICER1 expression by qPCR. Right: immunoblot of DICER1 expression (results are representative of 3 experiments). c Direct targeting of DICER1 3′UTR by miR-146b. Luciferase reporter activity relative to Renilla level was evaluated in cells 72 h after transfection of pIS1 DICER1 long UTR (WT) or DICER1 3′UTR mutated in the miR-146b binding site (MUT). d Representative images of crystal violet-stained cells 48 h after transfection with the DICER1 expression vector. e Representative images of a wound healing assay 0 and 48 h after scratching. f Relative quantification of the invasive capacity of cells was analyzed using Matrigel-coated Transwell assays. Left: representative images of the lower chamber (invading cells). Right: cell invasion relative to that of “Null” cells. Values represent mean ± SD (n = 3). **p < 0.01; ***p < 0.001; n.s. non-significant.

Fig. 2

miR-146b inhibition impairs established human thyroid tumor growth. a–c Tumor samples taken from mice treated with a control or an miR-146b inhibitor (Anti-146b) administered intratumorally to tumor xenografts generated as previously described [22]. Tumors were analyzed for DICER1 expression by a qPCR, b immunoblotting, and c immunohistochemistry. Actin was used as a loading control. d Cal62-luc cells were injected into the right thyroid lobe and an orthotopic thyroid tumor was generated. Then, a synthetic miR-146b inhibitor (Anti-146b) or a negative control was administered systemically via the retro-orbital vein. Left: endpoint (day 21) bioluminescent signal of the treated tumors. Right: tumor radiance quantification at the indicated time points in mice from treatment onset with the miRNA inhibitor (blue) or the negative control (green). e Representative immunohistochemistry with an anti-DICER1 antibody in orthotopic tumors. Values represent mean ± SEM. *p < 0.05

Fig. 3

DICER1 silencing increases proliferation, migration and invasion in vitro. a–f Cal62 and TPC1 cell lines were transfected with an siRNA against DICER1 (siDICER1) or a control siRNA (siControl) and analysis was performed 48 h later. a Relative expression levels of miRNAs. b Immunoblot of DICER1 and proliferating cell nuclear antigen (PCNA). Actin was used as a loading control. c Representative images of crystal violet-stained cells. d BrdU incorporation relative to siControl-transfected cells. e Representative images from a wound healing assay at 0 and 16 h (Cal62, left) or 0 and 48 h (TPC1, right) after scratching. f Quantification of the invasion rates. Left: representative images of the lower chamber (invading cells). Right: cell invasion rates relative to siControl cells. g, h Cells were transfected with an siRNA against DICER1 (siDICER1) or a control siRNA (siControl). g Relative mRNA levels assayed by qRT-PCR of EMT genes CDH1, EYA1, EYA2, FN, PAI1, SNAIL1, TWIST1, and ZEB1 72 h after siRNA transfection with RNAiMAX Lipofectamine. h Immunoblot for DICER1 and TWIST1 (left) and DICER1, ZEB1, and fibronectin (right) 48 h after siRNA transfection. Actin was used as a loading control. Values represent mean ± SD (n = 3). *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 4

DICER1 is downregulated in human PTC tumors. a DICER1 mRNA expression levels in the indicated cancer types obtained by Firebrowse analysis of the TCGA database. b Box plot of DICER1 mRNA expression levels in thyroid normal tissue, PTC, and metastases: data were obtained from the TCGA database (normal vs tumor p-value < 0.001) (Number of samples: 59 control, 501 primary solid tumors and 8 metastasis). c Left: relative DICER1 mRNA levels in 7 PTC patients (contralateral and normal thyroid tissue). Right: total average of DICER1 mRNA relative levels. d, e Correlation between DICER1 mRNA levels in high, intermediate or low risk (p-value < 0.001, adjusted p-value < 0.01) (d), or extrathyroid extension (p-value > 0.001, adjusted p-value = 0.3) (e), by analyzing the cBioPortal, where the TCGA datasets are deposited. Values represent mean ± SEM. **p < 0.01

Fig. 5

Downregulated miR-30a and miR-100 in thyroid cancer block DICER1-silencing-induced effects. Cal62 cells were silenced for DICER1 (siDICER1) and co-transfected with miR-30a or miR-100 plasmids. Control cells were transfected with siControl and the empty DICER1 vector. a Representative images of cells transfected with siControl, siDICER1, siDICER1 + miR-30a, and siDICER + miR-100 at 48 h post-transfection and stained with crystal violet. b BrdU incorporation. c Representative images from a wound healing assay 0, 14, and 24 h after scratching in the above-mentioned conditions. d Quantification of invasion capacity was analyzed in Matrigel-coated Transwells. Left: representative images of the lower chamber (invading cells). Right: cell invasion relative to siControl cells. Values represent mean ± SD (n = 3). *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 6

Enoxacin induces miRNA expression and decreases cell proliferation, migration and invasion in vitro. Cal62, TPC1, and SW1736 cell lines where treated with enoxacin or PBS + 5% DMSO (control) for 5 days. a Relative expression levels of miR-30a-5p, miR-146b-5p, miR-221-3p, miR-100-5p, miR-21-5p, miR-30a-3p, and miR-204-5p in cells. b Immunoblot of proliferating cell nuclear antigen (PCNA) and actin, as a loading control. c Representative images of cells treated as described and stained with crystal violet. d BrdU incorporation of cells treated with enoxacin relative to control-treated cells. e Representative images from a wound healing assay 0 and 14 h after scratching. f Quantification of invasion rates. Left: representative images of the lower chamber (invading cells). Right: cell invasion rates relative to control-treated cells. g Immunoblot for fibronectin, N-cadherin, ZEB1, and TWIST1. Values represent mean ± SD (n = 3). *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 7

Enoxacin impairs established human orthotopic thyroid tumor growth and decreases the expression of EMT-related genes in vivo. a–c Cal62-luc cells were injected into the right thyroid lobe. After the orthotopic model was established, enoxacin or vehicle was administered intraperitoneally. a Left: the image shows the endpoint (day 28) bioluminescent signal of the tumors imaged with the IVIS-Lumina II Imaging System. Right: tumor radiance quantification at the indicated time points in mice from treatment onset with enoxacin (blue) or vehicle (green). b RNA was obtained from the tumors and miRNA expression levels were determined. Shown are the relative intratumoral levels of miR-30a-5p, miR-146b-5p, miR-221-3p, mIR-100-5p, miR-21-5p, miR-30-5p, and miR-204-5p. c Immunoblot of intratumoral expression of fibronectin, N-cadherin, and PCNA in enoxacin- or vehicle-treated mice. Values represent mean ± SEM. *p < 0.05; n.s. non-significant

Fig. 8

Summary of established DICER1-miRNA regulatory network in thyroid cancer. Overexpressed (oncogenic) miRNAs in thyroid cancer target DICER1. miR-146b (bold) directly binds to DICER1 3′UTR; other miRNAs (grey) have a predicted biding site in DICER1 3′UTR. Also, DICER1 upregulates tumor suppressor miRNAs, which in turn inhibit proliferation, migration, invasion, and EMT. This effect can be phenocopied when increasing the DICER1 complex activity with the small-molecule enoxacin. The arrows denote induction and the truncated lines repression