METTL3 Facilitates Colorectal Carcinoma Progression via Regulating m6A-CRB3-Hippo Axis

Colorectal carcinoma (CRC) is the third most common cancer and the second most common cause of cancer-related death worldwide. RNA N6-methyladnosine (m6A) and methyltransferase-like 3 (METTL3) play an important role in cancer. However, the roles of m6A and METTL3 in CRC progression are still elusive. Adenoma and CRC samples were applied to detect m6A and METTL3 levels, and tissue microarrays were performed to evaluate their associations with survival of CRC patients. The biological functions of METTL3 were investigated by CCK8, wound healing and transwell assays. M6A epitranscriptomic microarray, RNA stability and luciferase reporter assays were performed to explore the mechanism of METTL3 in CRC.


Results
m6A and METTL3 levels were signi cantly upregulated in both adenoma and CRC tissues, and the CRC patients with high m6A or METTL3 level had both shorter overall survival. METTL3 knockdown markedly inhibited the proliferation, migration and invasion of CRC cells. M6A epitranscriptomic microarray revealed that the cell polarity regulator Crumbs3 (CRB3) was the downstream target of METTL3. METTL3 knockdown markedly inhibited the degradation of CRB3 mRNA to increase the CRB3 expression. In addition, CRB3 level was also markedly reduced in both adenoma and CRC tissues, and the CRC patients with high CRB3 level had higher overall survival and disease free survival. CRB3 knockdown signi cantly promoted the proliferation, migration and invasion of CRC cells. Finally, CRB3 knockdown inhibited Hippo pathway, and increased nuclear localization of YAP.

Conclusions
The m6A and METTL3 levels were signi cantly increased in both adenoma and CRC tissues. The CRC patients with high m6A or METTL3 levels had shorter overall survival. Mechanistically, METTL3 regulated the initiation and progression of CRC via regulating m6A-CRB3-Hippo pathway.

Background
Colorectal carcinoma (CRC) is the third most common cancer and the second most common cause of cancer-related death worldwide [1]. Due to tumor metastasis and other complications, the mortality rate of CRC remains high. Therefore, it is urgent to elucidate the molecular mechanism and effective therapeutic targets of CRC.
In this study, we rstly veri ed the m6A level in both adenoma, which was precancerous lesions of CRC, and CRC tissues, and further demonstrated the function of METTL3 in CRC progression. Moreover, we identi ed crumbs3 (CRB3) as a downstream target of METTL3, and proved the function of CRB3 in CRC. Finally, we found that CRB3 regulated CRC progression by Hippo pathway. Therefore, our ndings elucidated the roles of m6A and METTL3 and provided a new treatment strategy against CRC.

Materials And Methods
Clinical tissue specimens Thirty CRC, thirty adenoma and thirty adjacent normal tissues (normal) were obtained during surgery in the Longhua Hospital a liated with Shanghai University of Traditional Chinese Medicine. The diagnosis of CRC and adenoma was con rmed based on pathological evidence. Tissues were snap-frozen in liquid nitrogen and stored at -80°C before detection. The study was approved by the Ethics Committee of Longhua Hospital (2019LCSY020), and informed consent was obtained from all participants. RNA m6A quanti cation assay M6A level was assayed using an RNA m6A quanti cation kit (ab185912, abcam, USA) according to previous study [26]. In brief, 200ng RNA was incubated for 60 min with the capture antibody; and then the detection antibody and enhancer solution were added. Finally, samples were incubated with developer solution for 10 min. The absorbance was detected at a wavelength of 450 nm.
Tissue samples from the normal, adenoma, and CRC groups were xed, and then cut into 4-µm sections for immunohistochemistry (IHC). Tissue microarrays (TMAs) were obtained from Shanghai Outdo Biotech Co., Ltd (Shanghai, China), and IHC was performed. In brief, samples were incubated with the m6A antibody (56593, CST, USA), and METTL3 antibody (ab195352, Abcam, USA) overnight at 4 • C. Subsequently, the secondary antibodies were incubated for 1 h at 37•C. Finally, samples were stained, and then imaged. The scores of IHC were performed.

CCK8 assay
After transfection, HCT116 and SW620 cells were seeded into 96-well plates at a concentration of 1×10 4 cells and cultured for 0, 24, 48 and 72 h. Then, 10 µl CCK8 was added to each well. After incubation at 37°C for 1 h, the absorbance value was detected at 450 nm.
Wound healing assay After transfection, HCT116 and SW620 cells were seeded in a six-well dish with a culture insert (Ibidi, Germany) at a concentration of 3×10 4 cells. After 24 h, the culture insert was removed, and the cells were washed twice with phosphate buffer. Then 2ml serum-free medium was added to each dish for 48 h. Images were captured, and the wound area was measured using ImageJ software (National Institutes of Health, USA).

Transwell assay
Six-well plates with 8-µm chambers (Corning, USA) were used to assess cellular migration (without Matrigel) or invasion (with Matrigel). Brie y, transfected HCT116 and SW620 cells were seeded in 6-well plates at a concentration of 1×10 5 cells. 200µl serum-free medium was added to the upper chamber, and 600 µl of medium with 30% fetal bovine serum was added to the lower chamber for 48 h. Then, the cells were xed with 4% paraformaldehyde for 30 min and stained with 0.1% crystal violet solution for 15 min. Five elds were randomly selected to calculate the number of migrating or invading cells.

Quantitative real-time PCR
Total RNA was extracted using TRIzol reagent (Ambion, USA). cDNA was synthesized using an EVM-MLV reverse transcription kit (Aikeri Biotech, Hunan, China). The ampli cation reaction was performed using the SYBR-Green qPCR kit (Thermo Fisher Scienti c, MA, USA). Gene expression was normalized using βactin. The primers were listed in Additional le 3 Table S1.

Western blotting
Cells were collected and lysed. Protein concentration was determined. The protein was separated and transferred to a PVDF membrane followed by incubation with 5% milk at room temperature for 1 h. The the secondary antibody was added and incubated at room temperature for 1 h, and protein expression was observed using a chemiluminescence gel imaging system (Tanon 5200, China).

Human m6A Epitranscriptomic microarray analysis
Total RNA was quanti ed using the NanoDrop ND-1000. The sample preparation and microarray hybridization were performed based on the Arraystar's standard protocols. Brie y, the total RNAs were immunoprecipitated with anti-m6A antibody. The modi ed RNAs were eluted from the immunoprecipitated magnetic beads as the "IP". The unmodi ed RNAs were recovered from the supernatant as "Sup". The "IP" and "Sup" RNAs were labeled with Cy5 and Cy3 respectively using Arraystar Super RNA Labeling Kit. The RNAs were combined together and hybridized onto Arraystar Human m6A Epitranscriptomic Microarray (8x60K, Arraystar). After washing the slides, the arrays were scanned by an Agilent Scanner G2505C.

Data Processing and Analysis
Agilent Feature Extraction software was used to analyze acquired array images. Raw intensities of IP (Cy5-labelled) and Sup (Cy3-labelled) were normalized. After normalization, the probe signals were retained for further "m6A methylation level" and "m6A quantity" analysis. "m6A methylation level" was calculated for the percentage of modi cation based on the IP (Cy5-labelled) and Sup (Cy3-labelled) normalized intensities. "m6A quantity" was calculated for the m6A methylation amount based on the IP (Cy5-labelled) normalized intensities. Differentially m6A-methylated RNAs between two comparison groups were identi ed by ltering with the thresholds of fold change (FC) > 1.5 and P value < 0.05.

RNA stability assay
Page 7/23 HCT116 cells were seeded in 6-well plates for 24h, and then treated with 5 µg/mL actinomycin D (MCE, USA) at the 0, 2, 4, 8, 24 h. Total RNA was then isolated by TRIzol (Ambion, USA) and analyzed by qPCR. The mRNA expression for each group at the indicated time was calculated and normalized by β-Actin.

Statistical analysis
Statistical analysis was conducted using SPSS 24.0 software. Data were assessed using a two-tailed Student's t-test. Survival curves were generated using the Kaplan-Meier method and compared using the log-rank test. Survival data were performed by univariate and multivariate Cox regression analyses. The distribution differences of the variables were analyzed by the Pearson's chi-square test. P < 0.05 was considered statistically signi cant.

Results
M6A level was markedly increased in both adenoma and CRC, and associated with poor prognosis A recent study has indicated that approximately 85% of CRCs may be transformed from adenomas [29], so we rstly checked the level of m6A in both adenoma and CRC tissues. The results showed that the levels of m6A were markedly increased between both adenoma and CRC groups compared with the normal group (Fig. 1a-c). TMA was performed to explore the correlation between m6A with survival in CRC patients. The results indicated that the level of m6A was also markedly increased in the CRC group compared with the normal group (Fig. 1d-e). And the CRC patients with high m6A level had shorter overall survival (Fig. 1f, Additional le 3: Table S2), which suggests that m6A level might serve as a prognostic marker of CRC.
METTL3 was markedly increased in both adenoma and CRC, and associated with poor prognosis M6A modi cation is regulated by RNA methyltransferase and demethylase, so we further detected the expression of methyltransferase (METTL3, METTL14 and WTAP) and demethylase (FTO and ALKBH5) in both adenoma and CRC. The results showed that expression of METTL3 was signi cantly increased in both the adenoma and CRC groups compared with the normal group (Fig. 2a-c), consistent with the results in the database (Additional le 1: Fig. S1A). METTL14 was signi cantly increased in only adenoma, not in CRC (Additional le 1: Fig. S1B). FTO was signi cantly decreased in only adenoma, not in CRC (Additional le 1: Fig. S1C). WTAP and ALKBH5 had no difference among three groups (Additional le 1: Figure S1D and 1E). The mRNA and protein levels of METTL3 were also markedly increased in CRC cells (Fig. 2d-e). To explore the correlation between METTL3 with survival of CRC patients, TMA was performed. The results indicated that the level of METTL3 was also markedly increased in the CRC group compared with the normal group ( Fig. 2f-g). And the CRC patients with high METTL3 level had shorter overall survival (Fig. 2h, Additional le 3: Table S3), which suggests that METTL3 level might also serve as a prognostic marker of CRC.

METTL3 drives CRC proliferation and invasion
To investigate the function of METTL3 in CRC, the knockdown of METTL3 was established in HCT116 and SW620 cells (Fig. 3a). The knockdown of METTL3 signi cantly inhibited the proliferation of HCT116 and SW620 cells (Fig. 3b). Transwell assays showed that METTL3 knockdown markedly inhibited migration and invasion of HCT116 and SW620 cells (Fig. 3c-d), and METTL3 knockdown also signi cantly decreased the migration speed of HCT116 and SW620 cells (Fig. 3e-f).

CRB3 was regulated by METTL3-mediated m6A modi cation
To investigate the potential mechanism of METTL3 in CRC progression, the epitranscriptomic microarray was performed, and data were analyzed. In the m6A methylation level, 363 differentially m6A methylated sites (DMS) were identi ed, including 353 upregulated DMS and 10 downregulated DMS (Fig. 4a, Additional le 4). And further analysis found that these 363 DMS belonged to 349 differentially m6A methylated genes (DMG), including 341 upregulated DMG and 8 downregulated DMG (Fig. 4b). In the m6A quantity level, 248 DMS were identi ed, including 191 upregulated DMS and 57 downregulated DMS ( Fig. 4c, Additional le 5). And further analysis found that these 248 DMS belonged to 224 DMG, including 175 upregulated DMG and 49 downregulated DMG (Fig. 4d). Subsequently, the main functions were identi ed using GO and KEGG, including DNA binding, RNA polymerase II transcription factor activity, Toll-like receptor binding, beta-Alanine metabolism, alpha-Linolenic acid metabolism, Nicotinate and nicotinamide metabolism (Fig. 4e-h).
Seven overlapping DMG between the m6A methylation level and quantity level were ltered by a Venn diagram (Fig. 5a), and qPCR analysis indicated that CRB3 level was markedly increased after METTL3 knockdown (Fig. 5b-c). The result of The Cancer Genome Atlas (TCGA) database revealed that CRB3 level was markedly downregulated in the CRC group (Additional le 2: Fig. S2B). The survival analysis indicated that the CRC patients with high CRB3 levels had higher overall survival and disease free survival (Additional le 2: Fig. S2C). Although METTL3 knockdown also promoted bridging integrator 1 (BIN1) level, the BIN1 expression was markedly increased in the CRC group (Additional le 2: Fig. S2A and 2D). The survival analysis indicated that there was no correlation between BIN1 levels and CRC patient survival (Additional le 2: Fig. S2E). In addition, luciferase reporters were performed to determine the effect of m6A modi cation on CRB3 expression. For the mutant form of CRB3, the adenosine bases in m6A consensus sequences (GGAC) were replaced by cytosine, thus m6A modi cation was abolished. The results showed that transcriptional level of wild-type CRB3 signi cantly increased after METTL3 knockdown, but not the mutation (Fig. 5d). RNA stability assay revealed that METTL3 knockdown markedly inhibited the degradation of CRB3 mRNA (Fig. 5e). Moreover, METTL3 mRNA expression in CRC tissues was negatively associated with CRB3 levels according to TCGA database (Fig. 5f). Recent studies have reported that YTHDF2 could target mRNAs via recognizing m6A motif in CRC [5,12], so we explored the effect of YTHDF2 on CRB3. The results showed that expression of YTHDF2 was also signi cantly increased in the adenoma and CRC tissues (Fig. 5g), and YTHDF2 knockdown markedly increased the level of CRB3 (Fig. 5h). The results showed that METTL3 regulated the expression of CRB3 by an m6Adependent manner.

CRB3 inhibited CRC proliferation and invasion
Our results also showed that CRB3 level was markedly decreased in the adenoma and CRC groups, which was consistent with the TCGA database ( Fig. 6a-b, Additional le 2: Fig. S2B). To further investigate the function of CRB3 in CRC, CRB3 knockdown was established in HCT116 and SW620 cells. The knockdown of CRB3 signi cantly promoted the proliferation of HCT116 and SW620 cells (Fig. 6c). Transwell assays showed that CRB3 knockdown markedly increased migration and invasion of HCT116 and SW620 cells (Fig. 6d-e), and CRB3 knockdown also signi cantly increased the migration speed of HCT116 and SW620 cells (Fig. 6f-g). The results fully proved that CRB3 regulated CRC progression.

CRB3 inhibited CRC progression by regulating Hippo pathway
Previous studies have proved that CRB3 could regulate Hippo pathway [30,31]. In the present study, we found that CRB3 knockdown decreased MST1, LATS1, MOB1 and YAP phosphorylation levels, and also decreased SAV1 level in HCT116 and SW620 cells (Fig. 7a-b). In addition, CRB3 knockdown markedly increased YAP protein level ( Fig. 7a-b). Studies have proved that YAP enters the nucleus and acts as an oncogene. So we also detected the level of YAP in the nucleus, and results showed that CRB3 knockdown also markedly increased YAP protein level in the nucleus of HCT116 and SW620 cells (Fig. 7c-d). The results indicated that CRB3 could regulate CRC progression by Hippo pathway.

Discussion
As a malignant tumor, the incidence of CRC has recently increased. Although the ve-year survival rate of CRC patients is 65% [32], the ve-year survival rate of patients with advanced-stage CRC is very low. Therefore, it is urgent to develop techniques for the treatment strategy of CRC, which would improve patient survival [27]. Studies have indicated that m6A modi cation plays a critical role in CRC [4,5,12,33]. But the change of m6A in both adenoma and CRC are still unknown. In addition, the relationship between m6A level and survival of CRC patients is also unknown. In this study, we rst found that m6A level was signi cantly increased in both adenoma and CRC tissues, indicating that m6A modi cation could be involved in the adenoma-CRC process. Further studies indicated that the CRC patients with high m6A level had shorter overall survival, which suggests that m6A level might serve as a prognostic marker of CRC.
M6A modi cation is mainly mediated by the m6A methyltransferases, demethylases and reader proteins, and regulates pre-mRNA splicing, miRNA processing, translation and mRNA decay [34]. In this study, we rst found that the expression of METTL3 was signi cantly increased in both adenoma and CRC, showing that METTL3 could be involved in the initiation and progression of CRC. And the CRC patients with high METTL3 level had shorter overall survival, which suggests that METTL3 might also serve as a prognostic marker of CRC. The result was consistent with a previous study [4,35]. Then we further veri ed the function of METTL3 in CRC. METTL3 knockdown inhibited the proliferation of HCT116 and SW620 cells, and also markedly inhibited migration and invasion of HCT116 and SW620 cells. These results indicated that METTL3 acted as an oncogene to promote the progression of CRC. Previous studies also proved that METTL3 plays an important role in a variety of cancers. METTL3 could regulate MALAT1 stabilization via m6A modi cation, and activate NF-κB activity to promote the malignant progression of glioma [36]. METTL3 increased miR-1246 level via the m6A modi cation, thus to promote non-small cell lung cancer progression [37]. Moreover, METTL3 regulated m6A modi cation of SPHK2 to contribute the progression of gastric cancer [38]. These ndings fully proved the role of METTL3 in cancers, including CRC. Therefore, METTL3 could act as a new treatment target for cancers.
According to m6A Epitranscriptomic microarray analysis, we found that CRB3 might be the downstream target of METTL3. METTL3 knockdown markedly reduced the m6A level of CRB3 and increased the CRB3 expression. Previous studies have found that the m6A consensus sequences are GGAC [5]. Our study found that METTL3 knockdown increased the transcriptional level of CRB3. When the adenosine bases of GGAC in CRB3 were replaced by cytosine, thus the transcriptional level of CRB3 was no change. In addition, further study found that METTL3 knockdown markedly inhibited the degradation of CRB3 mRNA. These results indicated that CRB3 was a target of METTL3 in CRC. CRB3 is a protein of cell polarity, and associated with contact inhibition [39]. Previous studies proved that CRB3 played an important role in cancer, including CRC [30,40,41]. In present study, we found that CRB3 level was markedly reduced in both adenoma and CRC, and the CRC patients with high CRB3 levels had higher overall survival and disease free survival. CRB3 knockdown signi cantly promoted the proliferation, migration and invasion of HCT116 and SW620 cells. These results indicated that CRB3 regulated CRC progression.
The depletion of CRB3 could regulate hippo pathway and lead to increased nuclear localization of YAP/TAZ [30,31,42]. The hippo pathway plays a crucial role in regulating the CRC progression [43][44][45][46]. In present study, we found that CRB3 knockdown inhibited Hippo pathway, and increased nuclear localization of YAP, suggesting that CRB3 regulated the CRC progression by Hippo pathway. Finally, our study found that METTL3 facilitated CRC progression via regulating m6A-CRB3-Hippo pathway, which was a novel mechanism for regulating CRC. Nevertheless, although we proved the regulatory mechanism of METTL3 in CRC,further studies would be needed. First, although previous study has found that selective rst-in-class catalytic inhibitor of METTL3 (STM2457) could act as a strategy against acute myeloid leukaemia [47]. But so far, the inhibitor of METTL3 has been not found in the treatment of CRC, and further studies which nd the inhibitor of METTL3 are needed. Second, although we found that m6A and METTL3 levels were markedly increased in both adenoma and CRC the function in the adenoma-CRC transition are needed in future study.

Conclusion
In summary, we demonstrated that m6A and METTL3 levels were signi cantly increased in both adenoma and CRC. The CRC patients with high m6A or METTL3 level had shorter overall survival. Mechanistically, METTL3 regulates the initiation and maintenance of CRC via regulating m6A-CRB3-Hippo pathway. These ndings provide a new perspective for targeted therapy of CRC.

Availability of data and material
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate
This study was approved by the Ethics Committee of Longhua Hospital (2019LCSY020), and the informed consent was obtained from all participants.  cells with METTL3 knockdown were measured; (b) The proliferation of HCT116 and SW620 cells were measured after METTL3 knockdown; Transwell assays were performed with METTL3 knockdown in HCT116 cell (c), and SW620 cell (d); Wound healing assay were performed with METTL3 knockdown in HCT116 cell (e), and SW620 cell (f). Data were presented as means ±SD. *P < 0.05, **P < 0.01, ***P < 0.001.  CRB3 was regulated by METTL3-mediated m6A modi cation. (a) Overlapping DMG between the m6A methylation level and quantity level were ltered by a Venn diagram; (b) The CRB3 mRNA level was measured after METTL3 knockdown; (c) The CRB3 protein level was measured after METTL3 knockdown; (d) Luciferase reporters were performed to determine the effect of m6A modi cation on CRB3 expression; (e) The CRB3 mRNA expression were detected with or without treatment of actinomycin D at indicated times; (f) Correlation between METTL3 and CRB3 expression in TCGA database for COAD was analyzed; (g) The YTHDF2 mRNA level was measured in both adenoma and CRC; (h) The CRB3 protein level was measured after YTHDF2 knockdown. Data were presented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Figure 6 CRB3 inhibited CRC proliferation and invasion. CRB3 expression in both adenoma and CRC was assayed by qPCR (a), and Immuno uorescence (b); (c) The proliferation of HCT116 and SW620 cells were measured after CRB3 knockdown; Transwell assays were performed with CRB3 knockdown in HCT116 cell (d), and SW620 cell (e); Wound healing assay were performed with CRB3 knockdown in HCT116 cell (f), and SW620 cell (g). Data were presented as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.