MECP2 promotes migration and invasion of gastric cancer cells via modulating the Notch1/c-Myc/mTOR signaling pathways by suppressing FBXW7 transcription

Background: Methyl-CpG-binding protein 2 (MECP2), an epigenetic regulatory factor, promotes the carcinogenesis and progression of a number of cancers. However, its role in the migration and invasion of gastric cancer (GC) as well as the underlying molecular mechanisms remain unclear. Methods: Immunohistochemistry (IHC), Western blot, and quantitative real-time PCR (qRT-PCR) were performed to measure the expressions of MECP2, FBXW7, c-Myc, mTOR and Notch1 in GC tissues and cell lines, respectively. The effects of MECP2 silencing and overexpression on GC cell migration and invasion were detected by wound healing assay and transwell assay. The mechanisms of MECP2-mediated migration and invasion were further investigated using chromatin immunoprecipitation sequencing (ChIP-Seq) and luciferase reporter gene assay. Results: In this study, we found that MECP2 facilitated the migration and invasion of GC cells. Investigation of the molecular mechanism revealed that MECP2 restrained FBXW7 transcription in GC by binding to the methylated CpG sites of FBXW7’s promoter region. MECP2 expression was remarkably negatively correlated with the FBXW7 level in GC tissues. FBXW7 expression was signicantly downregulated in GC tissues and cell lines, and low FBXW7 expression was correlated with the clinicopathologic features. FBXW7 repressed cell migration and invasion by regulating the Notch1/c-Myc/mTOR signaling pathways, and knockdown of FBXW7 reversed the effects of silencing MECP2. Moreover, MECP2 upregulated the Notch1/c-Myc/mTOR signaling pathways by inhibiting FBXW7 expression at the transcription level. Conclusion: This study demonstrates that MECP2 promotes migration and invasion of GC cells via modulating the Notch1/c-Myc/mTOR signaling pathways by suppressing FBXW7 transcription. The ndings suggest that MECP2 may be a novel effective therapeutic target for GC patients. promoter region. Western blot and qRT-PCR revealed a lower expression of FBXW7 in GC tissues, which was correlated with the clinicopathologic features. MECP2 expression was negatively correlated with FBXW7 expression. FBXW7 restrained the migration and invasion of GC cells. Further molecular mechanistic investigations found that MECP2 promoted cell migration and invasion via regulating the Notch1/c-Myc/mTOR signaling pathways by inhibiting FBXW7 transcription. In brief, the ndings in this study indicate that MECP2 may be a novel effective therapeutic target for GC.

MECP2 promotes migration and invasion of gastric cancer cells via modulating the Notch1/c-Myc/mTOR signaling pathways by suppressing FBXW7 transcription Lingyu Zhao Xi  [10,11]. It is an X-linked gene whose mutation results in neurological disorders, such as Rett syndrome [12]. It has been reported that MECP2 not only suppresses gene transcription, such as inhibiting BDNF and Cdkl5 expressions, via binding to the methylated CpG islands and recruiting co-repressors (e.g., histone deacetylases and Sin3A), but also promotes gene transcription, such as facilitating GIT1 expression, through binding to the methylated CpG dinucleotides and recruiting activators (e.g., CREB1) [13][14][15][16][17]. In recent years, emerging evidence indicates the role of MECP2 as a crucial oncogene in several cancer types [18]. Our previous studies have found that MECP2 is upregulated in liver cancer and facilitates cell growth [19], it enhances breast cancer proliferation via facilitating ubiquitination-mediated P53 degradation by regulating RPL5/RPL11 expressions [20], and it promotes GC cell proliferation and suppresses cell apoptosis via restraining FOXF1/MYOD1 transcription and enhancing GIT1 transcription by binding to the methylated CpG sites of the promoter regions of FOXF1/MYOD1 [21,22]. Cancer cell migration and invasion are related to the metastatic potential of the cells, which may be independent of cell proliferation rates. By far, the function of MECP2 in cancer cell migration and invasion has not been precisely studied. In particular, the role of MECP2 in the migration and invasion of GC remains unknown.
In response to the above mentioned gap, we investigated the effect and the molecular mechanism of MECP2 on GC cell migration and invasion. The results demonstrated that MECP2 facilitated the migration and invasion of GC cells. Chromatin immunoprecipitation sequencing (ChIP-Seq) and luciferase reporter gene assay showed that MECP2 inhibited F-box and WD40 domain protein 7 (FBXW7) transcription by binding to the methylated CpG site of FBXW7's promoter region. Western blot and qRT-PCR revealed a lower expression of FBXW7 in GC tissues, which was correlated with the clinicopathologic features. MECP2 expression was negatively correlated with FBXW7 expression. FBXW7 restrained the migration and invasion of GC cells. Further molecular mechanistic investigations found that MECP2 promoted cell migration and invasion via regulating the Notch1/c-Myc/mTOR signaling pathways by inhibiting FBXW7 transcription. In brief, the ndings in this study indicate that MECP2 may be a novel effective therapeutic target for GC. RNA extraction and quantitative real-time PCR (qRT-PCR)

Patients and specimens
Total RNA was isolated from GC tissues and cell lines using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol. Reverse transcription was ful lled using a commercial kit (Takara, Dalian, China). qRT-PCR was performed with the SYBR Green PCR kit (Takara Biotechnology, Takara, Dalian, China) using an IQ5 Multicolor qRT-PCR Detection System (Bio-Rad, USA). All qRT-PCR reactions were implemented in triplicates for each specimen. The primer sequences were listed in Additional le 1: Table S1. The mRNA expressions of relative genes were normalized to β-Actin and the 2 −ΔΔCt method was used to analyze the expression levels.
Immunohistochemical (IHC) staining IHC was performed on the tissue specimens from GC patients. The tissue samples were xed in 4% paraformaldehyde, embedded in para n and sectioned at a thickness of 4 µm. After depara nization and hydration with graded alcohol, antigen retrieval and blocking were performed. The sections were then incubated with corresponding primary antibody (anti-FBXW7, Santa Cruz, CA, USA) at a dilution of 1: 100, followed by incubation with biotinylated secondary antibody (Santa Cruz, CA, USA). Subsequently, the sections were detected using the 3, 3′-diaminobenzidine (DAB) kit and hematoxylin, and were scored by two pathologists independently without knowing the patient outcome. FBXW7 protein expression was assessed semi-quantitatively. High expression was de ned as the proportion of positive cells being > 25% in 5 random elds; otherwise, the expression was considered low. The integral optical density (ISO) in 5 randomly selected elds was quanti ed using ImageJ v1.8.0 (Wayne Rasband, USA).

Wound healing assay
The scratch wound healing assay was performed to detect the migration capacity of human GC cells. Differently treated BGC-823 and MKN-45 cells were seeded in 6-well plates and reached approximately 60% con uence. A single wound was gently and slowly scratched with a 10 µl disposable pipette tip across the center of a well. Suspension cells were removed by washing with PBS, and the remaining cells were incubated in fresh medium. The wound closures were observed at 0 and 48 h under a microscope (Olympus, Japan). The gap distance was quantitatively evaluated using ImageJ v1.8.0 (Wayne Rasband, USA). The migration rate of cells was quanti ed by setting average migration distance in control group as 100%. Results were obtained from all the three independent experiments.

Transwell migration and invasion assays
Transwell assays were performed on BGC-823 and MKN-45 cells. For examining cell migration, differently treated cells were planted in the upper chambers of the 24-well transwell (8-µm pore size; Millipore, Billerica, MA, USA) at a density of 2.0 × 10 4 cells in 100 µl serumfree medium, and 600 µl of 10% serumcontaining medium was added to the lower chambers. Cells were then incubated for 24 h at 37 °C in a 5% CO 2 atmosphere and then removed from the upper surface of the lter by scraping with a cotton swab.
The migrated cells adhering to the bottom membrane were xed with 4% paraformaldehyde and stained with 0.1% crystal violet. For examining cell invasion, 4.0 × 10 4 cells were maintained in the matrigel (15 µg/ lter; BD Biosciences, Franklin Lakes, NJ, USA) coated chamber for 48 h. Cells that had migrated and invaded to the lower surface of the lter were gured with ImageJ v1.8.0 (Wayne Rasband, USA) in 5 randomly elds. The relative migration and invasion rates were measured by setting the number of cells in the control groups as 1. Each experiment was repeated three times.
A cell cracker was used to sonicate cells to obtain nuclear lysates. The chromatin was sonicated into approximate 200-bp DNA fragments. The obtained lysates were divided into two portions. One portion was used as input, and the other was incubated with 5 µg antibodies against GFP, MeCP2, or IgG (Additional le 1: Table S4)

Western blot analysis
The total protein from human GC cell lines and tissues were lysed for whole-cell extracts with RIPA lysis

Statistical analysis
All data were analyzed using Graphpad 7.0 (Graph-Pad Software, USA) and SPSS 19.0 (Abbott Laboratories, Chicago, IL, USA). Results are expressed as Mean ± SEM from at least 3 independent experiments. Student's t-test was used to compare two independent groups. One-way ANOVA followed by individual comparisons using Dunnett's test was performed to analyze differences among more than two groups. Chi-square test was employed to detect the relationships between FBXW7 expression and clinicopathologic characteristics. Pearson's correlation analyses were conducted to estimate the association of MECP2 with FBXW7. P < 0.05 was considered statistically signi cant.

MECP2 is overexpressed in GC and closely related with the clinicopathological characteristics
Analysis based on the Cancer Genome Atlas (TCGA) data identi ed signi cant overexpression of MECP2 in GC tissues (P < 0.001, Fig. 1a). qRT-PCR on the GC tissue samples and paired adjacent normal gastric tissue samples from 166 GC cases collected in this study revealed a remarkably upregulated MECP2 mRNA expression in cancer tissues (P < 0.001, Fig. 1b). This result is consistent with that of our previous study which was based on 76 GC patients [21]. Further analysis suggested a correlation between MECP2 expression and clinicopathologic characteristics. TCGA data also revealed an association of high MECP2 expression with tumor histology and T stage (P < 0.01, Fig. 1c, d).

MECP2 facilitates the migration and invasion of GC cells
To investigate the causal role of MECP2 in GC progression, a MECP2-overexpressing plasmid was constructed and MECP2-targeting siRNAs (si-MECP2-1 and si-MECP2-2) were designed and synthesized.
qRT-PCR showed that MECP2-overexpressing plasmid observably upregulated the mRNA expression of MeCP2 in BGC-823 and MKN-45 cells (P < 0.001, Fig. 2a), while si-MECP2-1 and si-MECP2-2 signi cantly downregulated the MeCP2 mRNA level (P < 0.001, Fig. 2b). Immuno uorescent assay revealed that a higher level of exogenous MECP2 protein in the GFP-MECP2-vector group than in the GFP-vector group in both BGC-823 and MKN-45 cells (Additional le 2: Figure S1A, B). Our previous study demonstrated that MECP2 promoted cell proliferation and inhibited cell apoptosis in GC [21,22]. To further study the effect of MECP2 on GC progression, a wound-healing assay and transwell assays were performed to explore the effect of MECP2 on the migration and invasion ability of GC cells. The wound-healing assay showed that MECP2-overexpressing plasmid signi cantly enhanced the migration of the cells into the scratched area in GC BGC-823 and MKN-45 cells, while si-MECP2-1 and si-MECP2-2 markedly suppressed cell migration (P < 0.01, Fig. 2c, d). Consistently, transwell assays also revealed a promoting effect of MECP2overexpressing plasmid and an inhibitive effect of si-MECP2-1 and si-MECP2-2 on cell migration (P < 0.01, Fig. 2e, f). Meanwhile, MECP2-overexpressing plasmid remarkably facilitated cell invasion, and si-MECP2-1 and si-MECP2-2 signi cantly restrained cell invasion (P < 0.01, Fig. 2g, h). Further exploration of the molecular mechanism of MECP2 modulation in GC progression revealed that the protein expressions of MECP2, MMP-2 and MMP-9 were upregulated by MECP2-overexpressing plasmid but downregulated by si-MECP2-1 and si-MECP2-2 (Fig. 2i, j). These data suggested that MECP2 might promote the migration and invasion of GC cells through regulating MMP-2 and MMP-9 expressions.

MECP2 suppresses FBXW7 transcription by binding to its promoter
To explore how MECP2 might regulate GC cell migration and invasion, ChIP-Seq assay was performed in BGC-823 cells to identify the genes regulated by MECP2. Altogether 8128 ChIP-Seq peaks with various fold enrichments were acquired, among which 220 peaks were located in the promoter regions of genes (Additional le 3: Table S5). Then, MethPrimer was used to predict potential CpG sites within the MECP2 binding region. The results showed that the MECP2 binding region in the promoter region of FBXW7 contained a CpG site (Cg01181485, Fig. 3a). ChIP RT-PCR also veri ed that MECP2 could directly bind to the promoter of FBXW7 (Fig. 3b). The wild type (WT) and mutation type (MT) GFP-MECP2 plasmids were constructed. After they were transfected into BGC-823 cells, we ful lled ChIP RT-PCR with anti-GFP antibody. The nding revealed that exogenous MECP2 could also bind to the CpG site of FBXW7 promoter region (Fig. 3c). GFP plasmid (Ctrl), GFP-MT1 and GFP-MT2 could not bind to the site, but GFP-WT could bind to it (Fig. 3d). TCGA data showed that the methylation level of the CpG site (Cg01181485) was inversely correlated with the expression of FBXW7 in GC (P < 0.01, Fig. 3e).
Subsequently, a promoter reporter assay was ful lled to determine whether MECP2 could bind to the CpG site of the promoter region of FBXW7. The binding sequences of FBXW7's promoter region from ChIP-Seq were subcloned into the upstream of the luciferase gene in the pGL3 reporter plasmid (Additional le 1: Table S5). The luciferase activity was detected at 48 h after transfection with different plasmids in BGC-823 cells. It was found that the luciferase activity was evidently reduced in the pGL3-FBXW7-luc and pGL3-FBXW7-luc + Methylation groups as against the pGL3 group, and the luciferase activity in the pGL3-FBXW7-luc + Methylation group was signi cantly lower than in the pGL3-FBXW7-luc group (P < 0.01, Fig. 3f).
Among the groups of transfection rst with pGL3-FBXW7-luc plasmid and then with NC-siRNA, si-MECP2-1, si-MECP2-2, control plasmid vector, MECP2-overexpressing plasmid, dimethyl sulphoxide (DMSO), and methylation inhibitor 5-aza-2′-deoxycytidine (Aza), respectively, the luciferase activity was observably enhanced in the si-MECP2-1 and si-MECP2-2 groups as compared with in the NC-siRNA group, remarkably increased in the methylation inhibitor 5-aza-2′-deoxycytidine (Aza) group as against the DMSO group, and signi cantly decreased in the MECP2-overexpressing plasmid group as against the vector group. The groups of rst transfection with pGL3-FBXW7-luc + Methylation plasmid and then the further treatment generated similar results, but the activity was lower than that in the pGL3-FBXW7-luc groups (P < 0.01, Fig. 3g). The MECP2 mRNA expression was signi cantly negatively correlated with the FBXW7 mRNA level in GC tissues (r = 0.6302, P < 0.0001, Pearson's correlation, Fig. 3h). Furthermore, the FBXW7 mRNA expression in BGC-823 and MKN-45 cells remarkably decreased after transfection with MECP2overexpressing plasmid and increased after treatment with the methylation inhibitor Aza, si-MECP2-1 and si-MECP2-2 (Additional le 4: Figure S2A-C). The protein expression of FBXW7 in the cells was also inhibited by MECP2-overexpressing plasmid (Fig. 3i), but facilitated by the methylation inhibitor Aza, si-MECP2-1 and si-MECP2-2 (Fig. 3j, k). These ndings suggested MECP2 as a transcription regulator of FBXW7 in GC cells.

FBXW7 is frequently downregulated in human GC tissues and is correlated with the clinicopathologic features
To further verify the ndings, FBXW7 expression in the tissue samples was measured. TCGA data revealed a remarkable downregulation of FBXW7 expression in GC tissues as compared with in normal gastric tissues (P < 0.001, Fig. 4a). In line with this, our statistical analyses showed that patients with lower FBXW7 levels had poorer overall survival (P < 0.05, Fig. 4b). The mRNA expression of FBXW7 was signi cantly downregulated in GC tissues (P < 0.0001, Fig. 4c). This trend was further veri ed by measuring the FBXW7 expression in GC BGC-823 and MKN-45cells. The mRNA expression of FBXW7 in BGC-823 and MKN-45 cells was evidently lower than in normal human gastric epithelial cells (GES-1) (P < 0.001, Fig. 4d). The protein expression of FBXW7 was markably lower in GC tissues than in normal gastric tissues (Fig. 4e) Table 1). The expression was not associated with age, gender, histology, and liver metastasis. Western blot also detected decrease of FBXW7 protein expression in 5 pairs of GC and ajacent normal tissues, and GC BGC-823 and MKN-45 cells (Fig. 4f, g). To investigate the potential molecular mechanisms of MECP2-regulated GC cell migration and invasion, mRNA and protein levels of the FBXW7-related downstream genes were measured by qRT-PCR and Western blot. FBXW7 is a substrate recognition subunit of the SKP1-CUL1-F-box protein (SCF) E3 ubiquitin ligase complex, which plays a crucial role in tumorigenesis and cancer progression by promoting degradation of ubiquitination-mediated oncoproteins, including mTOR, c-Myc, c-Jun, and Notch1 [24][25][26]. The results of qRT-PCR showed that the mRNA expressions of Hes1, MMP-2 and MMP-9 signi cantly increased in BGC-823 and MKN-45 cells after transfection with MECP2-overexpressing plasmid, and they dramatically decreased after the treatment with the methylation inhibitor Aza, si-MECP2-1, and si-MECP2-2 (P < 0.001, Fig. 5a-c). However, there were no signi cant differences in the mRNA levels of c-Myc, mTOR and Notch1 after the treatment (Additional le 5: Figure S3A-C). The Western blot analysis revealed that the protein expressions of c-Myc, mTOR, Notch1, Hes1, MMP-2 and MMP-9 were upregulated in whole-cells after transfection with MECP2-overexpressing plasmid, and they were downregulated after the treatment with methylation inhibitor Aza, si-MECP2-1 and si-MECP2-2 (Fig. 2i, j; Fig. 5d-f). In addition, the change of nuclear NICD1 protein level was consistent with the change of whole-cell Notch1 protein level (Fig. 5d-f). Based on these ndings, MECP2 might regulate the Notch1/c-Myc/mTOR signaling pathways via inhibiting degradation of ubiquitination-mediated Notch1, c-Myc and mTOR by suppressing FBXW7 transcription in GC.
Knockdown of FBXW7 reverses the effect of silencing MECP2 on GC cells To further con rm that MECP2 facilitates GC cell migration and invasion by repressing FBXW7 expression, FBXW7 siRNA was co-transfected with MECP2 siRNA into BGC-823 or MKN-45 cells. The wound-healing assay showed that silencing MECP2 inhibited GC cell migration, and this effect was eliminated by knockdown of FBXW7 (P < 0.01, Fig. 7a). The transwell assays also revealed that silencing MECP2 suppressed cell migration, which was reversed after co-transfection with MECP2 siRNA and FBXW7 siRNA (P < 0.01, Fig. 7b). In addition, co-transfection with MECP2 siRNA and FBXW7 siRNA rescued the effect of MECP2 knockdown on cell invasion (P < 0.01, Fig. 7c).
The downstream regulators involved in the promotion of cell migration and invasion by MECP2 were determined by qRT-PCR. MECP2 mRNA expression signi cantly decreased in BGC-823 and MKN-45 cells after transfection with MECP2 siRNA or MECP2 siRNA + FBXW7 siRNA (P < 0.01, Fig. 7d). The mRNA expression of FBXW7 markedly increased after transfection with MECP2 siRNA, and this effect was eliminated by transfection with MECP2 siRNA + FBXW7 siRNA (P < 0.01, Fig. 7e). There were no signi cant differences in the mRNA levels of c-Myc, mTOR and Notch1 among cells transfected with MECP2 siRNA alone, cells co-transfected with MECP2 siRNA and FBXW7 siRNA, and those transfected with NC-siRNA (Additional le 7: Figure S5A-C). Silencing MECP2 decreased the mRNA levels of Hes1, MMP-2 and MMP-9, while co-transfection with MECP2 siRNA and FBXW7 siRNA rescued these effects (P < 0.01, Additional le 7: Figure S5D-F). The protein expression of MECP2 decreased in GC cells after transfection with MECP2 siRNA or MECP2 siRNA + FBXW7 siRNA. Whereas, FBXW7 protein expression increased after transfection with MECP2 siRNA, and MECP2 siRNA + FBXW7 siRNA reversed the effect.
Knockdown of MECP2 also decreased the protein expressions of c-Myc, mTOR, Notch1, Hes1, MMP-2 and MMP-9 in whole-cells. Compared with those in cells transfected with MECP2 siRNA alone, the protein expressions of these genes were upregulated in the co-transfection cells. Moreover, the change of nuclear NICD1 protein level was consistent with that of the whole-cell protein level of Notch1 (Fig. 7f). The above ndings demonstrated that MECP2 facilitated GC cell migration and invasion through inhibiting FBXW7 transcription, thereby regulating the Notch1/c-Myc/mTOR signaling pathways.

Discussion
Oncogenesis and tumor progression are the multistep and multifactorial process involving different genes, which is accompanied by alterations in a variety of gene expression patterns that in turn affect cancer cell survival, growth, cycle, apoptosis, migration and invasion regulated by these genes [27]. In recent years, accumulating evidence has certi ed that MECP2, as a key epigenetic regulator, plays an important oncogene role in several cancer types [18,28]. For examle, MECP2 expression is upregulated and promotes tumor progression in breast cancer, lung cancer, cervical cancer and uterine cancer [18]. It enhances oral squamous cell carcinoma and colorectal cancer growth, and facilitates oncogenesis and development of osteosarcoma and neuroblastoma [29,30]. Silencing MECP2 reduces human prostate transformed cell proliferation [31]. Recent research has found that MECP2 regulates cancer cell migration and invasion in glioma and breast cancer [32,33]. Our previous studies have proved that MECP2 promotes cell proliferation and cell cycle G1-S transition, and restrains cell apoptosis in liver cancer, gastric cancer and breast cancer [19][20][21][22]. By expanding sample, the present study aims to further identify the effect and the molecular mechanism of MECP2 on GC cell migration and invasion. Our results again demonstrate that MECP2 expression is upregulated in primary GC and the high expression is closely related with tumor histology and T stage. The ndings suggest that MECP2 may play a key role in GC progression.
Our results demonstrated that high expression of MECP2 in GC signi cantly promoted cancer cell migration and invasion, while silencing MECP2 remarkably suppressed cell migration and invasion by upregulating the expressions of matrix metalloproteinase-2 (MMP-2) and MMP-9. The process of cancer metastasis is a multi-step biochemical reaction involving many molecular events. An essential step of tumor invasion and metastasis is the degradation of matrix proteins, while the matrix metalloproteinases (MMPs) are the most crucial proteolytic enzymes involved in the process. MMPs facilitate cancer cells accessing vasculature and then the target organs, forming tumor metastasis by degrading the basement membrane and extracellular matrix (ECM) [34]. MMPs can also stimulate cancer cell proliferation and movement to develop metastasis by enhancing the release of growth factors [35]. MMP-2 and MMP-9 can degrade type IV collagen, the major component of basement membranes separating the epithelial cells from the stroma [36,37], and they have been reported to promote GC cell migration and invasion [38,39]. Taking these together with the present study, we suggest that MECP2 facilitates GC cell migration and invasion by upregulating MMP-2/9 expression.
Our study identi es FBXW7 as a MECP2-targeting gene. MECP2 binds to the methylated CpG site in the promoter region of FBXW7, resulting in inhibition of FBXW7 transcription. This nding is consistent with that of some previous studies which show that MECP2 functions as a transcriptional repressor by binding to the methylated CpG sites of gene promoter regions and recruiting corepressors (e.g. histone deacetylases and Sin3A), to restrain the expression of some genes (e.g., MYOD1, FOXF1, BDNF, Cdkl5, RPL11 and RPL5) [17,20,21]. FBXW7, a member of the F-box family of proteins, has been characterized as a tumor suppressor gene that plays an important function in cancer cell survival, proliferation, cycle, apoptosis, differentiation, metabolism, tumor metastasis and drug resistance [40,41]. FBXW7 expression is frequently downregulated in multiple human cancers, such as lung cancer, breast cancer, colorectal cancer, liver cancer, gastric cancer, pancreatic cancer, cervical cancer, prostate cancer, and esophagus cancer [42]. Loss-of-function mutations of FBXW7 are frequently discovered in human cancers, and the total mutation rate is approximately 6% [43]. Recent studies have shown that FBXW7 suppresses cholangiocarcinoma and colorectal cancer cell migration and invasion [44,45]. The present study further demonstrates that FBXW7 is frequently downregulated in human GC tissues and FBXW7 exression is correlated with the clinicopathologic features of GC. Overexpression of FBXW7 suppresses GC cell migration and invasion, while silencing FBXW7 promotes cell migration and invasion. Co-transfection with MECP2 siRNA and FBXW7 siRNA rescues the effect of MECP2 knockdown on cell migration and invasion. These ndings suggest that MECP2 promotes GC cell migration and invasion via suppressing FBXW7 transcription by binding to the methylated CpG islands of the promoter region of FBXW7.
FBXW7 is a subunit of a SCF-type ubiquitin ligase complex that induces the ubiquitination and proteasomal degradation of oncoproteins, including SREBP1, Cyclin E, c-Jun, c-Myc, mTOR, Notch1, Notch4, MCL-1, KLF5 and MCL-1 [46][47][48][49]. This study showed that MECP2 increased the protein expressions of c-Myc, mTOR, and Notch1 by inhibiting FBXW7 transcription and then preventing ubiquitination degradation of oncoproteins. c-Myc is a member of Myc gene family involved in multiple biological processes, such as embryonic development, cell proliferation, cell cycle, apoptosis, differentiation, and protein synthesis [50], and is frequently ampli ed in many human cancers and promotes cell proliferation, migration and invasion [51]. mTOR acts as a serine/threonine protein kinase that modulates cell proliferation, motility, survival, protein synthesis, autophagy as well as cell cycle progression, and activation of the mTOR pathway promotes cancer cell proliferation, migration and invasion [52]. The Notch signaling pathways regulate cellular differentiation, proliferation and apoptotic events [53]. Notch1-4 are trans-membrane proteins that interact with ligands of the Delta-like and Jagged family. Binding of ligand to its receptor leads to the cleavage of Notch receptor [54] and the Notch1 intracellular domain (NICD1) is generated. NICD1 enters into the nucleus and promotes Hes1, MMP-2 and MMP-9 expressions in some cancers [55][56][57]. Previous studies have con rmed that Notch1 is upregulated in gastric cancer, liver cancer and ovarian cancer [58][59][60]. The present study reveals that, by regulating the FBXW7/Notch1 signaling pathway, MECP2 upregulates Hes1, MMP-2 and MMP-9 expressions. MECP2 promotes migration and invasion of GC cells by regulating the Notch1/c-Myc/mTOR signaling pathways through inhibiting FBXW7 transcription (Fig. 8).

Conclusion
In summary, this study demonstrates that MECP2 promotes GC cell migration and invasion through