The HKDC1-ASS1-ACSBG2 axis reprograms lipid metabolism to drive therapeutic resistance in hepatocellular carcinoma
Xiangyu Ling, Wenhu Zhao, Kuan Li, Litao Liang, Wenbo Jia, Jinyi Wang, Yanzhi Feng, Chao Xu, Qingpeng Lv, Zhiwen Feng, Deming Zhu, Lianbao Kong, Wenzhou Ding
TL;DR
This study identifies a new metabolic pathway in liver cancer that promotes tumor growth and drug resistance, offering a potential new treatment target.
Contribution
The study reveals a novel HKDC1-ASS1-ACSBG2 axis that drives lipid metabolism and lenvatinib resistance in hepatocellular carcinoma.
Findings
HKDC1 is highly expressed in HCC and linked to poor patient outcomes.
The HKDC1-ASS1-ACSBG2 axis promotes lipid biosynthesis and lenvatinib resistance.
This pathway operates independently of the FASN-dependent pathway.
Abstract
Hepatocellular carcinoma (HCC) is a lethal malignancy characterized by profound metabolic reprogramming, notably aberrant lipid metabolism. Current therapeutic strategies offer limited efficacy, highlighting the critical need to identify its metabolic vulnerabilities. This study aims to investigate the role of the HKDC1-ASS1-ACSBG2 axis in lipid metabolic reprogramming and therapeutic resistance in HCC. Using clinical HCC specimens and cell lines, we systematically evaluated the functions and mechanisms of HKDC1, ASS1, and ACSBG2 in lipogenesis, cell proliferation, and drug response via RNA sequencing, co-immunoprecipitation, immunofluorescence, ubiquitination assays, dual-luciferase reporter assays, cellular functional experiments, and nude mouse xenograft models. HKDC1 was highly expressed in HCC and correlated with poor patient prognosis. Mechanistically, HKDC1 interacts with ASS1…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8- —the National Natural Science Foundation of China (81871260)
- —the First Affiliated Hospital of Nanjing Medical University (PY2022007).
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsCancer, Lipids, and Metabolism · Cancer, Hypoxia, and Metabolism · Ferroptosis and cancer prognosis
Introduction
Hepatocellular carcinoma (HCC), the predominant form of primary liver malignancy, continues to pose significant clinical challenges with persistently poor prognosis [1]. Despite advancements in therapeutic interventions including targeted agents like lenvatinib, the clinical efficacy remains suboptimal [2]. This predicament underscores the urgent need to identify novel diagnostic biomarkers and develop innovative therapeutic strategies targeting HCC-specific vulnerabilities.
Hexokinase domain-containing 1 (HKDC1), a novel metabolic regulator containing conserved hexokinase domains, has recently emerged as a critical player in tumor glycolysis [3, 4]. Studies have shown that HKDC1 overexpression correlates with aggressive phenotypes and poor prognosis across multiple malignancies [5–9]. Beyond its canonical role in augmenting glycolytic flux, emerging evidence suggests HKDC1 may coordinate cross-talk between glucose and lipid metabolism, though its functional impact on HCC lipidome remodeling requires systematic investigation [10, 11].
Emerging evidence highlights metabolic reprogramming as a pivotal hallmark of hepatocarcinogenesis [12]. HCC cells exhibit enhanced glycolytic flux (the Warburg effect) to fuel their biosynthetic demands, generating glycolytic intermediates that serve as precursors for lipid biosynthesis [13, 14]. Notably, citrate-derived acetyl-CoA acts as the fundamental building block for fatty acid synthesis [15]. The coordinated regulation of lipid metabolism enzymes, including fatty acid synthase (FASN) and fatty acid β-oxidation machinery, supports membrane biogenesis, energy homeostasis, and oncogenic signaling [16]. These metabolic adaptations present attractive therapeutic targets, yet the molecular orchestrators governing HCC’s metabolic plasticity remain incompletely understood.
ACSBG2 (Acyl-CoA Synthetase Bubblegum Family Member 2), a crucial enzyme in the acyl-CoA synthetase superfamily, belongs to a conserved protein family comprising three distinct subfamilies: long-chain acyl-CoA synthetases (ACSLs), fatty acid transport proteins (FATPs), and the bubblegum acyl-CoA synthetases (ACSBGs). This enzyme catalyzes the activation of long-chain and very-long-chain fatty acids through CoA esterification [17]. ACSLs members have separate functions in fatty acid metabolism of different type ACSBG2, a crucial enzyme in the acyl-CoA synthetase superfamily, belongs to a conserved protein family comprising three distinct subfamilies: long-chain acyl-CoA synthetases (ACSLs), fatty acid transport proteins (FATPs), and the bubblegum acyl-CoA synthetases (ACSBGs). This enzyme catalyzes the activation of long-chain and very-long-chain fatty acids through CoA esterification.es of cells, which leads to liver diseases and metabolic diseases, such as fatty liver, obesity, atherosclerosis and diabetes. They are also related to neurological disorders and other diseases [18]. ACSBG2 may play a role in certain types of tumors, affecting the lipid metabolism and proliferation ability of tumor cells [19]. However, the specific mechanism of ACSBG2 in tumor occurrence and development, particularly in the context of therapy resistance, still needs further exploration.
Argininosuccinate synthase 1 (ASS1), traditionally recognized for its urea cycle function in ammonia detoxification, has recently been redefined as a metabolic nexus bridging nitrogen and lipid metabolism [20–22]. Pioneering work demonstrated that ASS1 mediates the conversion of glutamine-derived carbons into cytosolic acetyl-CoA, thereby fueling monounsaturated fatty acid (MUFA) synthesis and conferring ferroptosis resistance [23]. This paradigm-shifting discovery positions ASS1 as a critical regulator of redox homeostasis and lipid metabolism in malignant cells, suggesting its potential involvement in broader mechanisms of drug tolerance.
Lenvatinib is an established first-line therapy for HCC in clinical practice. [24] It is reported that reprogrammed lipid metabolism has been directly linked to therapeutic resistance. Enhanced de novo lipogenesis provides cancer cells with lipid droplets that act as energy reservoirs and signaling molecules, enabling survival under therapeutic stress and conferring resistance to targeted agents like lenvatinib [25, 26]. However, the specific mechanisms of Lenvatinib resistance in HCC needs further investigation.
Our current investigation unveils a novel HKDC1-ASS1-ACSBG2 signaling axis driving HCC lipogenesis and promoting resistance to lenvatinib. Mechanistically, HKDC1 stabilizes ASS1 protein by inhibiting ubiquitin-mediated degradation, leading to enhanced glutamine-to-acetyl-CoA conversion [23, 27]. The resultant acetyl-CoA accumulation transcriptionally activates ACSBG2. This metabolic cascade ultimately promotes lipid accumulation and confers proliferative advantage to HCC cells. We demonstrate that this axis is a key contributor to lenvatinib resistance, and its targeting can resensitize HCC cells to treatment. These findings not only elucidate the crosstalk between glucose and lipid metabolism in HCC but also identify HKDC1 as a master regulator of metabolic reprogramming and unveil a therapeutically targetable pathway to overcome lenvatinib resistance in HCC [28].
Materials and methods
Clinical samples and cell lines
The tissue samples utilized in this study were obtained from patients with HCC who underwent radical hepatectomy at the Hepatobiliary Center of the First Affiliated Hospital of Nanjing Medical University. This research was approved by the Ethics Committee of the First Affiliated Hospital of Nanjing Medical University, and all patients provided written informed consent prior to participation. The clinical data of patients are presented in Supplementary Table S1.
The cell lines employed in this investigation include Huh7, MHCC97H, MHCC97L, Hep3B, HCCLM3, SK-Hep1, HepG2 and the normal human liver cell line HHL-5. These cell lines were acquired from the Shanghai Institute of Cell Biology at the Chinese Academy of Sciences in Shanghai, China.
Cell culture
All mentioned cell lines were maintained in high-glucose (4.5 g/L) Dulbecco’s Modified Eagle’s Medium (DMEM) (Gibco, Carlsbad, CA, USA), supplemented with 10% fetal bovine serum (FBS) (WISENT, Montreal, Canada) and 1% penicillin-streptomycin solution. This standard DMEM formulation contains 4 mM L-glutamine.
Quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was isolated from cell cultures and tissue samples using TRIzol reagent (Invitrogen), with RNA concentration quantified by spectrophotometry. First-strand cDNA synthesis was performed using a reverse transcription kit (VA zyme, Nanjing, China). Quantitative PCR amplifications were conducted using Ace qPCR SYBR Green Master Mix (VA zyme) on an ABI 7900 Real-Time PCR System (Applied Biosystems). Gene expression levels were normalized to GAPDH, and relative quantification was calculated using the 2^−ΔΔCt^ method. Primer sequences are listed in Supplementary Table S2.
Transfection
The construction of cell lines with stable knockdown or overexpression of HKDC1, ASS1, or ACSBG2 is primarily achieved by transfecting them with lentiviruses containing the corresponding short hairpin RNAs (shRNAs) or overexpression plasmids (Genomeditech). The optimal multiplicity of infection (MOI) for each cell line was determined in a preliminary titration experiment. The final transductions were performed at an MOI of 20 for MHCC97H and an MOI of 15 for Hep-3B cells in the presence of 8 µg/mL polybrene (Invitrogen).
A kill-curve assay was performed prior to transduction. Briefly, parental cells were treated with a gradient of puromycin concentrations (0.5–10 µg/mL). The minimum concentration that achieved 100% cell death within 4 days (2 µg/mL) was selected for subsequent stable cell selection.
At 48 h post-transduction, the culture medium was replaced with fresh complete medium containing 2 µg/mL puromycin (Invitrogen). Selection was maintained for 7 days to establish stable polyclonal populations. The shRNA sequences used in this section are listed in Supplementary Table S3.
Western blot and quantification
Cellular and tissue proteins were extracted using RIPA lysis buffer (Beyotime Biotechnology) containing protease inhibitors. Protein lysates were resolved by 10% SDS-PAGE and electrophoretically transferred onto PVDF membranes (Millipore). After blocking with 5% non-fat milk in TBST for 2 h at room temperature, the membranes were probed with specific primary antibodies at 4 °C overnight. Following three washes with TBST, membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (1:5000 dilution) for 2 h at ambient temperature. Protein signals were developed using enhanced chemiluminescence detection reagents (ECL; Biosharp, China) and captured on X-ray film. Antibody information including sources is provided in Supplementary Table S4.
Density analysis of the Western blot bands was performed using ImageJ software (National Institutes of Health, version 1.53). Briefly, the original images were converted to 8-bit grayscale, and the background was subtracted. A fixed-size selection tool was then used to outline each target band, and its integrated density value (IntDen) was measured. The expression level of the target protein was normalized to the corresponding internal control by calculating the ratio of their respective IntDen values. The final data are presented and statistically analyzed as normalized ratios relative to the control group.
Cell counting kit-8 (CCK-8) assay
Transfected cells were seeded into each well of a 96-well plate at a density of 1,000 cells per well. Subsequently, the cells were incubated with 10 µl of CCK8 solution per well for 2 h. The absorbance of each group was then measured at a wavelength of 450 nm every 24 h.
Colony formation assay
A total of 500 individual HCC cells were meticulously seeded in 6-well plates and subjected to incubation within a cell culture incubator for a period of 14 days. Subsequently, the colonies were precisely fixed with paraformaldehyde and delicately stained with crystal violet stain (Beyotime, China). Upon drying, the colonies were accurately photographed and precisely counted through the utilization of Image J.
5-Ethynyl-20-deoxyuridine (EdU) assay
The EdU assay was carried out by employing an EdU Kit from Ribobio. Cells were positioned in a 24-well plate and subjected to an EdU solution for 2 h. Subsequent to fixation with 4% paraformaldehyde, cell permeabilization was accomplished through the utilization of Triton X-100. Thereafter, the cells were stained with the Apollo solution, and the nuclei were labeled with DAPI. Images were captured using a fluorescence microscope (DM4000B-1, Leica, Frankfurt, Germany).
Apoptosis analysis
Cell apoptosis was quantified using a FITC Annexin V Apoptosis Detection Kit (Vazyme). Following two PBS washes, approximately 5 × 10⁵ cells were resuspended in 100 µL Binding Buffer and stained with 5 µL Annexin V-PE and 5 µL PI for 10 min at room temperature. After adding 400 µL Binding Buffer, apoptotic populations were analyzed using a BD FACSCalibur flow cytometer.
Wound healing assay
Both MHCC97H and Hep3B cells (5 × 10^5^ cells) were seeded into six-well plates. When the cells reached 90% confluence, a 200 µL pipette tip was used to scratch the cell surface. After 48 h, the wound closure was calculated using the formula: wound closure = (wound width at 0 h − wound width at 48 h) / wound width at 0 h × 100%. Images were captured using an inverted phase-contrast microscope.
Transwell assay
Uncoated transwell chambers (8 μm pore size; Corning, USA) were used for migration assays. 2 × 10^4^ MHCC97H or Hep3B cells (suspended in 200 µL serum-free medium) were seeded into the upper chamber, while 600 µL medium supplemented with 10% FBS was added to the lower chamber. After 48 h of incubation, the cells on the lower surface of the membrane were fixed with 4% paraformaldehyde for 30 min and stained with crystal violet. Cells from three randomly selected fields were imaged and counted. The experiment was performed in triplicate.
RNA sequencing and bioinformatic analysis
Total RNA was extracted from MHCC97H cells in the negative control and HKDC1-OE groups using TRIzol reagent (Invitrogen). RNA purity and concentration were measured using a NanoDrop 2000 spectrophotometer (Thermo Scientific). RNA-seq libraries were prepared and sequenced on the Illumina platform by Hongxu Biotechnology Company (Shanghai, China), generating 150 bp paired-end reads. The target sequencing output was 20 million reads per sample (detailed raw data statistics are available in Supplementary Table S5). Clean reads were aligned to the human reference genome (GRCh38/hg38) using HISAT2, achieving an alignment rate of 92.26% to 92.65%. Differential expression analysis was performed using DESeq2, with genes exhibiting |log2(FC)| > 1 and an adjusted p-value < 0.05 classified as statistically significant DEGs. The raw data statistics of RNA-seq are displayed in Supplementary Table S5.
IP - mass spectrometry
Total protein was extracted from MHCC97H cells, and immunoprecipitation (IP) was performed using an anti-HKDC1 antibody and protein A/G agarose beads (ThermoScientific, MA, USA) as described above. Mass spectrometry analysis was conducted by BGI Tech Solutions Co., Ltd. (BGI, Shenzhen, Guangdong, China).
Co-immunoprecipitation (Co-IP) assay
The collected Hep3B cells were resuspended in 1 mL of RIPA protein lysis buffer containing 1% protease inhibitor and incubated on ice for a minimum of 30 min. The supernatants were then harvested and incubated with the respective primary antibodies or IgG at 4 °C for 24 h. Protein A and G agarose beads (Thermo Scientific, MA, USA) were subsequently added, and the mixture was incubated at 4 °C for an additional 3 h. The beads were washed five times with PBS buffer, and 40 µL of sample loading buffer was added followed by boiling for 5 min. Western blot analysis was performed on the processed samples.
Immunofluorescence
Coverslips were placed at the bottom of 24-well plates, and 40,000 treated HCC cells were added to each well. Once the cells adhered, they were fixed with 4% paraformaldehyde and blocked with goat serum for 30 min. The cells were incubated overnight at 4 °C with the primary antibody. After three washes with PBST, the cells were incubated with fluorescent secondary antibodies (Beyotime, Shanghai, China) for 1 h, followed by nuclear staining with DAPI (Beyotime, Shanghai, China). Images were captured using a confocal microscope.
Subcutaneous tumor model and drug treatment
Four-week-old male BALB/c nude mice were procured from Vital River (Beijing, China) and randomly allocated into four groups, each comprising five mice. Transfected cells were resuspended in PBS and administered subcutaneously into the flanks of the mice. Tumor dimensions were assessed every four days. After a period of four weeks, the mice were euthanized, and the volume and mass of the subcutaneous tumors were recorded. Ethical approval for all animal experiments was granted by the Institutional Animal Care and Use Committee (IACUC) at Nanjing Medical University. All animal procedures conformed to the guidelines established by the IACUC.
Lenvatinib treatment was initiated on Day 4 post-randomization. Mice received lenvatinib at a dose of 10 mg/kg, dissolved in a vehicle of 10% DMSO, via daily oral gavage for a duration of 24 days. Tumor volumes and body weights were measured every four days throughout the study.
To quantitatively assess the in vivo efficacy, the tumor growth inhibition (TGI) was calculated at the end of the treatment period (Day 28) using the formula: TGI (%) = [1 - (ΔT/ΔC)] × 100, where ΔT is the change in mean tumor volume of the treatment group and ΔC is the change in mean tumor volume of the control group over the same time period.
TG and T-CHO measurements
1 × 10^6^ cells of MHCC97H overexpressing HKDC1, Hep3B cells with HKDC11 knockdown, and corresponding control cells were collected. The cell samples were lysed with 100 µL of 2% Triton X-100 (Biosharp, China) lysis buffer. Total cholesterol and triglyceride levels were measured using the Total Cholesterol Assay Kit (Njjcbio, China) or Triglyceride Assay Kit (Njjcbio, China), with blank, calibration, and sample groups prepared. After adding 250 µL of working solution to each group, distilled water, calibrator, and cell samples were added to the blank, calibration, and sample groups, respectively, and incubated at 37 °C for 10 min. Absorbance at 500 nm was measured using a microplate reader. Total cholesterol and triglyceride contents were calculated using the following formula: T-CHO/TG (mmol/gprot) = [(Asample − Ablank) / (Acalibration − Ablank)] * Ccalibration / Cpr (Ccalibration: concentration of the calibration; Cpr: concentration of the sample). The standard curves of the concentration were displayed in Supplementary Material.
Nile red staining and quantification
The HCC cell lines (MHCC97H and Hep3B) were counted and seeded into six-well plates at a density of 1 × 10^5^ cells per well, 48 h post-transfection. After 24 h, the cells were fixed with 4% paraformaldehyde for 30 min. The cells were then stained using the Nile Red Staining Kit (Beyotime, Shanghai, China) according to the manufacturer’s instructions, and incubated at room temperature for 20 min. Fluorescent images were captured using an inverted fluorescence microscope (Nikon, Japan). The experiment was performed in triplicate. The mean fluorescence intensity (MFI) of Nile Red was quantified using ImageJ software. Values from multiple fields were averaged and normalized to the number of cells to yield the relative lipid content per cell.
Dual-luciferase reporter assay
The promoter region of ACSBG2 was cloned upstream of the firefly luciferase gene in the pGL3-Basic vector. The constructed reporter plasmid was then co-transfected with the pRL-TK Renilla luciferase control vector into Hep3B and MHCC97H cells. Luciferase activities were measured using a dual-luciferase reporter assay system, and the relative promoter activity was quantified by normalizing firefly luciferase activity to that of Renilla.
Ubiquitination assay
HCC cells were transfected with Flag-tagged ASS1, HKDC1 plasmid or vector for 48 h, followed by treatment with 10 µM proteasome inhibitor MG132 for an additional 8 h. After collecting the cells and extracting the proteins, they were suspended in SDS buffer diluted to a 1× SDS concentration and boiled to ensure complete protein denaturation. Ubiquitinated ASS1 was detected by Western blot using an anti-Flag antibody.
Statistical analysis
Quantitative data are shown as means ± SD. Prism 8 software (GraphPad Software, La Jolla, CA, USA) was used to perform the data analysis. Student’s t-test compared two samples, while one-way ANOVA analyzed multiple groups. All experiments were conducted in triplicate. Survival analysis used Kaplan–Meier for OS calculation and log-rank test for comparison. Statistical significance was indicated by *p < 0.05, **p < 0.01, and ***p < 0.001.
Multivariable Cox proportional hazards regression was performed to assess the independent prognostic value of HKDC1 expression in hepatocellular carcinoma. The model was adjusted for TNM stage, tumor size, and age. Results are presented as hazard ratios (HR) with 95% confidence intervals (CI). Statistical significance was defined as a two-sided p-value < 0.05. All analyses were conducted using R software (version 4.4.1).
Results
HKDC1 is highly expressed in HCC and correlates with poor prognosis
To understand the transcriptional changes occurring in HCC, we analyzed the transcriptomic differences between HCC tissues and adjacent non-cancerous tissues using the GEO database. In the GSE77314 dataset, we found that the expression of HKDC1 was upregulated in HCC tissues compared to adjacent non-cancerous tissues (P < 0.001) (Fig. 1A). Cross-validation using the GEPIA database confirmed that HKDC1 expression was significantly increased in HCC tissues compared to normal liver tissues (Fig. 1B). Additionally, pan-cancer analysis of HKDC1 indicated that it is upregulated in various cancers, including HCC (Fig. 1C). To further validate these findings, we measured the mRNA and protein levels of HKDC1 in paired HCC and adjacent non-cancerous tissue samples collected from the Department of Hepatobiliary Surgery at the First Affiliated Hospital of Nanjing Medical University using qRT-PCR and Western blotting. The qRT-PCR results showed that HKDC1 expression was significantly higher in the 60 HCC tissue samples compared to their paired adjacent tissues (Fig. 1D). Western blotting of representative cases showed the same trend (Fig. 1E). Western blot and qRT-PCR analyses of mRNA/protein extracts from normal hepatocytes (HHL-5) and multiple HCC lines (including SK-Hep1, MHCCLM3, Hep3B, Huh7, HepG2, MHCC97H, and MHCC97L) revealed elevated HKDC1 expression in HCC cells versus normal controls. Among the HCC cell lines, HKDC1 expression was highest in Hep3B and lowest in MHCC97H (Fig. 1F). Accordingly, these two cell lines were selected to establish HKDC1-knockdown and-overexpressing models for subsequent functional investigations. Furthermore, immunohistochemical (IHC) staining of 8 paired HCC and adjacent tissues revealed high HKDC1 expression in HCC tissues, the representative images are shown in Fig. 1G. Kaplan-Meier survival analysis (Kaplan-Meier plotter (kmplot.com)) showed that increased HKDC1 expression was associated with lower overall survival (OS) and disease-specific survival (DSS) in HCC patients (Fig. 1H). To further investigate whether HKDC1 is an independent prognostic factor for HCC, we performed a multivariate Cox regression analysis on the data of 60 patients in the follow-up cohort. After adjusting for tumor size, TNM stage, and age, the analysis results showed that the HR for HKDC1 was 2.929 (95% CI: 1.855–4.625), confirming it as an independent prognostic factor for HCC. In contrast, TNM stage and age were not independent prognostic factors for HCC (Supplementary Table S6). These analyses and experimental results suggest that HKDC1 is highly expressed in HCC and correlates with poor prognosis in patients.
Fig. 1. Expression and prognostic relevance of HKDC1 in HCC. A) Transcriptomic profiling of the GSE77314 dataset identified differential expression of HKDC1 between HCC and adjacent non-cancerous tissues. B) GEPIA database analysis confirmed distinct HKDC1 expression patterns in HCC versus normal liver tissues. C) Pan-cancer analysis revealed HKDC1 expression variations across cancer types, including HCC. D) qRT-PCR analysis of 60 paired clinical HCC/adjacent tissues demonstrated statistically significant differences in HKDC1 mRNA levels. E) Western blot analysis of paired tissues showed differential HKDC1 protein expression in HCC. F) Comparative analysis of HKDC1 expression in normal hepatocyte line HHL-5 and HCC cell lines (SK-Hep1, MHCCLM3, Hep3B, Huh7, HepG2, MHCC97H, MHCC97L) revealed cell line-specific expression profiles. G) Representative images of immunohistochemical (IHC) staining of paired HCC/adjacent tissues. Scale bars: 20 μm. H) Kaplan-Meier survival analysis demonstrated associations between HKDC1 expression levels and clinical outcomes (OS and DSS) in HCC patients. Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001
Overexpression of HKDC1 promotes cell proliferation in vitro and in vivo
To investigate the impact of HKDC1 on the malignant biological behavior of HCC, we transfected the MHCC97H cell line with a plasmid overexpressing HKDC1 and transfected the Hep3B cell line with short hairpin RNA (sh-HKDC1#1, sh-HKDC1#2) to knock down HKDC1 expression. The mRNA and protein levels of HKDC1 in the transfected cells were confirmed by qRT-PCR (Figure S1A) and Western blotting (Fig. 2A). The CCK-8 assay results showed that HKDC1 knockdown markedly reduced the proliferation ability of Hep3B cells compared to the control group. In contrast, overexpression of HKDC1 significantly enhanced the proliferation capacity of MHCC97H cells (Fig. 2B). Edu staining revealed that fluorescence intensity was significantly reduced in the HKDC1 knockdown group, while stronger Edu fluorescence intensity in the HKDC1 overexpression group (Fig. 2C). Furthermore, colony formation assays showed that the number of colonies formed by HKDC1-knockdown cells was significantly lower compared to the control group, while HKDC1 overexpression resulted in more colonies (Fig. 2D). We also found that HKDC1 knockdown promoted apoptosis in Hep3B cells, while overexpression of HKDC1 resulted in fewer apoptotic cells in MHCC97H cells (Fig. 2E). Subsequent wound healing assays (Figure S1B) and Transwell migration assays (Figure S1C) showed that higher HKDC1 expression enhanced the migration ability of HCC cells, whereas HKDC1 knockdown inhibited cell migration.
To verify these findings in vivo, we established subcutaneous xenograft models in nude mice. Following 28 days of tumor progression monitoring, xenografts were excised for evaluation. HKDC1 overexpression markedly enhanced tumorigenic capacity, exhibiting significantly increased tumor mass and volumetric parameters relative to control group. In contrast, HKDC1 knockdown suppressed tumor growth, with smaller volumes and lower weights observed (Fig. 2F, Figure S1D). IHC analysis of xenografts demonstrated elevated Ki-67 proliferation in HKDC1-overexpressing tumors but suppressed levels with HKDC1 knockdown (Fig. 2G).
Overall, these experimental results consistently indicate that HKDC1 overexpression promotes the proliferation of HCC cells both in vitro and in vivo.
Fig. 2. Modulation of HKDC1 expression alters HCC proliferation in vitro. A) Western blot analysis confirmed HKDC1 expression levels in MHCC97H cells transfected with an HKDC1-overexpressing plasmid and Hep3B cells transfected with shRNA targeting HKDC1 (sh-HKDC1#1, sh-HKDC1#2). B) CCK-8 assays demonstrated differences in cell proliferation between HKDC1-modulated groups and controls. C) EdU staining revealed variations in DNA replication activity across experimental groups. Scale bars: 50 μm. D) Colony formation assays showed quantitative differences in colony numbers following HKDC1 expression modulation. E) Apoptosis assays detected altered apoptotic rates in Hep3B and MHCC97H cells upon HKDC1 knockdown or overexpression. F) Subcutaneous xenograft tumors derived from HKDC1-modulated MHCC97H and Hep3B cells exhibited differential growth patterns over 28 days. G) H&E, Ki67, and HKDC1 IHC staining in xenograft tumors. Scale bars: 50 μm. Data are presented as mean ± SD
HKDC1 is involved in fatty acid metabolism in HCC cells
To explore the specific mechanisms through which HKDC1 affects the biological behavior of HCC cells, we performed RNA-seq on cell samples overexpressing HKDC1 and control cells. The heatmap of differentially expressed genes (DEGs) is shown in Fig. 3A. Gene Set Enrichment Analysis (GSEA) of DEGs indicated that the fatty acid metabolism pathway was upregulated upon HKDC1 overexpression (Fig. 3B), suggesting that HKDC1 may promote tumor cell proliferation through lipid metabolism. Using the TIMER 2.0 online tool, we analyzed the correlation between HKDC1 and the expression of key lipid synthesis enzymes, and found that HKDC1 expression was positively correlated with the expression of several lipid synthesis enzymes (Fig. 3C). To further validate these results, we performed qRT-PCR and Western blotting to confirm the relationship between HKDC1 and these key enzymes. These results indicated that HKDC1 knockdown suppressed key lipogenic enzymes, whereas overexpression enhanced their expression (Fig. 3D-E). Additionally, the measurement of triglyceride, cholesterol, and lipid droplet formation in cells revealed that HKDC1 promotes lipid synthesis (Fig. 3F-G). To determine whether HKDC1 also affects the lipid synthesis in vivo, we conducted nile red staining using the subcutaneous tumors harvested from the animal model. The results showed that HKDC1-overexpressed tumor tissue has stronger fluorescence intensity, while knocking down led to less lipid accumulated (Fig. 3H). These results suggest that HKDC1 can promote lipid synthesis in HCC cells.
Fig. 3. Association of HKDC1 with fatty acid metabolic pathways in HCC cells. A) Heatmap of RNA-seq data showing differentially expressed genes (DEGs) between HKDC1-overexpressing and control HCC cells. B) GSEA analysis of DEGs showing fatty acid metabolism pathway upregulated in HDKC1-overexpression cells. C) Correlation analysis revealed positive associations between HKDC1 and key lipid synthesis enzymes. D) qRT-PCR analysis demonstrated expression changes of lipid synthesis enzymes in HKDC1-modulated HCC cells. E) Western blot analysis confirmed altered protein levels of lipid synthesis enzymes upon HKDC1 knockdown or overexpression. F) Quantification of intracellular triglyceride and cholesterol levels showed metabolic variations between experimental groups. G) Nile Red staining detected differences in lipid droplet accumulation in HCC cells across groups. Scale bars: 50 μm. H) Nile Red staining and quantitative of lipid droplet accumulation in subcutaneous tumors. Scale bars: 100 μm. Data are presented as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001
HKDC1 regulates lipid synthesis and proliferation through ACSBG2
HKDC1 is a protein containing a hexokinase domain that regulates the glycolysis pathway [29]. Previous studies have shown that HKDC1 promotes the progression of various types of tumors through glycolysis, but its role in lipid metabolism has not been well explored.^[10–14]^ To further investigate the specific mechanism by which HKDC1 affects lipid synthesis in HCC cells, we intersected the DEGs from RNA-seq with lipid metabolism genes and identified five potential genes (ACSBG2, ACBD7, ACOX2, GGT1, and OLAH) regulated by HKDC1 (Fig. 4A), the heat map of these five genes is displayed in Figure S2A. We also labeled the top five most significant genes in the volcano plot of DEGs with particular attention paid to ACSBG2 (Figure S2B). To rule out the possibility of stochasticity in RNA-seq, we expanded the sample size and examined the expression of the aforementioned five genes in clinical HCC specimens. The results showed that among 60 paired HCC samples, only ACSBG2 exhibited a significant difference in expression levels between normal and tumor tissues (Figure S2C). ACSBG2 is an acyl-CoA synthetase that promotes lipid synthesis [30]. It is reported that ACSBG2 is related to lung adenocarcinoma progression, whether it is associated with HCC remains unclear [31]. We first analyzed the correlation between HKDC1 and ACSBG2 expression, the result showed that ACSBG2 has positive correlation with HKDC1 in 60-paired HCC tissues (Fig. 4B). Next, we analyzed the impact of ACSBG2 expression on patients’ survival. The result showed that high level of ACSBG2 expression is associated with poorer prognosis compared to those with lower expression (Fig. 4C).
Next, to further confirm the regulatory effects between HKDC1 and ACSBG2, we analyzed ACSBG2 protein level in HKDC1 knockdown and overexpression cell lines (Fig. 4D). To explore whether ACSBG2 participates in HKDC1-promoted lipid synthesis, we manipulated the expression of HKDC1 and ACSBG2 to perform Nile Red staining and examine the levels of key lipid synthesis enzymes (Fig. 4E-F). The results showed that the ACSBG2 knockdown reversed HKDC1-driven lipogenesis, demonstrating its essential role in mediating this process. Recent studies have shown that lipid metabolism reprogramming is a key factor in promoting tumor growth [32]. Through Edu staining (Fig. 4G), CCK-8 assays (Fig. 4H), colony formation assays (Fig. 4I), wound healing assays (Figure S2D), and Transwell assays (Figure S2E), we found that the promoting effect of HKDC1 on HCC cell functions could be reversed by ACSBG2 knockdown. Collectively, these results suggest that HKDC1 regulates HCC cell lipid synthesis and proliferation by controlling ACSBG2 expression.
Fig. 4HKDC1-associated lipid metabolic reprogramming and functional interactions with ACSBG2 in HCC. A) Intersection analysis of DEGs and lipid metabolism-related genes identified five candidate targets, with ACSBG2 showing consistent correlation with HKDC1 expression in expanded clinical samples. B) Correlation analysis between HKDC1 and ACSBG2 expression in 60 clinical samples. R^2^ = 0.366, P < 0.001. C) ACSBG2 expression and survival analysis in 60 clinical HCC patients. Log-rank P = 0.0073. D) Western blot analysis confirmed expression covariation between HKDC1 and ACSBG2 in HCC-modulated cells. E) Western blot analysis detected corresponding changes in lipid synthesis enzymes across experimental groups. F) Nile Red staining displayed altered lipid droplet accumulation patterns upon combinatorial modulation of HKDC1 and ACSBG2. Scale bars: 50 μm. G) EdU staining revealed proliferation differences linked to HKDC1/ACSBG2 expression states. Scale bars: 50 μm Photos of Edu and DAPI are shown in the Figure S2F. H) CCK-8 assays demonstrated proliferative profile changes associated with HKDC1/ACSBG2 interaction. I) Colony formation assays quantified clonogenic capacity variations in dual-modulation models. Data are presented as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001
HKDC1-ASS1 interaction drives ACSBG2 transcription with Acetyl-CoA-Dependent H3 acetylation
Previously, we found that HKDC1 upregulates ACSBG2 expression, but there is no evidence to suggest that HKDC1 directly promotes gene transcription and expression. Therefore, we hypothesized that there might be an intermediate regulatory factor between HKDC1 and ACSBG2. To clarify the specific mechanism by which HKDC1 upregulates ACSBG2, we performed immunoprecipitation (IP) using an anti-HKDC1 antibody on protein extracts from MHCC-97 H cells overexpressing HKDC1, followed by mass spectrometry analysis. The results of IP-MS analysis showed that ASS1 ranked highest (Fig. 5A). ASS1 is argininosuccinate synthase 1, a key enzyme in the urea cycle. It is reported that ASS1 converts glutamine to acetyl-CoA, which is a fundamental unit for lipid biosynthesis [23, 33]. We proposed that the interaction between ASS1 and HKDC1 may contribute to tumor cell lipid synthesis. Firstly, to further confirm the results from MS analysis, we performed co-immunoprecipitation (Co-IP) experiments and verified that ASS1 interacts with HKDC1 (Fig. 5B). Additionally, immunofluorescence images showed that ASS1 and HKDC1 co-localized in the cytoplasm, further confirming the interaction between these two proteins (Fig. 5C).
Given that ASS1 elevates lipid biosynthesis precursors acetyl-CoA, we hypothesize it contributes to HKDC1/ACSBG2-mediated lipid synthesis regulation by modulating precursor availability. To explore this, we examined the intracellular acetyl-CoA levels in ASS1 modulated cells. Results demonstrated that ASS1 knockdown significantly reduced acetyl-CoA levels, which were further decreased by combined HKDC1 knockdown, while HKDC1 overexpression partially restored them. Conversely, ASS1 overexpression elevated acetyl-CoA concentrations, an effect that was partially reversed by HKDC1 co-knockdown or synergistically enhanced with HKDC1 co-overexpression (Fig. 5D). The above results suggest that HKDC1 and ASS1 can synergistically regulate the levels of acetyl-CoA. We previously hypothesized that accumulated acetyl-CoA could promote the activation of the lipid synthesis process by stimulating the expression of lipid synthesis enzyme ACSBG2. To validate this, we treated cells with varying concentrations of acetyl-CoA and measured the expression of ACSBG2. The results showed that ACSBG2 expression increased with higher acetyl-CoA concentrations (Fig. 5E), indicating that acetyl-CoA activates ACSBG2 transcription.
Based on previous studies indicating that acetyl-CoA can influence histone acetylation levels, we hypothesized that the upregulation of ACSBG2 might be mediated through acetyl-CoA-induced histone acetylation [34]. To test this hypothesis, we first performed a luciferase reporter assay, which confirmed that acetyl-CoA positively regulates the transcriptional activity of ACSBG2 (Fig. 5F). Subsequently, we treated cells with citrate to modulate intracellular acetyl-CoA levels and assessed the corresponding histone acetylation status. The results showed that as acetyl-CoA levels increased, the acetylation level of histone H3 exhibited a gradual increase, while no significant change was observed for histone H4 (Fig. 5G). These results indicated that acetyl-CoA induces histone H3 acetylation in a dose-dependent manner, which is highly likely to be the underlying cause for the transcriptional activation of ACSBG2.
Fig. 5HKDC1-ASS1 Interaction Drives ACSBG2 Transcription with Acetyl-CoA-Dependent H3 Acetylation. A) HKDC1 IP-MS identifies ASS1 as top interactor. B) Co-immunoprecipitation (Co-IP) assays confirmed physical interaction between HKDC1 and ASS1. C) Confocal microscopy revealed cytoplasmic co-localization of HKDC1 and ASS1. Scale bar: 10 μm. D) Knockdown or overexpression of ASS1 and HKDC1 demonstrates their cooperative control over acetyl-CoA homeostasis. E) ACSBG2 expression in response to increasing acetyl-CoA concentrations F) Normalized ACSBG2 promoter activation fluorescence intensity ratio following treatment with different citrate concentrations G) Histone H3 and H4 acetylation levels after treatment with different citrate concentrations. Data are presented as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001
HKDC1 interacts with ASS1 to attenuate its Ubiquitin-Dependent degradation
To figure out the specific interaction region between HKDC1 and ASS1, we first employed computational protein–protein docking simulations. These analyses predicted two potential binding interfaces, notably involving residues HKDC1ᴰ²⁴⁶–ASS1ᴷ³¹⁶ and HKDC1ᴿ⁷⁹³–ASS1ᴱ¹²⁹ (Fig. 6A).
To experimentally validate these predictions, we designed a series of truncation constructs for both HKDC1 and ASS1, as schematically summarized in Fig. 6B. Subsequently, co-immunoprecipitation (Co-IP) assays were performed in HEK 293T cells. When HIS-HKDC1-WT was co-expressed with various FLAG-tagged ASS1 truncations, we observed that deletion of ASS1 amino acids 310–412 completely abrogated the interaction. Reciprocally, Co-IP experiments with FLAG-ASS1-WT and different HIS-tagged HKDC1 truncations demonstrated that the HKLS1 domain of HKDC1 is indispensable for binding (Fig. 6C-D). Collectively, these data confirm the critical role of the ASS1 region (310–412 aa) and the HKDC1 HKLS1 domain, and provide experimental evidence supporting the molecular docking prediction, specifically implicating the HKDC1ᴰ²⁴⁶–ASS1ᴷ³¹⁶ site as a key structural determinant for the interaction.
Next, we investigated whether there was a regulatory relationship between HKDC1 and ASS1, qRT-PCR and Western blot analyses revealed that HKDC1 overexpression upregulated ASS1 protein levels, whereas knockdown downregulated them (Fig. 6E). However, ASS1 mRNA remained unchanged across HKDC1-modulated groups (Fig. 6F). Thus, we hypothesized that HKDC1 regulates ASS1 expression through post-translational modifications. Ubiquitination is a widespread post-translational modification that marks proteins for degradation. This process is crucial in many biological processes, such as the degradation of oncogenes [35]. After treating cells with MG-132, we observed that the downregulation of ASS1 caused by HKDC1 knockdown was reversed (Fig. 6G). Furthermore, after treatment with cycloheximide (CHX) on HKDC1-overexpressing and control cells, we found that the stability of ASS1 protein was enhanced in the HKDC1-overexpressing group compared to the control group (Fig. 6H). These results suggest that HKDC1 participates in regulating ASS1 protein degradation. We further verified the effect of HKDC1 on the ubiquitination level of ASS1, and found that HKDC1 overexpression decreased the ubiquitination level of ASS1, while knockdown of HKDC1 showed the opposite effect (Fig. 6I). These results suggest that HKDC1 inhibits the ubiquitin-mediated degradation of ASS1.
Taken together, these results suggest that HKDC1 interacts with ASS1 and inhibits the ubiquitin-mediated degradation of ASS1.
Fig. 6HKDC1 Interacts with ASS1 to Attenuate Its Ubiquitin-Dependent Degradation. A) Schematic diagram of the HKDC1-ASS1 interaction based on molecular docking simulation. B) Schematic of the Flag-tagged ASS1 and His-tagged HKDC1 truncated variants. C-D) Co-IP analysis of His-HKDC1-WT with ASS1 truncation mutants and Flag-ASS1-WT with His-HKDC1 truncation mutants. E) Western blot analysis showed protein-level covariation between HKDC1 and ASS1 in genetically modulated cells. F) qRT-PCR analysis indicated no corresponding changes in ASS1 mRNA levels upon HKDC1 modulation. G) MG-132 treatment rescued ASS1 protein levels in HKDC1-knockdown cells. H) Cycloheximide (CHX) chase assays demonstrated delayed ASS1 degradation in HKDC1-overexpressing cells. I) Ubiquitination assays revealed inverse correlations between HKDC1 expression levels and ASS1 ubiquitination status. Data are presented as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001
HKDC1 promotes resistance to lenvatinib in HCC cells
Previous studies have shown that abnormal lipid metabolism is one of the major factors contributing to drug resistance in tumors [36]. We previously established that HKDC1 promotes lipid synthesis in HCC cells via interacting with ASS1, we sought to investigate whether the increased lipid synthesis caused by HKDC1 affects the efficacy of lenvatinib in HCC cells. We compared the IC50 values between normal MHCC97H and Hep-3B cell lines, as well as HKDC1 knockdown and overexpression cells. We found that HKDC1 knockdown significantly reduced the IC50 of lenvatinib in HCC cells, whereas overexpression of HKDC1 showed the opposite effect (Fig. 7A). Moreover, lenvatinib treatment induced significantly less apoptosis in HKDC1 overexpression cells compared to the control group (Fig. 7B). Colony formation assays showed that HKDC1 overexpression significantly weakened the sensitivity of HCC cell lines to lenvatinib (Fig. 7C). Additionally, we used xenograft tumor models to assess whether HKDC1 overexpression could affect the efficacy of lenvatinib in vivo. Consistent with the in vitro data, HKDC1 overexpression markedly promoted the resistance to lenvatinib (Fig. 7D-E). Moreover, quantitative analysis revealed a tumor growth inhibition (TGI) of 68.9% in the Vector Control + Lenvatinib group, which was significantly reduced to 46.9% in the HKDC1 OE + Lenvatinib group at the endpoint, underscoring the role of HKDC1 in driving therapeutic resistance. Throughout the study, all mouse groups exhibited comparable and steady body weight gain, indicating that the administered dose of lenvatinib was well-tolerated without systemic toxicity. This observation rules out potential off-target effects that could confound the tumor growth data (Fig. 7F). Microscopic examination revealed that HKDC1 overexpression enhanced the lipid accumulation in tumor tissue and rescued the tumor necrosis induced by lenvatinib, further demonstrating that HKDC1 confers drug resistance in HCC (Fig. 7G). Overall, these data suggest a potential mechanistic link between HKDC1 expression and lenvatinib resistance in HCC patients.
Fig. 7HKDC1 expression status correlates with therapeutic response in HCC models. A) Dose-response assays revealed differential IC50 values of lenvatinib between HKDC1-modulated HCC cells (knockdown vs. overexpression) and controls. The concentration of Lenvatinib treated with cells are ste as: 0.01, 0.1 0.5, 1, 2, 4, 10, 20µM. B) Apoptosis assays demonstrated decreased apoptotic rates in HKDC1-overexpression cells upon lenvatinib treatment. C) Colony formation assays quantified reduced clonogenic survival under lenvatinib exposure in HKDC1-overexpression groups. D) In vivo xenograft models showed weakened tumor growth suppression by lenvatinib in mice implanted with HKDC1- overexpression HCC cells. (Dose: 10 mg/kg/d for 28 days.) E) Weight and volume growth curve of the xenograft tumors. F) The comparable body weight gain across all groups demonstrates the absence of off-target toxicity. G) H&E staining of necrosis and Nile red staining in xenograft tumors. Scale bars: HE: 50 μm; Nile red: 100 μm. Data are presented as mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001
Discussion
In this study, we have elucidated a novel metabolic signaling axis in HCC comprising HKDC1, ASS1, and ACSBG2, which collectively drive de novo lipogenesis to support tumor growth and proliferation. Our results demonstrate that HKDC1 stabilizes ASS1 by inhibiting its ubiquitin-mediated degradation, enhancing the conversion of glutamine-derived carbons into acetyl-CoA [37]. Our molecular docking and co-IP assays with truncated variants pinpointed the HKDC1 HKLS1 domain and the ASS1 (310–412 aa) region as the critical structural basis for this interaction. This acetyl-CoA accumulation transcriptionally activates ACSBG2, a critical enzyme in fatty acid synthesis [38]. Importantly, we confirmed that acetyl-CoA promotes ACSBG2 promoter activity and induces specific histone H3 acetylation, establishing a direct epigenetic mechanism. This metabolic cascade not only meets the biosynthetic demands of HCC cells but also positions HKDC1 as a central regulator bridging glucose and lipid metabolism, extending its previously established role in glycolysis [39].
Metabolic reprogramming is a hallmark of HCC, characterized by enhanced glycolytic flux (the Warburg effect) and dysregulated lipid metabolism to fulfill the bioenergetic and anabolic needs of cancer cells [40, 41]. Prior research has underscored HCC’s dependence on glycolysis for energy and intermediates, with citrate-derived acetyl-CoA serving as a key substrate for fatty acid synthesis [33]. Our findings build on this foundation by identifying a specific mechanism whereby HKDC1, traditionally linked to glycolysis, orchestrates lipid metabolism through the ASS1-ACSBG2 axis [9]. This crosstalk aligns with evidence that metabolic enzymes often play multifaceted roles in cancer, as seen in ASS1’s shift from its canonical urea cycle function to a regulator of lipid metabolism and ferroptosis resistance. [2–38] By integrating glucose and lipid pathways, this axis highlights the metabolic plasticity driving HCC progression and therapeutic resistance [44].
The originality of our study lies in uncovering a previously unrecognized signaling network linking HKDC1, ASS1, and ACSBG2 in HCC. HKDC1 overexpression has been associated with aggressive tumor phenotypes across malignancies, including lung, gastric, and pancreatic cancers, yet its role in stabilizing ASS1 introduces a novel post-translational regulatory mechanism, akin to the stabilization of other metabolic enzymes in tumorigenesis.[13, 14, 37] Similarly, ASS1’s redirection of glutamine-derived carbons to acetyl-CoA production, beyond its ammonia detoxification role, reframes it as a metabolic hub in HCC [22, 37, 43]. Moreover, the transcriptional upregulation of ACSBG2 by acetyl-CoA adds a distinct regulatory layer to lipid homeostasis, differentiating this axis from pathways involving fatty acid synthase (FASN) [16, 33]. This integrated mechanism offers a comprehensive perspective on how HCC cells exploit metabolic networks to sustain proliferation, addressing a critical gap in hepatocarcinogenesis research.
The therapeutic potential of targeting the HKDC1-ASS1-ACSBG2 axis is significant, especially given the limited efficacy of current HCC treatments like lenvatinib. Our in vivo data directly demonstrate that HKDC1 overexpression confers lenvatinib resistance, with quantitative analysis revealing a significant reduction in tumor growth inhibition (TGI) from 68.9% to 46.9%. Disrupting this axis—e.g., by inhibiting HKDC1 or ASS1—could reduce acetyl-CoA availability, subsequently downregulating ACSBG2 and lipogenesis [45]. Given lipid metabolism’s roles in membrane biogenesis, energy homeostasis, and oncogenic signaling, such interventions could impede tumor growth and enhance sensitivity to existing therapies by inducing metabolic stress [16, 46]. The axis’s reliance on glutamine metabolism also suggests synergy with glutaminase inhibitors, a strategy gaining traction in cancer therapy [44]. However, therapeutic development must consider off-target effects and the preservation of normal liver function, which depends on metabolic balance [47]. Notably, ASS1’s dual role in suppressing HCC metastasis via glycolysis attenuation contrasts with its lipogenic function here, highlighting context-specific metabolic regulation that warrants further exploration [48].
However, our study has several limitations that warrant discussion and frame future research directions. First, although mechanistically insightful, our studies using HCC cell lines and immunodeficient mouse models cannot capture the full heterogeneity and microenvironment of human HCC. The use of nude mice enabled us to define the cell-autonomous role of this axis in resistance but excluded adaptive immune interactions. Future work should employ immunocompetent models or patient-derived organoids to validate these findings in a physiologically relevant context and explore immune and metabolic crosstalk. Second, while we have established that HKDC1 inhibits ASS1 ubiquitination, the specific E3 ligase involved remains to be fully elucidated, representing a key focus for ongoing structural studies. Third, our focus was on the anabolic arm of lipid metabolism; future work employing isotope tracer assays will be crucial to determine whether HKDC1 concurrently affects fatty acid β-oxidation, providing a holistic view of lipid remodeling. Previous work suggests ASS1 has dual roles in HCC, suppressing metastasis via glycolysis attenuation while promoting lipogenesis [48]. However, as our clinical cohort comprised only non-metastatic samples, we could not assess ASS1’s expression or function in metastasis. This precluded a deeper understanding of how its context-dependent roles are regulated. Future studies will incorporate metastatic HCC samples and combine systematic in vitro assays with ASS1-modulated models to directly verify its antimetastatic and metabolic functions, and to validate the prognostic value of this axis. Finally, and of paramount translational importance, our preclinical data nominate this axis as a biomarker; however, definitive validation requires correlating its expression with treatment outcomes in cohorts of lenvatinib-treated HCC patients, an endeavor we are actively pursuing through expanded clinical collaborations.
In conclusion, our study unveils a pivotal metabolic signaling axis in HCC that integrates multiple metabolic pathways to drive lipogenesis and tumor progression. By clarifying the interplay of HKDC1, ASS1, and ACSBG2, we offer fresh insights into HCC’s metabolic vulnerabilities, advancing our understanding of hepatocarcinogenesis. These findings highlight the therapeutic promise of targeting this axis and reinforce metabolic reprogramming as a cornerstone of HCC pathology, paving the way for innovative approaches to improve outcomes in this formidable malignancy.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1
Supplementary Material 2
Supplementary Material 3
Supplementary Material 4
Supplementary Material 5
The reference list from the paper itself. Each links out to its DOI / PubMed record.