RBM15/IGF2BP1-mediated m ⁶A methylation of AKR1B10 promotes the progression of cervical squamous cell carcinoma

Abstract

Objective

RNA-binding motif protein 15 (RBM15) is a methyltransferase involved in N ⁶-methyladenosine (m ⁶A) modification and has been shown to be associated with the initiation and progression of various malignancies. This study aims to investigate the biological function and underlying mechanism by which RBM15 promotes the progression of cervical squamous cell carcinoma (CSCC) through in vitro and in vivo experiments.

Methods

The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases were used to analyze differences in the mRNA expression levels of RBM15, aldo-keto reductase family 1 member B10 ( AKR1B10), and insulin-like growth factor 2 mRNA-binding protein 1 ( IGF2BP1) between CSCC and adjacent non-tumor tissues, as well as the associations of RBM15, AKR1B10, and IGF2BP1 expression with overall survival in patients with CSCC. Real-time PCR and Western blotting (WB) were used to determine the expression levels of RBM15, AKR1B10, and IGF2BP1 in CSCC tissues and adjacent normal tissues. Cell Counting Kit-8 (CCK-8), colony formation, 5-ethynyl-2’-deoxyuridine (EdU) staining, wound-healing, and Transwell migration and invasion assays were performed to investigate the effects of RBM15 knockdown on the proliferation, migration, and invasion of CSCC cells. A nude mouse xenograft model was established to evaluate the effect of RBM15 knockdown on tumor growth. Methylated RNA immunoprecipitation (MeRIP) was performed to assess m6A modification of AKR1B10. RNA immunoprecipitation (RIP) and actinomycin D assays were used to analyze the interactions of RBM15 and IGF2BP1 with AKR1B10 mRNA. Small interfering RNAs (siRNAs), including si-IGF2BP1 and si-YTHDF1, were used to knock down IGF2BP1 or YTH N6-methyladenosine RNA-binding protein F1 ( YTHDF1), respectively, and their roles in regulating AKR1B10 mRNA stability were analyzed.

Results

Analysis of TCGA and GEO data showed that the mRNA and protein expression levels of RBM15, AKR1B10, and IGF2BP1 were significantly up-regulated in CSCC tissues (all P<0.05), and high expression of RBM15, AKR1B10, and IGF2BP1 was significantly associated with poor prognosis in patients with CSCC (all P<0.05). In vitro experiments showed that RBM15 knockdown inhibited the proliferation, invasion, and migration of CSCC cells, while in vivo experiments showed that RBM15 knockdown inhibited tumor growth (all P<0.05). MeRIP and actinomycin D assays showed that RBM15 knockdown significantly reduced the m ⁶A methylation level of AKR1B10 and decreased its mRNA stability. RIP assays showed that RBM15 knockdown significantly reduced the binding of RBM15 and IGF2BP1 to AKR1B10 mRNA (all P<0.05). IGF2BP1-dependency experiments showed that si-IGF2BP1 treatment significantly decreased AKR1B10 mRNA expression ( P<0.05), whereas no significant change was observed in the si-YTHDF1 group ( P>0.05).

Conclusion

RBM15-mediated m ⁶A modification promotes CSCC progression through the IGF2BP1-AKR1B10 axis. RBM15 may serve as a potential biomarker for predicting outcomes in patients with CSCC.

Cervical cancer (CC) is the fourth most prevalent malignancy among females worldwide, causing significant morbidity and mortality, with over 600 000 new cases and more than 340 000 deaths annually ^{[ 1- 2]}. Approximately 10% of global CC-related deaths occur in China due to its large population. Cervical squamous cell carcinoma (CSCC) is the most common subtype of CC, accounting for about 90% of cases ^{[ 3]}. Early-stage CC patients who receive timely treatment often have a favorable prognosis. However, outcomes for recurrent and metastatic cases remain unsatisfactory ^{[ 4- 5]}. Thus, identifying novel therapeutic targets is of great clinical significance for reducing CC mortality and improving patient survival.

N ⁶-methyladenosine (m ⁶A) modification plays a critical role in RNA stability, splicing, export, and nuclear localization, processes essential for tumor proliferation, invasion, and metastasis ^{[ 6- 7]}. This modification is dynamic and reversible, regulated by methyltransferases (writers), demethylases (erasers), and RNA-binding proteins (readers) ^{[ 8- 12]}. RNA-binding motif protein 15 (RBM15), an “m ⁶A writer”, facilitates m ⁶A complex recruitment to specific RNA sequences ^{[ 13- 14]}. In recent years, RBM15 has been identified as a potential tumor biomarker in various malignancies. For instance, RBM15 knockdown suppresses cell proliferation, invasion, and liver metastasis in colorectal cancer ^{[ 15]}. Silencing RBM15 inhibits differentiation and induces apoptosis in acute myeloid leukemia cells through Notch signaling ^{[ 16- 17]}. Similarly, depletion of RBM15 reduces proliferation, migration, and invasion in laryngeal squamous cell carcinoma, thereby inhibiting tumor progression ^{[ 18]}. Additionally, RBM15 knockdown has been shown to reduce tumorigenesis and progression in CC via m ⁶A mRNA modification ^{[ 19- 20]}. However, the mechanism by which RBM15 mediates m ⁶A modification in CSCC remains poorly understood.

Aldo-keto reductase family 1 member B (AKR1B) 10, a member of the AKR1B subgroup, is a cytosolic reduced nicotinamide adenine dinucleotide phosphate (NADPH)-dependent reductase that catalyzes redox reactions involving various endogenous compounds ^{[ 21- 22]}. Emerging evidence ^{[ 23- 25]} indicates that AKR1B10 is up-regulated in multiple malignancies, including CSCC. Despite this, the role of AKR1B10 and its relationship with RBM15 in CSCC carcinogenesis have not been explored. In this study, we identify AKR1B10 as a key downstream target of RBM15, and demonstrate that AKR1B10 promotes CSCC progression in an insulin-like growth factor 2 mRNA-binding protein (IGF2BP) 1-dependent manner.

1. Materials and methods

1.1. Ethics statements

The study protocol was approved by the Ethics Committee of Changsha Hospital for Maternal & Child Health Care (approval number: EC-20230717-28).

1.2. Bioinformatic analysis

Clinical information and RNA sequencing (RNA-seq) data for CSCC patients were obtained from The Cancer Genome Atlas (TCGA, accessed on 24 October 2024). m ⁶A methylation-related regulatory factors (including KIAA1429, HNRNPC, RBM15, METTL14, YTHDC2, METTL3, ZC3H13, WTAP, YTHDF1, Y THDC1, FTO, YTHDF2, and ALKBH5) were analyzed in 253 CSCC patients by using the “limma” and “ggplot2” R packages. Survival curves were plotted by using the “survival” and “survminer” R packages. For survival analysis, tumor samples were dichotomized into high-expression and low-expression groups for each gene ( RBM15, AKR1B10, and IGF2BP1) based on the respective median expression levels within the CSCC cohort. The associations between RBM15 expression and clinicopathological features were analyzed by using the R software (version 4.4.1) with the “ggpubr” and “limma” packages. Meanwhile, the microarray dataset GSE63514 (including 24 normal cervical epithelium tissues and 28 CSCC tissues) was selected and downloaded from the Gene Expression Omnibus (GEO) to validate the expression of RBM15, AKR1B10, and IGF2BP1.

1.3. Tissue samples

33 CSCC tissues and 8 adjacent normal tissues were collected from patients undergoing surgery at Changsha Hospital for Maternal & Child Health Care. Patients did not receive radiotherapy or chemotherapy before surgery. All tissues were pathologically examined and then immediately stored at -80 ℃.

1.4. RNA extraction and real-time fluorogenic quantitative PCR

Total RNA was extracted from tissue specimens or cells using the FastPure Cell/Tissue Total RNA Isolation Kit V2 (Vazyme Biotech, Nanjing, China), and complementary DNA (cDNA) synthesis was conducted using the HiScript II Q RT SuperMix (Vazyme Biotech, Nanjing, China) for real-time PCR, following the manufacturer’s instructions. RNA expression levels were measured using the Bio-Rad QX100 Droplet Digital PCR system (Bio-Rad, USA) with SYBR Green (Vazyme Biotech, Nanjing, China). All results were normalized to glyceraldehyde-3-phosphate dehydrogenase ( GAPDH), and the relative mRNA expression was quantified using the 2 ^-∆∆Ct method. All primers were provided by Tsingke Biological Technology (Beijing, China). The primer sequences used in this study are listed in Table 1.

Table 1.

Primer sequences

Gene symbol	Forward primer (5'-3')	Reverse primer (5'-3')
GAPDH	GGAGCGAGATCCCTCCAAAAT	GGCTGTTGTCATACTTCTCATGG
RBM15	AGCCGCGAGTATGATACCG	GCCCGAAGAATTTTTGGTGCTC
AKR1B10	CATGAAGTGGGGGAAGCCAT	CGTTACAGGCCCTCCAGTTT
IGF2BP1	CAAAGGAGCCGGAAAATTCAAAT	CGTCTCACTCTCGGTGTTCA
YTHDF1	ACCTGTCCAGCTATTACCCG	TGGTGAGGTATGGAATCGGAG

GAPDH: Glyceraldehyde-3-phosphate dehydrogenase; RBM15: RNA-binding motif protein 15; AKR1B10: Aldo-keto reductase family 1 member B10; IGF2BP1: Insulin-like growth factor 2 mRNA-binding protein 1; YTHDF1: YTH m ⁶A RNA binding protein F1.

1.5. Western blotting

Total proteins were extracted from tissues or cells using precooled radioimmunoprecipitation assay (RIPA) buffer (Beyotime, Shanghai, China) supplemented with protease and phosphatase inhibitors (Thermo Fisher Scientific, USA). Protein concentrations were determined using the bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific, USA). Equal amounts of protein were separated on 4% to 12% sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels (GenScript, Nanjing, China) and then transferred onto 0.45 μm polyvinylidenefluoride (PVDF) membranes (Millipore, USA). After blocking with 5% skim milk in Tris-buffered saline with Tween-20 (TBST) for 1 h, the membrane was incubated with the corresponding primary antibody at 4 ℃ overnight. After washing with TBST for 3 times, the membrane was then incubated with enzyme-labeled secondary antibody at room temperature for 1 h. β-Actin was chosen as the loading control. All the antibodies used in Western blotting (WB) are listed in Table 2.

Table 2.

Antibodies employed in this study

Antibodies	Source	Identifier	RRID
β-actin	Bio-Rad, USA	Cat#12004163	AB_2861334
RBM15	Abcam, UK	Cat#A4936	AB_2765940
Anti-m 6A antibody	Synaptic Systems, Germany	Cat#202003	AB_2279214
AKR1B10	Abcam, UK	Cat#ab57547	AB_2242446
IGF2BP1	ABclonal, Wuhan, China	Cat#A13581	AB_2760443
YTHDF1	ABclonal, Wuhan, China	Cat#A13260	AB_2760113

m ⁶A: N ⁶-methyladenosine; RRID: Research resource identifier.

1.6. Cell culture and cell transfection

The human CSCC cell lines SiHa and CaSki were provided by the Chinese Academy of Sciences Cell Bank (Shanghai, China). Both SiHa and CaSki cells were cultured in RPMI 1640 medium (Gibco, USA) supplemented with 10% (vol/vol) fetal bovine serum (Thermo Fisher Scientific, USA), penicillin (100 U/mL; Gibco, USA), and streptomycin (100 μg/mL; Gibco, USA) at 37 ℃ in a humidified incubator with 5% CO₂. When the cell confluence reached 70% to 90%, transfection was performed using 50 nmol/L small interfering RNA (siRNA) or a negative control siRNA (si-NC) with Lipofectamine 3000 reagent (Thermo Fisher Scientific, USA) following the manufacturer’s protocol. si-NC and siRNAs targeting RBM15, AKR1B10, IGF2BP1, and YTHDF1 were designed by RiboBio (Guangzhou, China). The siRNA sequences are listed in Table 3.

Table 3.

RNA interference and plasmid sequences

Gene symbols	Forward primer (5'-3')	Reverse primer (5'-3')
si-RBM15-1	GCCUGUUUCAUGAGUUCAAAC	GUUUGAACUCAUGAAACAGGC
si-RBM15-2	GAGGAUGAUCAGCGAGCUAAC	GUUAGCUCGCUGAUCAUCCUC
si-AKR1B10	UCAAUGUGCCGAUAUCCUGCA	CAGGAUAUCGGCACAUUGACU
si-IGF2BP1	ATCGGAGCTGAGGTGGAATA	CTCGGGGAAAGTAGAACTGC
si-YTHDF1	GCUCCAUUAAGUACUCCAU	UGCUGAUAGAUGUUGUUCC
si-YTHDF2	GCCCAAUAAUGCAUAUACUTT	GCUCUGGAUAUAGUAGCAATT

1.7. Cell proliferation assay, colony formation and 5-ethynyl-2’-deoxyuridine incorporation assay

Cell proliferation ability was determined by cell counting kit-8 (CCK-8) according to the manufacturer’s protocol. For the colony formation assay, 1.5×10 ³ cells were seeded into 6-well plates with 3 repetitions. After 2 weeks of incubation, the plates were fixed with paraformaldehyde and stained with 1% crystal violet for 10 min. For the 5-ethynyl-2’-deoxyuridine (EdU) assay, an EdU assay kit (RiboBio, Guangzhou, China) was used to evaluate the proliferation ability of CSCC cells according to the manufacturer’s protocol. The results were visualized by a fluorescence microscope.

1.8. Wound healing assay

Cells were seeded into 6-well plates (1×10 ⁵ cells per well). Wounds were made with a 10 µL sterile pipette tip when the cell confluence reached approximately 90%. Subsequently, the plate was washed three times with phosphate-buffered saline (PBS) to remove cell debris. Fresh Dulbecco’s modified Eagle medium (DMEM) without fetal bovine serum was added for further culture. Wound closure was imaged at 0 h and 24 h. Image-Pro Plus 6.0 was used to assess the gap distance quantitatively.

1.9. Cell migration and invasion assays

For migration and invasion assays, 24-well plates with transwell filter inserts (8 μm pore size; Corning, USA) were used. For the invasion assay, the transwell insert was pre-coated with diluted matrix. 1×10 ⁵ CSCC cells in serum-free medium were seeded into the upper chamber, and then the medium containing 10% fetal bovine serum was added to the lower chamber. After incubation at 37 ℃ for 48 h (migration assay) or 72 h (invasion assay), cells on the lower surface of the membrane were fixed and stained with crystal violet. The number of cells was counted in 5 random fields under a microscope.

1.10. RNA total m ⁶A quantification

The m ⁶A quantification assay was performed utilizing the m ⁶A RNA Methylation Quantification Kit (Abcam, UK) following the manufacturer’s protocol. The appropriate solutions were added to each well as instructed and gently mixed by tilting the plate from side to side. m ⁶A RNA capture was then carried out, ensuring that all residual wash buffer was thoroughly removed during each washing step. Signals were detected at the end when the color of the positive control wells changed to medium blue. To terminate the enzyme reaction, 100 μL of stop solution was added to each well, which caused the solution to change color to yellow. Finally, the absorbance was measured at 450 nm using a microplate reader.

1.11. Methylated RNA immunoprecipitation assay

For the methylated RNA immunoprecipitation (MeRIP) assay, the MeRIP Kit (BersinBio, Guangzhou, China) was utilized. Briefly, the RNA samples were fragmented into approximately 300 bp fragments, with 50 µL set aside as the input control. The remaining RNA was divided into 2 groups and incubated with m ⁶A antibodies or IgG antibodies (IgG served as the negative control). After immunoprecipitation, the enriched RNA was extracted, and quantitative PCR (qPCR) was performed following RNA reverse transcription for further experimental analysis.

1.12. Dual-luciferase reporter assay

The 3’ untranslated region (3’ UTR) sequence of AKR1B10 cDNA was cloned into luciferase reporter vectors (pcDNA3.1 vector containing both firefly and Renilla luciferase). For mutant reporter plasmids, 2 adenosines (A) at the m ⁶A sites were replaced with cytosines (C). RBM15-knockdown SiHa and CaSki cells were transfected with either wild-type or mutant AKR1B10 reporter plasmids. After 24 h, luciferase activity was measured using the Dual Luciferase Reporter Assay Kit (Vazyme Biotech, Nanjing, China).

1.13. RNA immunoprecipitation assay

The Magna RNA immunoprecipitation (RIP) Kit (Millipore, Germany) was used to conduct the RIP assay according to the manufacturer’s instructions. Briefly, magnetic beads were conjugated with anti-IGF2BP1 (ABclonal, Wuhan, China) or anti-rabbit IgG (Millipore, Germany) antibodies and incubated with the prepared cell lysates. The RNA-protein complexes were then eluted, purified, and analyzed by qPCR.

1.14. RNA stability assays

An RNA decay assay was conducted to assess RNA stability. SiHa and CaSki cells were cultured in 6-well plates and subjected to RBM15 knockdown. Actinomycin D (5 μg/mL; MedChemExpress, USA) was added to each well to inhibit transcription. Cells were collected at 0, 12, and 24 h, and the relative abundance of AKR1B10 mRNA (normalized to the 0-hour time point) was quantified.

1.15. Xenograft experiments

Four-week-old female BALB/c nude mice were obtained from Shanghai Experimental Animal Center, Chinese Academy of Sciences (Shanghai, China). 5×10 ⁶ SiHa and CaSki cells resuspended in 100 μL PBS were subcutaneously injected into the left flank of the mice (5 mice/group). Tumor sizes were measured regularly. After more than 4 weeks of observation, mice were sacrificed, and tumors were surgically dissected for histological analysis. Tumor volumes were calculated using the formula: 0.5× a× b ² (where a and b represent the tumor’s long and short diameters, respectively).

1.16. Statistical analyses

All statistical analyses were conducted using Perl (5.30.1), R software (version 4.4.1), and GraphPad Prism 9. Data are presented as means±standard deviations. The Student’s t-test and chi-squared ( χ ²) test were used to evaluate differences between 2 groups, while one-way analysis of variance was applied for comparisons among multiple groups. Kaplan-Meier analysis with the log-rank test was performed to assess survival differences between high- and low-level groups. P<0.05 was considered statistically significant.

2. Results

2.1. RBM15 was up-regulated and served as a prognostic marker in CSCC based on the TCGA dataset

Four m ⁶A regulators showed differential expression between CSCC tissues and adjacent normal tissues, including the m ⁶A writers METTL3 and RBM15, the m ⁶A eraser FTO, and the m ⁶A reader YTHDF1. Among these, FTO was significantly down-regulated in CSCC tissues, while RBM15, YTHDF1, and METTL3 were significantly up-regulated (all P<0.05, Figure 1A). Notably, RBM15 was the most differentially expressed. Consistent with this, analysis of GEO datasets also showed that RBM15 expression was significantly up-regulated in tumor tissues ( P<0.05, Figure 1B). Similarly, real-time PCR and WB also demonstrated that RBM15 mRNA ( P<0.05, Figure 1C) and protein levels ( Figure 1D) were significantly higher in clinical CSCC tissues compared with adjacent normal tissues. Elevated RBM15 expression was significantly associated with tumor grade and T stage (both P<0.05, Figure 1E and 1F). Additionally, survival curve analysis further revealed that higher expression levels of RBM15 in CSCC patients were strongly associated with poor survival ( P<0.05, Figure 1G).

Figure 1. RBM15 is significantly up-regulated in CSCC, and up-regulation of RBM15 is associated with poorer prognosis.

A: Differential expression of m ⁶A-related regulatory factors between CSCC tissues and adjacent normal tissues in TCGA. B: Expression of RBM15 mRNA between CSCC tissues and adjacent normal tissues in GEO. C and D: Real-time PCR (C) and WB (D) assays verified the expression of RBM15 in CSCC samples (including 33 CSCC tissues and 8 adjacent normal tissues). T1, T2, T3, N1, N2, and N3 represent tumor samples 1 to 3 and matched adjacent normal tissue samples 1 to 3, respectively. E and F: Correlation of RBM15 expression with tumor grade (E) and T stage (F). G1, G2, and G3 represent tumor grades 1 to 3, respectively. G: Kaplan-Meier analysis of overall survival in CSCC patients. ** P<0.01, *** P<0.001. m ⁶A: N ⁶-methyladenosine; CSCC: Cervical squamous cell carcinoma; TCGA: The Cancer Genome Atlas; GEO: Gene Expression Omnibus; RBM15: RNA-binding motif protein 15; WB: Western blotting.

2.2. Knockdown of RBM15 inhibited CSCC progression in vitro and in vivo

siRNAs targeting RBM15 effectively reduced RBM15 protein levels in SiHa and CaSki cells (all P<0.05, Figure 2A and 2B), indicating that the cells had been successfully transfected. CCK-8 assays and EdU staining demonstrated that knockdown of RBM15 significantly suppressed the proliferative activity of SiHa and CaSki cells (all P<0.05, Figure 2C- 2F). Additionally, colony formation assays revealed that silencing RBM15 markedly reduced the colony-forming capacity of SiHa cells compared with the si-NC group ( P<0.05, Figure 2G). Transwell assays and wound healing assays further showed that knockdown of RBM15 significantly inhibited the migration and invasion abilities of SiHa and CaSki cells (all P<0.05, Figure 3A- 3C). Consistent with the in vitro findings, tumor volumes and weights were significantly lower in the si-RBM15 group compared with the si-NC group (all P<0.05, Figure 3D and 3E).

Figure 2. RBM15 knockdown inhibits CSCC progression in vitro.

A and B: WB analysis of RBM15 protein levels in SiHa (A) and CaSki (B) cells transfected with siRNA si-RBM15 or si-NC. C-F: Effect of RBM15 knockdown on the cell proliferation of SiHa and CaSki cells was assessed using the CCK-8 assay (C and D) and EdU staining (E and F). G: Colony formation assays were performed to evaluate the impact of RBM15 knockdown on the colony-forming ability of SiHa cells. * P<0.05, ** P<0.01, *** P<0.001 vs the si-NC group. siRNA: Small interfering RNA; si-RBM15: siRNAs targeting RBM15; si-NC: Negative control siRNA; D: Optical density; CCK-8: Cell counting kit-8; EdU: 5-Ethynyl-2’- deoxyuridine.

Figure 3. RBM15 knockdown inhibits CSCC progression in vitro and in vivo.

A-C: Scratch wound healing (A and B) and Transwell (C) assays assessed the effects of RBM15 knockdown on cell migration and invasion in CSCC cells. D and E: Xenograft tumor models were used to evaluate in vivo tumor growth. The tumor volume growth curves (D) and tumor weights (E) were compared between the si-NC and si-RBM15 groups. * P<0.05 , ** P<0.01, *** P<0.001 vs the si-NC group.

2.3. RBM15 regulated the methylation of AKR1B10 via m ⁶A modification

As shown in Figure 4A, knockdown of RBM15 significantly reduced the overall m ⁶A levels in SiHa cells ( P<0.05). To identify candidate genes affected by RBM15-mediated m ⁶A modification, an m ⁶A RNA RIP microarray was performed, and the top 10 genes with differential m ⁶A enrichment were identified ( Figure 4B). Among these candidate genes, real-time PCR analysis revealed that knockdown of RBM15 significantly decreased the expression of AKR1B10 and GBA1, with AKR1B10 exhibiting the most pronounced decrease (all P<0.05, Figure 4C). Consequently, AKR1B10 was selected for further investigation. MeRIP assays confirmed that AKR1B10 mRNA was enriched with m ⁶A modifications ( P<0.05, Figure 4D), and RBM15 knockdown significantly reduced the m ⁶A modification levels of AKR1B10 mRNA ( P<0.05, Figure 4E). At the mRNA level, silencing RBM15 caused a down-regulation of AKR1B10 expression ( P<0.05, Figure 4F). Dual-luciferase reporter assays further demonstrated a direct interaction between RBM15 and AKR1B10. The luciferase activity in the si-RBM15 group was significantly lower compared with the si-NC group ( P<0.05), while mutation of the AKR1B10 m ⁶A site abolished this effect ( Figure 4G), indicating that RBM15 regulates the activity of the AKR1B10 3’ UTR through m ⁶A modification. Consistent with TCGA and GEO analyses, real-time PCR and WB results confirmed that AKR1B10 was significantly up-regulated in CSCC tissues compared with adjacent normal tissues (all P<0.05, Figure 4H- 4K). Survival analysis indicated that patients with high AKR1B10 expression had significantly shorter overall survival compared with those with low AKR1B10 expression ( P<0.05, Figure 4L).

Figure 4. RBM15 regulates the methylation of AKR1B10 via m ⁶A modification.

Figure 4. RBM15 regulates the methylation of AKR1B10 via m ⁶A modification (continued).

A: m ⁶A levels were measured in SiHa cells following RBM15 knockdown. B: Hierarchical clustering of mRNA with differential m ⁶A enrichment. C: Real-time PCR analysis of selected mRNA in SiHa cells transfected with si-NC or si-RBM15. D: MeRIP assay confirming m ⁶A modification enrichment in the AKR1B10 sequence. E: MeRIP assay showing changes in m ⁶A modification enrichment of AKR1B10 after RBM15 knockdown in SiHa cells. F: Real-time PCR analysis of AKR1B10 mRNA expression after RBM15 silencing. G: Dual-luciferase reporter assay showing luciferase activity in SiHa cells co-transfected with AKR1B10-Wt or AKR1B10-Mut along with si-NC or si-RBM15. H and I: AKR1B10 mRNA expression levels in CSCC tissues from TCGA (H) and GEO (I). J and K: Real-time PCR (J) and WB (K) analysis of AKR1B10 mRNA and protein expression in clinical CSCC tissues and adjacent normal tissues. L: Kaplan-Meier survival analysis of AKR1B10 mRNA expression in CSCC patients using TCGA data. * P<0.05, ** P<0.01, *** P<0.001. MeRIP: Methylated RNA immunoprecipitation; AKR1B10: Aldo-keto reductase family 1 member B10; Wt: Wild-type; Mut: Mutant.

2.4. RBM15-mediated m ⁶A modification of AKR1B10 mRNA enhanced its stability via IGF2BP1

Silencing IGF2BP1 significantly reduced AKR1B10 expression, consistent with its role in stabilizing target mRNA ( P<0.05, Figure 5A). However, knockdown of YTHDF1 did not produce a comparable effect ( P˃0.05, Figure 5B). RIP assays confirmed that IGF2BP1 protein directly binds to AKR1B10 mRNA ( P<0.05, Figure 5C). Interestingly, anti-m ⁶A immunoprecipitation followed by real-time PCR analysis showed that knockdown of IGF2BP1 decreased the m ⁶A modification of AKR1B10 mRNA and the relative expression of AKR1B10 mRNA ( P<0.05, Figure 5D- 5E). Furthermore, RIP assays demonstrated that knockdown of RBM15 weakened the interaction between IGF2BP1 and AKR1B10 mRNA in SiHa cells ( P<0.05, Figure 5F), indicating a strong interrelationship among RBM15, IGF2BP1, and AKR1B10 mRNA. A luciferase reporter assay further revealed that the luciferase activity in the si-IGF2BP1 group was significantly lower than in the si-NC group ( P<0.05), whereas mutation of the m ⁶A site in AKR1B10 abolished this effect ( Figure 5G). RNA stability assays further demonstrated that knockdown of either RBM15 or IGF2BP1 shortened the half-life of AKR1B10 mRNA (all P<0.05, Figure 5H and 5I). Subsequently, real-time PCR and WB analysis showed that IGF2BP1 mRNA and protein were significantly up-regulated in CSCC tumor tissues compared with adjacent normal tissues (all P<0.05, Figure 6A- 6D), consistent with TCGA findings. TCGA survival analysis indicated that CSCC patients with high IGF2BP1 expression had significantly worse prognosis ( P<0.05, Figure 6E).

Figure 5. IGF2BP1 is involved in m ⁶A methylation modification in CSCC.

A: Real-time PCR analysis of IGF2BP1 and AKR1B10 expression in SiHa cells after IGF2BP1 knockdown. B: Real-time PCR analysis of YTHDF1 and AKR1B10 expression in SiHa cells after YTHDF1 knockdown. C: RIP-qPCR results show the enrichment of IGF2BP1 binding to AKR1B10 m ⁶A modification sites. D: MeRIP assay identifying changes in m ⁶A modification enrichment of AKR1B10 after IGF2BP1 silencing in CSCC cells. E: Real-time PCR analysis of AKR1B10 mRNA levels after IGF2BP1 silencing in CSCC cells. F: RIP-qPCR results show the interaction between IGF2BP1 and AKR1B10 mRNA after RBM15 silencing. G: Luciferase reporter assay showing luciferase activity in SiHa cells co-transfected with AKR1B10-Wt or AKR1B10-Mut and either si-NC or si-IGF2BP1. H and I: RNA stability assay in SiHa cells showing the effects of RBM15 (H) or IGF2BP1 (I) silencing on the half-life of AKR1B10 mRNA. * P<0.05 , ** P<0.01, *** P<0.001. YTHDF1: YTH m ⁶A RNA binding protein F1; RIP: RNA immunoprecipitation; qPCR: Quantitative PCR; IGF2BP1: Insulin-like growth factor 2 mRNA-binding protein 1; si-IGF2BP1: siRNAs targeting IGF2BP1.

Figure 6. IGF2BP1 is significantly up-regulated in CSCC and is associated with a worse prognosis.

A and B: Expression levels of IGF2BP1 mRNA in CSCC tissues from TCGA (A) and GEO (B). C and D: Real-time PCR (C) and WB (D) analysis of IGF2BP1 mRNA and protein levels in clinical CSCC tissues and adjacent normal tissues. E: Kaplan-Meier survival analysis of IGF2BP1 mRNA expression in CSCC patients using TCGA data. * P<0.05 , *** P<0.001.

3. Discussion

For patients with CSCC, traditional treatment modalities such as surgery, radiation, and chemotherapy have long been the primary therapeutic approaches ^{[ 26]}. In recent years, advancements, including immunotherapy ^{[ 27]} and antibody-drug conjugates ^{[ 28]}, have introduced new opportunities for improving patient outcomes. Despite these advancements, the risk of recurrence and metastasis remains significant, posing a major challenge to long-term survival of patients with CSCC. Recurrence and metastasis are currently the leading causes of mortality in CSCC patients ^{[ 29]}. Therefore, there is an urgent need to elucidate the molecular mechanisms underlying CSCC progression to develop more effective strategies for reducing its morbidity and mortality.

In recent years, m ⁶A RNA methylation has been recognized as a critical regulator in various physiological and pathological processes, particularly in the pathogenesis of many cancers ^{[ 30- 32]}. Tumor transcriptomes are influenced by dysregulated m ⁶A-associated writers, readers, and erasers ^{[ 33- 34]}. Targeting the m ⁶A machinery and its associated transcriptome could potentially modulate the fate of key transcripts, affecting various aspects of cancer biology ^{[ 35- 36]}. In this study, the expression of m ⁶A-related genes in CSCC was examined using the TCGA database. The results revealed that the m ⁶A writers METTL3 and RBM15, as well as the m ⁶A reader YTHDF1, were significantly up-regulated in CSCC tissues, while the m ⁶A eraser FTO was markedly down-regulated compared with normal tissues. Among these regulators, RBM15 was identified as the most differentially expressed m ⁶A regulator, with its over-expression associated with higher tumor grades, advanced T stage, and poor prognosis in CSCC patients. These findings suggest that RBM15 may play a pivotal role in CSCC development and progression.

RBM15, a member of the SPEN (split-end) protein family located on chromosome 1p13.3, is known to directly bind U-rich regions on mRNA and recruit the methyltransferase complex to target transcripts ^{[ 20]}. Previous studies have identified RBM15 as an oncogenic factor in various tumors. For instance, the RBM15-megakaryoblastic leukemia-1 (MKL1) fusion produced by chromosomal translocation t(1;22)(p13;q13) is implicated in acute megakaryocytic leukemia ^{[ 37]}. RBM15 has also been shown to enhance the growth, metastasis, and invasion of clear cell renal cell carcinoma ^{[ 38]}, and it interacts with circular RNA catenin beta 1 (circ-CTNNB1) to promote osteosarcoma glycolysis and progression via m ⁶A modification ^{[ 34]}. In our study, we found that RBM15 expression was significantly higher in CSCC tissues compared with adjacent normal tissues, as confirmed by real-time PCR and WB. Functionally, knockdown of RBM15 suppressed CSCC cell proliferation, colony formation, migration, and invasion in vitro, as well as tumor growth in vivo. These findings indicate that RBM15 plays an oncogenic role in CSCC.

Subsequently, we further investigated the molecular mechanism by which RBM15 exerts its oncogenic effects in CSCC. Previous studies have revealed that RBM15 promotes tumor progression by regulating target mRNAs. For instance, in laryngeal squamous cell carcinoma, RBM15 targets TMBIM6 mRNA and recruits the reader protein IGF2BP3 to stabilize TMBIM6, thereby promoting tumor growth and metastasis ^{[ 18]}. Similarly, in liver cancer, RBM15 recruits IGF2BP1 to regulate YES1 mRNA, driving tumor development ^{[ 13]}. In our study, by analyzing the real-time PCR, MeRIP and luciferase reporter assay data, we found that AKR1B10 was one of the target genes directly regulated by RBM15 in CSCC. AKR1B10 is up-regulated and plays oncogenic roles in multiple cancers, such as breast cancer, laryngeal carcinoma, colon cancer, and nasopharyngeal carcinoma ^{[ 39- 42]}. However, its role in CSCC remained unclear. Our results revealed that AKR1B10 was also significantly up-regulated in CSCC tissues compared with adjacent normal tissues, consistent with TCGA data, and its high expression is associated with poor prognosis in CSCC patients.

m ⁶A readers are considered regulators of mRNA metabolism, while the IGF2BP family is associated with stability of m ⁶A-methylated mRNA ^{[ 43- 44]}. Among RNA-binding proteins (RBPs), IGF2BP1 is particularly important in tumorigenesis and progression ^{[ 45- 46]}. A previous study ^{[ 47]} showed that IGF2BP1 played a vital part in esophageal squamous cell carcinoma invasion and metastasis. Another study ^{[ 48]} found that high expression of IGF2BP1 was associated with poor prognosis in lung adenocarcinoma. Our data revealed a positive correlation among IGF2BP1, RBM15, and AKR1B10 expressions in CSCC. We thus speculated that IGF2BP1 was involved in the process of the m ⁶A modification of AKR1B10 mediated by RBM15. Using MeRIP and luciferase reporter assays, we verified this speculation. Furthermore, silencing IGF2BP1 or RBM15 significantly reduced the stability and expression levels of AKR1B10 mRNA. These findings confirmed that IGF2BP1 served as a key regulator of AKR1B10 stability through m ⁶A modification. Finally, we investigated IGF2BP1 expression in CSCC. Consistent with TCGA data, IGF2BP1 expression was significantly higher in CSCC tissues than in adjacent normal tissues, and its over-expression was associated with poor prognosis. These results further underscore the critical role of IGF2BP1 in CSCC progression.

In brief, our data confirmed the oncogenic role of RBM15 in CSCC malignant progression. Mechanistically, RBM15 and IGF2BP1 participate in the m ⁶A modification of AKR1B10, thus regulating the expression of AKR1B10 in CSCC. These findings enhance our understanding of the functional value of m ⁶A in tumor hallmarks, and have essential value in the diagnosis and treatment of CSCC.

Funding Statement

This work was supported by the Natural Science Foundation of Changsha (kq2403188) and the Scientific Research Project of Health Commission of Hunan Province (W20243254), China.

Conflict of Interest

The authors declare that they have no conflicts of interest to disclose.

AUTHORS’CONTRIBUTIONS

Chen Yibo Research design, experimental operation, and paper writing; Jiang Yuanna and Wang Qian Data collection and analysis; He Zhouxiao, Xiang Xuanang, and Yang Ruichen Paper modification; Cao Lanqin Research design, paper supervision and revision. The final version of the manuscript has been read and approved by all authors.

Note

http://xbyxb.csu.edu.cn/xbwk/fileup/PDF/2026081661.pdf