Introduction
Hepatocellular carcinoma (HCC) is the most common primary liver cancer, accounting for over 85% of cases and ranking as the fourth leading cause of cancer-related deaths worldwide.1,2 Despite advancements in HCC treatment, prognosis remains poor, particularly in advanced-stage disease.3,4 For HCC caused by hepatitis B virus (HBV) and other etiological factors, surgical resection and liver transplantation are primary treatment options.5 However, many patients are diagnosed at an advanced stage, limiting curative interventions and reducing the five-year survival rate to less than 5%.6–8
Sorafenib, the first FDA-approved systemic therapy for advanced HCC, functions as a multi-targeted tyrosine kinase inhibitor that inhibits tumor growth and angiogenesis.9,10 However, its clinical efficacy is limited by the rapid development of drug resistance.11–13 HCC cells acquire resistance to sorafenib through altered cellular signaling, enhanced drug metabolism, and inhibition of apoptosis.14–16 Overcoming this resistance is critical for improving treatment outcomes.17,18 Therefore, identifying sorafenib sensitizers that counteract resistance mechanisms is an urgent clinical need.
Lipid metabolism plays a key role in HCC progression, as cancer cells reprogram lipid pathways to sustain rapid proliferation.19–21 Fatty acid-binding proteins facilitate lipid transport and metabolism, with fatty acid-binding protein 3 (fabp3) linked to HCC growth, invasion, and metastasis.22–24fabp3 influences cancer cell behavior by directing fatty acid deposition into lipid droplets, which may contribute to sorafenib resistance by mitigating lipid peroxidation-induced cell death.25–28. Additionally, reactive oxygen species (ROS), as metabolic by-products, play a key role in tumor progression and therapy responses.29–31
Oleanolic acid (OA), a natural pentacyclic triterpenoid, exhibits antioxidant, anti-inflammatory, and anticancer properties.32–34 OA has demonstrated anti-tumor effects in various cancers, including HCC.35–37 However, its role in sorafenib resistance and lipid metabolism remains poorly understood. This study investigates OA’s potential to modulate fabp3 expression, thereby influencing lipid metabolism and sorafenib sensitivity in resistant HCC cells. Our findings suggest that OA significantly inhibits migration and invasion in sorafenib-resistant HCC cells, highlighting its potential as a therapeutic adjunct to overcome drug resistance.
This study fills a crucial gap by elucidating OA’s impact on sorafenib resistance in HCC (HBV) through fabp3 modulation. By addressing this resistance mechanism, we aimed to enhance the therapeutic efficacy of sorafenib, offering novel treatment strategies for patients with advanced HCC (HBV) who respond poorly to current therapies.
Methods
Cell culture
The human HCC cell lines HepG2 (derived from hepatoblastoma) and Huh7 (derived from well-differentiated hepatocellular carcinoma, HBV-negative) were obtained from the Chinese Academy of Sciences Cell Bank (Shanghai). The cells were cultured in Dulbecco’s Modified Eagle Medium (G4650, Servicebio, China) supplemented with 10% fetal bovine serum (v/v) and maintained at 37°C with 5% CO2.
Generation of drug-resistant cells
Huh7 and HepG2 cell lines were treated with sorafenib (53468ES70, Yeasen Biotech, Shanghai, China) at a concentration of 10 µM for 48 h. Following the initial treatment, the sorafenib concentration was gradually increased. The cell lines’ resistance to sorafenib was assessed by determining the resistance drug index, with a value of ≥2.5 considered significant.
Transfection and stable cell line generation
The human fabp3 plasmid was cloned into the pcDNA3.1-HA-C vector (QP1397, Qiyunbio, China). The empty plasmid and pcDNA3.1-fabp3 plasmid were transfected into HCC cells using the PEI 40K transfection reagent (G1802, Servicebio, China), following the Lipofectamine 2000 transfection protocol (11668-019, Invitrogen, CA, USA). Transfected cells were used for subsequent experiments after 24 h or 48 h.
To establish stable cell lines, HCC cells were transfected with lentivirus-shfabp3 and lentivirus-shNC (China). Cells were selected with 1 µg/mL puromycin for 14 days. After 12 h of infection, the medium was replaced with fresh culture medium and incubated for 96 h. Cells were then selected with 10 µg/mL puromycin (MABE341, Sigma-Aldrich, USA) for at least one week to establish stable transfected cell lines.
Cell viability assay
Cell viability of Huh7 and HepG2 cells was assessed using the CCK-8 assay (C0037, Beyotime, China). Following treatment, cells (2 × 103) were seeded into 96-well plates and cultured at 37°C with 5% CO2. On days 1, 2, and 3, 10 µL of CCK-8 reagent and 90 µL of serum-free medium were added to each well. After 1 h of incubation, absorbance was measured at 450 nm.
Quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was extracted using a universal RNA extraction kit (R219-50, GeneBetter, China) and reverse transcribed using the PrimeScript™ RT reagent kit (Takara, China). The relative quantification of the fabp3 gene was standardized to the GAPDH gene (Sangon Biotech, China). Primer sequences are provided in Supplementary Table 1.
After recording the Ct values for the target gene in each well, the relative expression of the product was calculated using the 2−ΔΔCt method, with GAPDH as the internal control. The formula used was: ΔΔCt = (average Ct value of target gene in experimental group - average Ct value of housekeeping gene in experimental group) - (average Ct value of target gene in control group - average Ct value of housekeeping gene in control group).
Flow cytometry analysis
Treated cells were resuspended in PBS. The Annexin V-FITC Apoptosis Detection Kit (C1062S, Beyotime, China) was utilized to assess apoptotic cell populations. Gating and analysis were conducted using FlowJo v10 software (FlowJo, USA).
Colony formation assay
Cells were seeded at a density of 1 × 104 cells per well in six-well plates and cultured in Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum and specific concentrations of sorafenib. After one week, visible colonies were formed. Cells were then washed with PBS, fixed with 4% formaldehyde for 15 m, and stained with crystal violet for 25 m.
Transwell assay
Invasion and migration experiments were performed using Transwell chambers with 8.0 µm pore size membranes. The upper chamber was filled with 200 µL of serum-free medium containing suspended cells, while the lower chamber contained 600 µL of medium with 10% fetal bovine serum. For invasion assays, chambers were pre-coated with Matrigel. After 24 h for migration assays and 48 h for invasion assays, non-migratory cells on the top membrane surface were gently washed with PBS. Cells that penetrated the bottom membrane were stained with 1% crystal violet solution and observed under an optical microscope.
Wound healing experiment
Cells were seeded in six-well culture plates and incubated overnight until they reached approximately 95% confluence. The next day, a controlled scratch was made on the cell monolayer using a sterile 1 mL pipette tip, ensuring the desired scratch intensity. After removing cell debris, fresh medium was added for continued cultivation. Wound images were captured at 0, 12, and 24 h.
Antibodies and Western blot (WB)
The cells were lysed, protein extraction was performed using RIPA lysis buffer (KGB5203, Keygen Biotech, Jiangsu, China), and protein concentration was measured using the enhanced BCA protein assay kit (KGB5203, Keygen Biotech, Jiangsu, China). Protein transfer was conducted using SDS-PAGE electrophoresis (G2003, Servicebio, China) and nitrocellulose membranes. Specific primary antibodies included anti-fabp3 (10676-1-AP, 1:1,000, Proteintech, China), anti-AKT (60203-2-Ig, 1:1,000, Proteintech, China), anti-phospho-AKT (28731-1-AP, 1:1,000, Proteintech, China), anti-PI3K (60225-1-Ig, 1:1,000, Proteintech, China), anti-SNAIL (13099-1-AP, 1:1,000, Proteintech, China), anti-E-cadherin (20874-1-AP, 1:1,000, Proteintech, China), anti-N-cadherin (22018-1-AP, 1:1,000, Proteintech, China), anti-PARP1 antibody (Cat# ab191217, Abcam, USA), anti-cleaved PARP1 antibody (Cat# ab32064, Abcam, USA), anti-Caspase-3 antibody (Cat# ab32351, Abcam, USA), anti-cleaved Caspase-3 antibody (Cat# E83-77, Abcam, USA), and β-actin (81115-1-RR, 1:5,000, Proteintech, China), incubated overnight at 4°C. Following primary antibody incubation, rabbit or mouse HRP-conjugated secondary antibodies (KFA025, 1:10,000, Proteintech, China) were applied and incubated at room temperature for two hours. Finally, WB images were detected using ChemistarTM High-sig ECL WB Substrate. Uncropped WB gels are provided in Supplementary File 1.
Lipid analysis
This study employed the common BODIPY 493/503 fluorescence dye, commonly used for neutral lipid staining. Following cell washing with PBS, cells were co-incubated with 4 mg/mL BODIPY 493/503 solution (HY-W090090, MedChemExpress, China) at room temperature for 10 m. Nuclear staining was performed using DAPI (C1002, Beyotime, China) to visualize the cells, and observations were conducted using a fluorescence microscope. Intracellular triglyceride (TG) content was quantitatively measured using a TG assay kit (A110-1-1, Nanjing Jiacheng Bioengineering Institute, Nanjing, China).
Mitochondrial membrane potential analysis using JC-1 staining
JC-1 staining (G1515, Servicebio, China) was performed by incubating cells for 20 m in 200 nM JC-1 medium. The ratio of red fluorescence (indicating high mitochondrial membrane potential) to green fluorescence (representing total mitochondria) was calculated to assess mitochondrial membrane potential.
Detection of ROS
Mitochondrial ROS levels were assessed using DCFH-DA (G1515, Servicebio, China). After specific treatments, cells were washed three times with PBS and incubated with 5 µM DCFH-DA (1:1,000) in Dulbecco’s Modified Eagle Medium for 30 m. Fluorescence signals were recorded using the IBE2000 inverted fluorescence microscope, with an excitation wavelength of 510 nm and an emission wavelength of 580 nm.
Xenograft model
All animal care and experimental procedures were approved by the Animal Ethics Committee of Jiamusi University (2022-500-190). Four- to six-week-old BALB/c nude mice (HFK Bioscience Co., Ltd, Beijing, China) were randomly divided into four groups (N = 10 per group). The first group received a subcutaneous injection of 2 × 106 HepG2-SR cells. The cell suspension was injected subcutaneously into the right axillary region of the mice. The second group received a subcutaneous injection of 2 × 106 HepG2-SR cells and was administered sorafenib orally at 30 mg/kg daily, divided into two doses.
The third group received a subcutaneous injection of 2 × 106 HepG2-SR cells, oral administration of sorafenib at 30 mg/kg daily, divided into two doses, along with nobiletin at 4 mg/kg daily, divided into two doses.38 Nobiletin, a flavonoid compound typically found in the peel of citrus fruits, was dissolved in the aqueous solvent DMSO. The fourth group received a subcutaneous injection of 2 × 106 sh-fabp3-HepG2-SR cells, followed by oral administration of sorafenib at 30 mg/kg daily, divided into two doses.
After two months, the mice were euthanized by cervical dislocation, and tumor samples were subjected to immunohistochemical staining. All animal experiments were conducted following the ARRIVE guidelines.
Immunohistochemistry
Tissue sections were first defatted at 65°C for 1.5 h and dewaxed with xylene. Antigen retrieval was achieved by heating sections in boiling 0.01 M sodium citrate buffer for two minutes. To prevent nonspecific staining, tissue microarrays were blocked with 10% normal goat serum for one hour. Then, the primary antibody, rabbit anti-fabp3 (PA5-13461; 1:2,000; Thermo, USA), was incubated with the sections overnight at 4°C. After three washes with 0.1 M PBS (5 m each), a goat anti-rabbit IgG secondary antibody (ab6721, 1:1,000, Abcam, Cambridge, UK) was applied at 37°C for 20 m, followed by horseradish peroxidase-labeled streptavidin working solution (0343-10000U, Easybio, Beijing, China) for another 20 m at 37°C. Staining was performed using the DAB detection kit (G1212, Servicebio, China), and the sections were counterstained with hematoxylin and sealed with neutral resin.
After staining, tissue sections were mounted and photographed using a Nikon digital microscope camera. The expression levels of the staining results were scored on a proportional scale: the area of immunopositive staining (0%, 0; 1–25%, 1; 26–50%, 2; 51–75%, 3; 76–100%, 4) was multiplied by staining intensity (0 = negative, 1 = weakly positive, 2 = moderately positive, 3 = strongly positive). Two pathologists independently assigned scores. The staining intensity results of different groups were statistically analyzed to calculate the mean and standard deviation.
RNA-seq
Total RNA was first treated with the Ribo-Zero™ Magnetic Kit (MRZH116, Epicentre Technologies, USA) to remove ribosomal RNA. Sequencing libraries were then prepared using the NEB Next Ultra RNA Library Prep Kit (E7770, NEB, USA) according to the manufacturer’s protocol. RNA fragments (∼300 base pairs) were generated using the NEB Next First Strand Synthesis Reaction Buffer (5x). First-strand cDNA was synthesized using reverse transcriptase and random primers, followed by second-strand cDNA synthesis using the Second Strand Synthesis Reaction Buffer (10x) with dUTP Mix.
Subsequently, cDNA fragments were end-repaired, including adding polyA tails and sequencing adapters. After ligation of the Illumina sequencing adapters, the second strand of cDNA was digested using USER Enzyme (M5505S, NEB, USA) to construct strand-specific libraries. The library DNA was then polymerase chain reaction-amplified, purified, and enriched. The libraries were quantified and validated using the Agilent 2100 and the KAPA Library Quantification Kit (TAQDKB, KAPA Biosystems). Finally, paired-end sequencing was performed on the Illumina NextSeq CN500 sequencer.
Raw data was pre-processed for quality using the Trimmomatic tool, which included: (1) removal of adapters; (2) removal of low-quality reads; (3) trimming of low-quality bases from the 3′ and 5′ ends; (4) statistics on raw sequencing volume, effective sequencing volume, Q30, and GC content, with a summary of read counts throughout the quality control process. Filtered high-quality reads were then aligned to the genome in the NCBI database using hisat2 software.
Differential expression analysis of mRNA was performed using the R package “edgeR” based on mRNA read counts. Differential genes were screened with criteria set at |log2FC|>1 and p-value < 0.05.
Gene Ontology/Kyoto Encyclopedia of Genes and Genomes (KEGG) Analysis
Differentially expressed genes were identified using the DESeq tool. After clustering and filtering, Gene Ontology and KEGG pathway analyses were performed. Pathways with a false discovery rate < 0.05 were retained for further analysis.
Data analysis
Statistical analysis and image acquisition were conducted using GraphPad Prism 8. Inter-group differences were assessed using a paired two-tailed Student’s t-test or one-way analysis of variance (ANOVA). All results were based on at least three independent experiments, with qRT-PCR performed in triplicate. Data are presented as mean ± standard deviation. Normality and homogeneity of variance were tested first. For normally distributed data with equal variance, an unpaired t-test was used for two-group comparisons. For multiple group comparisons, one-way ANOVA or repeated measures ANOVA was applied, followed by Tukey’s post hoc test. Repeated measures ANOVA was employed for behavioral experiments, with Bonferroni corrections applied post hoc when appropriate. In omics data, p-values were adjusted using the false discovery rate method. Statistical significance was set at p < 0.05, with the following notations: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, and “ns” (not significant).
Results
OA inhibits the migration and invasion capabilities of sorafenib-resistant HCC cells and enhances the effectiveness of sorafenib
Huh7 and HepG2 cells were chronically exposed to sorafenib, successfully establishing sorafenib-resistant cell lines Huh7-R and HepG2-R. Cell viability assays were performed to determine the IC50 values of sorafenib and OA in both resistant and parental cell lines. The results showed that the IC50 of sorafenib was significantly higher in Huh7-R and HepG2-R compared to their parental counterparts. Similarly, the IC50 of OA was also elevated in the resistant cell lines (Fig. 1A). Further analysis assessed the sensitivity of Huh7-R and HepG2-R to different concentrations of OA (0 µM, 20 µM, 40 µM, 60 µM) and the combinatorial effects of OA with sorafenib (10 µM, 20 µM, 30 µM). The results indicated that at 40 µM OA, sorafenib-resistant cells exhibited enhanced sensitivity to sorafenib, with a significant decrease in cell viability. This suggests that OA exerts an inhibitory effect on resistant cells and acts synergistically with sorafenib (Fig. 1B). This observation was further validated by colony formation assays (Fig. 1C) and flow cytometry analysis (Fig. 1D). Additionally, Transwell and wound healing assays demonstrated that OA significantly inhibited the migration and invasion of sorafenib-resistant cells (Fig. 1E, F).
OA treatment decreases fabp3 expression, potentially contributing to the development of drug resistance
To explore the mechanisms underlying OA’s effect on resistance development, RNA sequencing was performed on HepG2-R cells, comparing OA-treated (60 µM, two days) and untreated groups. A total of 344 differentially expressed genes were identified, with 211 genes downregulated in the OA-treated group (Fig. 2A). KEGG pathway analysis revealed significant enrichment in ferroptosis, mineral absorption, and PPAR signaling pathways (Fig. 2B, see Supplementary File 2). Among these, six genes (pck2, fabp3, angptl4, plin4, sorbs1, and acadl) were functionally linked within the PPAR signaling pathway and associated with lipid metabolism.
TG levels were higher in resistant cells than in parental cells and decreased with increasing OA concentration (Fig. 2C). A similar trend was observed using BODIPY 493/503 staining (Fig. 2D). Bioinformatics analysis indicated that fabp3 was highly expressed in HCC tissues, particularly in HBV-related cases, and correlated with poor prognosis (Supplementary Fig. 1A). Based on these observations, fabp3 was selected for further investigation (Fig. 2E). qRT-PCR and Western blot analysis confirmed that fabp3 expression was significantly upregulated in sorafenib-resistant cells compared to parental cells. However, OA treatment reversed this upregulation (Fig. 2F, G). These findings suggest that OA may contribute to reversing sorafenib resistance in hepatocellular carcinoma cells by modulating the PPAR signaling pathway and downregulating fabp3 expression.
fabp3 mitigates sorafenib-induced apoptosis in HCC cells in vitro
Previous studies have identified fabp3 as a specific marker of adipocytes, with its fluorescence primarily localized in the cytoplasm, where lipid droplets (LDs) accumulate.39 Functionally, fabp3 facilitates fatty acid (FA) uptake and storage, promoting triglyceride synthesis and LD accumulation,40,41 which aligns with the findings of this study (Fig. 3A, B). To investigate the potential correlation between fabp3 and sorafenib resistance in HCC, Huh7 and HepG2 cells were transfected with a fabp3 plasmid, and its expression was confirmed by Western blot analysis (Supplementary Fig. 1B). The cells were then treated with Triacsin C (an acyl-CoA synthetase inhibitor that suppresses lipid synthesis), OA (which induces LD formation), or left untreated. Cell viability assays revealed that fabp3 overexpression attenuated sorafenib (10 µM)-induced cell death, an effect that was reversed by Triacsin C. However, OA treatment did not significantly affect the inhibitory action of Triacsin C (Fig. 3C). In hepatic lipid metabolism, triglycerides synthesized from FAs primarily serve as energy reserves in LDs. In this study, OA treatment did not affect cell viability but significantly increased LDs and lipid accumulation. This suggests that fabp3 may enhance sorafenib resistance in HCC by regulating intracellular FA transport rather than directly affecting triglyceride synthesis and LD accumulation. These findings were further supported by colony formation assays and flow cytometry apoptosis assays (Fig. 3D, E).
Inhibiting fabp3 in sorafenib-resistant HCC can restore sensitivity to sorafenib
fabp3 upregulation in sorafenib-resistant cells was assessed using shRNA plasmid interference, with WB analysis confirming the knockdown efficiency (Fig. 1C). Both fabp3 knockdown and Triacsin C treatment significantly reduced TG synthesis and LD accumulation (Fig. 4A, B). Cell viability and colony formation assays demonstrated that shfabp3 knockdown significantly decreased cell survival and colony formation compared to resistant cells, with minimal difference from the Triacsin C-treated group (Fig. 4C, D). Additionally, shfabp3 knockdown induced apoptosis in sorafenib-treated resistant cells, similar to Triacsin C treatment (Fig. 4E). Western blot analysis showed that shfabp3 knockdown increased the protein levels of total and cleaved Caspase-3 and PARP in sorafenib-treated resistant cells, consistent with the trend observed in the Triacsin C-treated group (Fig. 4F).
fabp3 alleviates sorafenib-induced oxidative stress in HCC cells
Following sorafenib treatment, JC-1 monomers (green fluorescence) increased, while JC-1 aggregates (red fluorescence) decreased in Huh7 and HepG2 cells, indicating enhanced mitochondrial depolarization. In resistant cells, JC-1 monomers remained the dominant fluorescent form regardless of sorafenib treatment, suggesting that mitochondrial depolarization is a characteristic feature of sorafenib-resistant cells (Fig. 5A). Further analysis of ROS levels showed that ROS levels in resistant cells were unaffected by sorafenib treatment, whereas OA combined with sorafenib significantly increased ROS-mediated cell death (Fig. 5A). Additionally, OA treatment increased ROS levels in Huh7-R and HepG2-R cells, while fabp3 overexpression suppressed ROS production. However, OA alone did not influence OA-induced ROS generation (Fig. 5B). Consistently, fabp3 overexpression inhibited sorafenib-induced ROS generation, whereas fabp3 knockdown or Triacsin C treatment enhanced ROS production (Fig. 5C).
fabp3 promotes the migration and invasion of human HCC by modulating the PI3K/AKT/Snail signaling pathway
Sorafenib resistance is closely associated with increased invasiveness in HCC cells. Given that OA treatment effectively suppressed fabp3 expression and the migratory capacity of resistant cells, fabp3 is likely to promote HCC cell migration and invasion.
Transwell migration and invasion assays revealed that fabp3 overexpression significantly enhanced migration and invasion in HepG2 and Huh7 cells (Fig. 6A), whereas fabp3 knockdown reduced the invasive capacity in HepG2-R and Huh7-R cells (Fig. 6B). Wound healing assays further confirmed that fabp3 overexpression accelerated wound closure, while fabp3 knockdown delayed the process in resistant cells (Fig. 6C,D). However, Triacsin C treatment did not significantly affect fabp3-mediated migration and invasion (Fig. 6A-D), suggesting the involvement of additional regulatory mechanisms.
TCGA data analysis indicated a correlation between fabp3 expression and the PI3K/AKT pathway (Supplementary Fig. 1D). Additionally, treatment with the AKT inhibitor GSK2141795 effectively suppressed migration and invasion induced by fabp3 overexpression or sorafenib resistance (Fig. 6A-D), suggesting that fabp3 promotes HCC cell migration and invasion through the PI3K/AKT pathway.
Western blot analysis further demonstrated that fabp3 overexpression upregulated EMT-related proteins (N-cadherin and Snail) while downregulating E-cadherin. These effects were reversed by GSK2141795 treatment (Fig. 6E). In resistant cells, fabp3 knockdown increased E-cadherin expression and reduced N-cadherin and Snail levels, whereas Triacsin C treatment had no significant effect on EMT markers (Fig. 6E).
In vivo, OA enhances the sensitivity of HCC cells to sorafenib by targeting fabp3
In the HepG2-R xenograft model, the effects of OA and fabp3 on sorafenib resistance in HCC were evaluated. Tumors were excised for analysis 27 days post-implantation (Fig. 7A). Tumor growth monitoring demonstrated that OA or fabp3 knockdown significantly inhibited tumor volume and weight increase, whereas sorafenib treatment alone failed to effectively suppress tumor growth (Fig. 7B,C). Immunohistochemical analysis revealed reduced fabp3 expression in tumors treated with OA or fabp3 knockdown. Additionally, N-cadherin and Snail expression were downregulated, while E-cadherin levels were increased (Fig. 7D), consistent with in vitro findings.
Discussion
This study reveals that fabp3 binds and transports fatty acids susceptible to ROS damage to LDs, protecting them from peroxidation and thereby reducing lipotoxicity in sorafenib (SOR)-resistant HCC cells induced by HBV and other pathogenic factors. Sorafenib inhibits tumor growth by increasing intracellular ROS levels27,42,43; however, fabp3, through its fatty acid-binding function, may mitigate ROS toxicity induced by SOR, making it a potential mechanism of SOR resistance.44,45 OA effectively inhibits fabp3 expression, overcoming SOR resistance and inducing apoptosis in HCC cells through ROS-mediated mechanisms.46 These findings provide new insights into HCC resistance mechanisms and emphasize the critical role of lipid metabolism in HCC progression and resistance formation.
fabp3, a fatty acid-binding protein, has been identified as a specific marker of adipocytes and is primarily localized in the cytoplasm, where it is closely associated with lipid droplet accumulation.25,47fabp3 facilitates fatty acid uptake and storage, promoting triglyceride synthesis and lipid droplet formation,40,48,49 which aligns with our study’s findings. Moreover, fabp3 promotes HCC cell migration and invasion through the PI3K/AKT/Snail signaling pathway, enhancing the invasive ability of SOR-resistant cells. Under normal physiological conditions, the PPAR signaling pathway regulates glucose and lipid metabolism to maintain energy homeostasis, but in cancer cells, its abnormal activation promotes tumor growth.50 Previous research indicates that simvastatin restores SOR sensitivity in HCC cells by inhibiting HIF-1α/PPAR-γ/PKM2-mediated glycolysis.51 This study confirms that fabp3 may regulate SOR resistance, migration, and invasion in HCC cells through the PPAR and PI3K/AKT/Snail signaling pathways, further emphasizing the clinical significance of fabp3 as a potential therapeutic target.
Our findings demonstrate that OA inhibits fabp3 expression, restores SOR sensitivity in resistant HCC cells, suppresses cancer cell migration and invasion, and reduces the risk of metastasis.46 Although this study did not directly assess the effects of OA and SOR combination therapy, previous research suggests that OA enhances SOR sensitivity in HCC cells by inhibiting HIF-1α/PPAR-γ/PKM2-mediated glycolysis.51 Thus, OA may serve as an important adjunctive therapy for SOR-resistant patients, warranting further investigation into its combined therapeutic potential with SOR. Additionally, in vitro studies indicate that OA enhances tumor cell radiosensitivity by inhibiting glutathione synthesis.52–54 The efficacy of radiotherapy in HCC is limited by tumor cell antioxidant capacity,55,56 but OA reduces GSH levels, weakening tumor cell antioxidant defenses and enhancing radiotherapy effectiveness. These findings suggest that OA may also be a valuable therapeutic option for radiotherapy-resistant HCC patients, a hypothesis that warrants further investigation.
fabp3 may be a predictive biomarker for SOR resistance, yet its expression levels in real patient samples require further validation. Previous studies have shown that fabp3 influences tumor cell metabolism and is closely associated with treatment response.57–59 Confirmation of fabp3 as a reliable biomarker for SOR resistance requires future studies to analyze fabp3 expression patterns in SOR-resistant patients using TCGA databases and other clinical samples, assessing its correlation with patient prognosis. If fabp3 levels are significantly upregulated in resistant patients, it could serve as a potential biomarker for predicting SOR treatment response in HCC and may help tailor personalized therapeutic strategies.
Although OA has demonstrated anticancer potential in various cancer cell lines,60,61 its application in humans remains challenging due to issues related to absorption, bioavailability, metabolism, dosing, and safety. Currently, there is no standardized recommended dose for OA, as most studies remain at the preclinical or early clinical trial stage.62 OA is primarily metabolized in the liver via the CYP450 system, with its metabolites excreted through urine.63,64 However, its metabolic pathways and clearance rates require further clinical validation. Additionally, OA exhibits lipophilicity, which enhances its bioavailability,62 but its oral absorption may be influenced by dietary intake, necessitating further research to optimize its administration.65 An animal study demonstrated that a single subcutaneous injection of 300 mg/kg OA did not cause significant toxicity,66 but data on human absorption and pharmacokinetics remain limited.
OA is generally considered safe, with no widespread reports of severe adverse effects,67 but long-term use may pose a risk of hepatotoxicity. Given that OA is predominantly metabolized in the liver, future research should focus on its potential impact on liver function (ALT/AST levels) and evaluate its long-term safety.68,69 Further studies should explore the safety of OA at different doses and administration routes and assess its potential synergy with SOR and other standard HCC treatments, ultimately advancing its clinical translation.
Despite providing new insights into the role of OA in reversing HCC resistance, this study has several limitations. First, the study is primarily based on in vitro cell models. Although the xenograft model provides some in vivo evidence, it does not fully replicate the real clinical environment. Future studies should incorporate in vivo animal models to validate the effects of OA in different microenvironments and assess its long-term impact on HCC treatment. Additionally, this study did not directly test the efficacy of OA+SOR combination therapy, although our results indicate that OA restores SOR sensitivity by suppressing fabp3 expression. The exact mechanisms underlying their combined therapeutic effect require further investigation through in vivo combination therapy studies or preclinical models to verify whether OA can enhance SOR efficacy in HCC.
Moreover, this study did not include a diverse range of HCC patient-derived cell lines. The current conclusions are primarily based on the Huh7 and HepG2 cell lines and their resistant sublines, which may not fully represent the biological characteristics of HCC patients. Future studies should incorporate a broader range of HCC cell lines from different pathological subtypes and resistance backgrounds to further validate the generalizability of OA’s effects on fabp3 inhibition. Furthermore, the role of fabp3 as a biomarker for SOR resistance still requires further validation, and future studies should leverage real patient datasets (such as TCGA) to assess fabp3 expression in HCC and its correlation with SOR resistance.
Future research should explore the clinical application of OA, optimizing its dosage, absorption, bioavailability, and long-term safety profile to facilitate its translation into practical treatment. To ensure its efficacy and safety, further investigations should evaluate OA’s impact on the HCC tumor microenvironment and metabolic pathways. Additionally, further validation of fabp3 as a predictive biomarker for SOR resistance is necessary to determine its role in guiding personalized treatment strategies. Ultimately, clinical trials will be crucial in confirming OA’s therapeutic potential in HCC, advancing its transition from experimental research to clinical application.