The m6A reader IGF2BP3 Facilitates Gastric Cancer Progression Through Activating EMT via Upregulating MYC mRNA Stability

Background: N6-methyladenosine (m6A) RNA methylation plays an important biological role in cancer progression. Even so, the role of m6A modication in gastric cancer (GC) still needs further research. Methods: Firstly, based on the bioinformatics databases and human GC tissues analysis. Secondly, the IGF2BP3 expression in GC cells was measured by the quanticational real-time polymerase chain reaction and Western Blot. Then, the IGF2BP3 knockdown stable cells model was successfully constructed with the specic lentivirus-mediated short-hairpin RNA to explore the functions and mechanism of IGF2BP3 in GC. Next, the functions of IGF2BP3 on the cell phenotypes, including proliferation, invasion, migration, and Epithelial-mesenchymal transition process were claried by the Cell Counting Kit-8, transwell, and WB experiments. Subsequently, RNA Immunoprecipitation analysis and mRNA stability experiments were used to verify the relationship between IGF2BP3 and MYC. Finally, in the rescue experiment, MYC was overexpressed and transfected into IGF2BP3 knockdown cells to further detect the inuences on the cell phenotypes and the EMT process. Results: IGF2BP3 was up-regulated in GC. Meanwhile, IGF2BP3 had diagnostic and prognosis values for GC. Functionally, knockdown IGF2BP3 repressed gastric cancer cells proliferation, migration invasion and EMT process. Mechanically, IGF2BP3 activated the EMT process by improving the expression of MYC via combining with MYC mRNA and promoting its stability. Conclusions: Taken together, IGF2BP3 could activate the EMT process via increasing the MYC mRNA stability and expression to promote GC development, which provided insight into promising early diagnose and treatment for gastric cancer. by m(6)A Reader IGF2BP3 Drives ccRCC Progression. Elevated expression was related to the diagnosis and the poor prognosis of patients. readers expression in based on databases, expression GEO The IGF2BP3 and GES-1, BGC-823, HGC-27 and SGC-7901 by using qRT-PCR experiment, **P<0.01 vs. GES-1. The IGF2BP3 protein levels were measured in GES-1 and BGC-823, HGC-27, SGC-7901 cells. a The protein expression was analyzed by WB. b The band densities of specic proteins were quantied by ImageJ software, **P<0.01 vs. GES-1. C The diagram was the IGF2BP3 Knockdown lentivirus empty vector. D The transfection eciency was detected after the HGC-27 GC cells were transfected with shNC and IGF2BP3 shRNA, namely Blank group (HGC27-shNC) and IGF2BP3 knockdown group (HGC27-shI3). E The IGF2BP3 mRNA levels were measured after transfection with shIGF2BP3 and shNC in HGC-27 cells by the qRT-PCR experiment, **P<0.01. F The IGF2BP3 protein levels and the relative grayscale value were measured after transfection with shIGF2BP3 and shNC in HGC-27 cells. a The protein expression was analyzed by WB. b The band densities of specic proteins were quantied by ImageJ software, **P<0.01. G The cell proliferative capacity was evaluated by CCK-8 array. **P<0.01. The transwell experiment displayed that IGF2BP3 knockdown repressed the HGC-27 GC cells migration and invasion. I The relative HGC-27 cells invasion and migration levels were measured in HGC27-shI3 cells and HGC27-shNC cells. a The relative cell migration levels, **P<0.01. b The relative cell invasion levels, **P<0.01. **P<0.01. B The MYC protein expression level was evaluated in HGC27-shI3 cells and HGC27-shNC cells. a The protein expression was analyzed by WB. b The band densities of specic proteins were quantied by ImageJ software, **P<0.01. C The MYC mRNA was abundant in IGF2BP3 RIP of HGC-27 compared with IgG RIP was veried by RIP-qRT-PCR experiment, **P<0.01 vs. IgG. D The MYC transcripts stability in ActD-treated cells after transfected with MYC overexpression plasmids in HGC-27 cells was displayed by the qRT-PCR experiment, **P<0.01. E The diagram was the overexpression MYC knockdown plasmids empty vector. F The MYC protein levels and relative grayscale value expression after transfected with MYC overexpression plasmids in HGC-27 cells was detected. a The protein expression was measured by WB. b The band densities of specic proteins were quantied by ImageJ software, **P<0.01.


Introduction
Gastric cancer (GC) is a life-threatening malignancy [1]. The latest global cancer statistics reported that GC ranks fth and fourth of all malignancies for morbidity and mortality in 2020 [2]. Unfortunately, people with early-stage GC often have no symptoms, while most GC patients are always in the advanced stage with malignant proliferation and metastasis at the time of diagnosis, resulting in a poor prognosis [3].
Thus, founding novel biomarkers for diagnosis and therapeutic targets for treatment of GC is an urgent need.
As one of the critical "epitranscriptomics" processes, the N6-methyladenosine (m6A) modi cation has attracted wide attention [4]. The m6A modi cation is a dynamic and reversible process, which is installed by m6A writers (namely m6A methyltransferases), removed by m6A erasers demethylases (namely m6A demethylases), and recognized by readers (namely m6A-binding proteins) [5]. Accumulating evidences indicated that the disorder of m6A methylation closely associates with the incidence and development of various cancers [6,7]. Furthermore, it speculated that the m6A readers determined the fate of m6Amodi ed mRNAs, to be speci c, m6A-modi ed mRNAs is recognized by different m6A readers may lead to divergent biological functions [8]. At present, although the importance of m6A binding protein in GC cancer has gradually gained wide attention [9][10][11][12], the underlying role and epigenetic regulation of m6A reader in gastric cancer still need to be explored.
In this study, we uncovered that insulin-like growth factor 2 messenger RNA-binding protein 3 (IGF2BP3) expression level was upregulated in GC and related to poor gastric cancer progression. We further revealed the crucial functions of IGF2BP3 in promoting GC progression in vitro via strengthening MYC mRNA stability. Our work uncovered the role and the novel mechanisms underlying GC tumorigenesis, which may provide a novel diagnosis predictor, a prognostic biomarker for patients, and a promising new therapeutic target of GC patients.

Materials And Methods
Clinical Specimens 35 pairs of GC patient specimens were used in the study. Patients with the following criteria were excluded: with any antitumor treatment; with other cancers diagnoses. The research was approved by the Ethics Committee for Clinical Research of the University hospital a liated with Southeast University and the informed consent was acquired from every patient.
RNA isolation and the quanti cational real-time polymerase chain reaction (qRT-PCR) Total RNAs were extracted by using TRIzol reagent (Genstar, China), and then reverse-transcribed to complementary DNA (cDNA) by using PrimeScript RT Reagent (Genstar, China). RNA expression levels were detected by qRT-PCR through using the SYBR Premix Ex Taq (Genstar, China) and analyzed by 2−ΔΔ Ct method, all the results were normalized by β-actin. The primer sequences were used in this study are listed in the Table 1.  The process of the WB experiment was operated as described in the previous research [13]. The primary antibodies against IGF2BP3, MYC, Snail, N-cadherin, vimentin, mouse-anti-β-actin, Tubulin and the goat anti-rabbit/mouse horseradish peroxidase-conjugated secondary antibody (HRP) (goat-anti-rabbit) (goatanti-mouse) were purchased from Cell Signaling Technology (CST, USA). The band densities of speci c proteins were quanti ed by ImageJ software. The proteins were normalized to β-actin or Tubulin.

Cell Proliferation Assay
Cells (3000 cells/well) were seeded in the 96-well plate in 6 replicates, and incubated for 12h, 24h, 48h and 72h, respectively. After that, the proliferation viability of cells was measured by the CCK8 kit (malunion, China). The Optical density (OD) was detected at 450nm by a microplate reader.

Transwell Assays
The migration and invasion assays were operated with a 3422 Transwell chamber (Corning, USA) which were coated without the Matrigel mix (BD Biosciences, USA) (migration assay) or with the Matrigel mix (invasion assay). After the cells incubated for 24h, we used 4% paraformaldehyde to x the cells located on the lower surface of chamber for 15 minutes. And then, the cells were stained with 0.1% crystal violet dye for 15 minutes. After that, the stained cells were photographed and the number of stained cells were calculated in ve randomly selected elds.

RNA binding protein immunoprecipitation (RIP)
RNA immunoprecipitation experiments were operated by using RNA Immunoprecipitation (RIP) Kit (Merck Millipore, USA), and the operating steps according to the manufacturer's instructions. The analysis of coprecipitated RNAs were detected by qRT-PCR.
mRNA stability assays The stability of MYC mRNA was determined by applying actinomycin D (ActD; Final concentration: 2.5 μg/mL; Selleck, USA) mRNA stability assays in HGC27-shI3, HGC27-NC cells. At the indicated time point, mRNA degradation was analyzed by quantitative RT-PCR as described above, normalized to β-actin. The half-life (t1/2) of mRNA was calculated.

Immunohistochemistry (IHC)
Immunohistochemistry was operated on the para n sections. And they were dewaxed, dehydrated and rehydrated ordinally. After washing with phosphate-buffered saline (PBS), the para n sections were adding citrate buffer and then performed with high pressure repairing antigen. Next, the para n sections were blocked endogenous peroxidase activity by adding 0.3% hydrogen peroxide. Then, washing with PBS 3 times. After that, the para n sections were blocked by adding 10% goat serum. Then the para n sections were incubated with the corresponding primary antibody at 4°C for 14-16 hours. Next, the

IGF2BP3 was up-regulated in GC and related with the diagnosis and prognosis of GC
To discover the potential target m6A reader in GC, we analyzed m6A readers expression in GC at rst, the analysis of the TCGA database showed that IGF2BP3, IGF2BP1, IGF2BP2 and YTHDF1 were increased with fold change (|FC|>2 and P<0.05) in GC tissues (Fig. 1A). The analysis of ve GEO datasets indicated that IGF2BP3(|FC|>2 and P<0.002) and YTHDF3 (|FC|>2 and P<0.001) were up-regulated, while IGF2BP2 and YTHDC1 were downregulated(|FC|>2 and P<0.001 in gastric cancer tissues (Fig. 1B). In keeping with the databases mentioned above, we further obtained that compared with the normal counterpart, IGF2BP3 expression was distinctively up-regulated (FC=7.73, P=0.046) in human GC tissues. (Fig. 1C).
Then, based on the TCGA database and human GC tissues, we investigated the clinicopathological signi cance of ten m6A readers in GC. First, the diagnostic value of m6A RNA readers in GC was investigated. The ROC analysis results based on human GC tissues (Fig. 1D) and the TCGA database ( Fig. 1E) consistently demonstrated that IGF2BP3 and YTHDF1 had diagnostic values for GC. Next, the prognostic role of m6A readers in GC was explored (Fig. 1F). The Kaplan-Meier plotter analysis based on the TCGA database revealed that the GC patients with the higher expression level of IGF2BP3, YTHDF1, YTHDF2, YTHDC1, and HNRNPA2B1 had a signi cantly worse prognosis, while the GC patients with the higher expression level of HNRNPC and YTHDF2 groups had longer survival times. In keeping with the databases mentioned above, we further obtained that compared with the normal counterpart, IGF2BP3 expression was distinctively up-regulated (FC=7.73, P=0.046) in human GC tissues. para n sections were washed with PBS 3 times and were added with the anti-rabbit IgG-horseradish peroxidase (HRP) for 10 minutes, at room temperature. Then, the para n sections were washed with PBS 3 times again. After that, the para n sections were visualized by using 3,3'-diaminobenzidine tetrahydrochloride (DAB). The reaction was terminated when dyeing was appropriate. Then, washing with PBS 3 times again. After the para n sections were washed with PBS 3 times, the para n sections were counterstained with hematoxylin, washed with running water, dehydrated, and mounted with neutral gum.

Statistical analysis
The SPSS 22.0 software was used to analyze the data. The mean ± standard deviation (Mean ± SD) was applied to represent the data. The 2 -△△Ct method was used to analyze the qRT-PCR results. The receiver operating characteristic (ROC) curve was applied to test the diagnostic ability, and the Kaplan-Meier (K-M) curves were used to evaluate the prognostic prediction. The difference of m6A readers expression based on the TCGA database was analyzed by Wilcox analysis via using R version 3.5.1. The comparison of data was analyzed by using variance analysis, chi-square test, or non-parametric test when appropriate.
The correlation between the target genes was explored by Pearson correlation analysis. The test level was α=0.05 (two-sided), P-value < 0.05 represented that the differences were statistically signi cant.
Above all, with the described results of TCGA, GEO databases and human GC tissues, we summarized that IGF2BP3 expression is remarkably up-regulated in GC. Meanwhile, IGF2BP3 had diagnostic and prognostic values of GC patients, which might act as an oncogene in GC tumors. Therefore, IGF2BP3 was selected as the target molecule for the subsequent research.
In addition, in order to further clarify the mechanism of IGF2BP3 in GC, based on the m6A2 Target database (http://m6a2target.canceromics.org) to preliminarily predict the downstream target mRNA of IGF2BP3, we found that MYC might be the downstream target gene of IGF2BP3. Moreover, based on the analysis of the TCGA database and the GES96668 dataset, we discovered that there was a positive correlation analysis of the correlation between IGF2BP3 and MYC mRNA expression levels(P<0.05) (Fig. 1G). In addition, the IGF2BP3, MYC, EMT markers protein expression in human gastric cancer and para-carcinoma tissues was measured by Immunohistochemistry (IHC). The IHC assay manifested that compared with para-carcinoma tissue, IGF2BP3, MYC and EMT markers (N-cadherin, Vimentin, α-SMA and Snail) protein expression all increased in gastric cancer tissue (Fig. 1H). These results prompted that MYC might be a potential target downstream gene for IGF2BP3, and the EMT process plays a vital role in the GC development.
IGF2BP3 was up-regulated in GC cells.
To validate the expression of IGF2BP3 in GC, the IGF2BP3 expression was detected in GC cell lines by applying qRT-PCR and WB. The results showed that IGF2BP3 expression was generally up-regulated in three types of GC cells (SGC-7901, BGC-823 and HGC-27) ( Fig. 2A, B), compared with GES-1 cells. What's more, IGF2BP3 expression in HGC-27 cells was the highest. Therefore, the HGC-27 cell was selected for further biological function studies and mechanism studies.
IGF2BP3 promoted the proliferation, migration, invasion and the EMT process of HGC-27 cells in vitro.
To investigate the possible functions of IGF2BP3 in gastric cancer cells, we established the stable IGF2BP3-knockdown HGC-27 cells with transfection the speci c lentivirus-mediated short-hairpin RNA (shIGF2BP3, abbreviated as shI3: IGF2BP3 knockdown group and shNC: control group) (Figure2C, D), respectively. qRT-PCR and WB consistently indicated that the IGF2BP3 mRNA expression was reduced in the HGC-27-shI3 cells, compared with the HGC-27-shNC cells (Fig. 2E, F). CCK-8 assays demonstrated that knockdown IGF2BP3 distinctly inhibited the proliferation ability of HGC-27 cells (Fig. 2G). The transwell assays also indicated that knockdown IGF2BP3 distinctly signi cantly repressed the migration and invasion ability of HGC-27 cells (Fig. 2H-2I). Additionally, the EMT markers protein expression was assessed by WB to further clarify the functions of IGF2BP3 on the regulation of GC cells progression. The WB indicated that EMT markers (N-cadherin, Vimentin, α-SMA and Snail) in the IGF2BP3 knockdown group were down-regulated in HGC-27 cells (P<0.01) (Fig. 2J). Above all, IGF2BP3 could facilitate gastric cancer cells proliferation, invasion, migration and EMT process.
IGF2BP3 could combine with MYC mRNA to increase the expression and stability of MYC mRNA To understand the underlying mechanism of IGF2BP3 exerted tumor-promoting effects in GC. Based on the preliminary exploration of m6A2Target, TCGA database and IHC assay analysis, we chose MYC as the downstream target gene to further clarify the correlation between IGF2BP3 and downstream target MYC. The MYC expression level after transfection with shIGF2BP3 and shNC in HGC-27 cells was measured by qRT-PCR and WB experiments. The results demonstrated that knockdown IGF2BP3 signi cantly inhibited the MYC mRNA level (Fig. 3A) and the protein level in HGC-27 cells (Fig. 3B). To verify whether the MYC expression was in uenced by IGF2BP3 combination with MYC, we performed the RIP assay and con rmed the binding. The qRT-PCR analysis indicated that knockdown IGF2BP3 repressed the combination of IGF2BP3 protein binding with MYC mRNA, compared to immunoglobulin G (IgG) in HGC-27 (Fig. 3C). Meanwhile, ActD was applied to explore whether IGF2BP3 can affect the MYC mRNA stability in HGC-27. The results suggested that knockdown IGF2BP3 signi cantly reduces MYC mRNA stability and shortens MYC mRNA half-life in HGC-27 (Fig. 3D). Therefore, the regulatory role of MYC in IGF2BP3 knockdown HGC-27 cells was investigated through transfecting with MYC overexpression plasmids (Fig. 3E) into HGC-27-shI3 cells. The Western blot results presented that MYC overexpression effectively reverses the inhibition of MYC protein expression level in IGF2BP3 knockdown HGC-27 (Fig. 3F). In summary, these results strongly suggested that IGF2BP3 can increase the MYC mRNA and protein expression via recognizing and combining with MYC mRNA to increase the MYC mRNA stability in GC cells.

Up-regulated MYC effectively restored IGF2BP3 knockdown-induced inhibition of GC Cell Progression
To further investigate whether MYC participated in IGF2BP3-induced effects on GC cell progression, we conducted the rescue assays. In the rescue experiment, MYC was overexpressed and transfected into IGF2BP3 knockdown HGC-27 to detect the in uence of the recovery MYC overexpression on the cellular phenotypes and the EMT process. The results presented that overexpression of MYC could restore the proliferation (Fig. 4A), invasion (Fig. 4B), migration ability (Fig. 4B) and EMT process (Fig. 4C) of GC cells tumor progression. In a word, IGF2BP3 activated the EMT process by recognizing and binding MYC and thus increasing the mRNA stability and expression of MYC, resulting in facilitating proliferation, invasion, migration ability of GC cells.

Discussion
It has been reported that the tumor cells' epi-transcriptome was indeed disturbed by abnormal m6A regulator (writers, erasers and readers) playing important roles in GC development [14]. The transcriptome-wide RNA m6A landscape was specially recognized by various m6A readers, which can ultimately control the destiny of many vital transcripts, in uencing distinct aspects of cancer malignancy biology [15].
In the present research, we systematically analyzed ten m6A readers and discovered that the m6A reader gene IGF2BP3 might be a critical prognostic danger factor of gastric cancer. More importantly, we proved that IGF2BP3 activated the EMT process by enhancing the combination with MYC mRNA and improving MYC mRNA stability to increase the MYC expression, accelerating the GC cells development in vitro.
At rst, to identify the critical m6A readers in GC, we analyzed the expression level of ten m6A readers (IGF2BP1/2/3, YTHDF1/2/3, YTHDC1/2, HNRNPC and HNRNPA2B1) based on the TCGA, GEO databases and human GC tissues, we recognized that IGF2BP3 is the most frequently overexpressed m6Aassociated gene in GC, consistent with the previous study. In addition, by integrating the Kaplan-Meier plotter and the ROC analysis, we discovered that IGF2BP3 overexpression correlates with the early diagnostic and prognosis, consistent with previous research [16-18], suggesting IGF2BP3 might be a potential candidate biomarker for GC.
Next, given that IGF2BP3 has been involved in the human tumor progression [19][20][21][22][23]. We successfully constructed the stably knockdown IGF2BP3 in GC HGC-27 cell lines to evaluate the in uence of IGF2BP3 on the GC cell phenotypes. The results indicated that knockdown IGF2BP3 inhibited the HGC-27 cells proliferation, migration and invasion in vitro, keeping with Zhang J et al. demonstrated that IGF2BP3 promoted GC cell proliferation in HGC-27 GC cell lines [24]. Likewise, Zhou Y et al. reported that knockdown IGF2BP3 could inhibit cell proliferation and invasion ability in MKN28 GC cells, in good agreement with our results [25]. In summary, these results further elucidated that IGF2BP3 plays an important role in promoting GC cells proliferation, invasion and migration, implying that IGF2BP3 could lead to poor prognosis by accelerating tumor growth of GC.
In addition, to explore the mechanisms of IGF2BP3 in GC in depth. By integrating results of the previous researches, the m6A2Target databases prediction, the TCGA databases, the GEO databases and the immunohistochemical (IHC) assays, we concluded that there was a positive correlation analysis of the correlation between IGF2BP3 and MYC mRNA expression levels, which indicated that IGF2BP3 may promote the progression of GC through upregulating MYC mRNA. It was reported that the MYC played a key role in cell proliferation, migration and homeostasis of GC development [26-28], and MYC was also often elevated in GC and related to poor survival and clinical late-stage [29][30][31]. What is more, through the RIP-qPCR validation method and mRNA stability experiment, we further con rmed that IGF2BP3 could promote MYC expression by direct combination with MYC mRNA to enhance the MYC mRNA stability in GC cells, consistent with the previous research [32]. This is the rst validation of direct binding between IGF2BP3 and MYC with a positive correlation in GC cells, to the best of our knowledge. Furthermore, the rescue experiment was applied to explore whether the IGF2BP3-MYC axis can regulate gastric cells' proliferation, invasion and metastasis. In conclusion, the above results revealed that MYC overexpression could restore the suppressing function of GC cells proliferation, migration and invasion induced by IGF2BP3 knockdown. Prompting that IGF2BP3 could facilitate GC development by enhancing the stability and expression of MYC.
It's worth noting that earlier research had reported that epithelial-mesenchymal transition (EMT) was the primary malignant progression mechanism leading to cell metastasis and invasion in GC [33,34].
Moreover, recent researches have indicated that m6A modi cation can in uence the EMT process in GC cells [35]. Cheng et al. indicated that IGF2BP3-mediated m6A modi cation on ZBTB4 was participated in CS-induced EMT and lung cancer through EZH2 [36]. However, few studies were about whether IGF2BP3 could activate EMT processes in GC cells. In this study, immunohistochemical results displayed that IGF2BP3 expression was positively related to the EMT markers (N-cadherin, Vimentin, α-SMA and Snail) expression. In addition, we discovered that knockdown IGF2BP3 inhibited the EMT markers protein expression levels in HGC-27 cells. The result was the same as reported in breast cancer [37]. The above results suggested that IGF2BP3 could activate the EMT process in GC cells. Notably, it was reported that MYC could activate the EMT process to facilitate proliferation, invasion and migration [38]. Furthermore, through the rescue experiment, we revealed that MYC overexpression could restore the EMT process of GC. Summing up, all the results manifested that IGF2BP3 activated the EMT process to accelerate the GC cells proliferation, migration and invasion via improving MYC expression.
In conclusion, based on the comprehensive analysis of m6A readers expression of GC population, we discovered that m6A reader IGF2BP3 was signi cantly overexpressed in gastric cancer and related to the diagnostic and prognosis for GC patients. Functionally, IGF2BP3 could promote the GC cells proliferation, migration and invasion ability in vitro. Mechanistically, we demonstrated that IGF2BP3 could strengthen MYC mRNA stability through identifying and combining MYC, result in activating the EMT process, which can play a carcinogenic role in GC progression.

Conclusions
This study innovatively validated that the IGF2BP3-MYC axis played a vital role in activating the EMT process in GC development. In addition, our researches strongly prompted that IGF2BP3 might be a potential biomarker of early diagnosis and prognostic for GC treatment. Which further revealed the regulatory mechanism of GC development and provided a novel target for gastric cancer treatment.     The illustration summarizes the function and mechanism of the IGF2BP3-MYC axis in gastric cancer.