TEC-mediated tRF-31R9J regulates histone lactylation and acetylation by HDAC1 to suppress hepatocyte ferroptosis and improve non-alcoholic steatohepatitis

Abstract

Background

Tectorigenin (TEC) is a monomer of anthocyanin, which we found exhibits hepatoprotective effects. tRNA-derived fragments (tRFs) and ferroptosis play important roles in the pathogenesis of non-alcoholic steatohepatitis (NASH). Recent discoveries have revealed that histone lactylation and acetylation play a crucial role in connecting cellular metabolism and epigenetic regulation through post-translational modification of histones. However, it is unclear whether TEC improves NASH by regulating histone lactylation, acetylation and hepatocyte ferroptosis through tRFs.

Results

In this study, we demonstrated that TEC significantly inhibits free fatty acids-induced hepatocyte ferroptosis both in vitro and in vivo. We identified tRF-31R9J (tRF-31-R9JP9P9NH5HYD) involved in TEC regulation of ferroptosis in steatosis hepatocytes. Overexpression of tRF-31R9J suppressed hepatocyte ferroptosis and enhanced cell viability in steatosis HepG2 cells. Knockdown of tRF-31R9J partially counteracted the inhibitory effect of TEC on ferroptosis in hepatocytes. Mechanistically, tRF-31R9J recruited HDAC1 to reduce the levels of histone lactylation and acetylation modifications of the pro-ferroptosis genes ATF3, ATF4, and CHAC1, thereby inhibiting their gene expression.

Conclusions

This study demonstrates that TEC-mediated tRF-31R9J inhibits hepatocyte ferroptosis through HDAC1-regulated histone delactylation and deacetylation, thereby improving NASH. These discoveries offer a theoretical foundation and new strategies for the medical management of NASH.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13148-025-01813-3.

Introduction

Non-alcoholic steatohepatitis (NASH) ranks high among the prevalent chronic liver diseases in children and adults worldwide, with an estimated global prevalence of 24% [1]. NASH is distinguished by the presence of hepatic steatosis, inflammation, hepatocellular injury, and variable degrees of fibrosis [2]. While the theory centered around oxidative stress and lipid peroxidation has been a hot topic in NASH research [3, 4], the pathogenesis of NASH remains incompletely understood. Consequently, it holds great importance to unravel the pathogenic mechanisms of NASH and investigate new potent drugs for clinical adjunctive therapy of NASH.

Blueberry is rich in nutrients such as anthocyanins, polyphenols and various trace elements [5], with the function of antioxidant, anti-inflammatory, hypolipidemic and so on [6]. Our preliminary research found that blueberries significantly reduced hepatic fat deposition and improved fatty liver [7–9]. We have found that tectorigenin (TEC), an abundant monomers of anthocyanins in blueberry, is the most effective inhibitor on the accumulation of lipid droplets within hepatocytes [10]. In rats with chemically induced hepatic fibrosis, TEC significantly inhibited the increase in the amount of collagen in the livers [11]. Furthermore, TEC enhances high-fat diet-induced non-alcoholic fatty liver disease (NAFLD) by exerting anti-inflammatory effects and regulating the gut microbiota [12]. These pieces of evidence preliminarily confirmed that TEC exerted a hepatoprotective effect. However, the regulatory mechanisms of TEC on diseases related to fatty liver had not been fully elucidated.

Ferroptosis is a type of cell death that relies on iron and is characterized by extensive lipid peroxidation [13]. An increasing body of research has demonstrated the significant involvement of ferroptosis in the development of several liver diseases including alcoholic liver disease, hepatitis C virus infection, NASH and hepatocellular carcinoma [14–17]. In NASH, hepatocyte ferroptosis plays a trigger to initiate inflammation. Weakening ferroptosis almost completely protects hepatocytes from necrosis and repressed subsequent necrotic immune cell infiltration and inflammatory responses [18]. Particularly, TEC inhibited ferroptosis in renal tubular epithelial cells via smad3 [19]. However, whether TEC controls ferroptosis in hepatocytes and the underlying mechanism is unknown.

tRNA-derived fragments (tRFs) are a novel category of conserved small non-coding RNAs derived from tRNAs. tRFs not only simulate the function of miRNA to inhibit the expression of target genes, but also regulate RNA binding proteins to regulate the transcription, mRNA stability and translation of downstream target genes [20]. Recent studies have shown that tRFs regulate ferroptosis, thereby influencing the presence and advancement of diseases like diabetic kidney disease, lung injury, and pancreatitis [21–23]. Our previous findings have indicated a significant role of tRF-3001b in both cellular and mouse models of NAFLD through the autophagy pathway [24]. The autophagy/pyroptosis pathway regulated by tRF-47-58ZZJQJYSWRYVMMV5BO is also involved in the improvement of NASH in mice [10]. The above examples suggested that tRFs represent a promising therapeutic target for NASH.

Histone lactylation refers to the lactylation modification of lysine residues on histones, also known as lysine lactylation (Kla). Lactylation is a post-translational modification of proteins that exerts its influence on various biological processes, including DNA replication, gene transcription, and DNA damage repair [25]. The report associated with lactylation mainly clustered in tumor progression, nerve excitation, and inflammation [26–28]. HDAC1 is a histone deacetylase, and liver Mettl3 or WTAP can prevent the progression of NASH by recruiting HDAC1 to inhibit gene expression [29, 30]. Danshensu-mediated HDAC1 overexpression alleviated NAFLD by initiation of TAF9 [31]. However, the role of histone lactylation in liver diseases remains unclear, and it is not yet known whether tRFs regulate histone lactylation mediated by HDAC1 to affect hepatocyte ferroptosis.

The objective of this study is to explore the role and molecular mechanisms of TEC-mediated tRFs in regulating ferroptosis in steatosis hepatocytes. We validated the effects of TEC on ferroptosis in NASH from both in vivo and in vitro perspectives. The analysis of our previous small RNA sequencing data revealed novel tRFs in NASH and their role in ferroptosis of steatotic hepatocytes was explored. Finally, we investigated the molecular mechanisms of tRF-31R9J in regulating hepatocyte ferroptosis by interacting with HDAC1. This study introduces novel molecular targets for the clinical management of NASH.

Methods

Cell culture and NASH cell model construction

MEM basal medium (10-010-CVR, Corning) supplemented with 10% FBS (10099–141, GIBCO) and 1% P/S was utilized for culturing the human hepatocellular carcinoma cell line HepG2. The cells were incubated at 37 °C with 5% CO₂. To construct the NASH cell model, HepG2 cells were exposed to a mixture of FFA comprising oleic acid and palmitic acid in a ratio of 2:1, at a concentration of 1 mmol/L, for a duration of 24 h. The steatosis HepG2 cells were subjected to treatment with varying concentrations of TEC: low concentration (25 μM), medium concentration (50 μM), and high concentration (75 μM) [10].

CCK-8 assay

HepG2 cells from each group were quantified by CCK-8 kit (Beyotime) to detect cell viability. To obtain a single-cell suspension, HepG2 cells were dissociated using Trypsin–EDTA (11668–500, Invitrogen). The cells were adjusted to a concentration of 3 × 10⁴ cells/mL for each group. Each well of a 96-well plate was seeded with 100 μL of the cell suspension, resulting in 3000 cells/well. Each sample had 6 replicate wells, and the edge wells were filled with 100 μL of sterile water or PBS. The plate was incubated at 37 °C with 5% CO₂ overnight. After the addition of 10 μL of CCK-8 to each well, the plate was incubated for 2 h. The OD value at 450 nm was quantified using the microplate reader (Infinite M1000, TECAN).

Detection of lipid oxidation

The lipid peroxidation levels in liver tissues and cells were quantitatively measured utilizing the Lipid Peroxidation MDA Assay Kit (Beyotime, S0131S). MDA (Malondialdehyde) is a natural byproduct of lipid peroxidation in living organisms. When cells from animals or plants undergo oxidative stress, lipid peroxidation occurs. Some fatty acids, after being oxidized, gradually break down into a complex series of compounds, including MDA [32]. Therefore, the level of lipid oxidation can be detected by measuring the level of MDA. Tissue and cells were lysed using cell lysis buffer (P0013). The ratio of tissue weight to lysis buffer was 10%, and for every 1 million cells, 0.1 mL of lysis buffer was utilized. The lysates were subjected to centrifugation at 10,000–12,000 g for 10 min. An appropriate amount of TBA was weighed, and TBA stock solution with a concentration of 0.37% was prepared. MDA can react with TBA (Thiobarbituric Acid) at higher temperatures and in acidic environments to form a red MDA-TBA adduct. To prepare the MDA detection working solution, 1500 μL of TBA dilution solution, 500 μL of TBA stock solution, and 30 μL of antioxidant were mixed. The solution was heated at 70 °C and vigorously vortexed to promote dissolution. Three centrifuge tubes were prepared: one with 0.1 mL of PBS, one with 0.1 mL of different concentrations of standard solutions for the standard curve, and one with 0.1 mL of the sample for measurement. Then, 0.2 mL of MDA detection working solution was added to each tube. The tubes were heated at 100 °C or in a boiling water bath for 15 min. The tubes were centrifuged at 1000 g for 10 min at room temperature. The absorbance was measured at 532 nm using the microplate reader.

Detection of Fe²⁺ levels

The levels of Fe²⁺ were quantitatively measured using the Intracellular Iron Colorimetric Assay Kit (Applygen, E1042). To prepare a mixture called Solution A, the lysis buffer was combined with a 4.5% potassium permanganate solution in a 1:1 ratio. Solution A was incubated with the test samples at 60 °C for one hour. Upon reaching room temperature, centrifugation was carried out, followed by the addition of 30 μL of the Fe²⁺ detection reagent. The absorbance of the samples was quantified within the range of 540–580 nm.

Detection of reactive oxygen species (ROS) levels

ROS levels in HepG2 cells were measured using the Reactive Oxygen Species Assay Kit (Solarbio, CA1410). The fluorescent probe DCFH-DA was diluted 1:1000 in serum-free culture medium to a final concentration of 10 μmol/L. An appropriate volume of diluted DCFH-DA was added to fully cover the cells. The cells were incubated at 37 °C for 30 min. Cells were washed three times and then observed and imaged using a fluorescence microscope.

Western blot (WB)

The HepG2 cells were centrifuged at 800 g for five minutes at 4 °C. Five times the volume of RIPA lysis buffer (Thermo) was added to the cells. The mixture was placed on ice for 10 min and vortexed for 30 s every five minutes. The lysate was centrifuged at 12,000 g for 10 min at 4 °C. SDS-PAGE gel electrophoresis was performed, and subsequently, the proteins were transferred onto a PVDF membrane. The PVDF membrane was then blocked with TBST solution containing 5% skim milk for three hours. Primary antibodies against GPX4 (Abcam, ab125066, 1:5000 dilution), ACSL4 (Abcam, ab155282, 1:50000 dilution), GAPDH (Proteintech, 60004-1-Ig, 1:10000 dilution), HDAC1 (Proteintech, 10197-1-AP, 1:10000 dilution), H3K18la (Jingjie PTM Biolab, PTM-1406RM, 1:1000 dilution), H3K18ac (Proteintech, 39129, 1:5000 dilution), Histone H3 (Abcam, ab8896, 1:15000 dilution), ATF3(Abcam, ab254268, 1:1000 dilution), ATF4(Abcam, ab270980, 1:1000 dilution), CHAC1(Proteintech, 15207-1-AP, 1:1000 dilution) were added and incubated at 4 °C overnight. The secondary antibodies Goat Anti-Rabbit IgG H&L(HRP) (Abcam, ab6721, 1:20,000 dilution), Goat Anti-Rabbit IgG H&L(HRP) (Beyotime, A0208, 1:1000 dilution), or Goat Anti-Mouse IgG H&L(HRP) (Beyotime, A0216, 1:1000 dilution) were added and incubated for two hours. An ECL substrate (Thermo) was utilized for chemiluminescent detection, and the images were acquired with a chemiluminescence imaging system.

Oil red O staining

Lipid content in HepG2 cells was examined using Oil Red O staining. Fresh staining solution, Oil Red O Staining Solution (SANGON BIOTECH, E607319), was prepared by mixing ddH₂O and the staining solution at a ratio of 6:4. The cells were fixed with 4% Paraformaldehyde Fix Solution (SANGON BIOTECH, E672002) at room temperature for 30 min. After two washes, the cells were treated with a freshly prepared staining solution. Then the cells were stained at room temperature for 60 min. After three washes, the stained cells were observed and photographed with a microscope.

Construction of NASH mouse model

Male C57BL/6 mice (8 weeks old) were purchased from SPF Biotech (BEIJING). There were two animal experiments conducted. In the first experiment, a random allocation was employed to divide the mice into five groups (n = 5). NC group: Mice were fed a standard diet for 18 weeks. NASH group: Mice were fed a composite high-fat diet (88.3% regular diet + 10% lard + 1.5% cholesterol + 0.2% sodium cholate) for 18 weeks. NASH + TEC group: After 10 weeks on the high-fat diet, mice were orally gavaged with three different concentrations of TEC (25 uM, 50 uM, 75 uM and 1.5 mL/100 g) [10] once daily for eight consecutive weeks. The second animal experiment consisted of four groups (n = 5). NC group and NASH group: These two groups were treated the same as before. NASH + TEC group: The concentration of TEC was 75 μM. NASH + TEC + tRF-31R9J antagomir/NC group: NASH mice were orally gavaged with TEC and subsequently injected with tRF-31R9J antagomir (5 μg/mouse in 1.5 mL saline) via the tail vein twice a week for 8 weeks. Upon completion of the experiment, the animals were humanely euthanized by administering an overdose of CO₂, and confirmation of death was obtained through cervical dislocation. Liver tissues were immediately subjected to pathological examination and biochemical analysis.

This study was approved and supervised by the Animal Care and Ethics Committee of Guizhou Medical University Affiliated Hospital. All animal handling procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals by the National Institutes of Health.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

RNA extraction from steatosis HepG2 was performed utilizing TRIzol™ (Invitrogen). A reverse transcription kit (Thermo, K1622) was utilized to perform the reverse transcription of RNA. The program was configured with the following parameters: 42 °C for 6 min, followed by 70 °C for 5 min. The obtained cDNA was further subjected to real-time PCR amplification using the 2 × Master Mix kit (Roche) in a real-time PCR machine (Applied Biosystems Inc, ABI Q6). U6 was employed as an endogenous control for normalization. The primers utilized in this study were obtained from SANGON BIOTECH and the primer sequences are provided in Table S1. The 2^−ΔΔCT method was employed to calculate the relative expression of the genes.

Cell transfection

Cell spreading was performed in 6-well plates with 5 × 10⁵ HepG2 cells per well. The cells were cultured in antibiotic-free medium (RC-003, STEMERY). Lipofectamine™ 2000 transfection reagent (Invitrogen), tsRNA-19724 mimics, tsRNA-19724 inhibitor, NC mimics, and NC inhibitor were diluted in OPTI-MEM (CORNING) and incubated for five minutes. The diluted RNA and transfection reagent were incubated at room temperature for 15 min before being added to the cells. The medium was replaced with antibiotic-free medium after six hours for subsequent experiments. The sequences of transfection fragments are provided in Table S2.

H&E staining

Liver tissues were fixed in 4% paraformaldehyde to prepare pathological sections. Each sample was embedded in paraffin and cut into sections that were 4 μm thick. The sections were immersed in hematoxylin staining solution for 10 min, followed by eosin staining for three minutes. Subsequently, dehydration and mounting were performed. The sections were observed using the microscope to evaluate lipid degeneration.

Immunofluorescence (IF)

For cellular immunofluorescence, the HepG2 cells were seeded in a 12-well plate with 5 × 10^4 cells per well. HepG2 cells were transfected with tRF-31R9J mimics and HDAC1 interfering fragments utilizing lipo2000 (Invitrogen). After six hours, each group was treated with FFA. To fix the cells, 300 μL of 4% paraformaldehyde was added and incubated for 30 min. Then, 300 μL of 0.5% Triton X-100 was added for permeabilization for 15 min. Subsequently, 300 μL of 3% BSA was added for blocking for 30 min. For each well, 500 μL of primary antibody against H3K18la (Jingjie PTM Biolab, PTM-1406RM, 1:100 dilution) and H3K18ac (Proteintech, 39129, 1:500 dilution) was added and incubated overnight at 4 °C. Then, 250 μL of secondary antibody Alexa Fluor 488-labeled Goat Anti-Rabbit IgG (H + L) (Abcam, ab150077, 1:250 dilution) was added and incubated in the dark for one hour. Subsequently, 300 μL of DAPI was added for dark staining for 10 min. The images were observed using a fluorescence microscope.

For tissue immunofluorescence, liver tissue sections were fixed with 4% paraformaldehyde. The sections were incubated with primary antibodies against ACSL4 (Abcam, ab155282, 1:200 dilution) and GPX4 (Abcam, ab125066, 1:150 dilution) at 37 °C for one hour. The sections were incubated with the secondary antibody, Alexa Fluor 488-labeled Goat Anti-Rabbit IgG (H + L) (Abcam, ab150077, diluted 1:250). The fluorescence signals were visualized using the fluorescence microscope.

Detection of triglyceride (TG) content

The TG Content Assay Kit (Enzyme-linked Biotechnology, ml076637) was employed for the detection of TG content in liver tissue. Approximately 0.1 g of tissue was weighed and combined with 1 mL of reagent 1. After homogenizing on ice, the mixture was centrifuged at 8000 g, 4 °C for 10 min. Reagent 2 and reagent 3 were added to the supernatant. The mixture was incubated for 20 min, and the absorbance was measured at 505 nm wavelength with a plate reader (ThermoFisher, FC).

Total cholesterol (TC) content detection

The measurement of TC content in liver tissue was performed using the TC Content Assay Kit (Solarbio, BC1980). The working solution was prepared by mixing reagent 1, reagent 2, and reagent 3 in a ratio of 3 mL:20 μL:3 μL, respectively. Approximately 0.1 g of tissue was weighed and added to 1 mL of extraction solution, followed by homogenization on ice. The mixture was then centrifuged at 10,000 g, 4 °C for 10 min. The working solution was added to the supernatant, and left to incubate for 15 min. The absorbance was measured at 500 nm wavelength with the plate reader (ThermoFisher, FC).

RNA pull-down

HePG2 cells were lysed using 100 μL of cell lysis buffer to obtain a protein concentration of 3 μg/μL. After centrifugation, the supernatant was collected, and 50 μL of streptavidin-coated magnetic beads were added. Then, 50 pmol of labeled negative control probe/target probe was added separately, and the mixture was incubated for 15–30 min. A total of 100 μL of 1 × Protein-RNA Binding Buffer was added to the magnetic beads and mixed. The RNA–protein binding system mixture was prepared and added to the magnetic beads. The mixture was incubated at 4 °C for 30–60 min. The 1 × wash buffer was added for washing and eluting the RNA-bound protein complex. Then, 20 μL of the pull-down complex obtained from the experiment was loaded onto a 10% SDS-PAGE gel. After electrophoresis, silver staining was carried out according to the instructions of the silver staining kit (Beyotime, P0017S).

IF/Fluorescence in situ hybridization (FISH)

Co-localization of HDAC1 and tRF-31R9J was performed using the IF/FISH dual staining method. According to the previously described method, HDAC1 localization was performed using a primary antibody against HDAC1 (Proteintech, 10197-1-AP) and the secondary antibody Cy3-labeled Goat Anti-Rabbit IgG (H + L) (Beyotime, A0516, 1:300 dilution). The tRF-31R9J was localized using the FISH kit (Bersinbio). HepG2 cells were fixed with 4% paraformaldehyde. Subsequently, the cells were hybridized with a specific probe for tRF-31R9J and counterstained with DAPI. Photographs of the samples were taken using the fluorescence microscope.

RNA immunoprecipitation (RIP)

The HDAC1-knockdown HepG2 cells were lysed and the lysates were aspirated into nuclease-free centrifuge tubes. To a new tube, 50 μL of magnetic beads and either 5 μg of HDAC1 antibody or 1 μg of IgG antibody were added and incubated at room temperature with rotation. The cell lysates were then centrifuged at 14,000 rpm for 10 min at 4 °C. A total of 100 μL of the supernatant was added to the magnetic bead-antibody tube and incubated overnight at 4 °C with rotation. RNA was extracted using the Trizol method and purified overnight using column chromatography.

Transcriptome sequencing

The steatosis hepatocarcinoma cells treated with TEC were divided into tRF-31R9J knockdown group and control group for transcriptome sequencing. TRIzol™ reagent (Invitrogen) was employed for the extraction of total RNA. mRNA was enriched and cDNA libraries were constructed. The libraries were then sequenced using Illumina novaseq6000. The raw data underwent quality control procedures to acquire clean reads. The clean reads were mapped to the reference genome sequence. RPKM (Reads Per Kilo Million) was used for normalization. Differential gene expression analysis was performed using the DESeq algorithm. In the screening process, genes with Log₂FC greater than 1 or less than −1, and FDR less than 0.05 were considered as differentially expressed genes (DEGs).

The selected DEGs were annotated with pathways in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.genome.jp/kegg) to obtain all the pathways in which the DEGs are involved.

Chromatin immunoprecipitation sequencing (ChIP-seq) data processing

The ChIP-seq data of H3K18la from the GSE208727 dataset in the GEO database were analyzed. The filtered clean reads were mapped to the reference genome of the corresponding species using the bowtie2 software. The alignment results were visualized. The macs2 software was used for peak calling and peak annotation analysis to identify regions of histone lactylation and the distribution of peaks on gene elements. The MAnorm2 software was used for differential analysis and annotation of histone modifications. In the screening process, peaks with a M.value greater than 1 or less than −1, and p-value less than 0.05 were considered as differentially modified peaks.

ChIP-qPCR

Approximately 5 × 10⁶ HepG2 cells were fixed in 1% formaldehyde. The cells were sonicated to fragment the DNA into 300 bp fragments. H3K18la antibody (PTMBIO, PTM-1406RM) and H3K18ac antibody (Proteintech, 39,129) were incubated with pre-treated Protein A/G magnetic beads, respectively. IgG was employed as a negative control. ChIP DNA was extracted and qPCR analysis was performed to detect the expression of ATF3, ATF4, and CHAC1. The primer sequences are provided in Table S1. All qPCR reactions were performed on a real-time fluorescence quantitative PCR instrument (ABI Q6, Applied Biosystems Inc., USA).

Statistical analysis

All experiments were conducted with a minimum of three biological replicates. The results are presented as mean ± standard deviation (SD). Statistical analysis was performed utilizing GraphPad Prism 8. For comparisons between two groups, a t test was employed; while, one-way analysis of variance (ANOVA) followed by Tukey's test was used for comparisons among multiple groups. A p value less than 0.05 was considered statistically significant.

Results

TEC attenuates FFA-induced ferroptosis in hepatocytes in vitro and in vivo

Various modes of cell death including ferroptosis, autophagy, pyroptosis, apoptosis and necrosis have now been reported to be closely associated with NASH. Our previous study showed that TEC attenuated FFA-induced hepatocyte death [10]. However, which type of death was mainly inhibited by TEC has not been clarified. To elucidate the phenomenon, we treated HepG2 cells with TEC and multiple death inducers (Erastin for ferroptosis; GSDMD overexpression plasmid for pyroptosis; Rapamycin for autophagy; Z-VAD for necrosis; CCCP for apoptosis). The results showed that, compared with the DMSO group, the cell survival fraction was significantly reduced after treatment with various death inducers, while TEC attenuated this effect, with the most significant effect observed in the ferroptosis inducer group (Fig. 1A). Therefore, we further examined whether TEC regulates hepatocyte ferroptosis. We induced a NASH cell model using FFA and treated with TEC. The CCK-8 results found that FFA significantly decreased cell viability, while different concentrations of TEC reversed this phenomenon (Fig. 1B). The levels of the final degradation product of lipid peroxidation MDA were measured in hepatocarcinoma cells. We found that FFA significantly increased the intracellular MDA levels, and this effect was significantly reversed after TEC treatment (Fig. 1C). Furthermore, we observed that FFA significantly elevated the levels of intracellular Fe²⁺ and ROS, which were mitigated by TEC (Fig. 1D and E). WB analysis of key molecules involved in ferroptosis, ACSL4, and GPX4, revealed that FFA significantly increased ACSL4 levels and decreased GPX4 levels. These effects were also counteracted by different concentrations of TEC (Fig. 1F). The Oil Red O staining results showed that FFA promoted lipid accumulation in hepatocarcinoma cells and TEC attenuated the effects of FFA (Fig. 1G). These results demonstrate that TEC attenuates FFA-induced ferroptosis in hepatocarcinoma cells.

Fig. 1.

In vitro, TEC attenuates FFA-induced ferroptosis in hepatocarcinoma cells. A HepG2 were exposed to various death inducers and TEC intervention was administered simultaneously. The measurement of cell viability was conducted by the CCK-8 assay (N = 6). DMSO: the vehicle control group. The death inducers from left to right were ferroptosis inducer (Erastin), pyroptosis inducer (GSDMD overexpression), autophagy inducer (rapamycin), necrosis inducer (Z-VAD), and apoptosis inducer (CCCP). B After treatment with FFA and TEC, cell viability was evaluated utilizing the CCK-8 assay (N = 6). C Measurement of the levels of MDA in hepatocarcinoma cells (N = 3). D Colorimetric assay was employed to measure the Fe²⁺ content. E Fluorescence was used to detect ROS levels in FFA-induced hepatocarcinoma cells (N = 3). F WB was performed to evaluate the ferroptosis key molecules ACSL4 and GPX4. G Lipid accumulation in NASH cells was assessed through Oil Red O staining (N = 3). *P < 0.05, **P < 0.01

Next, NASH mice were generated by feeding a high-fat diet, and then treated with varying concentrations of TEC (Fig. 2A). Oil Red O staining confirmed the same results, indicating a reduction in lipid accumulation in NASH mice following TEC treatment (Fig. 2B). Furthermore, we found that MDA and Fe²⁺ content were significantly upregulated in the liver tissue of NASH mice compared with the NC group, and the administration of TEC exhibited a pronounced counteractive effect on this phenomenon (Fig. 2 C, D). These results indicate that TEC effectively inhibits FFA-induced ferroptosis in NASH mice.

Fig. 2.

TEC alleviates ferroptosis in hepatocytes of NASH mice. A A high-fat diet was administered to mice to induce the development of NASH (N = 5). B Lipid accumulation in liver was detected by Oil Red O staining (N = 3). C A commercially available assay kit was utilized to measure the levels of MDA in liver tissue (N = 5). D Following TEC treatment in NASH mice, the Fe²⁺ content in the liver tissue was measured (N = 5). **P < 0.01

tRF-31R9J is activated by TEC and participates in ferroptosis-related pathways

We have previously reported that tRF is closely linked with fatty liver [24]. Therefore, we further explored whether TEC regulates the ferroptosis of NASH through tRFs. The differentially expressed tRFs in TEC-treated and untreated steatosis hepatocarcinoma cells were obtained using our previous small RNA sequencing results [10]. To identify the differentially expressed tRFs regulated by TEC and involved in ferroptosis, we analyzed the target proteins of the TEC-mediated differentially expressed tRFs using catRAPID database, and performed pathway analysis to screen for tRFs involved in ferroptosis-related signaling pathways, such as ferroptosis and the TGF-β signaling pathway [33]. Next, a network of “differentially expressed tRFs-target protein-ferroptosis-related signaling pathway” was constructed (Fig. 3A). Finally, seven tRFs (tRF-35-KY7343RX6NMH49, tRF-43-Z9HMI8W47W1R7HFEV, tRF-33-389MV47P596V03, tRF-47-RZYQHQ9M739P8WQ0D52, tRF-34-4R94SX73V2Y81W, tRF-29-KSR95R3J09FV, tRF-31-R9JP9P9NH5HYD) that involved in the ferroptosis-related signaling pathway were screened for RT-qPCR validation. The RT-qPCR results of these seven tRFs showed that tRF-31-R9JP9P9NH5HYD (“tRF-31R9J”) was down-regulated in the FFA group, and its expression showed the most significant recovery after TEC treatment (Fig. 3B). These findings suggest that tRF-31R9J may be involved in the regulation of ferroptosis by TEC. Therefore, we have chosen to focus on tRF-31R9J. tRF-31R9J is a 5 'tRF derived from ValTAC tRNA (Fig. 3C), and its function has not been reported in the literature. By analyzing the conservation of sequence, we observed that tRF-31R9J is highly conserved between humans and mice, suggesting its potential importance and functional relevance (Fig. 3D). Based on these findings, we have chosen to investigate tRF-31R9J in subsequent research.

Fig. 3.

TEC treatment results in an elevation of tRF-31R9J expression levels. A Network diagram of “tRF-target protein ferroptosis-related pathway”. B The expression of candidate differentially expressed tRFs in HepG2 cells were quantified by RT-qPCR (N = 3). C The biogenesis diagram of tRF-31R9J. D Conservation of tRF-31R9J sequence in human and mouse. *P < 0.05, **P < 0.01

tRF-31R9J significantly inhibits ferroptosis in hepatocarcinoma cells

To explore the effect of tRF-31R9J on ferroptosis of hepatocarcinoma cells, we employed knockdown and overexpression of tRF-31R9J in steatosis HepG2 cells (Fig. 4A). CCK-8 assay showed that tRF-31R9J knockdown reduced the cell viability; while, overexpression of tRF-31R9J enhanced cell viability (Fig. 4B). The ROS levels in hepatocarcinoma cells were detected. We found that knockdown of tRF-31R9J increased ROS levels, while tRF-31R9J overexpression resulted in a decrease in ROS levels (Fig. 4C). The levels of MDA and Fe²⁺ were significantly elevated in hepatocarcinoma cells after tRF-31R9J knockdown, and the opposite was true for tRF-31R9J overexpression (Fig. 4 D, E). The WB results indicated that knockdown of tRF-31R9J upregulated ACSL4 expression and decreased GPX4 expression (Fig. 4F). Overexpression of tRF-31R9J produced opposite effects (Fig. 4F). The Oil Red O staining exhibited that knocking down tRF-31R9J promoted lipid accumulation in liver cells, while overexpressing tRF-31R9J reduced lipid content (Fig. 4G). Furthermore, we overexpressed tRF-31R9J in steatosis HepG2 and treated cells with the ferroptosis inducer erastin. The results demonstrated that erastin not only reversed the inhibitory effects of tRF-31R9J on ROS and MDA levels but also reduced cell viability (Fig. 4H–J). These results suggest that tRF-31R9J significantly inhibits ferroptosis in steatosis HepG2 cells.

Fig. 4.

tRF-31R9J significantly inhibits ferroptosis in steatosis HepG2 cells. A The efficiency of tRF-31R9J overexpression and knockdown was quantified by RT-qPCR (N = 3). B The impact of tRF-31R9J knockdown or overexpression on steatosis hepatocarcinoma cell viability was carried out by utilizing CCK-8 assay (N = 6). C ROS levels in steatosis HepG2 was measured using fluorescent probe. D The levels of MDA were measured after overexpression or knockdown of tRF-31R9J (N = 3). E The Fe²⁺ content was detected after overexpression and interference of tRF-31R9J (N = 3). F The expression of ferroptosis key molecules was quantified utilizing WB (N = 3). G The effect of tRF-31R9J on lipid accumulation in steatosis HepG2 cells was assessed using Oil Red O staining. H The cellular ROS levels were measured after treatment with erastin. I Following erastin treatment, the levels of MDA were assessed (N = 3). J The effects of tRF-31R9J and erastin on cell viability in liver cells were assessed using the CCK-8 assay (N = 6). *P < 0.05, **P < 0.01

TEC is partially dependent on tRF-31R9J to suppress ferroptosis and NASH

To validate whether TEC suppresses ferroptosis in hepatocarcinoma cells through tRF-31R9J, we treated steatosis HepG2 cells with TEC and performed knockdown of tRF-31R9J. We observed that knockdown of tRF-31R9J significantly attenuated the inhibitory effect of TEC on the levels of MDA and Fe²⁺ (Fig. 5 A, B). The WB results revealed that knocking down tRF-31R9J reversed the inhibitory effect of TEC on ACSL4 and the promotive effect on GPX4 (Fig. 5C). Furthermore, knockdown of tRF-31R9J counteracted the decrease in ROS levels mediated by TEC in steatosis HepG2 cells (Fig. 5D). The Oil Red O staining revealed that knocking down tRF-31R9J significantly reduced the inhibitory effect of TEC on lipid accumulation (Fig. 5E). These findings indicate that TEC suppresses ferroptosis in hepatocarcinoma cells through tRF-31R9J.

Fig. 5.

TEC inhibits ferroptosis in hepatocarcinoma cells through tRF-31R9J. A The effects of tRF-31R9J and TEC on the MDA levels in steatosis HepG2 cells were detected. B The Fe²⁺ content was detected after knockdown of tRF-31R9J. C The expression of ferroptosis key molecules was determined by WB. D ROS levels in steatosis HepG2 were quantified using the fluorescent probe. E The lipid content in HepG2 cells was assessed using Oil Red O staining. N = 3; **P < 0.01

Next, we investigated the effects of TEC and tRF-31R9J on hepatocyte ferroptosis and NASH in vivo. NASH mice were orally gavaged with TCE and simultaneously injected with tRF-31R9J antagomir via the tail vein to knock down tRF-31R9J expression (Fig. 6A). H&E staining showed that liver tissues of the NC group had no lipid droplets, while NASH mice exhibited varying sizes and quantities of lipid droplets and lipid vacuoles in liver cells (Fig. 6B). TEC treatment reduced lipid accumulation, which was partially counteracted by tRF-31R9J knockdown (Fig. 6B). Oil Red O staining demonstrated similar findings, indicating that TEC reduced lipid accumulation in NASH mice, while knockdown of tRF-31R9J resulted in the opposite effect (Fig. 6C). IF was conducted to detect ferroptosis-associated molecules, and the results demonstrated that TEC decreased ACSL4 expression and increased GPX4 expression in NASH mice, while knockdown of tRF-31R9J attenuated this effect (Fig. 6D). Additionally, we measured the lipid metabolism indicators TC and TG. TEC significantly reduced the levels of TC and TG in the liver tissues of NASH mice, while knockdown of tRF-31R9J significantly increased their levels (Fig. 6 E, F). Collectively, TEC, as one of the monomers in anthocyanins, significantly attenuated ferroptosis and NASH, and knockdown of tRF-31R9J partially counteracted the inhibitory effect of TEC on ferroptosis. Thus, TEC partially relied on tRF-31R9J to inhibit ferroptosis.

Fig. 6.

TEC partially relies on tRF-31R9J to inhibit hepatocyte ferroptosis in vivo. A TEC treatment and knockdown of tRF-31R9J were performed in NASH mice. B Liver tissue from NASH mice was subjected to H&E staining to assess histopathological alterations (N = 5). C Oil Red O staining was performed to assess hepatic steatosis (N = 5). D IF was conducted to evaluate the regulation of ACSL4 and GPX4 by tRF-31R9J and TEC. E The effect of tRF-31R9J and TEC on the hepatic TG levels in NASH mice was measured (N = 5). F The impact of tRF-31R9J and TEC on the hepatic TC levels in NASH mice was assessed (N = 5). **P < 0.01

tRF-31R9J binds to HDAC1 to suppress histone lactylation and acetylation modification

To determine the mechanism of tRF-31R9J in inhibiting ferroptosis in hepatocarcinoma cells, we focused on the proteins in the network of “tRF-31R9J-target proteins ferroptosis-related pathway”, as shown in Fig. 3A above. Among these proteins, HDAC1 has captured our interest due to the liver specific recruitment of HDAC1 by METTL3 or WTAP to suppress gene expression and prevent the progression of NASH [30, 34]. We validated the binding of tRF-31R9J to HDAC1 through RNA pull-down and WB (Fig. 7 A, B). The interaction model between tRF-31R9J and HDAC1 was constructed using molecular docking analysis (Fig. 7C). RIP-qPCR results showed that tRF-31R9J was enriched by HDAC1 (Figure S1A). To determine the binding region of HDAC1 to tRF-31R9J, we analyzed the histone deacetylase domain of HDAC1 (9-321aa) (Figure S1B) and its predicted binding region with tRF-31R9J (150-250aa) (Figure S1C), and constructed three flag-tagged HDAC1 overexpression plasmids: flag-HDAC1-FL (full length), flag-HDAC1-△1 (9-321aa, removed non-functional domain, retained complete functional domain), flag-HDAC1-△2 (9–149 + 251-321aa, removed binding domain, retained part of the functional domain), and transfected them into NASH model cells, respectively. RNA pull-down was performed using a tRF-31R9J probe, and flag was detected by WB. The results showed that after removing the binding domain 150-250aa, the binding ability of tRF-31R9J to HDAC1 decreased (Figure S1D). IF and FISH revealed co-localization of tRF-31R9J and HDAC1 within the nucleus (Fig. 7D). GAPDH and U6 served as reference for the cytoplasm and nuclear, respectively (Figure S1E). To demonstrate the interaction between HDAC1 and endogenous tRF-31R9J in cells, we performed interference of HDAC1 in steatosis HepG2 cells. WB and RT-qPCR confirmed that the knockdown efficiency of HDAC1 was significant (Fig. 7E and Figure S1F). Furthermore, FISH results indicated that knockdown of HDAC1 had no significant effect on the expression and localization of nuclear tRF-31R9J (Figure S1G). These results indicate that HDAC1 interacts with tRF-31R9J.

Fig. 7.

tRF-31R9J weakens histone lactylation and acetylation modification by binding to the histone deacetylase HDAC1. A The interacting protein of tRF-31R9J was identified through RNA pull-down. B The enrichment of HDAC1 was validated by WB. C Molecular docking was employed to construct a model of the interaction between tRF-31R9J and HDAC1. D IF and FISH were employed to validate the co-localization of tRF-31R9J and HDAC1. E WB was performed to validate the knockdown efficiency of HDAC1 (N = 3). F The levels of histone lactylation modification were detected by WB (N = 3). G The lactylation level was further verified by IF. H WB was performed to assess the expression of ferroptosis key molecules (N = 3). *P < 0.05, **P < 0.01

Recent reports have shown that HDAC1 act as a de-lactylase for histone lactylation and acetylation [35, 36], which is a novel post-translational modification of histone that plays a part in gene transcriptional regulation [25]. Therefore, we investigated whether tRF-31R9J regulates histone lactylation and acetylation modification through HDAC1. The WB results displayed that tRF-31R9J overexpression resulted in decreased of H3K18la and H3K18ac levels, while knocking down HDAC1 counteracted the inhibitory effect of tRF-31R9J on H3K18la and H3K18ac levels (Fig. 7F and Figure S2A). These results were further verified by IF (Fig. 7G and Figure S2B). To determine the effect of tRF-31R9J recruiting HDAC1 on ferroptosis, we examined the expression of ACSL4 and GPX4. We found that overexpression of tRF-31R9J increased the levels of GPX4 and decreasing the levels of ACSL4, while knockdown of HDAC1 significantly attenuated this effect (Fig. 7H). These findings suggest that tRF-31R9J suppresses histone lactylation and acetylation modification by binding HDAC1.

Knockdown of tRF-31R9J resulted in upregulation of ferroptosis-related genes

To investigate the downstream genes regulated by tRF-31R9J, we divided TEC-treated steatosis hepatocarcinoma cells into a tRF-31R9J knockdown group and a control group for transcriptome sequencing. Heatmap analysis of the transcriptome sequencing data identified a total of 235 differentially expressed genes (DEGs) between the tRF-31R9J knockdown group and the control group. Among these DEGs, 154 genes were up-regulated and 81 genes were downregulated (Fig. 8A). We screened for ferroptosis-related genes by combining the ferroptosis database (FerrDb V2) that contains genes involved in ferroptosis driver and repressor genes (Fig. 8 A, B). The intersection of DEGs with ferroptosis genes in the ferroptosis database was taken, we found that tRF-31R9J knockdown resulted in upregulation of eight ferroptosis-related genes (CHAC1, CBS, ATF4, SLC7A1, SESN2, SLC3A2, ATF3, SLC7A11; Fig. 8B). Subsequently, pathway enrichment analysis was conducted for the up-regulated DEGs. We found that these genes were mainly involved in ferroptosis, PI3K-AKT, and MAPK signaling pathways (Fig. 8C). These findings indicate that tRF-31R9J reduces the expression levels of ferroptosis-related genes.

Fig. 8.

Transcriptome sequencing was performed to investigate the downstream genes and signaling pathways regulated by tRF-31R9J. A Heatmap displayed all DEGs in TEC-treated steatosis hepatocarcinoma cells with tRF-31R9J knockdown. B Transcriptome sequencing combined with ferroptosis database identified tRF-31R9J-regulated ferroptosis-related genes. C Pathway enrichment analysis revealed downstream signaling pathways regulated by tRF-31R9J

tRF-31R9J reduces histone lactylation of ferroptosis driver genes ATF3, ATF4 and CHAC1 by recruiting HDAC1

To investigated whether tRF-31R9J regulates the expression of the above eight genes through HDAC1-mediated histone lactation, we analyzed the ChIP-seq data of H3K18la from the GEO database (GSE208727). Between the control group and the lactate group, a total of 99 differential peaks were identified among the enriched genes, with 86 upregulated and 13 downregulated (|Mval|> 1, P < 0.05) (Fig. 9A). We observed an enrichment of H3K18la peaks near the promoters of pro-ferroptosis genes ATF3, ATF4, and CHAC1 (Fig. 9B). Overexpression of tRF-31R9J and knockdown of HDAC1 were performed in steatosis hepatocarcinoma cells, followed by ChIP-PCR to detect histone lactylation levels of the three genes (Fig. 9 C, D). We found that overexpression of tRF-31R9J reduced the levels of H3K18la near ATF3, ATF4, and CHAC1, while this effect was diminished when HDAC1 was knocked down (Fig. 9 C, D). However, tRF-31R9J and HDAC1 had no significant effect on the histone acetylation levels of ATF3, ATF4, and CHAC1 (Figure S3A). Importantly, compared with acetylation, HDAC1 and tRF-31R9J had a more significant effect on lactylation. The WB results revealed that tRF-31R9J overexpression reduced the protein expression of ATF3, ATF4, and CHAC1, while knockdown of HDAC1 increased their protein levels (Fig. 9E and Figure S3B). Furthermore, silencing of HDAC1 or the addition of lactate reversed the decreased expression of ATF3, ATF4, and CHAC1 induced by tRF-31R9J (Fig. 9F). These findings indicate that tRF-31R9J reduces histone lactylation modification levels of the ferroptosis driver genes ATF3, ATF4 and CHAC1 by binding HDAC1, thereby suppressing gene expression.

Fig. 9.

tRF-31R9J regulates histone lactylation levels near ferroptosis-inducing genes by interacting with HDAC1. A MA plot illustrated the changes in differential peaks of histone lactylation in genes enriched between control and lactate groups in GSE208727 (|Mval|> 1, p < 0.05). B The lactylation modification peaks on the ferroptosis-inducing genes ATF3, ATF4, and CHAC1 were visualized. C The regulation of histone lactylation by tRF-31R9J on ATF3, ATF4, and CHAC1 was verified by H3K18la ChIP-PCR and Gel electrophoresis. D H3K18la ChIP-qPCR was performed to validate the regulation of histone lactylation by tRF-31R9J on ATF3, ATF4, and CHAC1 genes (N = 3). E WB was conducted to examine the influence of HDAC1 on the expression of ATF3, ATF4, and CHAC1 (N = 3). F The effects of tRF-31R9J and lactate on the expression of ATF3, ATF4, and CHAC1 were validated by RT-qPCR (N = 3). *P < 0.05, **P < 0.01

Discussion

Currently, clinical practice faces a shortage of safe and efficacious medications for achieving definitive treatment of NASH. The monomer TEC, a type of anthocyanin, has been found to have hepatoprotective effects [37]. However, the regulatory mechanism of TEC in NASH is not clear. In this study, we discovered that TEC-mediated tRF-31R9J inhibits histone lactylation modification through HDAC1, thereby inhibiting hepatocyte ferroptosis and improving NASH (Fig. 10).

Fig. 10.

Mechanism diagram of TEC-mediated tRF-31R9J improving NASH through histone lactylation modification

Over the past few years, there has been growing recognition of the involvement of tRFs in the pathogenesis of various diseases, including neurodegenerative syndromes, metabolic disorders, and cancer [38]. A study has shown that three types of tsRNAs (tRF-Val-CAC-005, tiRNA-His-GTG-001, and tRF-Ala-CGC-006) in NAFLD patients are significantly elevated, and their expression levels are correlated with NAFLD activity scores and fibrosis staging. Further investigations in mice models have provided evidence of a positive association between the elevated levels of these three tsRNAs and the degree of liver fibrosis [39]. These findings suggest that tsRNAs present in the blood may serve as predictors of fibrosis risk in NAFLD patients. Ying et al. demonstrated that tRF-Gln-CTG-026 improves liver injury by weakening the interaction between TSR1 and the pre-40S ribosome, thus alleviating global protein synthesis [40]. Additionally, an RNA-seq analysis found that in both NAFLD mice and HepG2 cells with fatty liver, there is a decrease observed in the levels of LysTTT-5′tRFs and miR-194-5p, while after treatment with AM132, their levels are transiently restored [41]. These results indicate that LysTTT-5′tRFs and miR-194-5p collectively regulate hepatic steatosis. These findings support the results in this study that tRF-31R9J modulates hepatocyte activity and thereby improves NASH.

There have been no reports on whether tRFs can regulate histone acylation modification. This study for the first time reveals that tRF-31R9J mediates histone lactylation modification through HDAC1, thereby exerting regulatory control over the expression profiles of genes associated with ferroptosis and improving NASH. According to a study conducted, lactate produced by activated hepatic stellate cells promotes gene transcription and subsequently contributes to liver fibrosis through histone lactylation [42]. In the liver, Mettl3 or WTAP prevent the deterioration of NASH by negatively regulating gene transcription through their interaction with HDAC1 [30, 34]. Wang et al. discovered that Danshensu-mediated overexpression of HDAC1 mitigates NAFLD by activating TAF9 [31]. These studies indicate that HDAC1 influences liver diseases through the regulation of histone lactylation. Additionally, there are literature reports on the interactions between HDAC1 and non-coding RNAs. For example, the interaction between lncRNA SNHG1 and HDAC1/2 plays a significant role in the fate of neuroblastoma cells [43]. The non-coding RNA ASPACT interacts with HDAC1, affecting the expression and localization of the PACT transcript [44]. However, no literature has reported the specific regions of HDAC1 that interact with RNA. In this study, we found that the binding ability of HDAC1 to tRF-31R9J was decreased after removal of the binding domain 150-250aa. Together, these results offer strong support to the role of tRF-31R9J in improving NASH through HDAC1-mediated histone lactylation modification.

There are still some limitations in this study. To enhance the credibility of the findings in this study, it is imperative to detect the expression of tRF-31R9J, HDAC1, and lactylation modification in NASH clinical samples. Analyzing whether tRF-31R9J-regulated network molecules can serve as diagnostic biomarkers for NASH, and assessing their correlation with clinical parameters and the expression correlation among different molecules is also essential.

Conclusion

In summary, the findings of this study provide compelling evidence that TEC-mediated tRF-31R9J inhibits the transcription of ferroptosis driver genes ATF3, ATF4, and CHAC1 through the HDAC1/histone lactylation axis, thereby suppressing hepatocyte ferroptosis and improving NASH. This research offers a novel comprehension of the pathogenesis of NASH and provides new molecular targets for the practical implementation of therapeutic strategies targeting NASH in clinical settings.

Author contributions

Conceptualization contributed by J.Z, and X.W.; data curation and validation contributed by X.W.,M.M. and Q.Z.; writing—the original draft contributed by J.Z., writing—review & editing contributed by X.Z. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 82160119 and 82360122), the Regional Fund Cultivation Program of The National Natural Science Foundation of China (NSFC), Affiliated Hospital of Guizhou Medical University (gyfynsfc-2021-30), the Starting fund for doctoral research of Affiliated Hospital of Guizhou Medical University in 2021 (gyfybsky-2021-68), and the Basic Research Project of Science and Technology Department of Guizhou Province (Guizhou Science and Technology Foundation-ZK[2022] General 452).

Availability of data and materials

The data presented in the study (RNA-seq raw data) was deposited in the NCBI repository under accession numbers (PRJNA1184014).

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interest

The authors declare no competing interests.