ABSTRACT
-
Background/Aims
Hepatitis B virus (HBV) hijacks host cell metabolism, especially host glutamine metabolism, to support its replication. Glutamate dehydrogenase 1 (GDH1), a mitochondrial enzyme crucial for glutamine metabolism, can interact with histone demethylases to regulate gene expression through histone methylation. However, the mechanisms underlying GDH1-mediated glutamine metabolism reprogramming and the roles of key metabolites during HBV infection remain unclear.
-
Methods
Transcriptomic and metabolomic analyses of HBV-infected cell were performed. Both HBV-infected cells and humanized liver chimeric mice were used to elucidate the effect of glutamine metabolism on HBV.
-
Results
HBV infection leads to the abnormal activation of glutamine metabolism, including upregulation of key enzymes and metabolites involved in glutamine metabolism. The viral core protein (HBc) mediates the translocation of GDH1 into the nucleus, where GDH1 activates covalently closed circular DNA (cccDNA) transcription by converting glutamate to α-ketoglutarate (αKG). Mechanistically, the promoting effect of GDH1-derived αKG on cccDNA transcription is independent of its conventional role. Rather, αKG directly interacts with the lysine-specific demethylase KDM4A and enhances KDM4A demethylase activity to regulate αKG-dependent histone demethylation, controlling cccDNA transcription.
-
Conclusions
Our findings highlight the importance of glutamine metabolism in HBV transcription and suggest that glutamine deprivation is a potential strategy for silencing cccDNA transcription.
-
Keywords: Glutamine; GDH1; αKG; cccDNA; Methylation
Study Highlights
• HBV infection revealed a marked increase in glutaminolysis, and emphasized the role of glutamine metabolism in activating HBV transcription.
• HBV infection prompts GDH1 nuclear translocation, which relies on HBc; and nuclear GDH1 subsequently triggers αKG accumulation.
• αKG enhanced cccDNA transcription both in vitro and in vivo, via enhancing KDM4A enzymatic activity.
• GDH1 inhibitor exhibits potent antiviral activity in HBV-infected hepatocytes and a humanized liver mouse model.
Graphical Abstract
INTRODUCTION
Despite noteworthy advancements in both the prevention of hepatitis B virus (HBV) infection and treatment of HBV, an astonishing 316 million individuals still suffer from chronic hepatitis B, resulting in approximately 555,000 global fatalities each year [
1]. HBV covalently closed circular DNA (cccDNA) is the template for the transcription of all viral RNAs; HBV cccDNA organizes itself as a minichromosome and plays a pivotal role in HBV transcription and replication [
2]. As an obligate intracellular parasite, HBV does not carry out its own metabolism and relies solely on the host’s metabolism. Therefore, elucidating the metabolic alterations that are triggered by HBV infection is crucial for a deeper understanding of HBV replication mechanisms, particularly cccDNA transcription.
As the most abundant amino acid, glutamine plays a critical role in the process of viral infection. Various viruses, including herpes simplex virus 1, influenza A, and infectious spleen and kidney necrosis virus (ISKNV), trigger glutamine metabolism reprogramming to promote virus replication [
3,
4]. Furthermore, glutamine deprivation can hinder the replication of rabies virus [
5], snakehead vesiculovirus [
6], and hepatitis C virus (HCV) [
7], while glutamine supplementation can restore the replication of these viruses. Additionally, viruses modulate the expression levels of enzymes that govern glutamine metabolism, such as glutaminase (GLS) and glutamate dehydrogenase (GDH) [
7,
8]. Thus, virus-infected host cells exhibit a remarkable increase in glutamine utilization and metabolism. Notably, α-ketoglutarate (αKG), the product of glutamine deamination, plays a pivotal role in various cellular metabolic processes. In addition to the traditional function of αKG in the TCA cycle [
9,
10], recent studies have revealed the critical role of αKG in regulating gene expression through histone demethylation. Wang et al. [
11] demonstrated that αKG reduces the accumulation of H3K9me3 and H3K27me3 in the promoters of BMP signaling genes. Similarly, Chung et al. [
12] reported that αKG maintains low H3K27me3 levels in intrinsic pontine glioma cells. Furthermore, αKG reduces the trimethylation of H3K4 in the promoter of UCP1 [
13]. The above studies suggest that glutamine-derived αKG influences cccDNA histone modification and thereby participates in HBV transcription and replication.
In this study, we found a marked increase in glutaminolysis and enzyme levels in HBV-infected cells and humanized liver mice. Notably, glutamine deprivation or inhibition effectively suppressed HBV transcription and replication. Mechanically, HBV infection prompts GDH1 nuclear translocation, a process that relies on HBc. Nuclear GDH1 subsequently triggers αKG accumulation. Increased levels of αKG then interacts with KDM4A, leading to an increase in KDM4A enzymatic activity. Thereby, the GDH1-αKG-KDM4A downregulates histone methylation on the cccDNA minichromosome and ultimately controls cccDNA transcription. In summary, our data present a novel and intriguing role for GDH1 and its product αKG in mediating active cccDNA transcription.
MATERIALS AND METHODS
Cell culture
NTCP stable expressing HepG2 cell line (HepG2-NTCP) was constructed by our lab; primary human hepatocytes (PHHs) were obtained from ScienCell Research Laboratories (Carlsbad, CA, USA); Huh-7 cell line was purchased from HSRRB; HBV stable expression cell line HepAD38 was kindly provided by Prof. Ningshao Xia (The Xiamen University, Fujian, China). Huh-7 cells were maintained in DMEM (10-010-CVR; Corning, New York, NY, USA) supplemented with 10% fetal bovine serum (FBS) (10270; Corning). HepG2-NTCP cells were maintained in DMEM with 10% FBS and 2.5 μg/mL of doxycycline (ID0670; Solarbio, Beijing, China). HepAD38 cells were maintained in DMEM with 10% FBS and 400 μg/mL of G418 (345810; Merck Millipore, Darmstadt, Germany). PHHs were maintained in hepatocyte medium (5210; ScienCell Research Laboratories). All cells were cultured at 37°C within a humidified atmosphere with 5% CO2.
Plasmids, antibodies, and reagents
Plasmids pCMV6-Entry-GDH1 (RC211132) and pCMV6-Entry-GLS2 (RC212650) was purchased from OriGene (Rockville, MD, USA). Lentivirus expressing wild-type GDH1 or a catalytically inactive form GDH1 R443S were purchased from Shanghai Genechem Co., Ltd. (Shanghai, China). Short hairpin RNA targeting GDH1 (shGDH1-1 and shGDH1-2) and GLS2 (shGLS2-1 and shGLS2-2) or nontargeting shRNA (shCont) were purchased from Shanghai Genechem Co., Ltd. pcDNA3.1-EGFP-WT-GDH1 and pcDNA3.1-EGFP-NLS-GDH1 were purchased from Chongqing Wubai Biotechnology Co., Ltd. (Chongqing, China). pcDAN3.1-myc-HisB(-)-GDH1 WT, pcDAN3.1-myc-HisB(-)-GDH1ΔNAD, pcDAN3.1-myc-HisB(-)-GDH1ΔGLU, and pcDAN3.1-Flag-HBc were constructed by our lab. pCH9/3091, which harbors 1.1-ploid genome length of HBV, was kindly provided from Prof. Lin Lan (The Army Medical University, Chongqing, China). HBc-deleted HBV plasmid (HBVΔHBc) was constructed by introducing a stop codon at the beginning of the HBc gene based on pCH9/3091 (as the wild-type HBV, HBV WT).
Rabbit anti-GDH1 (ab89967), rabbit anti-GLS2 (ab113509), rabbit anti-KDM3B (ab70797), rabbit anti-KDM4A (ab191433P), and anti-KDM6A (ab36938) were purchased from Abcam (Cambridge, UK). Rabbit anti-KDM2A (24311-1-AP) was purchased from Proteintech (Chicago, USA). Rabbit anti-Flag (#14793S), rabbit anti-Histone 3 (2650S) and rabbit anti-KDM7B (#37377) were purchased from Cell Signaling Technology (Boston, MA, USA). Mouse anti-GAPDH (sc-47724) and rabbit anti-actin (sc-1616) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Rabbit anti-H3K4me3 (17-678), rabbit anti-H3K9me3 (17-10242), rabbit anti-H4K20me3 (17-671), rabbit anti-KDM3A (09-823), normal mouse IgG (I8765), and normal rabbit IgG (NI01) were obtained from Merck Millipore. Rabbit anti-HBsAg (NB100-62652), rabbit anti-KDM4B (NB100-74604), and rabbit anti-KDM4D (NBP1-03357) were purchased from Novus (Centennial, CO, USA). Rabbit Anti-KDM7A (A8266), rabbit anti-KDM3C (A20153), and rabbit anti-KDM6B (A12763) were purchased from ABclonal (Woburn, MA, USA). Mouse anti-HBcAg was kindly provided by Prof. Cai (Chongqing Medical University, Chongqing, China).
GDH inhibitor epigallocatechin gallate (EGCG, S2250), transaminases inhibitor aminooxyacetate (AOA, S4989), sodium oxamate (S6871), telaglenastat (CB-839, S7655), entecavir (ETV, S1252), and CP-2 (S8601) were obtained from Selleck Chemicals (Houston, TX, USA). NH4Cl (A9434), non-essential amino acid (M7145), dimethyl-αKG (349631), dimethyl-fumarate (242926), dimethyl-malate (S379239), and dimethyl-succinate (W239607) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Biotin labeled α-KG was purchased from Beijing Fluorescence Biotechnology Co., Ltd. (Beijing, China).
Virus production and cell infection
HBV particles (genotype D) were collected from the supernatant of HepAD38 cells. HBV wild-type virus (HBV WT) and HBV HBc-deficient virus were collected from the supernatant of Huh-7 cells transfected with HBV wild-type plasmids (HBV WT) or cotransfected with HBVΔHBc and Flag-HBc plasmids. Briefly, supernatants of cells were collected and concentrated with 5% polyethylene glycol 8,000 precipitation. Following gentle rotation overnight, HBV particles were precipitated by centrifugation. HBV particles in pellets after centrifugation were re-dissolved in serum-free opti-MEM and stocked at –80°C for further experiments.
For infection, cells were seeded in collagen-coated wells and infected with HBV particles (500 VGEs/cell) which were diluted in infection medium (Williams’ medium supplemented with 10% FBS, 2% dimethyl sulfoxide, 4% PEG8000, and 1% P/S). Twenty-four hours post-infection, infected cells were then washed by phosphate-buffer saline (PBS) twice for further experiments.
Targeted metabolomics profiling
HBV infected HepG2-NTCP cells were collected at indicated time points (day 0, 1, 2, and 3) post infection and the control (non-infected cells) were also harvested at corresponding time points. After removing the medium, each sample was then sequentially washed with each of the following solutions twice: prefilled PBS and prefilled 0.9% sodium chloride. Cells were scraped and resuspended in 1 mL of cold methanol/acetonitrile/water (2:2:1, v:v:v) and transferred to a 1.5-mL centrifuge tube. Following quenched in liquid nitrogen immediately, samples were delivered to Shanghai Applied Protein Technology Co., Ltd. (Shanghai, China). Targeted metabolomics profiling (including amino acid metabolism and energy metabolism) was performed using ultra-high-performance liquid chromatography (Agilent 1290 Infinity LC) coupled with 5500 QTRAP system (AB SCIEX) at Shanghai Applied Protein Technology Co., Ltd.
Glutamine detection
Concentrations of glutamine in cell lysate were determined by Glutamine Colorimetric Assay Kit (K556-100; Bio-Vision) according to the manufacturer’s instructions. Briefly, cultured cells were collected and resuspended in hydrolysis buffer. After centrifugation at 10,000 g for 10 minutes, the supernatant was collected and filtered by 10-kDa molecular weight cut-off spin column. Then, 10 μL of the filtrate and 10 μL of hydrolysis enzyme mix were added into the 96-well plate, subsequently. Following incubation for 30 minutes at 37°C, 50 μL of reaction mix was added into the wells. The plate was incubated at 37°C for 60 minutes, and concentration of glutamine was quantified by reading optical density (OD) value at 450 nm. The standard curve was used to calculate the concentration of glutamine in samples. Glutamine levels were calculated by standard curve.
Intracellular α-ketoglutarate detection
Intracellular levels of αKG were determined by using the α-Ketoglutarate Colorimetric/Fluorometric Assay Kit (K677-100; BioVision) following manufacturer’s instructions. Briefly, cells were collected and rapidly homogenized with 200 μL of ice cold αKG assay buffer. The extracted samples were further deproteinized by passing through a 10-kDa molecular weight cut-off spin column. Fifty microliters of samples were added into 96-well plate for reaction, and 50 μL of reaction mix was added into the sample and incubated for 30 minutes at 37°C. The concentration of αKG was quantified by reading OD value at 570 nm.
Nuclear α-ketoglutarate detection
To determine the level of αKG in nucleus, the nuclei were isolated and subjected to liquid chromatography mass spectrometry (LC-MS). The nuclei were isolated by using the NE-PER Nuclear and Cytoplasmic Extraction Reagents (78833; Thermo, Waltham, MA, USA) following the manufacturer’s instruction. Briefly, cells grown in 10-cm dishes were harvested and washed with PBS twice. The pellet was suspended by 1, 000 μL CRE Ⅰ reagent (containing protease inhibitor). Following vigorous vortexing at the highest setting for 15 seconds, samples were incubated on ice for 10 minutes. Then, 55 μL of ice-cold CER II was added and the mixture was vortexed at the highest setting for 5 seconds. After incubating on ice for 1 minute, samples were centrifuged at 16,000 g for 5 minutes at 4°C. The insoluble (pellet) fraction, which contains nuclei, was suspended in pre-cold 80% methanol and further subjected to LC-MS (AB SCIEX QTRAP 6500).
RNA extraction and Northern blot
Total RNA was extracted by TRNzol (DP424; TIANGEN, Beijing, China) methods. For Northern blot, DIG Northern Starter Kit (12039672910; Roche, Mannheim, Germany) was used. Briefly, the samples were separated by 1.4% formaldehyde-agarose gel and transferred to nylon membrane by capillary siphon method. After pre-hybridization, the membrane was hybridized with digoxigenin-labeled RNA probe at 68°C overnight. Next day, the membrane was washed at 68°C for 15 minutes by following buffers subsequently: washing buffer, 2× SSC/0.1% SDS buffer, and 1× SSC/0.1% SDS buffer. Then, the membrane was incubated in blocking solution and antibody solution at 37°C for 30 minutes, respectively. Finally, the signal was collected by X-ray film.
Nascent RNA synthesis assay
Cells are treated with EU (0.5 mM) for 3 hours to label newly transcribed RNA before harvest. Subsequently, the cells are lysed, and the nascent RNA is captured according to protocols of Click-iT® Nascent RNA Capture Kit (MP10365; Thermo). Then, newly transcribed HBV RNA was determined by real-time polymerase chain reaction (PCR).
RNA degradation experiment
To determine the degradation half-life of HBV RNA, HBV-infected HepG2-NTCP cells were treated with actinomycin D (5 μg/mL) for various durations: 0 hours, 6 hours, 12 hours, and 24 hours. Then, cells are harvested at each time point, and RNA is extracted using TRNzol (DP424; TIANGEN) methods. Finally, HBV RNA levels were determined by real-time PCR.
Protein extraction and Western blot
Cells were lysed by protein lysis buffer (RIPA) with protease inhibitor. After centrifugation, total protein concentration was quantified by using bicinchoninic acid (BCA) assay. Then, denatured protein samples were separated by SDS-PAGE and transferred to PVDF film (10600023; Amersham). Following blocking by 5% nonfat milk, the membrane was incubated with primary antibody overnight at 4°C. After incubation with corresponding secondary antibody, the membrane was subjected to signal visualization by ECL western blot reagents (WBKLS0500; Merck Millipore). GAPDH was used as a loading control.
HBV DNA extraction and Southern blot
HBV infected cells were lysed in 500 μL extraction buffer which contains 10 mM Tris-HCl pH 8.0, 1 mM EDTA, 1% NP-40, and 2% sucrose. After incubation for 15 minutes at 37°C, the nuclei were removed by centrifugation. Ten microliters of cytoplasmic cell lysate was put aside and β-actin was detected to serve as an internal control. Remaining cytoplasmic cell lysate was digested with 40 IU/mL DNase I and 10 mM MgCl2 for 4 hours at 37°C. Then, the HBV core capsids were precipitated with 5% PEG8000 and digested by proteinase K to release HBV DNA. Finally, the HBV DNA was purified by phenol/chloroform (1:1) and precipitated with ethanol. For liver samples, 10–20 mg liver tissues were homogenized in 500 μL of extraction buffer and incubated for 20 minutes at 37°C. The subsequent steps are same as the cells.
For absolute quantification PCR, Fast Start Universal SYBR Green Master (06924204001; Roche) was used and the specific primers are listed in
Supplementary Table 1.
For Southern blot, DIG-High Prime DNA Labeling and Detection Starter Kit (11585614910; Roche) were used. Briefly, the samples were separated by agarose gels and transferred onto nylon membrane by capillary siphon method. After fixed by UV cross-linking, the membrane was hybridized with digoxigenin-labeled DNA probe overnight at 42°C. Next day, the membrane was washed by washing buffer, 2× SSC/0.1% SDS buffer, 1× SSC/0.1% SDS buffer, and 0.1× SSC/0.1% SDS buffer, subsequently. Then, the membrane was blocked in blocking solution and incubated with anti-digoxin secondary antibody at 37°C for 30 minutes. Finally, the signal was collected by X-ray film.
cccDNA isolation and detection
cccDNA was isolated by using modified Hirt method. Briefly, cultured cells were lysed in 500 μL SDS lysis buffer which contains 50 mM Tris-HCl, pH 8.0, 10 mM EDTA, 150 mM NaCl, and 1% SDS. After incubation at 37°C for 20 minutes, 125 μL 2.5 M KCl was added and incubated at 4°C overnight. Next day, samples were centrifuged at 12,000 g for 20 minutes to collect the supernatant. Then, cccDNA was purified by phenol/chloroform (1:1) and precipitated by ethanol. For liver tissues, 10–20 mg liver tissues were homogenized in 500 μL SDS lysis buffer which contains 20 mM Tris·HCl, pH 8.0, 5 mM EDTA, 400 mM NaCl, and 1% SDS and incubated with 100 μg/mL protease K at 37°C overnight. Next day, DNA was purified by phenol/chloroform (1:1) and precipitated with ethanol.
For TaqMan probe qRT-PCR detection, samples were pretreated as follows: (1) rcDNA denaturation: heating at 80°C for 5 minutes and cooling immediately; (2) rcDNA digestion: incubating with exonuclease V at 37°C for 30 minutes; (3) cccDNA denaturation: heating at 100°C for 20 minutes. After pretreatment, samples were subjected to TaqMan probe qRT-PCR. The selective primers and probe are in
Supplementary Table 1.
For Southern blot, DIG-High Prime DNA Labeling and Detection Starter Kit (11585614910; Roche) was used. Briefly, the samples were separated by agarose gels and transferred on nylon membrane by capillary siphon method. After fixing by UV cross-linking, the membrane was hybridized with digoxigenin-labeled DNA probe overnight at 42°C. Next day, the membrane was washed by washing buffer, 2× SSC/0.1% SDS buffer, 1× SSC/0.1% SDS buffer, and 0.1× SSC/0.1% SDS buffer, subsequently. Then, the membrane was blocked in blocking solution and incubated with anti-digoxin secondary antibody at 37°C for 30 minutes. Finally, the signal was collected by X-ray film.
Proximity ligation assay (PLA)
PLA assays were performed using the Duolink PLA kit (Sigma-Aldrich) according to the manufacturer’s instructions. Briefly, Huh-7 cells were cultured on microcover glass in 24-well plates after transfection. Then, cells were fixed with 4% paraformaldehyde for 15 minutes at room temperature and washed in PBS. After 15 minutes of infiltration with 0.5% Triton X-100, cells were blocked in Duolink blocking buffer for 1 hour. Cells were incubated overnight at 4°C with primary antibodies (anti-Flag, F1804 [Sigma-Aldrich]; anti-His, M30111 [Abmart, Shanghai, China]; anti-GDH1, 12793S [Cell Signaling Technology]) diluted in Duolink antibody diluent. After washing, cells were incubated in appropriate Duolink secondary antibodies for 1 hour at 37°C. Then, the PLA ligation and amplification steps were performed using the Duolink in situ detection reagent FarRed. Cells were covered with DAPI mounting medium, and images were acquired using confocal microscopy (DMi8; Leica, Weztlar, Germany).
KDM4A enzymatic assay
To detect the enzymatic activity of KDM4A, we first employed protein immunoprecipitation to enrich KDM4A from the cells. Subsequently, 20% of the enriched product was subjected to Western blot analysis. Based on the Western blot results, an equal amount of protein was selected for further experiments. Assay was performed by using JMJD2A (KDM4A) Homogenous Assay Kit (#50413; BPS Bioscience, Milan, Italy) according to the manufacturer’s instructions. Briefly, 12 μL of cell lysate were added into a 96-well plate, with an equal volume of diluent solution added to the blank control wells. Subsequently, 16 μL of diluted KDM4A (25 ng/μL) were added to each well and incubated at room temperature for 30 minutes. After incubation, 12 μL of Master Mix were added to each well and incubated for an additional hour at room temperature. Then, mix of acceptor beads (diluted 1:500) with Primary Antibody 5 (diluted 1:200) and 10 μL of the mixture were added to each well, followed by incubation at room temperature for 30 minutes. Afterward, add 10 μL of donor beads and incubate for another 30 minutes at room temperature. Finally, fluorescence was measured with a BioTek Synergy H1 fully functional microplate detector (Agilent Technologies, Santa Clara, CA, USA) at excitation wavelength of 680 nm and emission wavelength of 615 nm. The intensity of these signals is directly proportional to the activity of KDM4A enzyme, allowing for quantitative assessment of enzyme activity.
Immunofluorescence staining
Cells grown on coverslip were fixed in 4% paraformaldehyde for 15 minutes and permeabilized in 0.5% Triton X-100 for 30 minutes. After blocking by normal goat serum for 1 hour at room temperature, cells were first incubated with mouse anti-HBcAg (1:200) at 4°C overnight. Next day, cells were incubated with anti-GDH1 (1:500) at room temperature for 2 hours. Subsequently, secondary antibodies conjugated with Alexa Fluor 680 (goat anti-mouse IgG, 1:1,000) were incubated at room temperature for 2 hours. Following washing by PBS, cells were incubated with DAPI for nuclear counterstaining and images were captured by confocal laser scanning microscope (DMi8; Leica).
Chromatin immunoprecipitation (ChIP)
ChIP assays were performed according to the manufacturer’s instructions (17-10086; Merck Millipore). Briefly, cells were scraped with prechilled PBS (containing protease inhibitor) and centrifuged at 800
g, 4°C for 5 minutes. After resuspension in ChIP cell lysis buffer, the nuclei were collected by centrifugation. For detection of the enrichment of histones on cccDNA, the nuclei pellet was resuspended in nuclear lysis buffer and subjected to sonication, directly. For detection of the enrichment of non-histone proteins on cccDNA, the nuclei were first fixed in 1% formaldehyde for 10 minutes at room temperature before sonication. After centrifugation, the sonication product was diluted by dilution buffer (1:10) and 5 μL of sample was put aside as input. The rest was subjected to immunoprecipitation. Finally, immunoprecipitated cccDNA was detected by TaqMan probe qRT-PCR. The selective primers are listed in
Supplementary Table 1. Normal IgG from same host was used as control to exclude the non-specific binding. The detailed information about antibodies used in ChIP assays were listed in
Supplementary Table 1.
His-Tag-KDM4A WT and His-Tag-KDM4Amut were labeled using the Mothlith His-Tag Labeling Kit RED-tris-NTA 2nd Generation (MO-L018; NanoTemper Technologies, München, Germany). Briefly, 90 μL of a solution of His-Tag protein (200 nM) was mixed with 90 μL dye (100 nM) in labeling buffer and incubated for 30 minutes at room temperature. Following centrifugation at 15,000 g for 10 minutes at 4°C, supernatant containing labeled protein was removed into a new tube for further experiments.
For MST assay (Monolith NT.115; NanoTemper Technologies), following three steps were conducted: (1) Pretest: 10 μL of labeled His-Tag protein was loaded into glass capillaries and the adsorption of labeled His-Tag protein to capillary walls was determined. (2) Binding check assay: 10 μL of labeled His-Tag protein were mixed with 10 μL of αKG (40 μM). Following incubation at room temperature for 10 minutes, the mixture was subjected to binding check assay. (3) Binding affinity: For αKG, the two-fold serial dilutions were made starting from 1,280 μM to 39.0625 nM in 16 steps. And 10 μL of labeled His-Tag protein were incubated with serial dilutions of αKG at room temperature for 10 minutes. Then, the mixture was loaded into glass capillaries and the MST analysis was performed using Monolith NT.115 (Nano-Temper Technologies) at 25°C.
Establishment of mouse model with HBV infection and drug administration
Alb-Cre transgenic mice (C57BL/6-Tg [Alb-cre] 21Mgn/J) were obtained from Shanghai Model Organisms Center, Inc. (Shanghai, China) and housed at Laboratory Animal Center of the Chongqing Medical University with pathogen-free conditions. For HBV infection, 4 μg prcccDNA was diluted in PBS as 8% of the mouse body weight and injected through tail veins within 5 to 8 seconds. Serum HBV DNA level was detected at one week post injection. Then, the mice were assigned to three groups randomly: vehicle group (n=8), EGCG low-dose group (10 mg/kg every other day, n=8), and EGCG high-dose group (20 mg/kg every other day, n=8). Serum samples were collected at indicated time points. At last time point, liver tissues were collected after sacrifice and frozen in liquid nitrogen or fixed by formalin for further study. Animal studies conformed to the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines and were approved by Laboratory Animal Center of the Chongqing Medical University (approval No. IACUC-CQMU-2024-0315).
Humanized mouse infection and drug administration
Human liver-chimeric Alb-uPA/SCID mice were purchased from Beijing Vitalstar Biotechnology Co., Ltd. (Beijing, China). Approved by the ethical committees at Chongqing Medical University (approval No. 2024010), serum of treatment-naïve hepatitis B patient was collected with written informed consent.
After one week of adaptive house, 50 μL of patient serum (genotype B, 2.42×106 copies/mL) were injected into mouse through tail vein. Six weeks post injection, the mice were treated with αKG or EGCG and corresponding vehicle. Blood samples were collected via orbital sinus at indicated time point. The mice were sacrificed post 42 days or 28 days of administration, and liver tissues were collected for further study. Animal studies conformed to the ARRIVE guidelines and were approved by the Laboratory Animal Center of Chongqing Medical University (approval No. IACUC-CQMU-2024-0315).
Immunohistochemistry (IHC)
Sections were deparaffinized, rehydrated, and incubated with indicated primary antibody at 4°C overnight. Followed counterstained with secondary antibody for 30 minutes at room temperature, slides were stained with diaminobenzidine (DAB) to determine immunoreactivity (Dako, Carpinteria, CA, USA). After washing, slides were counterstained with hematoxylin and images were captured by a microscope.
Statistical analysis
The results are expressed as the mean±standard deviation. Statistical analyses were performed using either Student’s t-test or the Mann–Whitney U test and by one-way ANOVA. Differences were considered significant when P<0.05. Statistical analysis and graphical presentation were performed by using Prism 8 (GraphPad Software).
RESULTS
HBV infection alters host cell glutamine utilization
To investigate the molecular mechanisms underlying HBV transcription, we first conducted a comprehensive analysis of whole-genome expression patterns in HBV-infected HepG2-NTCP cells over time. Our analysis revealed a total of 16,707 differentially expressed genes (DEGs) at four distinct time points (
Fig. 1A,
Supplementary Fig. 1A). Notably, these DEGs exhibited four distinct nonlinear expression patterns (
Fig. 1B). Specifically, the 4,045 genes in Cluster 4 were consistently upregulated from 12 to 72 hours, closely mirroring changes in cccDNA levels during HBV infection (
Supplementary Fig. 1B). Given this observation, we conducted a GO enrichment analysis of the specific DEGs in Cluster 4, which revealed a preponderance of the enrichment of mostly metabolism-related pathways in these DEGs (
Supplementary Fig. 1C). Notably, the glutamine family amino acid metabolic process, a typical amino acid metabolism pathway, was continuously enriched throughout HBV infection (
Fig. 1C,
1D), highlighting the potential importance of glutamine metabolism in HBV infection.
Consistent with these results, targeted metabolomics analysis (amino acids) revealed that glutamine levels were persistently elevated during HBV infection (
Fig. 1E). Additionally, a joint analysis of transcriptomic and metabolomic data was conducted, and glutamine metabolism-related genes were mapped to their respective metabolic pathways. Notably, glutamine levels increased during HBV infection, accompanied by a significant upregulation of key enzymes involved in glutamine metabolism, including glutamate dehydrogenase 1 (GDH1), glutaminase 2 (GLS2), carbamoyl-phosphate synthetase 2 (CAD), and glutamateammonia ligase (GLUL) (
Fig. 1F). The upregulation of those enzymes was further confirmed in HBV-infected HepG2-NTCP cells and PHHs (
Supplementary Fig. 1D), which were closely related to HBV transcription and replication (
Supplementary Fig. 1E,
1F). In summary, our study revealed that HBV infection modifies glutamine utilization in host cells.
To validate the effect of HBV infection on stimulating glutamine metabolism, we first observed a significant elevation of intracellular glutamine in HBV-infected HepG2-NTCP cells compared with mock-infected cells (
Fig. 2A). By culturing HBV-infected HepG2-NTCP or PHHs in either complete or glutamine-free medium for 24 hours, we found that glutamine deprivation led to a marked reduction in HBV 3.5-kb RNA (
Fig. 2B,
Supplementary Fig. 2A). Given that cccDNA functions as the template for HBV RNA, we specifically assessed the impact of glutamine deprivation on cccDNA transcription. Interestingly, glutamine deprivation significantly suppressed cccDNA transcription (ratio of HBV 3.5-kb RNA to cccDNA and total HBV RNA to cccDNA) without altering the cccDNA level (
Fig. 2C,
Supplementary Fig. 2C). Additionally, glutamine deprivation resulted in decreased HBV DNA levels (
Supplementary Fig. 2C), as well as decreased HBsAg and HBeAg levels (
Supplementary Fig. 2D). Notably, CB-839 (a specific GLS inhibitor) potently reduced HBV 3.5-kb RNA and DNA levels, whereas oxamate (an inhibitor of lactate dehydrogenase) caused a more moderate reduction in HBV 3.5-kb RNA and DNA (
Supplementary Fig. 2E,
2F), indicating the crucial role of glutamine metabolism in HBV replication. Furthermore, the knockdown of GLS2, an enzyme that catalyzes the conversion of glutamine to glutamate, significantly decreased HBV RNA, HBV DNA, HBsAg, and HBeAg levels in HBV-infected cells (
Supplementary Fig. 3A–
3E). Notably, knockdown of GLS2 also led to the profound suppression of cccDNA transcription (
Supplementary Fig. 3F), further confirming the critical role of glutaminolysis in HBV replication.
Glutamine serves as a carbon source via αKG, and also provides nitrogen for nonessential amino acid (NEAA), and ammonia (
Fig. 2D). HBV-infected HepG2-NTCP cells or PHHs were cultured in glutamine-free medium and supplemented with dimethyl-αKG (DM-αKG), a cell-permeable analog of αKG; NEAAs; or ammonia. Surprisingly, only DM-αKG restored HBV RNA, DNA, HBsAg, and HBeAg levels in glutamine-free medium (
Fig. 2E,
Supplementary Fig. 4A,
4B), whereas the cccDNA levels remained unchanged (
Fig. 2F). Furthermore, DM-αKG increased the levels of HBV RNA, DNA, HBsAg, and HBeAg in a dose-dependent manner in cells cultured with glutamine-free medium, whereas the cccDNA levels remained unchanged (
Fig. 2G–
2I,
Supplementary Fig. 4C–
4E). Using carbon-tracing experiments, we detected increased incorporation of
13C-labeled carbon into M+5 αKG and M+5 Glu during HBV infection, but no changes in other TCA cycle intermediates, such as succinate, fumarate, and malate (
Fig. 2J), as well as metabolites involved in lactate metabolism, including M+1, M+2, and M+3 pyruvate, as well as M+1, M+2, and M+3 lactate (
Supplementary Fig. 4F). These findings suggest that glutamine metabolism, especially the conversion of glutamine to αKG, plays a crucial role in HBV transcription.
Glutamine is metabolized to αKG via GDH1 or aminotransferases (
Supplementary Fig. 5A). Therefore, we studied their specific role in HBV replication via treatment with EGCG (a GDH1 inhibitor) or AOA (a transaminase inhibitor). EGCG reduced the levels of HBV RNA and DNA, but not AOA (
Fig. 3A,
Supplementary Fig. 5B–
5D). Notably, neither EGCG nor AOA treatments altered the cccDNA level (
Fig. 3B,
Supplementary Fig. 5E), indicating that glutamine affects HBV replication via GDH1-mediated αKG production. As a key enzyme in glutamine metabolism, GDH1 catalyzes glutaminolysis, which is essential for cell proliferation. We explored the effect of GDH1 on cellular glutamine level and growth in HepG2-NTCP cells without HBV infection. The data showed that GDH1 is involved in cellular glutamine metabolism and growth, which is independent of HBV infection (
Supplementary Fig. 5F–
5I). We also tracked GDH1 expression during HBV infection and found that HBV replication significantly increased the mRNA and protein levels of GDH1 (
Fig. 3C,
3D,
Supplementary Fig. 6A).
To further investigate the role of endogenous GDH1 in HBV transcription and replication, we transduced HBV-infected cells with lentiviruses expressing shRNAs specifically targeting GDH1 (shGDH1-1 or shGDH1-2) (
Supplementary Fig. 6B). We found that GDH1 silencing significantly reduced HBV RNA (
Fig. 3E). Further experiments revealed that GDH1 depletion did not affect the half-life of HBV 3.5-kb RNA (
Fig. 3F), but obviously decreased the synthesis of HBV 3.5-kb RNA synthesis (
Fig. 3G), indicating that GDH1 depletion impaired new HBV RNA synthesis without affecting RNA stability. Additionally, GDH1 knockdown reduced cccDNA transcription, without affecting cccDNA levels (
Fig. 3H,
3I). Moreover, GDH1 depletion reduced HBV DNA, HBsAg and HBeAg (
Supplementary Fig. 6C–
6E). Conversely, overexpression of wild-type GDH1, but not an inactive mutant of GDH1 (GDH1
mut, R443S), promoted HBV transcription and replication (
Supplementary Fig. 7). In summary, GDH1 stimulates HBV transcription and replication in an enzymatic activity dependent manner.
Since cccDNA transcription occurs within the nucleus, we analyzed the localization of GDH1 in HepG2-NTCP cells. Immunofluorescence microscopy revealed that GDH1 was localized primarily in the cytoplasm, but specific nuclear GDH1 localization was observed in HBV-infected cells (
Fig. 4A). Compared with wild-type GDH1 (WT-GDH1) overexpression, nuclear GDH1 (NLS-GDH1) overexpression led to higher HBV RNA and DNA levels (
Supplementary Fig. 8A–
8C). Notably, ChIP assays revealed that endogenous GDH1 can bind cccDNA, with
GAPDH and
MYH6 used as host controls (
Supplementary Fig. 8D). Moreover, ChIP-Seq experiments showed that the enrichment of GDH1 in HBV-specific reads (
Fig. 4B), indicating the role of GDH1 in cccDNA transcription. To understand the mechanism underlying GDH1 nuclear translocation, we overexpressed four viral proteins, the core protein (HBc), X protein (HBx), polymerase (HBp), and S antigen (HBs) in HepG2-NTCP cells. Notably, HBc overexpression specifically triggered the nuclear translocation of GDH1 (
Fig. 4C). Reciprocal coimmunoprecipitation and proximity ligation assays confirmed the interaction between GDH1 and HBc (
Fig. 4D,
4E). To identify the specific region in GDH1 that binds HBc, we then generated truncated clones of GDH1 designed on the basis of its structural domains: His-GDH1 (aa 1–558), His-ΔNAD (aa 1–261), and His-ΔGLU (aa 262–558). Intriguingly, we found that full-length GDH1 and His-ΔGLU, but not His-ΔNAD, bound to HBc (
Fig. 4F,
4G). These findings suggest that the NAD domain of GDH1 plays a pivotal role in mediating the interaction of GDH1 with HBc.
To explore the role of HBc in GDH1-mediated cccDNA transcription, we generated an HBc-deficient virus (HBVΔHBc) (
Supplementary Fig. 8E). We found that GDH1 nuclear translocation was significantly decreased in HBVΔHBc-infected cells compared with that in HBV-WT-infected cells (
Fig. 4H), and GDH1 enrichment on cccDNA was decreased (
Fig. 4I). Neither the overexpression nor the knockdown of GDH1 altered HBV RNA or DNA levels in HBVΔHBc-infected cells (
Fig. 4J,
Supplementary Fig. 8F–
8I). These results suggest that HBc-mediated GDH1 nuclear translocation is crucial for cccDNA transcription.
To study the functional role of αKG in nuclear GDH1-mediated cccDNA transcription, LC-MS-based
13C tracing experiments were conducted. We found reduced incorporation of
13C into M+5 αKG and glutamate was reduced in GDH1-knockdown cells but increased in GDH1-overexpressing cells. Glutamate serves as a precursor for the synthesis of non-essential amino acids and glutathione (
Supplementary Fig. 9A). Notably, the levels of M+1 aspartate, M+1 alanine and M+2 glutathione partially increase in GDH1 depletion cells (
Supplementary Fig. 9B). Therefore, we speculate that the decrease of M+5 glutamate in GDH1-silencing cells is caused by the activation of the metabolic pathways in which glutamate serves as a precursor for the synthesis of non-essential amino acids and glutathione. Conversely, the opposite results are also observed in GDH1-overexpression cells (
Supplementary Fig. 9C). However, succinate, fumarate, and malate levels were unchanged (
Fig. 5A). More importantly, nuclear αKG levels were decreased in GDH1-silenced cells but increased in GDH1-overexpressing cells (
Fig. 5B). Notably, DM-αKG reversed the decrease in HBV RNA and DNA levels induced by GDH1 depletion, but other TCA intermediates did not have this effect (
Fig. 5C,
Supplementary Fig. 9D). Additionally, the cccDNA levels remained unchanged (
Fig. 5D).
Then, we treated HBV-infected HepG2-NTCP cells or PHHs with increasing concentrations of DM-αKG. As expected, DM-αKG treatment increased the HBV 3.5-kb RNA, DNA, HBsAg, and HBeAg levels in a dose-dependent manner, but did not affect cccDNA level (
Fig. 5E,
Supplementary Fig. 9E–
9G). Further experiments showed that DM-αKG promoted new HBV RNA synthesis without affecting its stability (
Fig. 5F,
5G). Next, human liver chimeric uPA/SCID mice were given dietary αKG 6 weeks after HBV virion injection (
Fig. 5H,
Supplementary Fig. 9H). Blood samples were collected every 7 days, and liver tissues were harvested after 6 weeks of DM-αKG treatment. Body weight was monitored every two days, and blood parameters (including white blood cells, red blood cells, hemoglobin, platelets, total protein, albumin, alanine transaminase, aspartate transaminase, alkaline phosphatase, γ-glutamyl transpeptidase, total bilirubin, creatinine and blood urea nitrogen) were examined on day 42. The results revealed no significant cytotoxicity after DM-αKG administration (
Fig. 5I,
Supplementary Fig. 9I,
Supplementary Table 2). DM-αKG supplementation increased HBsAg, HBV DNA levels in serum, and HBV DNA and RNA levels in liver, but did not affect cccDNA levels (
Fig. 5J–
5L,
Supplementary Fig. 9J,
9K). IHC revealed a significant increase of HBc in hepatocytes treated with DM-αKG (
Fig. 5M). Collectively, these findings suggest that αKG supports HBV transcription and replication both in vitro and in vivo.
Histone lysine demethylases (KDMs) are grouped into αKG-dependent (KDM2-KDM8) and FAD-dependent (KDM1) KDMs. We explored the influence of GDH1-derived αKG on cccDNA histone methylation. GDH1 silencing significantly reduced the enzymatic activity of KDM4A, but not other αKG-regulated KDMs, or FAD-dependent KDM1A-B (
Fig. 6A). Conversely, GDH1 overexpression enhanced the enzymatic activity of KDM4A (
Fig. 6B). However, KDM protein levels remained unchanged (
Supplementary Fig. 10A). Furthermore, reciprocal coimmunoprecipitation experiments revealed that GDH1 specifically interacts with KD M4A (
Fig. 6C,
Supplementary Fig. 10B). Moreover, a PLA confirmed that GDH1 and KDM4A interact primarily in the nucleus (
Fig. 6D), indicating that GDH1-derived αKG regulates cccDNA transcription by activating KDM4A enzymatic activity. To further investigate the role of KDM4A in HBV transcription and replication, we transduced HBV-infected cells with lentiviruses expressing shRNAs specifically targeting KDM4A (shKDM4A-1 or shKDM4A-2) (
Supplementary Fig. 11A). We found that KDM4A silencing significantly reduced HBV RNA, HBV DNA and cccDNA transcription (
Supplementary Fig. 11B–
11D). In contrast, overexpression of KDM4A promoted HBV transcription and replication (
Supplementary Fig. 11E–
11H). To further explore whether the impact of αKG on cccDNA transcription is mediated by KDM4A, we depleted KDM4A in DM-αKG treated cells. As expected, depletion of KDM4A almost abolished DM-αKG-induced enhancement of HBV transcription and replication (
Supplementary Fig. 11I–
11K). Similar results were also obtained in DM-αKG-treated cells which further exposed to KDM4A inhibitor CP-2 (
Supplementary Fig. 11L–
11N), confirming the requirement of KDM4A for αKG-mediated cccDNA transcription. We also used microscale thermophoresis to investigate the binding of αKG with KDM4A. Briefly, fluorescent His tagged wild-type KDM4A (KDM4A
WT) or inactive mutant form of KDM4A (KDM4A
mut, E190A) was mixed with 10 μL of αKG (40 μM), followed by measurements of MST traces. The results showed that the migration of fluorescent KDM4A
WT changed significantly in the presence of αKG, with a signalto-noise ratio of 18.9 (a signal-to-noise ratio ≥5 is consid-ered positive binding) (
Fig. 6E, left panel). Subsequently, the binding curve for the interaction of KDM4A
WT with αKG yielded a Kd of 15.3 (
Fig. 6E, right panel). However, MST experiments showed no binding between KDM4A
mut protein and αKG (
Fig. 6F), indicating that the residue glutamate 190 (E190) within KDM4A is critical for αKG binding. Moreover, we elucidated the functional consequence on KDM4A enzymatic activity. Expectedly, αKG is unable to effectively modulate the enzymatic activity of mutant type KDM4A in which the binding site (glutamate 190) is disrupted (
Supplementary Fig. 12A). Importantly, only wild-type KDM4A (KDM4A
WT), not inactive mutant KDM4A (KDM4A
mut, E190A), promoted HBV transcription and replication (
Supplementary Fig. 12B–
12E).
Next, a ChIP assay showed that KDM4A binds to cccDNA (
Fig. 6G), indicating that KDM4A demethylates cccDNA-bound histones. HBV-infected HepG2-NTCP cells overexpressing GDH1 or treated with DM-αKG presented reduced recruitment of H3K4me3, H3K9me3, and H4K-20me3 (demethylated by KDM4A) to cccDNA, whereas H3K4me1/2 levels (demethylated by KDM1) were unaffected (
Fig. 6H). Conversely, GDH1 knockdown increased these methylations on cccDNA, which could be abolished by DM-αKG (
Supplementary Fig. 12F). Furthermore, ChIP-seq analysis revealed that the regions on cccDNA enriched with H3K4me3, H3K9me3, and H4K20me3 in GDH1-overexpressing cells were consistent with those in DM-αKG-treated cells (
Fig. 6I). Additionally, a dual-luciferase reporter assay showed that GDH1 overexpression enhanced the activities of HBV core promoter, X promoter, SpI and SpII, and had no significant effect on enhancer I and II activity (
Supplementary Fig. 13A). It is worth noting that although GDH1 overexpression did not significantly affect the expression levels of transcription factors, including HNF1α, HNF4α, CEBP, CREB, and FXRα (
Supplementary Fig. 13B,
13C), it can increase the enrichment of transcription factors on cccDNA (
Supplementary Fig. 13D). The same results were also obtained in DM-αKG treatment cells (
Supplementary Fig. 13A–
13D). Conversely, GDH1 silencing decreased the enrichment of those transcription factors on cccDNA, which can be restored by DM-αKG treatment (
Supplementary Fig. 13E). In summary, those findings suggested that their regulatory effects on HBV transcription are partly dependent on the recruitment of transcription factors to cccDNA.
To comprehensively investigate the impact of glutamine on HBV transcription and replication in vivo, human liver-chimeric mice injected with HBV virions were divided into the following groups: the control group, the glutamine-deprived group (no Gln), and DM-αKG replenished group (no Gln+DM-αKG). Throughout the experiment, we closely monitored serum human albumin levels (
Fig. 7A). Glutamine deprivation reduced serum HBsAg, HBV DNA (
Fig. 7B,
7C), liver HBV RNA, 3.5-kb RNA, and HBV DNA (
Fig. 7D,
7E), which was restored by DM-αKG replenishment. However, cccDNA levels remained stable (
Fig. 7F).
To assess the antiviral activity of EGCG in vivo, we tested the effect of EGCG in Alb-Cre transgenic mice and human liver-chimeric uPA/SCID mice. Alb-Cre mice were injected with prcccDNA to establish an HBV replication mouse model (
Supplementary Fig. 14A). The mice subsequently received 0, 10, or 20 mg/kg EGCG every other day for 21 days. Body weight and blood parameters suggested that EGCG showed no significant cytotoxicity, even at 20 mg/kg (
Supplementary Table 3). Importantly, EGCG reduced serum HBV DNA levels, as well as HBV RNA and DNA in liver (
Supplementary Fig. 14B–
14E). Owing to the effective ability of 10 mg/kg EGCG to effectively inhibit HBV with minimal cytotoxicity, EGCG was chosen to be used at this dose for further study.
Human liver-chimeric uPA/SCID mice were injected with HBV virions and split into four groups: control group, EGCG group, ETV group, and EGCG+ETV group. Serum human albumin was tracked (
Fig. 7G). Compared to the control, EGCG and the combined group lowered serum HBsAg and HBV DNA levels (
Fig. 7H,
7I). They also reduced HBV RNA, 3.5-kb RNA, and DNA in the liver. However, cccDNA levels remained unchanged (
Fig. 7J,
7K). Notably, the EGCG+ETV combination excelled in inhibiting HBV DNA. In summary, GDH1 inhibitor EGCG suppressed HBV transcription and replication in vivo.
DISCUSSION
Viruses, particularly oncogenic viruses, can hijack host metabolic pathways, including glycolysis, glutaminolysis, and fatty acid synthesis (FAS), to secure energy and biosynthetic necessities for their efficient replication and the survival of infected cells. Research has shown that glycolysis, glutaminolysis, and FAS are indispensable for optimal Kaposi’s sarcoma-associated herpesvirus (KSHV) production, and inhibitors of these cellular metabolic pathways effectively block the generation of infectious viruses from infected cells [
14-
17]. Similarly, the reprogramming of glutamine metabolism through MYC activation induced by adenovirus or HCV fuels virus replication [
4,
7,
18].
However, the metabolic landscape during HBV infection remains largely unexplored. Although few studies have offered insight into the metabolic alterations that are triggered by HBV, there are notable gaps in our understanding. For example, metabolic profiling of HepG2.2.15 cells revealed that HBV infection significantly increases the biosynthesis of hexosamine and phosphatidylcholine [
19]. Additionally, HBV stimulates central carbon metabolism and nucleotide synthesis [
20], whereas HBV antigen expression in transgenic mice affects lipid metabolism in the liver [
21]. Nevertheless, these studies have certain limitations. Given that HBV DNA integrates into the host cell chromosomes, models such as the HepG2.2.15 cell line or HBV transgenic mice do not fully recapitulate natural infection with HBV. Consequently, analyzing the metabolic profiles of such cells or mouse models may not accurately reflect the metabolic alterations in hepatocytes infected with HBV particles. In this study, we demonstrated that glutamine metabolism is enhanced in HBV-infected cells. Glutamine deprivation or GLS knockdown led to a significant inhibition of HBV replication, particularly affecting cccDNA transcription. These results align with those of previous studies, which indicated a positive correlation between elevated glutamine levels and HBV DNA levels in both acute and chronic hepatitis patients [
22]. Furthermore, our study revealed that HBV replication in hepatocytes triggers the expression of key enzymes involved in glutamine metabolism, including GLS2, GOT1, and GDH1. However, the exact mechanism by which HBV elevates the levels of these enzymes to activate glutamine metabolism remains elusive. While the specific mechanisms by which HBV infection upregulates glutamine metabolism enzymes are yet to be elucidated, the mechanism have been explored in several other viruses. For instance, ISKNV upregulates the gene expression of GLS1 and GDH through mTOR pathway [
8]. Additionally, c-Myc has been implicated in upregulating GDH1 protein levels in Epstein–Barr virus or human herpesvirus 4 infected cells [
23]. Prior studies have demonstrated that Myc activates the transcription of genes vital for glutamine uptake and metabolism, leading to cellular dependency on glutamine as an energy source [
18]. Therefore, we hypothesize that HBV might also upregulate GDH1 and GLS2 through similar mechanisms involving mTOR or Myc.
Glutamine is initially converted into glutamate by GLS. Glutamate is subsequently metabolized into αKG, a crucial intermediate in the TCA cycle, primarily through the action of GDH1 [
24]. Like other TCA cycle intermediates, αKG, is actively transported between the mitochondrial matrix and the cytoplasm [
25]. Given that the regulation of HBV cccDNA occurs within the nucleus, a pivotal question arises regarding the availability and transport of αKG into the nuclear compartment, which links cellular metabolism to the epigenetic regulation of HBV. Recently, the mitochondrial enzyme GDH was reported to be relocated to the nucleus by the αKG-consuming protein Tet3 in murine neurons, indicating on-site αKG production [
26]. Additionally, the local accumulation of fumarate, which is generated by chromatin-associated fumarase, plays a vital role in DNA repair by competitively inhibiting αKG-dependent lysine demethylases such as KDM2B in regions of DNA damage [
27]. Therefore, we investigated the subcellular localization of GDH1 in HBV-infected HepG2-NTCP cells. Our findings revealed that GDH1 was conspicuously present in the nuclear fraction of HBV-infected cells, suggesting that GDH1 translocates to the nucleus and elevates the local level of αKG. Accumulating evidence suggests that HBc plays a pivotal role in mediating the binding of histone-modifying enzymes, including methyltransferase PRMT5 [
27] and histone desuccinylase SIRT7 [
28], to cccDNA. Intriguingly, our study discovered that HBc facilitates the nuclear translocation of GDH1, which subsequently leads to an elevation in nuclear αKG levels. In addition to increasing nuclear αKG levels, GDH1 also regulates the metabolic pathways in which glutamate serves as a precursor for the synthesis of amino acids and glutathione. In this study, we found that GDH1 depletion could prevent the conversion of glutamate to αKG, and further activate the synthesis of alanine, aspartate and glutathione. Additionally, GDH1 and αKG are crucial components in the metabolic control of viral replication, including HBV, HSV-1, influenza A, and KSHV [
29,
30]. By upregulating GDH1 activity, viruses can enhance the conversion of glutamate to αKG, thereby augmenting the metabolic intermediates and energy needed for their replication. Consequently, this metabolic manipulation supplies viruses with the necessary precursors and energy, which in turn facilitates the completion of their replication cycle within the host cell.
Methylation of histone proteins H3 and H4 is intricately linked to chromatin structure, and regulates the transcription of cccDNA. Histone KDMs, which are categorized as αKG-dependent dioxygenases, rely on αKG as a crucial cofactor and play pivotal roles in various biological processes, particularly the epigenetic regulation of gene demethylation. Previous studies have shown that a reduction in αKG levels in embryonic stem cells leads to increased methylation of H3K27me3 and H3K9me3, which are involved in regulating the expression of pluripotency-associated genes [
31]. Additionally, fumarate has been found to inhibit αKG-dependent H3K36me2 demethylation mediated by KDM2B, resulting in enhanced H3K36me2 levels and subsequent accumulation of the DNA-PK complex at double-strand break regions [
32]. In bone marrow mesenchymal stromal/stem cells, αKG downregulation has been associated with reduced global H3K9me3 and H3K27me3 levels [
11]. Nevertheless, it is unclear whether metabolites can modulate histone modifications within cccDNA minichromosomes. In our investigation, we made a fascinating discovery that the supplementation of HBV-infected cells with αKG reduced the binding of repressive histone modifications, including H3K4me3, H3K9me3, and H4K20me3, to cccDNA. On the other hand, our dual-luciferase assay revealed that GDH1 and DM-αKG enhance the activities of HBV promoters. Given the roles of GDH1 and DM-αKG in promoting transcription factor enrichment on cccDNA, we speculated that the enhanced activity of HBV promoters may stem from reduced histone methylation, thereby facilitating transcription factor binding to cccDNA. However, we cannot rule out the possibility that GDH1 and DM-αKG may directly regulate the binding between transcription factors and promoters.
Collectively, our findings establish a correlation between HBV infection and glutaminolysis, particularly highlighting a novel dependency on GDH1 for the production of αKG during HBV infection. In HBV-infected cells, the HBc redirects GDH1 to the nucleus, facilitating the on-site generation of αKG. This nuclear αKG subsequently modifies the methylation profile of the cccDNA minichromosome by activating the histone demethylase activity of KDM4A, thereby activating cccDNA transcription. The necessity of αKG in mediating active cccDNA transcription suggests the intersection of cellular metabolism and epigenetic modification of HBV cccDNA. Consequently, our study identifies GDH1 as a crucial cellular target for the development of innovative epigenetic therapies aimed at treating HBV infection. To develop the strategies that specifically disrupt the role of GDH1-derived αKG in epigenetic regulation without compromising its essential function in glutamine metabolism will be the future research directions and significant challenge. In addition, our data were derived from HBV-infected cell models and have not been corroborated in other cell models. Moreover, despite our efforts to imitate the natural HBV infection process using PHH and humanized mouse chimeric liver models, there remains a certain difference between our experimental outcomes and the complexities of natural HBV infection. Also, our current study does discuss histone methylation but does not extensively elaborate on the precise mechanisms by which these modifications impact cccDNA transcription, and the specific mechanisms require further investigation.
FOOTNOTES
-
Authors’ contribution
JC, STC, and WXC designed the study. STC, HJD, XH, HZ, MT, JHR, HBY, MLY, DPZ, ZHL, and WXC performed the experiments and analyses. ZZZ and WXC provided the materials. JC and STC wrote the manuscript. JC critically reviewed the manuscript. JC supervised the study.
-
Acknowledgements
This work was supported by the Key Technologies R&D Program (2022YFA1303600 to JC); National Natural Science Foundation of China (U23A20472 and 82273423 to JC, 82202501 and 82473089 to STC); Chongqing Natural Science Foundation (CSTB2024NSCQ-QCXMX0003 to STC, CSTB2023NSCQ-LZX0053 to JC, CSTB2022NSCQ-MSX0864 to JHR); Scientific and Technological Research Program of Chongqing Municipal Education Commission (KJQN202300465 to STC, KJQN202300483 to ZHL).
-
Conflicts of Interest
The authors have no conflicts to disclose.
SUPPLEMENTAL MATERIAL
Supplementary material is available at Clinical and Molecular Hepatology website (
http://www.e-cmh.org).
Supplementary Figure 1.
HBV infection alters host cell glutamine utilization. (A) HBV-infected HepG2-NTCP cells were collected at indicated time points and subjected to RNA-seq. PCA shows the correlation and dispersion among groups. (B) HBV-infected HepG2-NTCP cells were collected at indicated time points and HBV cccDNA levels were detected by TaqMan probe qRT-PCR. (C) Pie chart shows the proportion of metabolism related pathways in Cluster 4, which is enriched by Gene Ontology (GO) analysis. (D) The mRNA levels of GDH1, NIT2, AADAT, GLUL, GLS2, GFPT1, GFPT2, PPAT, CAD and CPS1 in HBV-infected HepG2-NTCP cells or PHHs were tested by real time PCR using specific primers. (E, F) HBV-infected HepG2-NTCP or PHHs cells were transduced with short hairpin RNAs targeting specific genes. The silencing efficiency was confirmed by RT-PCR (E). The level of HBV 3.5 kb RNA was tested by RT-PCR (F). β-actin was used as an internal control. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. *P<0.05; **P<0.01; ***P<0.001.
Supplementary Figure 2.
Glutaminolysis is benefit to HBV replication. (A) Cell viability was detected by CCK8 in HepG2-NTCP cells and PHHs which cultured with glutamine-free medium supplemented with 10% dialyzed FBS or complete media at indicated time points. (B–D) HBV-infected HepG2-NTCP cells or PHHs were cultured with glutamine-free medium supplemented with 10% dialyzed FBS for 24 hours before harvest. Complete media was used as control. (B) cccDNA was tested by Southern blot. (C) HBV DNA was analyzed by real-time PCR and Southern blot. (D) The HBsAg and HBeAg levels were examined by ELISA. (E, F) HBV-infected HepG2-NTCP cells were treated with oxamate (50 mM) or CB-839 (8 μM) at indicated time course. HBV RNA and DNA were detected by real time PCR. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test. *P<0.05; **P<0.01.
Supplementary Figure 3.
Silencing of GLS2 inhibits cccDNA transcription. HBV-infected HepG2-NTCP cells or PHHs were transduced with lentivirus expressing shGLS2 for 5 days. (A) The silencing efficiency was confirmed by Western blot. (B) HBV RNA was analyzed by real-time PCR and Northern blot. (C) HBV DNA was analyzed by real-time PCR and Southern blot. (D) The cccDNA was quantified by Southern blot. (E) The HBsAg and HBeAg levels were determined by ELISA. (F) The cccDNA was quantified by TaqMan probe qRT-PCR and the ratios of total RNAs to cccDNA and 3.5-kb RNA to cccDNA were calculated. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test. *P<0.05; **P<0.01.
Supplementary Figure 4.
Glutamine mediated the promotion of HBV replication depends on glutamine metabolite αKG. (A, B) HBV-infected HepG2-NTCP cells or PHHs were cultured with glutamine-free medium supplemented with DM-αKG (4 mM), NH Cl (5 mM) or NEAA (0.1 mM) for 24 hours before harvest. (A) HBV DNA was analyzed by real-time PCR. (B) The HBsAg and HBeAg levels were determined by ELISA. (C–E) HBV-infected HepG2-NTCP cells or PHHs were cultured with glutamine-free medium supplemented with a series of concentration of DM-αKG (0 mM, 2 mM, 4 mM, 8 mM) for 24 hours before harvest. (C) HBV DNA was analyzed by real-time PCR and Southern blot. (D) The cccDNA was quantified by Southern blot. (E) The HBsAg and HBeAg levels were determined by ELISA. (F) HBV-infected cells were incubated with U5-[13C]-glutamine for 24 hours and studied using LC-MS profiling. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test or one-way ANOVA. *P<0.05; **P<0.01. Gln, glutamine; DM-αKG, dimethyl alpha-ketoglutarate; NEAA, non-essential amino acid.
Supplementary Figure 5.
The effect of EGCG on cccDNA transcription. (A) Schematic overview of the target of GDH1 inhibitor EGCG or transaminase inhibitor AOA among glutamine metabolism. (B–E) HBV-infected HepG2-NTCP cells or PHHs were treated with a series of concentration of EGCG (0 μM, 25 μM, 50 μM, 100 μM) or AOA (0 mM, 0.25 mM, 0.50 mM, 1.00 mM). (B) HBV DNA was detected by real-time PCR and Southern blot. (C) HBV RNA was analyzed by real-time PCR. (D) HBV DNA was analyzed by real-time PCR. (E) The cccDNA was quantified by TaqMan probe qRT-PCR. For HBV infection, cells were infected with 500 VGEs/cell. (F–I) HepG2-NTCP cells were transduced with lentivirus expressing shGDH1 or GDH1. (F, G) Intracellular glutamine levels were measured using a colorimetric assay. (H, I) Cell proliferation was detected by CCK-8. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test or one-way ANOVA. *P<0.05; **P<0.01. EGCG, epigallocatechin gallate; AOA, aminooxyacetate.
Supplementary Figure 6.
Silencing of GDH1 inhibits cccDNA transcription. (A) HBV-infected HepG2-NTCP cells were harvested at indicated days. GDH1 protein was detected by Western blot. (B–E) HBV-infected HepG2-NTCP and PHHs were transduced with lentivirus expressing shGDH1 for 5 days. (B) The silencing efficiency was confirmed by Western blot. (C) HBV DNA was analyzed by real-time PCR and Southern blot. (D, E) The HBsAg and HBeAg levels were determined by ELISA. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test. *P<0.05.
Supplementary Figure 7.
GDH1 overexpression facilitates cccDNA transcription. HBV-infected HepG2-NTCP cells or PHHs were transduced with lentivirus expressing GDH1 or GDH1mut for 5 days. (A) The overexpression efficiency was confirmed by Western blot. (B) HBV RNA was analyzed by real-time PCR and Northern blot. (C, D) The cccDNA was quantified by TaqMan probe qRT-PCR and Southern blot. The ratios of total RNAs to cccDNA and 3.5-kb RNA to cccDNA were calculated. (E) HBV DNA was detected by real-time PCR and Southern blot. (F, G) The HBsAg and HBeAg levels were determined by ELISA. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test. **P<0.01.
Supplementary Figure 8.
GDH1 is transported into nuclei by interacting with core protein to facilitate cccDNA transcription. (A) HepG2-NTCP cells were transfected with Vector, EGFP-NLS-GDH1 or EGFP-WT-GDH1, respectively. The localization of GDH1 was observed by using fluorescence microscopy. (B, C) HBV-infected HepG2-NTCP cells were transfected with Vector, GFP-NLS-GDH1 or GFP-WT-GDH1 for 5 days. HBV RNA and HBV DNA were detected by real-time PCR. (D) The level of GDH1 associated with cccDNA, GAPDH or MYH6 promoter were examined by ChIP assay. IgG antibody was used as a negative antibody control. GAPDH and histone 3 were used as markers for cytoplasmic and nuclear fractions, respectively. ChIP results are expressed as % of input. (E) HepG2-NTCP cells were infected with HBV WT (500 VGEs/cell) or HBV∆HBc (500 VGEs/cell) for 5 days. HBV RNA was detected by real-time PCR (left panel) and the cccDNA was quantified by TaqMan probe qRT-PCR (right panel). (F, G) HepG2-NTCP cells were infected with HBV WT (500 VGEs/cell) or HBV∆HBc (500 VGEs/cell) for 24 hours and then transduced with vector or GDH1. HBV DNA was detected by Southern blot and real-time PCR on 15th post infection. (H, I) HepG2-NTCP cells were infected with HBV WT (500 VGEs/cell) or HBV∆HBc (500 VGEs/cell) for 24 hours and then transduced lentivirus expression shGDH1 for 5 days. HBV RNA and HBV DNA were detected by real-time PCR. For HBV DNA (C–I), representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test. *P<0.05; **P<0.01; ***P<0.001.
Supplementary Figure 9.
αKG promotes HBV replication both in vitro and in vivo. (A) Diagram showing glutamate metabolism pathways (Glu, glutamate; Gln, glutamine; GSH, glutathione). (B, C) HBV-infected HepG2-NTCP cells transduced with shGDH1 or GDH1 lentivirus for 4 days. Cells were incubated with U5-[13C]-glutamine for 24 hours. Metabolites were analyzed by LC-MS. (D) HBV infected HepG2-NTCP cells transduced with lentivirus expressing shGDH1 for 4 days and supplemented with indicated metabolites (DM-αKG, 4 mM; DMF, 50 μM; DMM, 200 μM; DMS, 200 μM) for 24 hours. HBV DNA was determined by real time PCR. (D–F) HBV-infected HepG2-NTCP cells were treated with a series of concentration of DM-αKG (0 mM, 2 mM, 4 mM, 8 mM) for 24 hours before harvest. (D) HBV DNA was analyzed by real-time PCR and Southern blot. (E) The HBsAg and HBeAg levels were determined by ELISA. (G) The cccDNA was quantified by TaqMan probe qRT-PCR. (H–K) HBV infected human liver-chimeric Alb-uPA/SCID mice were treated with DM-αKG. (H) αKG levels in mice serum were detected by α-ketoglutarate colorimetric assay. (I) Body weight was monitored every week. (J) HBV DNA in liver tissues was tested by real-time PCR. (K) The cccDNA in liver tissues was quantified by TaqMan probe qRT-PCR. For (A–G), representative data are from at least three independent experiments. Values are presented as mean±standard deviation. For (H–K), n=6 in each group. Statistical analyses were performed using the Mann-Whitney U test or one-way ANOVA. *P<0.05; **P<0.01. DM-αKG, dimethyl alpha-ketoglutarate.
Supplementary Figure 10.
The functional role of GDH1 on KDMs. (A, B) HBV infected HepG2-NTCP cells transduced with lentivirus expressing GDH1 for 5 days. (A) The effect of GDH1 silencing on KDM3A, KDM3B, KDM3C, KDM4A, KDM4B, KDM4D, KDM6A, KDM6B, KDM7A and KDM7B expression levels was determined by Western blot. (B) The interaction between GDH1 and KDMs (KDM3A, KDM3B, KDM3C, KDM4B, KDM4D, KDM6A, KDM6B, KDM7A, and KDM7B) was detected by immunoprecipitation.
Supplementary Figure 11.
KDM4A involved in αKG mediated cccDNA transcription promotion. (A–H) HBV-infected HepG2-NTCP cells or PHHs were transduced with lentivirus expressing shKDM4A or KDM4A for 5 days. (A) The silencing efficiency was confirmed by Western blot. (B) HBV RNA was analyzed by real-time PCR. (C) HBV DNA was detected by real-time PCR. (D) The cccDNA was quantified by TaqMan probe qRT-PCR and the ratios of total RNAs to cccDNA and 3.5-kb RNA to cccDNA were calculated. (E) The overexpression efficiency was confirmed by Western blot. (F) HBV RNA was analyzed by real-time PCR. (G) HBV DNA was detected by real-time PCR. (H) The cccDNA was quantified by TaqMan probe qRT-PCR and the ratios of total RNAs to cccDNA and 3.5-kb RNA to cccDNA were calculated. (I-K) HBV-infected HepG2-NTCP cells or PHHs were treated with DM-αKG, and transduced with lentivirus expressing shKDM4A for 5 days. (I) HBV RNA was analyzed by real-time PCR. (J) HBV DNA was detected by real-time PCR. (K) The cccDNA was quantified by TaqMan probe qRT-PCR and the ratios of total RNAs to cccDNA and 3.5-kb RNA to cccDNA were calculated. (L–N) HBV-infected HepG2-NTCP cells or PHHs were treated with DM-αKG (8 mM), and exposed to KDM4A inhibitor CP-2 (5 μM) for 48 hours before harvest. (L) HBV RNA was analyzed by real-time PCR. (M) HBV DNA was detected by real-time PCR. (N) The cccDNA was quantified by TaqMan probe qRT-PCR and the ratios of total RNAs to cccDNA and 3.5-kb RNA to cccDNA were calculated. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test or one-way ANOVA. *P<0.05; **P<0.01; ns, not significant.
Supplementary Figure 12.
KDM4A promotes cccDNA transcription relying on histone demethylation. (A) The effect of αKG on the enzymatic activity of wild-type KDM4A (KDM4AWT) and mutant type KDM4A (KDM4Amut, E190A) by using KDM4A Homogenous Assay Kit. (B–E) HBV-infected HepG2-NTCP cells or PHHs were transduced with lentivirus expressing KDM4AWT or KDM4Amut for 5 days. (B) The overexpression efficiency was confirmed by Western blot. (C) HBV RNA was analyzed by real-time PCR. (D) HBV DNA was detected by real-time PCR. (E) The cccDNA was quantified by TaqMan probe qRT-PCR and the ratios of total RNAs to cccDNA and 3.5-kb RNA to cccDNA were calculated. (F) HBV-infected HepG2-NTCP cells were transduced with lentivirus expressing shGDH1 for 5 days, and exposed to DM-αKG. The level of H3K4me3, H3K9me3, H4K20me3, H3K4me1, and H3K4me2 associated with cccDNA, GAPDH or MYH6 promoters were examined by ChIP assay. IgG antibody was used as a negative antibody control. GAPDH and histone 3 were used as markers for cytoplasmic and nuclear fractions, respectively. ChIP results are expressed as % of input. For (F), representative data are from at least three independent experiments. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test. **P<0.01; ***P<0.001; ns, not significant.
Supplementary Figure 13.
GDH1-αKG enhances HBV promoters’ activity via increasing the enrichment of transcription factors on cccDNA. (A–C) HepG2-NTCP cells were transduced with lentivirus expressing GDH1 or treated with DM-αKG (8 mM) for 24 hours. (A) Different luciferase reporter plasmids as indicated were cotransfected into cells. The activities of four HBV promoters and two HBV enhancers were detected by dual-luciferase reporter assay system. (B) The mRNA levels of indicated transcription factors were determined by real-time PCR. (C) The protein levels of indicated transcription factors were determined by western blot. (D) HBV-infected HepG2-NTCP cells were transduced with lentivirus expressing GDH1 or treated with DM-αKG (8 mM) for 24 hours. The level of HNF1α, HNF4α, CEBP, CREB, and FXRα associated with cccDNA, GAPDH or MYH6 promoters were examined by ChIP assay. (E) HBV-infected HepG2-NTCP cells were transduced with lentivirus expressing shGDH1 and further treated with DM-αKG (8 mM) for 24 hours. The level of HNF1α, HNF4α, CEBP, CREB, and FXRα associated with cccDNA, GAPDH or MYH6 promoters were examined by ChIP assay. IgG antibody was used as a negative antibody control. GAPDH and histone 3 were used as markers for cytoplasmic and nuclear fractions, respectively. ChIP results are expressed as % of input. For HBV infection, cells were infected with 500 VGEs/cell. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test. *P<0.05; **P<0.01; ns, not significant.
Supplementary Figure 14.
Antiviral activity of GDH1 inhibitor EGCG in Alb-Cre transgenic (Tg) mice. (A) Flowchart of experiments in Alb-Cre transgenic (Tg) mice treated with EGCG. n=8 in each group. (B) Serum HBV DNA was measured by real-time PCR. (C, D) HBV RNA in liver tissues was measured by real-time PCR. (E) HBV DNA in liver tissues was determined by real-time PCR. Statistical analyses were performed using Student’s t-test or one-way ANOVA. **P<0.01. EGCG, epigallocatechin gallate.
Figure 1.HBV infection alters host cell glutamine utilization. (A–D) HBV-infected HepG2-NTCP cells were collected at indicated times and underwent RNA-sequencing. (A) The volcano map shows genes with differing expression levels from RNA-seq (P<0.05, |logFC|>0.585). (B) Cluster analysis of DEGs based on Mfuzz. (C) Gene Ontology (GO) terms in Cluster 4 at four indicated time points. (D) Protein-Protein Interaction Network showing the expression of glutamine metabolism-related genes at indicated time points. (E) HBV-infected cells were subjected to targeted metabolomics analysis of amino acids. Heatmap represents change of amino acid level at indicated time points relative to 0-days post infection. (F) Glutamine metabolism changes during HBV infection are shown. Green/red boxes indicate gene down/up-regulation. Yellow circles indicate metabolite up-regulation. Cells were infected with 500 VGEs/cell. HBV, hepatitis B virus; DEG, differentially expressed gene; VGEs, virion genome equivalents.
Figure 2.Glutamine promotes HBV transcription and replication which depends on glutamine metabolite αKG. (A) After HBV infection, intracellular glutamine levels were measured using a colorimetric assay. Non-infected cells served as a control. (B, C) HBV-infected cells were grown in glutamine-free media with 10% dialyzed FBS for 24 hours before testing. Complete media was used as a control. (B) HBV RNA was detected by PCR and Northern blot. (C) cccDNA was quantified using TaqMan probe qRT-PCR. Ratios of RNAs to cccDNA were calculated. (D) Diagram showing glutamine metabolism pathways. (E, F) HBV-infected cells were grown in glutamine-free media with DM-αKG, NH4 Cl, or NEAA for 24 hours. HBV 3.5-kb RNA and cccDNA were tested. (G–I) HBV-infected cells were grown in glutamine-free media with varying concentrations of DM-αKG for 24 hours. (G) HBV RNA was detected by real-time PCR. (H, I) cccDNA were tested by TaqMan PCR and Southern blot. (J) Diagram showing conversion of U-[13C]-glutamine into metabolites. HBV-infected cells were incubated with U5-[13C]-glutamine for 24 hours and studied using LC-MS profiling. For HBV infection, 500 VGEs/cell were used. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test or one-way ANOVA. *P<0.05; **P<0.01; ns, not significant. HBV, hepatitis B virus; αKG, α-ketoglutarate; FBS, fetal bovine serum; PCR, polymerase chain reaction; cccDNA, covalently closed circular DNA; qRT-PCR, quantitative reverse transcription polymerase chain reaction; LC-MS, liquid chromatography mass spectrometry; VGEs, virion genome equivalents; Gln, glutamine; PHH, primary human hepatocyte; DM-αKG, dimethyl-αKG; NEAA, nonessential amino acid; Cont, control.
Figure 3.GDH1 is responsible for glutamine-αKG mediated HBV replication and transcription. (A, B) HBV-infected HepG2-NTCP cells were treated with varying concentrations of EGCG or AOA. (A) HBV RNA levels were measured by real-time PCR and Northern blot. (B) The cccDNA was quantified by TaqMan probe qRT-PCR. (C, D) HBV-infected HepG2-NTCP cells were harvested at indicated days. (C) HBV 3.5-kb RNA and GDH1 mRNA was analyzed by real-time PCR. (D) GDH1 protein were detected by Western blot and the bands were quantified by Image J. (E–H) HBV-infected HepG2-NTCP cells were transduced with lentivirus expressing shGDH1 for 5 days. (E) HBV RNA was analyzed by real-time PCR and Northern blot. (F) Cells were treated with actinomycin D (5 μg/mL) at indicated time points. HBV RNA were quantified by real-time PCR. (G) Cells were incubated with EU (0.5 mM) for 3 hours. The newly synthesized EU-labelled RNA was purified and EU-labelled HBV RNA were quantified by real-time PCR. (H, I) The cccDNA was quantified by TaqMan probe qRT-PCR and Southern blot. The ratios of RNAs to cccDNA were calculated. For HBV infection, 500 VGEs/cell were used. Representative data are from at least three independent experiments. Values are presented as mean±standard deviation. Statistical analyses were performed using the Mann–Whitney U test or one-way ANOVA. *P<0.05; **P<0.01. GDH1, glutamate dehydrogenase 1; αKG, α-ketoglutarate; HBV, hepatitis B virus; EGCG, epigallocatechin gallate; AOA, aminooxyacetate; PCR, polymerase chain reaction; cccDNA, covalently closed circular DNA; PHH, primary human hepatocyte; 5-EU, 5-ethynyl uridine.
Figure 4.GDH1 is transported into nuclei by interacting with core protein to facilitate cccDNA transcription. (A) Immunofluorescence staining examined subcellular localization of HBc and GDH1 in HBV-infected and non-infected HepG2-NTCP cells. (B) The HBV-specific reads in ChIP-Seq pilot experiments were quantified and normalized by the reads mapped to the human genome (per kilobase per million reads mapped to human genome). Localization of GDH1 along the HBV genome was exhibited. (C) Indicated plasmids were transduced into HepG2-NTCP cells for 3 days and the cytoplasmic and nuclear fractions were extracted. The level of GDH1 was examined by western blot. Histone 3 and GAPDH were used as controls for cytoplasmic and nuclear fractions, respectively. (D, E) Flag-HBc expression was transduced into HepG2-NTCP cells, the interaction between HBc and GDH1 was detected by protein immunoprecipitation and PLA. (F, G) Plasmid expression His-GDH1, His-∆NAD and His-∆GLU were transduced into HepG2-NTCP cells for 3 days, and the interaction between HBc and truncated clones of GDH1 was detected by protein immunoprecipitation and PLA. (H) HepG2-NTCP cells infected with HBV∆HBc or HBV WT, and GDH1 localization were examined. (I) HepG2-NTCP cells infected with HBV WT or HBV∆HBc, GDH1 associations with cccDNA, GAPDH, and MYH6 promoter were tested by ChIP assay. IgG antibody was used as a negative antibody control. GAPDH and histone 3 were used as markers for cytoplasmic and nuclear fractions, respectively. ChIP results are expressed as % of input. (J) HepG2-NTCP cells were infected with HBV WT (500 VGEs/cell) or HBV∆HBc (500 VGEs/cell) for 24 hours and then transduced with vector or GDH1, respectively. HBV RNA was determined by real-time PCR and Northern blot. **P<0.01. ***P<0.001. GDH1, glutamate dehydrogenase 1; cccDNA, covalently closed circular DNA; HBc, HBV core protein; HBV, hepatitis B virus; ChIP, chromatin immunoprecipitation; PLA, proximity ligation assay; WT, wild-type; VGEs, virion genome equivalents; PCR, polymerase chain reaction; HBx, HBV X protein; HBp, HBV polymerase; HBs, HBV S antigen.
Figure 5.Nuclear GDH1 contributes to HBV transcription and replication through αKG. (A–D) HBV-infected HepG2-NTCP cells were transduced with shGDH1 or GDH1 lentivirus for 4 days. (A) Cells were incubated with U5-[13C]-glutamine for 24 hours. Metabolites were analyzed by LC-MS. (B) The nuclei were isolated and αKG level in nuclei were detected by LC-MS. (C, D) Cells were supplemented with indicated metabolites for 24 hours. HBV RNA and cccDNA were examined. (E–G) HBV-infected HepG2-NTCP cells were treated with a series of concentration of DM-αKG for 24 hours before harvest. (E) HBV RNA was analyzed by real-time PCR and Northern blot. (F) Cells were treated with actinomycin D (5 μg/mL). HBV RNA were quantified by real-time PCR. (G) Cells were incubated with EU (0.5 mM) for 3 hours. The newly synthesized EU-labelled RNA was purified and EU-labelled HBV RNA were quantified by real-time PCR. (H) Flowchart of the experiments in human liver-chimeric Alb-uPA/SCID (n=6). (I) Serum human albumin levels were determined by ELISA. (J) Serum HBsAg was quantified by chemiluminescent microparticle immunoassay. (K) Serum HBV DNA was measured by real-time PCR. (L) HBV RNA in liver tissues was determined by real-time PCR. (M) HBc in liver tissues was analysed by immunohistochemistry. *P<0.05; Vector GDH1 **P<0.01; ns, not significant. GDH1, glutamate dehydrogenase 1; HBV, hepatitis B virus; αKG, α-ketoglutarate; LC-MS, liquid chromatography mass spectrometry; cccDNA, covalently closed circular DNA; DM-αKG, dimethyl-αKG; PCR, polymerase chain reaction; 5-EU, 5-ethynyl uridine; HBc, HBV core protein.
Figure 6.GDH1-derived αKG interacts with KDM4A and downregulates histone methylation of H3K4, H3K9 and H4K20 on cccDNA minichromosome. (A) HBV-infected HepG2-NTCP cells were transduced with lentivirus-expressed shGDH1 for 5 days to measure KDMs demethylase activity. (B) HBV-infected cells transduced with GDH1-expressing lentivirus subjected to detect KDM4A demethylase activity. (C, D) The interaction between KDM4A and GDH1 was detected by immunoprecipitation and PLA in HepG2-NTCP cells. (E, F) MST analysis was performed by using Monolith NT.115 (NanoTemper Technologies). MST traces for single dose of αKG (right panel) and dose response curve (left panel) for KDM4A WT (E) and KDM4Amut (F) were provided. (G) ChIP assay revealed the enrichment of KDM4A, KDM1A, KDM1B on cccDNA, GAPDH, MYH6 promoter. (H) HBV-infected cells transduced with GDH1 or treated with DM-αKG, ChIP assay showed the enrichment of H3K4me3, H3K9me3, H4K20me3 on cccDNA, GAPDH, MYH6 promoters. (I) Localization of H3K4me3, H3K9me3, H4K20me3 along the HBV genome was exhibited by using ChIP-Seq. **P<0.01; ***P<0.001. GDH1, glutamate dehydrogenase 1; αKG, α-ketoglutarate; cccDNA, covalently closed circular DNA; HBV, hepatitis B virus; KDMs, histone lysine demethylases; PLA, proximity ligation assay; MST, micro scale thermophoresis; WT, wild-type; ChIP, chromatin immunoprecipitation; DM-αKG, dimethyl-αKG; LSD, lysine-specific histone demethylase.
Figure 7.Deprivation of glutamine or loss of GDH1 enzymatic activity inhibit HBV replication and transcription in human liver-chimeric Alb-uPA/SCID mice. (A–F) Human liver-chimeric uPA/SCID mice were injected with HBV virions for 6 weeks and grouped (n=6). (A) Serum human albumin levels were determined by ELISA. (B) Serum HBsAg was quantified by chemiluminescent microparticle immunoassay. (C) Serum HBV DNA was measured by real-time PCR. (D–F) HBV DNA, RNA, and cccDNA in liver tissues were measured. (G–L) Human liver-chimeric uPA/SCID mice were injected with HBV virions through tail vein for 6 weeks and grouped randomly (n=6). (G–I) Serum human albumin, HBsAg and HBV DNA were measured. (J–L) HBV DNA, cccDNA and HBV RNA in liver tissues were measured. (M) The functional role of GDH1-derived αKG on HBV transcription and replication. *P<0.05; **P<0.01; ns, not significant. GDH1, glutamate dehydrogenase 1; HBV, hepatitis B virus; PCR, polymerase chain reaction; cccDNA, covalently closed circular DNA; αKG, α-ketoglutarate; Gln, glutamine; DM-αKG, dimethyl-αKG; EGCG, epigallocatechin gallate; ETV, entecavir; HBc, HBV core protein.
Abbreviations
carbamoyl-phosphate synthetase 2
covalently closed circular DNA
chromatin immunoprecipitation
differentially expressed gene
glutamate dehydrogenase 1
infectious spleen and kidney necrosis virus
histone lysine demethylases
Kaposi’s sarcoma-associated herpesvirus
liquid chromatography mass spectrometry
micro scale thermophoresis
polymerase chain reaction
quantitative reverse transcription polymerase chain reaction
virion genome equivalents
REFERENCES
- 1. GBD 2019 Hepatitis B Collaborators. Global, regional, and national burden of hepatitis B, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Gastroenterol Hepatol 2022;7:796-829.
- 2. Gish RG, Given BD, Lai CL, Locarnini SA, Lau JY, Lewis DL, et al. Chronic hepatitis B: virology, natural history, current management and a glimpse at future opportunities. Antiviral Res 2015;121:47-58.
- 3. Fu X, Guo X, Wu S, Lin Q, Liu L, Liang H, et al. Non-targeted UHPLC-Q-TOF/MS-based metabolomics reveals a metabolic shift from glucose to glutamine in CPB cells during ISKNV infection cycle. Metabolites 2019;9:174.
- 4. Thai M, Thaker SK, Feng J, Du Y, Hu H, Ting Wu T, et al. MYC-induced reprogramming of glutamine catabolism supports optimal virus replication. Nat Commun 2015;6:8873.
- 5. Gou H, Bian Z, Li Y, Cai R, Jiang Z, Song S, et al. Metabolomics exploration of pseudorabies virus reprogramming metabolic profiles of PK-15 cells to enhance viral replication. Front Cell Infect Microbiol 2021;10:599087.
- 6. Li C, Sun L, Lin H, Qin Z, Tu J, Li J, et al. Glutamine starvation inhibits snakehead vesiculovirus replication via inducing autophagy associated with the disturbance of endogenous glutathione pool. Fish Shellfish Immunol 2019;86:1044-1052.
- 7. Lévy PL, Duponchel S, Eischeid H, Molle J, Michelet M, Diserens G, et al. Hepatitis C virus infection triggers a tumor-like glutamine metabolism. Hepatology 2017;65:789-803.
- 8. Fu X, Li K, Niu Y, Lin Q, Liang H, Luo X, et al. The mTOR/PGC-1α/SIRT3 pathway drives reductive glutamine metabolism to reduce oxidative stress caused by ISKNV in CPB cells. Microbiol Spectr 2022;10:e0231021.
- 9. Wagner M, Bertero E, Nickel A, Kohlhaas M, Gibson GE, Heggermont W, et al. Selective NADH communication from α-ketoglutarate dehydrogenase to mitochondrial transhydrogenase prevents reactive oxygen species formation under reducing conditions in the heart. Basic Res Cardiol 2020;115:53.
- 10. Yang Y, He P, Hou Y, Liu Z, Zhang X, Li N. Osmundacetone modulates mitochondrial metabolism in non-small cell lung cancer cells by hijacking the glutamine/glutamate/α-KG metabolic axis. Phytomedicine 2022;100:154075.
- 11. Wang Y, Deng P, Liu Y, Wu Y, Chen Y, Guo Y, et al. Alpha-ketoglutarate ameliorates age-related osteoporosis via regulating histone methylations. Nat Commun 2020;11:5596.
- 12. Chung C, Sweha SR, Pratt D, Tamrazi B, Panwalkar P, Banda A, et al. Integrated metabolic and epigenomic reprograming by H3K27M mutations in diffuse intrinsic pontine gliomas. Cancer Cell 2020;38:334-349.e9.
- 13. Kang HS, Lee JH, Oh KJ, Lee EW, Han BS, Park KY, et al. IDH1-dependent α-KG regulates brown fat differentiation and function by modulating histone methylation. Metabolism 2020;105:154173.
- 14. Liu X, Zhu C, Wang Y, Wei F, Cai Q. KSHV reprogramming of host energy metabolism for pathogenesis. Front Cell Infect Microbiol 2021;11:621156.
- 15. Li T, Gao SJ. Metabolic reprogramming and metabolic sensors in KSHV-induced cancers and KSHV infection. Cell Biosci 2021;11:176.
- 16. Fröhlich J, Grundhoff A. Epigenetic control in Kaposi sarcomaassociated herpesvirus infection and associated disease. Semin Immunopathol 2020;42:143-157.
- 17. Sanchez EL, Pulliam TH, Dimaio TA, Thalhofer AB, Delgado T, Lagunoff M. Glycolysis, glutaminolysis, and fatty acid synthesis are required for distinct stages of Kaposi’s sarcoma-associated herpesvirus lytic replication. J Virol 2017;91:e02237-16.
- 18. Wise DR, DeBerardinis RJ, Mancuso A, Sayed N, Zhang XY, Pfeiffer HK, et al. Myc regulates a transcriptional program that stimulates mitochondrial glutaminolysis and leads to glutamine addiction. Proc Natl Acad Sci U S A 2008;105:18782-18787.
- 19. Li H, Zhu W, Zhang L, Lei H, Wu X, Guo L, et al. The metabolic responses to hepatitis B virus infection shed new light on pathogenesis and targets for treatment. Sci Rep 2015;5:8421.
- 20. Thaker SK, Ch’ng J, Christofk HR. Viral hijacking of cellular metabolism. BMC Biol 2019;17:59.
- 21. Yang F, Yan S, He Y, Wang F, Song S, Guo Y, et al. Expression of hepatitis B virus proteins in transgenic mice alters lipid metabolism and induces oxidative stress in the liver. J Hepatol 2008;48:12-19.
- 22. Murad H, Tayeb H, Mosli M, Rafeeq M, Basheikh M. Blood levels of glutamine and nitrotyrosine in patients with chronic viral hepatitis. Int J Gen Med 2021;14:8753-8762.
- 23. Krishna G, Soman Pillai V, Valiya Veettil M. Upregulation of GLS1 isoforms KGA and GAC facilitates mitochondrial metabolism and cell proliferation in Epstein-Barr virus infected cells. Viruses 2020;12:811.
- 24. Cai WF, Zhang C, Wu YQ, Zhuang G, Ye Z, Zhang CS, et al. Glutaminase GLS1 senses glutamine availability in a nonenzymatic manner triggering mitochondrial fusion. Cell Res 2018;28:865-867.
- 25. Abla H, Sollazzo M, Gasparre G, Iommarini L, Porcelli AM. The multifaceted contribution of α-ketoglutarate to tumor progression: An opportunity to exploit? Semin Cell Dev Biol 2020;98:26-33.
- 26. Traube FR, Özdemir D, Sahin H, Scheel C, Glück AF, Geserich AS, et al. Redirected nuclear glutamate dehydrogenase supplies Tet3 with α-ketoglutarate in neurons. Nat Commun 2021;12:4100.
- 27. Zhang W, Chen J, Wu M, Zhang X, Zhang M, Yue L, et al. PRMT5 restricts hepatitis B virus replication through epigenetic repression of covalently closed circular DNA transcription and interference with pregenomic RNA encapsidation. Hepatology 2017;66:398-415.
- 28. Yu HB, Cheng ST, Ren F, Chen Y, Shi XF, Wong VKW, et al. SIRT7 restricts HBV transcription and replication through catalyzing desuccinylation of histone H3 associated with cccDNA minichromosome. Clin Sci (Lond) 2021;135:1505-1522.
- 29. Girdhar K, Powis A, Raisingani A, Chrudinová M, Huang R, Tran T, et al. Viruses and metabolism: the effects of viral infections and viral insulins on host metabolism. Annu Rev Virol 2021;8:373-391.
- 30. Shi F, He Y, Li J, Tang M, Li Y, Xie L, et al. Wild-type IDH2 contributes to Epstein-Barr virus-dependent metabolic alterations and tumorigenesis. Mol Metab 2020;36:100966.
- 31. Carey BW, Finley LW, Cross JR, Allis CD, Thompson CB. Intracellular α-ketoglutarate maintains the pluripotency of embryonic stem cells. Nature 2015;518:413-416.
- 32. Jiang Y, Qian X, Shen J, Wang Y, Li X, Liu R, et al. Local generation of fumarate promotes DNA repair through inhibition of histone H3 demethylation. Nat Cell Biol 2015;17:1158-1168.
