HBO1 overexpression is important for hepatocellular carcinoma cell growth

Hepatocellular carcinoma (HCC) is a common primary liver malignancy lacking effective molecularly-targeted therapies. HBO1 (lysine acetyltransferase 7/KAT7) is a member of MYST histone acetyltransferase family. Its expression and potential function in HCC are studied. We show that HBO1 mRNA and protein expression is elevated in human HCC tissues and HCC cells. HBO1 expression is however low in cancer-surrounding normal liver tissues and hepatocytes. In HepG2 and primary human HCC cells, shRNA-induced HBO1 silencing or CRISPR/Cas9-induced HBO1 knockout potently inhibited cell viability, proliferation, migration, and invasion, while provoking mitochondrial depolarization and apoptosis induction. Conversely, ectopic overexpression of HBO1 by a lentiviral construct augmented HCC cell proliferation, migration and invasion. In vivo, xenografts-bearing HBO1-KO HCC cells grew significantly slower than xenografts with control HCC cells in severe combined immunodeficient mice. These results suggest HBO1 overexpression is important for HCC cell progression.


Introduction
Liver cancer is the fifth most common cancer and the 2th leading cause of cancer-related human mortalities globally, with over 840,000 new cases and 780,000 deaths reported each year 1,2 . Of which 75-80% of liver cancer is hepatocellular carcinoma (HCC), the latter is considered as a major health threat globally 3,4 . The overall five-year survival of HCC is far more from satisfactory 5 . The current treatments for HCC are limited, and sorafenib is the only molecularly targeted agent 3,4 . Therefore, patients with advanced HCCs, including those with unresectable, recurrent, and metastatic HCCs, often have very poor prognosis 3,4 . In the past years, HCC's incidence has been steadily rising, especially in Western countries 3,4 . Thus, it is extremely urgent to explore and develop novel anti-HCC strategies and molecularly-targeted agents 6,7 .
Acetylation is an universal protein modification vital for almost all cellular behaviors, from cell cycle progression, gene transcription and expression, signaling transduction, RNA splicing and cellular metabolism, among others 8,9 . The histone acetylation will lead to chromatin unfold, which facilitates proteins accessing DNA being replicated and/or transcribed 10,11 . Histone acetylation on lysine residues is a reversible step regulated by the antagonistic actions of two enzymes: histone acetyltransferases and histone deacetylases 12 . Dysregulation of histone acetylation is commonly detected in HCC, which is associated with HCC tumorigenesis, progression, and therapy resistance 13 . However, the underlying mechanisms are still largely unknown.
The multifunctional HBO1 (lysine acetyltransferase 7/ KAT7) is a primary and essential member of MYST (MOZ, Ybf1/Sas3, Sas2, and Tip60) family histone acetyltransferases 10 . It is responsible for the acetylation of histone H4 and H3K14 10 . HBO1 is vital for the prereplication complex (pre-RC) formation, DNA replication, and cell proliferation via acetylation of histone H4 and H3 14 . HBO1-dictated H3K14ac initiates de novo activation of key embryonic patterning genes during embryonic development 15,16 . HBO1 is actively involved in regulating multiple and key cellular and physiological functions, including DNA replication, gene transcription, and protein ubiquitination as well as immune regulation, stem cell pluripotent, self-renewal maintenance, and embryonic development 10,17,18 .
HBO1 associates with chromatin licensing and DNA replication factor 1 and is required for G1 phase cell cycle progression 16,19 . In addition, HBO1 functions as a positive regulator of centromeric CENPA (Centromere Protein A) by preventing SUV39H1-mediated centromere inactivation 20 . MacPherson et al., discovered that HBO1 is essential for the acetylation of H3K14 (H3K14ac), thereby promoting the processivity of RNA polymerase II to maintain high expression of key oncogenic genes (MYLK, HOXA9, HOXA10, and several others) in leukemia stem cells 21 .
Recent studies have implied an essential role of HBO1 in cancer cell progression 22 . HBO1 silencing by targeted short hairpin RNA (shRNA) resulted in significant proliferation inhibition in MCF7 and HeLa cancer cell lines 14,15 . HBO1 activated Wnt/β-catenin signaling to assist bladder cancer cell proliferation 23 . Taniue et al., found that HBO1-mediated downregulation of tumor suppressor candidate 3 (TUSC3) is essential for colon cancer cell proliferation 24 . Kueh and colleagues however showed that HBO1 did not have an essential role in cell proliferation and DNA replication in HEK293T, MCF7, and HeLa cancer cell lines 25 . Expression and potential function of HBO1 in HCC have not been studied thus far. In the present study, we show that HBO1 overexpression is important for HCC cell growth.

Cell culture
HepG2 HCC cell line and the HL-7702 hepatocytes were obtained from the Cell Bank of CAS Shanghai (Shanghai, China) and cultured as described 26 . The primary human HCC cells, derived from five primary HCC patients, HCC-1/HCC-2/HCC-3/HCC-4/HCC-5, were provided by Dr. Lu at Nanjing Medical University 26,27 . The primary human HCC cells were cultured as described 26 . Human primary adult hepatocytes, purchased from the Cell Bank of Fudan University (Shanghai, China), were derived from the liver of a partial hepatectomy patient. Human hepatocytes were cultured in primary cell culture medium as described previously 26 . In the present studies, protocols testing human tissues and primary human cells were approved by Ethics Review Board of Guangdong Academy of Medical Sciences. The written-informed consent was obtained from each primary HCC patient. All studies were conducted according to the principles expressed in the Declaration of Helsinki and international guidelines.

Human tissues
Human HCC tissues and the surrounding normal liver tissues were from 10 individual written-informed primary HCC patients, enrolled at Guangdong Provincial People's Hospital. HCC patients enrolled received no chemotherapy and radiotherapy before surgery. Tissues were incubated with the described lysis buffer (Biyuntian, Wuxi, China) and tissue lysates stored in liquid nitrogen. The written-informed consent was obtained from each participant. The protocols of this study were approved by the Ethic Committee of Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences.

Quantitative real-time PCR (qPCR)
Total RNA was extracted by Trizol reagents, reversetranscribed and qPCR were performed using the SYBR green kit by an ABI-7500 system (Applied Biosystems, Shanghai, China) 28 . For data quantification, a 2 ΔΔCt method was utilized, and GAPDH tested as the reference gene and internal control. All the primers were synthesized by Genechem (Shanghai, China).

HBO1 shRNA
A pair of lentiviral GV369 constructs (containing GFP gene and puromycin selection gene), encoding nonoverlapping HBO1 shRNA sequences, namely HBO1-shRNA-1/2, were designed, synthesized, and verified by Shanghai Genechem (Shanghai). The construct together with the lentivirus-packing plasmids (psPAX2 and pMD2. G, Shanghai Genechem Co.) were co-transfected to HEK-293T cells, generating HBO1-shRNA lentivirus. The viruses were added to human HCC cells (cultured into six-well tissue plates at 2 × 10 5 cells per well). After 24 h, virus-containing medium was replaced with fresh complete medium, and cells were subjected to FASC sorting to generate monoclonal cells (GFP-positive). Stable HCC cells were further selected by puromycin (5 μg/mL, Sigma) for 10 days. HBO1 silencing in stable cells was verified by qPCR and Western blotting assays.

HBO1 knockout
The small guide RNA (sgRNA) targeting human HBO1 (Target DNA Sequence: GATGAACGAGTCTGCCGA AG. PAM Sequence: AGG) was inserted into a lenti-CRISPR-GFP-puro plasmid (from Dr. Chen at Jiangsu University 29 ). The construct was transfected to HCC cells by using Lipofectamine 2000. Afterwards, GFP-positive cells were sorted by FACS and resulting monoclonal cells were selected by puromycin (5 μg/mL)-containing medium. HBO1 knockout in stable cells was screened by qPCR and Western blotting assays.

Western blotting
In brief, the protein lysates, from human tissues or cultured cells, were separated by 10-12% SDS-PAGE gels (40 μg protein in each lane), and transferred to polyvinylidene difluoride (PVDF) blots (EMD Millipore, Shanghai, China). The blots were blocked and incubated with the applied primary and secondary antibodies, with antibody-antigen binding examined by an ECL kit (GE Healthcare, Chicago, IL, USA). The same set of lysates were run in sister gels to test different proteins. The ImageJ software was utilized for data quantification.

Cell-counting kit 8 (CCK-8) assay
Cells were trypsinized and inoculated into the 96-well tissue-culture plates at 3500 cells per well. After incubation at 37°C for 96 h, 10 µL of CCK-8 reagent (Dojindo, Kumamoto, Japan) was added into each well for 2 h. CCK-8 absorbance, the optical density (OD), was always examined at 450 nm.

Colony formation
HCC-1 primary cells, with applied genetic modifications, were initially seeded at 10,000 cells per well into 10cm tissue-culture plates. The complete medium was renewed every two days (total culture for 10 days), and large colonies (>100 cells/per colony) stained and manually counted.

Migration and invasion assays
The established and primary human HCC cells were trypsinized and suspended into serum-free medium. "Transwell" chambers with 8 μm pore-size were utilized (BD Biosciences, Shanghai, China). For each condition, 30,000 cells were added to the upper surface of the chamber, with the lower chamber filled with complete medium (10% FBS). Cells were allowed to migrate for 16 h, excluding the possible influence from proliferation/ viability change. Afterwards, the migrated cells, in the lower chamber, were fixed, stained and counted. For the invasion assays, "Matrigel" (Sigma, Shanghai, China) was coated to the "Transwell" chambers.

EdU (5-ethynyl-20-deoxyuridine) staining
The established or primary human HCC cells, with or without the applied genetic modifications, were seeded into twelve-well tissue culture plates (at 0 × 10 5 cells per well), cells were cultured for 72 h. An EdU Apollo-567 kit (RiboBio, Guangzhou, China) was utilized, and the cell proliferation ratio (EdU/DAPI×100%) calculated from at least 500 nuclei in five random views per treatment.
Annexin V fluorescence activated cell sorting (FACS) HCC cells, with the applied genetic modifications, were seeded into six-well tissue culture plates (at 2 × 10 5 cells per well), cells were cultured for 48 h and stained with Annexin V-FITC and propidium Iodide (PI) (each at 10 μg/mL). Cells were then subjected to flow cytometry (Beckman Coulter, Brea, CA). The Annexin V-positive cells were gated, and its ratio recorded.

TUNEL assay
HCC cells, with the applied genetic modifications, were seeded into 24-well tissue culture plates (at 0.3 × 10 5 cells per well), cells were further cultured for 48 h and incubated with TUNEL (Invitrogen) for 3 h and DAPI for 5 min. TUNEL and DAPI staining was visualized under a fluorescent microscope (Leica). TUNEL ratio (TUNEL/ DAPI×100%) was calculated from at least 500 nuclei in five random views per treatment.

Mitochondrial depolarization
In stressed cells with mitochondrial depolarization, the JC-1 fluorescence dye can aggregate in mitochondria, forming green JC-1 monomers 31 . HCC cells, with or without the applied genetic modifications, were seeded into 24-well tissue culture plates (at 0.2 × 10 5 cells per well), cells were further cultured for 48 h and incubated with JC-1 (5 μg/mL) for 30 min under the dark at room temperature. Afterwards, cells were washed and tested under a fluorescence spectrofluorometer at 488 nm. The representative JC-1 images, merging both green and red fluorescence channels, were presented as well.

Ectopic HBO1 overexpression
A GV369 construct with HBO1-cDNA was designed, synthesized and verified by Shanghai Genechem (Shanghai, China). The construct together with the lentiviruspacking plasmids were co-transfected to HEK-293T cells, generating HBO1 expression lentivirus. HCC-1 cells were transfected with the viruses for 12 h, and the viruscontaining medium was then replaced with fresh complete medium. Cells were subjected to FASC sorting to generate monoclonal cells (GFP-positive cells). Stable cells were selected by puromycin, and two stable cell lines were established. HBO1 overexpression was verified by qPCR and Western blotting analyses.
Tumor growth in vivo SCID mice (half male half female, 5-6 week old) were purchased from the Experimental Animal Center, School of Medicine of Zhejiang University (Hangzhou, China), maintained under specific pathogen-free (SPF) conditions. All animal studies in the present study are in accordance with IACUC as well as international guidelines and regulations, with the protocols approved by Ethics Review Board of Guangdong Academy of Medical Sciences. The exponentially growing HCC-1 primary cells (1 × 10 7 cells per mice), with CRISPR/Cas9 HBO1-KO construct or the empty vector, were subcutaneously injected into the loose skin in the right front leg of the recipient SCID mice. Within 2 weeks the xenograft HCC-1 tumors were established and recordings were started. The animal studies were approved by Institutional Animal Care and Use Committee (IACUC) and Ethics Committee of Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences. Tumors were cut into small pieces and lysed in tissue lysis buffer (Biyuntian, Wuxi, China). Protein lysates were tested by Western blotting 32 .

Statistical analysis
In vitro experiments were repeated at least three times and similar results were obtained. Quantitative results were expressed as mean ± standard deviation (SD). Statistical analyses among different groups were performed by one-way ANOVA with Scheffe's test using SPSS23.0 software (SPSS Inc., Chicago, IL). The twotailed unpaired T-test (Excel 2007) was utilized when testing the significance between two treatment groups. P values of <0.05 were considered statistically significant.

HBO1 is upregulated in human HCC tissues and cells
First, we tested the expression of HBO1 in human HCC. UALCAN is a comprehensive and interactive web resource for analyzing cancer data. It performs in-depth analysis of gene expression data in different cancer databases, including TCGA, MET500, and CPTAC. It also provides graphs and plots depicting expression profile and patient survival information for protein-coding genes 33 . Through searching the UALCAN database (http://ualcan. path.uab.edu), we found that HBO1 mRNA expression in liver cancer tissue specimens is significantly higher than that in normal liver tissues (P < 0.01, Fig. 1A).
Furthermore, HBO1 overexpression is associated with poor overall-survival in liver cancer patients (P = 0.027 Fig. 1B).
To verify the bioinformatics results, HCC tissues, derived from ten human patients with primary HCC, as well as the matched surrounding normal liver tissues, were obtained. Tissue lysates were achieved by incubation tissues with the tissue lysis buffer. Testing mRNA expression, using qPCR assays, show that HBO1 mRNA expression in HCC tissues ("T") is over four folds of that in the surrounding normal liver tissues ("N") (Fig. 1C). Western blotting assay results, Fig. 1D, confirmed HBO1 protein upregulation in HCC tissues of representative human patients, while its expression is relatively low in matched surrounding normal liver tissues (Fig. 1D). Quantitative analyses of HBO1 protein expression from all ten pairs of tissues confirmed that HBO1 protein upregulation in HCC tissues is significant (P < 0.05 vs. normal liver tissues) (Fig. 1E).
Further studies were performed to test HBO1 expression in human HCC cells. Primary human HCC cells, derived from three primary HCC patients, HCC-1/HCC-2/HCC-3 (from Dr. Lu at Nanjing Medical University 26,27 ), were tested. Results in Fig. 1F demonstrated that HBO1 mRNA is elevated in primary human HCC cells and established HepG2 cells. Conversely, HBO1 mRNA expression is relatively low in primary human hepatocytes and HL-7702 hepatocytes 26,27 (Fig. 1F). HBO1 protein expression was elevated in HepG2 and primary human HCC cells as well (Fig. 1G). Yet, a low HBO1 protein expression is detected in primary human hepatocytes (Fig. 1G). These results imply that HBO1 is upregulated in human HCC tissues and cells.
nuclear localization of β-catenin in cancer cells 23 . We found that β-catenin protein levels were decreased in HCC-1 cells with HBO1 silencing (Fig. 2B).
To study the functional results of HBO1 silencing, a CCK-8 viability assay was performed. As shown, HBO1 silencing resulted in significant viability reduction in HCC-1 cells (Fig. 2C). Furthermore, in HCC-1 cells HBO1 shRNA largely inhibited colony formation (Fig. 2D) and nuclear EdU incorporation (Fig. 2E), indicating proliferation inhibition. Cell cycle progression assessed by flow cytometry with propidium iodine (PI) showed that HBO1 shRNA induced an increase in the G1 cell population with a concomitant decrease of the cell population in S phase in HCC-1 cells. Thus, HBO1 silencing induced G1-S cell cycle arrest in HCC-1 cells (Fig. 2F). Additionally, HBO1-silenced HCC-1 cells showed significantly inhibited cell migration and invasion, which were tested by "Transwell" assay ( Fig. 2G) and "Matrigel Transwell" assay ( Fig. 2H), respectively.
In the primary human hepatocytes, infection with the HBO1-shRNA-2 lentivirus resulted in potent HBO1 mRNA reduction (Fig. S2A). However, HBO1 silencing failed to inhibit cell viability (Fig. S2B) and proliferation (Fig. S2C) in hepatocytes, indicating a specific effect in the cancerous cells. Collectively, these results show that HBO1 silencing inhibited HCC cell viability, proliferation, migration, and invasion.

HBO1 silencing provokes apoptosis in HCC cells
RNA-seq microarray analysis in HBO1-depleted cells identified over 250 differentially regulated genes. A number of them are anti-apoptosis genes 34 . Furthermore, recent studies have reported apoptosis activation in cells with HBO1 depletion 21 . CRISPR/Cas9-induced HBO1 KO was shown to induce apoptosis activation in leukemia stem cells 21 . Similarly, in AML cells HBO1 depletion by CRISPR/Cas9 method reduced proliferation and increased apoptosis activation 45 .

Ectopic overexpression of HBO1 promotes HCC cell progression in vitro
The results have shown that HBO1 silencing or KO will result in significant inhibition of HCC cell progression. We proposed that forced overexpression of HBO1 should exert opposite functions. To test this hypothesis, a HBO1expressing GV369 construct (with GFP) was transduced to HCC-1 primary cells. GFP-positive cells were sorted by FACS, and puromycin was added to select stable cells. Two stable cell lines, HBO1-OE-L1 and HBO1-OE-L2, were established. As compared to the vector control cells, HBO1 mRNA levels increased over 8-10-folds in HBO1-OE HCC-1 cells (Fig. 5A), and HBO1 protein overexpression detected as well (Fig. 5B). H3K14 acetylation was enhanced in HBO1-overexpressed cells (Fig. 5B). MYLK, VEGFR2, PBX3, CCR2, HOXA10, and FRZB mRNAs were upregulated in HBO1-OE HCC-1 cells (Fig. S1C). β-catenin protein levels were increased as well (Fig. 5B).
These results together showed that ectopic overexpression of HBO1 promoted HCC cell proliferation, migration, and invasion, further supporting the role of HBO1 in HCC cell progression.

HBO1 KO inhibits HCC xenograft growth in SCID mice
At last, we tested the potential effect of HBO1 on HCC cell growth in vivo. Primary HCC1 cells, with the CRISPR/ Cas9 HBO1-KO construct ("KO-HBO1") or the empty vector ("Cas9-C vector") (see Fig. 4), were inoculated via s. c. injection to SCID mice. Recordings were started when the volume of each xenograft tumor close to 100 mm 3 ("Day-0"). As demonstrated, KO-HBO1 HCC1 xenografts grew significantly slower than the control xenografts (Fig.  6A). Volumes of KO-HBO1 HCC1 xenografts were significantly lower than those of control tumors (Fig. 6A). When calculating the estimated daily tumor growth via the formula: (Tumor volume at Day-35 − Tumor volume at Day-0)/35, we show that HCC1 xenograft growth in vivo was significantly inhibited with HBO1 KO (Fig. 6B). At Day-35 all tumors were isolated and weighted individually. Results demonstrated that KO-HBO1 HCC1 xenografts weighted much lower than Cas9-C vector control xenografts (Fig. 6C). On the contrary, the mice body weights were not significantly different between the two groups (Fig. 6D). The xenograft tumor tissues were also analyzed through qPCR (Fig. 6E) and Western blotting (Fig. 6F), and results confirmed HBO1 depletion in the KO-HBO1 HCC1 xenograft tissues (Fig. 6E, F). Collectively, these results demonstrated that HBO1 KO inhibited HCC xenograft growth in SCID mice.

Discussion
Recent studies have reported gene amplification and protein overexpression of HBO1 in human malignancies, which is linked to tumorigenesis and cancer progression. Studies have shown that HBO1 overexpression promoted bladder and breast cancer cell proliferation and tumorigenesis 16,23,46 . In pancreatic cancer cells, HBO1 expression contributed to gemcitabine resistance 47 . Wang et al., reported that HBO1 overexpression is associated with poor prognosis in gastric cancer 48 . A very recent study has demonstrated that HBO1 is required for maintenance of leukemia stem cells (LSC) in acute myeloid leukemia (AML) 21 . Interestingly, Kahali et al., demonstrated that HBO1 activity can induce the expression of anti-cancer genes such as Brahma 49 . It is therefore important to examine HBO1 expression and its functions in specific types of cancers with different genetic backgrounds 10 .
Because of the direct interaction between HBO1 and DNA pre-replication complex proteins, HBO1 is proposed to be required for DNA replication and cell proliferation 50 . However, Kueh et al., have suggested that cancer cell lines with an established transcription profile were relatively insensitive to HBO1 depletion, and did express genes required for cell proliferation 25 . We therefore tested HBO1's function mainly in primary HCC cells.
The results of the present study implied that HBO1 is an important oncogenic gene and therapeutic target of HCC. HBO1 mRNA and protein expression is elevated in human HCC tissues as well as in established and primary human HCC cells. While its expression is low in liver tissues and hepatocytes. In HepG2 and primary human HCC cells, HBO1 silencing, by targeted shRNA, potently inhibited cell viability, proliferation, migration and invasion, while provoking apoptosis. Additionally, CRISPR/ Cas9-induced HBO1 KO inhibited HCC cell progression and induced apoptosis activation. Conversely, forced overexpression of HBO1, through a lentiviral construct, promoted HCC cell proliferation, migration, and invasion. In vivo, the growth of xenografts bearing HBO1-KO HCC cells was largely inhibited in SCID mice. These results suggest that targeting HBO1 could be a novel and valuable strategy to inhibit HCC growth.
Studies have proposed different mechanisms of HBO1induced cancer progression. Quintela et al., reported that Primary HCC1 cells, bearing CRISPR/Cas9 HBO1-KO construct ("KO-HBO1") or the empty vector ("Cas9-C vector"), were s.c. inoculated to the SCID mice (with 11 mice per group). Recordings were started when the volume of each tumor close to 100 mm 3 ("Day-0"). Tumor volumes (A) and mice body weights (D) were recorded every seven days for a total of 35 days; Estimated daily tumor growth was calculated as described (B); At Day-35 all tumors were isolated and weighted individually (C). The tumor lysates were subjected to qPCR (E) and Western blotting (F) assays. Error bars indicate mean ± standard deviation (SD, n = 5). *P < 0.05 vs. "Cas9-C vector" tumors.
HBO1 directly acetylated histone H4 to promote expression of a key oncogene YAP1, required for mechanotransduction and membrane elasticity in ovarian cancer cells 19 . Chen et al., found that HBO1 can activate Wnt/ β-catenin signaling pathway to promote bladder cancer cell progression 23 . Iizuka and colleagues demonstrated that HBO1 destabilized estrogen receptor α by ubiquitination to promote breast cancer cell growth 51 . Future studies will be needed to understand the underlying mechanisms of HBO1 in promoting HCC cell progression.
The majority of HCC are in Asia-Pacific counties, yet rising incidence has been reported in the Western World, possibly due to increasing non-alcoholic fatty liver disease (NAFLD) 3,4 . For clinical practices, current therapeutic options for HCC include liver resection, immunotherapy, and chemotherapy drugs 3,4 . For advanced HCC patients, a multitarget kinase inhibitor sorafenib is possibly the only available systemic therapy, which can slightly increase the survival of certain unresectable HCC patients 52,53 . Since the approve of sorafenib, many other targeted therapies, including sunitinib, tivantinib, brivanib, erlotinib, linifanib, and bevacizumab have been tested, but showing no meaningful improvement in treatment of HCC 54 . Song et al., found that polo-like kinase 1 (Plk1)-induced phosphorylation of HBO1 transcriptionally increased expression of cFos and multidrug resistance 1 (MDR1), essential for gemcitabine's resistance in pancreatic cancer cells 47 . Future studies will be needed to explore the potential role of HBO1 in overwhelming chemoresistance in HCC 7,55,56 . These results suggest HBO1 overexpression is important for HCC cell progression in vitro and in vivo. It could be a promising oncogenic gene and therapeutic target of HCC.