FTO-mediated m6A demethylation of CSF3 suppresses NETosis via downregulation of RLN2 expression in colorectal cancer.

Introduction
Colorectal cancer (CRC) is one of the most popular cancers worldwide, with probably 1.9 million new cases and nearly 900000 deaths annually (Andrei et al. 2022; Siegel et al. 2024). The conventional treatment for CRC includes surgery, radiotherapy, chemotherapy, and immunotherapy (Ganesh et al. 2019). Although CRC has improved in screening and treatment, its incidence rate, prevalence and mortality are still high, especially for patients diagnosed with advanced CRC, and the 5-year survival rate is below 10% (Schmitt and Greten 2021). Thus, it’s imperative to uncover the molecular mechanisms of genes participating in CRC progression and deepen our insights into the CRC progression process.
Neutrophils are crucial in innate immunity (Mantovani et al. 2011). When the host is infected with pathogenic bacteria, activated neutrophils can engulf and kill the pathogen, while releasing neutrophil extracellular traps (NETs) to defend the body (Papayannopoulos 2018). The process of neutrophils with NET formation is called NETosis (Liang et al. 2024). It’s reported that NETosis is closely bound up with the occurrence of tumors, and NETosis levels are significantly higher in patients with lung cancer, pancreatic cancer and bladder cancer, compared to healthy volunteers (Oklu et al. 2017). After chemotherapy, patients with increased NETosis markers in tumor patients are more likely to metastasize and undergo surgery again, leading to poor prognosis (Cools-Lartigue et al. 2014). CSF3, also known as G-CSF18, is a regulatory factor for neutrophil recruitment. In our previous study, we found that CSF3 accelerates the progression of CRC (Xu et al. 2025). Notably, a recent study has found that CSF3 can regulate NETosis in malignant glioma (Dai et al. 2024). However, it is uncertain whether CSF3 promotes the progression of CRC by regulating NETosis.
RNA modification is a momentous part of epigenetics and exerts a significant function in cell regulation along with gene and protein modification (Lin and Kuang 2024). N6-methyladenosine (m6A) methylation is the commonest and most abundant form of mRNA modification in eukaryotes, and it’s a dynamic and reversible process of mRNA modification (An and Duan 2022). In recent years, increasing evidence suggests that m6A RNA methylation regulatory genes are strongly bound up with the development of various malignant tumors. He et al. have confirmed that ALKBH5 plummets the activity of pancreatic cancer cells by RNA KCNK15-AS1 demethylation (He et al. 2018). Li et al. implied that m6A RNA methylation regulatory genes lead to genetic alterations in different forms of cancer (Li et al. 2019). However, how m6A regulates CSF3 function and its influence on the pathogenesis of CRC remain elusive. In this work, the molecular mechanisms of CSF3 involved in CRC progression were explored. This study revealed that CSF3 promotes the tumorigenesis of CRC and NETosis via upregulating RLN2, and FTO mediates m6A demethylation of CSF3 and regulates its expression. This study will provide a novel therapeutic strategy for CRC treatment.

Methods

Isolation of neutrophils
For neutrophils isolation in bone marrow, the mice were euthanized by cervical dislocation. Then the femur and tibia were extracted, and a 1 mL syringe was used to aspirate PBS solution. Then the needle was inserted from one end of the bone, and the bone marrow was rinsed repeatedly, followed by washed cell suspension collection. Subsequently, a 70 μm filter was applied to filter the cell suspension, and the filtrate was collected and centrifuged. Then 5 mL red blood cell lysis buffer (#C3702-120 mL, Beyotime) was added for 5 min and centrifuged. The 2 mL cell suspension was then added carefully into the top of a three-layer Percoll (#17–0891-09, Pharmacia) gradient of 78%, 64% and 52% diluted in PBS. After centrifugation, neutrophils were collected from the 64/78% interface. Then neutrophils with high purity were acquired after residual erythrocyte lysis. The obtained neutrophils were washed with PBS, and a 2 mL RPMI-1640 medium (#C11875500BT, Gibco) was employed to resuspend. The purity and survival rate of neutrophils were assessed with FACS utilizing antibodies against CD11b (#101,211, Biolegend) and Ly6G (#127,605, Biolegend).

Cell culture and transfection
Human CRC cell lines HCT116 (#CL-038 h) and RKO (#CL-356 h) purchased from SAIOS (Wuhan, China) were cultivated in DMEM medium (#C11885500BT, Gibco) supplemented with 10% FBS (#S9030, Solarbio) at 37℃ with 5% CO2. CCD 841 CoN cells (#CL-513 h, SAIOS) were maintained in DMEM/F12 medium (#11,320,033, Thermo) containing 10 mM 4-(2-hydroxyethyl)−1-piperazineethanesulfonic acid (HEPES), 10% FBS, 10 ng/mL cholera toxin, 100 ng/mL hydrocortisone, 10 mM insulin, and 10 mM transferrin. CSF3 overexpression (oe-CSF3) and FTO overexpression (oe-FTO) sequences were acquired from the NCBI website, and the sequence for the sh-CSF3, sh-FTO and sh-RLN2 design obtained from the VectorBuilder website (https://www.vectorbuilder.cn/): CSF3 shRNA: 5’-TGCAGGCTCTATCGGGTATTT-3’, FTO shRNA: 5’-CAGCAGUGGCAGCUGAAAUAU-3’, RLN2 shRNA: 5’-AGCACTGGAGGCTTGAATAAA-3’. To downregulate CSF3, FTO, and RLN2, the shRNA sequences and negative control shRNA (shNC) were cloned into the pLVX-shRNA2-Puro vector. The coding sequences region of CSF3 and FTO were synthesized and cloned into the pCDH-CMV-MCS-EF1-T2A-Neo vector for overexpressing CSF3 and FTO. The HCT116 and RKO cells were grown into a 6-well plate, with 2 × 105 cells per well, and maintained until the cell density reached 70–90%. To produce lentiviruses, 293 T cells were transfected with the lentiviral vector along with packaging plasmids. The culture media was collected, pooled, and filtered at 48 h after transfection. Subsequently, the designated lentivirus was used to infect the HCT116 and RKO cells, and the expression levels of CSF3, FTO, and RLN2 were assessed.

Animals
Our study exclusively examined female mice. It is unknown whether the findings are relevant for male mice.
Female C57BL/6 mice (8–10 weeks, 17–22 g) were acquired from the Laboratory Animal Center of Yangzhou University. After one week of adaptation, mice were intraperitoneal injected 12.5 mg/kg azoxymethane (AOM) (#HY-111375, MCE), and the mice were supported drinking water supplemented with 2.5% DSS (#HY-116282, MCE) after 5 days. Then the mice received regular drinking water for 16 days, then two additional DSS treatment cycles were carried out. On the 100th day, the colon was removed, and the tumors were counted. In addition, oe-FTO and oe-NC lentivirus were tail vein injected 7 days before AOM injection, and the mice received another injection of lentivirus on the 35th day after the first injection. All research procedures associated with mice were approved by the Experimental Animal Ethics Committee of Yangzhou University (Ethics No. 202409049). This study was implemented in line with the Animal Research: Reporting of In Vivo Experiments (ARRIVE).

Cell proliferation
CCK-8 and EdU staining were utilized to detect the cell proliferation. For CCK-8, cell viability was assessed at 1, 2, 3, and 4 days utilizing CCK-8 kits (#C0037, Beyotime). HCT116 and RKO cells were grown in a 96-well plate, with 100 μL per well. Then 10 μL CCK-8 solution was put in each microwell for 2 h. The microplate reader (DR-3518G, Wuxi Hiwell Diatek) was employed to determine the absorbance at 450 nm. Besides, for EdU staining, the BeyoClick™ EdU-488 cell proliferation kit (#C0071S, Beyotime) was utilized. HCT116 and RKO cells were maintained in a 96-well plate, with 5 × 105 per well. The EdU solution was employed to label the cells at 37 ℃ for 2 h, then fixed and permeabilized. Subsequently, the cells were maintained at Click Additive solution and DAPI solution, and the results were determined using a fluorescence microscope (BX53, Olympus, Japan).

Transwell assay
After the HCT116 and RKO cells were diluted with basic culture medium into a cell suspension of 1 × 105 cells/mL, cells were seeded in 24-well transwell chambers coated with or without 50 mg/mL Matrigel to assess cell migration and invasion. Subsequently, 200 µL cells were maintained on transwell chambers under 5% CO2 at 37 °C, and medium containing 10% FBS was added. After 24 h, 4% formaldehyde was employed to fix cells for 30 min and 0.1% crystal violet was utilized to stain for 20 min. The cells in each group were measured with an inverted microscope (CKX53, Olympus, Japan).

Quantification of total RNA m6A
The total RNA m6A was evaluated utilizing a m6A RNA Methylation kit (Colorimetric) (#P-9005, Amyjet Scientific). In short, total RNA was collected from HCT116 and RKO cells. Then 80 μL binding solution and 200 ng RNA were added in each well and maintained at 37 °C. Next, 50 μL capture antibody, 50 μL detection antibody, 50 μL enhancer, 100 μL developer solution and 100 μL stop solution were in turn put in each well. A microplate reader (DR-3518G, Wuxi Hiwell Diatek) was applied to determine the absorbance at 450 nm.

ELISA
The content of RLN2 in HCT116 and RKO cells culture medium supernatant or tissues was detected by commercial kits (#SEKH-0304, Solarbio), as well as MPO-DNA complexes level (#NLM510, NEW LAI BIOCAIENCES). Absorbances at 450 and 630 nm were estimated by a microplate reader (DR-3518G, Wuxi Hiwell Diatek).

Immunofluorescence
The following treatment according to grouping, HCT116 and RKO cells and 4% paraformaldehyde was employed to fix the tissues for 20 min. After blocking with goat serum (#SL038, Solarbio) for 30 min, cells were probed with primary antibodies, containing MPO (1:50, #ab208670, Abcam) and H3cit (1:250, #ab309551, Abcam), at 4℃ overnight. Subsequently, cells were maintained with secondary antibody goat anti-rabbit IgG (1:200, #ab6721, Abcam) for 1 h. After 200 μL DAPI (#C1005, Beyotime) staining for 10 min, the immunofluorescence micrographs were determined utilizing a fluorescence microscope (BX53, Olympus, Japan).

Western blot
RIPA lysis buffer (#P0013B, Beyotime) was added to HCT116 and RKO cells to obtain total protein, followed by separation by SDS-PAGE and transferred to PVDF membranes (Beyotime, FFP24). The blocked membranes were incubated with primary antibodies FTO (1:1000, #27,226–1-AP, proteintech), CSF3 (1:1000, #ab181053, abcam), RLN2 (1:1000, #ab183505, abcam) and GAPDH (1:10,000, #ab181602, Abcam) at 4℃ overnight. Membranes were then probed with secondary antibody goat anti-rabbit IgG (1:2000, #ab6721, Abcam) for 1 h. Protein bands were developed by ECL (Applygen, P1000).

qRT-PCR
The cells and tissues were lysed with Trizol reagent (#15,596,018, Invitrogen, USA), and then 5 × FastKing RT SuperMix (#KR118; TIANGEN, China) was used for cDNA synthesis. SYBR Green Power Master Mix (#A4004M, Lifeint, China) was utilized for qRT-PCR. Primer sequences contained: Human CSF3 (F) 5’-CCCTGGAGCAAGTGAGGAAG-3’ and (R) 5’-TACGAAATGGCCAGGACACC-3’; Mouse CSF3 (F) 5’-CATGAAGCTAATGGCCCTGC-3’ and (R) 5’- GATGTCTTGTCCCCCGGAAG-3’; Human RLN2 (F) 5’-ATTGTGCCATCCTTCATC-3’ and (R) 5’-GCGGCTTCACTTTGTCTA-3’; GAPDH (F) 5’-TGTGGGCATCAATGGATTTGG-3’, and (R) 5’-ACACCATGTATTCCGGGTCAAT-3’. GAPDH was applied as the internal control.

Methylated RNA immunoprecipitation (MeRIP) assay
The GenSeq®m6AMeRIP kit (#GS-ET-001, GenSeq) was employed to assess the m6A level of CSF3 in HCT116 and RKO cells. In short, after total RNA collection from HCT116 and RKO cells, 2 μL fragmentation buffer was added to 10 μL RNA in each tube. Subsequently, the fragmented RNA and magnetic beads conjugated with m6A antibody were maintained in the MeRIP reaction mixture, followed by qRT-PCR carried out.

RNA-binding protein immunoprecipitation (RIP)
A RIP kit (#Bes5101, BersinBio) was employed to conduct RIP analysis. Briefly, RIP lysis buffer was utilized to lyse the HCT116 and RKO cells. Then the lysate products with magnetic beads preconjugated to anti-IgG, anti-CSF3, or anti-FTO antibodies overnight. Then, phenol–chloroform‒isoamyl alcohol method was applied to collect and purify the RNA, followed by detection the expression level of CSF3 by qRT-PCR.

Immunohistochemistry
Immunohistochemistry was applied to determine the expression of FTO and Ki-67. The 4% paraformaldehyde was utilized to fix the tissue sections for 48 h, then deparaffinized, and hydrated. Subsequently, sections were probed with primary antibodies including FTO (1:200, #27,226–1-AP, proteintech) and Ki-67 (1:1000, #28,074–1-AP, proteintech) at 4℃ overnight. Next, sections were incubated with secondary antibody for 10–15 min. Following DAB (#P0202, Beyotime) color development and hematoxylin (#G1080, Solarbio) counterstaining, sections were observed under an Olympus microscope (DP27, Tokyo, Japan).

Bioinformatics analysis
TCGA-CRC cohort was obtained from TCGA database, and GSE47756 dataset was downloaded from GEO database. TCGA-CRC cohort included 635 CRC and 51 normal samples, and GSE47756 contained 55 CRC and 38 normal samples. Wilcoxon test was used to evaluate the differences of expression levels of CSF3, RLN2, and FTO between the CRC and normal groups. In addition, molecular docking was employed to explore the associations between CSF3 and RLN2 or FTO. First, the protein accession numbers of CSF3, RLN2 and FTO were obtained from the Uniprot database, and the crystal structures of the proteins (in PDB format) were downloaded from PDB database. The HDOCK Server database was used to obtain the residue pairs and atomic distances involved in the interaction between Model1-receptor (CSF3) and ligands (FTO or RLN2). Atomic distances less than 3.5 Å was considered a strong interaction.

Statistical analysis
All experimental data (means ± SD) were analyzed with GraphPad 7.0. Data normality was evaluated by Shapiro–Wilk test, and the variance homogeneity was assessed using Levene test. Differences between the two groups were assessed using unpaired Student’s t-test. For multiple-group data, one-way analysis of variance (ANOVA) was employed, followed by Tukey’s post-hoc test. The normality tests were conducted before performing the unpaired Student’s t-test and ANOVA if applicable. Otherwise, Mann–Whitney U test and Kruskal–Wallis H test were carried out for non-parametric data. All experiments were repeated at least three times. Statistical significance was deemed when P < 0.05.

Results

CSF3 facilitates NETosis in CRC
Considering the dysregulation of CSF3 in CRC (Xu et al. 2025), the expression level of CSF3 was explored based on TCGA-CRC dataset. As shown in Supplementary Fig.  1 A, CSF3 was significantly overexpressed in CRC group when compared to that in normal group (P < 0.0001). Then oe-CSF3 and sh-CSF3 lentivirus were used to infect the HCT116 and RKO cells to explore the involved molecular mechanism, and the transfection efficiency was measured (Figs. 1A and B). To explore the specific influence of CSF3 on the function of neutrophil, the neutrophils were isolated from bone marrow, and the purity and survival rate of neutrophils were assessed. FACS revealed that the purity and survival rate of neutrophils were 88.07% and 94.84%, respectively (Fig. 1C). After cell transfection, a co-culture system was constructed utilizing transwell chambers with a 0.4 μm porous membrane separating the two chambers to induce NETs in the presence of CRC cells. Neutrophils were placed in the bottom chamber, while the CRC cells were placed on the membrane of the upper chamber. Then the neutrophils were collected after 48 h. Immunofluorescence was applied to evaluate the markers of neutrophils (MPO) and NETosis (H3cit) (Yang et al. 2020), and CRC cells co-cultured with neutrophils induced NETosis, and Sytox Green was used to detect the NETs formation. Of note, the induced NETosis was obviously enhanced and inhibited by oe-CSF3 and sh-CSF3 treatment, respectively (both P < 0.01; Figs. 1D-F), as well as the MPO-DNA complexes level (Fig. 1G). In addition, to determine the direct effects of CSF3 on NETosis, 2 ng/mL G-CSF was used to stimulate neutrophils. As illustrated in Figs. 1H-I, G-CSF-stimulated obviously enhance the NETs formation, as well as MPO-DNA complexes level. These findings imply that CSF3 facilitates NETosis in CRC.

CSF3 promotes the tumorigenesis of CRC and NETosis via upregulating RLN2
It’s reported that RLN2 is closely associated with tumor occurrence (Bialek et al. 2021; Kotwal et al. 2024; Tevz et al. 2016). In this study, the expression level of RLN2 was obviously increased in CRC group relative to normal group (P < 0.0001; Supplementary Fig. 1B), and a strong interaction also found between CSF3 and RLN2 (atomic distances less than 3.5 Å; Supplementary Fig.  1 C and Supplementary Table 1). Thus, the influence of CSF3 on RLN2 in CRC was assessed in this study. The results found that the RLN2 level significantly increased with the expression of CSF3 in CRC cells culture medium supernatant (P < 0.05; Figs. 2A-C). To explore the influence of RLN2 in the molecular mechanism of CSF3 in CRC, the RLN2 was added, then the RLN2 level was detected after 24 h, and the results showed that RLN2 implemented significantly elevated the RLN2 expressed in CRC cells culture medium supernatant after sh-CSF3 treatment (P < 0.05; Figs. 2D-F). In addition, RLN2 administration significantly increased the cell proliferation, migration and invasion of CRC cells after sh-CSF3 treatment (both P < 0.05; Figs. 2G-J). However, opposite results were found in oe-CSF3 treated CRC cells after sh-RLN2 treatment (Supplementary Figs. 2A-D). Of note, after co-culturing RKO and HCT116 cells with neutrophils, co-cultured significantly elevated the RLN2 expression (P < 0.05; Supplementary Fig. 2E). More interestingly, RLN2 is highly expressed in the culture medium supernatant of CSF3-treated neutrophils (Sheng et al. 2023). In this study, immunofluorescence found that RLN2 treatment could obviously enhance the NETs formation in the co-culture system after sh-CSF3 implementation (both P < 0.01; Figs. 3A-C), as well as the MPO-DNA complexes level (Fig. 3D). However, opposite results were found in oe-CSF3 treated CRC cells after sh-RLN2 treatment (Supplementary Fig. 3). All these suggest that CSF3 promotes the tumorigenesis of CRC and NETosis via upregulating RLN2.

FTO mediates m6A demethylation of CSF3 and mitigates its expression
In recent years, accumulating studies have demonstrated the crucial effect of m6A modification on RNA metabolism and function of tumor cells at the post-transcriptional level (Huang et al. 2020; Wang et al. 2014). We want to know whether the modification of m6A is associated with the CSF3 in CRC. Firstly, the total m6A levels of RNAs in CRC cells were examined, and the results revealed that levels of RNA m6A in CRC cells were lower than that in control (Fig. 4A), suggesting that m6A might have participated in the pathogenesis of CRC. Then, the transcripts of CSF3 in CRC cells enriched with m6A were obviously elevated when compared to the CCD 841 CoN cells (P < 0.01; Fig. 4B), indicating that m6A modification may exert a significant function in the regulation of CSF3. Besides, it’s reported that knocking down FTO can promote m6A methylation of CSF3 and regulate its mRNA expression (Dai et al. 2024), indicating that FTO is an upstream molecule of CSF3. In this study, the expression of FTO protein was remarkedly decreased in CRC cells compared to CCD 841 CoN cells (P < 0.01; Fig. 4C), as well as the downregulated FTO in CRC in GEO dataset (Supplementary Fig. 1D), which further highlighted the critical role of FTO in the pathogenesis of CRC. Also, a strong interaction was uncovered between CSF3 and FTO (atomic distances less than 3.5 Å; Supplementary Fig. 1E and Supplementary Table 2). To deeply investigate the molecular mechanism, the designed lentivirus was used to infect the HCT116 and RKO cells, and the transfection efficiency was measured (Fig. 4D). Also, upregulation of FTO resulted in the downregulation of the CSF3 mRNA expression, the overall m6A modification level, and transcripts of CSF3 enriched with m6A in CRC cells (Figs. 4E-G). However, opposite results were found in sh-FTO treated CRC cells (Supplementary Fig. 4). The RIP assay result revealed that FTO can specifically bind to CSF3 (Fig. 4H), which further suggested that FTO mediated m6A demethylation of CSF3 and inhibited its expression.

FTO/CSF3/RLN2 axis regulates NETosis and tumorigenesis of CRC in vitro
To explore the influence of FTO/CSF3/RLN2 axis in tumorigenesis of CRC, the in vitro experiments were conducted. The results revealed that oe-CSF3 administration could fully abolished inhibitory impact of FTO on the levels of CSF3 and RLN2 (Figs. 5A-C), while didn’t change the protein expression of FTO (Fig. 5D). Upregulation of CSF3 effectively relieved the inhibitory action of FTO overexpression on the proliferation, migration and invasion of CRC cells (Figs. 5E-H). Immunofluorescence found that oe-FTO treatment could obviously inhibit the NETs formation in the co-culture system (both P < 0.05; Figs. 6A-C), as well as the MPO-DNA complexes level (Fig. 6D), while this phenomenon was reversed after CSF3 upregulation. Collectively, these data elucidate that the FTO/CSF3/RLN2 axis regulates NETosis and tumorigenesis of CRC in vitro.

FTO/CSF3/RLN2 axis mediates NETosis and tumorigenesis of CRC in vivo
Furthermore, the role of FTO/CSF3/RLN2 axis was also assessed in CRC mice. After FTO upregulation, the mice exhibited clinical signs of CRC relief, such as body weight, tumor length, and tumor number (Figs. 7A and B). Also, the expression of FTO was obviously elevated after FTO overexpression, while Ki-67 expression showed an opposite trend (both P < 0.05; Fig. 7C). Notably, the expression levels of CSF3 and RLN2 in CRC mice both significantly decreased after FTO overexpression (both P < 0.05; Figs. 7D and E). Similarly, immunofluorescence found that oe-FTO treatment could obviously mitigate the NETs formation in CRC mice (P < 0.01; Fig. 7F). All these findings elucidate that the FTO/CSF3/RLN2 axis mediates NETosis and tumorigenesis of CRC in vivo.

Discussion
CRC is a frequent malignant tumor with high morbidity and mortality (Bray et al. 2024). In our previous study, we found that CSF3 accelerates the progression of CRC (Xu et al. 2025). However, the involved molecular mechanisms require further in-depth exploration. In this study, we found that FTO-mediated m6A demethylation of CSF3 suppresses NETosis via downregulation of RLN2 expression in CRC.
NETosis is a special type of neutrophil death that exerts a significant function in promoting tumor function and attenuating tumor immunotherapy (Mowery and Luke 2024; Yang et al. 2024). Lei et al. revealed that A2AR-mediated CXCL5 overexpression on macrophages facilitates the development of non-small cell lung cancer via NETosis (Lei et al. 2024). In this study, we found that the induced NETosis was obviously enhanced and inhibited by oe-CSF3 and sh-CSF3 treatment, respectively. Dai et al. also uncovered that IGF2BP3-induced NETosis was dependent on CSF3, and CSF3 could promote the progression of malignant glioma exacerbated by IGF2BP3-induced NETosis (Dai et al. 2024). Thus, these findings highlight that CSF3 facilitates CRC via NETosis, suggesting the function of immunosuppressive neutrophils in the pathogenesis of CRC treatment.
RLN is considered a central pregnancy hormone (Ivell and Einspanier 2002). As is well known, three peptide forms of RLN, containing RLN1, RLN2, and RLN3, exist in humans (Patil et al. 2017). Among these, RLN2 is found to be closely bound up with the development of cancer. For instance, Fue et al. have demonstrated that RLN2/RXFP1 signaling induces cell invasion of endometrial cancer cells via the β-Catenin pathway (Fue et al. 2018). Ren et al. have found that RLN2 regulates cell migration, invasiveness and viability of osteosarcoma cells in vitro through S100A4/MMP-9 signaling (Ren et al. 2015). In this study, molecular docking found a strong interaction between CSF3 and RLN2, and RLN2 content was significantly increased with the expression of CSF3 in CRC cells culture medium supernatant, and RLN2 administration significantly increased the tumorigenesis of CRC cells after sh-CSF3 treatment. However, opposite results were found in oe-CSF3 treated CRC cells after sh-RLN2 treatment, which indicating that CSF3 promotes the tumorigenesis of CRC via upregulating RLN2. Besides, it’s reported that RLN2 is overexpressed in the culture medium supernatant of CSF3 treated neutrophils (Sheng et al. 2023). This study found that RLN2 treatment could obviously enhanced the NETs formation in the co-culture system after sh-CSF3 implement, as well as the MPO-DNA complexes level. However, contrary trends were uncovered in oe-CSF3 treated CRC cells after sh-RLN2 treatment. These data confirmed that CSF3 promotes the tumorigenesis of CRC and NETosis via upregulating RLN2.
The m6A modification is now considered the main driving factor of RNA function for maintaining homeostasis in cancer cells (Alam and Giri 2024; Shu et al. 2024; Wang et al. 2024). The dynamic m6A modification of mRNA has been reported to be essential for maintaining cell growth, differentiation and metabolism (Ries et al. 2019; Zhang et al. 2017). Tumor cells retain proliferation and self-renewal by regulating abnormal m6A modifications of specific mRNAs under different conditions (Liu et al. 2018; Vu et al. 2019). Numerous researches have focused on investigating the function of m6A-methylated mRNA in cancer cells, such as SOCS2, EphA2 and VEGFA (Chen et al. 2018; Liu et al. 2022). However, how m6A mediates the CSF3 effect and its influence on the pathogenesis of CRC is still elusive. This study revealed that levels of RNA m6A in CRC cells were lower than that in control, and the transcripts of CSF3 in CRC cells enriched with m6A were obviously elevated when compared to the CCD 841 CoN cell, which was mediated by the m6A demethylase FTO. FTO has been reported to be associated with various types of cancer (Gholamalizadeh et al. 2020; Ye et al. 2023). This study also observed that the FTO expression was remarkedly decreased in CRC cells compared to CCD 841 CoN cells. Besides, molecular docking found a strong interaction between CSF3 and FTO, and the upregulation of CSF3 effectively relieved the inhibitory action of FTO upregulation on the tumorigenesis of CRC cells. Immunofluorescence found that oe-FTO treatment could obviously mitigate the NETosis in CRC, as well as the MPO-DNA complexes level, while this phenomenon was reversed after CSF3 upregulation. Similarly, the function of FTO in CRC was also confirmed in CRC mice. Collectively, these data elucidate that FTO-mediated m6A demethylation of CSF3 suppresses NETosis via downregulation of RLN2 expression in CRC.
Although this study confirmed the stability of CSF3 in CRC can be mediated by m6A, this study also has some deficiencies. Firstly, the influences of CSF3, FTO, and RLN2 in CRC tumorigenesis should be confirmed at the clinical level, as well as the function of FTO/CSF3/RLN2 axis in vivo. Secondly, more NETosis inhibitors should be employed to confirm the results obtained in this study. Besides, the specific molecular mechanism by which CSF3 regulates RLN2 still needs to be further investigated. Furthermore, only female mice are used for the in vivo experiments in this study. Gender differences may affect the stability of the results, and future research should be conducted to explore the gender-specific differences. Also, this study has not yet precisely identified the specific m6A binding sites of FTO on the CSF3 transcript, and the direct effect of methylation modification on the stability of CSF3 has not been verified through site mutation experiments. Additionally, further exploration is warranted to explore the role of immunosuppressive neutrophils in the pathogenesis of CRC treatment.

Conclusion
In summary, this study found that FTO-mediated m6A demethylation of CSF3 suppresses NETosis via downregulation of RLN2 expression in CRC (Fig. 8). CSF3, FTO, and RLN2 could be served as the potential biomarkers for CRC. This study might offer novel perspectives on CRC treatment.

Supplementary Information
Below is the link to the electronic supplementary material.