Insights into the role of the N6-methyladenosine reader IGF2BP3 in the progression of oral squamous cell carcinoma and its connection to cell-cycle control

Highlights

• •Overexpression of IGF2BP3 in promotes tumor development and metastasis in vitro and in vivo.
• •IGF2BP3 associated with G2/M transition phase through regulating cyclin A, cyclin E, and CDK-related genes.
• •IGF2BP3 is a potential biomarker for poor prognostic in OSCC.

Abstract

The genome of oral squamous cell carcinoma (OSCC) has been extensively characterized via bulk sequencing, revealing a multitude of genetic changes. The gene IGF2BP3, which encodes for the insulin-like growth factor 2 mRNA-binding protein 3, has been observed to be highly expressed in several types of cancer. This finding suggests that IGF2BP3 may play a significant role in the initiation and advancement of cancer. Nevertheless, the mechanisms by which IGF2BP3 contribute to OSCC are yet to be fully understood. In this study, we have observed that IGF2BP3 exhibits overexpression in OSCC. Based on our findings from bulk sequencing analysis, we have concluded that IGF2BP3 could potentially serve as a biomarker for predicting poor prognosis in OSCC. Moreover, it has been demonstrated that IGF2BP3 exhibits a significant association with the initiation and advancement of tumors both in vivo and in vitro. The evaluation of IGF2BP3 expression levels in relation to the cell cycle stage was conducted using single-cell RNA sequencing data. Tumor cells characterized by elevated IGF2BP3 expression demonstrated a higher percentage of cells in the G2/M transition phase. This study presents new findings indicating that the molecular target IGF2BP3 can serve as a prognostic indicator for tumors and has an impact on the development and progression of OSCC by influencing the regulation of the cell cycle.

Introduction

Oral squamous cell carcinoma (OSCC) is recognized as the primary cause of mortality associated with oral disease [1,2]. Significant patient heterogeneity has posed challenges in the advancement of enhanced diagnosis and innovative therapy [3], [4], [5]. In recent years, considerate advancements have been achieved in the elucidation of the functions of RNA alteration in cancer by bulk RNA sequencing [6]. However, the precise impact of RNA alteration on the formation and progression of OSCC has yet to be fully understood. Dynamic RNA modification is classified as an epigenetic phenomenon and is frequently observed in relation to cancer, malignancy, and resistance to therapeutic interventions [7,8].

There is a growing body of information that supports the notion that dynamic RNA modification pathways are dysregulated in various types of human malignancies, including gynecological tumors, bladder cancer, and GBM. These findings suggest that targeting these pathways could potentially serve as an effective strategy for cancer therapy [9], [10], [11]. RNA post-transcriptional modifications, including methylation, are a prominent area of interest and research in the field of epigenetics. Among the several types of methylation, N6-methyladenosine (m6A), 5-methylcytosine (m5C), N1 methyladenosine (m1A), and 7-methyladenosine (m7G) methylation are the most prevalent [12]. Among these modifications, m6A is the predominant form of methylation modification observed in eukaryotic RNA and has been extensively integrated as the most well-studied type of RNA modification. The regions surrounding m6A modification sites in messenger RNA (mRNA) exhibit a high degree of conservation, with this modification primarily occurring in patterns known as RRACH (where R represents a purine, A represents m6A, and H represents a non-guanine base) within the adenine molecule [13], [14], [15]. m5C modification is prevalent in several cellular contexts and plays significant functions in the regulation of gene expression and the maintenance of RNA stability [16,17]. The m1A modification has an impact on the initial nitrogen atom of the adenine base, resulting in a positive charge when subjected to physiological circumstances. Additionally, the m1A modification has an impact on the structural composition of ribosomes and the process of gene translation. This modification plays a crucial role in the regulation of gene expression and the control of cellular fate. Consequently, it has implications for the onset and progression of diseases [18]. According to existing research, it has been found that m7G is found in the 5′caps eukaryotic mRNA, as well as in certain internal locations within transfer RNA (tRNA) molecules and ribosomal RNAs in all domains of life [19,20].

The gene encoding insulin-like growth factor 2 mRNA-binding protein 3 (IGF2BP3) is situated on chromosome 7p15.3 in human genome [21]. This gene, initially discovered by Mueller-Pillasch et al. (1997), has been found to have increased expression levels in pancreatic cancer [22]. The encoded protein has a molecular weight of 69 kDa. The protein in question belongs to the IGF2 mRNA-binding protein family, which is distinguished by its unique structure consisting of six RNA-binding domains. The N-terminal section of the protein contains two RNA recognition motifs, while the C-terminal region consists of four K-homology domains. These domains are organized into three pairs and are connected by flexible linkers [23,24]. The available evidence in humans is rather constrained; however, it dose support the significant involvement of IGF2BP3 as an embryonic regulator. In line with this, it has been observed that fetal hematopoietic progenitors, namely megakaryocytes, have elevated levels of IGF2BP3 compared to their adult counterparts. Furthermore, the presence of IGF2BP3 has been found to play a role in preserving the molecular and phenotypic characteristics of fetal-type cells [25]. As of now, there is a lack of documented information on mutations in the IGF2BP3 gene in OSCC [26].

In order to investigate the molecular mediator(s) responsible for RNA post-transcriptional modifications in the advancement of OSCC, we obtained expression data for 60 genes associated with m6A/m5C/m1A/m7G alterations from The Cancer Genome Atlas (TCGA) database. These data were subsequently subjected to bioinformatic analytic tools for additional examination. Our study specifically elucidated the significant involvement of IGF2BP3 in the prognosis and metastasis of OSCC. We further validated its correlation with tumor growth and distant metastasis by conducting in vitro and in vivo. The application of single-cell RNA sequencing research has revealed that the gene IGF2BP3 is frequently upregulated in epithelial cells and has a regulatory role in the G2/M cell cycle phase.

Materials and methods

Samples collection and cell lines

OSCC and adjacent normal fresh tissues were collected from patients at Fudan university affiliated Zhongshan Hospital throughout the period of 2022–2023. This study was carried out with agreement of the Ethics Committee of Zhongshan Hospital (registration number: KY2022065). In light of the fact that all specimens utilized for testing purposes were anonymized, the Medical Ethics Committee has granted patients an exemption from the obligation to provide informed consent. The exemption was bestowed in adherence to the standards delineated in the Declaration of Helsinki, and the investigation received official approval. The SCC-9 and Cal-27 cell lines were acquired from the American Type Culture Collection (ATCC) through the Beijing Beina Chuanglian Biotechnology Institute. These cell lines were grown individually in F12 and DMEM media supplemented with 10% fetal bovine serum (FBS) (Gibo, Carlsbad, CA, USA). The cell lines were cultured in a humidified incubator at a temperature of 37 °C with a 5% concentration of CO₂.

Lentivirus construction and infection

Three human short hairpin RNA (shRNA) sequences were demonstrated in Supple. Table S1. The negative control shRNA (shNC) was cloned in pLKO.1-puro vector to endogenously down-regulate IGF2BP3. The coding sequences (CDS) portion of the human IGF2BP3 gene was artificially produced and inserted into the pLVX-puro vector in order to induce overexpression of IGF2BP3. The core plasmid that was manufactured underwent confirmation through DNA sequencing (Majorbio, Shanghai, China). Lentiviral production was conducted by transfecting Cal-27 cells with the lentiviral vector and packaging plasmids using Lipofectamine 2000 (Invitrogen, Shanghai, China). The overexpression and knockdown cells were subjected to screening using western blotting and real-time quantitative PCR techniques.

Western blot analysis

The cellular proteins from each group were extracted using RIPA lysis buffer containing 1% phenylmethanesulfonylfluoride (PMSF). Subsequently, equivalent quantities (20 μg) of protein, ascertained through use of the BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA), were subjected to separation using 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were subsequently deposited onto polyvinylidene difluoride (PVDF) membranes with a thickness of 0.45 mm, obtained from Solarbio, Beijing, China. The membranes were obstructed using a 5% nonfat milk solution for a duration of 1 h at room temperature. Subsequently, the membranes were subjected to incubation with anti-IGF2BP3, Cyclin D1 (1:1000, proteintech Group. Inc) rabbit polyclonal antibodies, and anti-β-actin rabbit polyclonal antibody (1:4000, Proteintech Group. Inc) at a temperature of 4 °C for a duration of 12 h. The proteins of interest were seen with the ECL Chemiluminescence system (Santa Cruz, Santa Cruz,CA).

RNA extraction and quantitative real-time PCR

The extraction of total RNA from transiently transfected cells was performed using Trizol Regent (Tiangen, Beijing, China). Subsequently, cDNA synthesis was carried out using a Superscript Reverse Transcriptase kit (Thermo, USA), following the directions provided by the manufacture. The information on primer sequences can be found in Supplementary Table S1. The SYBR Green PCR Master Mix Kit (TOYOBO) was utilized for the PCR amplification procedure. The expression data were computed utilizing the 2^−△△Ct approach and normalized by employing GAPDH as an internal reference to regulate the relative expression levels.

Immunohistochemistry (IHC) staining

Immunohistochemistry (IHC) was conducted on tissue sections that had been fixed in formalin and embedded in paraffin (FFPE). The sections underwent deparaffinization using xylene and subsequent hydration with ethanol at decreasing concentration (100, 90, 80, 75%) for a duration of 3 min each. Following this, the sections were subjected to microwave heating in sodium citrate buffer to facilitate antigen retrieval. Subsequently, the sections were treated with a blocking solution containing 5% bovine serum albumin (BSA) and then exposed to an anti-IGF2BP3 rabbit polyclonal antibody (dilution 1:200; ProtenTech Group., Inc., Wuhan, China) at a temperature of 4 °C for an overnight incubation period. Subsequently, the sections were treatment with a rabbit secondary antibody conjugated with horseradish perxidase (HRP) at a dilution 1:200 (ProteinTech Group, Inc.) for a duration of 60 min at ambient temperature. Following this, the 3,3′-diaminobenzidine development (DAB, Substrate Chromogen System; Dako, Denmark) and hematoxylin staining were employed. The images were acquired using an inverted microscope (Olympus IX71, Japan). The IHC score was determined by multiplying the staining intensity value with the staining percentage score. The staining intensity ratings were scaled on a range of 0–3, with each value signifying the absence of staining, accordingly. The percentage was scored using the subsequent scale: 0, < 1% positive cells, 1 as less than 25% positive cells, 2 as 25 - 50% positive cells, 3 as 50–75% positive cells, 4 as more than 75% positive cells. A positive cell was operationally defined as a cell exhibiting a nucleus that had faint, moderate, or strong staining. The photos were reviewed separated by three pathologists. The IHC scores of IGF2BP3 were employed for the purpose of categorizing the patients. Patients with IGF2BP3 levels were below the median were categorized as having ‘low expression’, while those with scores above the median were categorized as having ‘high expression’.

Cell proliferation assay

The cell proliferation rate was determined by doing the Cell counting Kit-8 (CCK-8) experiment. During the assay, cells that were either overexpressed or knocked down, as well as control cells, were separately seeded into 96-well plates at a density of 1000 cells per cell. Subsequently, these cells were grown for a duration of seven days. The plates were treated with the CCK-8 solution (10 μl/well) and incubated for 1 h on days one, three, five, and seven days. The measurement of absorbance at a wavelength of 450 nm was conducted to ascertain the quantity of cells. The experiment was conducted with three replicates for each assay.

Cell migration assay

The assessment of cell migration rates was conducted utilizing the Transwell migration assay. In a concise manner, a total of 2 × 10⁴ cells were introduced into the upper chamber, which was then filled with a medium containing 10% FBS. The separation of the top and bottom chambers was achieved by employing a permeable polycarbonate membrane that had been covered with Matrigel basement membrane matrix at a concentration of 1 μg/μl (BD Bioscience). After a 24-hour incubation period, the cells that had migrated were treated with 4% paraformaldehyde for fixation and subsequently stained using crystal violet. The migration assay was performed in triplicate. The quantification of migrating cells was conducted by enumerating the number of cells in three randomly chosen fields on each membrane. The mean value was utilized as the outcome.

Animal study

The in vivo investigations were conducted using male BLAB/c nude mice aged 4 to 6 weeks. A total of ten mice were acquired from the from Fudan university affiliated Zhongshan hospital Laboratory Animal Co. (Shanghai, China). The mice were subsequently housed in a controlled environment with regulated temperature, provided with unrestricted access to food and water, and subjected to a natural day/night cycle. The animal protocols utilized in this study were granted approval by the Ethical Committee of Fudan university affiliated Zhongshan hospital. These protocols were found to be in accordance with the principal standards outlined in the Guide for the Care and Use of Laboratory Animals.

In the xenograft model, the mice were injected subcutaneously with 200μL solution of Cal-27-EV and Cal-27-IGF2BP3 cells (1 × 10⁶ cells/mL). The measurement of tumor volume was conducted at regular intervals of 3 days, with a minimum of 3 distinct diameters being reordered using a caliper. The calculation of tumor volume was performed using the formula Volume (mm³) = Length × Width²/2. Following a period of 21 days of dietary intake, the mice were euthanized using cervical dislocation. The tumors were collected and subsequently weighed, followed by their collection for further study.

In order to investigate liver metastasis, the naked mice were randomly assigned to two groups, each consisting of eight individuals, for the metastasis model. The mice were subjected to anesthesia with intraperitoneal administration of a combination of ketamine (70 mg/kg) and xylazine (10 mg/kg). A 1-cm incision was performed in the upper left lateral abdomen, followed by the extraction of the spleen from the abdominal cavity. A cohort of Cal-27-control and Cal-27-IGF2BP3 cells, each consisting of 2 × 10⁶ cells, was intracelomically administered in a gradual manner with an insulin syringe. The duration of the injection process was roughly 3 min, during which the spleen exhibited signs of swelling and discoloration, appearing whitened. Following the administration of the injection, a cotton ball saturated with 75% alcohol was applied to the injection site and maintained in compression for a duration of 15 min. Subsequently, the spleen was excised, followed by the closure of the abdominal wall using sutures. The mice in the liver metastasis experiment were euthanized 21 days post-operation. The liver specimen was obtained, subsequently immersed in a 10% formalin solution for fixation, and then its weight was measured. Metastatic nodules were visually detected based on their color and appearance, and subsequently quantified using a dissecting microscope. The presence of metastatic liver nodules was subsequently verified through the use of haematoxylin eosin (H&E) staining.

Data processing of single-cell transcriptomic sequencing

We obtained eight samples of single-cell RNA-seq datasets for OSCC, among which six samples from Gene Expression Omnibus database (GSE172577) [27] and two samples from Fudan university affiliated Zhongshan hospital. The single-cell RNA-seq data was integrated in accordance with the methodology described in a previously reported [28]. Eight individual who were diagnosed with OSCC in the tongue were ultimately chosen for the study. Our main focus was on eliminating genes that were expressed in less than three cells and cells that had fewer than 200 genes. During the quality control process, cells that had mitochondrial reading above 5% and had a minimum of 7000 nFeature_RNA and 18,000 nCount_RNA were eliminated.

Data processing of bulk sequencing

The transcriptome expression data and clinical information for OSCC were obtained from the TCGA database (https://tcga-data.nci.nih.gov/tcga/), while the transcriptome expression data and survival information were acquired from the GEO database (http://www.ncbi.nlm.nih.gov/geo/). A total of 340 OSCC samples and 31 normal samples was selected form the TCGA-HNSC corhort after filtering based on tumor sites, including oral tongue, base of tongue, floor of mouth, buccal mucosa, hard palate, oral cavity, and alveolar ridge. The mRNA expression matrices of GSE31056, and GSE30784, were retrieved using the R package ‘GEOquery’ and utilized as the validation group for this study. The clinical information of the TCGA sets can be found in Table 1. Data visualization was performed using the R package ‘ggpubr’ and ‘VennDiagram’.

Table 1.

The clinical information of the TCGA sets.

Cohort	TCGA, OSCC(n = 340)
Gender
Male	235
Female	105
Age(years)
<65	200
≥65	140
Stage
Stage Ⅰ	222
Stage Ⅱ	55
Stage Ⅲ	63
Node
N0	123
N1	51
N2	107
N3	3

Cell cycle assay

The analysis of the cell cycle was conducted utilizing the TriCycle R package [29]. The expression matrix, after normalization, was initially transformed into ‘cell cycle space’ by the utilization of the ‘project_cycle_space’ function. The function ‘estimate_cycle_position’ was utilized to approximate the stages of the cell cycle. The interval from 0 to 0.6π was designated as the ‘G1/S’ phase, the interval from 0.6 to 1π as the ‘S’ phase, the interval from 1 to 1.25π as the ‘G2/M” phase, the interval from 1.25 to 1.75π as the ‘M’ phase, and the interval from 1.75 to 2π as the ‘G0/G1’ phase. The stage proportions of several clusters were computed and depicted using the R package ggpubr.

Statistical analysis

The experiments were conducted a minimum of three times. The data are presented in the form of the mean±SD. The data from two groups were analyzed using the Student's t-test, which can be applied to paired or non-paired data. The contingency tables were analysis and visualization tasks were performed using GraphPad Prism v9.3 software. The threshold for statistical was established at a significance level of p < 0.05.

Results

Screening the methylation related genes using bulk sequencing techniques

In order to obtain a deeper understanding of the molecular mechanisms behind the involvement of methylation in the development and advancement of OSCC, we incorporated a comprehensive set of seventy-three genes associated with m6A, m5C, m1A, and m7G [12]. Initially, we conducted a comparative analysis of gene expression levels in tumor tissues and healthy normal tissues. The analysis was performed using bulk sequencing data obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). As a result, a total of sixty genes were identified to be differentially expressed in OSCC (Fig. 1A). Principal Component Analysis (PCA) was utilized to effectively differentiate between the healthy control and tumor groups (Fig. 1B). Subsequently, differential expression analysis was conducted using the limma software. A total of 3392 genes were identified as exhibiting differential expression in the tissues of patients with OSCC compared to normal tissues. These genes met the statistical criteria of having a p-value less than 0.05 and an absolute of log2 fold change more than 1 ((p < 0.05 and |log2 FC| > 1). Among these genes, 1265 were found to be up-regulated, while 2142 were down-regulated, as visually represented in volcano plot (Fig. 1C). In the context of OSCC, it was anticipated that the genes related with methylation would be identified among the genes exhibiting differential expression. In order to address this objective, a Venn analysis was conducted to examine the overlap between genes that were differentially expressed in the TCGA dataset and a set of 73 genes associated with m5C, m6A, m7G, and m1A modification. The overlap region contains a total of seven genes, consisting of four genes (IGF2BP1, IGF2BP3, DNMT1, and DNMT3) that are up-regulated, and three genes (EIF4E3, NSUN7, and NUDT4) that are down-regulated to varying degrees, as depicted in Fig. 1D. In aggregate, two genes associated with m6A modification (IGF2BP1, IGF2BP3) and two genes associated with m5C modification (DNMT1, DNMT3B) were identified as having considerably high expression levels in OSCC. Conversely, two genes associated with m7G modification (EIF4E3, NUDT4) and one gene associated with m5C modification (NSUN7) had low expression levels in OSCC.

Fig. 1.

Overexpression of IGF2BP3 in OSCC tissue. (A) Differential expression of 60 genes of m7G/m6A/m1A/m5C genes in TCGA between tumor and adjacent normal tissues. (B) Principal component analysis (PCA) of differential gene analysis. Red represent normal samples, blue represents OSCC samples. (C) Vocal plot differential gene analysis of TCGA, red presents up-regulated genes and blue represents down-regulated genes. (D)Venn analysis of m7G/m6A/m1A/m5C genes and OSCC differential gene, up-regulated genes and down-regulated genes.

IGF2BP3 is upregulated in oral squamous cell carcinoma

A total of seven genes exhibiting statistically significant differential expression were selected for further analysis. This was achieved by comparing the gene expression profiles of tumor tissues and normal tissues from two independent cohorts, namely GSE30784 and GSE31056. The selected genes were next subjected to validation procedures. The study findings indicate that the levels of expression for four specific genes, namely DNMT1, DNMT3B, IGF2BP3, and NSUN7, were notably elevated in the oral tissues of patients diagnosed with OSCC in comparison to individuals without the condition. The difference in gene expression was shown to be statistically significant (p < 0.0001) in both cohorts (Fig. 2A and B). In order to assess the potential use of the aforementioned seven genes as potential biomarkers for OSCC, we conducted receiver operating characteristic (ROC) curve analysis on oral tissue samples obtained from both OSCC patients and individuals without the disease. The ROC curve is a graphical representation of the relationship between sensitivity and specificity by the vertical coordinate. The area under the curve (AUC) is a metric used to evaluate the accuracy of a diagnostic model, with a bigger AUC indicating a higher level of accuracy. The AUC values for IGF2BP3, NSUN7, and NUDT4 were 0.921, 0.808, 0.812, respectively. Conversely, the AUC values for the remaining genes were below 0.5 (Fig. 2C). To conduct a more comprehensive investigation into the relationship between the expression levels of seven specific genes and the prognostic of tumors, we conducted an analysis on patients from the TCGA datasets. Our findings indicate that individuals with greater expression levels of IGF2BP3 exhibited significantly shorter overall survival compared to those with lower expression levels (p = 0.029, Fig. 2D, Supplementary Fig. S1). The protein interactions network consisting of 34 proteins revealed that IGF2BP3 had direct interaction with YBX1, IGF2BP1, and HNRNPA2B1 (Fig. 2E). Hence, the findings of this study suggest a potential correlation between elevated levels of IGF2BP3 and worse clinical features and poor prognosis in patients with OSCC. Consequently, IGF2BP3 may serve as a promising biomarker for poor prognostic in OSCC.

Fig. 2.

Relationship between IGF2BP3 and the survival prognosis. (A and B) Differential expression of 7 filtered genes in GSE30784 (A) and GSE31056 (B) between tumor and adjacent normal tissues. (C) The ROC curve of the hub gene, the AUC values of IGF2BP3 is greater than 0.921. (D) Kaplan -meier's survival curve of TCGA cohort. High and low groups are determined by the mean IGF2BP3 expression. (E) The network plots the interaction of the m7G/m6A/m1A/m5C genes.

IGF2BP3 facilitated cellular proliferation and migration in vitro

Based on the aforementioned findings, it is postulated that IGF2BP3 potentially functions as an oncogene, facilitating the advancement and dissemination of OSCC. In order to gain a deeper understanding of the role of OSCC in cellular processes, we created cell lines SCC-9 and Cal-27 with overexpression and knockdown of IGF2BP3, respectively. The effective establishment was validated using Western blot analysis, as depicted in Fig. 3A. It is worth mentioning that the cells that exhibited an over-expression of IGF2BP3 had a notable increase in proliferation, as evidenced by the results of the CCK-8 assay (Fig. 3B). Conversely, the suppression of IGF2BP3 resulted in a considerable decrease in proliferation (Fig. 3B). The evaluation of cell migration was conducted utilizing the Transwell assay. The Cal-27 cells that overexpressed IGF2BP3 exhibited a greater proportion of cells that had moved. In contrast, the inhibition of cell migration was observed upon the knockdown of IGF2BP3 (Fig. 3C and D). The findings of this study provide empirical evidence that supports our initial hypothesis, which posits that IGF2BP3 plays a significant role in the progression of disease by facilitating cellular proliferation and migration.

Fig. 3.

IGF2BP3 promotes cell proliferation and migration. (A and B) Verification of overexpression and knockdown (A) SCC-9 and (B) Cal-27 cell lines by western blot analysis. (C and D) CCK-8 assays to assess cell growth. IGF2BP3 overexpressed and knockdown in (C) SCC-9 and (D) Cal-27 cell lines. (E) Transwell migration assay of Cal-27 cell lines. EV: empty vector.

Overexpression of IGF2BP3 promotes tumor development and metastasis in vivo

In this study, we investigated the potential role of IGF2BP3 in facilitating tumor development and metastasis in vivo. For the xenograft tests, nude mice were injected subcutaneously with Cal-27 cells that were either overexpressing IGF2BP3 or under control conditions. IHC of xenograft tumors revealed heightened expression levels of IGF2BP3 (Fig. 4A and B). It was revealed that tumors exhibiting an overexpression of IGF2BP3 demonstrated a considerably greater weight compared to tumors in the control groups (Fig. 4C and D, p < 0.01). The growth patterns of tumors on different time points of observation revealed that tumors exhibiting overexpression IGF2BP3 displayed a more pronounced rate of growth compared to control tumors (Fig. 4B). It was observed that the group of subjects transplanted with cells overexpressing IGF2BP3 exhibited a higher number of liver metastatic nodules compared to the group transplanted with control cells (Fig. 4E). The presence of these nodules was then verified through the use of H&E staining (Fig. 4F). Collectively, our findings offer more substantiation for the notion that IGF2BP3 amplifies neoplastic proliferation and dissemination in vivo experiment.

Fig. 4.

IGF2BP3 promotes tumor growth and metastasis in vivo. (A) Tumor growth in xenograft mice from IGF2BP3 overexpressed and control Cal-27 cell lines. (B) Tumor growth curves of control and IGF2BP3 overexpressed groups. (C) Weight of IGF2BP3 overexpressed tumors and control tumors of xenograft mice. (D) Representative IHC of xenograft tumors of control group and IGF2BP3 overexpressed group. (E) Livers from xenograft mice. Visible metastases on the liver are marked with arrows. (F) H&E staining of the normal livers and liver with metastases.

IGF2BP3 promotes the advancement of the cell cycle

In order to gain a deeper understanding of the underlying mechanism by which IGF2BP3 facilitates the advancement of OSCC, we conducted a single-sample gene set enrichment analysis (ssGSEA) on a dataset consisting of 348 tumor samples obtained from the TCGA database. This study involved the utilization of the 50 most typical hallmarks derived from the MsigDB database. The findings indicated a strong association between IGF2BP3 and various biological processes, including the estrogen response, adipogenesis, apical surface, bile acid metabolism, and epithelial-mesenchymal transition (Fig. 5A). Moreover, there exists a strong correlation between IGF2BP3 and some signaling pathways, namely NOTCH, WNT, and the reactive oxygen species pathway (Fig. 5A). In order to conduct a more comprehensive examination of the impact of heightened levels of IGF2BP3 on neoplastic cells, we incorporated single cell transcriptome sequencing data from a total of 32,621 OSCC cells from eight patients into our analysis. The cells were categorized into 24 clusters (Fig. 5B). The observed association between IGF2BP3, estrogen response, and NOTCH signaling was shown to be statistically significant (p < 0.001, Fig. 5A). After the clusters were annotated, it was shown that IGF2BP3 had a high level of expression in epithelial cells, primarily originating from primary tumor tissues. Furthermore, a strong correlation was found between IGF2BP3 expression and lymph node metastasis, as depicted in Fig. 5C and D. On the other hand, there exists a strong association between epithelial-mesenchymal transition and the initiation of the cell cycle, which holds true in both normal and abnormal biological processes [30,31]. The modulation of the cell cycle is also a crucial factor in the regulation of adipogenesis [32,33]. Consequently, it was postulated that IGF2BP3 participates in the regulation of the cell cycle. However, the importance of this correlation was confirmed through examination of single-cell sequencing data. Next, we employed Tricycle [29], a highly accurate and robust cell cycle stage estimator, to conduct cell cycle analysis on the 32,621 tumor cells. Clusters exhibiting elevated expression levels of IGF2BP3 were shown to possess a comparatively greater fraction of cells in the G2/M transition phase, while displaying a diminished proportion in the G1/S transition phase. In contrast, the fraction of cells in the M-phase remained generally consistent across the various clusters, while cells in other phases exhibited varying patterns (Fig. 5E). An link between IGF2BP3 and the proteins involved in the cell cycle was also discovered. The expression of IGF2BP3 demonstrated a notable and robust association with cyclin A, cyclin E, and CDK-related genes (Fig. 5F). The findings of this study suggest that IGF2BP3 primarily influences the G2/M transition phase of the cell cycle by modulating the expression of cyclin A, cyclinE, and CDK- related genes.

Fig. 5.

IGF2BP3 and the progression of the cell cycle. (A) The correlation of IGF2BP3 with ssGSEA results of 50 Hallmark gene sets of TCGA-OSCC datasets. (B) The UMAP plot showed 32,621 OSCC cells were grouped into 24 clusters. (C) Average expression of IGF2BP3 of different clusters. (D) The UMAP plot showed the annotation of 24 cluster. (E) 24 clusters were split by primary tumors (PT) and normal tissue (NT). (F) 24 clusters were split by whether lymph node metastases existed in the samples. (G) Mean expression of IGF2BP3 in different groups. (H) The proportion of cells at different cell cycle stages in different clusters. (I) The correlation between the expression of IGF2BP3 and cell cycle genes. LNM(-): Negative lymph node metastases, LNM(+): Positive lymph node metastases.

Discussion

Currently, cancers are the primary cause of mortality among individuals in the middle-aged and older population. OSCC poses a significant threat to human health and well-being as it is one of the most prevalent malignant tumors affecting the head and neck region. This is primarily attributed to its elevated rates of recurrence and metastasis [34]. Prior research has indicated that the development of tumors may be influenced by various variables, including tumor microenvironment (TME), aberrant gene expression, and immunological infiltration [35], [36], [37]. Nevertheless, the precise mechanisms by which methylation contributes to the etiology and progression of OSCC have not yet fully understood.

IGF2BP3 is a constituent of the IGF2 mRNA-binding protein family. Numerous investigations have demonstrated aberrant expression of IGF2BP3 in diverse malignancies, including lung adenocarcinoma, pancreatic, and bladder cancers [26,38,39]. The overexpressed of IGF2BP3 has been shown to enhance tumor cell proliferation, promote migration and invasion, inhibit cell apoptosis, and expedite tumor growth [24,40,41]. In a study conducted by Yu Y et al., it was observed that the circular RNA Hsa_circ_003258 has the ability to interact with IGF2BP3 and sequester miR-653-5p, thereby facilitating the spread of prostate cancer [42]. In a previous study, it was observed that IGF2BP3 possesses the capacity to facilitate the proliferation of cancer cells by interacting with circARID1 and subsequently generating a complex known as circARID1A-IGF2BP3-SLC7A5 RNA-protein ternary complex [38]. Several studies have indicated that the overexpression of IGF2BP3 may serve as a potentially valuable biomarkers for prognostic evaluation and therapeutic interventions across different types of malignancies, operating through diverse molecular mechanisms [43], [44], [45]. To date, there is a paucity of research conducted on the subject of IGF2BP3 in OSCC. Qin et al. [46] and Chou et al. [47] have demonstrated that m6A-RNA-methylation has a regulatory role in the processes of migration, invasion, and metastasis in OSCC. Within this particular family, IGF2BP2 has been implicated in the facilitation of proliferation and invasion in OSCC. In their study, Wang et al. conducted a bioinformatic analysis to validate the correlation between elevated expression of IGF2BP2 and enhanced cell invasion capability as well as tumorigenicity in human OSCC [48]. The current investigation has substantiated that the upregulation of IGF2BP3 was markedly observed in OSCC tumor tissue subsequent to the integration of TCGA and GEO datasets, single-cell transcript sequencing, and clinical samples. The findings of the study indicate a significant statistical association between IGF2BP3 and both lymph node invasion and the clinical stage of OSCC. Moreover, the atypical expression of IGF2BP3 is strongly correlated with unfavorable prognosis and reduced overall survival rate in OSCC [49]. Taken together, the findings of this study indicate that IGF2BP3 may function as an oncogene, contributing to the development of OSCC. Additionally, it is anticipated that IGF2BP3 could serve as a promising prognostic indicator for OSCC patients.

In order to investigate the molecular function and probable mechanism of IGF2BP3 in OSCC, the samples were categorized into two groups based on the expression levels of IGF2BP3: IGF2BP3-high and IGF2BP3-low. Subsequently, a comprehensive analysis was conducted using the GSEA software (version 4.1.0). The analysis of the C2 set from MSigDB revealed a significant association between IGF2BP3 and both the estrogen response and the WNT/β-catenin signaling pathway. Prior research has also demonstrated that IGF2BP3, functioning as a posttranslational mRNA, plays a crucial role in the regulation of β-catenin. The dysregulation of IGF2BP3 expression in renal tubular cells has been found to contribute to kidney injury through its regulation of the β-catenin signaling pathway, as reported in a previous study [50]. Collectively, the findings of this study suggest that IGF2BP3 has a role in both the tumor and immune-related KEGG pathway. The results of the current study indicate that the increased expression of IGF2BP3 plays a role in the modulation of OSCC. This modulation influences several signaling pathways such as NOTCH and WNT signaling pathways, ultimately leading to a negative prognosis for patients with OSCC.

In addition, an investigation was conducted to determine the potential impact of IGF2BP3 on the various stages of the cell cycle in OSCC. Prior research has established that the reduction of IGF2BP3 in clone cancer has effectively augmented the proportion of S phage within the entire cell cycle, hence impeding DNA replication and impeding the development of cancerous cells [51]. IGF2BP3 has been demonstrated to modulate the expression levels of cell cycle proteins, including CCND1, in clone cancer [51]. Furthermore, several pharmaceutical substances, including triptolide, have the capability to induce cell cycle arrest through the inhibition of IGF2BP3 expression [52]. The findings of our study consistently indicate that there is a direct relationship between the expression levels of IGF2BP3 and the proportion of cells at the G2/M transition. Furthermore, our results demonstrate a significant positive correlation between the expression levels of IGF2BP3 and those of CCNA2, CCNE2, and CDK2. These findings provide strong evidence for the association between IGF2BP3 and the cell cycle.

Nevertheless, it is important to acknowledge the limitation of our study. The clinical examination was conducted with a limited sample size of patients, potentially introducing bias into our conclusions. The deficiency was largely mitigated through our examination of reputable public databases. However, our study is innovative in elucidating the involvement of IGF2BP3 in OSCC and contributes to the advancement of precision medicine approaches for OSCC.

In order to facilitate future research, it would be advantageous to development animal models that exhibit conditional knockout or point mutations specifically targeting oral mucosal IGF2BP3. These models would serve to further our understanding of the phenotypic manifestations associated with IGF2BP3 mutations in vivo, and enable the investigation of potential deregulation of the cell cycle in the pathogenesis of OSCC.

Informed consent statement

Patient consent was waived due to all specimens tested were anonymous.

Institutional review board statement

The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Institutional Ethics Committee of Zhongshan Hospital of Fudan University. (Protocol code KY2022065 and date of approval: 2022-12-08).

Data availability

Transcriptome expression data and clinical OSCC information were downloaded from TCGA database (https://tcga-data.nci.nih.gov/tcga/), and transcrpiptome expression data and survival information were obtained from the GEO database through Accession Number GSE172577, GSE31056, and GSE30784. Part of the data can be obtained from the corresponding author.

CRediT authorship contribution statement

Dandan Song: Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft. Xiaofeng Dai: Software, Validation, Writing – review & editing. Minna Fu: Software, Validation, Writing – review & editing. Yang Sun: Resources, Visualization. Xingwen Wu: Resources, Visualization. Qianrong Zhou: Visualization. Wei Bi: Visualization. Jian Sun: Visualization. Fei Yang: Funding acquisition, Project administration. Youcheng Yu: Conceptualization, Funding acquisition, Project administration, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.