This article is mentioned in:

Introduction

Hepatocellular carcinoma (HCC) is one of the most common malignant tumors. It is the fifth most common cancer and ranks third in cancer-related deaths. In developing countries the incidence rate is even higher (1), and in developed countries the incidence rate is gradually increasing (2). Due to higher and earlier metastasis, HCC has a very poor prognosis with a 5-year survival of only ~3–5% (3,4). Several options are available for HCC therapy, such as local treatment, surgical resection, targeted biological therapy, systemic chemotherapy and liver transplantation. However, the cure rate and overall survival have been unsatisfactory. Many cancer patients are unable to undergo operational therapy surgery due to poor liver function resulted resulting from chronic liver diseases (5). Therefore, there is an urgent need for early diagnosis and treatment of HCC. Identification of HCC-related genes is essential to have a better understating understanding of the pathogenic mechanisms in order to develop new diagnostic and therapeutic strategies.

Chromatin assembly factor 1(CAF-1) is a trimer histone chaperone located in the nucleus, which mediates chromatin assembly after DNA replication and repair (6). CAF-1 is a heterotrimeric complex composed of three subunits p150, p60 and p48. It recruits histones H3 and H4 to DNA to facilitate nucleosome assembly, participates in the regulation of DNA repair and epigenetic changes in embryonic stem cellsas well as cell proliferation. It has been revealed to be an important histone chaperone in eukaryotic cells such as yeast and multicellular organisms (7). CAF1B is a p60 subunit of CAF-1, which is positively related to cell proliferation, and is located in the nucleolus during interphase and in the site of DNA replication in the S phase (8). It has been proposed as a clinical marker to distinguish quiescent from proliferating cells (9). High expression of CAF-1 has been revealed to be associated with the invasion and metastasis of oral squamous cell carcinoma (10). High expression of CAF1B was demonstrated to be an indicator of poor prognosis in neuroblastoma (11). In the DNA synthesis phase CAF-1 is reversibly phosphorylated by Cyclin/Cdk2 (12). These findings indicated that the expression level of CAF-1 as well as CAF1B is closely related to tumor metastasis, invasion and poor prognosis (11,13). Numerous studies have shown that CAF-1 controls specific chromatin recombination in the S phase to promote cell cycle progression and influences cell proliferation and apoptosis. It has been demonstrated that CAF-1 is essential for differentiation and proliferation of higher eukaryotic animal cells. Recently, CAF-1, particularly CHAF1B, was proposed as a new marker for tumor proliferation and prognosis and overexpression of CAF-1/p60 was used as a clinical marker for malignant tumor progression in some human malignant tumors (14,15).

Previous studies have confirmed that CHAF1B is expressed in HCC cell lines Hep3B, HUH-7, HepG2 and SMMC-7721. Analysis of the expression data in the TCGA database revealed that the gene had higher mRNA abundance in a large number of HCC tissues than in the adjacent tissues. Therefore, we speculated that CHAF1B may be related to the biological behavior of HCC. In the present study, we knocked down the CHAF1B gene in HUH-7 cells using an RNAi lentiviral vector to investigate its effect on proliferation, apoptosis and the cell cycle of the cells. The findings would provide insight into the role of the gene in HCC, help in the identification of new signaling molecules for HCC research and validate its potential in HCC diagnosis and therapy.

Materials and methods

Cell line, reagents and equipment

Human HCC cell line HUH-7 was purchased from Yingniurui Biological Co. (Wuxi, Jiangsu, China). DMEM high glucose medium was obtained from Sijiqing Biotech (Hangzhou, China) (batch no. KGM12800S-500), DMSO was purchased from Amresco (Solon, OH, USA) (batch no. 302A0325), 0.25% trypsin (containing EDTA) was purchased from Kaiji Biotech (Jiangsu, China) (batch no. 20160818), penicillin and streptomycin mix solution and GoldView I nucleic acid stain were obtained from Beijing Solarbio Science & Technology Co., Ltd. (Beijing, China) (batch no. 20160909), fetal bovine serum (FBS) was purchased from BI Biotech (Beijing, China) (batch no. 1552680) and Transwell cells were obtained from Falcon (Atlanta, GA, USA). GoldStar Taq MasterMix and UltraSYBR Mixture were products from CWBiotech (Beijing, China). Primers were synthesized at General Biotech (Beijing, China). The ultra-sensitive chemiluminescence imaging system (ChemiDoc™XRS+) and real-time PCR instrument CFX Connect™ were purchased by Bio-Rad Laboratories, Inc. (Hercules, CA, USA).

Experimental grouping

The cells were divided into the control that did not receive any treatment, the empty-vector group that was transfected with a vector without an insert andthe CHAF1B group that was transfected with CHAF1B.

Construction of lentivirus for CHAF1B knockdown

Interference sequences (upstream 5′-CACCGCCACTTAGAAGATGTGTATGCGAACATACACATCTTCTAAGTGGC and downstream 5′-AAAAGCCACTTAGAAGATGTGTATGTTCGCATACACATCTTCTAAGTGGC) for CHAF1B were designed and synthesized and ligated to lentivirus vector pDS019_pL_shRNA_F. The lentivirus containing vector pDS019-PL-shRNA-GFP-homo-CHAF1B was packed and verified using a fluorescence microscope. The virus titer was determined using the TCID50 method and its expression was confirmed using western blot and PCR analyses. The virus was then amplified and purified using CsC1 gradient centrifugation. Cells without any treatment were used as the blank control, while the empty vector was used as a control.

Cell invasion and migration assays

Prior to the assessment of the invasion ability, the cells were infected with the CHAF1B-knockdown virus or control for 24 h. Then 5×105 cells in serum-free medium were placed into the upper chamber of an insert (8 µm pore size; BD Biosciences (San Jose, CA, USA). After 48 h of incubation, the cells remaining on the upper membrane were removed with cotton wool, whereas the cells that had migrated or invaded through the membrane were stained with 0.1% crystal violet in 25% methanol/PBS, imaged and counted under an inverted microscope (Life Technologies; Thermo Fisher Scientific, Inc., Waltham, MA, USA). The experiments were independently repeated three times.

For the assessment of migration ability, an in vitro scratch assay was used as previously reported (16).

Colony-formation ability assay

Cells (0.4 ml) were inoculated into culture medium in the wells of a 96-plate (400 cells/well) in a total volume of 3 ml and cultured for 2 to 3 weeks. The medium was refreshed every 3 days. Once the colonies became visible, they were fixed in methanol, and then washed and stained with Giemsa staining solution for 30 min. The number of colonies were counted and the colony-forming ability was calculated. The experiments were repeated three times independently.

Flow cytometry

Cells were washed with pre-chilled PBS three times and fixed in 1 ml 70% pre-chilled ethanol overnight at 4°C and washed again with PBS three times. The cells were re-suspended in 1 ml PI/Triton X-100 staining solution containing 0.2 mg RNase A for 15 min at 37°C and then analyzed by flow cytometry. Cells (2×104) were assayed at each cell cycle.

Expression profiling and annotation

Total RNA was extracted using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and an RNeasy Kit (Qiagen GmbH, Hilden, Germany). The first strands of cDNA were synthesized using a cDNA synthesis kit (Invitrogen, Thermo Fisher Scientific, Inc.) and labelled using a monochromatic DNA labeling kit (Roche NimbleGen, Inc., Madison, WI, USA). The labeled cDNA was hybridized to GeneChip Human Gene 2.0 ST Array (Affymetrix; Thermo Fisher Scientific, Inc.) using a hybridization kit (Roche NimbleGen, Inc.) and the signals were scanned, with the GenePix 4000B chip scanner. The data were normalized using NimbleScan software (v. 2.0) and analyzed for differentially expressed genes using Agilent GeneSpring GX software (v. 11.5.1). Pathway analysis was performed against the latest KEGG (Kyoto Encyclopedia of Genes and Genomes) database.

Real-time PCR

Total RNA was extracted from the cells using TRIzol reagent according to the supplier's instructions, reversely transcripted into cDNA and used for PCR assays using primers for CHAF1B as shown in Table I. PCR was carried out in a total volume of 20 µl containing 1 µl of diluted and pre-amplified cDNA, 12.5 µl of 2X UltraSYBR Mixture and 1 µl of each fluorescence TaqMan probe using primers listed in Table I. The cycling conditions were 50°C for 2 min, 95°C for 3 min followed by 40 cycles, each one consisting of 15 sec at 95°C and 30 sec at 50.6°C. The samples were run in triplicate and the mean value was calculated for each case.

Table I. Primers for RT-PCR.

Table I.

Primers for RT-PCR.

Gene	Primer sequence
DNAJB1	F 3′-GATGGCTCTGATGTCATTTATC-5′
	R 3′-GCCTTCTCCAGGAACTTTT-5′
SBNO1	F 3′-ATCCAGTGTTACTCCTCCTG-5′
	R 3′-TGCTATCGTCCTTCCTTT-5′
BLM	F 3′-TGAATCCAGAAACCAGCAC-5′
	R 3′-AAGCAGTTCGTTCCCACA-5′
PSMB6	F 3′-CTGGGAAAGCCGAGAAGT-5′
	R 3′-ATGCGGTCGTGAATAGGT-5′
SLC30A7	F 3′-TAGGCTTGATTTCCGACTC-5′
	R 3′-CCGCTCTAACATACCCATA-5′
DDX3X	F 3′-GGAGCGAATGCGTAAGGT-5′
	R 3′-TTAGGGTGCGAAATGCTG-5′
TWF2	F 3-′AGGTTGTGATTGAGGACGAG-5′
	R 3′-GGAGTTATCAGGCGACCAG-5′
SMC3	F 3′-TGTGATGAACCTCCTTGA-5′
	R 3′-TCTGAGAATCTGGTGCTG-5′
TMEM259	F 3′-CGAGACGCCCACCAAAGT-5
	R 3′-CATCGTAGCCCAGGAACTCAT-5′
β-actin	F 3′-CCTGTATGCCTCTGGTCG-5′
	R 3′-GGCGTAACCCTCGTAGAT-5′

[i] F, forward; R, reverse.

The data were managed using the Applied Biosystems software RQ Manager v. 1.2.1 (Applied Biosystems, Foster City, CA, USA). The relative expression was calculated using the comparative Ct method and by obtaining the fold change value (2−ΔΔCt) according to a previously described protocol (17).

Western blot analysis

Cells were lysed using RIPA buffer (50 mM Tris, pH 7.2; 150 mM NaCl; 1% Triton X-100; and 0.1% SDS) containing protease (1:100; Roche Diagnostics, Indianapolis, IN, USA) and phosphatase (1:100; Sigma-Aldrich, St. Louis, MO, USA) inhibitors. The protein concentrations were determined using a bicinchoninic acid assay (Pierce; Thermo Fisher Scientific, Inc.). Proteins (60 µg) were separated by SDS-PAGE and transferred (Bio-Rad Laboratories, Inc.) to PVDF membranes (EMD Millipore, Billerica, MA, USA). The membranes were incubated with rabbit polyclonal antibodies specific for SMC3 (1:2,000; cat. no. ab9263) and β-actin (1:1,000; cat. no. ab8266; both from Abcam, Cambridge, MA, USA), PSMB6 (1:2,000; cat. no. A4053) TWF2 (1:2,000; cat. no. A5860) BLM (1:2,000; cat. no. A6535) and SLC30A7 (1:2,000; cat. no. A5172; all from ABclonal Biotech Co., Ltd., Woburn, MA, USA). The expression levels of these three proteins were standardized to human α-actin using a mouse polyclonal anti-α-actin antibody (1:1,000; cat. no. MAB1501R; EMD Millipore). Primary antibodies were detected using goat anti-rabbit or goat anti-mouse horseradish peroxidase (HRP)-conjugated secondary antibodies (1:1,000; AP189P; EMD Millipore). Immunoreactive bands were visualized using the ultra-sensitive chemiluminescence imaging system (ChemiDocXRS+) according to the manufacturer's instructions, and then quantified by densitometry using a ChemiGenius Gel Bio Imaging System (Syngene, Frederick, MD, USA).

Construction of HCC nude mouse models

Infected and control HUH-7 cells were digested, washed three times with PBS and re-suspended in PBS. Cells (200 µl; 3×106/mouse) were subcutaneously injected into nude mice. The animals were scarified three weeks later and the tumor tissues were collected for analysis.

H&E staining

Slides were dewaxed, dehydrated and stained with hematoxylin staining solution for 5 min. The stained slides were soaked in hydrochloric acid alcohol for 30 sec, washed twice with PBS, briefly soaked in 0.5% ammonia for 1 to 2 min. After being washed with PBS, the slides were counterstained with eosin for 2 min, washed 2 times with PBS, dehydrated, sealed with neutral balsam and images were captured under a E400 Nikon microscope at ×400 magnification (Nikon, Tokyo, Japan).

Statistical analysis

Statistical analyses were performed using SPSS 20 software (SPSS, Inc., Chicago, IL, USA). All experiments were repeated at least three times and performed in triplicate. The means were compared using the Student's t-test or one-way ANOVA with the corresponding post hoc test. P≤0.05 was considered to indicate a statistically significant result.

Results

Transfection of HUH-7 cells with CHAF1B-silencing lentivirus

We examined the transfection efficiency by observing the intensity of GFP fluorescence 24 h after the transfection. Observations revealed that >95% of cells were fluorescent after the cells were transfected with a lentivirus harboring CHAF1B-shRNA (Fig. 1A). No fluorescence was observed in the untransfected cells (data not shown). Western blot analysis revealed that CHAF1B was significantly downregulated at the protein level (Fig. 1B).

Figure 1. GFP fluorescence and expression of CHAF1B from HUH-7 cells transfected with a lentivirus harboring CHAF1B-shRNA. (A) GFP fluorescence and (B) the protein level of CHAF1B. **P<0.01.

Invasion ability of HUH-7 cells after CHAF1B knockdown

We then examined the invasion ability of HUH-7 cells after the knockdown of the CHAF1B gene. The number of invaded cells was significantly lower after knockdown as compared to the control (16.33 vs. 42.13, P<0.01), while the negative vector did not change the invasion ability (Fig. 2A), indicating that the downregulation of CHAF1B expression reduces the invasion ability of HUH-7 cells, suggesting that it may be used as a new approach for HCC therapy.

Figure 2. Invasion, migration, colony-forming ability and the cell cycle of HUH-7 cells after the knockdown of CHAF1B. (A) Right panel, microimages of the Transwell assay; left panel, the number of invaded cells. (B) Microimages of the scratch assay. (C) Colony formation rate. (D) Right panel, flow cytometric results; left panel, the percentage of cells in different cell phases. *P<0.05 and **P<0.01 vs. the control, respectively.

Migration and colony-forming ability of HUH-7 cells after CHAF1B knockdown

We next investigated the effect of CHAF1B knockdown on migration and colony-forming ability of HUH-7 cells. The scratch assays revealed that compared with the control, CHAF1B knockdown significantly reduced the migration distance of HUH-7 cells (98.6±3.29 µm vs. 41.2±2.59 µm, P<0.01) (Fig. 2B). We also compared the colony formation ability as a measure of tumor-forming ability. The results revealed that CHAF1B knockdown significantly reduced the colony formation rate as compared with the control (Fig. 2C, P<0.01). These findings revealed that silencing of CHAF1B may reduce the tumorigenic ability of HUH-7 cells and inhibit tumor occurrence and development.

Effect of CHAF1B knockdown on the cell cycle of HUH-7 cells

Flow cytometric analysis revealed that there were significantly less HUH-7 cells in the G0/G1 phase and more in the S phase after CHAF1B knockdown (36.10 and 59.7 vs. 54.10 and 40.8%). However, the percentage of cells at the G2/M phase were similar (4.20 vs. 5.10%), indicating that knockdown of the CHAF1B gene reduced the number of cells in the G1 phase and increased the cells in the S phase, resulting in changes in the distribution of cells in different phases (Fig. 2D).

Effect of CHAF1B knockdown on the apoptosis of HUH-7 cells

Flow cytometric analysis revealed that following CHAF1B knockdown, HUH-7 cells had a significantly higher apoptosis rate as compared with the control (Fig. 3).

Figure 3. Apoptosis of HUH-7 cells after CHAF1B knockdown. Upper panel, flow cytometric results; lower panel, percentage of apoptotic cells. **P<0.01.

Expression of downstream genes in HUH-7 cells after CHAF1B knockdown. We profiled the expression of genes using the gene chip analysis. Several genes were found differentially expressed in the knockdown cells as compared with the control (Fig. 4), among them, the expression levels of PSMB6, SLC30A7, SMC3, TWF2 and BLM were the most markedlyaltered. RT-PCR results revealed that the mRNA levels of TWF2, TMEM259, DDX3X, SBNO1 and BLM were downregulated while the levels of PSMB6, DNAJB1, SLC30A7 and SMC3 were upregulated after CHAF1B knockdown (Fig. 5). Western blot analyses revealed that the protein levels of PSMB6, SLC30A7 and SMC3 were significantly increased and the levels of BLM and TWF2 were significantly decreased following CHAF1B knockdown (Fig. 6).

Figure 4. Expression profiling of genes in HUH-7 cells after CHAF1B knockdown.

Figure 5. mRNA levels of downstream genes in HUH-7 cells after CHAF1B knockdown. (A) Expression profiling. *P<0.05 and **P<0.01 vs. the control, respectively.

Figure 6. Western blot analyses of downstream genes in HUH-7 cells after CHAF1B knockdown. Right panel, representativewestern blots; left panel, relative protein content. *P<0.05 and **P<0.01 vs. the control, respectively.

Effect of CHAF1B knockdown on tumor growth

We then compared the growth of tumors in the mouse models prepared using CHAF1B-knockdown HUH-7 cells. Twenty-eight days after the injections, the volume and tumor formation rate were similar between the control and the empty vector. However, these figures were significantly lower in mice injected with CHAF1B knockdown-cells as compared with the control (Fig. 7A). H&E staining further confirmed that tumors were present in rats in all groups (Fig. 7B). Western blot analyses revealed that CHAF1B was significantly lower after CHAF1B-knockdown (Fig. 8A) and TUNEL assay revealed that there were more apoptotic cells in the CHAF1B-knockdown HUH-7 cells (Fig. 8B). These findings demonstrated that CHAF1B knockdown reduced the tumorigenicity of HUH-7 cells.

Figure 7. Development of the tumors and pathological changes in mice injected with CHAF1B-knockdown HUH-7 cells. (A) Tumor volume and (B) H&E staining of the tumor tissues. *P<0.05 vs. control.

Figure 8. CHAF1B expression and HUH-7 cell apoptosis after CHAF1B knockdown. (A) Left panel, representative western blot; right panel, relative CHAF1B level. (B) TUNEL assay. **P<0.01.

Discussion

Chromatin assembly factors (CAFs) have subunits of different sizes (150 kDa, CAF-1-p150 or CHAF1a), (60 kDa, CAF-1-p60 or CHAF1B) and (48 kDa, CAF-1-p48 or RbAp48, p48) (18). Verreault et al (19) purified the complex and revealed that CHAF1B can reposition nucleosomes, and change the position and structure of nucleosomes to regulate the level of chromatin during DNA replication, transcription, repair and recombination (20). They determined that transcriptional factors (TFs) and CAFs bind to specific sites in the promoter to change the position and structure of nucleosomes and sensitize chromatin to ribozymes. The synthesis of DNA is the basis for nucleosome assembly and plays a key role in maintaining genome stability. During the replication of DNA, the synthesis of CAF-1 in histones H3 and H4 is turned off directly in many tissues, and the large subunit (CAF-1-P150, CHAF1a) and small subunit (RbAp48, P48) of CAF-1 have a variety of functions, for instance, as histone chaperones (21).

In recent years, CAF1B has been demonstrated to be highly expressed in a variety of malignant tumors, and has been identified as a marker of poor prognosis in cancer patients (7,10). The expression level of CHAF1B has been demonstrated to be related to tumor invasiveness (11). It has been revealed that from radial growth in the early stage to pituitary growth, the expression level of CAF1B is significantly elevated, and the level also increases during the transition from benign nevus to malignant melanoma (21). Polo et al revealed that in breast, cervical and endometrial cancer as well as in renal cell carcinoma, higher expression of CAF-1 was related to histological grading and may be used as an independent factor of poor prognosis (22). Our experimental results also revealed that knockdown of the CHAF1B gene reduced the invasiveness, migration and colony-forming ability of HCC cells, further confirming that CHAF1B is involved in HCC biology.

CHAF1B may play a role in predicting therapeutic efficacy and monitoring drug response (22). Inhibition of this protein in invasive tumors can lead to tumor cell death and therefore it may become a new therapeutic target (23). CAF-1 is a p48 complex composed of three subunits p150, p60, and p48 (21,24–26). Its compositions and functions are highly conserved in genome replication (27). The primary role of CAF-1 in DNA replication is to facilitate the first step in assembling nucleosomes to the newly synthesized DNA (9,28,29). At any time in the cell cycle, DNA damage could lead to the initiation of DNA damage response (DDR). In eukaryotic tissues, DDR coordinates DNA damage repair and repair of altered chromatin (30). For example, CAF-1 may recruit a large number of damage-repairing proteins to the damaged chromatin regions as histone chaperones and/or the protein platform to participate in DNA and chromatin damage repair (9,31–33). Recent studies also demonstrated that CAF-1 may serve as a protein platform of epigenetics to inhibit chromatin markers or signal transduction of active histone, which are important for normal cell activities (34,35). In addition, numerous studies have revealed that CAF-1 promoted cell cycle progression by adjusting the specific chromatin reorganization in the S phase to affect cell proliferation (34).

In addition, CAF-1 has been demonstrated to be associated with human diseases. The main function of the p48 subunit is to support the role of retinoblastoma protein RbAp48 in the inhibition of cell growth. p150 is directly related to proliferating cell nuclear antigen (PCNA), and actively participates in the DNA repair process during reproduction (28). The CAF1B subunits are highly expressed in breast, oral, tongue, prostate and salivary gland cancer, as well as in skin melanoma and other malignant tumors and downregulation of p150 in tongue cancer and head and neck cancer is an indicator for poor prognosis (36). The present study also demonstrated that the knockdown of the CHAF1B gene significantly inhibited cell proliferation and promoted apoptosis.

Collectively, CAF1B elevation was associated with poor prognosis in HCC, and related to tumor stage, grade and distant metastasis. Therefore, understanding the expression level of CAF1B is helpful for the diagnosis and treatment of HCC, although more studies are warranted to elucidate the specific mechanism. Our study demonstrated that downregulation of CHAF1B inhibited the proliferation of HCC cells and induced apoptosis. These findings provided insight into the involvement of the CHAF1B gene in tumor signal transduction pathways and indicates the possible use of the gene for clinical diagnosis and targeted treatment.

Acknowledgements

This study was supported bythe Department of Science and Technology of Jiangxi Province (grant no. 20152ACG70020).

Competing interests

The authors declare that they have no competing interests.

References