Lenvatinib (HY-10981), gefitinib (HY-50895), and copanlisib (HY-15346A) were purchased from MedChemExpress (Shanghai, China), and elacridar (S7772) was purchased from Selleck Chemicals (Shanghai, China). Stock solutions of 20 mM lenvatinib, 100 mM elacridar, and 20 mM gefitinib were dissolved in 100% dimethyl sulfoxide (DMSO), and the stock solution of 10 mM copanlisib was dissolved in Milli-Q water. Antibodies against total epidermal growth factor receptor (EGFR; A11577, ABclonal), phospho-EGFR (AP0820, ABclonal), total PI3K (ab32089, Abcam), phospho-PI3K (4228, CST), total AKT (9272, CST), phospho-AKT (4060, CST), total MEK1/2 (A4868, ABclonal), phospho-MEK1/2 (AP0209, ABclonal), total ERK1/2 (4695, CST), phospho-ERK1/2 (4376, CST), caspase-3 (T40051, Abmart), Bcl-2-associated X (Bax; T40044, Abmart), multidrug resistance protein 1 (MDR1; 13978, CST), breast cancer resistance protein (BCRP; 130244, Abcam), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH; 5174, CST) were used. Horseradish peroxidase (HRP)-conjugated goat anti-rabbit and goat anti-mouse antibodies were purchased from Beyotime Biotechnology (Shanghai, China). Alexa Fluor-conjugated goat anti-rabbit (647 nm) and goat anti-mouse (488 nm) antibodies were purchased from Invitrogen (Shanghai, China).

The Huh7 parental (Huh7 P) cell line was obtained from the Cell Bank of National Biomedicine Research (Beijing, China) and cultivated in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% foetal bovine serum and 1% antibiotics (penicillin and streptomycin) at 37°C and 5% CO₂. To generate the Huh7 LR cell line, Huh7 P cells were exposed to lenvatinib at an initial dose of 1 μM. Thereafter, the stable cell line was exposed to a lenvatinib concentration that was gradually increased by 1.0–2.0 μM per week. Approximately 10 months later, the Huh7 LR cell line was established and maintained in culture medium containing 20 μM lenvatinib.

Total RNA was extracted using TRIzol from a 10 cm cell culture plate when the cells reached 70–80% confluence. Three independent samples from each group (Huh7 P and Huh7 LR) were used for RNA-seq by Biomarker Technologies (Beijing, China). Log₂ (mRNA fold change) was used to assess differentially expressed mRNAs, with the calculated value of < -1 or > 1 deemed statistically significant (p < 0.001). The online bioinformatics database (DAVID Bioinformatics Resources 6.8, NIAID/NIH; website, https://david.ncifcrf.gov/tools.jsp) was used to analyse the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and biological processes based on the RNA-seq results.

Cells were plated at a density of 4,000 cells per well in a 96-well plate and cultivated overnight. The cells were then exposed to drugs suspended in DMEM (10% FBS) for 96 h. Thereafter, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution at a working concentration of 5 mg/mL was added to the culture medium. After 4 h of incubation, the upper medium was removed, and 100 μL of DMSO was added to dissolve the crystals formed in the lower medium. After 10 min of incubation and shaking, the absorbance was measured at a wavelength of 490 nm. A real-time cell analyser (RTCA) S16 (Celligence, China) was used to compare cell proliferation ability. During this process, 4,000 cells per well were seeded in a 16-well plate and cultivated in medium containing 20 μM lenvatinib. During the next 72 h, a detector connected to a computer constantly calculated and displayed relative cell proliferation by measuring the electrical resistance of the plate bottom.

To compare the clonogenicity of Huh7 P and Huh7 LR cell lines, 1,000 cells per well were seeded in a 6-well plate and continuously exposed to culture medium containing lenvatinib (20 μM) for two weeks. To assess the clonogenicity of the Huh7 LR cell line after different drug treatments, 2,000 cells were seeded per well in a 6-well plate, drug-containing media was removed from the cells after 72 h of exposure, and the medium without drugs was changed every 3 days for the next 11 days. After fixing in methyl alcohol for 15 min and staining with crystal violet for 20 min, the colonies were photographed using a camera and analysed using Image J Software.

Cells were seeded at a density of 2×10⁵ cells per well in a 6-well plate and cultivated overnight. The cells were then treated with the control or drug-containing media for 72 h. Subsequently, the cells were collected and stained with Annexin V-FITC and propidium iodide (Beyotime Biotechnology, China) for 20 min and then analysed using flow cytometry (Beckman Coulter, USA). At least 5×10⁴ cells were analysed for each sample.

Total RNA was extracted from Huh7 P and Huh7 LR cells using TRIzol reagent, and cDNA synthesis was conducted using a reverse transcription kit (Toyobo FSQ 301, Japan) following the manufacturer’s protocol. Subsequently, qRT-PCR was performed in a total volume of 20 μL, containing Milli-Q water (2 μL), c-DNA (6 μL), forward and reverse primers (2 μL), and q-PCR mix (10 μL; Toyobo QPS-201, Japan). The primers used in this study were manufactured by Ruibiotech (Beijing, China) with the following sequences: β-actin forward: 5′-ATCGTCCACCGCAAATGCTTCTA-3′ and reverse: 5′-AGCCATGCCAATCTCATCTTGTT-3′, MDR1 forward: 5′-GGGAGCTTAACACCCGACTTA-3′ and reverse: 5′-GCCAAAATCACAAGGGTTAGCTT-3′, and BCRP forward: 5′-GCCACAGAGATCATAGAGCCT-3′ and reverse: 5′-TCACCCCCGGAAAGTTGATG-3′. The results were normalised to β-actin expression and are presented as relative mRNA expression levels.

Cells (Huh7 P and Huh7 LR) were seeded at a density of 30,000 cells per well in an 8-well plate (BD Falcon 354108, USA) overnight. The cells were then washed thrice with PBS, fixed with 4% paraformaldehyde (PFA) for 20 min, blocked with 10% goat serum, and incubated with primary antibodies MDR1 (Rabbit mAb #13978 CST) and BCRP (Mouse mAb #130244 Abcam) for another day. After washing thrice with PBS, the cells were incubated with conjugated secondary antibodies for two hours at room temperature. Subsequently, 4′,6-diamidino-2-phenylindole was added and incubated for 15 min, and images were captured using a VS200 SlideView (Olympus, Japan).

After incubation with different drugs, the cells were collected and lysed using radioimmunoprecipitation assay buffer supplemented with a protease and phosphatase inhibitor cocktail (Beyotime Biotechnology, China). Equal amounts of protein from each sample were loaded on 8% or 10% sodium dodecyl-sulphate-polyacrylamide gel electrophoresis and transferred onto an Immobilon^®-P Transfer membrane (Merck Millipore Ltd.). After blocking with 5% non-fat milk, the labelled membrane was incubated with relevant primary antibodies at 4°C overnight. The membranes were then incubated with HRP-conjugated secondary antibodies for two hours at room temperature. Finally, the membranes were incubated with enhanced chemiluminescence reagent (Applygen, China), and the bands were detected. GAPDH was used as an internal reference.

BALB/C nude mice (male, 6 weeks old) were obtained from Charles River (Beijing, China). Huh7 LR cells (1.0×10⁷ cells per mouse) were injected into the flanks of the mice. After tumour establishment, the mice were randomly assigned to six groups (five mice per group): the vehicle, lenvatinib (5 mg/kg), gefitinib (80 mg/kg), elacridar (80 mg/kg), lenvatinib (5 mg/kg) combined with elacridar (80 mg/kg), and lenvatinib (5 mg/kg) combined with gefitinib (80 mg/kg) groups. The drugs were suspended in 5‰ carboxymethylcellulose sodium (powder dissolved in Milli-Q water). In the lenvatinib and elacridar group, elacridar was administered two hours prior to lenvatinib. All indicated treatments were orally administered to the mice 5 days per week. Tumour length and width were measured using callipers, and their volumes were calculated using the following formula: tumour volume = ½ length × width². All animal experiments were conducted in accordance with the approved protocol from Charles River (No. P2021049).

Harvested tumours were fixed in 4% PFA, dehydrated gradually, embedded in paraffin, and sliced into 4 μm thick sections. Some sections were subjected to haematoxylin-eosin (H&E) staining, whereas other sections were used for immunohistochemistry (IHC). After routine IHC procedures, the samples were incubated with primary antibodies against Ki67 (14-5698-80, Invitrogen) and proliferating cell nuclear antigen (PCNA; 13110, CST) at 4°C overnight. The samples were then incubated with secondary antibodies using the VECTASTAIN ^® Elite ^® ABC Universal Kit, Peroxidase (Horse Anti-Mouse/Rabbit IgG; PK-6200, Vector Laboratories, Inc., USA).

OriginPro 2021 software was used to perform data analysis. Data are presented as the mean ± standard deviation based on triplicate experiments, and the final results are representative of more than two independent experiments, excluding the xenograft tumour experiment. All p values are denoted as significant at p < 0.05. Following the Chou-Talalay method (11), CompuSyn software (ComboSyn, Inc., Paramus, NJ, USA) was used to calculate combination index (CI) values. The CI values reflect the interaction between two drugs (CI < 1, synergism; CI = 1, additive effect; CI > 1, antagonism).

After continuous exposure to lenvatinib (1–20 μM) for approximately 10 months, the Huh 7 LR cell line was normally passaged and maintained in medium containing 20 μM lenvatinib ( Figure 1A ). In contrast to Huh7 P cells, Huh7 LR cells were smaller in size and grew aggressively ( Figure 1B ). The MTT results revealed that Huh7 LR cells exhibited a higher proliferation rate than Huh7 P cells did in cultivation medium containing different lenvatinib concentrations (1.25, 2.5, 5, 10, and 20 μM), and the half maximal inhibitory concentration (IC₅₀) of lenvatinib in Huh7 LR cells (IC₅₀ > 20 μM) was significantly higher than that in Huh7 P cells (IC₅₀ 5.34 ± 1.07 μM) ( Figure 1C ). Meanwhile, Huh7 LR cells exhibited a relatively higher proliferation rate ( Figure 1D ) and higher colony forming ability ( Figure 1E ) than Huh7 P cells did in medium containing 20 μM lenvatinib. Moreover, Huh7 LR cells exhibited a higher anti-apoptotic activity than Huh7 P cells did in cultivation medium containing 20 μM lenvatinib ( Figure 1F ). These results confirmed that the Huh7 LR cell line was resistant to lenvatinib.

RNA-seq results were obtained to assess differentially expressed mRNAs between the Huh7 P and Huh7 LR cell lines ( Supplementary Table 1 ). Three independent samples were examined for each cell line ( Figure 2A ), and the Huh7 LR cell line exhibited 728 upregulated and 274 downregulated genes compared with those in the Huh7 P cell line ( Figure 2B ). KEGG pathway enrichment analysis revealed that pathways related to metabolism and ATP-binding cassette (ABC) transporters and the ERBB signalling pathway were enriched after LR induction ( Figure 2C ). Additionally, gene annotation analysis of biological processes demonstrated that cellular efflux and metabolic processes were increased after LR induction ( Figure 2D ).

The expression of MDR1 and BCRP, important ATP-binding cassette (ABC) transporters, was upregulated according to the RNA-seq results ( Supplementary Table 2 ). We used qRT-PCR, western blotting, and immunofluorescence staining to further verify MDR1 and BCRP expression levels. First, the qRT-PCR results revealed that Huh7 LR cells exhibited significantly higher MDR1 and BCRP mRNA levels than Huh7 P cells did ( Figure 3A ). Second, western blotting demonstrated that Huh7 LR cells exhibited significantly higher MDR1 and BCRP levels than Huh7 P cells did ( Figure 3B ). Third, immunofluorescence staining revealed that MDR1 and BCRP were located in the cell membrane, and their expression was significantly higher in Huh7 LR cells than in Huh7 P cells ( Figure 3C ).

The EGFR signalling pathway is an important branch of the ERBB signalling pathway, and the activation of EGFR signalling pathway is a hallmark of human malignancies (12–14). Supplementary Table 3 demonstrates that the transcriptional levels of EGFR and the downstream RAS/RAF/MEK/ERK and PI3K/AKT/mTOR pathways were normal or upregulated after LR induction. Importantly, among the upregulated mRNAs, PIK3R2, an oncogene involved in the physiological activation of PI3K (15), ranked first in fold-change (Log₂FC = 5.71) ( Supplementary Tables 1 , 2 , Figure 2B ). However, RNA-seq only reflects transcriptional level but cannot comprehensively reflect protein levels and functional alterations. Therefore, western blotting was performed to determine the levels of total and phosphorylated proteins involved in EGFR signalling and its downstream pathways. The western blot results revealed that phosphorylated EGFR, PI3K, AKT, MEK1/2, and ERK1/2 were significantly upregulated in Huh7 LR cells compared to those in Huh7 P cells ( Figure 3D ).

The in vivo antitumour effects of lenvatinib in combination with elacridar or gefitinib were assessed using xenografts derived from the Huh7 LR cell line. One week after tumour cell injection, the average xenograft size reached approximately 6 mm in diameter, and therapeutic treatment was initiated accordingly. Subsequently, the tumour volume and mouse body weight were measured once every two to three days. The mice were orally administered with the drugs for two weeks and then sacrificed. The harvested tumours were imaged and their corresponding weights were measured. Compared with that in the vehicle group, neither elacridar nor gefitinib inhibited tumour growth, and lenvatinib-based co-treatments significantly suppressed tumour growth ( Figures 6A–C ). Intriguingly, lenvatinib-based co-treatments exerted a much better antitumour effect than lenvatinib treatment alone did. Particularly, the co-administration of lenvatinib with elacridar exhibited the most potent antitumour efficacy ( Figures 6A–C ). During the treatment, no significant side effects were observed, as the mouse body weights were comparable in different groups ( Figure 6D ).

Regarding the histological analysis, lenvatinib alone and lenvatinib-based co-administrations significantly inhibited tumour angiogenesis, which could be easily determined by observing the general shape of the harvested tumours ( Figure 6B ). Upon further analysis using pathological H&E staining, remaining tumour micro-vessels were observed in the group treated with lenvatinib alone, but not in the group treated with co-administration of gefitinib or co-administration of elacridar ( Figure 7A ). Lenvatinib alone could inhibit cell proliferation; however, lenvatinib-based co-administration enhanced the inhibition of cell proliferation, which was assessed using IHC for Ki67 and PCNA ( Figures 7B, C ).