DNA methylation status of the SPHK1 and LTB genes underlies the clinicopathological diversity of non-alcoholic steatohepatitis-related hepatocellular carcinomas

For the present analysis, we used 26 paired samples of non-cancerous liver tissue (N) and the corresponding tumorous tissue (T) obtained by partial hepatectomy from 26 HCC patients whose N samples showed histological features compatible with NASH. All 26 patients were negative for HBV surface antigen and anti-HCV antibody. NASH stage was evaluated microscopically on the basis of the histological scoring system for NASH and the Brunt classification criteria (Brunt et al. 2011). The HCCs were diagnosed histologically in accordance with the World Health Organization classification (Torbenson et al. 2019) and the Tumor-Node-Metastasis classification (Brierley et al. 2017). Moreover, in each tumor, the percentage of tumor cells with lipid droplets among all observed tumor cells was evaluated in 10 fields of view at 100x. The presence or absence of fibrous tissue spreading along the sinusoidal space within each tumor was evaluated in 10 fields of view at 20x, and the areal ratio of the scirrhous HCC component was calculated as the ratio of such spreading-positive fields among the 10 evaluated fields. For comparison, 36 samples of normal control liver tissue (C), obtained by partial hepatectomy from 36 patients with liver metastases of primary colorectal cancers without HBV or HCV infection, chronic hepatitis, liver cirrhosis or HCC, were examined.

None of the patients had received preoperative treatment, and all underwent surgery at the National Cancer Center Hospital, Tokyo, Japan. The age, sex and clinicopathological parameters of the 26 patients from whom N and T samples were obtained and the 36 patients from whom C samples were obtained are summarized in Supplementary Table 1. Immediately after surgical removal, tissue specimens were frozen and stored in liquid nitrogen at the National Cancer Center Biobank, Tokyo, Japan, until use in research, in accordance with the Japanese Society of Pathology Guidelines for the handling of pathological tissue samples for genomic research (Kanai et al. 2018). This study was approved by the Ethics Committees of the National Cancer Center, Tokyo, Japan, and Keio University, and was performed in accordance with the Declaration of Helsinki. All of the patients provided written informed consent prior to inclusion of their specimens in the study.

High-molecular-weight DNA was extracted from fresh-frozen tissue samples using phenol–chloroform, followed by dialysis. Five hundred nanograms of genomic DNA was subjected to bisulfite treatment using an EZ DNA Methylation-GoldTM Kit (Zymo Research, Irvine, CA) in accordance with the manufacturer’s protocol. DNA methylation status at 485,577 CpG loci was examined at single-CpG resolution using the Infinium HumanMethylation450 BeadChip (Illumina, San Diego, CA) (Bibikova et al. 2009). After hybridization, the specifically hybridized DNA was fluorescence-labeled by a single-base extension reaction and detected using an iScan reader (Illumina) in accordance with the manufacturer’s protocol.

The data were then assembled using GenomeStudio methylation software (Illumina). At each CpG site, the ratio of the fluorescence signal was measured using a methylated probe relative to the sum of the methylated and unmethylated probes, i.e. the so-called β-value, which ranges from 0.00 to 1.00, reflecting the methylation level of an individual CpG site. Some of the results of the Infinium assay had been used in our previous study focusing on comparison with viral hepatitis-related HCCs and deposited in the Gene Expression Omnibus (GEO) database (https://​www.​ncbi.​nlm.​nih.​gov/​geo/​, Accession number GSE183468).

MetaCoreTM software (version 19.3) (Thomson Reuters, NY) is a pathway analysis tool based on a proprietary manually curated database of human protein–protein, protein-DNA and protein compound interactions. Using genes showing significant differences in DNA methylation levels between epigenomic clusters, MetaCore pathway analysis by GeneGo was performed. Such genes were considered significantly enriched in pathways for which the false discovery rate (FDR) was less than 0.05.

Total RNA was isolated from all 26 paired N and T samples, and 31 C samples from which tissue samples were available even after DNA extraction, using TRIzol reagent (Life Technologies, Carlsbad, CA). cDNA was generated from total RNA using random primers and SuperScript IV Reverse Transcriptase (Invitrogen, Carlsbad. CA). Levels of expression of mRNA for the SPHK1 (sphingosine kinase 1), INHBA (inhibin, beta A), LTB (lymphotoxin beta) and PDE3B (phosphodiesterase 3B) genes were determined using the PowerUp SYBR Green Master Mix (Applied Biosystems, Foster City, CA) on the 7500 Fast Real-Time PCR system (Applied Biosystems) employing the relative standard curve method. PCR primers were designed using the Primer Designer software (Thermo Fisher Scientific, UK, https://​www.​thermofisher.​com/​order/​genome-database/​) and Primer-BLAST software (https://​www.​ncbi.​nlm.​nih.​gov/​tools/​primer-blast/​index.​cgi). The sequences of the PCR primer sets employed are shown in Supplementary Table 2. Experiments were performed in triplicate, and the mean value for the three determinations was used as the threshold cycle (Ct) value. All Ct values were normalized to that of the GAPDH gene in the same sample.

The human HCC cell line Hep3B (Knowles et al. 1980) was purchased from the American Type Culture Collection (Manassas, VA) in February 2020. The human HCC cell lines PLC/PRF/5 (MacNab et al. 1976), JHH-7 (Fujise et al. 1990) and HLF (Dor et al. 1975) were purchased from the Japanese Collection of Research Bioresources (JCRB) (Osaka, Japan) in January 2020. Hep3B, PLC/PRF/5, JHH-7 and HLF were authenticated based on short tandem repeat analysis by JCRB Cell Bank in August 2022 (certification numbers: KBN0850-01, KBN0850-02, KBN0850-03 and KBN0850-04, respectively). It was confirmed that all cell lines used were mycoplasma-free. JHH-7, HLF and PLC/PRF/5 were maintained in D6046 medium (Sigma-Aldrich, St. Louis, MO) supplemented with 10% fetal bovine serum, under 95% air and 5% CO2 at 37 °C. Hep3B was maintained in M4655 medium (Sigma-Aldrich), supplemented with S8636 sodium pyruvate solution (Sigma-Aldrich), M7145 non-essential amino acid solution (Sigma-Aldrich) and 10% fetal bovine serum, under 95% air and 5% CO₂ at 37 °C.

JHH-7 and HLF cells were seeded at a density of 9 × 10⁵ cells per 15-cm dish on day 0 and then allowed to attach for 24 h. Then, 5-aza-dC (Sigma-Aldrich, St. Louis, MO) was added to a final concentration of 1 μM. Cells were passaged on day 3. At 24 h after replacing, 5-aza-dC was added again to the same final concentration. Control cells were treated with dimethyl sulfoxide. Genomic DNA and total RNA were extracted from both cell lines on day 6.

Hep3B cells were seeded in 96-well plates at a concentration of 5 × 10⁶ cells/well and PLC/PRF/5 cells at 1 × 10⁶ cells/well. When the cells had reached about 60% confluence, the medium was replaced with Opti-MEM^® I Reduced Serum Medium (Thermo Fisher Scientific). The cells were then transfected with either the negative control siRNA (siNC), SPHK1-specific siRNA (s16958 and s16959) or LTB-specific siRNA (s194597, s8311 and s8312) (Thermo Fisher Scientific) using LipofectamineTM RNAiMAX reagent (Thermo Fisher Scientific). At 48 h after transfection, the levels of expression of mRNAs for SPHK1 or LTB were determined by quantitative real-time RT-PCR analysis, using GAPDH as the reference gene. Transfected cells were then subjected to the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) cell viability assay, cell apoptosis assay and cell migration assay.

The MTS cell viability assay was performed as described previously (Hamada et al. 2021; Arai et al. 2015). Briefly, 48 h after transfection with negative control siRNA and SPHK1- or LTB-specific siRNAs, cells were treated with CellTiter 96 Aqueous One Solution Reagent (Promega, Madison, WI). After 1 h of treatment, the optical density was measured at 490 nm on a GloMax^®-Multi + Detection System Glomax (Promega). Results were presented as the mean ± standard deviation for three separate determinations.

The apoptosis assay was performed as described previously (Hamada et al. 2021; Arai et al. 2015). Briefly, 48 h after transfection with negative control siRNA and SPHK1- or LTB-specific siRNAs, cells were treated with a Caspase-Glo 3/7 assay kit (Promega). After 1 h of incubation, the luminescent signal was measured on a GloMax-Multi + Detection System Glomax (Promega). Results were presented as the mean ± standard deviation for three separate determinations.

The cell migration assay was performed as described previously (Hamada et al. 2021; Arai et al. 2015). Cell migration was determined using 24-well transwell chambers with an 8-μm pore polycarbonate filter (Corning Inc., Corning, NY). Forty-eight hours after transfection with the negative control siRNA and SPHK1- or LTB-specific siRNAs, 5 × 10⁴ PLC/PRF/5 and Hep3B cells were seeded onto the upper-side transwells in 100 μl of serum-free medium, and 500 μl of the complete medium was added to the lower chamber. The cells were incubated to allow migration for 48 h at 37 °C and 5% CO₂. At the end of the assay, the non-motile cells on the top surface of the inserts were removed with cotton swabs. Cells that had passed through the polycarbonate membrane were fixed with 10% formalin and stained with 0.5% crystal violet to visualize the attached cells. The crystal violet was eluted with 10% acetic acid and the optical density was measured at 600 nm on a GloMax Multi Detection System (Promega). Results were presented as the mean ± standard deviation for three separate determinations.

In the Infinium assay, the call proportions (P < 0.01 for detection of signals above the background) for 859 probes in all of the examined tissue samples were less than 90%. Since such a low proportion may due to polymorphism at the probe for CpG sites, these 859 probes were excluded from subsequent analysis, as described previously (Fujimoto et al. 2020; Tsumura et al. 2019). In addition, 45 probes with missing β values in more than 10% of the samples were excluded. Finally, probes on chromosomes X and Y were removed to avoid any gender-specific methylation bias, leaving a final total of 473,332 autosomal CpG sites.

Differences in levels of DNA methylation and mRNA expression between sample groups were examined by Welch’s t test. To correct for multiple testing, we used Bonferroni correction. The DNA methylation profiles were analyzed using principal component analysis (PCA) and hierarchical clustering (Euclidean distance, Ward’s method). Correlations between epigenetic clustering and clinicopathological parameters were tested by Welch’s t test and Fisher’s exact test. All statistical analyses were performed using the programing language R. Differences at P values of less than 0.05 were considered statistically significant.

A total of 64, 027 probes, for which DNA methylation levels differed significantly between C and T samples, were identified (Welch’s t test, P < 0.05 after Bonferroni correction, Δβ_T−C value of more than 0.2 or less than − 0.2), indicating that DNA methylation alterations are associated with NASH-related hepatocarcinogenesis. To examine the DNA methylation profiles during multistage NASH-related hepatocarcinogenesis, principal component analysis of C, N and T samples was performed using the 64,027 probes (Fig. 1a). Since N samples were obtained from non-cancerous liver tissue that had already become the origin of NASH-related HCCs, such samples were considered to be at the precancerous stages. Such precancerous N samples showed distinct DNA methylation profiles that clearly differed from those of normal control C samples (Fig. 1a). Moreover, T samples themselves were scattered over a wider area, indicating heterogeneity of the DNA methylation profiles of such samples (Fig. 1a).

To further investigate the heterogeneity of T samples, we performed hierarchical clustering using the DNA methylation levels of the 64,027 probes (Euclidean distance, Ward’s method) (Fig. 1b): T samples were separated into Cluster I (n = 8) and Cluster II (n = 18). Correlations between such epigenetic clustering of NASH-related HCCs and clinicopathological parameters, such as age, sex and histopathological findings of N and T samples, were examined (Table 1). Poorly differentiated HCCs were significantly accumulated in Cluster I, whereas most of the well to moderately differentiated HCCs belonged to Cluster II (P = 0.020). The average percentage of tumor cells with lipid droplets (tumor steatosis) in Cluster II (15.2 ± 18.0%) was significantly higher than that in Cluster I (3.88 ± 5.25%) (P = 0.027). On the other hand, the average areal ratio of the scirrhous HCC component in Cluster I (41.1 ± 42.6%) was significantly higher than that in Cluster II (0.66 ± 1.57%) (P = 0.040). Representative photos of poorly differentiated HCC, HCC with tumor cell steatosis (steatotic HCC) and scirrhous HCC are shown in Fig. 2.

Since epigenetic clustering was significantly correlated with the histological diversity of NASH-related HCCs, i.e. poorer tumor differentiation, tumor steatosis and development of the scirrhous HCC component, we identified the 140,241 probes showing significant differences in DNA methylation levels between Clusters I (n = 8) and Cluster II (n = 16) (Welch’s t test, P < 0.05). Among the 140,241 probes, 63,523 were located within CpG islands, island shores (2000-bp regions adjacent to a CpG island) or island shelves (2000-bp regions adjacent to an island shore) based on the University of California, Santa Cruz (UCSC) genome browser (https://​genome.​ucsc.​edu/​). Among the 63,523 probes, 22,512 are annotated with TSS1500 (from 200 bp upstream of the transcription start site [TSS] to 1500 bp upstream of it), TSS200 (from TSS to 200 bp upstream of it), the 5′ untranslated region (UTR) or the 1st exon based on the RefSeq database (http://​www.​ncbi.​nlm.​nih.​gov/​refseq/​).

To focus on DNA methylation alterations possibly resulting in changes of expression, using datasets for 412 samples of non-cancerous and cancerous liver tissue deposited in the Cancer Genome Atlas (TCGA) database (https://​www.​cancer.​gov/​about-nci/​organization/​ccg/​research/​structural-genomics/​tcga), correlations between DNA methylation and mRNA expression levels for the 22,512 probe CpG sites were examined. Among the 22,512 probes, 3888 designed for 1890 genes showed a significant inverse correlation between the levels of DNA methylation and mRNA expression (r < − 0.2, P < 0.05) (Supplementary Table 3). Inverse correlations between DNA methylation and mRNA expression levels on representative genes are shown in Supplementary Fig. 1.

The 3888 probes were then subjected to MetaCore pathway analysis. Genes showing significant differences in DNA methylation levels between the two clusters were significantly accumulated in 353 signaling pathways (FDR < 0.05). After elimination of pathways solely participating in diseases other than cancer or organs other than the liver, Table 2 summarizes the top 25 signaling pathways. Among them, representative pathways clearly participate in cell adhesion and cytoskeletal remodeling, such as “Cytoskeleton remodeling_Regulation of actin cytoskeleton organization by the kinase effectors of Rho GTPases (FDR = 5.49 × 10⁻⁶)” and “Cell adhesion_Histamine H1 receptor signaling in the interruption of cell barrier integrity (FDR = 1.62 × 10⁻⁴)”, cell proliferation and death, such as “Development_Positive regulation of STK3/4 (Hippo) pathway and negative regulation of YAP/TAZ function (FDR = 8.64 × 10⁻⁵)” and “Signal transduction_FGFR3 signaling (FDR = 3.97 × 10⁻⁴)” and epigenetic regulation, such as “CHDI_Correlations from Discovery data_Causal network (FDR = 2.09 × 10⁻⁴)” and “Development_H3K27 demethylases in differentiation of stem cells (FDR = 3.97 × 10⁻⁴)”. Representative pathway maps are shown schematically in Supplementary Fig. 2. The top 25 pathways in Table 2 consist of 149 genes in total. Among these 149 genes, taking into consideration their literature-based implications in the process of carcinogenesis, we further focused on the SPHK1, INHBA, LTB and PDE3B genes.

Levels of mRNA expression for the SPHK1, INHBA, LTB and PDE3B genes were examined in the present liver tissue samples using real-time quantitative RT-PCR analysis. To confirm that DNA methylation profiles participate in histological diversity characterizing the epigenetic clustering, correlations between DNA methylation levels based on the Infinium assay and mRNA expression levels based on real-time quantitative RT-PCR analysis of the focused genes on one hand and poorer tumor differentiation were examined (Fig. 3). DNA hypermethylation of INHBA and PDE3B resulting in their reduced expression was observed in poorly differentiated HCCs (n = 5) relative to well to moderately differentiated HCCs (n = 21). On the other hand, DNA hypomethylation of SPHK1 and LTB resulting in their overexpression was observed in poorly differentiated HCCs, although such overexpression did not reach a statistically significant level due to a few outliers. These data indicated that DNA methylation profiles participate in determining the histological diversity of HCCs, such as poorer differentiation, via alterations of gene expression.

To further reveal details of DNA methylation status-related transcriptional regulation, we focused on the SPHK1 and LTB genes. Based on the Infinium assay, the DNA methylation levels of the SPHK1 gene in the human HCC cell lines, Hep3B, PLC/PRF/5, JHH-7 and HLF, are shown in Fig. 4a. In the top two cell lines showing the highest levels of DNA methylation for SPHK1, JHH-7 and HLF, the levels of mRNA expression were low (Fig. 4a). Similarly, in the top two cell lines showing the highest DNA methylation levels for LTB, JHH-7 and HLF, the levels of mRNA expression were also low (Fig. 4c). These cell lines were then subjected to 5-aza-dC treatment. This led to a marked reduction in the levels of DNA methylation and restoration of the expression levels of SPHK1 and LTB mRNA (Fig. 4b, d, respectively), indicating that the mRNA expression levels of these genes are regulated by DNA methylation in HCC cells.

Knockdown of SPHK1 using siRNA transfection was performed in the top two cell lines showing the highest mRNA expression levels, Hep3B and PLC/PRF/5. After transfection with the SPHK1-specific siRNAs, s16958 and s16959, marked reduction of SPHK1 expression was confirmed in both cell lines by quantitative real-time RT-PCR (Fig. 5a). Decreased cell growth was observed in s16959-treated Hep3B and PLC/PRF/5 cells (Fig. 5b). Moreover, caspase-3/7 activities were increased (Fig. 5c) and cell migration ability was repressed (Fig. 5d) using all SPHK1-specific siRNAs in both Hep3B and PLC/PRF/5 cells.

Knockdown of LTB using siRNA transfection was performed in the top two cell lines showing the highest mRNA expression levels, PLC/PRF/5 and Hep3B. After transfection, reduction of LTB expression was confirmed in PLC/PRF/5 (s194597) and Hep3B (s8311 and s8312) by quantitative real-time RT-PCR (Fig. 5e). Although an increase of both cell growth and caspase-3/7 activities was observed in only Hep3B cells (Fig. 5f, g), cell migration ability was clearly repressed by knockdown of LTB in both PLC/PRF/5 and Hep3B cells (Fig. 5h).