m6A methyltransferase METTL3 promotes retinoblastoma progression via PI3K/AKT/mTOR pathway

Abstract Retinoblastoma (RB) is a common intraocular malignancy in children. Due to the poor prognosis of RB, it is crucial to search for efficient diagnostic and therapeutic strategies. Studies have shown that methyltransferase‐like 3 (METTL3), a major RNA N (6)‐adenosine methyltransferase, is closely related to the initiation and development of cancers. Nevertheless, whether METTL3 is associated with RB remains unexplored. Therefore, we investigated the function and mechanisms of METTL3 in the regulation of RB progression. We manipulated METTL3 expression in RB cells. Then, cell proliferation, apoptosis, migration and invasion were analysed. We also analysed the expression of PI3K/AKT/mTOR pathway members. Finally, we incorporated subcutaneous xenograft mouse models into our studies. The results showed that METTL3 is highly expressed in RB patients and RB cells. We found that METTL3 knockdown decreases cell proliferation, migration and invasion of RB cells, while METTL3 overexpression promotes RB progression in vitro and in vivo. Moreover, two downstream members of the PI3K/AKT/mTOR pathway, P70S6K and 4EBP1, were affected by METTL3. Our study revealed that METTL3 promotes the progression of RB through PI3K/AKT/mTOR pathways in vitro and in vivo. Targeting the METTL3/PI3K/AKT/mTOR signalling axis could be a promising therapeutic strategy for the treatment of RB.

cells of the retina and often extends along the optic nerve into the brain or metastasizes distally to other organs. 4 Retinoblastoma can lead to devastating consequences, including blindness and even death. 5 Recently, Yang et al found that self-generated oxygen could mediate the therapy of tumours through scavenging in hypoxic conditions. 6 Studies show that establishment of hypoxia has been identified as a significant step in RB tumour progression. [7][8][9] Currently, although it is widely accepted that mutations in the 2 alleles of the RB tumour suppressor gene RB1 are triggers for this cancer, 10 12,13 Emerging evidence suggests that METTL3 has diverse functions in different cancers. METTL3 regulates oncogene expression by affecting mRNA processing, stability and translation, and it facilitates the progression of different types of cancers, including lung cancer, 14 breast cancer, 15 colorectal carcinoma, 16 bladder cancer 17 and hepatoblastoma. 18 In addition,

METTL3-depleted pancreatic cancer cells showed higher sensitivity
to anticancer reagents such as gemcitabine, 5-fluorouracil, cisplatin and irradiation, suggesting that METTL3 is a potent target for enhancing therapeutic efficacy. 19 However, the function of METTL3 in RB and the mechanism of METTL3 in the progression of RB remain unknown.
In this study, we investigated the function and mechanism of METTL3 in the pathogenesis of RB and revealed that METTL3 promotes RB progression in vitro and in vivo, suggesting that METTL3 may be a novel therapeutic target for RB treatment. Retinoblastoma   up-regulated lentivirus. The plasmids were gifts from Professor Shuibin Lin. 14 Lentiviral transfection was performed as described previously. 20 Briefly, cells (2 × 10 5 /mL) were seeded in six-well plates in 2 mL of culture media and then were infected with the lentiviruses.

| Plasmids and transfection
Polybrene (10 μg/mL; Sigma, USA, New York) was added to the lentiviruses to enhance infection efficiency. Pooled stable populations of Y79 and WERI-Rb-1 cells were generated by treatment with puromycin (20 mg/mL; Solarbio Life Science, China, Beijing) for 1 week.

| Cell Counting Kit-8 assays
For the CCK-8 (Cell Counting Kit-8) assays, 5000 cells/well were seeded in 96-well plates and cultured for 0, 24 and 48 hours. 10 μL of CCK-8 (Dojindo, Japan, Tokyo) solution was added to the cells for a 4 hours incubation at 37°C in 5% CO 2 atmosphere, and then, the absorbance at 450 nm was detected by a microplate reader (BioTek Instruments, Winooski, VT).

| Apoptosis
Cells (1 × 10 6 ) were seeded in six-well plates, and after being cultured for 24 hours, they were collected and centrifuged. 5 μL of PI(Propidium Iodide) 5 μL of FITC(Fluorescein Isothiocyanate) (BD Pharmingen, USA, New York) were added after the cells were pelleted and resuspended in 100 μL of 1× binding buffer. Apoptosis was detected by flow cytometry (LSRFortessa, BD, USA, New York) after 15 minutes of incubation in the dark at room temperature. The results were analysed by FlowJo 7.6.2. were analysed by ImageJ software.

| Colony formation assay
For the colony formation assay, 6-well plates were coated with poly-D-lysine (Solarbio Life Science, China, Beijing) overnight at room temperature, and then 2000 cells were seeded. Subsequently, the culture medium was changed every three days. After culturing for 14-21 days, the colonies were fixed with methanol (Boster Biological Technology, China,Beijing) before staining with 5% crystal violet (Solarbio Life Science, China,Beijing). Finally, the plates were photographed with a camera (Canon, Japan, Tokyo).

| Reverse transcription-quantitative polymerase chain reaction and western blotting
These assays were conducted as described previously. 21 The primer sequences of reverse transcription-quantitative poly-

| Subcutaneous xenograft mouse models
Male BALB/c nude mice (4-6 weeks, 18-20 g) were purchased from SPF (Beijing Biotechnology Co, Ltd, China, Beijing) and were fed in the Ophthalmic Animal Laboratory, Zhongshan Ophthalmic Center, Sun Yat-sen University. To establish a subcutaneous tumour model in nude mice, 2 × 10 7 Y79 cells (METTL3 knockdown group: shNC, shRNA1 and shRNA2; METTL3 up-regulated group: NC and METLL3) were resuspended in 1 mL of pre-cooled PBS, and 200 μL of the cell suspension was injected subcutaneously into the left side of the armpit to investigate tumour growth (4 × 10 6 per mouse).
Approximately 35 days later, the mice were euthanized by cervical dislocation, and the tumours were removed and weighed. All experimental procedures were conducted in line with the Guide for the Care and Use of Laboratory Animals and were approved by our institutional ethical guidelines. Tumour volume was calculated using the formula V = 1/2 × larger diameter × (smaller diameter) 2 .

| Haematoxylin and eosin staining
Subcutaneous neoplasms fixed with 10% formalin were imbedded in paraffin, cut into 5-mm-thick sections and mounted on glass slides.

| Statistical analysis
All experiments shown in the figures were conducted at least in triplicate, and IBM SPSS Statistics 21 software was used for statistical analysis. Data are presented as the average ±SD. Significant differences among the down-regulated METTL3 groups were determined by multiple LSD's multiple comparison test (one-way ANOVA).

| The expression of METTL3 in patient samples and different cell lines
To study the function of METTL3 in the regulation of RB, we first analysed the expression of METTL3 in RB patients and found that METTL3 is expressed in RB tumour samples ( Figure 1A). Since RB originates from the developing retina, we further compared the mRNA and protein levels of METTL3 in the normal ARPE-19 and RB cell lines Y79 and WERI-Rb-1 by quantitative real-time PCR and Western blot, respectively. Our results showed that the mRNA and protein levels of METTL3 in two different RB cell lines, Y79 and WERI-Rb-1, were higher than they were in normal ARPE-19 cells ( Figure 1B,C). Overall, our study revealed that METTL3 potentially regulates the progression of RB.

| Down-regulating METTL3 negatively regulates cell biological processes in tumour cells
To investigate the function of METTL3 in RB, we down-regulated METTL3 in Y79 and WERI-Rb-1 cells. The control group transfected with a vector (shNC) and the knockdown groups were two different shRNAs targeting diverse METTL3 sequences (shRNA1, shRNA2). The knockdown efficiency of METTL3 was verified by RT-qPCR and Western blot (Figure 2A-D). Then, we evaluated cell proliferation using a CCK-8 kit and found that cell proliferation was impaired upon METTL3 depletion ( Figure 2E,F). Furthermore, we showed that compared to the control levels, the apoptosis ratio increased by more than 50% in both METTL3 knockdown RB cell lines ( Figure 2G-J), suggesting that METTL3 is essential for RB cell survival. In METTL3-down-regulated RB cells, we found that the features of migration and invasion were obviously weakened, as measured by transwell assays (Figure 2K,N). The areas of migration and invasion were analysed by ImageJ software (Figure 2L-P). The colony formation assay also revealed that down-regulated METTL3

| Up-regulating METTL3 positively impacted the biological processes of tumour cells
We explored the influence of up-regulated METTL3 in RB cells.

| METTL3 influences the PI3K/AKT/mTOR signalling pathway in RB cells
We further studied the underlying mechanisms of METTL3 in RB regulation. Since METTL3 was reported to regulate mRNA translation in cancer cells, we decided to explore the role of METTL3 in the regulation of the PI3K/AKT/mTOR signalling pathway. We found F I G U R E 4 Methyltransferase-like 3 (METTL3) influences the expression of the PI3K/AKT/mTOR pathway. A, Down-regulated METTL3 decreases the expression of PI3K-p85, AKT, mTOR and P70S6K but increases 4EBP1 mRNA levels. B, Western blot data show that the expression of p-PI3K-p85, p-AKT, p-mTOR and p-P70S6K is decreased, but p-4EBP1 is increased; non-phosphorylated PI3K-p85, AKT, mTOR, P70S6K and 4EBP1 have no difference. C, The optical density ratio of the Western blot. D, Up-regulated METTL3 stimulates PI3K-p85, AKT, mTOR and P70S6K expression at the mRNA level but reduces 4EBP1. E, The protein levels of p-PI3K-p85, p-AKT, p-mTOR and p-P70S6K are elevated, but p-4EBP1 is decreased; the level of non-phosphorylated total proteins is steady. F, The optical density ratio of the Western blot. Data are shown as the average ± SD (n = 3). P < 0.05(*), P < 0.01(**), and P < 0.001(***) F I G U R E 5 Methyltransferase-like 3 (METTL3) regulates cell proliferation, apoptosis, migration and invasion via the PI3K/AKT/mTOR pathway. A, Rapamycin inhibits the expression of p-PI3K-p85, p-AKT, p-mTOR and p-P70S6K but elevates the expression of p-4EBP1. The total protein levels of PI3K-p85, AKT, mTOR, P70S6K and 4EBP1 were not different. B, The statistical histogram of the phosphorylated proteins/total proteins. C, The cell proliferation results show that the stimulatory function of METTL3 is lost after rapamycin treatment. D and E, The apoptosis ratio changes were higher in the METTL3 Rapa group than in the METTL3 group and NC group. F, The cell migration and invasion abilities in the METTL3 Rapa groups were weaker than those in the METTL3 groups and had no statistical difference with NC groups. G and H, Quantification of migratory and invasive areas was analysed by ImageJ software. Data are shown as the average ± SD (n = 3). P < 0.05(*), P < 0.01(**), P < 0.001(***), P < 0.0001(****) and P > 0.05(#). The scale bars represent 50 μm that METTL3 knockdown decreased the mRNA levels of PI3K-p85, AKT, mTOR and P70S6K, but it increased 4EBP1 mRNA expression ( Figure 4A). In addition, Western blot analysis revealed that there was no significant difference in the levels of non-phosphorylated PI3K-p85, AKT, mTOR and P70S6K and 4EBP1. Moreover, while phosphorylated PI3K-p85, AKT, mTOR and P70S6K decreased, p-4EBP1 increased in METTL3-down-regulated cells ( Figure 4B,C).
This indicated that the activity of this pathway was decreased in METTL3-depleted RB cells. On the other hand, we found that the mRNA levels of PI3K-p85, AKT, mTOR and P70S6K increased in METTL3-up-regulated cells, but 4EBP1 decreased ( Figure 4D).

| METTL3 regulates cell proliferation, migration and invasion through the PI3K/AKT/mTOR signalling pathway
We used 10 μmol/L rapamycin, an inhibitor of mTOR, to restrict PI3K/AKT/mTOR signalling in up-regulated METTL3 cells and then tested the changes in biological processes in cells. First, we examined the inhibitory effect of rapamycin on the PI3K/AKT/mTOR signalling pathway ( Figure 5A,B). Compared with the METTL3 group, the cell proliferation in the METTL3 Rapa group decreased, and there was nearly no difference from the NC group ( Figure 5C).
The apoptosis ratio was increased in the METTL3 Rapa group compared with that of the METTL3 group and NC group ( Figure 5D,E).
Cell migration and invasion were also repressed after rapamycin treatment ( Figure 5F-H). We found that in up-regulated METTL3 cells, the positive effects of RB cells were eliminated by inhibiting the PI3K/AKT/mTOR signalling pathway using rapamycin. These results indicate that METTL3 regulates the proliferation, apoptosis, migration and invasion of RB cells through the PI3K/AKT/ mTOR signalling pathway.

| METTL3 promotes the growth of RB cells in vivo
To determine whether METTL3 affects the growth of RB cells in vivo, we established a subcutaneous tumour model by injecting

METTL3-down-regulated Y79 cells and control cells (shNC, shRNA1
and shRNA2) into nude mice. Compared with the control (shNC), the tumours in the METTL3 knockdown groups (shRNA1 and shRNA2) did not grow or were gradually absorbed under the skin ( Figure 6A).
Next, we further investigated the function of METTL3 overexpression in RB tumorigenesis in the subcutaneous tumour model.
Our results showed that METTL3 overexpression significantly

| D ISCUSS I ON
In this manuscript, we first revealed that METTL3 is a critical factor promoting RB progression. Our results revealed that METTL3 Studies have shown that RB is prone to invade adjacent tissues. 4,5 Therefore, preventing metastasis is of great significance for con-  26 This pathway is pivotal in modulating cancer proliferation, migration and invasion by altering some genes that could impact biological processes in tumour cells. 27,28 Members of the PI3K/AKT/mTOR pathway were inactivated in METTL3down-regulated cells but activated in METTL3-up-regulated cells.
After treatment with an mTOR inhibitor, the PI3K/AKT/mTOR pathway was inactivated. At the same time, the function of METTL3, which can promote cell biological processes, disappeared. Our data suggested that METTL3 mediated the biological features of RB cells through the PI3K/AKT/mTOR/P70S6K/4EBP1 pathway. Therefore, we think that the METTL3/PI3K/AKT/mTOR signalling axis may be an efficient target for the treatment of RB.
In summary, we revealed that METTL3 could regulate oncogenesis in RB, suggesting that METTL3 is an oncogene in RB. Our findings uncovered novel insights into the function and mechanism of METTL3 in promoting RB progression.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data included in this study are available upon reasonable request by contact with the corresponding author.