Targeting epigenetics in cancer has emerged as a promising anticancer strategy. p300/CBP is a central regulator of epigenetics and plays an important role in hepatocellular carcinoma (HCC) progression. Tumor-associated metabolic alterations contribute to the establishment and maintenance of the tumorigenic state. In this study, we used a novel p300 inhibitor, B029-2, to investigate the effect of targeting p300/CBP in HCC and tumor metabolism. p300/CBP–mediated acetylation of H3K18 and H3K27 increased in HCC tissues compared with surrounding noncancerous tissues. Conversely, treatment with B029-2 specifically decreased H3K18Ac and H3K27Ac and displayed significant antitumor effects in HCC cells in vitro and in vivo. Importantly, ATAC-seq and RNA-seq integrated analysis revealed that B029-2 disturbed metabolic reprogramming in HCC cells. Moreover, B029-2 decreased glycolytic function and nucleotide synthesis in Huh7 cells by reducing H3K18Ac and H3K27Ac levels at the promoter regions of amino acid metabolism and nucleotide synthesis enzyme genes, including PSPH, PSAT1, ALDH18A1, TALDO1, ATIC, and DTYMK. Overexpression of PSPH and DTYMK partially reversed the inhibitory effect of B029-2 on HCC cells. These findings suggested that p300/CBP epigenetically regulates the expression of glycolysis-related metabolic enzymes through modulation of histone acetylation in HCC and highlights the value of targeting the histone acetyltransferase activity of p300/CBP for HCC therapy.
This study demonstrates p300/CBP as a critical epigenetic regulator of glycolysis-related metabolic enzymes in HCC and identifies the p300/CBP inhibitor B029-2 as a potential therapeutic strategy in this disease.
Liver cancer was the sixth most common cancer and the fourth leading cause of cancer-related death worldwide in 2018, and there are approximately 841,000 new cases and 782,000 deaths annually (1). Hepatocellular carcinoma (HCC) is the main type of primary liver cancer, accounting for 75%–85% of cases. Although currently available treatments (2, 3) have significantly improved the survival of patients with HCC, the prognosis of advanced HCC remains poor. Therefore, new treatment methods for HCC are needed to improve overall survival.
Histone acetylation has been recognized as an important posttranslational modification (PTM) of core nucleosomal histones that changes access to chromatin to allow gene transcription, DNA replication, and repair (4). E1A-binding protein (p300) and its paralog CREB-binding protein (CBP) are two prominent members of the histone acetyltransferase (HAT) and transcriptional coactivator families (5), which play a central role in transcriptional activation through acetylating histone H3K18/K27 (6). The HAT activity of p300/CBP is often under aberrant control in human disease, particularly in cancer (7). Diwakar and colleagues reported that p300/CBP interferes with oncogene-driven transcriptional programmes in leukemogenesis through interaction with c-Myb (8). p300/CBP is also required to maintain the growth of castration-resistant prostate cancer (9). Targeting p300/CBP resulted in antitumor effects in several hematologic malignancies, prostate cancer, and colorectal cancer in vitro (10, 11). Recent studies found that high expression of p300 in HCC was correlated with enhanced vascular invasion, intrahepatic metastasis, and shortened survival (12–15), indicating the therapeutic potential of p300/CBP inhibitors in HCC.
It is known that tumor-associated metabolic alterations contribute to the establishment and maintenance of the tumorigenic state. Cancers utilize aerobic glycolysis to accommodate rapid cell growth and proliferation in spite of high-oxygen conditions (16). During aerobic glycolysis, cancer cells metabolize glucose to lactic acid to produce ATP and generate metabolic intermediates for the synthesis of lipids, proteins, and nucleic acids. Many oncogenes, such as KRAS (17) and c-Myc (18), alter the metabolism of cancer cells in several ways, including increased glucose uptake and glycolysis even in the presence of abundant oxygen. However, the role that p300/CBP plays in shaping metabolic patterns in cancer cells remains undefined.
Previous studies have demonstrated the antitumor effect of A-485, a selective p300 inhibitor, in lineage-specific tumor types, including several hematologic malignancies and androgen receptor–positive prostate cancer (11). Our recent study reported the discovery of novel, highly selective, potent small-molecule inhibitors of p300/CBP HAT derived from A-485 through the artificial intelligence–assisted drug discovery pipeline and further optimization (19). According to the properties of the reported compounds, we acquired another compound, B029-2, with potent activity through further systematic optimization. In this study, we investigated the effect of B029-2 on HCC malignancy. Our data revealed that inhibition of p300/CBP in HCC significantly decreased the proliferation and metastatic ability of HCC cells. Moreover, we found that p300/CBP regulated the alteration of cancer metabolism and the transcription of the enzymes in amino acid metabolism and nucleotide synthesis by acetylating histone H3K18/K27.
Materials and Methods
Human liver tissue samples were obtained from patients who received surgical resection at the Eastern Hepatobiliary Surgery Hospital (Shanghai, China). Written informed consent was issued by all patients. HCC tissues with typical macroscopic features were collected from tumor nodules, which were examined using hematoxylin and eosin (H&E) staining to confirm the diagnosis. Paired adjacent noncancerous tissues without histopathologically identified tumor cells were collected from 5 cm from the tumor border. All human experiments were approved by the Ethics Committee of the Second Military Medical University (Shanghai, China).
Cell lines and cell cultures
Human HCC cell lines (Huh7, Hep3B, HepG2, MHCC-H, PLC) were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). We routinely tested cell lines for Mycoplasma contamination using the MycoAlert Detection Kit (Lonza) and authenticated them by short-tandem repeat analysis. Primary human hepatocytes were obtained from Xeno Tech and cultured in William's E medium (Xeno Tech) containing 50 μg/mL penicillin–streptomycin and 6 μg/mL insulin. Huh7, HepG2, MHCC-H, and PLC cells were cultured in DMEM containing 10% heat-inactivated FBS. Hep3B cells were cultured in Eagle's minimum essential medium (MEM) supplemented with 10% FBS and nonessential amino acids (NEAA).
The compound B029-2 was designed and synthesized by Drug Discovery and Design Center, Shanghai Institute of Materia Medica (Shanghai, China). Compound A-485 was obtained from the Drug Discovery and Design Center, Shanghai Institute of Materia Medica (Shanghai, China). Antibodies against H3K27ac (#8173), H3K18ac (#13998), H3K9ac (#9649), H3K14ac (#7627), H3 (#4499), p300 (#54062), and BRD4 (#13440) were purchased from Cell Signaling Technology; antibodies against GAPDH (KGAA002) were purchased from KeyGEN BioTECH; antibodies against PSPH (#14513–1-AP) and DTYMK (#15360–1-AP) were purchased from ProteinTech; and antibodies against p300 (#sc-32244) were purchased from Santa Cruz Biotechnology.
Radioactive acetyltransferase activity assay
The radioactivity assay was performed in 50 mmol/L Tris-HCl, pH 7.5, and 0.01% Tween-20 buffer. p300 proteins (0.2 nmol/L) were preincubated with a range of compound concentrations for 15 minutes at room temperature before adding 600 nmol/L H3 (1–21) substrate and 250 μmol/L [3H] Ac-CoA. After 60 minutes of incubation at 37°C, the reaction systems were transferred to a MultiScreen HTS Filter Plate (Millipore), and the radioactivity was determined by liquid scintillation counting (MicroBeta, PerkinElmer).
tRNA was isolated from cells or tissues following the standard TRIzol (Takara) protocol. First-strand cDNA of cells was synthesized from tRNA using RT Master Mix (Takara). First-strand cDNA of tissues was synthesized from tRNA using the SuperScript III First-Strand Synthesis System (Invitrogen). The expression of various mRNAs was detected using SYBR Green–based RT-PCR performed on the ABI StepOne Real-time PCR Detection System (Life Technologies). The mRNA levels were normalized to those of β-actin mRNA. At least three independent experiments were performed using each condition. Primer sequences are listed in Supplementary Table S1.
Western blotting analysis
Proteins were isolated from whole cells and tissues with lysis buffer (125 mmol/L Tris-HCl pH 6.8, 25% glycerol, 5% SDS) supplemented with protease inhibitor (Roche), separated by SDS-PAGE, and then transferred onto nitrocellulose membranes (HAHY00010, Millipore). The membranes were blocked in PBST containing 5% skim milk and then incubated with primary antibody overnight. After incubation with a secondary antibody (donkey-anti-mouse or donkey-anti-rabbit, IRDye 700 or IRDye 800, respectively) for 1 hour, signals were quantified using an Odyssey Infrared Imaging System (LI-COR) at 700 nm or 800 nm.
Flow cytometry analysis
Cell-cycle distribution was identified by flow cytometry analysis. Huh7 and Hep3B cells were seeded in 6-well plates, starved for 12 hours, and then treated with DMSO (0.2‰) or B029-2 (0.1 μmol/L, 0.2 μmol/L) for 72 hours. For cell-cycle distribution analysis, following the treatments, the cells were harvested by 0.25% trypsin and washed twice with PBS. Approximately 1.0 × 106 cells were incubated with 100 μL 0.1% FBS–PBS containing 2 μL 10% Triton X-100, 10 μg propidium iodide (PI), and 40 μg RNase A for 30 minutes in the dark. The fixed cells were then diluted with PBS to 600 μL, and the cell-cycle distribution was assessed by flow cytometry with FACS (Navios, Beckman Coulter). At least three independent experiments were performed for each condition.
Cell proliferation assay, migration, and invasion assay
Huh7 and Hep3B cells were plated in 96-well plates at 3 × 103 cells/well and cultured in medium containing B029-2. After transfection overnight, Huh7 and Hep3B cells were plated in 96-well plates at 2 × 103 cells/well. The number of metabolically active cells was determined using the Cell Counting Kit-8 (CCK-8, Dojindo) every one or two days for approximately 1 week. In vitro migration and invasion assays were performed using Transwell Chambers (BD Bioscience) without or with Matrigel, according to the manufacturer's instructions. In brief, 3 × 104 HCC cells in serum-free medium containing B029-2 were seeded in the upper chamber, and medium supplemented with FBS and the same concentration of B029-2 was added into the lower chamber. A total of 4 × 104 HCC cells were transfected overnight in serum-free medium in the upper chamber, and medium supplemented with FBS was added into the lower chamber. After incubation for 48/72 hours at 37°C, cells remaining on the upper membrane were removed with cotton swabs. Cells on the lower surface of the membrane were fixed and stained with 0.1% crystal violet and 20% methanol. Five fields of cells on the lower membrane were photographed and counted to estimate cell density. Image analysis software (Image-Pro Plus 6.0, Media Cybernetics) was used to measure the stained area. At least two independent experiments were performed for each condition.
Six thousand cells per condition were plated in 96-well plates in triplicate. Regular media was replaced by 2% FBS supplemented media with increasing concentrations of B029–2, A-485, B026 (0 μmol/L, 0.005 μmol/L, 0.015 μmol/L, 0.046 μmol/L, 0.14 μmol/L, 0.41 μmol/L, 1.23 μmol/L, 3.7 μmol/L, 11.11 μmol/L, 33.33 μmol/L and 100 μmol/L) 24 hours later. The number of metabolically active cells was determined using CCK-8 (Dojindo) after 48 hours of treatment. IC50 was defined as the concentration resulting in a 50% reduction in absorbance.
The pharmacokinetic properties of B029–2 were evaluated in Male ICR (CD-1) mice. The animals were fasted for 12 hours before administration (p.o., 5 mg/kg) and remained fasting for 2 hours. Aliquots were sampled at 0.25, 0.5, 1, 2, 4, 8, and 24 hours. The plasma sample (20 μL) in a centrifuge tube was added to 200 μL of the precipitant (acetonitrile/MeOH = 1:1). Samples were vortexed for 1 minute and then centrifuged at room temperature (15,000 rpm) for 5 minutes. The supernatant was separated and mixed with an equal volume of water before analysis. Samples were analyzed by HPLC.
For overexpressing PSPH and DTYMK, PSPH and DTYMK were cloned into the lentiviral vector pCDH-CMV-MCS-EF1-copGFP (System Biosciences) to generate the lentivirus LV-PSPH and LV-DTYMK, respectively. The lentiviral vectors were transfected into subconfluent HEK 293T cells along with the packaging plasmid psPAX2 (Addgene) and envelope plasmid pMD2.G (Addgene) to produce lentiviral particles. All vectors were verified by sequencing.
Human HCC xenograft model
Male athymic BALB/c nude mice (5 weeks old) were purchased and maintained under specific pathogen-free conditions with a 12-hour on/off light cycle. To generate tumor xenografts, 1.0 × 106 Huh7 cells were inoculated subcutaneously in the right flank. Tumors were measured with Vernier callipers in two dimensions, and the volumes were calculated using the following equation: volume = length × (width)2 × π/6. When tumors reached an average volume of approximately 60 mm3, the mice were randomly divided into two groups (6 animals/group) and injected with DMSO (100 μL) or 5 mg/kg B029–2 (0.1 mg in 100 μL DMSO) intratumorally every 2 days for 1 week. The mice were sacrificed, and the tumors were excised and weighed after the fourth injection. For rescue experiments, 1.0 × 106 Huh7 cells infected with LV-Vector, LV-PSPH, or LV-DTYMK for 7 days were inoculated subcutaneously in the right flank (3 groups, 16 animals/group). At 14th day of inoculation, each group of mice was injected with DMSO (100 μL) or 2.5 mg/kg B029-2 (0.05 mg in 100 μL DMSO), respectively. RT-PCR was used to verify PSPH and DTYMK overexpression. All animal experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee at the Second Military Medical University (Shanghai, China).
IHC staining was performed on 4-μm thick paraffin sections of tissues fixed in buffered formalin. Sections were deparaffinized in xylene and rehydrated in graded alcohols. Endogenous peroxidase was blocked by 3% H2O2 followed by antigen retrieval. Slides were incubated with primary antibodies overnight at 4°C and incubated with a secondary antibody at room temperature for 30 minutes. The following primary antibodies were used: Ki67 (#DR3005, TRGET) and GS (#sc-166043, Santa Cruz Biotechnology). Staining was developed using an EnVision Detection Rabbit/Mouse Kit (GK500710, GeneTech).
RNA interference and transfection
siRNAs were purchased from GenePharma (Shanghai GenePharma Co.) and were transfected using Lipofectamine 2000 (Invitrogen) reagent in 6-well plates according to the manufacturer's protocols. The target sequences were listed in Supplementary Table S1. After transfection overnight, the cells were subjected to proliferation, migration, and invasion assays. After transfection for 48 hours, the cells were collected, and the specific silencing of p300, PSPH, and DTYMK expressions was assessed using RT-PCR, and Western blotting was performed at 72 hours after transfection.
Chromatin immunoprecipitation assay
Chromatin immunoprecipitation (ChIP) analysis was performed essentially as described previously (20). Immunoprecipitation (IP) was carried out using the following antibodies: H3K27ac (Cell Signaling Technology, #8173), H3K18ac (Cell Signaling Technology, #13998), H3K9ac (Cell Signaling Technology, #9649), and IgG (Cell Signaling Technology, #7074P2). DNA was amplified by RT-PCR and normalized to the input. The RPL6 primer was used as a positive control (21). The primer for a nonspecific site was used as the negative control. Primer sequences are shown in Supplementary Table S1.
Huh7 cells plated in 10 cm dishes were treated with DMSO or 0.2 μmol/L B029-2 for 72 hours, respectively. Then, the cells were washed with PBS three times and lysed with TRIzol at room temperature. Samples from two independent treatments were sequenced using the Illumina HiSeq 2500, and raw reads were aligned to hg38_genecode_genome using the star program. Read counts for each transcript were calculated using htseq-count. We applied the DESeq2 algorithm to filter the differentially expressed genes after the significance and FDR analysis under the following criteria: (i) |log2FC| > 0; (ii) FDR < 0.05. Pathway analysis was used to determine the significant pathways of the differentially expressed genes according to the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. We turned to Fisher exact test to select the significant pathway, and the threshold of significance was defined by the P value and FDR. The DAVID bioinformatics functional annotation tool was used to identify enriched KEGG pathway terms. The significance of fold enrichment was calculated using a BH adjusted P < 0.05. (https://david.ncifcrf.gov/). GSEA was used for cellular pathway analysis. The data are accessible at NCBI-BioProject under the accession number PRJNA613779.
Assay for transposase-accessible chromatin using sequencing
Assay for transposase-accessible chromatin using sequencing (ATAC-seq) was carried out as described previously (22). Briefly, Huh7 cells were treated with 0.2 μmol/L B029-2 or DMSO control for 3 days. Samples from two independent treatments were harvested for ATAC-seq. ATAC-seq reads were mapped to the hg38 reference genome using the bwa program and using macs2 software for peak calling with a cut-off q-value < 0.05. HOMER's findMotifsGenome.pl tool was used for motif analysis. The deeptools tool was used for ATAC-seq read distribution analysis. Peaks were annotated using homer's annotatePeaks.pl. We downloaded the Gene Oncology (GO) annotations from NCBI (http://www.ncbi.nlm.nih.gov/), UniProt (http://www.uniprot.org/), and Gene Ontology (http://www.geneontology.org/) databases. Fisher exact test was applied to identify the significant GO categories, and FDR was used to correct the P values. Pathway analysis was used to determine the significant pathways of the genes according to KEGG database. We used Fisher exact test to determine the significant pathways, and the threshold of significance was defined by the P value and FDR. The data are accessible at NCBI-BioProject under the accession number PRJNA613779.
Nucleotide analysis by HPLC
Adenine (J&K, #LM30T47), uracil (J&K, #L240T20), and thymine (J&K, #LMB0Q20) were purchased for obtaining standard curve. After treatment with B029-2 (0.2 μmol/L) for 72 hours, Huh7 cells were harvested using 0.25% trypsin and washed twice with PBS. Approximately 5.0 × 106 cells were extracted with ice-cold 60% methanol. HPLC analysis was performed on DIONEX UltiMate 3000 with a Waters XSELECT HSS T3 column (3.0 × 100 mm, 2.5 μmol/L) at 35°C. Solvent A was water with 1% FA, and solvent B was methanol. The gradient was as follows: 98% A for 4.5 minutes at 0.3 mL/min, 10% A at 4.5 minutes at 0.3 mL/min, 98% A at 6.6 minutes at 0.3 mL/minutes until 9.1 minutes.
Glycolytic function test
Seahorse glycolytic stress test analyses were performed on a Seahorse XFe96 Analyzer according to manufacturer's instructions (Agilent Technologies). Briefly, the cells were treated with 0.2 μmol/L B029-2 in the culture medium for 72 hours or after transfection with siRNA overnight. Then, 8 × 103 Huh7 cells were plated in a 96-well assay plate the day before the analysis in the complete medium to ensure 80%–90% confluence on the next day, then washed and incubated in a freshly prepared XF assay medium containing 2 mmol/L glutamate. For rescue experiments by overexpression of PSPH and DTYMK, Huh7 cells were treated with 0.05 μmol/L B029-2 for 24 hours and Hep3B cells were treated with 0.1 μmol/L B029-2 for 24 hours before analysis. Extracellular acidification rate (ECAR) was measured under basal conditions and after injection of 10 mmol/L glucose, 1 μmol/L oligomycin, and 50 mmol/L 2-deoxyglucose.
Data analyses were performed with Prism 5 (GraphPad software). For experiments involving only two groups, data were analyzed with Student unpaired t tests. FDR was adjusted using the Benjamini–Hochberg method for RNA sequencing (RNA-seq) data and ATAC-seq data. All measurements are shown as the mean ± SD or mean ± SEM where appropriate. Statistical significance was set at *, P ≤ 0.05; **, P ≤ 0.01; and ***, P ≤ 0.001. P ≤ 0.05 was considered statistically significant. RNA-seq data of patients with liver hepatocellular carcinoma (LIHC; n = 369) from The Cancer Genome Atlas (TCGA) and/or adjacent noncancerous tissues (n = 160) were obtained and downloaded from http://gepia.cancer-pku.cn.
B029–2 potently inhibits p300/CBP activity in HCC cells
We first examined p300/CBP–mediated acetylation of histones in HCC samples. The levels of acetylated H3K27 and H3K18 in HCC tissues were markedly higher than that in surrounding noncancerous tissues (Fig. 1A), indicating the enhanced activity of p300/CPB in HCC. The protein expression of p300 was also increased in HCC tissues compared with the surrounding noncancerous tissues (Supplementary Fig. S1A). In addition, p300 expression in hepatoma cell lines was also increased in comparison with that in primary human hepatocytes (Supplementary Fig. S1B). Moreover, knockdown of p300 significantly reduced the levels of H3K27Ac and H3K18Ac in HCC cells (Supplementary Fig. S1C).
Previously, we reported highly selective, potent small-molecule inhibitors of p300/CBP HAT. Among them, compound B026 (IC50 = 1.8 nmol/L to p300), the most potent p300/CBP HAT inhibitor, showed significant and dose-dependent tumor growth inhibitory ability in an animal model of human cancer (19). On the basis of these results, we obtained another compound, B029-2, through further systematic optimization (Fig. 1B). B029-2, which has a chemical structure similar to compound B026, displayed more potent inhibitory activity against p300, with an IC50 of 0.52 ± 0.17 nmol/L, as detected by a radioactive acetyltransferase activity assay (Fig. 1B).
Moreover, B029–2 inhibited the activity of the CBP with an IC50 value of 11 nmol/L, whereas it exhibited no inhibitory activities against HAT1, GCN5, NatD, p300/CBP–associated factor (PCAF), KAT7, and MOF, with all of the IC50 values greater than 20 μmol/L (Supplementary Fig. S2A). The binding mode of B029-2 is very similar to A-485, which occupied the binding pocket of Ac-CoA and competed for its binding (Supplementary Fig. S2B). Pharmacokinetic property of B029-2 was evaluated in mice after oral administration (Supplementary Fig. S2C). We then evaluated the inhibitory effect of B029-2 on p300/CPB activity in HCC cells. As expected, the levels of acetylated histones H3K27 and H3K18 were decreased in a dose-dependent manner in Huh7 and Hep3B cells treated with B029-2 for 72 hours, while no significant alteration was detected in the acetylation of H3K9 and H3K14 (Fig. 1C). In addition, B029-2 robustly decreased the levels of H3K27Ac and H3K18Ac at 0.2 μmol/L, whereas A-485 only had a mild effect (Supplementary Fig. S2D and S2E). In addition, the IC50 of B029-2 was slightly lower than that of B026 (3.751 μmol/L vs. 10.51 μmol/L in Huh7 cells and 6.572 μmol/L vs. 15.49 μmol/L in Hep3B cells) and much lower than that of A-485 (3.751 μmol/L vs. 37.18 μmol/L in Huh7 cells and 6.572 μmol/L vs. 26 μmol/L in Hep3B cells; Supplementary Fig. S2F). Moreover, B029-2 exhibited low toxicity on primary human hepatocytes (Supplementary Fig. S2G), indicating its translational value in cancer therapy.
B029-2 suppresses the malignant phenotypes of HCC cells
It has been reported that silencing p300 suppresses the proliferation and invasion of HCC cells (15). However, the therapeutic potential of p300/CBP inhibitors in HCC cells has been less explored. Herein, we found that B029-2 inhibited the proliferation of HCC cells in a dose-dependent manner (Fig. 2A). In addition, knockdown of p300 also decreased the proliferation of HCC cells and abolished the effect of B029–2 on cell proliferation (Supplementary Fig. S3A). The transwell assays showed that B029-2 markedly suppressed the migration and invasion of HCC cells (Fig. 2B and C). Previous studies found that p300/CBP increases cell growth by promoting cell-cycle progression (23). B029-2 treatment also significantly induced G0–G1 arrest and decreased the expression of cyclin D1 in HCC cells (Supplementary Fig. S3B and S3C). These data suggest that B029-2 significantly inhibits the malignant phenotypes of HCC cells in vitro.
We then explored the potential therapeutic effect of B029-2 on HCC xenografts in mice. The size and weight of HCC xenografts treated with B029-2 were significantly reduced compared with those treated with the vehicle control (Fig. 2D–F). Regarding amino acid homology between human and mouse p300/CBP proteins, B029-2 is able to inhibit activity of both human and mouse p300/CBP. We found no significant alteration was detected on the body weight of the mice, indicating the low adverse effects of B029-2 (Supplementary Fig. S3D). IHC staining of tumors showed that B029-2 treatment resulted in a marked decrease in Ki67 and glutamine synthetase (GS), a biomarker for HCC diagnosis (Fig. 2G; ref. 24). The levels of H3K27Ac and H3K18Ac also decreased in treated tumors (Fig. 2H), suggesting that B029-2 efficiently inhibits the HAT activity of p300/CBP in HCC xenografts. B029-2 also reduced the expression of cyclin D1 in HCC xenografts (Supplementary Fig. S3E). These results revealed a potent inhibitory effect of B029-2 on HCC in vivo.
Targeting p300/CBP disturbs the metabolic reprogramming of HCC cells
To elucidate the antitumor mechanism of B029-2, RNA-seq analysis was used to detect the alterations in the gene expression profile of Huh7 cells treated with B029-2. We identified 3,297 genes that were downregulated and 2,699 genes that were upregulated in B029-2–treated cells compared with DMSO-treated cells (Supplementary Table S2, FDR < 0.05). All of the upregulated and downregulated genes are repeatedly observed in the duplicate sequencing analysis. KEGG pathway analysis of the differentially expressed genes revealed that the dysregulated genes were involved in ribosome, metabolic pathways, lysine degradation, carbon metabolism, biosynthesis, and metabolism of amino acids (Fig. 3A; Supplementary Table S3). Gene set enrichment analysis (GSEA) showed that downregulated genes were involved in ribosomes (Supplementary Fig. S4A), citrate cycle, fatty acid metabolism, and amino acid metabolism (Fig. 3B), suggesting that p300/CBP regulates essential enzyme gene expression in the metabolic process of HCC cells. The ribosome is a cellular machine for protein synthesis that is essential for sustained growth of cells (25). Strikingly, 101 genes encoding ribosomal proteins were downregulated, whereas none of the ribosomal genes were upregulated in cells treated with B029-2 (Fig. 3A). Downregulation of 10 of these ribosomal genes was validated by RT-PCR in HCC cells treated with B029-2 (Supplementary Fig. S4B). These findings indicate that B029-2 inhibits protein synthesis in HCC cells.
Chromatin accessibility plays a pivotal role in modulating gene expression during many biological and pathologic processes. To further investigate the effect of targeting p300 on chromatin accessibility, we performed ATAC-seq of Huh7 cells treated with B029-2 (0.2 μmol/L) versus DMSO-treated controls. GO analysis of annotated genes associated with the open chromatin regions identified active biological processes, including transport, phosphorylation, and metabolic process (Fig. 3C; Supplementary Table S4). To examine the correlation between depot-specific open chromatin regions from the ATAC-seq analysis with depot-specific gene expression signatures, we integrated the ATAC-seq and RNA-seq datasets and found that 215 upregulated genes and 459 downregulated genes overlapped (Fig. 3D; Supplementary Table S5). Because histone acetylation is mostly involved in gene activation, we focused on the genes downregulated by B029-2. Consistently, KEGG pathway analysis showed that the downregulated genes were mainly involved in glycolysis-related metabolic pathways (amino biosynthesis and metabolism, fatty acid biosynthesis and metabolism, and so on; Fig. 3E; Supplementary Table S6). Seahorse analysis further confirmed that Huh7 cells displayed reduced glycolysis, glycolytic capacity, and glycolytic reserve after B029-2 treatment (Fig. 3F). With HPLC analysis, we confirmed that the levels of adenine, uracil, and thymine were decreased in Huh7 cells upon B029-2 treatment (Fig. 3G; Supplementary Table S7). Together, our data reveal that targeting p300/CBP disturbs the metabolic reprogramming of HCC cells.
Epigenetic regulation of metabolic enzymes in HCC via the histone acetyltransferase activity of p300/CBP
Next, we searched the TCGA database for the prognosis of the 61 downregulated genes in the metabolic pathways, and found that the higher levels of 13 genes in the 61 metabolic genes were associated with poor outcomes in the LIHC patient cohort (Supplementary Table, S6 genes with blue background; Supplementary Fig. S5A). Then, six genes (PSPH, PSAT1, ALDH18A1, ATIC, DTYMK), which are related to amino acid metabolism and nucleotide synthesis, were chosen for verification. Most of the six genes were markedly upregulated in LIHC tumor samples compared with noncancerous samples (Supplementary Fig. S5B). Downregulation of these genes was validated by RT-PCR in Huh7 and Hep3B cells treated with B029-2 as well as in HCC cells with knockdown of p300 (Fig. 4A and B). Among these genes, the expression of PSPH, DTYMK, ALDH18A1 and PSAT1 is closely related to p300 (Supplementary Fig. S5C).
It is well known that histone acetylation increases chromatin accessibility (26). Interestingly, the ChIP-seq data from the Cistrome Data Browser indicated the binding of H3K27Ac and H3K9Ac with the promoters of PSPH, DTYMK, ALDH18A1, PSAT1, TALDO1, and ATIC genes in hepatoma cells and liver tissues (Supplementary Fig. S6A). To further unravel the mechanism by which p300/CBP regulates these genes, we examined whether acetylated histone H3 occupied the promoter region of these genes. Quantitative ChIP experiments showed that H3K18Ac, H3K27Ac, and H3K9Ac bound to the promoter region of these genes (Supplementary Fig. S6B). However, the enrichment of H3K18Ac and H3K27Ac at the promoters of these genes was reduced by treatment with B029-2, while no reduction of H3K9Ac at these regions was detected (Fig. 4C). It is known that the binding of Bromodomain protein 4 (BRD4) with H3K27Ac sites are associated with gene activity (27). Consistently, we found that the decrease of H3K18Ac and H3K27Ac was accompanied with the reduction occupations of p300 and BRD4 at the promoters of these genes (Supplementary Fig. S7A and S7B). These results indicated that p300 promoted the transcription of PSPH, PSAT1, ALDH18A1, ATIC, TALDO1, and DTYMK by regulating the binding of H3K18Ac and H3K27Ac to their promoters, and B029-2 reduced the expression of the six metabolic genes at least partially through inhibiting the p300/CBP–induced histone acetylation.
Targeting p300/CBP attenuates HCC progression by regulating amino acid metabolism and nucleotide synthesis
PSPH is the final rate-limiting enzyme (RLE) in the serine synthesis pathway (SSP), while DTYMK is a kind of RLE that is critical for dTTP biosynthesis, suggesting that both PSPH and DTYMK are crucial for nucleotide synthesis. Consistent with the results from the TCGA database, the levels of PSPH and DTYMK were increased in most HCC tumors compared with their surrounding noncancerous tissues (Fig. 5A; Supplementary Fig. S8A). RT-PCR and Western blotting also confirmed the decreased expression of PSPH and DTYMK in HCC cells and xenografts with B029-2 treatment (Fig. 5B and C; Supplementary Fig. S8B). We next investigated their functions in HCC using siRNAs targeting PSPH (siPSPH) and DTYMK (siDTYMK) (Supplementary Fig. S8C and S8D). The results showed that siPSPH and siDTYMK inhibited the proliferation, migration, and invasion of HCC cells (Fig. 5D–I; Supplementary Fig. S8E–S8H). In agreement with the alteration in biological phenomena, knockdown of PSPH and DTYMK resulted in a significant decrease in the baseline levels of glycolysis and total glycolytic capacity in Huh7 cells (Fig. 5J and K; Supplementary Fig. S8I).These data suggest that both PSPH and DTYMK play important roles in HCC progression.
To evaluate whether B029-2 inhibited HCC by regulating the metabolism of HCC cells, PSPH and DTYMK were overexpressed in HCC cells treated with B029-2 (Supplementary Fig. S9A and S9B). Enhanced expression of PSPH and DTYMK partly attenuated the inhibitory effect of B029-2 on the proliferation, migration, and invasion of HCC cells (Fig. 6A–C; Supplementary Fig. S9C and S9D). In addition, overexpression of PSPH and DTYMK also partly restored glycolysis, glycolytic capacity, and the production of adenine, uracil, and thymine in B029-2–treated HCC cells (Fig. 6D and E; Supplementary Fig. S9E and S9F). We also detected whether PSPH and DTYMK affected the tumorigenesis of Huh7 cells in mice. The data showed that overexpression of either PSPH or DTYMK increased the growth of HCC xenografts and moderately blocked B029-2–induced retardation of tumorigenesis (Supplementary Fig. S10A–S10D). More interestingly, Ki67 staining revealed that the inhibitory effect of B029-2 on the tumor proliferation was reversed by ectopic expression of PSPH and DTYMK (Fig. 6F and G). Taken together, these data suggest that B029-2 attenuates hepatocellular cancer progression partially through regulating amino acid metabolism and nucleotide synthesis.
The role of p300/CBP in cancer is controversial. Loss-of-function mutations of p300/CBP were found in cutaneous squamous cell carcinoma, suggesting that p300/CBP may act as a tumor suppressor (28). However, gain-of-function alterations of p300/CBP could contribute to cancer (including melanoma, endometrial, colorectal carcinoma, hematologic malignancies, and esophageal squamous carcinoma) development (7, 29), indicating that p300/CBP is a driver of cancer growth. The data from the publicly available COSMIC database (cancer.sanger.ac.uk) showed a low frequency of p300 mutations (1.22%, 12/987) in HCC. In addition, increased expression of p300 is reported to correlate with poor survival and aggressive phenotypes in HCC (12–15). Here, we observed that p300 expression was significantly increased in HCC samples and HCC cell lines. Moreover, H3K18Ac and H3K27Ac, two known histone acetylation types catalyzed by p300/CBP, were abundant in HCC tissues. These data provide a theoretical basis for targeting p300 for HCC therapy.
It is known that epigenetic regulation of gene expression occurs at the DNA, histone, and RNA levels. Epigenetic modifications are marks connected to differentiation and disease pathogenesis, including human cancers (30). Histone acetylation, the most common epigenetic modification, is almost invariably associated with the activation of transcription and plays a role in oncogenesis. H3K18Ac plays an important role in driving the progression of many types of cancer, including breast, colon, lung, hepatocellular, pancreatic, prostate, and thyroid cancer (31). H3K27Ac-enriched enhancers are involved in hematologic malignancies (32). In Merkel cell carcinoma, H3K27Ac is enriched in the promoter of c-Myc and promotes c-Mycexpression (33). In this study, we also found that H3K18Ac and H3K27Ac, two known histone acetylation types catalyzed by p300/CBP, were highly expressed in HCC tissues. A recent study reported that A-485, a selective p300 inhibitor, attenuated proliferation in lineage-specific tumor types (11). However, the effect of p300 inhibitors in HCC remains unknown. Herein, we assessed the effect of B029-2, which is modified from A-485 and serves as a more potent inhibitor, on HCC cells. B029-2 decreased the levels of H3K18Ac and H3K27Ac and exhibited a therapeutic effect on HCC cells and xenografts, suggesting that targeting p300/CBP might be a promising clinical strategy for HCC therapy and that B029-2 has potential applications in HCC therapy.
Cancer cells utilize altered bioenergetics to fuel-uncontrolled proliferation and progression. Aerobic glycolysis, also called “Warburg effect,” is one of the hallmarks of cancer-associated metabolic changes (34). Previous studies showed that glycolytic activities dynamically regulate the cellular levels of acetyl-CoA, which acts as a substrate for HATs and modulates the HAT-mediated acetylation of histone proteins (35). However, it is unknown whether histone acetylation directly affects tumor metabolism via epigenetic alterations. In this study, we revealed that p300/CBP promoted the transcription of metabolic enzyme genes by regulating H3K27Ac and H3K18Ac. B029-2 reduced glycolytic function and decreased the expression of the genes involved in amino acid metabolism and nucleotide synthesis. This finding indicates cross-talk between metabolism and epigenetic modifications, leading to a new understanding of the interrelationships between metabolism and epigenetics.
PSPH catalyzes the generation of serine from 3-phosphoglycerate (3-PG). Serine and glycine are the main sources of one-carbon unit production, which is important for nucleotide synthesis (36). DTYMK is a kind of RLE of nucleotide biosynthesis. A previous study showed that PSPH is critical for cMyc-driven HCC progression both in vitro and in vivo (37). In addition, upregulated DTYMK in HCC was associated with poor patient survival in multiple cohorts (38). Herein, we also found that PSPH and DTYMK exerted oncogenic effects in HCC. Moreover, the inhibitory effect of B029-2 on HCC was partially reversed by ectopic expression of PSPH and DTYMK, suggesting that p300/CBP may promote HCC progression through epigenetic regulation of metabolism in cancer cells. Interestingly, the most recent study revealed ALDH18A1, a key protein in the proline biosynthetic pathway, was also upregulated in HCC models and inhibition of ALDH18A1 leads to reduced tumor burden (39). Therefore, the roles of other p300/CPB–regulated metabolism enzymes in HCC, including PSAT1, ATIC, and TALDO1, merit further study.
In summary, our study confirmed the epigenetic alteration of H3K27Ac and H3K18Ac in HCC, providing more evidence for HAT as a therapeutic target of HCC. Our data also demonstrated the anticancer effect of the novel p300 inhibitor B029-2 in vivo and in vitro. Using the combined application of ATAC-seq and RNA-seq, a powerful platform to discover epigenetic regulation, we revealed that p300/CBP is involved in metabolic reprogramming of cancer cells by regulating H3K27Ac and H3K18Ac. B029-2 inhibits HCC partially through the reduction of glycolytic function and nucleotide synthesis. Conceptually, our findings underscore the novel interplay between epigenetics and metabolomics, and highlight the value of therapeutically targeting the HAT activity of p300/CBP in HCC.
No disclosures were reported.
L.-Y. Cai: Conceptualization, data curation, software, formal analysis, validation, methodology, writing-original draft. S.-J. Chen: Resources, supervision, funding acquisition, methodology, writing-review and editing. S.-H. Xiao: Resources, validation. Q.-J. Sun: Data curation, validation. C.-H. Ding: Funding acquisition, methodology. B.-N. Zheng: Validation, methodology. X.-Y. Zhu: data curation, software. S.-Q. Liu: Data curation, validation. F. Yang: Methodology. Y.-X. Yang: resources. B. Zhou: Resources. C. Luo: Resources. X. Zhang: Resources, supervision, funding acquisition, methodology, writing-review and editing. W.-F. Xie: Conceptualization, resources, supervision, methodology, project administration, writing-review and editing.
This work was supported by the National Natural Science Foundation of China (81530019, 81772523, 81703415, 81802324, 82072641); the Top-Level Clinical Discipline Project of Shanghai Pudong (PWYgf2018-04); Shanghai Sailing Program (17YF1423100); Youth Innovation Promotion Association (2017333); Natural Science Foundation of Shanghai (17ZR1436500); National Science & Technology Major Project “Key New Drug Creation and Manufacturing Program” (2018ZX09711002–008).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.