Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Oct 1.
Published in final edited form as: Mol Cancer Res. 2020 Jan 15;18(4):612–622. doi: 10.1158/1541-7786.MCR-19-0805

AGO2 mediates MYC mRNA stability in hepatocellular carcinoma

Kai Zhang 1,2, Yotsawat Pomyen 3,4, Anna E Barry 1,2, Sean P Martin 3, Subreen Khatib 3, Lucy Knight 3, Marshonna Forgues 3, Dana A Dominguez 3, Ravinder Parhar 3, Ashesh P Shah 1, Adam S Bodzin 1, Xin Wei Wang 3,*, Hien Dang 1,2,*
PMCID: PMC7127970  NIHMSID: NIHMS1549631  PMID: 31941754

Abstract

Deregulated RNA binding proteins (RBPs), such as Argonaute 2 (AGO2) mediate tumor-promoting transcriptomic changes during carcinogenesis, including hepatocellular carcinoma (HCC). While AGO2 is well characterized as a member of the RNA-induced silencing complex (RISC), which represses gene expression through miRNAs, its role as a bona fide RBP remain unclear. In the current study, we investigated AGO2’s role as an RBP that regulates the MYC transcript to promote HCC. Using mRNA and miRNA arrays from HCC patients, we demonstrate that HCCs with elevated AGO2 levels are more likely to have the mRNA transcriptome deregulated and are associated with poor survival. Moreover, AGO2 overexpression stabilizes the MYC transcript independent of miRNAs. These observations provide a novel mechanism of gene regulation by AGO2 and provide further insights into the potential functions of AGO2 as an RBP in addition to RISC.

Keywords: RNA binding protein, AGO2, HCC, liver cancer, MYC

Introduction

Hepatocarcinogenesis is a complex process that requires the selection of oncogenic pathways and deregulation of tumor-suppressive mechanisms. In addition to genetic and epigenetic alterations, transcriptomic changes are often observed during tumorigenesis, all of which are highly regulated by RNA binding proteins (RBPs) (1). RBPs make up 7.5% of all protein-coding genes and regulate the transcriptome through many different mechanisms, including splicing, localization, stability, translation and degradation of RNAs (2). Due to their functional diversity, it is inevitable that deregulated RBPs are selected during hepatocellular carcinoma (HCC) evolution. In fact, genomic profiling studies have revealed that RBPs are altered across many cancer types, including HCC (2–7). Interestingly, the gene expression changes of RBPs in cancers compared to non-tumors are small, suggesting that RBPs are not true cancer drivers, but potential amplifiers or suppressors of oncogenic or tumor-suppressive signaling. Indeed, many studies have demonstrated that various RBPs enhance or suppresses oncogenic or tumor-suppressive signaling to promote tumorigenesis (2, 3, 5, 8).

Argonaute 2 (AGO2), an RBP involved in post-transcriptional regulation, is a key member of the RNA-induced silencing complex (RISC), which directly interacts with miRNAs to repress gene expression. The AGO2-miRNA-RISC complex directly recognizes their mRNA targets through seed complementary at the 3’ untranslated regions (3’UTR) of mRNAs, subsequently triggering repression of its target. This regulatory mechanism is required to maintain transcriptomic homeostasis and often deregulated in cancers. Moreover, the overexpression of AGO2 and its interactions with miRNA has been described to play an important role in promoting hepatocarcinogenesis (9–11). While the AGO2-miRNA interactions are well established as a mechanism of transcriptomic maintenance, recent works demonstrate that AGO2 also interacts with mRNAs in a sequence specific manner, independent of miRNAs (12–14). Most recently, AGO2 has been demonstrated to directly interact with the mRNA ACOT7, repressing its expression with the help of a RNA binding zinc finger protein, WIG1 (12). Together, these works indicate an important AGO2-mRNA regulatory mechanism in which AGO2 can affect gene expression independent of RISC. However, the functional impact of AGO2/mRNA interactions in a RISC-independent manner remain unclear.

HCC is a highly heterogeneous tumor type with poor outcome due to its diverse etiological background and is on the rise in the United States (15–17). Like many other solid tumor types, diverse genomic alterations contribute to HCC progression (3). Using integrative analyses of mRNA and gene copy number alterations from human HCC samples, we have identified AGO2 as an important RBP selected during HCC progression (3). In the current study, we demonstrate that AGO2 enhances HCC progression by stabilizing the proto-oncogene MYC, which has been known to drive tumorigenesis (18). Furthermore, impaired miRNA processing via DICER1 knockout enhanced AGO2 binding to the MYC mRNA transcript and subsequently, increased MYC gene expression. Together, our work reveals a complex post-transcriptional regulation of MYC by AGO2 in HCC.

Materials and Methods

See Supplementary Experimental Material Tables for list of plasmids, siRNAs, antibodies, and Taqman assays used for this study.

Clinical Liver Samples and Array Data

The datasets used are: LCI cohort (N = 488), available on Gene Expression Omnibus (GEO) GSE14520 (http://www.ncbi.nlm.nih.gov/geo). For the TCGA Data, LIHC RNASeqV2 and Copy Number datasets were downloaded using the R (v3.12 package TCGA Assembler, http://www.compgenome.org/TCGA-Assembler, downloaded date: 03–27-2015) (19). TIGER-LC data can be found on GEO (GSE76311). All data were then log2 transformed for analyses.

Statistical analysis

Statistical analysis was performed using GraphPad Prism (v6-v8) and R (v3.30). Enrichment analysis was performed as follows: a specific gene list (observed) is different from a gene list randomly selected from all genes in the analysis (expected) using Chi-square or Fisher’s Exact test followed by Bonferonni correction testing. Two-sided student’s t-tests were performed when there are two groups and one-way ANOVA was performed amongst different groups followed by Tukeys posthoc test to identify differences among groups. Microarray analysis was performed using the paired t-test to look for differentially expressed genes with a p < 0.001 as a cutoff using BRB-ArrayTools (v4.3). Kaplan-Meier curves are calculated using the Cox’s log-rank Mantel. GSEA analysis was performed using GSEA/MSigDB (Broad Institute, http://www.broadinstitute.org/gsea/msigdb/index.jsp). For global mRNA and miRNA expressions between AGO2-Low and AGO2-High samples, patient samples were first divided into tertiles (low, medium, and high) based on AGO2 expression. Second, two-sided Student’s t-test between AGO2-Low (lowest tertile) and AGO2-High (highest tertile) samples were performed on all the genes to identify differentially expressed genes based on AGO2 expression. Third, the list of overlapped genes between differentially expressed genes (DEmRNAs) and all predicted miRNA targets was identified. Finally, density plots of the list of DEmRNAs that are also predicted miRNA genes for AGO2-Low and AGO2-High samples were plotted for both tumor and non-tumor tissues. For global miRNA expression analysis, similar analysis was performed. The list of differentially expressed miRNA (DEmiRNA) based on AGO2-Low and AGO2-High was identified through Student’s t-test, then the density plots of all DEmiRNA between AGO2-Low and AGO2-High sample tertiles were plotted for both tumor and non-tumor tissues. The lists of all miRNAs and their predicted target genes (highly conserved predictions across mammal species) used in these analyses were retrieved from TargetScan version 7.1 (June 2016) (20).

Cell lines, plasmids and chemicals

Huh1 HCC cells were attained at ATCC and MHCC97H cells were a gift from Prof. Zhao-You Tang of the Liver Cancer Institute of Zhongshan Hospital. Mycoplasm testing was performed every 6 months on cultured cells using LookOut Mycoplasm PCR detection kit (Sigma). Cells that are positive are treated with plasmocin (InvivoGen) at 25 ug/ml and retested after two weeks treatment. All cell lines were authenticated using STR DNA profiling analyses (03/2018 and 12/2019). Huh1 cells were cultured in Dulbecco’s modified Eagle Medium (Life Technologies) supplemented with 10% fetal bovine serum (FBS), penicillin, streptomycin and L-glutamine. HCC cells were passed at most 15X and discarded. Newly thawed cells were passed twice before they were used for experiments. MHCC97H cells were cultured in Dulbecco’s modified Eagle Medium (Life Technologies) supplemented with 10% defined FBS, penicillin, streptomycin and L-glutamine. HHT4 cells were cultured on fibronectin (BD Biosciences) coated plates. pLKO.1 shAGO2 and pLKO.1 shCtrl were purchased from OpenBiosystems (TRC lentivirus). pGFP-C-shlentil (shCtrl, shAGO2 1, and shAGO2 2) were purchased from Origene and packaged using HEK293T cells. MOI 5 was used for transduction of 1×105 HCC cells. AGO2 open reading frame (ORF) was cloned into the pLenti-C-mGFP backbone (Origene) and validated using Sanger Sequencing. Lentiviral particles were generated and HCC cells were transduced with the backbone followed by immunoblotting to confirm overexpression.

Generation of DICER1 and AGO2 HCC cell models using CRISPR/Cas9 genome editing.

sgRNAs F’caccgTCACCAATGGGTCCTTTCTT and R’aaacAAGAAAGGACCCATTGGTGAc of DICER1 were cloned into lenti-CRISPRv2 (Addgene, plasmid #52961) and packaged using HEK293T cells using GeneCopia. Huh1 cells were transduced with lenti-CRISPRv2-sgCtrl (empty vector) or lentiCRISPRv2-sgDICER1 virus supplemented with 8 μg/mL polybrene and incubated overnight followed by media change. After 72 hr, cells were selected using 2 μg/mL puromycin, changing the media and adding fresh puromycin every 2–3 days. After cells reached 80% confluence, single cells were passed into 96-well plates, and clonal selection was continued. Single clones were selected, expanded and verified via immunoblotting and Sanger sequencing (See Supplementary Materials and Methods for other related information). sgCtrls were not clonally selected and represents a pooled cell population.

To generate Huh1 and MHCC97H AGO2 knockdown cells, the sgRNA F’caccgTCAAGCCAGAGAAGTGCCCG and R’aaacCGGGCACTTCTCTGGCTTGAc of AGO2 were cloned into the lentiCRISPRv2 plasmid and packaged using HEK293T cells using GeneCopia. Both HCC cells were transduced with lenti-CRISPRv2-sgCtrl (empty vector) and lentiCRISPRv2-sgAGO2 virus supplemented with 8 μg/mL polybrene and incubated overnight followed by media change. Cells were selected with 2 ug/ml puromycin. Validation was performed by immunoblotting for anti-AGO2. Since we did not perform single cell cloning, we considered these cell models AGO2 knockdown systems.

Organotypic culture

AlgiMatrix™ 3D Culture System (6-well plate) (Life Technologies) was used for the organotypic cell culture as developed by our group. 1×105 cells were plated in triplicates in 6-well plates for 24 hr followed by lentiviral transfection with 8 μg/ml polybrene supplementation and incubated overnight. 0.5–1.0×105 cells were seeded to Algimatrix 3D 6-well plates and cultured in the regular cell culture medium for 10 days. Media was changed every 3–4 days. Spheres were collected by dissolving the matrix with 5 ml of Algimatrix Dissolving Buffer (Life Technologies) followed by centrifugation at 300 xg for 5 min. Pellets were then resuspended with 2 ml serum free medium and plated on fibronectin coated 6 well plates and incubated at 37 C for 10 min. Media were aspirated and cells were then fixed using 10% cold formalin followed by ice cold 1X PBS wash, and 2 hour staining with 0.5% crystal violet stain. Colonies were then manually counted using the Bel-Art hand held colony counter from Scienceware.

Western blotting

Protein lysates (20–30μg) were separated on Bis-tris 4–12% SDS-polyacrylamide gels (ThermoScientific) and transferred to a nitrocellulose membrane. Protein detection was performed using anti-AGO2 (Abcam, cat#. ab186733), anti-MYC (Abcam, cat#. 32072), anti-DICER1 (CST, cat#. 5362S), anti-GAPDH (CST, cat#.5174), and anti-β-ACTIN (Sigma-Aldrich, cat#. A5316).

Animal Study

Four-week-old NOD/SCID mice (NOD.CB17-Prkdcscid /NcrCrl) and athymic nude mice were purchased from Charles River Laboratories, Inc. (Wilmington, MA). The animal study protocol was approved by the National Cancer Institute-Bethesda Animal Care and Use Committee. For subcutaneous tumorigenic studies, Huh1 and MHCC97H cells were transfected with pLKO.1 shCtrl or pLKO.1 shAGO2 lentivirus with 8 μg/ml polybrene for 24 hours. Media was changed and 2 μg/ml puromycin was supplemented for selection for 7 days. Cells were then trypan blue counted and 2.5×105 Huh1 or MHCC97H cells were subcutaneously injected into bilateral regions of the athymic nude mice. Tumor volumes were measured weekly until sacrifice at week 12. Tumors were weighed and fixed with 10% formalin and embedded using paraffin.

Histology and immunohistochemistry

The paraffin blocks were sectioned at 5 μm and stained with hematoxylin and eosin. Anti-c-Myc (Santa Cruz, cat. Sc-764) and anti-AGO2 (Abcam, Cat. 186733) and Envision™+ Kit (Dako, Glostrup, Denmark) were used for immunohistochemistry analyses.

Cell proliferation and Migration/Invasion Assays

XCELLigence (ACEA Biosciences) assays were performed for cell proliferation and cell migration and invasion assays. 1×105 HCC cells were plated in triplicates in 6-well plates and incubated overnight. Cells were then transfected with either shCtrl or shAGO2 lentivirus with 8 μg/ml polybrene supplementation and incubated overnight. 3×103 cells were plated onto 8-well E-Plates with appropriate medium supplemented with on with 2 μg/ml of puromycin for Huh1 and MHCC97H cells. For migration/invasion assays, 3×104- 4×104 cells were plated on 8-well CIM Plates (ACEA Biosciences) with or without matrigel in quadruplicates. In addition to XCELLigence, we used CCK8 (Sigma, cat#.96992) to measure cell proliferation. 2.5 × 103 cells were plated in 96-well plates and measured per the manufacturer’s recommendation. All values were subtracted from media only for background and normalized to the 24 hr time point.

Cell colony formation

1×105 HCC cells were plated onto a 6-well plate overnight followed by shAGO2 or shCtrl lentivirus transduction supplemented with 8 μg/ml polybrene. After 24 hr cells were then trypsinized and 5×103 cells were plated in triplicates using 6-well plates with 2 μg/ml puromycin selection and cultured for 10 days. Cells were then washed with ice cold 1X PBS, fixed with ice cold 10% formalin for 30 min at room temperature, washed with dH2O, stained with 0.05% crystal violet for 2 hr, and washed 2X with dH2O. Colonies were then manually counted. For overexpression studies, 5×103 cells were plated in triplicates using 6-well plates with antibiotic selection.

Quantitative RT-PCR

RNA isolation via Trizol was performed and the quality of RNA was assessed using the Nanochip on the Bioanalyzer (Agilent Technologies). cDNA was prepared using the ABI High Capacity Kit with 2 μg of total RNA to generate cDNA via Reverse Transcription Kit (Applied Biosystems), and PCR reactions were carried out with TaqMan Gene Expression assay probes (Applied Biosystems). mRNA (See TaqMan Assays Table in Supplementary Materials and Methods Table). The data were acquired using the ABI SDS 2.4 Software Package (Applied Biosystems) and analyzed using the 2-ΔΔ Ct method. The Ct value of these genes were normalized by subtracting the Ct of the endogenous control 18S mRNA. For miRNAs, total RNA was isolated using miRNeasy Mini Kit (Qiagen, Cat. 217004) per manufacturer’s protocol. 200 ng of RNA was reversed transcribed using ABI High Capacity RT kit (Life Tech) and the corresponding RT primer for each miRNA and SYBR green analysis was performed for each miRNA and U6 was used as an endogenous control. Relative fold change was analyzed using the 2 -ΔΔ Ct method.

mRNA stability assay

1×106 HCC cells were plated onto a 6-well plate overnight followed by shAGO2 or shCtrl lentivirus transduction supplemented with 8 μg/ml polybrene. After 24 hr, cells were then trypsinized and 1.5×105 cells were plated in triplicates using 6-well plates with 2 μg/ml for selection and incubated for a total of 72 hr. At time point 72, cells were washed with 1X PBS, cultured in serum free medium for 2 hr, followed by treatment of Actinomycin D or 5,6-dichloro-1-beta-D-ribofuranosylbenzimidazole (DRB) for 0, 0.5, 1, 1.5, 2 hr in regular media. During each time point, cells were collected by adding 250 μL of Trizol and stored in 4C until RNA isolation.

RNA Immunoprecipitation (RIP)-PCR

HCC cells were plated in 10 cm plates and cultured until 80%−90% confluence. Cells were washed 3X with ice cold 1X PBS. Plates were placed on ice and 5ml of ice cold 1X PBS was added to each plate. Cells were scraped, collected into a 15 ml conical tube and centrifuged for 5 min at 1000 rpm. Cell pellets were resuspended in 1X RIPA lysis buffer and incubated on ice for 20 min. Whole lysates were then centrifuged for 20 min at 12,000 rpm. Supernatant was then transferred to a new microcentrifuged tube. RBP pull down was performed per the manufacturer’s protocol with some changes (MagnaRIP, Millipore (Cat. 1700–700)). Briefly, 2.5 μg of anti-AGO2 and Rabbit IgG antibodies were incubated with 50 μL of Protein G beads for 1 hour at 4C spinning. Beads were then washed and 500 μL of whole cell lysate were then added to the beads with antibodies and incubated overnight at 4C in rotation. 50 μL of the whole lysate was transferred to a new tube and stored at −80 C as input. Before RNA purification, the beads with samples were washed 6X followed by phenol: chloroform extraction and DNAse digestion (Promega, Cat. M6101) per manufacturers protocol. RNA is then reverse transcribed (see RT-PCR) and PCR was performed using primers for Let-7 or MYC using PowerUp SYBR green master mix (Applied Biosystems, Cat. A25778). See Supplementary Materials and Methods for primer sequences.

Results

AGO2 is altered in hepatocellular carcinoma

We have previously demonstrated that RBPs are globally dysregulated in HCC (3). Among the most altered RBPs is Argonaute 2 (AGO2), whose gene expression is significantly elevated in tumors compared to non-tumors in two different HCC cohorts, including the Liver Cancer Institute (LCI) and Cancer Genome Atlas (TCGA) (Supplementary Fig. S1A). Consistent with previous works, AGO2 gene copy number alterations are correlated with increased gene expression in HCC, suggesting that AGO2 is selected during tumor evolution (Supplementary Fig. S1B) (9). To determine whether elevated AGO2 expression is associated with clinical outcome, we performed Kaplan-Meier survival analyses in HCC patients, using the median AGO2 gene expression as a cutoff. We found that HCC patients with elevated AGO2 mRNA levels had a worse prognosis in both the LCI and TCGA cohorts (Fig. 1A). Together, these data suggest that AGO2 is functionally linked to HCC progression.

Figure 1.

Figure 1.

Open in a new tab

The mRNA transcriptome is altered in AGO2-driven hepatocellular carcinoma (HCC). (A) Kaplan-Meier analyses of HCC samples with high AGO2 mRNA expression compared to low AGO2 levels (using the median gene expression as a cutoff). LCI: Liver Cancer Initiative, TCGA-LIHC: the Cancer Genomic Atlas-liver hepatocellular carcinoma. (B) Differentially expressed mRNAs (DEmRNA) that are also a miRNA target according to TargetScan, from the LCI cohort of AGO2-High (High, black) vs. AGO2-Low (Low, red) HCC samples of Tumor (left) and Non-Tumor (right). (C) Differentially expressed miRNAs (DEmiRNAs) from the LCI cohort of AGO2-High (High, black) vs. AGO2-Low (Low, red) HCC samples of Tumor (left) and Non-Tumor (right). Data are z-scored transformed from log 2 values. (D) Bubble plot of the 120 DEmiRNAs between AGO2-High vs. AGO2-Low. Red represents DEmiRNAs negatively correlated (high miRNAs, low mRNAs; vice versa). Blue represents DEmiRNAs positively correlated (high miRNAs, high mRNAs; vice versa). The size of the circle represents the adjusted (Benjamini-Hochberg) p value.

We hypothesized that AGO2 is oncogenic by altering the transcriptome in HCC. Since AGO2 represses mRNAs to regulate gene expression through the RNAi-silencing complex (RISC), we sought to evaluate whether elevated AGO2 levels were more likely to affect the mRNA transcriptome through RISC. We reasoned that if AGO2 acts as an oncogene through the deregulation of RISC, HCC tumors that are driven by AGO2 will have significantly more differentially affected mRNAs that are miRNA targets. Thus, we expect HCC tumors with high AGO2 levels will have lower mRNA levels than AGO2 low tumors. To test this hypothesis, we divided the LCI cohort into two groups: AGO2-High vs. AGO2-Low, using the median gene expression of AGO2 as a cutoff. We next determined differentially expressed mRNAs (DEmRNAs) between AGO2-High and AGO2-Low HCC patients and used TargetScan to determine DEmRNAs that are also predicted miRNA targets (Supplementary Fig. S1C) and plotted their gene expression. In AGO2-High HCCs, the expression of DEmRNAs that is also a miRNA target is higher than AGO2-Low HCCs; whereas, in the non-tumors there were no difference between the two groups (Fig. 1B). This was also observed in the TCGA dataset (Supplementary Fig. S1D). When we plotted the distribution of miRNAs between these two groups, we found the expression (z-score expression) of miRNAs in AGO2-Low HCCs were higher compared to AGO2-High HCCs (Fig. 1C). However, there was no significant difference in miRNA expression between AGO2-High and AGO-Low HCCs in the TCGA dataset (Supplementary Fig. S1E). These analyses suggest AGO2 drives HCC progression predominantly through the alterations of mRNAs.

It is likely that the deregulation of some miRNAs through RISC contributes to the mRNAs that were altered in AGO2-High HCCs. If AGO2 is inducing its tumorigenic activity in a RISC-dependent manner, we would expect the gene expression of DEmRNAs that is also a miRNA target to be negatively correlated with DEmiRNAs expression (high mRNA expression, low miRNA expression). To test this, we investigated the relationship between differentially expressed miRNAs (DEmiRNAs) and its DEmRNA targets in the TCGA dataset using Spearman rank correlation (r>0.2, Benjamini-Hochberg FDR>0.05) and identified 120 DEmiRNAs with at least one DEmRNA target (Supplementary Table S1). Among the 120 DEmiRNAs, 70.8% (85/120) of DEmiRNAs were positively correlated with their DEmRNA targets and 16.7% (20/120) of DEmiRNAs were negatively correlated with their DEmRNA targets. We found 12.5% (15/120) of DEmiRNAs had predicted targets that were both positively and negatively correlated with their mRNA targets (Fig. 1D and Supplementary Table S1). These data suggest that the DEmRNAs altered in AGO2-driven HCCs are not likely regulated by miRNAs through RISC.

AGO2 is important for HCC progression

To test whether AGO2 amplification drives HCC progression, we performed immunoblotting and q-PCR for AGO2 gene copy number in 10 HCC cell lines and HHT4, a telomerase immortalized normal human hepatocyte line that exhibits a near diploid karyotype and expresses many hepatocyte-specific genes (21). We found that most HCC cell lines have two AGO2 gene copies, except for Huh1, MHCC97H and SNU-398, which has more than four copies of AGO2 (Supplementary Fig. S2A–B). Thus, we selected Huh1 and MHCC97H cells to investigate the oncogenic roles of AGO2 in HCC. To address the relevance of AGO2 overexpression in HCC, we knocked down AGO2 in Huh1 and MHCC97H cell lines using two lentiviruses expressing AGO2 shRNA and performed cell proliferation, colony formation, and oncosphere formation assays (Supplementary Fig. 2C and Supplementary Fig. S3). The knockdown of AGO2 decreased cell proliferation as measured over time by xCELLigence and colony formation compared to shCtrl (Supplementary Fig. S3A–B). In addition, we examined the ability of HCC cell self-renewal via the HCC organoid culture model and cell migration and invasion using xCELLigence. HCC cells transduced with AGO2 shRNA lentivirus formed significantly less oncospheres (Supplementary Fig. S3C) and lost their ability to migrate (Supplementary Fig. S3D) (22). However, the knockdown of AGO2 had no effect on cell invasion (Supplementary Fig. S3E). To determine the effect of AGO2 on HCC tumor growth in vivo, we subcutaneously injected Huh1-shCtrl, Huh1-shAGO2, MHCC97H-shCtrl, or MHCC97H-shAGO2 cells into athymic/nude mice and followed tumor growth up to six weeks. The knockdown of AGO2 in HCC cells decreased tumor volume and weight compared to shCtrl (Supplementary Fig. S3F).

AGO2 regulates MYC mRNA stability

To identify potential signaling pathways that are associated with AGO2-driven HCCs, we determined DEGs (differentially expressed genes) between AGO2-High and AGO2-Low HCCs in the LCI and TCGA cohorts. Using the overlapped DEGs (686 genes, p<0.001) from both cohorts, we performed gene set enrichment analysis (GSEA) (Fig. 2A and Supplementary Fig. S4A). Among the top ten altered signaling pathways, we found MYC signaling to be significantly enriched (Fig. 2B). Moreover, we found that the expression of MYC in AGO2-High HCCs were more than two-fold higher than AGO2-Low HCCs in both the LCI (2.5FC, p<0.001) and TCGA (2.4FC) cohort (Supplementary Fig. S4B). Notably, the significant up-regulation of MYC in AGO2-High HCCs were also tumor-specific (Supplementary Fig. S4B). We found that AGO2 gene expression or gene copy number alteration was significantly correlated with MYC gene expression or copy number, suggesting that AGO2 and MYC were selected during hepatocarcinogenesis (Fig. 2C and Supplementary Fig. S4C). This observation is also evident in HCC cell lines, including Huh1 and MHCC97H, where increased MYC gene copy number is significantly correlated with increased AGO2 copy number (Supplementary Fig. S4D).

Figure 2.

Figure 2.

Open in a new tab

AGO2 driven HCCs are associated with MYC signaling. (A) Schematic of overview of gene expression analysis between AGO2-High vs. AGO2-Low in the LCI and TCGA cohort. 686 genes overlap both cohorts and directional (up-regulated of down-regulated in both lists). Hierarchal clustering (Pearson, complete linkage) heatmap of 686 genes. (B) GSEA analysis of 686 AGO2 associated genes showing only the top 10 most enriched genesets. (C) MYC and AGO2 mRNA levels (log 2) are correlated in three different HCC datasets (Spearman rank test).

To understand the relationship between AGO2 and MYC, we measured MYC mRNA expression after AGO2 knockdown in Huh1 and MHCC97H cells. The knockdown of AGO2 significantly reduced MYC mRNA and protein expression compared to shCtrl (Fig. 3A). Moreover, immunohistochemistry (IHC) analyses in mouse tumor tissues for anti-MYC or anti-AGO2 showed loss of AGO2 also decreased MYC protein levels (Fig. 3B). We next evaluated whether AGO2 interacts with MYC mRNA by performing RNA binding protein Immunoprecipitation (RIP) followed by PCR of the MYC mRNA. In both HCC cell lines, we found that MYC was significantly enriched in AGO2 IP, indicating that AGO2 interacts with the MYC mRNA (Fig. 3C). To investigate the effects of AGO2 knockdown on the MYC transcript, we used luciferase-based reporter constructs carrying the MYC transcript 3’UTR. Consistently, we observed decreased MYC luciferase signals after AGO2 knockdown (Fig. 3D). To evaluate how AGO2 regulates the MYC transcript, we inhibited transcription using Actinomycin D or 5,6-Dichlorobenzimidazole (DRB) in HCC cells treated with shCtrl or shAGO2. In HCC cells with AGO2 knockdown, the MYC transcript levels were less stable compared to shCtrl, indicating that AGO2 stabilizes MYC mRNA (Fig. 3E).

Figure 3.

Figure 3.

Open in a new tab

AGO2 regulates MYC mRNA transcript. (A) Relative mRNA expression of MYC mRNA of Huh1 and MHCC97H cells transduced with shCtrl and shAGO2 lentivirus. (**p<0.005, Data are mean ± SEM). (B) Immunohistochemistry analyses of MYC and AGO2 in tumor xenograft tissues. (Images are at 200X magnification and scale bar is 20 μM). (C) RBP-RNA immunoprecipitation analysis. Anti-AGO2 or mouse anti-IgG were used to pull down RNAs cross-linked to AGO2 followed by qRT-PCR of MYC. Fold enrichment of transcript levels first adjusted to 2% input and normalized to IgG (**p<0.005, Data are mean ± SD of triplicates). (D) Relative luciferase (normalized to Renilla) activity of Huh1 and MHCC97H cells transduced with shAGO2 compared to shCtrl (**p<0.005, Data are mean ± SD of triplicates). (E) MYC mRNA transcript levels in shCtrl or shAGO2 after Actinomycin D (ActD) or 5,6-dichloro-1-beta-D-ribofuranosylbenzimidazole (DRB) treatment over specified time (Data are mean ± SEM).

It is possible that MYC is regulated by other mechanisms, including other RBPs, not specific to AGO2 (23–25). Thus, we overexpressed AGO2 in Huh1 and MHCC97H cells that have AGO2 genomically edited by CRISPR/Cas9. Since these cells were not clonally selected for AGO2 knockout, they are considered knockdown models (Supplementary Fig. S5A–B). The overexpression of AGO2 partially rescued MYC expression in both HCC cell lines (Fig. 4A). Additionally, AGO2 overexpression rescued cell proliferation, colony and spheroid formation (Fig. 4B–D and Supplementary Fig. S5C). We also determined MYC mRNA stability in these CRISPR models and found the overexpression of AGO2 increased MYC mRNA stability compared to HCC cells with sgAGO2 (Fig. 4E).

Figure 4.

Figure 4.

Open in a new tab

Overexpression of AGO2 rescues HCC associated phenotypes. (A) Relative mRNA expression of MYC mRNA in Huh1 and MHCC97H cells with sgCtrl, sgAGO2, sgAGO2+AGO2 overexpression (OE) (**p<0.01, Data are mean ± SEM). (B) Cell proliferation rates of measured by CCK8. (Data are mean ± SD). (C) Colony formation in Huh1 and MHCC97H cells of different AGO2 protein level (*p<0.05, Data are mean ± SD of triplicates). (D) Knockdown of AGO2 followed by the overexpression of AGO2 (OE) partially rescues oncosphere formation as measured by Algimatrix 3D oncosphere assay (*p<0.05, Data are mean ± SD triplicates). (E) MYC mRNA transcript levels after Actinomycin D (ActD) treatment over specified time (Data are mean ± SEM).

AGO2 regulates MYC mRNA independent of RISC

Contrary to our observations, the knockdown of AGO2 or miRNA processing has been demonstrated to increase MYC mRNA expression (24, 26, 27). Moreover, it is known that miRNAs, including Let-7, directly regulate MYC post-transcriptionally through a RISC-dependent manner (24, 26, 27). To determine whether the regulation of MYC through AGO2 is independent of miRNAs, we measured MYC expression following siDICER1 treatment. As expected, siRNA mediated inhibition of DICER1 in Huh1 and MHCC97H cells increased MYC mRNA, luciferase signal, and protein expression and had no effect on AGO2 expression (Fig. 5A–C and Supplementary Fig. S5D) (26). We next measured the relative amount of MYC transcripts bound to AGO2 in DICER1 siRNA treated HCC cells (Supplementary Fig. S5E). The inhibition of DICER1 decreased Let-7 miRNA expression and had no effect on AGO2 binding to the MYC transcript in Huh1 cells (Fig. 5D). In MHCC97H cells, DICER1 inhibition enhanced AGO2 binding to MYC transcripts, whereas there was no significant change in Huh1 cells (Fig. 5E). This could be due to the many RBPs that can directly interact with the MYC transcripts which may be more active in Huh1 than MHCC97H cells.

Figure 5.

Figure 5.

Open in a new tab

AGO2 stabilizes MYC transcript independent of miRNAs. (A) siRNA mediated inhibition of DICER1 increases MYC mRNA expression in Huh1 and MHCC97H cells (**p<0.005, Data are mean ± SEM). (B) Relative luciferase (normalized to Renilla) activity of Huh1 and MHCC97H cells treated with siCtrl or siDICER1 (*p<0.01 Data are mean ± SD of triplicates). (C) Representative Immunoblot of Huh1 and MHCC97H treated with siCtrl and siDICER1. DICER1 KO Huh1 and MHCC97H cells. (D) DICER1 inhibition reduces the expression of the miRNA Let-7 in HCC cells (**p<0.005, Data are mean ± SEM). (E) RNA immunoprecipitation of anti-AGO2 and anti-IgG in HCC cells treated with siCtrl or siDICER1 in HCC cells. MYC mRNA was measured using qPCR (**p<0.005, ns: not significant, Data are mean ± SEM).

To investigate the regulation of MYC mRNA stability by AGO2 independent of miRNAs, we employed CRISPR/Cas9 to knockout DICER1 in Huh1 cells (Supplementary Fig. S6A). Consistent with the above data, Huh1 DICER1 knockout (−/−) cells had an increase MYC mRNA and protein expression and significant reduction of Let-7 miRNA expression compared to Huh1.sgCtrl (Ctrl) (Fig. 6A–C). Since MYC mRNA expression is significantly increased following impaired miRNA processing, we first determined whether increased MYC expression is due to impaired miRNA/AGO2 interactions via RNP-IP. In DICER1−/− cells, we observed enhanced AGO2 binding to MYC compared to wildtype, which showed no significant change in our siRNA model (Fig. 6D). These differences may be due to the loss of miRNAs, which causes more free MYC transcripts. We next determined whether shAGO2 affects MYC mRNA stability in Huh1 DICER1−/− cells with Actinomycin D treatment (Fig. 6E). We found shAGO2 decreased AGO2 and MYC expression (Supplementary Fig. S6B). Moreover, the half-life of MYC in DICER1−/− cells increased compared to Ctrl, indicating that DICER1 and its role in miRNA processing is important for the post-transcriptional regulation of MYC transcripts (Fig. 6E). This was most evident within the first hour after ActD treatment. However, after 1 hr, the half-life of the MYC mRNA transcript decreased in DICER1−/− cells treated with shAGO2 compared to shCtrl (Fig. 6E). Although DICER1 is important for MYC mRNA stability, AGO2 remains important for MYC mRNA stability in HCC cells with impaired miRNA processing. Furthermore, AGO2 knockdown in Huh1 DICER1−/− cells decreased MYC mRNA expression, suggesting that without miRNAs, AGO2 predominantly stabilizes the MYC transcript (Fig. 6F). In addition to its role in repressing MYC mRNA through miRNAs/RISC, these data indicate that AGO2 also stabilizes MYC mRNA expression in HCC cells.

Figure 6.

Figure 6.

Open in a new tab

AGO2 stabilizes MYC transcript in Huh1 cells with impaired miRNA processing. (A) Representative immunoblot of Huh1.sgCtrl (Ctrl) and sg.DICER1 (DICER1−/−) cells. (B-C) Huh1 DICER1−/− cells express higher levels of MYC mRNA and lower Let-7 miRNA expression (**p<0.005, Data are mean ± SEM). (D) RNA immunoprecipitation of anti-AGO2 and anti-IgG in wildtype (Ctrl), DICER1−/− HCC cells (**p<0.005, Data are mean ± SEM). (E) MYC mRNA levels in Ctrl and DICER1−/− cells following Actinomycin D (ActD) treatment. (F) Ctrl and DICER1−/− cells were treated with shCtrl or shAGO2 lentivirus and MYC mRNA expression was measured (**p<0.005, Data are mean ± SEM).

Discussion

Global transcriptomic alterations have been observed in cancers, possibly through the deregulation of RBPs (3). Since RBPs regulate a multitude of RNA species in a spatial- or temporal-dependent manner, it is likely that during tumor evolution, deregulated RBPs are selected for cell survival. We have previously identified AGO2 as a potential oncogenic RBP that is selected during HCC evolution using an integrative genomics approach (3, 9). Consistent with previous work, we show that AGO2’s genomic imbalance enhances HCC progression, further supporting its oncogenic role in HCC (9, 28). Interestingly, our work revealed AGO2’s complex role in regulating gene expression. We found that while AGO2 suppresses gene expression through miRNAs, deregulated AGO2 stabilizes MYC stability in hepatocellular carcinoma (HCC). Moreover, we showed that impaired miRNA processing by DICER1 knockout up-regulates MYC, which subsequently increases MYC availability. As a result, free MYC transcripts are more likely to be bound by AGO2 and other RBPs. Since AGO2/RISC maintains the transcriptome homeostasis by mediating mRNA degradation and translational repression, our current work provide a novel mechanism of gene regulation by AGO2 in HCC (2, 29).

MicroRNA-mediated gene repression is often triggered by the interaction between the miRNA-RISC complex and its target’s 3’UTR. These interactions are dependent on a seed sequence found in the miRNA that is complementary to the mRNA target. Interestingly, miRNAs do not have an AGO2 seed sequence and AGO2-mRNA interactions can occur in the absence of miRNAs (12–14). Moreover, recent studies revealed AGO proteins preferentially bind to mRNAs with sequence specificity, independent of the presence of a miRNA sequence, suggesting that AGO2 is an RBP that can interact and regulate mRNAs in a sequence specific manner independent of RISC (14). For example, WIG1, a p53-inducible RNA-binding zinc-finger protein, facilitates AGO2 binding to target mRNAs independent of miRNAs to mediate translational repression (12). While this is an example of how AGO2 regulates mRNA repression independent of miRNAs, whether AGO2 can stabilize mRNAs remains unclear. Consistent with the above observations, our work demonstrates that AGO2 impacts mRNA expression more than miRNAs in HCC, suggesting a potential regulatory role for AGO2 in the mRNA transcriptome. Moreover, this regulatory mechanism may not require miRNAs as there are more mRNAs affected independent of miRNA expression. This regulatory role is evident with the AGO2/MYC interplay, where the downregulation of AGO2 leads to decreased MYC mRNA and protein expression. According to our results, AGO2 stabilizes MYC mRNA in HCC cells by interacting with its 3’UTR region, independent of the presence of mature miRNAs. Although the observation that AGO2 can stabilize mRNAs is limited to the MYC transcript in this study, our work provide evidence that AGO2 is also a generic RBP that can modulate different RNA targets. Future work will focus on identifying and characterizing other potential mRNA transcripts regulated by AGO2.

Post-transcriptional control of MYC is well documented. They include cap-independent translational repression by TIAR and cytoplasmic polyadenylation element–binding protein (CPEB), translational activation by AUF1, translational inhibition by CELF1, and miRNA-mediated transcript repression by ELAV1 recruitment to the Let-7/RISC complex (23, 24, 30, 31).

Adding to these RBPs are other miRNAs, including miR130, and miR24, that have been known to repress MYC transcripts through RISC (27, 32, 33). Notably, DICER1, which is important for the processing of pri-mRNAs to mature miRNAs, indirectly regulates MYC transcripts. This is evident in our study, since the loss of DICER1 increased the half-life of MYC, including in the context of shAGO2 treatment. DICER1 has been demonstrated to be mutated in HCC and acts as a haploinsufficient tumor suppressor in cancers (34, 35). Consistent with our observation, the deletion of one DICER1 gene copy significantly increases MYC and other oncogenes, indicating that DICER1 is important for MYC regulation (35). These data and ours indicate that impaired miRNA processing, through DICER1, is important for oncogenesis.

In diseases, including cancers, the downstream targets of MYC are often deregulated and MYC gene amplification is most commonly observed (36). Since high levels of MYC can induce cellular apoptosis or rapid proliferation, MYC expression must be tightly regulated throughout tumor evolution for cell survival (37). To maintain this homeostasis, cells regulate the levels of MYC through recruitment of different RBPs, including ELAV1, CELF, and AUF1 through a RISC-dependent manner leading to subsequent degradation. However, MYC is also required to promote cellular proliferation, thus AGO2’s ability to stabilize MYC allows for a sustainably active MYC signaling without induction of cell death (38, 39). Thus, in HCC cells, MYC can be regulated by AGO2 through a RISC-dependent and RISC-independent manner simultaneously. Therefore, it is conceivable that the selection of AGO2 in HCC with high MYC may serve as a protective mechanism that allows cells to avoid apoptosis and induce cell proliferation in HCC. This mechanism may account for the 30% of all HCCs with deregulated MYC signaling (3). Together, our work demonstrates that AGO2 also regulates MYC post-transcriptionally as a stabilizer in HCC, providing further evidence that AGO2-mRNA interactions can also co-occur without miRNAs.

Supplementary Material

Supplemental Figures
Supplemental Figure Legends
Supplemental Materials and Methods
Supplemental Table S1

Implications:

Authors demonstrate that the RNA binding protein Argonaute 2 stabilizes the MYC transcript to promote HCC.

Acknowledgements:

This work was supported by grants (Z01 BC 010877 and Z01 BC 010876) from the Intramural Research Program of the Center for Cancer Research, National Cancer Institute, Bethesda, MD and the startup fund from Thomas Jefferson University and Sidney Kimmel Cancer Center, Philadelphia, PA to H.D. We would like to thank Dr. Mu-Shui Dai for sharing the pGL3-MYC-FL plasmid and Miss Emmie Chacker for her assistance in the immunoblots.

Footnotes

Conflict of interest statement: The authors declare no conflict of interests.

REFERENCES

  • 1.Castello A et al. , Insights into RNA biology from an atlas of mammalian mRNA-binding proteins. Cell 149, 1393–1406 (2012). [DOI] [PubMed] [Google Scholar]
  • 2.Pereira B, Billaud M, Almeida R, RNA-Binding Proteins in Cancer: Old Players and New Actors. Trends Cancer 3, 506–528 (2017). [DOI] [PubMed] [Google Scholar]
  • 3.Dang H et al. , Oncogenic Activation of the RNA Binding Protein NELFE and MYC Signaling in Hepatocellular Carcinoma. Cancer Cell 32, 101–114 e108 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Blanco FF et al. , The mRNA-binding protein HuR promotes hypoxia-induced chemoresistance through posttranscriptional regulation of the proto-oncogene PIM1 in pancreatic cancer cells. Oncogene 35, 2529–2541 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Chen L et al. , Recoding RNA editing of AZIN1 predisposes to hepatocellular carcinoma. Nature medicine 19, 209–216 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Amin EM et al. , The RNA-editing enzyme ADAR promotes lung adenocarcinoma migration and invasion by stabilizing FAK. Sci Signal 10, (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Degrauwe N et al. , The RNA Binding Protein IMP2 Preserves Glioblastoma Stem Cells by Preventing let-7 Target Gene Silencing. Cell Rep 15, 1634–1647 (2016). [DOI] [PubMed] [Google Scholar]
  • 8.Han W et al. , RNA-binding protein PCBP2 modulates glioma growth by regulating FHL3. The Journal of clinical investigation 123, 2103–2118 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Cheng N, Li Y, Han ZG, Argonaute2 promotes tumor metastasis by way of up-regulating focal adhesion kinase expression in hepatocellular carcinoma. Hepatology 57, 1906–1918 (2013). [DOI] [PubMed] [Google Scholar]
  • 10.Ye ZL et al. , Argonaute 2 promotes angiogenesis via the PTEN/VEGF signaling pathway in human hepatocellular carcinoma. Acta Pharmacol Sin 36, 1237–1245 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Ye Z, Jin H, Qian Q, Argonaute 2: A Novel Rising Star in Cancer Research. J Cancer 6, 877–882 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Lee HC et al. , WIG1 is crucial for AGO2-mediated ACOT7 mRNA silencing via miRNA-dependent and -independent mechanisms. Nucleic Acids Res 45, 6894–6910 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Li J et al. , Identifying mRNA sequence elements for target recognition by human Argonaute proteins. Genome Res 24, 775–785 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Leung AK et al. , Genome-wide identification of Ago2 binding sites from mouse embryonic stem cells with and without mature microRNAs. Nat Struct Mol Biol 18, 237–244 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Global C Burden of Disease Cancer et al. , Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-years for 32 Cancer Groups, 1990 to 2015: A Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol 3, 524–548 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Weir HK, Thompson TD, Soman A, Moller B, Leadbetter S, The past, present, and future of cancer incidence in the United States: 1975 through 2020. Cancer 121, 1827–1837 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Ferlay J et al. , Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 136, E359–386 (2015). [DOI] [PubMed] [Google Scholar]
  • 18.Dang CV, MYC on the path to cancer. Cell 149, 22–35 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Zhu Y, Qiu P, Ji Y, TCGA-assembler: open-source software for retrieving and processing TCGA data. Nature methods 11, 599–600 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Agarwal V, Bell GW, Nam JW, Bartel DP, Predicting effective microRNA target sites in mammalian mRNAs. Elife 4, (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Jiang W et al. , Cooperation of tumor-derived HBx mutants and p53–249(ser) mutant in regulating cell proliferation, anchorage-independent growth and aneuploidy in a telomerase-immortalized normal human hepatocyte-derived cell line. Int J Cancer 127, 1011–1020 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Takai A et al. , Three-dimensional Organotypic Culture Models of Human Hepatocellular Carcinoma. Sci Rep 6, 21174 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Liao B, Hu Y, Brewer G, Competitive binding of AUF1 and TIAR to MYC mRNA controls its translation. Nat Struct Mol Biol 14, 511–518 (2007). [DOI] [PubMed] [Google Scholar]
  • 24.Kim HH et al. , HuR recruits let-7/RISC to repress c-Myc expression. Genes Dev 23, 1743–1748 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Liu L et al. , Competition between RNA-binding proteins CELF1 and HuR modulates MYC translation and intestinal epithelium renewal. Mol Biol Cell 26, 1797–1810 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Kumar MS, Lu J, Mercer KL, Golub TR, Jacks T, Impaired microRNA processing enhances cellular transformation and tumorigenesis. Nat Genet 39, 673–677 (2007). [DOI] [PubMed] [Google Scholar]
  • 27.Challagundla KB et al. , Ribosomal protein L11 recruits miR-24/miRISC to repress c-Myc expression in response to ribosomal stress. Mol Cell Biol 31, 4007–4021 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Zhang J et al. , MiRNA-99a directly regulates AGO2 through translational repression in hepatocellular carcinoma. Oncogenesis 3, e97 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Gregory RI, Shiekhattar R, MicroRNA biogenesis and cancer. Cancer Res 65, 3509–3512 (2005). [DOI] [PubMed] [Google Scholar]
  • 30.Bernstein PL, Herrick DJ, Prokipcak RD, Ross J, Control of c-myc mRNA half-life in vitro by a protein capable of binding to a coding region stability determinant. Genes Dev 6, 642–654 (1992). [DOI] [PubMed] [Google Scholar]
  • 31.Mazan-Mamczarz K, Lal A, Martindale JL, Kawai T, Gorospe M, Translational repression by RNA-binding protein TIAR. Mol Cell Biol 26, 2716–2727 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Lal A et al. , miR-24 Inhibits cell proliferation by targeting E2F2, MYC, and other cell-cycle genes via binding to “seedless” 3’UTR microRNA recognition elements. Mol Cell 35, 610–625 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Pei ZJ, Zhang ZG, Hu AX, Yang F, Gai Y, miR-122–5p inhibits tumor cell proliferation and induces apoptosis by targeting MYC in gastric cancer cells. Pharmazie 72, 344–347 (2017). [DOI] [PubMed] [Google Scholar]
  • 34.Caruso S et al. , Germline and somatic DICER1 mutations in familial and sporadic liver tumors. J Hepatol 66, 734–742 (2017). [DOI] [PubMed] [Google Scholar]
  • 35.Kumar MS et al. , Dicer1 functions as a haploinsufficient tumor suppressor. Genes Dev 23, 2700–2704 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Kawate S, Fukusato T, Ohwada S, Watanuki A, Morishita Y, Amplification of c-myc in hepatocellular carcinoma: correlation with clinicopathologic features, proliferative activity and p53 overexpression. Oncology 57, 157–163 (1999). [DOI] [PubMed] [Google Scholar]
  • 37.McMahon SB, MYC and the control of apoptosis. Cold Spring Harb Perspect Med 4, a014407 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Vazquez-Chantada M et al. , HuR/methyl-HuR and AUF1 regulate the MAT expressed during liver proliferation, differentiation, and carcinogenesis. Gastroenterology 138, 1943–1953 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Frau M et al. , Role of transcriptional and posttranscriptional regulation of methionine adenosyltransferases in liver cancer progression. Hepatology 56, 165–175 (2012). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Figures
NIHMS1549631-supplement-Supplemental_Figures.pdf (458.3KB, pdf)
Supplemental Figure Legends
NIHMS1549631-supplement-Supplemental_Figure_Legends.docx (26.4KB, docx)
Supplemental Materials and Methods
NIHMS1549631-supplement-Supplemental_Materials_and_Methods.xlsx (19KB, xlsx)
Supplemental Table S1
NIHMS1549631-supplement-Supplemental_Table_S1.xlsx (17.6KB, xlsx)

RESOURCES