Cell culture and lines
We obtained the SCC9, SCC15, and CAL27 OSCC cell lines from the American Type Culture Collection (ATCC, USA). Both SCC4 cell lines and human oral epithelial cell (HOEC) were purchased from the China Center for Type Culture Collection. Previous work established the establishment of OSCC cell lines that are radioresistant (SCC9-RR and CAL27-RR) [13]. All cell lines were checked for mycoplasma contamination through short tandem repeat (STR) profiling before any studies were performed. Cells were cultured in a 37°C, 5% CO2 incubator with DMEM /F12 medium (Gibco, USA) supplemented with 10% fetal bovine serum (FBS), streptomycin (100 µg/mL), and penicillin (100 µg/mL).
Magnetic cell sorting
Cell suspensions were treated in a sorting buffer, which contained MACS Buffer A (autoMACS® Rinsing Solution) and Buffer B (MACS® BSA Stock Solution) in a 20:1 ratio. Centrifuge cell suspensions at 1000g for 10 min. Add 20 µL of CD44 Microbeads (MiltenyiBiotec, Bergisch Galdbach, Germany) and 80 µL of sorting buffer per 1×107 total cells. Mix well and incubated at 4℃ for 15 min in the dark, followed by MACS buffer washes and centrifugation for 10 min at 1000 rpm. Cells were suspended in 500 µL of sorting buffer per 1×107 total cells. CD44(+) cells and CD44(-) cells were dissociated using a magnetic cell sorting device (MiltenyiBiotec, Bergisch Galdbach, Germany). Briefly, CD44 microbead stain was then applied before they were run through an LS magnetic column, CD44(+) cells were eluted when the magnet was taken away from the column.
Irradiation
Via a medical linear accelerator (Varian, MA, USA), CD44(+)-OSCC cells or CD44(-)-OSCC cells were irradiated at varying doses (Gy) under the specific conditions: Dose rate = 2 Gy/min, SSD = 100 cm, irradiation field = 25 cm x 25 cm, and vertical irradiation. The cell culture flask was covered with a 5 cm dose compensation plates and placed in the middle of the irradiation field.
Sequencing and bioinformatic analysis
Sequencing was performed to analyze DEGs of parental and radioresistant cells, with analysis of significance using the EdgeR package at a cutoff of |log2(fold change)| >1 and P < 0.05. Heatmap was used to represent genes showing a significant change and those associated with the maintenance of CSCs phenotype. The database of Cancer Genome Atlas (TCGA) (portal.gdc.cancer.gov) of 330 tumors and 32 normal samples were used to obtain OSCC mRNA sequencing data. Moreover, METTL3-linked patient survival was evaluated via the GEPIA platform (gepia.cancer-pku.cn). The UALCAN database (ualcan.path.uab.edu) was approached for analyzing the lymph node and tumor metastasis. The correlation between CSCs-related genes and SALL4 mRNA in OSCC was evaluated using the GEO data set (www.ncbi.nlm.nih.gov/geo).
Quantitative Real-Time PCR
All OSCC cell lines, as well as CD44(+)-OSCC and CD44(-)-OSCC cells, total RNA was extracted with the use of the TaKaRa MiniBEST Universal RNA Extraction kit (Cat#9767, TaKaRa, Inc., Otsu, Japan). The Epoch™ spectrophotometer (BIO-TEK, Vermont, USA) was used to measure the concentration and purity of the RNA. Following the manufacturer's instructions, cDNA was synthesized via a PrimeScript™ RT Master Mix (Perfect Real Time) kit (Cat# RR036A, TaKaRa, Inc.; Otsu, Japan). For quantitative real-time PCR, we utilized the TB Green Premix Ex Taq™ II kit (RR820A, TaKaRa, Inc.; Otsu, Japan). The 7500 Real-Time PCR System (Applied Biosystems; Carlsbad, CA, USA) was employed for gene expression detection. The relative expression of target genes was determined via the 2−△△CT method, with GAPDH serving as an internal reference. Sangon Biotechnology Inc. (Shanghai, China) produced the primers, which are listed in Supplemental Table 1.
Western blotting
Radioimmunoprecipitation assay (RIPA) buffer (Beyotime; Shanghai, China) containing protease and phosphatase inhibitors was utilized for extracting total protein from each sample. The BCA protein assay kit (Beyotime, Shanghai, China) was approached for determining the protein concentration in the samples. SDS-PAGE (4–12%) was employed for separating 20 µg of proteins in each lane, and the separated proteins were then transferred to PVDF membranes (Millipore; Billerica, MA, USA). The membranes were then incubated with the primary antibody (Supplemental Table 2) overnight after being blocked in 5% skim milk in TBST for 1 hour. Proteins were detected via a Chemi-DocTMXRS+Image Lab™ system (BIO-RAD; Thermo, USA) after incubation with goat anti-rabbit IgG H&L (HRP) (Abcam, Cambridge, UK, ab6721, 1:5000) secondary antibody at 37°C for 1 hour. Protein bands were analyzed with enhanced chemiluminescence (ECL) detection reagent.
Knockdown or overexpression of target genes
Cells were transfected with plasmids and siRNAs from Sangon Biotechnology Inc. (Shanghai, China) via Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's procedure. qRT-PCR and western blotting were employed for analyzing the level of transfection success. After 48 and 72 hours of transfection, mRNA and protein expression levels were measured. The siRNA primers tested in this work are detailed in Supplement Table 1. Stably transfected cell lines with METTL3 knockdown and SALL4 overexpression were constructed by Hanheng Biological Company (Shanghai, China).
Sphere formation assay
Spheroid media, which consisted of DMEM/F12 (Gibco, USA), 2×N2 supplements (Gibco, USA), 1×B27® supplements (Gibco, USA), insulin (5 µg/mL; Invitrogen, USA), bFGF (20 ng/mL; PeproTech, Rocky Hill, NJ, USA), and EGF (20 ng/mL; PeproTech, Cranbury, NJ, USA), were employed for cell resuspension. Cells were seeded at a density of 1000 cells per well in 6-well ultra-low attachment culture plates (PlatesCorning Inc., Corning, NY), The medium was added with fresh medium every 5, and then photographed using a microscope after 14 days of constant incubation in the same conditions.
CD44 + /ALDH1 + cell sorting
This experiment was conducted by setting up the following groups: blank control, CD44(PE), ALDH1(FITC), and CD44/ALDH1 double-stained; 1×106 cells/mL were used. Since ALDH1 is an intracellular protein, membrane breaking was required for staining. First, the cells were centrifuged for 10 min at 1000 rpm before an incubation period of 10 min at room temperature in a 500 µL pre-cooled fixative. Subsequently, an appropriate amount of membrane-breaking reagent was added, and the samples centrifuged for 5 min at 1000 rpm before an incubation period of 15 min at room temperature. Eventually, samples could be incubated with FITC-ALDH1A1 (Abcam, ab275646) and PE-CD44 (BioLegend, cat#338808) fluorescent antibodies at 4℃ for 30 min and detected using flow cytometry.
MeRIP-qPCR assay
The m6A immunoprecipitation qPCR (MeRIP-qPCR) was performed using GenSeq®m6A MeRIP Kit (Cat No. GS-ET-001, Cloud-Seq Biotech, Shanghai, China). Briefly, RNA (1 µg/µL) was fragmented to a size of nearly 200 nt, and a portion of fragmented RNA (1−3 µg) was used as the input group, whereas the rest of the fragmented RNA was employed in immunoprecipitation experiment. Anti-m6A or anti-IgG antibody-coated magnetic beads (2 µL) were utilized for immunoprecipitation. Purified RNA was then subjected to qPCR using SALL4-specific primers designed based on a motif-dependent m6A site predictor SRAMP (http://www.cuilab.cn/sramp) (Forward: GAGCTGTACTGAGCCACCAG; reverse: GCATCCGGCTTCTCCTTCAT). Normalizing m6A enrichment could be achieved via the value generated using the input group as follows: %Input = 2Ct [input] – Ct [IP] x Dilution multiplier x 100.
Flow cytometry assay
Cells were stained with the Annexin V/FITC Apoptosis Detection Kit I (BD Pharmingen; Franklin Lakes, NJ, USA) 48 hours after receiving a 10 Gy radiation dose to look for signs of apoptosis. Using a Coulter-XL flow cytometer (Beckman Coulter Inc., Brea, CA, USA) and an EXPO32 ADC, we were able to capture and analyze pictures of apoptosis. The total apoptosis rate of tumor cells was calculated by adding the apoptosis rates observed in the early (lower-right) and late (upper-right) time points. After 24 hours, cells that had been irradiated with 10 Gy were collected and preserved in cold ethanol (1 mL PBS + 2 mL absolute ethanol) for studying the cell cycle. The cells were stained with a PI/RNase staining solution (BD Pharmingen; Franklin Lakes, NJ, USA). After that, we analyzed the cells cycles distributions using flow cytometry (Coulter-XL; Beckman Coulter Inc.; Brea, CA, USA) and Mod Fit 3.0 (Verity Software House, Inc.; USA).
Colony formation assay
CD44(+)-OSCC cells (200 cells/well) were plated in 6-cm tissue culture plates and irradiated at 0, 4, and 8 Gy, after 2 days of attachment. After 2 weeks, the cells were fixed in 4% paraformaldehyde for 20 minutes, stained with crystal violet, and then washed in PBS. Survival rate was calculated as (number of colonies/number of cells plated) irradiated / (number of colonies/number of cells plated) nonirradiated, where 50 cells was used as the cutoff for excluding colonies.
Caspase-3 activity assay
The day following cell seeding in 6-well plates, the cells were irradiated (10 Gy; single treatment) when they reached 70–80% confluence. At 24 hours post-irradiation, cells were analyzed using a Caspase-3 apoptosis detection kit (Beyotime; Shanghai, China). Ac-DEVD-pNA (2 mM) was added to the cell lysis detection solution. After incubation for 2 hours at 37°C, absorbances (at 405 nm) were measured using an Epoch™ spectrophotometer (BIO-TEK instruments Inc., Vermont, USA).
Immunofluorescence assay
The experiment was conducted as previously described [13]. Briefly, we first prepared and dewaxed the paraffin-embedded tumor samples. Subsequently, the samples were rehydrated, followed by antigen retrieval and blocking. Cells were processed by permeabilization using Triton-X100 (1%) for 30 min and blocking in BSA (1%) for 30 min after fixing in paraformaldehyde (4%) at room temperature for 20 min. Fluorescent rabbit secondary antibody (1:100, Abbkine) was used to detect the primary antibodies β-catenin (1:50, GTX101435, GeneTex), SALL4 (1:50, GTX109983, GeneTex), CD44 (1:50, #15675-1-AP, Proteintech Group) after an overnight incubation at 4°C with the primary antibodies. Fluorescent pictures were taken via the FluoView FV1000, a confocal laser scanning microscope (Olympus, Tokyo, Japan) after the nuclei were stained with a DAPI-based anti-fluorescence quencher overnight at 4°C.
In vivo tumorigenicity assay
At the Laboratory Animal Center of the Fourth Military Medical University in Xi'an, China, we bought female BALB/c nude mice aged 4–6 weeks. Mice were injected subcutaneously in the left forelimb with a solution containing 1 x 107 stably transfected CAL27-CD44(+) cells in 0.2 mL. After the tumor reached a volume of 100 mm3, the mice were randomized into two groups: those who received 10 Gy of radiation (5 Gy, 2 times) and those that did not. In addition, 12 mice were split into 4 groups (n = 3 in each) and randomly allocated to receive either a vector injection, SALL4 overexpression, vector injection plus irradiation, or SALL4 overexpression injection plus irradiation. The volume of the tumor was measured using a Vernier caliper every two days (length x width2 x 1/2). Over the course of two weeks, the tumor's development was tracked weekly. The Research at the Fourth Military Medical University and the Committee on the Use of Live Animals in Teaching and authorized all procedures using mice.
Statistical analysis
GraphPad Prism 8.0 (GraphPad Software, California, USA) was employed for all the statistical work. There were at least three sets of each experiment done. The student's unpaired t-test was employed to examine the correlation between the two groups. One-way ANOVA with Dunnett's multiple comparisons was utilized instead for analyzing data from more than three groups. Means and standard deviations are approached for summarizing the data. Statistical significance was defined as a P-value less than 0.05(*p < 0.05, **p < 0.01, ***p < 0.001).