LncRNA GABPB1-AS1 and GABPB1 regulate oxidative stress during erastin-induced ferroptosis in HepG2 hepatocellular carcinoma cells

Ferroptosis is a non-apoptotic, iron-dependent oxidative form of cell death that is specifically induced by erastin in RAS mutant cancer cells. Ferroptotic cell death is the result of membrane lipid peroxide damage caused by the accumulation of hydroxyl radicals derived from H2O2 by the Fenton reaction. Peroxidases are key cellular antioxidant enzymes that block such damaging processes. Few studies have examined the roles of long non-coding RNAs (lncRNAs) in the regulation of cellular oxidative stress, especially in ferroptosis. Here, we demonstrated that erastin upregulated the lncRNA GABPB1-AS1, which downregulated GABPB1 protein levels by blocking GABPB1 translation, leading to the downregulation of the gene encoding Peroxiredoxin-5 (PRDX5) peroxidase and the eventual suppression of the cellular antioxidant capacity. Such effects critically inhibited the cellular antioxidant capacity and cell viability. Additionally, high expression levels of GABPB1 were correlated with poor prognosis of hepatocellular carcinoma (HCC) Patients, while high GABPB1-AS1 levels in HCC patients correlated with improved overall survival. Collectively, these data demonstrate a mechanistic link between GABPB1 and its antisense lncRNA GABPB1-AS1 in erastin-induced ferroptosis and establish GABPB1 and GABPB1-AS1 as attractive therapeutic targets for HCC.

defence systems to prevent oxidative damage caused by ROS 12 . The PRDX gene family includes six isoforms in mammals: Typical 2-cysteine PRDXs (PRDXs 1-4), atypical 2-cysteine PRDX (PRDX5), and atypical 1-cysteine PRDX (PRDX6) 13 . PRDX1 is mainly located in the cytoplasm, and knockdown of PRDX1 in mice leads to haemolytic anaemia and a shortened life span 14 . PRDX3 is located in the mitochondria, protects mitochondria against oxidative damage and affects diverse cellular processes, including growth, differentiation, carcinogenesis and apoptosis 15 . PRDX6 was identified as a secretory antioxidant protein of the olfactory epithelium 16 . A new study has revealed that PRDX6 plays an essential role in protecting cells against ferroptosis 17 . PRDX5 is localized widely in cells, including in the cytosol, mitochondria, peroxisome and nucleus and is able to reduce hydrogen peroxide, alkyl hydroperoxides, and peroxynitrite 18 . Although PRDX5 is involved in redox regulation in various metabolic functions, it remains unknown whether the function of this protein is related to ferroptosis.
Nuclear respiration factor-2 (NRF2), also called GA-binding protein, is a transcription activator 19 that not only regulates the expression of genes encoding mitochondrial respiratory chain proteins and mitochondrial transport proteins [20][21][22] but also affects the cellular antioxidant capacity by regulating the expression of genes encoding peroxidases, such as PRDX5. NRF2 has three subunits, alpha, beta and gamma, and GABPB1 protein is the activation subunit of NRF2 23 . GABPB1 forms a tetrameric complex with the alpha subunit and stimulates the transcription of various genes 24,25 including the antioxidant gene PRDX5 26 . PRDX5 is expressed at a high level in a variety of mammalian cell lines and normal tissues 27 . The transcription factor GABP/NRF-2 is required for the basal expression of the human PRDX5 gene, and it was also found that mutation-mediated inactivation of all potential GAPB binding sites in the PRDX5 promoter led to 80% inhibition of its basal activity in a reporter gene assay 26 . The GABPB1-AS1 lncRNA is the antisense RNA of GABPB1 mRNA. The expression level of GABPB1-AS1 changes in response to various stimuli such as stress 28,29 but the physiological significance and molecular mechanism are not clear.
In this study, we analysed the regulatory effects of lncRNA GABPB1-AS1 on GABPB1 and its potential functions in HepG2 hepatocellular carcinoma cell ferroptosis induced by erastin. We found that GABPB1-AS1 was upregulated by erastin which inhibited the translation of GABPB1, leading to the inhibition of peroxidase gene expression and causing the accumulation of ROS and MDA and HepG2 cell death. These findings revealed that GABPB1-AS1 may be a key molecule in HCC cell ferroptosis induced by erastin and enrich our understanding of lncRNA regulation of oxidative stress.
Vector construction, small interfering RNA (siRNA) synthesis and cell transfection. The cDNA encoding the coding sequence of GABPB1 was PCR-amplified using PrimeStar Max PCR mix(Takara) and subcloned into the NheI and BamHI (Thermo Scientific) sites of the pcDNA3.1 vector. For the expression of GABPB1-AS1 RNA, the cDNA sequence was PCR-amplified using PrimeStar Max PCR mix and subcloned into the NheI and BamHI sites of the pcDNA3.1 vector. The cDNA encoding PRDX5 amplified by PCR was amplified using PrimeStar PCR mix and subcloned into the EcoRI and BamHI sites of the pcDNA3.1 vector. The siRNAs targeting GABPB1, GABPB1-AS1 and the control siRNA were synthesized by Genepharm (Shanghai, China). Transfection of plasmids and siRNAs was conducted using Lipofectamine 3000 (Thermo Scientific) according to the manufacturer's protocol. The sequences of the cloning primers and siRNAs are listed in Supplementary Table 1.
Isolation of cytoplasmic and nuclear RNA. Cytoplasmic and nuclear proteins were separated using the Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime, Shanghai, China) according to the manufacturer's instructions. RiboLock RNase Inhibitor 100 U/ml (Thermo Scientific) was added to the lysis buffer from the kit. The nuclear precipitate was directly resuspended in TRIzol reagent and subjected to RNA extraction.
RNase protection assay (RPA). RNA was isolated from HepG2 cells and incubated at 37 °C for 1 h and then treated with RNase A + T (1 U/ml, Thermo Scientific) for 30 min at 37 °C. RNase A + T digests single-stranded RNA but not RNA duplexes. After RNase A + T treatment, the RNA was extracted using TRIzol. cDNA was generated, and real-time qPCR was performed as described above. The primers used are shown in Supplementary Table 1.
Polysome analysis. Polysome analysis was performed as previously described 30 . At least 50 million HepG2 cells were treated with cycloheximide (CHX) (0.1 mg/ml) (Solarbio, Shanghai, China) for 15 min. Cells were collected and washed with ice-cold PBS containing CHX (0.1 mg/ml) three times, followed by centrifugation at 600 g for 5 min at 4 °C. Cells were then lysed in 1 ml of polysome extraction buffer (15 mM Tris-Cl [pH 7.4], 15 mM MgCl 2 , 0.3 M NaCl, 0.1 mg/ml CHX, 1 mg/ml heparin and 1% Triton X-100) and then centrifuged at 10,000 × g for 20 min at 4 °C. The supernatant was loaded onto a sucrose gradient of 10-50% (w/v) in an ultra-centrifuge (Beckman Coulter) at 100,000 g for 4 h. Absorbance at 254 nm was measured. RNA was isolated from the fractions by phenol-CHCl 3 extraction and ethanol precipitation.

RNA immunoprecipitation (RIP).
HepG2 cell lysate was prepared as described in the RNA pull-down assay section. Cell lysates from 1 × 10 6 cells were immunoprecipitated using 2 μg eIF4A antibody or IgG at 4 °C overnight with rotation. The immunoprecipitates were digested with proteinase K (Thermo Scientific), and then the immunoprecipitated RNA was purified by TRIzol regent (Thermo Scientific). cDNA was synthesized with random primers using the RevertAid RT Kit (Thermo Scientific) and subjected to RT-qPCR for GABPB1 transcripts.
Cell viability assay. Cell viability was measured using the CCK-8 kit (KeyGEN BioTECH, Nanjing, China) according to the manufacturer's protocol. Briefly, HepG2 cells were seeded in 96-well plates at 5000 cells/well in triplicate. After 2 h incubation with fresh media containing 10% (v/v) CCK-8 reagent, the absorbance at 450 nm of each well was measured on a microplate reader. The mean value of the wells with media alone was used as background. Tissue samples from HCC patients. A total of 12 human HCC tissue samples were obtained from the Affiliated Hospital of Changchun University of Traditional Chinese Medicine. The tissues were snap-frozen in liquid nitrogen immediately after dissection and stored in liquid nitrogen until further analysis by western blotting and RT-qPCR. Studies using human tissues were approved by the Human Research Ethics Committees of the Changchun University of Traditional Chinese Medicine, and all participants provided written informed consent. This study followed ICH-GCP guidelines, the Declaration of Helsinki, internationally recognized ethical standards, and relevant Chinese laws and regulations. All collected samples were eligible for experimental use.

Decreased GABPB1 expression levels during erastin-mediated ferroptosis. Erastin is commonly
used as an inducer of ferroptosis 31 , and previous studies have confirmed that 10 μM erastin induces an inhibitory effect on HCC cell growth 32 . CCK-8 assays showed that treatment of HepG2 cells with erastin for 24 h resulted in an ~40% reduction in HepG2 cell survival and that cell death was blocked by ferrostatin-1 (a potent ferroptosis inhibitor) (Fig. 1A). We next analysed the protein and mRNA levels of GABPB1 in HepG2 cells treated with erastin. Remarkably, erastin treatment decreased GABPB1 mRNA and protein expression in HepG2 cells (Fig. 1B,C). The human PRDX5 gene encodes a cytoprotective antioxidant enzyme that inhibits endogenous or exogenous peroxide accumulation and is regulated by the transcription factor GABP/NRF-2 33 . Bioinformatic analysis showed the potential GABP protein binding sites in the PRDX5 promoter 26 (Fig. 1D). Chromatin immunoprecipitation (ChIP) and qPCR assays confirmed the binding loci of GABPB1 on the PRDX5 promotor (Fig. 1E). We also detected reduced protein levels of PRDX5 in HepG2 cells treated with erastin (Fig. 1F). To evaluate the role of GABPB1 in erastin-induced ROS and MDA generation, we overexpressed GABPB1 in HepG2 cells. Overexpression of GABPB1 reduced the ROS generation (Fig. 1G,K) and MDA levels induced by erastin (Fig. 1L) and restored cell viability in the presence of erastin (Fig. 1M). We further examined the role of the peroxidase protein PRDX5 in erastin-induced oxidative stress by overexpression of PRDX5 and observed reduced ROS generation and MDA levels and restored cell viability (Fig. 1I,N-P). To confirm that this observation was not cell line specific, we analysed the protein and mRNA levels of GABPB1 in two other HCC cell lines (Huh7 and Hep3B). The results showed that treatment with erastin significantly reduced GABPB1 expression in Huh7 and Hep3B cell lines ( Supplementary Fig. 1a,b). Together, these results suggest that erastin may promote ROS generation and MDA levels by inhibiting the expression of GABPB1 and its target gene PRDX5.
GABPB1-AS1 is a short-lived lncRNA. The expression levels of GABPB1-AS1 have been shown to be elevated because of prolonged decay rates in response to chemical stressors and interruption of RNA degradation pathways 29 . To determine whether the addition of erastin prolonged the half-life of GABPB1-AS1, HepG2 cells were treated with α-amanitin to block new RNA synthesis, and then the loss of GABPB1-AS1 mRNA was measured. The results showed that erastin prolonged the half-life of GABPB1-AS1 (Fig. 2F). Consistent with the findings in HepG2 cells, erastin also increased the expression levels of GABPB1-AS1 and prolonged the half-life of GABPB1-AS1 in both Huh7 and Hep3B cells ( Supplementary Fig. 2a,b).

GABPB1-AS1 regulates GABPB1 expression at the translational level. The above results
showed that erastin decreased the expression of GABPB1 mRNA and protein while increasing the expression of GABPB1-AS1. These findings suggest that GABPB1-AS1 may have an inhibitory effect on the expression of GABPB1. We next constructed a GABPB1-AS1 expression vector for transfection into HepG2 cells to determine the effects of GABPB1-AS1 on GABPB1 mRNA and protein levels by RT-PCR and western blot analysis, respectively. We confirmed the strong upregulation of GABPB1-AS1 in HepG2 cells transfected with the GABPB1-AS1 vector (Fig. 3A). Overexpression of GABPB1-AS1 had a slight effect on the expression of GABPB1 mRNA (Fig. 3B). However, the protein level of GABPB1 was inhibited in cells overexpressing GABPB1-AS1 (Fig. 3C), along with decreased PRDX5 protein levels ( Supplementary Fig. 3a).
We next designed siRNAs targeting GABPB1-AS1. While RT-qPCR results showed effective knockdown of GABPB1-AS1 using the siRNAs (Fig. 3D), no significant effect on the expression of GABPB1 mRNA was observed (Fig. 3E). However, western blot showed that GABPB1-AS1 knockdown resulted in increased GABPB1 protein levels (Fig. 3F). Since GABPB1-AS1-siRNA2 demonstrated the most efficient knockdown, this siRNA was used in subsequent experiments. To further elucidate the regulation of GABPB1 and GABPB1-AS1, we knocked down GABPB1 using two independent siRNAs. RT-qPCR results and western blot analysis showed effective knockdown of GABPB1 using the siRNAs (Fig. 3G,H). Knockdown of GABPB1 reduced PRDX5 protein levels ( Supplementary Fig. 3b). However, there was no effect on the expression of GABPB1-AS1 (Fig. 3I). GABPB1-siRNA1 demonstrated the most efficient knockdown, and this siRNA was used in subsequent experiments. These results indicate that GABPB1-AS1 regulates the expression of GABPB1 through a post-transcriptional mechanism.
We further investigated whether knockdown of GABPB1 lead to cell apoptosis and autophagy. Interestingly, results from Annexin V-FITC/PI double staining and activation of caspase-3 further verified the roles of GABPB1 in cell apoptosis ( Supplementary Fig. 3c,d). It has been shown that PRDX5 suppresses p53-dependent apoptosis 36 , and p53 can also enhance ferroptosis by inhibiting the expression of SLC7A11 (solute carrier family 7 member 11) 37 . We hypothesized that GABPB1 may participate in apoptosis by regulating PRDX5 and p53; however, whether the involvement of GABPB1 in ferroptosis is related to p53 needs to be further proven. It has been www.nature.com/scientificreports www.nature.com/scientificreports/ previously reported that GABP binds to the promoter regions of BECN1-PIK3C3 complex genes and activates their transcriptional activities. Knockdown of GABP reduced BECN1-PIK3C3 complex transcripts and reduced autophagy 38 . To investigate the effect of GABPB1 knockdown on HepG2 autophagy, AVOs (acidic vesicular organelles) were visualized using MDC staining, and western blot analysis was performed of microtubule-associated proteins 1 A/1B light chain 3 (LC3) II protein, the lipidated form of LC3 that is associated with the autophagosomal membrane. However, the results showed that knockdown of GABPB1 caused neither the formation of acidic vesicular organelles nor a change in LC3 II protein levels ( Supplementary Fig. 3e-g). Overall, knockdown of GABPB1 leads to cell apoptosis but not autophagy.

GABPB1-AS1 interacts with GABPB1 and suppresses GABPB1 translation.
Analysis of the GABPB1-AS1 and GABPB1 gene sequences revealed complementary fragments in the 812-1235 nt region of GABPB1-AS1 and the first exon (1-424 nt) of GABPB1 (Fig. 4A). To determine whether GABPB1 mRNA and GABPB1-AS1 overlap, we performed an RNase protection assay. The results showed protection of the GABPB1-AS1 and GABPB1 overlapping regions from RNase degradation and the formation of RNA duplexes (Fig. 4B). Furthermore, affinity pulldown of endogenous GABPB1-AS1 mRNA using in vitro transcribed biotin-labelled GABPB1 first exon showed that GABPB1 mRNA interacted with GABPB1-AS1 mRNA (Fig. 4C).
To further clarify how GABPB1-AS1 affects GABPB1 translation, polysomes were extracted from cells either depleted of GABPB1-AS1 or overexpressing GABPB1-AS1. RNA from the polysome and non-polysome fractions was extracted, and GABPB1 mRNA levels were evaluated by RT-qPCR. We found that the polysome-associated GABPB1 mRNA level was significantly increased in GABPB1-AS1 siRNA-transfected cells compared with controls, while the polysome-associated GABPB1 mRNA level was decreased in GABPB1-AS1-overexpressing cells compared with controls (Fig. 4D). These data indicated that GABPB1-AS1 suppressed GABPB1 translation by blocking the recruitment of GABPB1 mRNA to polysomes.
During translation, the eukaryotic initiation factor-4A (eIF4A) is required for the binding of mRNA to 40 S ribosomal subunits 39 . To further elucidate the functional mechanisms, we examined whether eIF4A binding with GABPB1 mRNA was inhibited in the regulation of GABPB1 translation by GABPB1-AS1. We performed anti-eIF4A RIP and found that the interaction between GABPB1 mRNA and eIF4A was significantly increased in cells depleted of GABPB1-AS1 and that GABPB1-AS1 overexpression reduced eIF4A binding to GABPB1 mRNA www.nature.com/scientificreports www.nature.com/scientificreports/ (Fig. 4E,F). Based on these results, we propose that GABPB1-AS1 suppresses GABPB1 translation by blocking GABPB1 mRNA recruitment to polysomes and binding with eIF4A.

Abnormal expression of GABPB1-AS1 affects ROS and MDA levels in HepG2 cells and GABPB1 expression in clinical tissues. Ferroptosis is triggered by inactivation of cellular GSH-dependent antiox-
idant defences, leading to the accumulation of toxic lipid ROS 40 . We next examined the role of GABPB1-AS1 on ROS and MDA levels in cells stimulated by erastin. We found that GABPB1-AS1 knockdown could partially reduce the ROS and MDA levels of erastin-treated cells and could inhibit cell death. Overexpression of GABPB1-AS1 increased ROS and MDA levels as well as promoted cell death (Fig. 5A,C-E). Moreover, western blot analysis showed that GABPB1 protein was increased after GABPB1-AS1 knockdown in erastin-treated cells, and overexpression of GABPB1-AS1 inhibited the protein expression of GABPB1 (Fig. 5B). These results demonstrated that overexpression of GABPB1-AS1 may increase ROS generation, leading to membrane lipid peroxide damage, which increases cell death. www.nature.com/scientificreports www.nature.com/scientificreports/ Cancer cells need higher antioxidant-producing capacity that enables them to survive oxidative stress 41,42 . We found that overexpression of GABPB1 could reduce ROS and MDA levels, which indicated that GABPB1 could enhance the cellular antioxidant capacity. We then investigated GABPB1 expression in human HCC tissues. Consistent with our assumption, the results showed higher GABPB1 expression in HCC tissues than in normal tissue 43 (Fig. 5F). Low levels of ROS can activate various signalling pathways to stimulate survival and cell proliferation 44 . Kaplan-Meier analysis further demonstrated that high GABPB1 mRNA levels in HCC tissues correlated with reduced overall survival 45 . Excess levels of ROS irreversibly damage cellular macromolecular components and result in cancer cell death 46 . Our results showed that HCC cell death induced by high GABPB1-AS1 levels correlated with increased ROS and MDA levels. Kaplan-Meier analysis also revealed that high GABPB1-AS1 levels in HCC tissues correlated with improved overall survival 47 . Interestingly, RT-qPCR results showed a positive correlation between the levels of GABPB1-AS1 and GABPB1 transcripts in 12 human HCC tissues (Fig. 5H). This may be due to the overlap between GABPB1 mRNA and GABPB1-AS1, which could enhances the stability of GABPB1 mRNA.

Discussion
Our study demonstrated that erastin upregulated the expression of the GABPB1-AS1 lncRNA and, consequently, inhibited the translation of GABPB1, which resulted in reduced expression of the PRDX5 peroxidase gene and subsequent rapid accumulation of ROS levels in HepG2 cells (Fig. 6A). Overexpression of GABPB1-AS1 had similar effects.
A key process in ferroptosis is the accumulation of hydroxyl radicals produced from hydrogen peroxide by the Fenton reaction, which initiates the lipid peroxidation process and results in membrane lipid damage 48 . Peroxidase is the key metabolic enzyme that controls the excessive accumulation of hydroxyl radicals 49,50 . Downregulation of peroxidase expression directly and negatively impacts the cellular antioxidative ability, which exerts an effect on the antioxidant capacity of hepatocellular carcinoma cells prior to that of reducing equivalent. Our work suggests that the GABPB1-AS1 lncRNA may play an important role in ferroptosis induced by erastin. In further support of the important role of GABPB1 and GABPB1-AS1 in HCC tumourigenesis, our data indicate that high expression levels of GABPB1 are positively correlated with poor prognosis of HCC patients, while high levels of GABPB1-AS1 correlated with improved overall survival.
In addition to the effects of erastin on inhibiting GABPB1 translation by upregulating GABPB1-AS1 lncRNA, we found that erastin could directly inhibit the transcription of GABPB1, leading to decreased mRNA levels. We speculate that GABPB1 may play a role in regulating the interaction between mitochondria and the nucleoplasm www.nature.com/scientificreports www.nature.com/scientificreports/