Human normal liver cell lines, L02 and Chang liver, and human hepatoma cell lines, HepG2 and Huh7, were cultured in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% fetal bovine serum (FBS) and 1% antibiotics [penicillin-streptomycin solution (PSN)], under identical conditions (37°C in humidified 5% CO₂/95% air), respectively. For drug treatment, cell lines were treated with 5 μmol/l 5-aza-2′-deoxycytidine (Aza) (Sigma, St. Louis, MO, USA) for 3 days, changing Aza and medium every 24 h. Control cells were incubated with culture medium.

There were clear similarities between the results from the Chang liver and L02 cell. The results from Huh 7 were similar to those of the HepG2 cells. Therefore, the results from Chang liver and Huh7 cells are not presented.

HCC paraffin blocks and frozen tissues were obtained from the archives of the Department of Pathology, Beijing Ditan Hospital, Capital Medical University, Beijing, China, according to institutional review board-approved protocol. The use of human specimens in this research was approved by the ethics committee of our hospital according to the Declaration of Helsinki. We clearly confirm that we had all the necessary consent from any patient involved in the study, including consent to participate in the study and consent to publish.

RNA was extracted from cells or patient tissues using an RNA isolation reagent (TRIzol; Invitrogen Life Technologies, Carlsbad, CA, USA). To prevent DNA contamination, total RNA was treated with RNase-free DNase II (Invitrogen Life Technologies).

The human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene (forward primer, 5′-TCACCAGGGCTGCTT TTA-3′ and reverse, 5′-TTCACACCCATGACGAACA-3′) was used as an internal control in the PCR amplification. A two-step RT-PCR procedure was performed in all experiments. First, total RNA samples (2 μg/reaction) were reversely transcribed into cDNAs by RT II reverse transcriptase (Invitrogen Life Technologies). Then, the cDNAs were used as templates in PCR with NOLC1 gene-specific primers (5′-AGCTGGCCTGACGGTATG-3′ and 5′-TTGGTCTGGCTGAGTACCG-3′). The primers for cyclin D3 were 5′-ATTCCTCTTTGCTTTG CTTTC-3′ and 5′-CAGCAGCAAAGCTGTCAATC-3′. The primers that were used for amplification of MDM2 were 5′-CGGCACGAGCTAGGATCT and 5′-ACGGCAGCTCCA TGAGTC-3′. The amplification reactions were performed using AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, CA, USA). The PCR was programmed as follows: 2 min at 95°C and 30–32 cycles of 30 sec at 94°C, 30 sec at 58–62°C, and 30 sec at 72°C, with an extension for 10 min at 72°C. The PCR bands were visualized under UV light and photographed. Quantitative real-time PCR was used to measure mRNA levels of urea cycle genes for the HCC cell lines and NOLC1 mRNA levels for tissues using the StepOne Plus Real-Time PCR system with SYBR-Green Master Mix (Applied Biosystems).

HCC biopsy specimens were subjected to routine immunohistochemical staining using a monoclonal antibody directed against NOLC1 (13), according to a previously described method (14). Immunoreactivity, defined as the number of positive tumor cells over the total tumor cells, was scored independently by two researchers. The number of NOLC1-positive and -negative HCC cells was counted under a light microscope at a magnification of ×400. Only the cells displaying brown nucleoli on the section were considered NOLC1-positive. For each slide, 7–10 microscopic fields were randomly chosen. Positive scores were categorized into weak staining (only one nucleolus was stained), moderate staining (more than one nucleolus was stained), and strong staining (both the nucleus and nucleolus of the tumor cells were stained). The average percentage of NOLC1-positive HCC cells was then calculated for each group.

Lysates from the cultured cells were subjected to routine western blotting as described previously (15). The antibodies used were monoclonal anti-mouse antibodies against NOLC1 (13) and β-actin (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), and polyclonal rabbit antibodies against cyclin D3 and caspase-8 (Abcam, Cambridge, MA, USA). The results shown are representative of 2 independent experiments.

Genomic DNA was purified from cells with the Wizard Genomic DNA purification kit (Promega Corporation, Madison, WI, USA). DNA (2 μg) was bisulfite modified with the EZ DNA methylation direct kit (Zymo Research, Orange, CA, USA). Sequence-specific primers to amplify the CpG-rich regions of interest were designed using a computer program (MethPrimer; http://www.urogene.org/methprimer). The primers that were used for amplification were as follows: NOLC1 upstream CpG island-1 forward, 5′-TTGGGATGGTATTAGAAAGGGT-3′ and reverse 5′-CTCAAAAACCCAACAAAAACAAT-3′; NOLC1 upstream CpG island-2 forward, 5′-GTTTTTGTTGGGTTTTTGAGT-3′ and reverse 5′-CCTCAATAAAACAAAACTCTA CCTC-3′. The PCR products were amplified, purified and cloned into a vector (pGEM-T Easy Vector; Promega Corporation). Clones were selected through blue-white screening. Finally, the colonies harboring the insert were sequenced in a 96-well plate using the M13 reverse and/or forward primers.

The bisulfite-treated DNA was amplified using primers that specifically amplify either the methylated or unmethylated sequence of the NOLC1 promoter containing four CpG dinucleotides, respectively. The PCR was performed for 40 cycles, with annealing temperatures of 58°C for the methylated reaction and 52°C for the unmethylated reaction. The human methylated and unmethylated DNA was used as a control to verify the specificity of the primers (Qiagen, Valencia, CA, USA).

Cells were subcultured and transfected as previously described (16,17). The cDNA encoding NOLC1, flanked by BamHI and SalI restriction sites, was cloned into the mammalian expression vector pCDNA4 (Stratagene Inc., La Jolla, CA, USA) to generate pCDNA-NOLC1, which expresses an N-terminal myc-tagged NOLC1 fusion protein. The promoter of NOLC1 was cloned into the pGL3 plasmid. Subconfluent cells were transiently transfected with pCDNA-NOLC1 DNA (4 μg/dish) mixed with Lipofectamine and Plus™ reagent (Invitrogen Life Technologies), according to the manufacturer’s protocol. Cells were harvested ~48 h after transfection (18).

The shRNA constructs described in Table I were purchased from Open Biosystems (Huntsville, AL, USA). When the HCC cultured cells had reached 70–80% confluence, the shRNA constructs were transfected into the HCC cells using the Arrest-In transfection reagent for RNAi (Open Biosystems) (12).

The cells were transfected with 0.6 μg of firefly luciferase reporter plasmid and 0.05 μg of control plasmid containing Renilla luciferase (pRL-TK; Promega Corporation). A promoterless basic vector (pGL3; Promega Corporation) was used as a negative control. To confirm the efficiency of transfection (Lipofectin; Invitrogen Life Technologies), a luciferase expression vector (pGL3-control; Promega Corporation) was used as a positive control. After 48 h, the cells were harvested for analysis. Luciferase enzyme assays and colorimetric β-galactosidase assays were performed according to the manufacturer’s instructions (Promega Corporation). Luciferase activity was normalized to β-galactosidase activity to assess the transfection efficiency. When indicated, the firefly and Renilla luciferase activities were measured using the Dual-Luciferase reporter assay system (Promega Corporation), according to the manufacturer’s instructions. The Renilla luciferase activity of pRL-TK was used to normalize the firefly luciferase activity of the reporter construct. Each transfection experiment was repeated 3 times.

Cell growth was determined by the colorimetric tetrazolium derived sodium 3′-[1-(phenylamino-carbonyl)-3,4-tetrazolium]-bis(4-methoxy-6-nitro)benzene-sulfonic acid hydrate assay (XTT), and DNA synthesis of cells was assessed by the bromodeoxyuridine (BrdU) incorporation assay (both from Roche Applied Science, Mannheim, Germany). For the cell growth and proliferation assays, the cells of each group at 48 h after treatment were re-seeded onto 96-well plates at a density of 3×10³ cells/well. Then XTT and incorporated BrdU were measured colorimetrically using a microtiter plate reader (Bio-Rad, Hercules, CA, USA) at a wavelength of 450 nm (19,20).

Cells (3×10⁵) were cultured onto 60-mm SWNHs-coated and non-coated dishes for 48 h. Then apoptotic cells were identified by using fluorescence-activated cell sorting (FACS) Annexin V-Fluos (BioLegend) following the protocol of the manufacturer. Cells having been cultured were washed at 4°C for 30 min in PBS and then stained with Annexin-V staining solution at 4°C for 3 h. Gels were washed 4 times in PBS at 4°C and fixed at room temperature with 1% paraformaldehyde (Sigma-Aldrich, St. Louis, MO, USA) in PBS for 15 min. For counterstaining, 7-AAD (2 μg/ml) was added to the first washing step. The numbers of total, Annexin-V-positive, 7-AAD-positive, and double-positive cells were determined respectively using FACS technique. Apoptosis was verified by detection of activated caspases (21).

Significance was determined using the one-way ANOVA test for the mean of three different experiments. Significance was determined using the paired Student’s t-test for the mean of 3 different experiments. Probabilities of ≤0.05 were considered to indicate statistically significant results.

To understand the potential mechanisms by which NOLC1 is regulated in HCC, we detected the NOLC1 expression levels at the transcription and translation levels in HCC cell lines and human HCC tissues using immunohistochemical staining and RT-PCR approach. The results showed that NOLC1 was moderately expressed in the cytoplasm and the nucleus (Fig. 1A). It was expression at a low level in HCC tissues and was localized in the nucleus (Fig. 1B). We then examined NOLC1 mRNA expression in 38 pairs of patient tissues. The NOLC1 expression was decreased in tumor tissues when compared with that in the matched noncancerous tissues after 32 cycles of PCR amplification (P<0.01) (Fig. 1C). As shown in Fig. 1D, strong expression of NOLC1 protein was detected in the normal liver cell lines, L02 and Chang liver, but NOLC1 protein was weakly expressed in hepatoma liver cell lines HepG2 and Huh7. To correlate the mRNA transcription with protein expression, Western blot analysis was performed to examine the NOLC1 protein expression in patient tissues. As shown in Fig. 1E, NOLC1 protein in the tumor samples was decreased at different degrees when compared with that in the adjacent noncancerous specimens, particularly in patients 1–7. Hwang et al(12) found that NOLC1 plays a role in the regulation of tumorigenesis of nasopharyngeal carcinoma (NPC) and demonstrated that both NOLC1 and tumor protein 53 synergistically activate the MDM2 promoter in NPC cells. The frequent downregulation of miR-138 regulates cyclin D3 and functions as a tumor-suppressor in HCC (22). We next examined MDM2 and cyclin D3 mRNA expression in patient tissues. MDM2 expression in tumor tissues was similar to that in the matched noncancerous tissues (Fig. 2A), while the tumor samples exhibited an increase in cyclin D3 expression when compared with that in the adjacent noncancerous specimens (Fig. 2B).

Since many cancer cells exhibit aberrant epigenetic regulation, it is possible that NOLC1 expression is regulated by epigenetic modification. To confirm that DNA methylation regulates NOLC1 expression, detailed methylation analysis of the NOLC1 gene sequence was performed using genomic DNA extracted from cell lines and liver tissues. Sequence analysis (GenBank accession no. GI 470595142; http://www.ncbi.nlm.nih.gov/mapview) (23) indicated that the NOLC1 promoter has 2 typical CpG islands (CpG1 and CpG2) upstream of it. We predicted the CpG islands of the NOLC1 promoter: the length of CpG island 1 (CpG1) was 616 bp (−1987 and −2602), and the length of CpG island 2 (CpG2) was 148 bp (−2678 and −2825) (Fig. 3A). To determine whether NOLC1 promoter methylation is associated with the control of NOLC1 expression in liver cell lines and clinical specimens, we performed methylation-specific PCR and compared the promoter methylation status in the hepatoma cell lines with that in the normal liver cell lines, and in tumor tissues with that in paired noncancerous tissues. Methylation of the 2 CpG dinucleotides in the promoter region was detectable. As shown in Fig. 3, the results revealed that 4 dinucleotides at the start of CpG1 were strongly methylated (Fig. 3A and D), particularly in the hepatoma cell lines and tumor tissues when compared with that in normal liver cell lines and adjacent noncancerous tissues (Fig. 3D). The methylation frequencies were 58.07% in CpG1-57, 47.8% in CpG1-72, 44.9% in CpG1-82 and 37.7% in CpG1-93 (Fig. 3B). But CpG2 was not methylated (Fig. 3C). The methylation status in the promoter CpG1 start region appeared to be correlated with NOLC1 expression levels in the liver cell lines and tissues specimen.

To investigate the possible effect of methylation of the CpG1 on promoter activity and determine the functional significance of the CpG1, we generated a reporter gene construct using a NOLC1 promoter sequence containing CpG1. The reporter construct (pNOLC1) together with pRL-TK Renilla luciferase expression vector were transiently transfected in L02 and HepG2 cells, followed by Aza treatment. The promoter activity was determined by luciferase assay. Firefly and Renilla luciferase activities were measured at the points indicated. Renilla luciferase activity was used to normalize firefly luciferase activity of the reporter constructs. As shown in Fig. 4, Aza treatment caused a significant increase in promoter activity in both cell lines, particular in the HepG2 cells (Fig. 4B). Luciferase activity of pGL3-basic, which has no promoter element, was not affected by Aza treatment. The CpG1 dinucleotides in the plasmid were not methylated at transfection (data not shown). Thus, the data suggest that the 4 CG dinucleotides at the CpG1 island start site appear to be critical for NOLC1 promoter activity.

It is uncertain how NOLC1 affects cellular function. Cells were synchronized at the G1/S boundary by double thymidine block, and then released into mitosis. After 24 h, BrdU was added into the medium at the indicated time points to evaluate DNA synthesis. As shown in Fig. 5A, incorporation of BrdU into the control, and accumulation of mitotic L02 cells was significantly promoted by shRNA interfere NOLC1 (P<0.05). In contrast, overexpression of NOLC1 in HepG2 cells were significantly delayed at 36 and 48 h, particularly in cells treated with Aza (P<0.05) (Fig. 5B).

As determined by XTT assays, NOLC1 silencing of L02 cells resulted in a significant increased in cell growth when compared to that in the control and other cells (P<0.05) (Fig. 5C). However, Aza treatment in HepG2 cells significantly inhibited cell growth when compared with than the control group (P<0.01) (Fig. 5D).

The effect of NOLC1 expression on apoptosis was determined in L02 and HepG2 cells by flow cytometry. Downregulation of NOLC1 in L02 cells significantly suppressed apoptosis (P<0.01) (Fig. 6A). Furthermore, HepG2 cells treated with Aza exhibited increased apoptosis (P<0.01) (Fig. 6B), which may be associated with overexpression of caspase-8 and low levels of cyclin D3 (Fig. 7B).