AKR1C1 overexpression leads to lenvatinib resistance in hepatocellular carcinoma

Introduction

Hepatocellular carcinoma (HCC) has a high mortality rate and has been more prevalent worldwide over the past several decades. The cancer fatality and incidence rates are both considerably greater among the Chinese population compared to those in developed nations. Surgical methods are the most common treatment for HCC, although sorafenib, lenvatinib, etc. are also frequently utilized (1). Lenvatinib, a multi-target tyrosine kinase inhibitor (TKI), has shown promising results as a new first-line option in the treatment of HCC; its increased application has provided an alternative to the use of sorafenib as the sole first-line TKI treatment for HCC, which has been the case for years. Preclinical studies have demonstrated that lenvatinib has anticancer proliferation and immunomodulatory effects (2). However, only 42.9% and 25% of HCC patients with Child-Pugh A and B respond to lenvatinib (3). Similarly to sorafenib resistance, lenvatinib therapy resistance is also becoming increasingly widespread (4,5). Additionally, the immune microenvironment has an important role in the progression and development of HCC (6), but the correlation between lenvatinib-related targets and immune checkpoints remains unclear. Given the limitations of the currently available clinicopathological biomarkers, a study suggests that molecular biomarker should be used in addition to clinicopathological marker (7). There is a limited number of molecular markers that can be used to guide lenvatinib therapy and spot early lenvatinib resistance. Finding new therapeutic targets and prognostic biomarkers will facilitate more targeted approaches for the prevention and treatment of HCC (8).

Aldo-keto reductases comprise AKR1C1-AKR1C4 are four enzymes that catalyze nicotinamide adenine dinucleotide phosphate hydride (NADPH)-dependent reductions and have been implicated in biosynthesis, intermediary metabolism, and detoxification. Recent research has confirmed a significant relationship between these enzymes’ expression levels and both malignant transformation and cancer therapy resistance (9). In the present study, we found that AKR1C1 was significantly enhanced in HCC cells resistant to lenvatinib and was correlated with poor patient prognosis. According to the results of this study, AKR1C1 contributes to the development of lenvatinib resistance in oncogenic HCC. Consequently, it may be inferred that AKR1C1 could be used as a molecular biomarker for assessing lenvatinib sensitivity and as a potential therapeutic target for individuals with HCC. We present this article in accordance with the ARRIVE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-23-277/rc).

Methods

Cell culture

HCC cell lines (MHCC-97H, LM3, HepG2, and Hep3B) and human normal liver cells (QSG-7701) were respectively cultured in dulbecco’s minimum essential medium (DMEM) or Roswell Park Memorial Institute (RPMI)-1640 medium (Thermo Fisher Scientific, Waltham, MA, USA) containing 1% streptomycin/penicillin and 10% FBS (Fetal Bovine Serum) (Thermo Fisher Scientific). The cell cultures were grown at 37 ℃ with a continuous supply of 5% carbon dioxide (CO₂). The HCC cell lines were provided by the Cell Bank at UCAS (Chinese Academy of Sciences, China). The QSG-7701 was obtained from Beyotime (Shanghai, China). Lenvatinib (Beyotime, SF5346-10 mM) concentration was progressively increased weekly up to 10 µM. After 30 weeks of treatment, HepG2 lenvatinib-resistant (LR) cell lines (with mixed cell populations, not from a single clone) were obtained.

Colony formation assay

Treatments were applied to cells that had been seeded at a density of 1,000 per well in a six-well plate. After 2 weeks, the medium was discarded, followed by washing twice with phosphate buffered saline (PBS) to remove any remaining debris. After a 30-min methanol fixation, the clones were stained with crystal violet. After being cleaned and dried, the clones were photographed.

RNA-seq and bioinformatics analysis

The extraction of total RNA was performed using a TRIzol (Invitrogen, USA)-based technique according to the manufacturer’s guidelines. RNA-seq was provided by Sangon Biotech (Shanghai) Co., Ltd. According to the manufacturer’s provided guidelines, the RNA concentration was measured using the Qubit2.0 (Thermo Fisher Scientific, USA), and the integrity of the RNA and the presence of genomic contamination were both examined using an agarose gel. Sequencing was performed on an Illumina HiSeq 2500 (Illumina, USA) with paired-end 150-bp read lengths. Furthermore, mRNA clean reads were mapped using TopHat2 (Johns Hopkins University, Johns Hopkins University, USA) software to the UCSC (University of California, Santa Cruz) hg38 primary assembly genome. Transcript counts were calculated using HTSeq (High-throughput sequence analysis in Python). Read normalization, size factor estimation, and differential expression analysis were performed using the Bioconductor package DESeq2 (version 3.15) (10). The datasets used in this study were uploaded to the Gene Expression Omnibus (GEO) database (GSE214324).

Analysis of differentially expressed genes (DEGs)

The Limma package (Version 3.42.2) of R programming language was used to analyze the DEGs and visualize them using a heatmap and volcano plot with an adjusted P value of less than 0.05 in each cohort (11). Furthermore, GENEMANIA software (http://genemania.org/) was employed to generate a protein-protein interaction (PPI) network of the DEGs. The degree rank of hub genes was determined using Cytoscape (National Institute of General Medical Sciences, USA) (version 3.6.1) software, and these genes were considered for further analysis (12).

Analysis of immune infiltration

CIBERSORT is an online web portal (http://cibersortx.stanford.edu/) that uses gene expression data to estimate the abundances of individual cell types in a heterogeneous cell population (13). HCC patient data were downloaded from The Cancer Genome Atlas (TCGA) database. Furthermore, the immune response of 22 tumor-infiltrating immune cells (TIICs) was assessed using CIBERSORT to evaluate the links between different groups in HCC and to determine whether there were any correlations among the TIICs.

Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis

KEGG analysis was performed using R-clusterProfiler (version 3.14.3) and the enrichplot package to identify probable signaling pathways associated with the DEGs (14). The HCC patient data used for analysis was downloaded from TCGA database. Pathways with adjusted P value of less than 0.05 were considered significant, and the complete details of significant pathways are displayed in Tables S1,S2.

qRT-PCR analysis

The extraction of total RNA was performed using a TRIzol (Invitrogen)-based technique. To obtain cDNA (complementary DNA), the extracted RNA was reverse transcribed using a RevertAid First Strand cDNA Synthesis Kit (TaKaRa, Japan). According to the established method, RT-qPCR was performed on the LightCycler 480 II (Roche, USA) using the SYBR Master Mixture (TaKaRa). β-actin and U6 were used as the internal control for AKR1C1 or miR-4644. The 2^−∆∆Ct method was applied for the analysis of the obtained data. The primers used in this study are listed in Table 1.

Transduction of AKR1C1 small interfering RNAs (siRNAs)

Two AKR1C1 siRNAs, i.e., siRNA-AKR1C1#1 (CTACCTTATTCATTTTCCA) and siRNA-AKR1C1#2 (CCGTGGAGAAGTGTAAAGA), were designed and then synthesized. AKR1C1 siRNA transduction was carried out using riboFECTTM CP (RiboBio, Guangzhou, China) according to the manufacturer’s provided guidelines.

Cell proliferation assay

A cell counting kit-8 (CCK-8; Sigma-Aldrich, St. Louis, MO, USA) was employed to evaluate cellular proliferation. For each condition, the results are represented as the mean ± standard error of three different experiments. When the cells in six-well plates reached 80% confluency, three scratch wounds were induced in each well using 100 µL pipette tips. After 48 h, the wound had closed, and photos were taken at 0 and 48 h. For the EdU test, the number of cells in each group was counted and observed under a fluorescence microscope (Olympus Corporation, Japan) following the application of the Cell-LightTM EdU Apollo567 kit (Ribo) in-vitro according to the manufacturer’s provided guidelines.

Drug sensitivity analysis

The anticancer drug dataset was downloaded from the Genomics of Drug Sensitivity in Cancer (GDSC) website, and we then used the oncoPredict package to determine whether there was a link between the IC₅₀ (half maximal inhibitory concentration) values of various drugs reported with anticancer activity and the different risk groups (15). The screening of drugs was performed with a standard mean IC₅₀ <1 for all liver cancer samples, which were considered to be highly effective drugs for liver cancer treatment. Furthermore, statistical analyses of drug sensitivity for these drugs in the high- and low-risk subgroups were performed to determine the various levels of response to drugs in patients in different risk groups.

Human sample and immunohistochemistry

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Informed consent was obtained from the patients, and this research was approved by the Ethics Committee of Affiliated Hospital of Nantong University (No. 2018-L006). The data and samples were collected from patients at the Affiliated Hospital of Nantong University. Diluted AKR1C1 monoclonal antibodies (1:400, SANTA CRUZ BIOTECHNOLOGY, sc-166297) were dropped on a tissue chip, which contained 90 pairs of HCC tissues and tumor-adjacent liver tissue. The chip was then placed in a moisture chamber at 4 ℃ overnight. Secondary antibodies (Thermo Fisher Scientific) at a 1:1,000 dilution were added to the tissue chip and incubated at room temperature for 30 min. Three representative images were achieved from each sample, and we invited a senior pathologist for further analysis.

Tumor xenografts

Females BALB/c nude mice (5 weeks old, weight: 18–21 g, Cyagen Biosciences lnc.) were used for the xenograft models. All of the mice were housed in SPF (Specific pathogen Free) conditions. For the in vivo siRNA treatment experiment, LEN-HeG2 cells (5×10⁵ in 100 µL sterile PBS) were subcutaneously injected in the right flanks of each mouse. Next, the mice were randomized into two groups (n=4 in each group) when the tumor size reached approximately 100 mm³, in vivo siRNA was delivered by intratumor injection twice a week at a dose of 5 nmol in each mouse. After 31 days, the mice were sacrificed and the tumors were obtained for further study. Animal experiments were performed under a project license (No. S20220228-004) granted by the Institute Ethics Committee at the Affiliated Hospital of Nantong University, in compliance with the guidance of the care and use of laboratory animals issued by the Ministry of Science and Technology of the China. A protocol was prepared before the study without registration.

Statistical analysis

R program (version 4.1.2) was used for all data visualization and statistical analysis. The mean and standard deviation (SD) from at least three different experiments were used to represent all of the data. One-way Analysis of Variance (ANOVA) or the student’s t-test was utilized to assess the differences between several groups. Statistical significance was defined as P<0.05.

Results

Lenvatinib inhibits the proliferation of HCC cells

To investigate the role of lenvatinib in HCC cells, a colony formation assay was used to determine the influence of lenvatinib on the proliferation of liver cancer cells. Different concentrations of lenvatinib were co-cultured with liver cancer cells for 2 weeks. At a low dose of 20 µM, lenvatinib dramatically decreased the proliferation of LM3 cells (Figure 1A) but only suppressed cell proliferation in MHCC-97H cells at doses higher than 40 µM (Figure 1B). Analysis of the growth curves of CCK-8 cells showed that lenvatinib suppressed the growth and proliferation of the LM3 and MHCC-97H HCC cell lines (Figure 1C,1D).

Figure 1 Inhibitory effect of lenvatinib on HCC cell proliferation. Different lenvatinib concentrations, i.e., 0, 20, 40, 60, 80, and 100 µM, and stain with crystal violet. Results display lenvatinib inhibited the proliferation of HCC cells. (A) Lenvatinib decreases the proliferation of LM3 cells at a low dose of 20 µM. (B) Lenvatinib decreases the proliferation of MHCC-97H cells at 40 µM. After methanol fixation, clones were stained with crystal violet. According to the CCK-8 assay, the proliferative rates of LM3 (C) and MHCC-97H (D) cells exposed to various concentrations of lenvatinib were evaluated. *P<0.05; **P<0.01. OD, optical density; HCC, hepatocellular carcinoma; DMSO, dimethyl sulfoxide.

Identification and validation of targets for lenvatinib in HCC

To determine the drug target of lenvatinib, RNA-seq was employed to evaluate the transcriptome differences between lenvatinib-treated MHCC-97H cells and their parental counterparts. MHCC-97H cells treated with 40, 60, 80, and 100 µM lenvatinib were used for sequencing. The results showed that different significant genes affect the molecular function (MF, Figure 2A), cellular component (CC, Figure 2B), biological process (BP, Figure 2C), and KEGG (Figure 2D) pathways. Moreover, compared to their parental counterparts, LR MHCC-97H cells (LEN-MHCC-97H) had 42 up-regulated and 38 down-regulated DEGs (Figure 3A, Table S3). The PPI network of DEGs demonstrated that AKR1C1 was the most prominent up-regulated gene in LEN-MHCC-97H cells, and serpin family E member 1 (SERPINE1) was a considerably down-regulated gene in LEN-MHCC-97H cells (Figure 3B,3C, Tables 2,3). SERPINE1 promotes the proliferation, invasion, migration, and epithelial-mesenchymal transition (EMT) of HCC cells (16,17). However, the function of AKR1C1 in HCC is not clear.

Figure 2 Identification and analysis of GO and KEGG terms of differentially expressed genes following RNA-seq. Bubble diagram of the top 10 activated or suppressed terms in biological process (A), cellular component (B), molecular function (C), and Kyoto Encyclopedia of Genes and Genomes enrichment (D). GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Figure 3 AKR1C1 and SERPINE1 were the key differentially expressed genes of lenvatinib targets. (A) RNA-seq was used to examine the transcriptome alterations in MHCC-97H cells treated with varying doses of lenvatinib. Genes with high or low expression in the four treatment groups were simultaneously shown by a volcano plot. Blue represents downregulated genes, red represents upregulated genes, and grey represents no difference genes. (B) The PPI network of the hub genes from upregulated genes. (C) PPI network of the hub genes from downregulated genes. (D,E) High expressions of AKR1C1 and SERPINE1 predicted poor prognosis in HCC patients. (F) The KEGG analysis revealed that AKR1C1 is involved in the Wnt signaling pathway, calcium signaling pathway, gastric cancer, etc. (G) The KEGG analysis also showed that AKR1C1 is involved in the PI3K/AKT signaling pathway, the Wnt signaling pathway, the IL-17 signaling pathway, and other pathways. (H) According to the expression level of AKR1C1, the data were divided into two groups. The ratios of activated dendritic cells, macrophages (M0), activated memory CD4 T cells, and eosinophils were characterized in two groups. (I) Similarly, for SERPINE1, the ratios of the proportions of activated NK cells, neutrophils, macrophages (M0), resting memory CD4 T cell, resting mast cells, and activated mast cells showed marked changes. (J) AKR1C1 and SERPINE1 are significantly correlated with macrophages (M0) (red indicates that AKR1C1 and SERPINE1 expression are positively correlated with immune cells. Blue indicates that AKR1C1 and SERPINE1 expression are negatively correlated with immune cells. The larger the difference in the ratios of the proportions of immune cells, the larger the circle size). *P<0.05, **P<0.01, ***P<0.001. PPI, protein-protein interaction; KEGG, Kyoto Encyclopedia of Genes and Genomes; LIHC, liver cancer; HCC, hepatocellular carcinoma; NC, negative control; NS, no sense.

The UALCAN (The University of ALabama at Birmingham CANcer data analysis Portal, http://ualcan.path.uab.edu) database was used to evaluate the link between AKR1C1 and SERPINE1. Furthermore, their expressions with clinical features were also assessed (Figure 3D,3E). In HCC patients, high expression of AKR1C1 and SERPINE1 was linked to worse overall survival (OS). To learn more about the interconnected roles and processes of DEGs, KEGG pathway enrichment analysis was employed for the functional annotation of these genes. To further understand the biological significance of co-regulated proteins, KEGG pathway enrichment analyzed the collected expression data of AKR1C1 and SERPINE1 messenger RNA (mRNA) for HCC samples from TCGA dataset. The obtained results showed that AKR1C1 is correlated with several tumor progression cascades, such as the Wnt signaling cascade, the calcium signaling system, the cascades that lead to gastric cancer, and so on (Figure 3F). In addition to its role in the PI3K/AKT signaling cascade, SERPINE1 is also engaged in the Wnt signaling cascade, the IL-17 signaling cascade, etc. (Figure 3G). The Wnt signaling pathway is shared by AKR1C1 and SERPINE1, suggesting that lenvatinib may affect the progression of HCC through the Wnt signaling pathway.

The gene expression of downloaded samples was analyzed using CIBERSORT to estimate the levels of 22 different types of immune cells. This helped find differences in the expression levels between the low- and high-expression groups. For AKR1C1, the ratios of activated dendritic cells activated memory CD4 T cells, and eosinophils were different between the two groups, and the proportions of macrophages (M0) were significantly higher in the high-expression group (Figure 3H). Moreover, for SERPINE1 the ratios of the proportions of activated NK cells (natural killer cells), neutrophils, and activated mast cells were found to be changed in two groups. The proportions of macrophages (M0) and resting memory CD4 T cells, were significantly elevated in the high-expression group, while the ratios of resting mast cells were considerably elevated in the low-expression group (Figure 3I). AKR1C1 and SERPINE1 were significantly correlated with at least one type of immune cell and were positively correlated with macrophages (M0) (Figure 3J). Taken together, the underlined data suggested that AKR1C1 and SERPINE1 may be crucial contributors to the development of HCC.

AKR1C1 was increased in LR HCC cells and HCC tissues

Hou et al. developed two LR (LR HCC cell lines, i.e., Hep3B_LR and Huh7_LR (18). Venn diagrams indicated that the XDH, RAB20, ITGA3, GALNT6, NYNRIN, DCHS1, and FERMT1 genes were co-upregulated in both LR cell lines, and the LINC00941, INHBB, and MYBPC3 genes were down-regulated collectively in both LR cell lines (Figure 4A,4B, Table 4).

Figure 4 The AKR1C1 expression was elevated in LR-HCC cells. (A) Seven up-regulated genes in LR Hep3B and Huh7 cells. (B) Three down-regulated genes in LR Hep3B and Huh7 cells. AKR1C1 was increased in HCC cell lines that were resistant to lenvatinib compared with the control group, (C) Hep3B, and (D) Huh7, respectively. (E) AKR1C1 was elevated in HCC cells and LEN-HepG2 (LR-HepG2 cells). **P<0.01, ****P<0.0001. LR, lenvatinib-resistant; HCC, hepatocellular carcinoma; NC, negative control; LEN, lenvatinib; NS, no sense.

Using in vitro experiments, drug-resistant cell lines can be established using either the low concentration gradient rise or high dosage intermittent impact. In our previous study, we found that the expression of AKR1C1 was increased in MHCC-97H cells co-cultured with different concentrations of lenvatinib, especially when co-cultured with lenvatinib at 100 µM. Simultaneously, the expression of AKR1C1 was also enhanced in the drug-resistant hepatoma cells induced by low concentrations of lenvatinib from the GSE186191 dataset (Figure 4C,4D). To further verify these conclusions, HepG2_LR cell lines were established and the results demonstrated that the expression of AKR1C1 was higher in HCC cells and also higher in LR cells than in the corresponding wild-type cells (Figure 4E). These findings suggest that AKR1C1 can be used as a marker of lenvatinib resistance in HCC.

Immunohistochemistry was applied to detect the expression of AKR1C1 in HCC tissue chips. We observed that the expression of AKR1C1 was different between the HCC tissue and the tumor-adjacent liver tissue (Figure 5A). AKR1C1 has an immunostaining score of 0 to 12, using the product of the percentage of tumor cells with positive staining and staining intensity as the criterion. The expression level of AKR1C1 was determined by staining intensity (no staining, 0; Light staining, 1 point; Medium staining, 2 points; Heavy staining, 3 points) (Figure 5B), percentage of stained cells in cell count (<5%, 0 points; 5–25%, 1; 26–50%, 2 points; 51–75%, 3 points; >75%, 4 points). Multiply the two points to get a positive level. A total score greater than 6 indicates high AKR1C1 expression, while a score less than or equal to 6 indicates low AKR1C1 expression. In the 90 pairs of tissue sections in this study, there was a statistically significant difference in AKR1C1 expression between HCC tissues and liver tissues adjacent to tumors (Figure 5C). In addition, patients were divided into high AKR1C1 expression group and low AKR1C1 expression group based on the expression of AKR1C1 in HCC tissues relative to tumor adjacent liver tissues of HCC patients. Combined with the clinicopathological data of the patients, the results showed that AKR1C1 expression was closely related to cirrhosis, tumor number, lymph gland metastasis, and the Tumor Node Metastasis (TNM) stage of HCC patients (P<0.05, Table 5). The overall survival rates of patients in the high AKR1C1 expression group were significantly lower than those in the low AKR1C1 expression group (Figure 5D).

Figure 5 Analysis of AKR1C1 expression in human HCC tissues and the survival of HCC patients. (A) Expression of AKR1C1 in human HCC tissues and paracancerous tissues (IHC staining). (B) Staining scores of AKR1C1 expression in HCC tissue samples and surrounding noncancerous tissues (IHC staining). (C) The immunostaining scores of AKR1C1 in 90 pairs of human HCC tissues and paracancerous tissues (****P<0.0001). (D) Kaplan-Meier analysis of the overall survival outcomes for 90 pairs of human HCC tissues (P=0.0198). *P<0.05, **P<0.01. HCC, hepatocellular carcinoma; TNM, International Classification of Clinical Stages for Malignant Tumors; AFP, alpha-fetoprotein.

Low expression of AKR1C1 inhibits the proliferation of HCC cells

The roles of AKR1C1 in the proliferation of LEN-HepG2 cells were respectively analyzed by a CCK-8, wound healing assay, and EdU. First, we found that in comparison to the siNC group (control group), AKR1C1 mRNA expression in LEN-HepG2 was downregulated after transduction with AKR1C1 siRNA (Figure 6A). The CCK-8 assays showed that LEN-HepG2 cell viability was notably lower in the knockdown group compared with the siNC group (Figure 6B). Furthermore, the wound healing and EdU assays (Figure 6C,6D) displayed the same results. Meanwhile, in MHCC-97H cells, the knockdown of AKR1C1 also decreased the proliferation and migration ability of MHCC-97H cells (Figure 7A). CCK-8 assays revealed that cell proliferation of MHCC-97H cells was notably reduced after transduction (Figure 7B). In addition, the wound healing (Figure 7C) and EdU (Figure 7D) assays also showed the same results.

Figure 6 Knockdown of AKR1C1 inhibits the proliferation of LEN-HepG2 cells. (A) The mRNA expression level of AKR1C1 in LEN-HepG2 cells was determined by qRT-PCR analysis after transduction. The proliferation of LEN-HepG2 cells was assessed by CCK-8 assays (B), wound healing assay (C; ×100), and EdU (D; ×100), respectively. **P<0.01, ***P<0.001. NC, negative control; OD, optical density; EdU, 5-ethynyl-2'-deoxyuridine.

Figure 7 Knockdown of AKR1C1 inhibits the proliferation of MHCC-97H cells. (A) The mRNA expression level of AKR1C1 in MHCC-97H cells was determined by qRT-PCR analysis after transduction. The proliferation of MHCC-97H cells was assessed by CCK-8 assays (B), wound healing assay (C), and EdU (D), respectively (×20). **P<0.01, ***P<0.001, ****P<0.0001. NC, negative control; OD, optical density; EdU, 5-ethynyl-2'-deoxyuridine; DAPI, 4'-6-diamidino-2-phenylindole.

We then calculated the weights and volumes of the tumors to explore the roles of AKR1C1 in the tumor growth of HCC cells in vivo. Our results showed that the AKR1C1 siRNA group had a reduced tumor weight and volume compared with the siNC group (Figure 8A-8C). Also, immunohistochemistry staining revealed elevated expressions of proliferating cell nuclear antigen (PCNA) and N-cadherin and a reduced expression level of E-cadherin in the siNC group (Figure 8D).

Figure 8 Knockdown of AKR1C1 inhibits the proliferation of HCC cells in vivo. (A) Images of tumors in the AKR1C1 knockdown and the control groups. (B,C) Knockdown of AKR1C1 causes a significant decrease in tumor weight and volume (**P<0.01). (D) IHC staining images of E-cadherin, N-cadherin, and PCNA in the xenograft tumor tissues from each group of mice. HCC, hepatocellular carcinoma; NC, negative control; IHC, immunohistochemistry; PCNA, proliferating cell nuclear antigen.

MiRNA prediction

MiRNAs are epigenetic regulators that have shown enormous potential as therapeutic targets for a variety of human diseases and are the subject of extensive research (19). They play essential roles in processes related to tumor growth, including proliferation, angiogenesis, and invasion (20). The microRNA data integration portal (mirDIP) and starBase databases were utilized to predict the interactions between miRNAs and AKR1C1 (Table S4), and only one miRNA, i.e., miR-4644, was identified (Figure 9A). It was also found that miR-4644 expression was elevated in LR cells compared with the corresponding wild-type cells (Figure 9B). According to reports from Zhao et al., miR-4644 has higher stability and expression abundance than other regularly used reference controls, making it an effective endogenous normalizer for circulating microRNA quantification in HCC (21). Considering the elevated level of miR-4644 expression in LR cells, circulating miR-4644 may be a promising biomarker for the early diagnosis of LR.

Figure 9 MiR-4644 may be a biomarker for the early diagnosis of LR in HCC. (A) The mirDIP and starBase databases predicted that miR-4644 may bind to AKR1C1; (B) in cells that have developed resistance to lenvatinib, miR-4644 expression was significantly higher than in their wild-type counterparts; (C) prediction of the lncRNA-miRNA network in HCC. *P<0.05. LR, lenvatinib-resistant; HCC, hepatocellular carcinoma.

Extensive evidence from numerous sources confirms that long non-coding RNAs (lncRNAs) are crucial in the development of HCC treatment resistance (22,23). It was previously reported that the expression of LINC00941 was reduced in LR cells (Table 4); therefore, we used the database to further investigate the lncRNAs that interacted with regulator miRNAs and generated lncRNA-miRNA subnetworks (Figure 9C). Interestingly, it was determined that several lncRNAs enhance cell proliferation in the development of hepatic carcinogenesis, such as taurine up-regulated 1 (TUG1) (24), and human leukocyte antigen (HLA) complex group 18 (HCG18) (25).

Erastin inhibits the expression of AKR1C1 in HepG2 cells

Studies have revealed that AKR1C1 is linked to numerous ferroptosis-associated cancers, such as glioblastoma (26), colon adenocarcinoma (27), non-small cell lung cancer (28), breast cancer (29) and clear cell renal cell carcinoma (30). Recent research has indicated that sorafenib can elicit ferroptosis, a type of programmed cell death characterized by iron-dependent lipid peroxide accumulation (31,32). Our results showed that erastin treatment inhibits the AKR1C1 expression in HepG2 cells from the GSE104462 dataset (Figure 10A) (33). Our experiment demonstrated that different concentrations of erastin can inhibit the expression of AKR1C1 in LEN-HepG2 cells (Figure 10B). In addition, sorafenib, which was the earliest approved first-line treatment for advanced HCC (34), also inhibits the expression of AKR1C1 in LEN-HepG2 cells (Figure 10C). This suggests that LR patients who exhibit elevated AKR1C1 expression can be treated with sorafenib, but more experiments to verify this viewpoint.

Figure 10 Erastin or sorafenib suppressed AKR1C1 expression in HCC cells. (A) GSE104462 shows that erastin suppressed AKR1C1 expression in HepG2 cells. (B) Different concentrations of erastin were co-cultured with LEN-HepG2 cells for 24 h. According to the results, erastin suppressed AKR1C1 expression in LEN-HepG2 cells. (C) Sorafenib suppressed AKR1C1 expression in LEN-HepG2 cells after being co-cultured for 24 h. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. NC, negative control.

Prediction of drug sensitivity associated with AKR1C1

Considering the probable role of AKR1C1 in regulating medication sensitivity, we investigated whether individuals in different risk categories had varying levels of sensitivity to various cancer drugs. Each HCC patient’s IC₅₀ value was predicted using the prediction model. Taking IC₅₀ <1 as the screening criterion, we were able to narrow down the number of prospective anticancer medicines to 11. This means that these compounds have a strong inhibitory effect on HCC (Figure 11A). The response to five of these medications varied in the various risk categories with a statistically significant P value <0.05 (Figure 11B). Other than SN-38 (7-ethyl-10-hydroxycamptothecin), the majority of the medications, such as trametinib, vinorelbine, sepantronium bromide, and luminespib, displayed greater IC₅₀ values in the high-risk groups than in the low-risk groups, indicating that SN-38 patients were more likely to respond appropriately.

Figure 11 A drug sensitivity analysis. (A) Drug sensitivity analyses (IC₅₀<1). (B) Potential drugs with remarkable treatment variations between the high- and low-risk populations. The P value of each group is less than 0.05.

Discussion

Patients with cirrhosis have an increased risk of developing HCC, and those with a single tumor and healthy liver function make the best resection candidates. The conventional treatments for advanced HCC are sorafenib, lenvatinib, and regorafenib, as they increase the patient’s survival rate (35). Lenvatinib, an oral small molecule inhibitor of several receptor tyrosine kinases, has been approved in the United States, the European Union, Japan, and China as the first-line therapy in patients with unresectable HCC. Lenvatinib significantly improves the objective response rate (ORR), progression-free survival (PFS), and time to progression (TTP) compared to sorafenib therapy (36). At present, HCC treatment is challenging because of the lack of understanding of the molecular mechanisms of lenvatinib resistance in HCC, which occurs often in both primary and acquired forms.

In this study, we first confirmed that lenvatinib inhibits the proliferation of HCC cells. To explore the underlying molecular mechanism that mediates the phenotypic changes, RNA-seq was used to compare the transcriptomes of MHCC-97H cells exposed to varying doses of lenvatinib with those of their parental counterparts. We identified two of the most valuable genes, i.e., AKR1C1 and SERPINE1. When HCC cells were exposed to various doses of lenvatinib, the expression of AKR1C1 was upregulated, while that of SERPINE1 was downregulated. According to the obtained results, lenvatinib can inhibit the expression of SERPINE1 (an oncogene in HCC) in HCC cells, regardless of the concentration (16,17) This indicated that SERPINE1 was the target of lenvatinib, and in HCC patients with a high expression of SERPINE1, lenvatinib can be used for targeted treatment. In contrast, AKR1C1 was increased in HCC cells treated with varying doses of lenvatinib, and the prognosis of HCC patients with high expression of AKR1C1 was poor. Therefore, it is suggested that AKR1C1 may induce lenvatinib resistance in HCC cells.

The KEGG links genomic data with higher-order functional data to analyze genes in a systematic way (37). Studies have confirmed that lenvatinib can regulate programmed cell death 1 ligand 1 (PD-L1) levels and regulatory T cells (Treg) differentiation in HCC (38), and CIBERSORT can accurately estimate the immune composition of a tumor biopsy (39). In this study, KEGG pathway enrichment analysis was performed to retrieve the expression data of AKR1C1 and SERPINE1 mRNA in HCC samples from TCGA database to better understand the biological function of HCC. Moreover, we also assessed the immune response of 22 TIICs using CIBERSORT. The Wnt/β-catenin pathway is typically upregulated in HCC and has been linked to tumor-initiating cell maintenance, tumor progression, drug resistance, and metastasis (40). KEGG analysis revealed that the Wnt signaling pathway is shared by AKR1C1 and SERPINE1, which indicates that this pathway plays an important role in the treatment of HCC using lenvatinib. Compared to healthy liver tissues, the infiltration of M0 macrophages was noticeably higher in HCC tissues. Correlation analysis showed that M0 macrophage-related genes (M0RGs) had positive associations with immune cell infiltration, immune checkpoint inhibitory targets, clinicopathological characteristics, and the efficacy of common drugs (41). Resting-state macrophages (M0), which originate in the bone marrow, are typically regarded as precursors of polarized macrophages. M0-like macrophages are considered a feature of tumor malignancy (42). Additionally, these macrophages may act as a biomarker for HCC carcinogenesis, progression, and clinical results (43). CIBERSORT showed AKR1C1 and SERPINE1 are significantly correlated with M0 macrophages, which revealed that these two genes act as potential characteristic markers of HCC that are enriched in M0 macrophages.

The drug resistance of lenvatinib in HCC has recently gained increasing attention (18,44). By investigating the underlying molecular basis of lenvatinib resistance, we detected genes with abnormal expression in the GEO186191 dataset and the RNA-seq data for parental and LR HCC cells. We found that XDH, RAB20, ITGA3, GALNT6, NYNRIN, DCHS1, and FERMT1 were co-upregulated genes, and LINC00941, INHBB, and MYBPC3 were co-downregulated genes in both LR cell lines. This finding demonstrated that the XDH, RAB20, ITGA3, GALNT6, NYNRIN, DCHS1, and FERMT1 genes may be targets of lenvatinib resistance, as genes with up-expression in drug-resistant cell lines possess a possible mechanism of producing drug resistance. However, we discovered that XDH (45), RAB20 (46), and ITGA3 (47) were downregulated and inhibited HCC cell proliferation; previous evidence suggests that these genes exert a tumor-suppressive function. Similarly, LINC00941’s activity as an oncogene is decreased in drug-resistant cell lines (48). Guan et al. considered that under the stimulation of sorafenib, the tumor suppressor gene TSC2 becomes a “tumor-promoting factor” and then mediates the transmission of drug resistance signals (49), which implies that, like sorafenib resistance, the tumor microenvironment of LR patients has changed and requires personalized targeted therapy.

In the GEO186191 dataset, we found that the expression of AKR1C1 was elevated in the drug-resistant HCC cells. Moreover, our results also showed that AKR1C1 was highly expressed in HCC cells and LR cells. In our study, AKR1C1 was overexpressed in HCC tissues, and high expression of AKR1C1 predicts a poor outcome. Furthermore, in vitro experiments confirmed that the knockdown of AKR1C1 inhibits the proliferation of LEN-HepG2 and MHCC-97H cells. According to these results, AKR1C1 plays a pivotal role in LR in HCC cells. Moreover, miR-4644 was identified and is highly expressed in LR cells compared with the corresponding wild-type cells. This is contrary to the competitive endogenous RNAs (ceRNA) network. Therefore, this study proposes two hypotheses. Firstly, miRNA may also function to induce gene expression (50), and miR-4644 may target the AKR1C1 promoter site and induce AKR1C1 expression. Secondly, transcription factors involved in the regulation of AKR1C1 are regulated by miR-4644; miR-4644 and AKR1C1 do not bind directly. According to the results of this study, miR-4644 may act as a circulating microRNA to predict LR in HCC. However, there is still a need for further study to confirm this.

Additionally, this study revealed that AKR1C1 was associated with multiple pathways that were linked to ferroptosis in different types of cancers (28,29). Ferroptosis is an iron-dependent regulatory cell death that is induced by severe lipid peroxidation (51). Erastin (an inducer of ferroptosis) inhibits the expression of AKR1C1 in HepG2 cells. Considering that sorafenib can induce ferroptosis in HCC (52), we speculate that sorafenib can treat HCC patients with an elevated level of AKR1C1 expression and individuals who are resistant to lenvatinib. Meanwhile, the GDSC database indicated that SN-38 was more likely to respond well in the high AKR1C1 expression group. SN-38 is a semi-synthetic hydroxycamptothecin anti-tumor active substance, which can specifically bind to topoisomerase I to inhibit the synthesis of DNA in tumor cells. Irinotecan (CPT-11) as a clinical application preparation was first developed in Japan; it can release the prototype drug SN-38 with anti-cancer activity in the human body under the catalysis of carboxylesterase (53). Moreover, a study has shown that SN-38 and sorafenib have synergistic anticancer activity on HCC cells in vitro by augmenting apoptosis (54). According to our previous study, sorafenib can inhibit the expression of AKR1C1. The results obtained in this study demonstrated that a combination of SN-38 and sorafenib is a possible approach to treat LR in HCC. Other study found that neurofibromin 1 (NF1), and dual specificity phosphatase 9 (DUSP9), as critical drivers for lenvatinib resistance in HCC (55). However, these are trial treatments following LR; the role of LR sensitivity in HCC needs to be validated by extensive studies. Next, we will use more experiments to verify our viewpoint.

Conclusions

Taken together, this study found that AKR1C1 has potential prognostic significance in the prediction of LR in HCC. AKR1C1 could be a promising therapeutic target for patients with LR-type liver cancer. Moreover, it is necessary to explore the molecular mechanism of LR and evaluate the therapeutic effect on these HCC patients.

Acknowledgments

Funding: This study was supported by grants from the National Natural Science Foundation of China (No. 81871927) and the Postgraduate Research & Practice Innovation Program of Jiangsu Province (No. KYCX20_2798).

Footnote

Reporting Checklist: The authors have completed the ARRIVE reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-23-277/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-23-277/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-23-277/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-23-277/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). Informed consent was obtained from the patients, and this research was approved by the Ethics Committee of Affiliated Hospital of Nantong University (No. 2018-L006). Animal experiments were performed under a project license (No. S20220228-004) granted by the Institute Ethics Committee at the Affiliated Hospital of Nantong University, in compliance with the guidance of the care and use of laboratory animals issued by the Ministry of Science and Technology of the China.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Chidambaranathan-Reghupaty S, Fisher PB, Sarkar D. Hepatocellular carcinoma (HCC): Epidemiology, etiology and molecular classification. Adv Cancer Res 2021;149:1-61. [Crossref] [PubMed]
2. Zhao Y, Zhang YN, Wang KT, et al. Lenvatinib for hepatocellular carcinoma: From preclinical mechanisms to anti-cancer therapy. Biochim Biophys Acta Rev Cancer 2020;1874:188391. [Crossref] [PubMed]
3. Komatsu S, Yano Y, Sofue K, et al. Assessment of lenvatinib treatment for unresectable hepatocellular carcinoma with liver cirrhosis. HPB (Oxford) 2020;22:1450-6. [Crossref] [PubMed]
4. Jin H, Shi Y, Lv Y, et al. EGFR activation limits the response of liver cancer to lenvatinib. Nature 2021;595:730-4. [Crossref] [PubMed]
5. Tang W, Chen Z, Zhang W, et al. The mechanisms of sorafenib resistance in hepatocellular carcinoma: theoretical basis and therapeutic aspects. Signal Transduct Target Ther 2020;5:87. [Crossref] [PubMed]
6. Ruf B, Heinrich B, Greten TF. Immunobiology and immunotherapy of HCC: spotlight on innate and innate-like immune cells. Cell Mol Immunol 2021;18:112-27. [Crossref] [PubMed]
7. Llovet JM, Zucman-Rossi J, Pikarsky E, et al. Hepatocellular carcinoma. Nat Rev Dis Primers 2016;2:16018. [Crossref] [PubMed]
8. Vogel A, Qin S, Kudo M, et al. Lenvatinib versus sorafenib for first-line treatment of unresectable hepatocellular carcinoma: patient-reported outcomes from a randomised, open-label, non-inferiority, phase 3 trial. Lancet Gastroenterol Hepatol 2021;6:649-58. [Crossref] [PubMed]
9. Zeng CM, Chang LL, Ying MD, et al. Aldo-Keto Reductase AKR1C1-AKR1C4: Functions, Regulation, and Intervention for Anti-cancer Therapy. Front Pharmacol 2017;8:119. [Crossref] [PubMed]
10. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 2014;15:550. [Crossref] [PubMed]
11. Smyth GK, Michaud J, Scott HS. Use of within-array replicate spots for assessing differential expression in microarray experiments. Bioinformatics 2005;21:2067-75. [Crossref] [PubMed]
12. Chin CH, Chen SH, Wu HH, et al. cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 2014;8:S11. [Crossref] [PubMed]
13. Newman AM, Liu CL, Green MR, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods 2015;12:453-7. [Crossref] [PubMed]
14. Yu G, Wang LG, Han Y, et al. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 2012;16:284-7. [Crossref] [PubMed]
15. Yang W, Soares J, Greninger P, et al. Genomics of Drug Sensitivity in Cancer (GDSC): a resource for therapeutic biomarker discovery in cancer cells. Nucleic Acids Res 2013;41:D955-61. [Crossref] [PubMed]
16. Li G, Du P, He J, et al. CircRNA circBACH1 (hsa_circ_0061395) serves as a miR-656-3p sponge to facilitate hepatocellular carcinoma progression through increasing SERBP1 expression. Biochem Biophys Res Commun 2021;556:1-8. [Crossref] [PubMed]
17. Li LM, Chen C, Ran RX, et al. Loss of TARBP2 Drives the Progression of Hepatocellular Carcinoma via miR-145-SERPINE1 Axis. Front Oncol 2021;11:620912. [Crossref] [PubMed]
18. Hou W, Bridgeman B, Malnassy G, et al. Integrin subunit beta 8 contributes to lenvatinib resistance in HCC. Hepatol Commun 2022;6:1786-802. [Crossref] [PubMed]
19. Hayes J, Peruzzi PP, Lawler S. MicroRNAs in cancer: biomarkers, functions and therapy. Trends Mol Med 2014;20:460-9. [Crossref] [PubMed]
20. Lee YS, Dutta A. MicroRNAs in cancer. Annu Rev Pathol 2009;4:199-227. [Crossref] [PubMed]
21. Zhao J, Zhu XC, Wu XS, et al. Identification of miR-4644 as a suitable endogenous normalizer for circulating miRNA quantification in hepatocellular carcinoma. J Cancer 2020;11:7032-44. [Crossref] [PubMed]
22. Al-Noshokaty TM, Mesbah NM, Abo-Elmatty DM, et al. Selenium nanoparticles overcomes sorafenib resistance in thioacetamide induced hepatocellular carcinoma in rats by modulation of mTOR, NF-κB pathways and LncRNA-AF085935/GPC3 axis. Life Sci 2022;303:120675. [Crossref] [PubMed]
23. Zhang Y, Luo M, Cui X, et al. Long noncoding RNA NEAT1 promotes ferroptosis by modulating the miR-362-3p/MIOX axis as a ceRNA. Cell Death Differ 2022;29:1850-63. [Crossref] [PubMed]
24. Tang K, Lv D, Miao L, et al. LncRNA TUG1 functions as a ceRNA for miR-1-3p to promote cell proliferation in hepatic carcinogenesis. J Clin Lab Anal 2022;36:e24415. [Crossref] [PubMed]
25. Zou Y, Sun Z, Sun S. LncRNA HCG18 contributes to the progression of hepatocellular carcinoma via miR-214-3p/CENPM axis. J Biochem 2020;168:535-46. [Crossref] [PubMed]
26. Tian Y, Liu H, Zhang C, et al. Comprehensive Analyses of Ferroptosis-Related Alterations and Their Prognostic Significance in Glioblastoma. Front Mol Biosci 2022;9:904098. [Crossref] [PubMed]
27. Miao YD, Kou ZY, Wang JT, et al. Prognostic implications of ferroptosis-associated gene signature in colon adenocarcinoma. World J Clin Cases 2021;9:8671-93. [Crossref] [PubMed]
28. Huang F, Zheng Y, Li X, et al. Ferroptosis-related gene AKR1C1 predicts the prognosis of non-small cell lung cancer. Cancer Cell Int 2021;21:567. [Crossref] [PubMed]
29. Zhang Z, Qiu X, Yan Y, et al. Evaluation of Ferroptosis-related Gene AKR1C1 as a Novel Biomarker Associated with the Immune Microenvironment and Prognosis in Breast Cancer. Int J Gen Med 2021;14:6189-200. [Crossref] [PubMed]
30. Hong Y, Lin M, Ou D, et al. A novel ferroptosis-related 12-gene signature predicts clinical prognosis and reveals immune relevancy in clear cell renal cell carcinoma. BMC Cancer 2021;21:831. [Crossref] [PubMed]
31. Chen S, Zhu JY, Zang X, et al. The Emerging Role of Ferroptosis in Liver Diseases. Front Cell Dev Biol 2021;9:801365. [Crossref] [PubMed]
32. Gao R, Kalathur RKR, Coto-Llerena M, et al. YAP/TAZ and ATF4 drive resistance to Sorafenib in hepatocellular carcinoma by preventing ferroptosis. EMBO Mol Med 2021;13:e14351. [Crossref] [PubMed]
33. Zhang X, Du L, Qiao Y, et al. Ferroptosis is governed by differential regulation of transcription in liver cancer. Redox Biol 2019;24:101211. [Crossref] [PubMed]
34. Faivre S, Rimassa L, Finn RS. Molecular therapies for HCC: Looking outside the box. J Hepatol 2020;72:342-52. [Crossref] [PubMed]
35. Forner A, Reig M, Bruix J. Hepatocellular carcinoma. Lancet 2018;391:1301-14. [Crossref] [PubMed]
36. Al-Salama ZT, Syed YY, Scott LJ. Lenvatinib: A Review in Hepatocellular Carcinoma. Drugs 2019;79:665-74. [Crossref] [PubMed]
37. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res 2000;28:27-30. [Crossref] [PubMed]
38. Yi C, Chen L, Lin Z, et al. Lenvatinib Targets FGF Receptor 4 to Enhance Antitumor Immune Response of Anti-Programmed Cell Death-1 in HCC. Hepatology 2021;74:2544-60. [Crossref] [PubMed]
39. Chen B, Khodadoust MS, Liu CL, et al. Profiling Tumor Infiltrating Immune Cells with CIBERSORT. Methods Mol Biol 2018;1711:243-59. [Crossref] [PubMed]
40. Vilchez V, Turcios L, Marti F, et al. Targeting Wnt/β-catenin pathway in hepatocellular carcinoma treatment. World J Gastroenterol 2016;22:823-32. [Crossref] [PubMed]
41. Zhang Y, Zou J, Chen R. An M0 macrophage-related prognostic model for hepatocellular carcinoma. BMC Cancer 2022;22:791. [Crossref] [PubMed]
42. Huang L, Wang Z, Chang Y, et al. EFEMP2 indicates assembly of M0 macrophage and more malignant phenotypes of glioma. Aging (Albany NY) 2020;12:8397-412. [Crossref] [PubMed]
43. Zhang Z, Wang Z, Huang Y. Comprehensive Analyses of the Infiltrating Immune Cell Landscape and Its Clinical Significance in Hepatocellular Carcinoma. Int J Gen Med 2021;14:4695-704. [Crossref] [PubMed]
44. Zhang P, Sun H, Wen P, et al. circRNA circMED27 acts as a prognostic factor and mediator to promote lenvatinib resistance of hepatocellular carcinoma. Mol Ther Nucleic Acids 2022;27:293-303. [Crossref] [PubMed]
45. Chen GL, Ye T, Chen HL, et al. Xanthine dehydrogenase downregulation promotes TGFβ signaling and cancer stem cell-related gene expression in hepatocellular carcinoma. Oncogenesis 2017;6:e382. [Crossref] [PubMed]
46. Liu BHM, Tey SK, Mao X, et al. TPI1-reduced extracellular vesicles mediated by Rab20 downregulation promotes aerobic glycolysis to drive hepatocarcinogenesis. J Extracell Vesicles 2021;10:e12135. [Crossref] [PubMed]
47. Hou J, Wang L, Wu D. The root of Actinidia chinensis inhibits hepatocellular carcinomas cells through LAMB3. Cell Biol Toxicol 2018;34:321-32. [Crossref] [PubMed]
48. Chen J, Tang D, Li H, et al. Expression changes of serum LINC00941 and LINC00514 in HBV infection-related liver diseases and their potential application values. J Clin Lab Anal 2022;36:e24143. [Crossref] [PubMed]
49. Guan DX, Shi J, Zhang Y, et al. Sorafenib enriches epithelial cell adhesion molecule-positive tumor initiating cells and exacerbates a subtype of hepatocellular carcinoma through TSC2-AKT cascade. Hepatology 2015;62:1791-803. [Crossref] [PubMed]
50. Place RF, Li LC, Pookot D, et al. MicroRNA-373 induces expression of genes with complementary promoter sequences. Proc Natl Acad Sci U S A 2008;105:1608-13. Erratum in: Proc Natl Acad Sci U S A 2018;115:E3325. [Crossref] [PubMed]
51. Chen X, Kang R, Kroemer G, et al. Broadening horizons: the role of ferroptosis in cancer. Nat Rev Clin Oncol 2021;18:280-96. [Crossref] [PubMed]
52. Nie J, Lin B, Zhou M, et al. Role of ferroptosis in hepatocellular carcinoma. J Cancer Res Clin Oncol 2018;144:2329-37. [Crossref] [PubMed]
53. Kunimoto T, Nitta K, Tanaka T, et al. Antitumor activity of 7-ethyl-10-(4-(1-piperidino)-1-piperidino)carbonyloxy-camptothec in, a novel water-soluble derivative of camptothecin, against murine tumors. Cancer Res 1987;47:5944-7. [PubMed]
54. Xu L, Yuan-run Z, Jian C, et al. Anticancer effect of SN-38 combined with sorafenib on hepatocellular carcinoma in vitro and its mechanism. Zhejiang Da Xue Xue Bao Yi Xue Ban 2015;44:486-92. [PubMed]
55. Lu Y, Shen H, Huang W, et al. Genome-scale CRISPR-Cas9 knockout screening in hepatocellular carcinoma with lenvatinib resistance. Cell Death Discov 2021;7:359. Erratum in: Cell Death Discov 2022;8:74. [Crossref] [PubMed]

(English Language Editor: A. Kassem)