Cyclin B2 overexpression promotes tumour growth by regulating jagged 1 in hepatocellular carcinoma

Clinical specimens and cell lines

From January 2018 to September 2020, tissues were collected from 82 patients with HCC and 20 patients with corresponding paracancerous lesions at Haikou Hospital Affiliated to Xiangya Medical College of Central South University. All patients were diagnosed to have HCC by pathological examination. The clinical and pathological data of 82 patients with HCC were collected. All patients had not received preoperative treatment such as hepatic arterial chemoembolization, local radiofrequency ablation, targeted therapy, chemotherapy or radiotherapy and were pathologically confirmed to have HCC after surgery. Patients with HCC together with other neoplasms or with positive tumour margins were excluded. The study was approved by the ethics committee of Haikou Hospital Affiliated to Xiangya Medical College of Central South University.

The human HCC cell line (Huh-7) used in this study was obtained from the Cell Bank of the Chinese Academy of Sciences in Shanghai, China. Huh-7 cells were cultured in Dulbecco’s Modified Eagle’s Medium (Gibco, Carlsbad, CA, USA).

Cell transfection

CCNB2 small interfering ribonucleic acid (siRNA) (RNAi:5ʹ-CAAGAATGTGGTGAAAGTA-3ʹ) and negative lentivirus were obtained from Shanghai GeneChem Co., Ltd., (China). The experimental groups were as follows: KD group: Huh-7 cells infected with CCNB2-siRNA lentivirus; NC group: Huh-7 cells infected with CON077 negative control virus; and CON group: Uninfected empty Huh-7 cells.

The expression of the CCNB2 gene was determined by fluorescence microscopy approximately 72 h after infection. The fluorescence rate indicated the positive infection rate. The cells were stable, and the infection efficiency reached that of the standard group, which could be used for later experimental research.

Quantitative real-time PCR (qPCR)

TRIzol reagent (Shanghai Pufei Biotech Co., Ltd., China) was used to extract total R-NA, and the Bulge-Loop™ miRNA qPCR Primer Set (Guangzhou Ruibo Biotechnology Co., Ltd., China) was used for reverse transcription in accordance with the manufacturer’s protocol. qRT-PCR was performed on a real-time PCR instrument (Roche, Switzerland) with SYBR Master Mix (TaKaRa, Dalian, China). ACTB was used as an internal control. The primer information is as follows: CCNB2-forward primer: 5ʹ-CAACCCACCAAAACAACA-3ʹ; CCNB2-reverse primer: 5ʹ-AGAGCAAGGCATCAGAAA-3ʹ; ACTB-forward primer: 5ʹ-GCGTGACATTAAGGAGAAGC-3ʹ; ACTB-reverse primer: 5ʹ-CCACGTCACACTTCATGATGG-3ʹ.

Nude mice tumorigenesis experiment

The animal experiment was approved by the ethics committee of Haikou Hospital Affiliated to Xiangya Medical College of Central South University. All animal experiments were conducted in accordance to ARRIVE guidelines for animal experiments. Twenty nude mice were randomly distributed into two groups. The experimental groups were as follows: (1) NC group (control group): mice (numbers 1–10) were injected subcutaneously with Huh-7 cells without CCNB2 knockdown and (2) KD group (experimental group): mice (numbers 11–20) were injected subcutaneously with Huh-7 cells with CCNB2 knockdown. Tumour development was observed over the next 4 weeks, and tumour size and the weight of mice were measured every 5 days. Tumour volume was calculated using the formula: 0.5 × (length × width²) [9]. Based on actual tumour formation and the guidelines of the animal ethics committee, the tumour-bearing nude mice were euthanised, and the tumours were entirely removed and weighed.

Tandem mass tag marker quantification

SDT lysate (Shanghai GeneChem Co., Ltd.) was added to the sample for protein extraction. Next, FASP enzymatic hydrolysis was performed for the extracted protein after SDS-PAGE. Tandem mass tag (TMT) labelling was performed according to the instructions provided in the TMT labelling kit (Thermo Fisher Scientific, USA). The labelled peptides of each group were mixed, and high-pH reversed-phase fractionation was performed using an HPLC system (Thermo Fisher Scientific). The samples were eluted with the nanoscale flow rate EASY-nLC system (Thermo Fisher Scientific), and the chromatographically separated samples were analysed with a Q-Exactive Plus mass spectrometer (Thermo Fisher Scientific). Mascot 2.6 (Matrix Science, UK) and Proteome Discoverer 2.2 (Thermo Fisher Scientific) were used for quantitative analysis.

Bioinformatics analysis

Blast2GO [10] was used for Gene Ontology(GO) annotation, and KOALA (KEGG Orthology And Links Annotation) [11] software was used for Kyoto Encyclopedia of Genes and Genomes(KEGG) annotation. Subsequently, GO and KEGG enrichment analysis was performed. The software WoLF PSORT (https://wolfpsort.hgc.jp/) [12] was used for performing projection analysis of different proteins. The Interpro database was used to annotate the functional domains of differential proteins. The expression levels of the samples were graded by Matplotlib software [13], and the hierarchical heat map of clustering was established. An IPA pathway analysis (https://www.ingenuity.com) was conducted on differential proteins by using keywords such as “proliferation,” “invasion,” “apoptosis,” “cell cycle progression,” and “tumour cell immune response”.

Parallel reaction monitoring targeted quantification

The extracted proteins were analysed by mass spectrometry. Parallel reaction monitoring (PRM) master ion screening was performed using Proteome Discoverer 2.1 software (Thermo Fisher Scientific) and Mascot 2.6 server. PRM was analysed by the EASY-nLC system (Thermo Fisher Scientific) and a Q-Exactive mass spectrometer (Thermo Fisher Scientific).

Western blot (WB)

Total protein extraction was performed using RIPA lysate (Beyotime, Shanghai, China), and SDS-PAGE was then performed. The protein was transferred to a PVDF membrane using an electrical transfer device (Shanghai TANON Technology Co., Ltd., China). After blocking the membrane, it was incubated with primary antibody and secondary antibody successively. The following primary antibodies were used: rabbit anti-human CCNB2 antibody (Abcam, USA) and rabbit anti-human JAG1 antibody (CST, USA). X-ray imaging was performed using the CST 20X Lumiglo® Reagent and 20X spectroscopy #7003 kit (CST). The relative optical densities of all bands were quantified by GelPro Analyzer (Media Cybernetics, Silver Spring, MD, USA). All results were normalized to the internal standard (β-actin) [14].

Cell function experiment

The experimental groups were as follows: KD+OE group: Huh-7 cells were infected by LV-JAG1 and LV-CCNB2-RNAi lentiviruses; OE+NC group: Huh-7 cells were infected with LV-JAG1 and CON077 negative control virus; KD+NC group: Huh-7 cells were infected with LV-CCNB2-RNAi and Con382 negative control virus; and NC+NC group: Huh-7 cells were infected with CON382 negative control virus and CON077 negative control virus.

The CCK-8 kit (Dojindo Laboratories, Japan) was used to detect cell viability. The cells were seeded onto 96-well plates (Corning, USA) at the density of 2000 cells/well. Next, 10 μL CCK-8 reagent was added before the termination of cell culture. After 4 h of incubation, the OD value was measured at 450 nm by using a microplate analyser (Thermo Fisher Scientific).

The appropriate numbers of chambers were placed in a new 24-well plate (Corning, USA). In the upper chamber, a serum-free cell suspension of 10⁵ cells/well was added, while in the lower chamber, 600 μL of 30% FBS medium (Ausbian, Australia) was added, and the plate was incubated for 24 h at 37° C. All compartments were fixed in a 4% paraformaldehyde fixation solution for 30 min. The cells were stained with crystal violet solution for 3 min and then photographed and counted using an inverted microscope (Shanghai Caikon Optical Instrument Co., Ltd., China).

Immunohistochemistry (IHC)

The samples were dewaxed and hydrated with xylene and ethanol successively. An antigen repair solution was used to repair antigen in the Lab Vision™ PT Module. After blocking, the samples were incubated successively with primary antibody and secondary antibody. The rabbit anti-human JAG1 antibody was used as the primary antibody. After colour development with the DAB chromogenic solution, counterstaining with haematoxylin was performed, after which the sections were dehydrated with ethanol and xylene and sealed with resin.

All specimens were histologically reviewed by two independent pathologists. A positive immune reaction was indicated by brown-stained cells [14]. The percentage of positive cells was scored as follows: 0, 0% positive cells; 1, 1%–25% positive cells; 2, 26%–50% positive cells; 3, 51–75% positive cells; 4, 76%–100% positive cells. The staining intensity was scored as follows: 0, negative; 1, weak; 2, medium; and 3, strong. The final expression score was calculated by multiplying the percentage of positive cells by the fraction of staining intensity on the scale of 0 to 12. Samples with a score of > 6 were considered to have high expression, while those with a score ≤ 6 were considered to have low expression [14].

Statistical analysis

Statistical analysis was conducted with GraphPad Prism 8.0 (GraphPad Prism, La Jolla, CA, USA) and SPSS 25.0 (SPSS, Inc. Chicago, IL, USA). Quantitative data with normal distribution were expressed as mean ± standard deviation (x±s), and comparison between two groups was performed using two independent sample t-test. Quantitative data with skewed distribution were expressed as median (first quartile, third quartile) [M (QL, QU)], and comparison between two groups was performed using the Mann-Whitney U test. The relationship between JAG1 expression and clinicopathological characteristics of patients with HCC was determined using the chi-square test. A p value of < 0.05 was considered to be statistically significant.

CCNB2 enhances tumour growth in nude mice

As shown in Figure 1, the expression level of CCNB2 was considerably reduced in Huh-7 cells transfected with CCNB2 siRNA as determined by qPCR (KD 0.121 ± 0.036 vs. NC 1.012 ± 0.190; p < 0.01), and the interference efficiency of the KD group was 87.9%.

Huh-7 cells with CCNB2 knockdown and Huh-7 cells without CCNB2 knockdown were injected subcutaneously into the experimental group (KD group) and the control group (NC group), respectively, to observe tumour formation. After 25 days, eight nude mice in the experimental group developed tumours, and the tumour formation rate of this group was 80%. All nude mice in the control group developed tumours, and the tumour formation rate of this group was 100% (Figure 2A). These results suggest that CCNB2 knockdown could reduce the tumour formation rate in nude mice and may inhibit the occurrence of HCC.

After the inoculation of Huh-7 cells in nude mice, the tumour volume of the NC group at five time points (6, 11, 16, 21 and 26 days) was 186.58 (99.91-240.34) [M(P25-P75)] mm³, 265.74 (139.55-326.00) [M(P25-P75)] mm³, 431.78 (216.69-528.06) [M(P25-P75)] mm³, 549.67 (291.7-719.26) [M(P25-P75)] mm³ and 1201.39 (619.57-1363.32) [M(P25-P75)]mm³, respectively, while the tumour volume of the KD group was 6.21 (0.00-14.45)[M(P25-P75] mm³, 13.47 (0.00-33.22) [M(P25-P75)] mm³, 28.03 (3.91-69.06) [M(P25-P75)] mm³, 51.5(4.98-105.61) [M(P25-P75)] mm³, 215.11(45.47-383.62) [M (P25-P75)] mm³, respectively. A significant difference in tumour volume was observed between the two groups (p < 0.05, Figure 2B). This finding suggested that the tumour volume was reduced after knockdown of CCNB2. The tumour growth was inhibited following CCNB2 knockdown, which indicates that CCNB2 may activate HCC progression.

The tumour weight of the KD group was 0.27 (0.06–0.49) [M (P25-P75)] g and that of the NC group was 1.33 (0.69-1.74)[M (P25-P75)] g (p < 0.05, Figure 2D). Tumour weight decreased after CCNB2 was knockdown, which suggested that tumour proliferation ability decreased following CCNB2 knockdown. Thus, CCNB2 may be associated with promoting the development of HCC.

Proteomic results

A total of 7271 proteins were identified by TMT mass spectrometry. Bioinformatics analysis was used to screen proteins with expression fold change of > 1.2 (up-regulated or down-regulated) and p < 0.05 (t-test) as differentially expressed proteins. There were 130 differentially expressed proteins between the KD group and NC group, including 78 up-regulated proteins and 52 down-regulated proteins. A volcano plot was mapped to show differential proteins (Figure 3A).

GO functional annotations and enrichment analysis showed that the cell components were primarily enriched in cell surface, extracellular space, integrin alpha2-beta1 complex, anchored component of cell membrane and glial cell projection. Molecular functions were primarily involved in the structural constituent of the extracellular matrix conferring tensile strength, long-chain fatty acid transporter activity, collagen binding involved in cell-matrix adhesion, L-lysine transmembrane transporter activity and heparin binding. The biological processes were significantly enriched in negative regulation of lamellipodium assembly, cellular response to heparin, integrin-mediated signalling pathway, axon extension involved in axon guidance and aortic valve morphogenesis (Figure 3B).

KEGG pathway annotation and enrichment analysis showed that the differential proteins were primarily enriched in various metabolic pathways, especially in the biosynthesis of amino acids, cell adhesion molecules, protein digestion and absorption, PPAR signalling pathway and ABC transporters (Figure 3C).

The results of subcellular localisation analysis showed that the majority of the differential proteins were located in the plasma membrane (27.7%), followed by the cytosol (24.6%) and nucleus (21.5%), and the remaining proteins were localised in the extracellular matrix, mitochondria and endoplasmic reticulum (Figure 3D).

The results of domain annotation and enrichment analysis indicated that the differential protein domains were enriched in Von Willebrand factor type A domain signature, laminin G domain, collagen triple helix repeat, amino acid/polyamine transporter I and concanavalin A-like lectin (Figure 3E).

The cluster analysis results showed that the differential proteins could differentiate between samples with and without CCNB2 knockdown (Figure 3F).

We selected 30 proteins from the differential proteins for PRM analysis. In the PRM analysis, proteins with the expression fold change of > 1.2 (up-regulated or down-regulated) and p < 0.01 (t-test) were considered as significant differential proteins. The KD/NC value of RAB5A was > 1, suggesting that RAB5A was up-regulated when CCNB2 was knocked down (Table 1). The KD/NC values of PPCS and JAG1 were < 1, indicating that when CCNB2 was knocked down, PPCS and JAG1 were down-regulated. The results of IPA for significant differential proteins showed that JAG1 was correlated with the proliferation and apoptosis of HCC cells. Based on the above results, JAG1 could probably be the downstream target of CCNB2, and it was selected for further experimental confirmation.

CCNB2 knockdown inhibits JAG1 expression

Two nude mouse tumour tissues were randomly selected from the experimental group and control group for the WB assay. The results showed that knockdown of CCNB2 inhibited the expression of the JAG1 protein as compared with that in the control group (Figure 4).

CCNB2 regulates JAG1 to promote tumour cell proliferation and migration

We transfected LV-JAG1 into the Huh-7 cell line to construct a JAG1 functionally overexpressing cell line (OE group). The transfection efficiency was confirmed before the functional experiment. qPCR showed that the expression abundance of the JAG1 gene in the OE group was 3.239 times higher than that in the NC group, and the expression level of the JAG1 gene was significantly increased (p < 0.01, Figure 5A). The WB assay indicated that as compared with the NC group, the expression of JAG1 in the OE group was significantly increased at the protein level (Figure 5B). We then performed cell infection according to the experimental groups, and the experimental results showed that cell transfection was successful.

The CCK-8 assay showed that the proliferation function of Huh-7 cells in the OE+NC group (DAY4: 6.85 ± 0.09, DAY5: 7.49 ± 0.03) was enhanced as compared with that in the NC+NC group (DAY4: 6.07 ± 0.05, DAY5: 6.85 ± 0.02) on day 4 and day 5 after Huh-7 cell infection (p < 0.05), respectively. The proliferation ability of the KD+NC group (DAY4: 2.28 ± 0.09, DAY5: 2.14 ± 0.03) was weakened as compared with that of the NC+NC group (p < 0.05), while the proliferation ability of the KD+OE group (DAY4: 2.73 ± 0.04, DAY5: 2.83 ± 0.08) was partially restored as compared to that of the KD+NC group (p < 0.05, Figure 5C). The results of the CCK-8 assay indicated that CCNB2 silencing restrained the proliferation of HCC cells, while JAG1 overexpression partially restored proliferation inhibition induced by CCNB2 silencing. Functional recovery experiments further confirmed that CCNB2 promoted cell proliferation by increasing JAG1 expression.

The Transwell assay showed that the number of migratory cells in the OE+NC group (149 ± 2.91) was higher than that in the NC+NC group (131 ± 4.61, p < 0.05). The number of migratory cells in the KD+NC group (29 ± 4.50) was lower than that in the NC+NC group (p < 0.05), and the number of migratory cells in the KD+OE group (89 ± 3.04) was partially restored as compared with that in the KD+NC group (p < 0.05, Figure 5D, 5E). The Transwell assay showed that the migration function of HCC cells was inhibited by CCNB2 silencing, and the overexpression of JAG1 partially restored the inhibition of migration ability following CCNB2 silencing. Thus, the Transwell assay further confirmed that CCNB2 promotes cell migration by increasing JAG1 expression.

JAG1 expression is up-regulated in HCC tissues

IHC revealed that among the 82 patients with HCC, 56 patients (68.29%) showed high expression of JAG1, while 26 patients (31.71%) showed low expression of JAG1. JAG1 expression was low in 17 patients (85%) and high in three patients (15%) with paracancerous lesions (Figure 6A). The expression of JAG1 was higher in HCC tissues (6.78 ± 3.12) than in the adjacent tissues (3.05 ± 2.67), and the difference was statistically significant (p = 0.000, Figure 6B). Studies on the correlation between the expression of JAG1 and clinicopathological indicators showed that the expression of JAG1 was significantly correlated with tumour size and number, and tumour thrombus formation, but not associated with gender, age, hepatitis B virus infection, Child-Pugh class, alpha fetoprotein level and tumour stage and grade (Table 2).