Carfilzomib was obtained from MedChemExpress (cat. no. HY-10455) and dissolved in DMSO. Antibodies were procured from Proteintech Group, Inc., including cyclin-dependent kinase (CDK)2 (cat. no. 10122-1-AP), CDK4 (cat. no. 11026-1-AP), cyclin A2 (cat. no. 66391-1-Ig), cyclin E1 (cat. no. 11554-1-AP), GAPDH (cat. no. 60004-1-Ig), phosphorylated (p)-JNK (Tyr185) recombinant antibody (cat. no. 80024-1-RR), JNK monoclonal antibody (cat. no. 66210-1-Ig), p38 MAPK polyclonal antibody (cat. no. 14064-1-AP), p-p38 MAPK (Thr180/Tyr182) polyclonal antibody (cat. no. 28796-1-AP). Additionally, anti-GADD45α antibody (cat. no. A1797) was purchased from ABclonal Biotech Co., Ltd. A total of 12 4-week-old female BALB/c nude mice (15–16 g) were purchased from Beijing Vital River Laboratory Animal Technology Co.

MHCC-97H and Huh7 cells were acquired from Wuhan Pricella Biotechnology Co., Ltd. and cultured in Modified Eagle's medium (Gibco; Thermo Fisher Scientific, Inc.) containing 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.), 1×10⁵ U/ml penicillin and 100 mg/ml streptomycin (Beijing Solarbio Science & Technology Co., Ltd.) in a constant temperature incubator at 37°C and 5% CO₂. Cells were treated with different concentrations of carfilzomib and control cells treated with the same concentration of DMSO.

The cytotoxic effects of carfilzomib on MHCC-97H and Huh7 cells were assessed using Cell Counting Kit-8 (CCK-8; Beyotime Institute of Biotechnology). Cells were cultured in 96-well plates (3×10⁴ cells/well) and subsequently treated with different concentrations of carfilzomib or DMSO (0, 10, 20, 30, 40, 50, 60, 80 and 100 µM) for 24 h at 37°C. Following treatment, 10 µl CCK-8 solution was added to each well and incubated for an additional 1 h at 37°C. Absorbance was measured at 450 nm using a Multiskan FC microplate reader (Thermo Fisher Scientific, Inc.) (16).

The evaluation of MHCC-97H and Huh7 cell proliferation was performed using an 5-ethynyl-2-deoxyuridine (EdU) staining kit (Beyotime Institute of Biotechnology) following the manufacturer's instructions (17). Briefly, 5,000 cells were cultured into a 96-well plate. The following day, the cells were treated with carfilzomib (30 µM for MHCC-97H cells and 60 µM for Huh7 cells) at 37°C for 24 h. For the EdU staining assay, each well was treated with 10 µM EdU working solution and was incubated for 2 h at 37°C. The nuclei were stained with DAPI for 10 min at room temperature and EdU-positive cells were visualized using a fluorescence microscope (Olympus Corporation). ImageJ software (version 1.51, National Institutes of Health) was used for sample analysis.

MHCC-97H and Huh7 cells were seeded in 6-well plates (1×10⁴ cells/well) and incubated for 24 h at 37°C. Subsequently, the cells were treated with carfilzomib (30 µM for MHCC-97H cells and 60 µM for Huh7 cells) at 37°C and cultured for 14 days. The culture medium containing carfilzomib was replaced every 3 days throughout this period. Colonies were washed with PBS, fixed with 4% paraformaldehyde at room temperature for 10 min and crystal violet (0.1%) was then used to stain the colonies at room temperature for 20 min (18). After washing, colony formation was observed using an inverted light microscope. Colonies consisting of ≥50 cells were counted. ImageJ software (version 1.51; National Institutes of Health) was used for data analysis.

Cell migration and invasion were assessed using Transwell chambers (pore size, 8 µm) in 24-well plates with (for invasion) or without (for migration) Matrigel (BD Biosciences). A total of 100 µl Matrigel (diluted 1:8 with serum-free DMEM medium) was added to the upper chambers of the Transwell inserts and incubated at 37°C for 2 h before use. MHCC-97H and Huh7 cells (1×10⁵ cells/well) were seeded in the upper chambers of the Transwell inserts with serum-free DMEM medium containing carfilzomib (30 µM for MHCC-97H cells and 60 µM for Huh7 cells). Medium containing 20% FBS was added into the lower chamber. After incubation for 24 h at 37°C, non-migrating cells on the top of the filter were eliminated using a cotton swab. Subsequently, the migrated cells on the bottom surface of the membrane were fixed and stained with 0.1% crystal violet at room temperature for 10 min (19). After washing, attached cells were counted under an inverted light microscope. The number of cells that had migrated was counted using ImageJ software (version 1.51; National Institutes of Health).

MHCC-97H and Huh7 cells were seeded in 6-well plates at a density of 1×10⁶ cells/well until the confluency reached 90%. Subsequently, the cells were scraped with the tip of a standard 200 µl sterile pipette to create a wound and further incubated in fresh serum-free DMEM medium containing carfilzomib (30 µM for MHCC-97H cells and 60 µM for Huh7 cells) for 48 h at 37°C. Cells were imaged with a light microscope at 0 and 48 h and the migratory capacity of the cells was assessed using the size of the healed wound area. ImageJ software (version 1.51l National Institutes of Health) was used for data analysis. Wound closure ratio=(0 h cell migration area-48 h cell migration area)/0 h cell migration area.

MHCC-97H and Huh7 cells were seeded in 6-well plates at a density of 5×10⁵ cells/well and treated with carfilzomib (30 µM for MHCC-97H cells and 60 µM for Huh7 cells) at 37°C for 24 h after 12 h. After 24 h of treatment, the cells were collected and fixed in 70% ethanol at 4°C overnight. Subsequently, the collected cells were washed with PBS, resuspended in staining buffer containing 50 µl PI and 450 µl RNase A (Beijing Solarbio Science & Technology Co., Ltd.) and incubated at room temperature for 30 min in the dark. The cells were analyzed and calculated using BD FACSCanto^™ Flow Cytometer (BD Biosciences) and MoD FIT software (version 3.0; BD Biosciences) (20).

Protein samples were extracted from MHCC-97H and Huh7 cells using RIPA buffer supplemented with a mixture of 1% PMSF and 1% phosphatase inhibitors (Beijing Solarbio Science & Technology Co., Ltd.) and the concentration was measured using a BCA Protein Assay kit (Beijing Solarbio Science & Technology Co., Ltd.). Subsequently. 30 µg of total protein was separated by 10% SDS-PAGE and transferred onto a PVDF membrane (Merck KGaA). After blocking with 5% non-fat milk at room temperature for 1 h, the membrane was incubated with primary antibodies at 4°C overnight, including anti-CDK2 (1:5,000), anti-CDK4 (1:2,000), anti-cyclin A2 (1:5,000), anti-cyclin E1 (1:1,000), anti-p-JNK (1:1,000), anti-JNK (1:1,000), anti-p-p38 (1:1,000), anti-p38 (1:1,000), anti-GADD45α (1:1,000) and anti-GAPDH (1:10,000), followed by incubation with horseradish peroxidase (HRP)-conjugated Goat Anti-Rabbit IgG (1:10,000; cat. no. SA00001-2, Proteintech Group, Inc.) or HRP-conjugated Goat Anti-Mouse IgG (1:10,000; cat. no. SA00001-1, Proteintech Group, Inc.) secondary antibodies at room temperature for 1 h. The signal was developed with an enhanced chemiluminescence reagent (New Cell & Molecular Biotechnology Co., Ltd.), and images were captured using an Amersham ImageQuant 800 imaging system (Cytiva). The quantification of protein abundance was performed using ImageJ software (version 1.51; National Institutes of Health), with GAPDH as an endogenous control.

GADD45α knockdown plasmids containing pPLK GFP+Puro plasmid backbone [short hairpin (sh)-negative control (NC) and shGADD45α, pPLK GFP+Puro], were provided by Beijing Tsingke Biotech Co., Ltd. The 2nd generation system was used for plasmid packaging. A total of 1.5 µg knockdown plasmids, 1 µg pCMV–VSV-G and 0.5 µg pCAG-dR8.9 (packaging plasmid provided by Beyotime Institute of Biotechnology) were transfected into 293T cells in 6-well plates using Lipofectamine 3000^® transfection reagent (Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions at 37°C. Lentiviral particles were harvested at 48 h after transfection, followed by infection of HCC cells for 24 h and screening of the cells for 7 days with medium containing 2 µg/ml puromycin (Beyotime Institute of Biotechnology). The target sequences of the knockdown lentiviral plasmids were as follows: shNC, 5′-TGTCGCGGTAAGTGCCTCATA-3′; GADD45α sh1, 5′-GCTGGAGAGCAGAAGACCGAA-3′; GADD45α sh2, 5′-GAAGACCGAAAGGATGGATAA-3′; and GADD45α sh3, 5′-CAATGGGTTCCAGTGATTAAT-3′.

A total of 12 four-week-old female BALB/c nude mice (weight, 15–16 g) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. and housed in the animal facility under standard conditions with a controlled temperature of 25±2°C, a humidity of 55±5%, a light/dark photoperiod of 12 h and free access to water and food. The study was conducted in compliance with the ‘Animal Research: Reporting of In Vivo Experiments’ guidelines. Animal experiments were approved by the Animal Care and Use Committee of The First Affiliated Hospital of Zhengzhou University (Zhengzhou, China; approval no. 2023-KY-1008-002). Huh7 cells were collected in the logarithmic growth phase and resuspended in PBS at a density of 5×10⁷ cells/ml. Each mouse received a subcutaneous injection of 100 µl (5×10⁶) cell suspension into the right flank. The mice were then randomly divided into two groups and treated with DMSO or 4 mg/kg carfilzomib. Carfilzomib (dissolved in DMSO and diluted in PBS) was administered intraperitoneally twice a week. Tumor volume and mouse weight were measured every 3–4 days. Tumor volume was calculated using the following formula: Tumor volume=(width² × length)/2. After 4 weeks of treatment, mice were euthanized by cervical dislocation. Subsequently, the subcutaneous tumors were excised and processed for further experiments (21).

VeZol reagent (cat. no. R411-01; Vazyme Biotech Co., Ltd.) was used to extract RNA from MHCC-97H and Huh7 cells. A NanoDrop 2000 spectrophotometer was used to measure the RNA concentration, and cDNA was synthesized from RT of 1 µg total RNA according to the manufacturer's protocol of the RT kit (cat. no. R233-01; Vazyme Biotech Co., Ltd.). A qPCR machine was used to amplify the cDNA using the SYBR Green qPCR Mix (cat. no. Q711-02; Vazyme Biotech Co., Ltd.). The thermocycling conditions for amplification were as follows: 95°C for 5 min, followed by 40 cycles at 95°C for 10 sec, 60°C for 30 sec and a final step at 95°C for 15 sec, 60°C for 60 sec and 95°C for 15 sec. GAPDH was used as an internal control. Subsequently, the 2^−ΔΔCq method was used to analyze the gene expression data (22). Primer sequences were as follows: GADD45α forward (F), 5′-CTGGAGGAAGTGCTCAGCAAAG-3′ and reverse (R), 5′-AGAGCCACATCTCTGTCGTCGT-3′; GADD45β F, 5′-GCCAGGATCGCCTCACAGTGG-3′ and R, 5′-GGATTTGCAGGGCGATGTCATC-3′; GADD45γ F, 5′-CGTCTACGAGTCAGCCAAAGTC-3′ and R, 5′-CGATGTCGTTCTCGCAGCAGAA-3′; and GAPDH F, 5′-GTCTCCTCTGACTTCAACAGCG-3′ and R, 5′-ACCACCCTGTTGCTGTAGCCAA-3′.

Total RNA from Huh7 cells was extracted using VeZol reagent (cat. no. R411-01; Vazyme Biotech Co., Ltd.). Furthermore, a NanoDrop 2000 spectrophotometer was used to determine the purity and quantity of RNA. The RNA 6000 Nano Kit (cat. no. 5067-1511; Agilent Technologies, Inc.) and an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.) was used to determine RNA integrity. For each sample, 1 µg of the total RNA was used for RNA-seq library preparation using the VAHTS Universal V6 RNA-seq Library Prep Kit (cat. no. NR604-01; Vazyme Biotech Co., Ltd.) per the manufacturer's instructions. Then mRNA was enriched and purified using VAHTS mRNA Capture Beads (cat. no. N401; Vazyme Biotech Co., Ltd.). The purified mRNA was cut into short fragments using fragmentation buffer and reverse transcribed into cDNA using random primers. Second-strand cDNA was synthesized using DNA polymerase I, RNase H, dNTP and buffer oncluded in the VAHTS Universal V6 RNA-seq Library Prep Kit. After performing end repair and adding poly (A), the cDNA fragments were ligated using VAHTS RNA Adapters (cat. no. N803/N804; Vazyme Biotech Co., Ltd.). The ligation products were enriched via PCR amplification to construct the cDNA library template (NovaSeq 6000 S4 Reagent Kit v2.5; 300 cycles; cat. no. 20028312). The quality of these cDNA libraries was evaluated with the Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.). An Illumina Novaseq 6000 (Illumina Inc.) was used for 150 bp paired-end sequencing. The loading concentration was 4 nM, and the concentration was quantified using the Qubit 3.0 (Thermo Fisher Scientific, Inc.). The sequencing was performed by Shanghai OE Biotech Co., Ltd. DESeq (2012) package (version 1.46.0) was used to identify transcriptome differentially expressed genes (23,24). The Hierarchical clustering analysis, volcano map and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were produced using R (https://www.r-project.org/) on the OECloud platform (25,26). To complete these analyses, the plate corresponding to the class volume was selected on the OECloud platform for hierarchical clustering analysis, volcano map and KEGG analyses. Subsequently, the sequenced gene expression data, included in the raw data from the Gene Expression Omnibus database (accession no. GSE284346; http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE284346) was used for further analysis.

Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database (version 11.5; http://version-11-5.string-db.org/) is an online search tool for the analysis of PPIs and functional protein networks, which contains confirmed and predicted direct and indirect PPI biological data. In the present study, GADD45α, P38 and JNK were used as query conditions, and the minimum required interaction score was set to >0.4, which is a medium level of confidence.

The xenograft tumor tissue was fixed in 10% formalin for 48 h at room temperature, paraffin-embedded and cut into 5 µm-thick paraffin sections. Tissue sections were incubated at 60°C for 2 h, deparaffinized through xylene, dehydrated in ethanol gradients and subjected to thermal antigen retrieval as samples were heated to boiling in a microwave oven for 10 min with 0.01 mol/l citrate buffer (pH 6.0). The sections were blocked in 10% goat serum (Wuhan Servicebio Technology Co., Ltd.) for 30 min at room temperature and incubated overnight at 4°C with proliferating cell nuclear antigen (PCNA) primary antibodies (1:200; cat. no. GB11010; Wuhan Servicebio Technology Co., Ltd.). Subsequently, the sections were incubated with CoraLite594-conjugated Goat Anti-Rabbit IgG fluorescent secondary antibodies (cat. no. SA00013-4; 1:500; Proteintech Group, Inc.) for 1 h at room temperature, counterstained with DAPI for 10 min at room temperature and examined using a fluorescence microscope (27).

The xenograft tumor tissue was fixed in 10% formalin for 48 h at room temperature, paraffin-embedded and cut into 5 µm-thick paraffin sections. Tissue sections were incubated at 60°C for 2 h, deparaffinized using xylene for 5 min at room temperature and dehydrated in 5 and 95% ethanol gradient 5 min at room temperature. Then, the sections were soaked in hematoxylin solution for 5 min, rinsed in tap water for 1 min, incubated with 1% hydrochloric acid alcohol solution for 3 sec, rinsed in tap water for 1 min and hematoxylin reblue solution was added for 10 sec. Next, the sections were soaked in eosin solution for 5 min. The sections were dehydrated using 75 and 95% gradient ethanol for 5 min at room temperature, then incubated with xylene for 5 min at room temperature and sealed with neutral gum. Observation and analysis were performed under an inverted light microscope. H&E staining reagent was provided by Wuhan Servicebio Technology Co., Ltd. and all incubations were performed room temperature, unless otherwise stated.

The Kaplan-Meier (KM) plotter (http://kmplot.com/analysis/) can perform survival analysis on >54,000 genes, including mRNA, microRNA and proteins, across 21 tumor types, including liver cancer (28,29). This database consolidates information from the Gene Expression Omnibus (GEO), the Cancer Genome Atlas (TCGA) and European Genome-phenome Archive (EGA), providing gene expression profiles and prognostic details for a wide range of tumors (28,29). The datasets used for liver cancer survival analysis in Kaplan Meier plotter are available at TCGA (https://cancergenome.nih.gov/) and GEO databases (dataset GSE9843 is available in the https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE9843; dataset GSE20017 is available at http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE20017) (28). In the present study, the KM plotter was used to evaluate the prognostic value of GADD45α in patients with HCC. GADD45α was entered into the database and the number-at-risk was displayed under the main panel of the OS. Based on the Kaplan-Meier plotter Online Tool, 322 patients with HCC (follow-up of 60 months) were included in the online database and were divided into two groups (high vs. low expression levels) according to the median expression levels of GADD45α. All possible cut-off values were auto selected. Statistical parameters such as survival plot, hazard ratio and log-rank P-values were obtained from the KM plotter Online Tool.

Statistical analysis was performed using SPSS 16.0 (SPSS, Inc.). Data are presented as the mean ± SD. An unpaired Student's t-test was used to compare two groups and one-way ANOVA followed by Tukey's post hoc test was used to compare multiple groups. Kaplan-Meier analysis followed by the log-rank test was used to evaluate the prognosis of patients. P<0.05 was considered to indicate a statistically significant difference.

To assess the effect of carfilzomib on HCC cells, MHCC-97H and Huh7 cells were treated with several concentrations of carfilzomib. As shown in Fig. 1A and B, carfilzomib decreased the viability of MHCC-97H and Huh7 cells in a dose-dependent manner. Based on the results of the CCK-8 assay, the activity of MHCC-97H and Huh7 cells decreased by ~50% at carfilzomib concentrations of 30 and 60 µM, carfilzomib concentrations of 30 and 60 µM were used for subsequent experiments on MHCC-97H and Huh7 cells, respectively. These observations collectively suggested that carfilzomib inhibits the viability of HCC cells.

The progression of HCC is characterized by the increased motility and invasiveness of cancer cells, which contribute to the high incidence and mortality rate of patients with HCC (30). Therefore, a series of experiments were performed to investigate the effect of carfilzomib on the migration and invasion of HCC cells. The results of the wound-healing assay showed that carfilzomib inhibited the migration of MHCC-97H and Huh7 cells compared with that of the control group (Fig. 2A and B). Transwell assays with (for invasion) or without (for migration) Matrigel were used to corroborate the role of carfilzomib in the migration and invasion of HCC cells. As shown in Fig. 2C and D, carfilzomib significantly suppressed both invasion and migration in MHCC-97H and Huh7 cells compared with those of the control groups. These findings indicated that carfilzomib exhibits strong anti-migratory and anti-invasive properties against HCC cells.

Uncontrolled cell proliferation caused by cell cycle dysregulation is a fundamental feature of cancer cells (31,32). To investigate the anti-proliferative effect of carfilzomib, MHCC-97H and Huh7 cells were treated with carfilzomib, and EdU-staining and colony formation assays were performed. The EdU proliferation assay demonstrated that carfilzomib significantly decreased the number of EdU-positive MHCC-97H and Huh7 cells (Fig. 3A). Consistent with these findings, the colony formation experiments showed that carfilzomib inhibited colony formation in HCC cells compared with that in the control group (Fig. 3B).

To investigate the antiproliferative mechanism of carfilzomib, cell cycle analysis was performed. The results revealed that carfilzomib increased the percentage of cells in the G₀/G₁ phase compared with that in the control group, indicating that carfilzomib caused cell cycle arrest at the G₀/G₁ phase, which reduced entry into the S and G2 phases, particularly in Huh7 cells (Fig. 3C and D). Additionally, the expression of cycle-associated proteins was examined by western blotting. The results demonstrated that carfilzomib significantly decreased the expression of cyclin A2, cyclin E1, CDK2 and CDK4 (Fig. 3E and F) compared with that in the control group in both cell lines. These results indicated that carfilzomib suppressed cell proliferation by inducing G₀/G₁ phase arrest in HCC cells.

To assess the antitumor effect of carfilzomib in vivo, a Huh7 ×enograft tumor model using nude mice was established. In the present study, nude mice were randomly divided into two groups (n=6 per group), which were treated with vehicle control or carfilzomib for 4 weeks. Tumor volume was measured to evaluate the antitumor efficacy of carfilzomib. Carfilzomib significantly reduced tumor growth, volume and weight compared with that in the control group after 17, 21, 24 and 28 days of treatment (Fig. 4A-C). Notably, there was no significant difference in the body weight of carfilzomib-treated mice compared with the control mice (Fig. 4D). Furthermore, analysis of the xenograft tumor tissues using H&E staining, and PCNA immunofluorescence staining showed that the positive rate of PCNA in the carfilzomib group was significantly lower compared with that in the control group (Fig. 4E and F). These findings indicated that carfilzomib significantly inhibited tumor cell proliferation in vivo.

To investigate the mechanism of carfilzomib action in HCC cells, transcriptome sequencing was performed in Huh7 cells treated with or without carfilzomib (Fig. 5A and B). The results showed that GADD45α exhibited the highest expression among the most significant top five differentially expressed cell cycle-related genes (Fig. 5C). Carfilzomib also exhibited a significant effect on the expression of GADD45β and GADD45γ (Fig. S1); however, carfilzomib exhibited the most marked effect on GADD45α expression; therefore, in the present study the effect of carfilzomib on GADD45α was investigated. Western blot analysis demonstrated that carfilzomib upregulated the protein expression level of GADD45α (Fig. 5D and E). Simultaneously, analysis of data from The Cancer Genome Atlas showed that higher expression of GADD45α was associated with improved survival of patients with HCC compared with low GADD45α expression (Fig. 5F). These findings indicated that GADD45α may mediate the effects of carfilzomib in HCC cells.

To investigate the role of GADD45α in the carfilzomib-mediated antitumor effect on HCC, stable GADD45α knockdown cell lines were constructed using lentiviral shRNA targeting GADD45α (Fig. S2). Based on the expression of the GADD45α protein, sh2 was the most effective shRNA tested and was used for the subsequent experiments. The results demonstrated that knockdown of GADD45α counteracted the cell cycle inhibitory effects of carfilzomib on the G₀/G₁ phase (Fig. 6A and B). Furthermore, the EdU proliferation assay showed that carfilzomib decreased the number of positive cells, while silencing GADD45α abolished the anti-proliferative effect of carfilzomib (Fig. 6C and D). Moreover, the expression of proteins involved in cell cycle regulation was examined using western blotting, including cyclin A2, cyclin E1, CDK2 and CDK4. As shown in Fig. 6E and F, silencing GADD45α increased the expression levels of cyclin A2, cyclin E1, CDK2 and CDK4 compared with those in the carfilzomib-treated group. Collectively, these findings indicated that GADD45α serves a crucial role in carfilzomib-mediated antitumor therapy.

The MAPK signaling pathway serves a crucial role in cell cycle arrest (33). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis indicated significant enrichment of the ‘MAPK signaling pathway’ with carfilzomib treatment in Huh7 cells (Fig. 7A). STRING network analysis revealed a strong association between GADD45α and proteins involved in the MAPK signaling pathway (Fig. 7B). To assess the role of GADD45α in the MAPK signaling pathway, the expression levels of JNK, p-JNK, p38 and p-p38 were examined using western blotting. As shown in Fig. 7C and D, carfilzomib upregulated the phosphorylation levels of JNK and p38, suggesting that carfilzomib activated the MAPK signaling pathway. Moreover, silencing GADD45α significantly decreased the levels of p-JNK and p-p38 (Fig. 7C and D). These findings indicated that carfilzomib may induce cell cycle arrest in HCC cells via GADD45α-mediated MAPK activation.