Human THLE-2 and HepG2 cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, United States; Cat #CRL-2706 and HB-8065, respectively). JHH-4 and Huh7 cell lines were obtained from the Japanese Collection of Research Bioresources (JCRB) Cell Bank (Osaka, Japan; Cat #JCRB0435 and JCRB0403, respectively). THLE-2 cells were maintained in bronchial epithelial cell-growth medium (Lonza, Basel, Switzerland; Cat #CC-3171) with growth supplements according to the ATCC-recommended protocol. HepG2 and Huh7 cells were cultured in low glucose DMEM (Gibco, Detroit, MI, United States), supplemented with 10% fetal bovine serum (FBS) (Cytiva, Marlborough, MA, United States) and 1X Gibco™ Antibiotic-Antimycotic (Gibco, Detroit, MI, United States) in a humidified atmosphere of 5% CO₂ at 37°C. The culture medium and conditions for JHH-4 cells were similar to those of HepG2 and Huh7 cells, except for using Eagle’s minimal essential medium (EMEM) as the basal medium.

To inhibit the function of H19, we employed lentivirus-mediated short-hairpin RNA (shRNA) expression that can be temporally controlled by doxycycline (Dox) treatment [22]. Annealed oligonucleotides of H19-specific and control shRNA sequences were cloned into a lentiviral vector, Tet-LKO-Puro (AddGene Plasmid #21915), and confirmed for proper insertion by Sanger sequencing. Lentivirus was produced in HEK293FT cells and used to establish Dox-inducible shRNA cell lines as previously described [23]. To suppress the activity of microRNA-107 (miR-107), we cloned a sequence of antagomiR against miR-107 into Tet-LKO-Puro and established transgenic HCC lines with Dox-inducible expression of antagomiR. All primers used for cloning can be found in Suppl. Table 1.

A plasmid for overexpression of H19 (pcDNa3.1(+)_A009-H19, AddGene Plasmid #122473) or a negative control plasmid was transfected into HCC cells using Lipofectamine® 3000 reagent (Invitrogen, Carlsbad, CA, United States) according to the manufacturer’s protocol. A total of 500 ng of plasmid was used for transfection per well of a 24-well plate. Transfected cells were employed for subsequent analysis after 48 h posttransfection. For overexpression of miR-107, the pre-miRNA sequence of miR-107 was cloned into a lentiviral vector as previously described [23]. Primers used for cloning can be found in Suppl. Table 1.

Cell proliferation was determined using an MTT-based colorimetric assay (Invitrogen, Carlsbad, CA, United States) according to the manufacturer’s protocol. Briefly, cells were incubated in their basal medium containing 0.5 ug/mL of MTT for 60 min. Then, the insoluble formazan crystals were dissolved in DMSO and quantified by measuring the absorbance at 570 nanometers with a spectrophotometer. Detection of apoptosis was accomplished by the Caspase-Glo® 3/7 assay system (Promega, Madison, WI, United States) and staining with Apopxin Green Indicator (AbCam, Waltham, MA, United States) according to the manufacturers’ protocols. For BrdU incorporation, cells were incubated with culture medium supplemented with 10 µM BrdU (AbCam, Waltham, MA, United States) for 30–60 min before fixation with 4% paraformaldehyde. Cells were permeabilized and denatured with 2 N hydrochloric acid prior to detection by immunofluorescence with an anti-BrdU antibody (1:200 dilution; Cat #sc-32323; Santa Cruz, Dallas, TX, United States), followed by incubation with Alexa Fluor 488-conjugated secondary antibodies (1:500; Cat #A-21206; Invitrogen, Carlsbad, CA, United States) and 300 nM DAPI (Invitrogen, Carlsbad, CA, United States). Cells were visualized by the EVOS M7000 cell imaging system (Invitrogen, Carlsbad, CA, United States), and DAPI⁺ and BrdU⁺ cells were counted by ImageJ software. Quantification of BrdU⁺ cells were performed from 10 random microscope fields, including >2,000 nuclei for each group.

Cell cycle analysis was performed as previously described [24]. Briefly, HepG2 or JHH-4 cells were harvested, washed twice with phosphate-buffered saline (PBS), and then treated with 70% ice-cold ethanol overnight at −20°C. Fixed cells were centrifuged at 600 g for 2 min, washed twice with PBS, and resuspended in a dye mixture containing 100 μg/mL RNase A (Thermo Scientific, Carlsbad, CA, United States; Cat #EN0531) and 50 μg/mL Propidium Iodide (Thermo Scientific, Carlsbad, CA, United States; Cat #P1304MP) in PBS. After overnight staining at 4°C, cells were analyzed for DNA content by the MACSQuant® X Flow Cytometer (Miltenyi Biotec, Gaithersburg, MD, USA). Flow cytometry data were analyzed and processed using FlowJo software (ver. 10.4, BD Biosciences Systems, San Diego, CA, USA) for differentiation of cells into G0/G1, S, and G2/M phases.

RNA isolation, reverse transcription, and qRT-PCR analysis of mRNA and miRNA were performed as previously described [23]. Data were analyzed using the 2^−ΔΔCt method and normalized to expression of RPL19 and U6 for analysis of mRNA and miRNA, respectively. All primers used for qRT-PCR can be found in Suppl. Table 1.

Protein lysates were prepared in RIPA lysis buffer, separated by 12% SDS-PAGE gel, and transferred to a nitrocellulose membrane (Cytiva, Marlborough, MA, United States) using a semi-dry method. Membranes were then probed overnight with the following primary antibodies against CDK6 (1:1,000; Santa Cruz; Cat #sc-7961), Cyclin A2 (CST, Danvers, MA, United States; Cat #4656T), PLK1 (CST; Cat #4513T), GAPDH (1:2,500; Invitrogen; Cat #437000), total Rb (1:500; Santa Cruz; Cat #sc-102), and phosphorylated Rb (1:1,000; CST; Cat #8516T for Ser807/811, #9307T for Ser780, and #9301T for Ser795). HRP-conjugated goat anti-mouse IgG and goat anti-rabbit IgG (1:5,000; CST; Cat #7076S and #7074, respectively) were used as secondary antibodies. Blots were imaged using a UVP ChemStudio instrument (Analytik Jena, Beverly, MA, United States). Uncropped blots were shown in Suppl. Fig. 1.

Putative miRNA targets of H19 were obtained from LncBase (www.microrna.gr/LncBase) and LncBook (https://ngdc.cncb.ac.cn/lncbook) databases [25,26]. The hybridization pattern between miRNA and the 3′ untranslated region (3′ UTR) of CDK6 was predicted by using the DIANA-microT web server (http://www.microrna.gr/webServer) and the RNA hybrid tool (http://bibiserv.techfak.uni-bielefeld.de/rnahybrid/) [27,28]. Wild-type and mutant sequences of H19 and CDK6 3′ UTR were cloned into pmirGLO plasmid, and the miR-107 sequence was cloned into pSilencer plasmid as previously described [23]. Primers used for the cloning can be found in Suppl. Table 1. The interaction between miR-107 and its putative binding sites was tested by the Dual-Luciferase® Reporter Assay System (Promega, Madison, WI, United States) following the manufacturer’s instruction. Briefly, HEK293T cells were seeded in a 96-well plate and transfected with 300 ng of the luciferase reporter constructs (pmirGLO plasmid) and 300 ng of the pSilencer-miR-107 plasmid using Lipofectamine® 3000 Transfection Reagent (Invitrogen, Carlsbad, CA, United States) according to the manufacturer’s protocol. After 48 h, cells were lysed, incubated with firefly and Renilla luciferase substrates, and measured for enzyme activity using a luminometer. Firefly luciferase activity is normalized against Renilla luciferase to represent the miRNA binding efficiency.

Data are presented as the mean ± standard deviation of the mean (SD) and statistically analyzed by the GraphPad Prism software version 9.5.1 (GraphPad) using unpaired Student’s t-test or one-way ANOVA. Three or more replicates from at least two independent experiments were used for the determination of significant differences. A p value of <0.05 was considered statistically significant.

To explore the role of H19 in cell growth, we determined the effect of knockdown of H19 on the proliferation of HCC cells. First, we confirmed the upregulation of H19 in HCC cells in comparison with a normal hepatocyte cell line, consistent with previous clinical data where H19 is found upregulated in HCC tissues (Suppl. Fig. 2A) [10–12]. Then, we established a doxycycline (Dox)-inducible H19 knockdown system in two independent HCC cell lines (HepG2 and JHH-4). Upon Dox treatment, levels of H19 transcripts were significantly downregulated (Suppl. Fig. 2B). To determine cell growth, we performed the MTT assay and found that H19 suppression impaired cell proliferation (Fig. 1A and Suppl. Fig. 3A). Consistent with the results of the MTT assay, we observed a significantly reduced number of BrdU-positive (BrdU⁺) cells upon downregulation of H19 (Fig. 1B). Cell cycle analysis revealed that H19 knockdown significantly decreased the distribution of cells in S and G2/M phases while increasing the number of cells arrested in the G1 phase (Fig. 1C). Moreover, apoptosis detection by caspase-3/7 activation and Apopxin staining was significantly elevated in H19-knockdown cells (Fig. 1D and Suppl. Fig. 3B). Conversely, overexpression of H19 promoted cell growth, BrdU incorporation, and cell cycle progression while reducing apoptosis in HCC cells (Figs. 2A–2E and Suppl. Fig. 3C). Taken together, these results suggest the oncogenic role of H19 in HCC, and its upregulation may be crucial to drive the proliferation and the cell cycle of HCC cells.

As BrdU incorporation and cell cycle progression were impaired in H19-knockdown cells (Figs. 1B and 1C), we hypothesized that suppression of H19 could induce G1 phase arrest in the cell cycle. Thus, we surveyed the expression of key cell cycle genes that promote the S-phase entry by qRT-PCR analysis. We found that upon H19 depletion, levels of CDK6 transcripts were significantly decreased in two independent HCC cell lines (Fig. 3A). Downregulation of CDK6 in H19-knockdown cells was further confirmed at the protein level by Western blot (Fig. 3B). CDK6 is a critical mediator in the cellular transition into the S phase, and its overexpression plays an important role in driving tumorigenesis in many cancer types [29,30]. CDK6 forms a protein complex with CDK4 and D-type cyclins to drive cell proliferation. A well-known substrate of this protein complex is the retinoblastoma protein (Rb), which inhibits cell cycle progression into the S phase until its inactivation by phosphorylation. Upon H19 knockdown expression, we observed a significant reduction in Rb phosphorylation at the Ser780, Ser795, and Ser807/811 residues, supporting the impaired function of CDK6 (Fig. 3C and Suppl. Fig. 4A). Rb phosphorylation weakens its association with the E2F transcription factor, allowing for activation of E2F-mediated transcription and the S-phase entry [31]. Accordingly, we observed downregulation of E2F target genes such as Cyclin A2 and PLK1 in H19-knockdown cells (Suppl. Fig. 4B). Together, these data indicate that H19 depletion could impair the proliferation of HCC cells at least by suppressing the activity of CDK6.

LncRNAs can exert their regulatory function by acting as sponges of microRNAs (miRNAs), which control gene expression post-transcriptionally by binding target mRNAs at the 3’ UTR, resulting in mRNA degradation and/or translational repression [32]. Thus, lncRNAs can inhibit the activity of miRNAs and allow upregulation of miRNA target genes. Given this mode of action, we hypothesized that H19 could sequester a miRNA that negatively regulates the expression of CDK6. To identify such miRNAs, we obtained two lists of miRNAs that were previously reported or predicted to interact with H19 from the lncBase and lncBook databases [25,26]. A total of 54 miRNAs were found to overlap from the two sources (Suppl. Table 2). Then, we employed the DIANA-microT web server to rank the interaction of these miRNAs with the sequence of CDK6 3’UTR and found that miR-107 achieved the highest miRNA target gene (miTG) score (Fig. 4A) [27]. Using the RNA hybrid tool, we analyzed the sequence of H19 and revealed multiple binding sites of miR-107 with minimal free energy (MFE) <−20 kcal/mol (Suppl. Table 3) [28]. This bioinformatic analysis indicated that H19 could serve as a molecular sponge for miR-107 and hence promote the expression of CDK6. Supporting this notion, we observed that levels of miR-107 were slightly increased, although not significantly, in HCC cells, and overexpression of H19 could be a way to counter the antiproliferative activity of miR-107 (Suppl. Fig. 1A and Fig. 4B).

To test whether miR-107 was sponged by H19, we performed a dual luciferase reporter gene assay in HEK293 cells, a common model used for testing miRNA-target interactions [33,34]. Results showed that miR-107 significantly reduced the luciferase activity of the H19 wild-type (WT) reporter, while this inhibition was alleviated with the H19 mutant (MUT) sequence (Figs. 4C and 4D). To test if CDK6 is a direct target of miR-107, we performed another set of dual luciferase reporter gene assay and demonstrated that miR-107 directly interacted with the 3′ UTR of CDK6 mRNA (Figs. 4E and 4F). Collectively, we illustrated that H19 could sequester miR-107 to promote CDK6 expression.

To confirm that miR-107 is crucial in the regulation of the H19/miR-107/CDK6 axis, we introduced control antagomir (anta-miR-con) or miR-107 antagomir (anta-miR-107) into H19-knockdown HCC cells. We first showed that anta-miR-107 efficiently blocked the activity of miR-107 (Fig. 5A). Then, we tested the proliferation of HCC cells with antagomirs by the MTT assay, and the results suggest that inhibition of miR-107 partially attenuated the antiproliferative effect mediated by H19 depletion (Fig. 5B). Consistently, expression of CDK6 was significantly upregulated at both mRNA and protein levels upon miR-107 inhibition in H19-knockdown cells (Figs. 5C and 5D). To confirm the antiproliferative role of miR-107, we overexpressed miR-107 in HepG2 and JHH-4 cells and observed growth suppression (Figs. 5E and 5F). Taken together, these results indicate that H19 could play an important role in sequestering miR-107 and promoting CDK6 expression, thus driving proliferation in HCC cells.