Bioinformatics

To find out the genes associated with THCA, the GSE87410 (https://www.ncbi.nlm.nih.gov/search/all/?term=GSE87410%20) and GSE201365 (https://www.ncbi.nlm.nih.gov/search/all/?term=GSE201365%20) datasets were used with the screening criteria of adjusted P-value < 0.05 and |LogFC|> 2. KEGG pathway enrichment analysis of these differentially expressed genes (DEGs) was performed using the DAVID bioinformatics (https://david.ncifcrf.gov/) with a modified Fisher's exact test with Benjamini–Hochberg false discovery rate correction (FDR≦0.05) and FDR-adjusted P-value < 0.05. The available databases TIMER2.0 (http://timer.cistrome.org/), GEPIA version 2 (http://gepia2.cancer-pku.cn/#index, |Log2FC| Cutoff:1 and q-value Cutoff: 0.01) and TCGA (https://ualcan.path.uab.edu/analysis.html) were also used to confirm the upregulation of RASGRF1 in THCA. The LinkedOmics Suite web tool was used to predict RASGRF1-related m6A modifiers at https://www.linkedomics.org/#/, which was used to analyze the association of RASGRF1 and macrophage M1/M2 polarization (Pearson’s correlation analysis). The SRAMP web tool (http://www.cuilab.cn/sramp) was used for the prediction of the putative m6A sites in RASGRF1 mRNA based on the sequence from the UCSC database (https://genome.ucsc.edu/). The online TIMER2.0 project (Spearman’s correlation analysis) was used to observe the association of RASGRF1 and macrophage infiltration level in THCA.

Human clinical specimens

Primary THCA tumors (n = 33) and paired normal thyroid tissues (n = 33) were harvested from 33 patients with THCA from the Second Affiliated Hospital of Xi’an Jiaotong University after acquisition of informed consent. Analysis of human specimens for RASGRF1 expression was approved by the Second Affiliated Hospital of Xi’an Jiaotong University Ethics Committee.

Cell type and culture conditions

In this study, BCPAP (#IM-H149, Immocell, Xiamen, China), IHH4 (#CL-0803, Procell, Wuhan, China), TPC1 (#IM-H153, Immocell), K1 (#IM-H513, Immocell), Nthy-ori3-1 (#IM-H512, Immocell), and THP-1 (#CL-0233, Procell) cells were used. For cell cultivation, RPMI-1640 (for BCPAP, IHH4, Nthy-ori3-1 and THP-1; Beyotime, Shanghai, China), DMEM (for TPC1; Beyotime), and F12K (for K1; Lonza, Basel, Switzerland) were procured, all of which were replenished with 1% penicillin/streptomycin (Invitrogen, Wesel, Germany) and 10% FBS (Bovogen Biologicals, Springvale, Victoria, Australia). For THP-1 cells, the additional addition of 0.05 mM β-mercaptoethanol (Solarbio, Beijing, China) was needed.

ShRNAs, plasmids, and lentivirus constructs

For in vitro depletion experiments, pLV3-U6-IGF2BP3(human)-shRNA2-CopGFP-Puro (sh-IGF2BP3), pLV3-U6-IGF2BP2(human)-shRNA3-Puro (sh-IGF2BP2) and pLV3-U6-IGF2BP1(human)-shRNA1-Puro (sh-IGF2BP1) were purchased from Miaoling (Wuhan, China), as well as sh-RASGRF1 and shRNA harboring a scrambled sequence (sh-NC) from Tsingke (Xian, China). For in vitro upregulation assays, pCMV-FTO(human)−3 × FLAG-Puro (oe-FTO), pCMV-EGFP-RASGRF1(human)-Neo (oe-RASGRF1) and the matched control vector were from Miaoling. Sh-RASGRF1, sh-NC, oe-FTO, and oe-RASGRF1 lentivirus particles were generated by Genomeditech (Shanghai, China).

Transient transfection and stable transduction

Using FuGene, transient transfection of shRNAs or plasmids into K1 and TPC1 THCA cells was performed as suggested by the supplier (Promega, Leiden, The Netherlands). For stable transduction as described elsewhere (Lee et al. 2023), lentivirus particles were used to infect TPC1 THCA cells as per the manufactory protocols. After 24 h of infection, the media were changed with the fresh growth medium containing 2 µg/mL puromycin (Solarbio) to select stable cell lines.

Animal xenograft experiments and immunohistochemistry (IHC)

All procedures involving animals were conducted following the protocols approved by the Second Affiliated Hospital of Xi’an Jiaotong University Animal Care and Use Committee. To establish subcutaneous TPC1 THCA xenografts, female BALB/c nude mice aged from 5 to 7 weeks (Vital River Laboratory, Beijing, China) were injected with lentivirus-transduced TPC1 cells (6 × 106 cells/mouse) into their left flanks. Each group contained five mice. Tumor growth (volume = width2 × length × 0.5) was detected every five days. Upon reaching the predetermined tumor volume threshold (100 mm3), the oe-RASGRF1 lentivirus construct was administered via direct intratumoral injection to the assigned treatment cohort. Thirty days post the implantation, xenografts were harvested for photograph, weight and IHC assay. Sections of formalin-fixed, paraffin-embedded TPC1 THCA xenografts were subjected to IHC assays under standard methods (Lee et al. 2023), with antibodies (Proteintech, Wuhan, China) against Ki67 (rabbit pAb, #27,309–1-AP, 1 to 8,000), RASGRF1 (rabbit pAb, #12,958–1-AP, 1 to 300), PCNA (rabbit pAb, #10,205–2-AP, 1 to 5,000), MMP2 (rabbit pAb, #10,373–2-AP, 1 to 200), BAX (rabbit pAb, #50,599–2-Ig, 1 to 3,000), and GPX4 (mouse mAb, #67,763–1-Ig, 1 to 2,000).

Quantitative PCR

Using the RNA/Protein Co-extraction Kit and the accompanying protocols (Beyotime), total RNA was prepared from cultivated cells and harvested samples. Set-up first strand cDNA generation and quantitative PCR were conducted using PrimeScript RT Reagent kit and SYBR Green Mix, respectively, as suggested by the producer (TaKaRa, Dalian, China). Gene expression was determined using the formula 2−ΔΔCt relative to the level of β-actin as the reference gene. Sequences of primers were listed in Supplementary Table 1.

Immunoblotting

Using the RNA/Protein Co-extraction Kit (Beyotime), total protein was prepared from cultivated cells and harvested samples, followed by quantification of protein concentration by BCA assay (Thermo Fisher Scientific, Saint-Aubin, France). Protein (20 µg/lane) was gel electrophoresed on Bis–Tris gels (Thermo Fisher Scientific), electroblotted to PVDF membranes (Servicebio, Wuhan, China), and probed with antibodies specific for RASGRF1 (rabbit pAb, #12,958–1-AP, 1 to 800), PCNA (rabbit pAb, #10,205–2-AP, 1 to 30,000), MMP2 (rabbit pAb, #10,373–2-AP, 1 to 1000), BAX (rabbit pAb, #50,599–2-Ig, 1 to 10,000), GPX4 (mouse mAb, #67,763–1-Ig, 1 to 3000), and β-actin (mouse mAb, #66,009–1-Ig, 1 to 50,000). Following signal development by EZ-ECL Kit (Biological Industries, Beit-Haemek, Israel), data were analyzed using ChemiDoc XRS + imaging system (Bio-Rad, Glattbrugg, Switzerland).

Evaluation of colony formation ability and proliferation

For colony formation assay, ~ 200 K1 and TPC1 THCA cells after transfection of sh-Ctrl, sh-RASGRF1, vector, oe-FTO or oe-FTO + oe-RASGRF1 were plated in a 6-well culture plate (Corning Costar, Lindfield, NSW, Australia). Cell colony was allowed to grow for 10–14 days. After crystal violet (1%) staining, pictures were captured and quantified the generated colonies (more than 50 cells) with ImageJ (NIH, Bethesda, MD, USA).

For EdU proliferation assay, K1 and TPC1 THCA cells grown in a 96-multiwell culture plate were subjected to the relevant transfection. Forty-eight hours post the transfection, a Yefluor 488 EdU Imaging Kit was employed to assess the proliferating cells under the protocols of the producer (Yeasen, Shanghai, China). The percentage of EdU positive cells (showing a green fluorescence) was determined relative to total nuclei (showing a blue fluorescence).

Detection of ROS, MDA, GSH and Fe2+ contents

The commercial MDA Assay Kit (Spbio, Wuhan, China), GSH Assay Kit (Spbio), DCFH-DA probe ROS assay Kit (Elabscience, Wuhan, China), and Ferrous Iron Colorimetric Assay Kit (Elabscience) were utilized to detect the levels of MDA, GSH, ROS and Fe2+ in transfected K1 and TPC1 THCA cells, as suggested by the vendors.

Incubation of the conditioned medium (CM)

K1 and TPC1 THCA cells were subjected to the relevant transfection, followed by the collection of the CMs. Meantime, THP-1 cells were exposed to 100 ng/mL of PMA (Selleck, Shanghai, China) to induce differentiation to macrophages (THP-1-M0). The CM was used to incubate THP-1-M0 cells in a mixture of CM:complete culture medium (1:1). Twenty-four hours later, incubated THP-1-M0 cells were harvested for expression analysis, CD206+ macrophage evaluation, and migration assay.

Flow cytometry

Using an LSRII flow cytometer (BD Bioscience, Heidelberg, Germany), flow cytometry analyses were conducted. For apoptosis assay, a Double Staining Apoptosis Kit and protocols (Vazyme Biotech, Nanjing, China) for Annexin V-FITC and propidium iodide double staining were applied. Within 1 h, the apoptotic rate was analyzed. To characterize macrophages, single-cell suspensions of THP-1 and THP-1-M0 cells were stained with FITC-labeled CD11b antibody (#FITC-65055, Proteintech). To detect the percent of CD206+ macrophages, single-cell suspensions of CM-incubated THP-1-M0 cells were stained with FITC-labeled CD206 antibody (#FITC-65155, Proteintech) and PC5.5-labeled 7AAD (for dead cell elimination; Biolegend, San Diego, CA, USA).

Transwell assays for cell migration and invasion

24-Transwell inserts (Corning Costar) pre-coated with (for invasiveness evaluation) or without (for motility analysis) Matrigel (MCE, Shanghai, China) were applied for these assays. Briefly, serum-starved K1 and TPC1 THCA cells after the relevant transfection as well as CM-incubated THP-1-M0 cells were seeded onto the inserts at 1 × 105 (invasiveness assay) or 4 × 105 (migration assay) cells per well. The inserts into were then placed in the matched 24-well culture plates containing 600–700 µL of 10% FBS complete medium. Twenty-four hours later, following crystal violet (1%) staining, pictures were obtained at 100 × magnification under an invert microscope (Axiovert 200 M, Zeiss, Oberkochen, Germany) and the invasive or migratory cells were quantified at more than five random fields.

RNA immunoprecipitation (RIP) and MeRIP assays

Using a BeyoRIP™ RIP Assay Kit (Beyotime), these experiments were conducted. In accordance with the manufactory protocols, untransfected, vector-, or oe-FTO-transfected K1 and TPC1 cells were lysed and subjected to incubation of prepared complex of Protein A/G Agarose and antibodies to m6A (mouse mAb, #68,055–1-Ig, Proteintech), FTO (rabbit pAb, #27,226–1-AP, Proteintech), IGF2BP2 (rabbit pAb, #11,601–1-AP, Proteintech), or Isotype IgG (rabbit pAb, #30,000–0-AP, Proteintech). The precipitated complex was harvested to isolate related RNA for the quantification of the enrichment amount of RASGRF1 mRNA.

Analysis of RASGRF1 mRNA stability

Vector- or oe-FTO-transfected K1 and TPC1 cells were subjected to Actinomycin D exposure (Selleck) at 50 µg/mL. At 0, 4 and 8 h after treatment, treated cells were harvested to isolate RNA and examine the amount of RASGRF1 mRNA.

Data analysis

All studies were performed at least 3 times (in triplicate). Two-tailed t-test or one- or two-way ANOVA were applied to evaluate significant differences. A P value less than 0.05 indicated statistical significance.