Cell lines and culture

The SiHa and CaSki cell lines, both containing HPV16 integration, were procured from the American Type Culture Collection (ATCC). The HaCaT cell line, an HPV16 non-integrated immortalized human cervical keratinocyte cell line, was obtained from the China Center for Type Culture Collection (CCTCC) and maintained according to standard protocols. The S12 cell line, an HPV16-integrated immortalized human cervical keratinocyte cell line, was graciously provided by Prof. Hu Zheng (Precision Medicine Institute, the First Affiliated Hospital, Sun Yat-sen University, Guangzhou, China) [3], with permission from the original owner, Prof. Margaret Stanley (Department of Pathology, University of Cambridge, Cambridge, United Kingdom) [18]. S12 cells were cultured in a 1:3 mixture of DMEM and Ham’s F-12 medium supplemented with 5% FBS, 8.4 ng/ml cholera toxin, 5 μg/ml insulin, 0.5 μg/ml hydrocortisone, 24.3 μg/ml adenine, and 10 ng/ml epidermal growth factor. Further details regarding cell line authentication are available in the supplementary materials.

Clinical specimens

Clinical specimens were collected from the Department of Gynecological Oncology at the First Affiliated Hospital of Guangzhou Medical University, comprising a total of 100 cervical samples. The specimens included 20 normal cervix, 20 cervical intraepithelial neoplasia (CIN) I, 30 CIN II-III, and 30 CSCC tissues. Lesion samples were sourced from cervical biopsies (8 cases), loop electrosurgical excision procedures (22 cases), cone biopsies (25 cases), and radical hysterectomies (25 cases). Additionally, 20 normal cervical tissue specimens were obtained from patients undergoing hysterectomy for non-cancerous conditions, with no history of CIN or abnormal Pap smears. Pregnant women, individuals with acute pelvic infections, and those who had undergone radiotherapy/chemotherapy were excluded from the study. The study involving human tissue samples obtained ethical approval from the Ethics Committee of the First Affiliated Hospital of Guangzhou Medical University (approval number: 2022-58, granted on 14 April 2022) and adhered to the principles outlined in the Declaration of Helsinki. Informed consent was waived due to the use of residual specimens and routine medical records. Each sample was assessed by two experienced pathologists.

High­throughput viral integration detection (HIVID)

HPV16 integration was analyzed using the HIVID method, as presented in Table 1. HIVID is a next­generation sequencing and computational method developed by our collaborative research group [19]. While primarily used for HBV integration, HIVID has also been successfully applied to detect integration of other viruses, including HPV, as demonstrated in our previous study [3]. The detailed experimental procedure has been previously documented in our previous study. Briefly, the methodology encompassed the design of sequence-capture probes targeting 17 distinct HPV genome sequences (6, 11, 16, 18, 31, 33, 35, 39, 45, 52, 56, 58, 59, 66, 68, 69, and 82) supplied by MyGenostics. The HIVID pipeline was deployed for breakpoint identification, with RNA-seq employed for precise breakpoint localization. Subsequent PCR amplification was facilitated using the GeneAmp PCR System 9700 thermal cycler, with Sanger sequencing conducted on the Applied Biosystems 3730 × DNA analyzer (Life Technologies Inc.). The identified integrated breakpoints were annotated via ANNOVAR [20].

Table 1 HPV16 integration, c-Myc and SLC7A11 expression in different cervical tissuesRNA extraction and RT-qPCR

RNA extraction from cells was performed using the miRNeasy Mini Kit (Qiagen), following the manufacturer’s guidelines. For miRNA, reverse transcription adopted the Mir-X™ miRNA First-Strand Synthesis Kit (TaKaRa), while for mRNA, the PrimeScript™ RT Master Mix (TaKaRa) was employed. RT-qPCR was performed as previously described [21]. Primer sets for miR-142-5p and U6 were sourced from RiboBio Inc. [16]. Expression levels of both miRNAs and mRNAs were standardized against U6 and GAPDH, respectively. Refer to Supplemental Table 1 for the primer sequences.

Western blot analysis

Western blot analysis was conducted following established protocols [21]. Primary antibodies included anti-SLC7A11 (#ab37185, Abcam), anti-HOXA5 (#ab140636, Abcam), and anti-GAPDH (#60004-1-Ig, Proteintech). Secondary antibodies applied were horseradish peroxidase-conjugated anti-rabbit (#ab6721, Abcam) and anti-mouse (#ab6789, Abcam) immunoglobulin-G. In order to avoid interference caused by simultaneous exposure of different protein molecules with different expression levels, we performed complete cleavage and exposure analysis of the bands based on different molecular weights.

Dual luciferase assays

To explore the interaction between c-Myc and the miR-142-5p promoter, the coding region of c-Myc wild sequence (WT) or its mutated biding sequence (MT), and the 2-kb region upstream of the miR-142-5p transcription start site were PCR amplified and inserted into pcDNA3.1 (+) and PGL3 vectors, respectively. For assessing miR-142-5p targeting of HOXA5, the potential miR-142-5p complementary site in the 3′-UTR of HOXA5 or its mutant sequence was cloned into the pmiR-RB-Report vector (RiboBio Inc.). Subsequently, pmiR-RB-Report-HOXA5-3′UTR-WT or pmiR-RB-Report- HOXA5-3′UTR-MT were co-transfected into S12/SiHa/CaSki cells with miR-142-5p mimics. To investigate HOXA5 targeting of SLC7A11, SLC7A11 promoter plasmids with firefly luciferase reporters were co-transfected with an internal control pRL-TK containing a full-length Renilla luciferase gene (GeneChem Inc.) into S12/SiHa/CaSki cells overexpressing HOXA5, following the manufacturer’s guidelines. After 48 h post-transfection, the cells were analyzed using the Dual-Luciferase Reporter Assay System (Promega). Firefly luciferase activity was normalized to Renilla luciferase activity for each well. All experiments were performed in triplicate and repeated thrice. The cloning sequences can be found in Supplemental Table 2.

Chromatin immunoprecipitation (ChIP) assay

Cells were fixed using 1% formaldehyde, quenched with glycine, and then subjected to a ChIP assay using an Enzymatic ChIP Kit (#9003, CST) following the provided protocols [16]. Immunoprecipitation was carried out with anti-HOXA5 (#sc-365784X, SantaCruz) and control IgG (#2729, CST) antibodies. The enrichment of HOXA5-binding sites (HBS) within the SLC7A11 promoter region was assessed through qPCR. The results were presented as the relative enrichment normalized to control IgG. Detailed primer sequences utilized for ChIP-PCR are listed in Supplemental Table 1.

miRNA target prediction

The prediction and analysis of miRNA targets were performed utilizing the algorithms TargetScan (http://www.targetscan.org/), PicTar (http://pictar.bio.nyu.edu/), and miRWalk (http://zmf.umm.uni-heidelberg.de/apps/zmf/mirwalk2/) [16].

In situ hybridization (ISH) and immunohistochemical (IHC) staining

ISH was conducted on cervical tissues utilizing an ISH kit and miR-142-5p synthetic oligonucleotide probes (Boster Biological Technology Co., Ltd) according to the manufacturer’s instructions [16]. Briefly, slides were processed with 3% H2O2, digested with pepsin, and fixed with 1% PFA in DEPC. Subsequently, the slides were pre-hybridized with hybridization buffer at 42 °C using miR-142-5p or U6 probes, followed by incubation with streptavidin–biotin complex and horseradish peroxidase polymer. For IHC staining [21], sections were treated with 3% H2O2 to block endogenous peroxidase activity, and incubated with primary antibodies overnight at 4 °C. A horseradish peroxidase-conjugated anti-rabbit secondary antibody was applied for 1 h at 26 °C. The expression of c-Myc (#ab32072, Abcam), SLC7A11 (#ab37185, Abcam), and HOXA5 (#ab140636, Abcam) was visualized using DAB and counterstained with hematoxylin. ISH and IHC tissue sections underwent independent evaluation and scoring by two experienced pathologists. The H-score algorithm was applied as previously detailed [21]. The median H-score values were used to categorize groups with low and high expression of c-Myc or SLC7A11.

Transient transfection with oligonucleotides and plasmids

Mimic/inhibitor chemically modified oligonucleotides were utilized to imitate or suppress endogenous miRNAs. The miR-142-5p mimic/inhibitor and corresponding negative controls were designed and synthesized by RiboBio Inc. Transfection with the miR-142-5p mimic or inhibitor was employed to regulate the levels of miR-142-5p. The HOXA5 coding sequence (without the 3′-UTR) was inserted into the pCDNA3.1 (+) vector (RiboBio Inc.). An empty vector served as the control. Cells were assessed for RNA extraction, western blot analysis, and in vitro assays 48 h after transfection.

Lipid reactive oxygen species (ROS) measurements

Cellular lipid ROS levels were assessed using a ROS Assay Kit (#S0033S, Beyotime) as per the manufacturer’s protocol [22]. Cells were plated in a 6-well plate and treated with serum-free medium containing 10 μmol/L 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) in the absence of light for 20 min at 37 °C with periodic agitation. Subsequently, cells were collected, rinsed with serum-free medium, and re-suspended in serum-free medium. Measurement of fluorescence intensity at an excitation wavelength of 488 nm and an emission wavelength of 525 nm allowed for the evaluation of intracellular ROS levels. Serum-free medium was utilized to prevent potential interference from endogenous esterase activity in serum [23].

Malondialdehyde (MDA) assay

The MDA content in cell lysates was quantified using a Lipid Peroxidation MDA Assay Kit (#S0131S, Beyotime) following the manufacturer’s protocol [24]. Cells were harvested and lysed with cell RIPA Lysis Buffer (#KGP702, KeyGEN). The resulting supernatant, obtained post-centrifugation at 10,000–12,000 g for 10 min, was utilized for the analysis. TBA was added to the samples to form MDA-TBA adducts, and the relative MDA concentration was determined colorimetrically at an absorbance of 535 nm.

Transmission electron microscopy (TEM)

To prepare for electron microscopy, cells were first fixed with 2% glutaraldehyde for 5 min [25]. Subsequently, postfixation was performed using osmium tetroxide, followed by dehydration with ethanol, propylene oxide treatment, and embedding in Spurr’s epoxy resin. Thin sections of 90 nm were stained with uranyl acetate and lead citrate before being examined under an H-7500 transmission electron microscope (Hitachi) at a magnification of 40,000 ×. Ferroptotic cells typically display an increase in mitochondrial membrane density along with a reduction or loss of cristae and outer mitochondrial membranes [26].

Iron assay

The levels of intracellular ferrous iron (Fe2 +) were assessed using the iron assay kit (ab83366, Abcom) as per the instructions provided by the manufacturer [24]. Samples were gathered, rinsed with chilled PBS, and homogenized in 5 × volumes of iron assay buffer on ice. The resulting supernatant was then treated with iron reducer and iron probe, and each sample was subsequently incubated. The measurement was taken using a colorimetric microplate reader, reading the output at an optical density of 593 nm.

Colony formation assay

For the colony formation assay [27], S12 and SiHa cells were plated in 60 mm dishes. Upon reaching 70–80% confluence, the cells were either transduced with lentivirus carrying c-Myc overexpression or exposed to the c-Myc antagomir (IZCZ-3). Following trypsinization and cell counting, the cells were re-seeded at appropriate dilutions to allow for colony formation. After an incubation period of 10–14 days, the formed colonies were fixed, stained with crystal violet, washed, and air-dried. The plating efficiency was calculated for each cell line, and the survival rate was determined based on the number of colonies formed post-treatment. Each experiment was conducted in triplicate.

Establishment of mCherry overexpression SiHa cells

To establish SiHa cells overexpressing mCherry, lentiviral vector containing the mCherry sequence (lenti-mCherry) was obtained from GeneChem Inc. SiHa cells were transfected with lenti-mCherry as per the manufacturer’s protocol [25]. Purinomycin (2 μg/ml) was added 48 h after virus transfection for selection. The culture medium with purinomycin was refreshed every 2–3 days over a 4-week period. The transfection efficiency was subsequently assessed using a fluorescence microscope.

In vivo xenograft model

Six-week-old immunodeficient Female nude mice were obtained from Guangdong Animal Center (Guangzhou, China) and randomly divided into four groups (n = 3 mice/group): (a) NC group; (b) IZCZ-3 group; (c) erastin group; and (d) IZCZ-3 + erastin group. To establish subcutaneous tumors in mice, SiHa-mCherry cells (5 × 106 cells/mouse) were subcutaneously injected into the right posterior flanks of nude mice without matrigel. Tumor size was recorded every 4 days, and tumor volume was calculated using the formula: volume (mm3) = 1/2 (length × width2). When the tumor volume reached approximately 50 mm3, mice were intraperitoneally treated with 10 mg/kg IZCZ-3 (MCE, #HY-111411) and/or 10 mg/kg erastin (MCE, #HY-15763) daily for 20 days. IZCZ-3 and erastin were dissolved in a solution of 5% dimethyl sulfoxide (DMSO) and corn oil. Animal experiments were approved by the Committee Review of Animal Experiments in Guangzhou Medical University (approved number: 2019-081, on 26 February 2019) and were performed in accordance with the Basel Declaration.

Statistical analysis

Statistical analysis was performed using SPSS software (version 20.0). The results are presented as the mean ± SEM values and were analyzed using t-tests or one-way ANOVA. Frequency tables were analyzed using the chi-squared test, while the Pearson correlation coefficient was employed to evaluate the significance of links between categorical variables. Statistical significance was defined as a P-value below 0.05.