Cell lines and culture
In this research, the human normal liver cell line LO2, along with a range of liver cancer cell lines, specifically HepG-2, Hep3B, Huh7, Hepa1-6, H22, and SMMC-7721, were utilized. These cell lines, sourced from the Cell Bank of the Chinese Academy of Sciences in Shanghai with STR authentication, underwent cultivation in distinct mediums. The LO2 and H22 lines were propagated in RPMI-1640 medium (Gibco, Life Technologies, USA), in contrast to the HepG-2, Hep3B, Huh7, Hepa1-6, and SMMC-7721 lines, which were cultivated in Dulbecco’s Modified Eagle Medium (DMEM) medium (Gibco, Life Technologies, USA). Both media were supplemented with 10% v/v fetal bovine serum (FBS, Hyclone, USA) and 1% penicillin/streptomycin (Gibco, Life Technologies, USA). The cell cultures were maintained under controlled conditions at 37 °C within a humidified incubator, supplemented with a 5% CO2 atmosphere.
Liver cancer cells sorting
In this experiment, Liver cancer cells were initially incubated with FITC-conjugated anti-CD44 mouse antibodies, procured from Stem Cell Technologies, USA. After this step, the cells were intricately labeled with dextran-coated magnetic nanoparticles, a process facilitated by the employment of bispecific Tetrameric Antibody Complexes (TAC) [37]. Following this, immuno-magnetic separation was conducted, leading to the successful isolation and definitive identification of CD44-positive cells, unequivocally characterized as purified cancer stem cells (CSCs) [38].
Cell transfection and reagents
ALKBH5 and CIITA overexpression and knockdown plasmids were sourced from Gene Corporation in Suzhou, China. The full-length open reading frames (ORFs) of ALKBH5 (NCBI accession NM_001363894.1) and CIITA (NCBI accession NM_003394.4) were inserted into the pcDNA3.1 eukaryotic expression vector to develop their expression vectors. Additionally, knockdown constructs for ALKBH5 (sense sequence: 5’-GGGUCAUUACAGAGAUUAAAU-3’; antisense sequence: 5’-UUAAUCUCUGUAAUGACCCUG-3’) and CIITA (sense sequence: 5’-CGAAUAGACUAAGAUGAAAUG-3’; antisense sequence: 5’-UUUCAUCUUAGUCUAUUCGAG-3’) were constructed and cloned into the pLVX-shRNA vector. In the experimental procedures, 4 μg of these plasmids were transfected into liver cancer CSC cell cultures using Lipofectamine® 2000 (Thermo Fisher Scientific, Inc.) and were subsequently incubated for 24 h in six-well plates. For comparison, control cells were treated with an empty vector.
In vivo tumorigenicity assay
Six-week-old male C57BL/6 mice, sourced from Beijing Huafukang Biotechnology Co., Ltd. (license SCXK 2014-0004, Beijing, China), were housed in a controlled barrier facility. They were allocated to ventilated cages, with no more than four mice per cage, and provided with sterile hardwood bedding. The facility maintained a specific pathogen-free environment at a stable temperature range of 22–26 °C, complemented by a 12-h light-dark cycle. The mice had continuous access to sterile pellet food and autoclaved water. After a seven-day acclimatization period, animals were randomly grouped into four groups: control (Model) group, radiotherapy (Radio) group, anti-PD-1 (PD-1) group and Radio+PD-1 group (n = 5 per group). Hepa1-6 cells (1 × 106 cells per mouse) were subcutaneously implanted into the right axilla of each mouse. Mice in the Radio treatment groups underwent irradiation at 8 Gy per day for three consecutive days. Meanwhile, the mice allocated to the PD-1 treatment groups received tail vein injections of camrelizumab with a frequency of once every three days. Tumor volume was calculated using the formula (volume = width × width × length/2). After a period of 28 days, the mice were euthanized through cervical dislocation following anesthesia with isoflurane. The researcher responsible for outcome measurement was blinded during the trial. This study has been approved by the Animal Research Committee at Mongolia Medical University, with ethical approval number YKD202002012, and all experimental protocols were conducted in accordance with institutional guidelines. All experiments adhered to the ARRIVE guidelines. In this experiment, hepatocellular carcinoma cells were cultured in DMEM supplemented with 10% v/v inactivated calf serum. The cells were incubated at a constant temperature of 37 °C in an atmosphere containing 5% CO2 and with saturated humidity. The culture medium was replaced at 48-h intervals. During their logarithmic growth phase, cells were treated with 0.25% trypsin for one minute to facilitate detachment. After trypsin removal, the cells were resuspended in a small volume of the same medium to ensure a homogeneous mixture. The cell concentration was then determined using a hemocytometer and adjusted to 1 × 106 cells/mL. Mice were subjected to isoflurane-inhalation anesthesia. Once anesthesia was confirmed effective, the cell suspension was carefully injected into the thoracic cavity beneath the liver capsule of the mice [39].
Reverse transcription-quantitative PCR (RT‒qPCR)
Extraction of total RNA was accomplished utilizing the EasyPure RNA Kit (Cat# H30828; TransGen Biotech, Beijing, China), followed by the generation of single-stranded cDNA from the extracted RNA. This process was facilitated using oligo (dT) priming in accordance with the Thermoscript RT kit protocol (Invitrogen, Carlsbad, CA, USA) and employing the iQ™5 Real-Time PCR System (Bio-Rad Laboratories, Inc., USA). The RT-qPCR reactions were conducted in strict compliance with the manufacturer’s instructions. The thermocycling parameters comprised an initial denaturation at 95 °C for 30 s, followed by 40 sequential cycles of 95 °C for 15 s and 60 °C for 30 s each. The assessment of mRNA relative expression levels utilized the 2-ΔΔCt method (cited as reference [16]), with a normalization baseline set against GAPDH mRNA levels. To ensure reproducibility, each assay was conducted in triplicate and independently replicated three times. The quantitative analysis of the data was executed using the relative standard curve method, normalizing gene expression to GAPDH mRNA levels. The experiment incorporated primer sequences synthesized by Sangon (Shanghai, China).
RNA stability detection
In this investigation, 40,000 CSCs were cultivated in a six-well plate configuration over a period of 24 hours. Cells from the initial well were isolated and stored at −80 °C in 1 ml of TRI Reagent for subsequent analysis. The cells in the five remaining wells were exposed to actinomycin D (Act-D; concentration: 10 μg/mL; Sigma-Aldrich, USA). Following Act-D exposure, cell samples were systematically collected at intervals of 1, 2, 4, 6, and 8 hours, each preserved in 1 ml of TRI Reagent as per the established protocol. The extraction of total RNA was followed by its reverse transcription into cDNA, in accordance with documented methods [40]. RNA expression from the initial well was employed as a baseline for normalization. The primer sequences used for amplifying ALKBH5 pre-mRNA were as follows: forward 5’-TGAGCTCGGTGAGAGCAAAG-3’ and reverse 5’-TTAAACCGTGGGGAGACTGC-3’. For ALKBH5 mature-mRNA, the forward primer was 5’-GTGAGCGAGACCCCACTATG-3’, and the reverse primer was 5’-CACATAGCAGCACCAGTGGA-3’.
Total m6A measurement
Cells from each experimental group were subjected to a series of post-drug treatment procedures, beginning with the aspiration of the existing culture medium and followed by a single wash using 1× PBS. This was followed by the administration of 150 μL of reporter gene cell lysate into the wells. The cells were then lysed and centrifuged at 12,000 rpm for five minutes, a step crucial for the extraction of the supernatant for assay purposes. Subsequently, total RNA was extracted using Trizol reagent. This RNA underwent a selective enrichment process for polyA mRNA via Oligo-dT magnetic beads, isolating mRNA with polyA tails. This enriched mRNA was then fragmented into segments of 50–100 nucleotides using a specific fragmentation agent. The fragmented RNA was split into two equal portions. The first portion was subjected to an additional enrichment process for m6A-methylated mRNA fragments, employing m6A antibody-coated immunomagnetic beads. The second portion acted as a control, utilized in the direct construction of a standard transcriptome sequencing library. Following the m6A-modified mRNA fragment enrichment and recovery, a conventional sequencing library was compiled, adhering to the protocol for transcriptome library construction. Finally, both the m6A-seq and RNA-seq libraries underwent partial high-throughput sequencing on a Hiseq 2500 platform, employing the PE100 sequencing modality.
MeRIP-qPCR
To explore the methylation dynamics within RNA on a transcriptome-wide scale, the selective binding of antibodies to methylation-modified bases is employed. This technique involves the use of RNA immunoprecipitation to enrich methylation-modified fragments as the foundation, followed by high-throughput sequencing, ensuring the efficient acquisition of results. In the experimental design, cellular models closely related to the disease are utilized to establish knockdown, knockout, or overexpression cell lines of RNA modification-related factor ALKBH5, collectively referred to as the knockdown/knockout/overexpression group. Each group is compared to wild-type cells, with each sample containing 5 × 10 or more cells. Alternatively, additional samples can be utilized, each containing 5 × 106 to 107 cells. Additionally, a comparative analysis is conducted between normal human cells/tissues and diseased counterparts, constituting two distinct groups. Each group comprises a minimum of three samples, with each sample containing 5 × 106 to 107 cells or 10–20 mg of tissue.
MTT assay
Cellular viability was rigorously assessed post-treatment via the MTT assay. Cells were incubated for 30 min in an MTT solution, standardized to a 5 mg/ml concentration. This incubation facilitated the formation of dark blue formazan crystals in viable cells, which were then solubilized using ethanol (100% concentration) as a lysis agent. Spectrophotometric analysis of the resultant mixture was conducted in the 540 to 595 nm wavelength range, utilizing a microplate reader from Molecular Devices (Sunnyvale, CA, USA). Comparative data analysis was performed, with results expressed as percentages relative to a baseline established by the control group of untreated cells.
Flow cytometry
In this study, single-cell suspensions isolated from both neoplastic and adjacent benign tissues were methodically prepared as detailed in the preceding sections. These cells underwent a preliminary incubation with FcBlock (clone 93, Biolegend) for 15 min, subsequently followed by a 45-minute incubation at 4 °C with fluorophore-conjugated antibodies. Post-incubation, the cells were subjected to a thorough washing in a phosphate-buffered saline (PBS) solution augmented with 0.5% Bovine Serum Albumin (BSA) and 2 mM Ethylenediaminetetraacetic Acid (EDTA), culminating in their analysis via a BD LSRII flow cytometer. The exclusion of non-viable cells was accomplished using 7-Aminoactinomycin D (7AAD), and cell doublets were efficiently segregated and removed based on their distinctive forward and side scatter profiles.
Apoptosis analysis
Quantitative assessment of cellular apoptosis was conducted using an Annexin V-FITC/PI Apoptosis Detection Kit (Hangzhou Lianke Biotechnology Co., Ltd., China), conforming rigorously to the manufacturer’s specified protocols. For this assay, precisely 40,000 cells were suspended in a solution consisting of 5 μL Annexin V-FITC and 10 μL Propidium Iodide (PI). This suspension underwent a 15-min incubation in an environment shielded from light, at standard room temperature. The analysis, executed within a critical timeframe of 60 min post-incubation, was performed using a BD FACSCantoII™ flow cytometer (BD Biosciences), ensuring the accuracy and reliability of the apoptotic rate measurements.
Cell-cycle analysis
In this study, a precise quantity of 10,000 cells was subjected to fixation in 75% ethanol, maintained at a constant temperature of 4 °C throughout the night. Post-fixation, these cells were subjected to centrifugal separation, followed by a single PBS wash. They were then resuspended in 1 milliliter of DNA Staining Solution (sourced from Hangzhou Lianke Biotechnology Co., Ltd., China) and incubated for 30 min in an environment devoid of light. The final stage entailed the assessment of the cell-cycle distribution of the treated cells, employing a Beckman Coulter flow cytometer for this purpose.
TUNEL assay
Quantitative assessment of apoptosis in tendon tissues was conducted using a TUNEL assay (Boster, China), in strict compliance with the manufacturer’s detailed instructions. The protocol commenced with the fixation of cryopreserved tissue sections in 4% paraformaldehyde for 60 minutes. This was followed by a 10-min proteolytic treatment with proteinase K. Subsequently, the sections underwent a sequential incubation: initially with terminal deoxynucleotidyl transferase (TdT), and then with biotinylated deoxyuridine triphosphate (BIO-dUTP). Post-incubation, application of a blocking solution preceded the introduction of streptavidin-biotin complex (SABC) to facilitate fluorescent tagging. The final step involved the detailed visualization of these fluorescently labeled sections utilizing a laser confocal microscope (Olympus, Japan).
Colony formation assay
Investigating anchorage-independent growth, a suspension of 1 × 104 cells/ml was established in 2 ml of 0.3% agar. This medium was augmented with 1% N2 Supplement, 2% B27 Supplement, 20 ng/ml human platelet growth factor, 100 ng/ml epidermal growth factor, and 1% antibiotic-antimycotic, all procured from Invitrogen. The suspension was carefully layered over a 0.5% agar base in 6-well plates. After an incubation period of 21 days, the resultant colonies were stained with 2% crystal violet. Any colony exceeding 0.2 mm in diameter was meticulously counted under a microscope.
Hematoxylin-eosin (H&E) staining
Procured for detailed histopathological scrutiny, allograft tumor specimens were initially fixed in a 4% paraformaldehyde solution and subsequently encased in paraffin. Thin sections, precisely 5 μm in thickness, were methodically stained employing H&E. The intricate examination of areas harboring cancer emboli was conducted using the high-resolution capabilities of an Olympus BX51 microscope.
Immunofluorescence staining
In this experiment, we first permeabilized the tissue sections using a 0.2% Triton X-100 solution for a period of 15 min. This was followed by a blocking stage, where 3% Bovine Serum Albumin (BSA) was applied for an hour. For primary antibody incubation, the sections were maintained overnight at 4 °C with a series of antibodies: anti-CD3 (ab135372, Abcam, UK) at 1:50 dilution, anti-CD8 (53-6.7, Novusbio, USA) also at 1:100 dilution, and anti-CD4 (ab288724, Abcam, UK) at 1:50 dilution. Following this, the sections were cleansed with PBS and subsequently exposed to secondary antibodies (ab150077, ab150167, or ab150157, Abcam, UK) at a 1:500 dilution for a duration of two hours at ambient temperature. Finally, nuclear staining was executed using DAPI (Solarbio), and the resultant images were captured utilizing an Olympus BX51 fluorescence microscope (Olympus Corporation).
Clinical samples
Five pairs of liver cancer tissues and matched adjacent normal tissues were collected from patients with liver cancer at Affiliated Hospital of Inner Mongolia Medical University Hospital. The study protocol was approved by the Ethical Committee of Medical Ethics Committee of Inner Mongolia Medical University (approval number YKD202201227), and informed consent was obtained from all participants.
Immunohistochemistry staining
Tumor tissues and matched adjacent normal tissues were fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned into 5 μm thick slices. After dewaxing, hydration, and antigen retrieval, the sections were blocked with normal goat serum. They were then incubated overnight at 4 °C with the following primary antibodies: anti-ALKBH5 (16837-1-AP, Proteintech, China) and anti-CIITA (PA5-21031, Thermo Fisher, USA). Following primary antibody incubation, the sections were treated with corresponding secondary antibodies for 20 min at room temperature. Finally, observation was carried out using an Olympus BX51 high-resolution optical microscope.
Western blotting assay
The analysis of protein levels in tumor tissues and cells involved lysis using RIPA buffer (Thermo Fisher Scientific, Inc., USA), enhanced with 1% protease inhibitors (Pierce). Protein concentrations were subsequently measured using a BCA protein assay kit (Thermo Fisher Scientific, MA, USA). For each sample, 30 mg of protein was loaded onto PVDF membranes. These membranes were first incubated with primary antibodies at 4 °C overnight, then with HRP-conjugated secondary antibodies at room temperature for 2 h. Detailed antibody specifications are provided in the Supplementary Material (Table S1). Protein signal detection utilized the enhanced chemiluminescence (ECL) reagent (Millipore, Bedford, MA, USA) and analysis was conducted using NIH ImageJ software (Scion, Frederick, MD), with data normalization against glyceraldehyde-3-phosphate dehydrogenase (GAPDH).
Bioinformatics assay
This research leveraged publicly accessible datasets, specifically miRNA and mRNA expression profiles from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA). The acquisition of these datasets was facilitated through their respective official websites, including GEO, TCGA, and StarBase. Analysis of the collected data was meticulously conducted using the GEO2R tool [41]. In the realm of survival analysis, we utilized methodologies provided by the GEPIA website and Kaplan-Meier Plotter [42, 43]. Furthermore, the study engaged in Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, deploying the advanced capabilities of GSEA software.
Statistical analysis
Data compilation employed a Microsoft Excel worksheet, while SPSS version 23.0 underpinned the statistical analysis. Categorical data were delineated as rates or composition ratios. Statistical comparisons hinged on specific thresholds: (1) Pearson’s χ2 test for N ≥ 40 and T ≥ 5, (2) a corrected χ2 test for N ≥ 40 with 1 < T < 5, and (3) Fisher’s exact probability method for N < 40 or T in the range of 1 to 5. The nonparametric rank sum test facilitated comparison of categorical data across different groups. To discern factors influencing tumor marker positivity, an unconditional logistic regression model was deployed. The Kaplan–Meier method served for initial univariate prognosis assessment in liver cancer patients, followed by a Cox regression model for multifaceted prognostic factor analysis, adhering to an α threshold of 0.05. For in vitro and in vivo experiments, ANOVA was employed to assess statistical significance across three or more groups, while the Student’s t-test was utilized for comparisons between two groups. Data are presented as mean values ± standard deviation. Differences were deemed statistically significant when p < 0.05.