YY1 is regulated by ALKBH5-mediated m6A modification and promotes autophagy and cancer progression through targeting ATG4B

Cell culture

Human gastric cancer cells (MGC803, AGS, MKN45 and BGC823) and normal human gastric mucosal cells (GES-1) were purchased from the American Type Culture Collection and Nanjing Cobioer Biosciences (China). All cells were cultured with 10% fetal bovine serum (FBS), 100 u/ml streptomycin and 100 μg/ml penicillin under standard conditions (5% CO 2 at 37°C).

Real-time quantitative PCR (RT-qPCR)

First, total RNA was isolated using the EZ-press RNA Purification Kit (B0004D), and was then reverse transcribed to cDNA with a common product (Vazyme, R212-01, China). Real-time quantitative PCR was performed using the SYBR qPCR Master Mix kit. Primers used for qPCR were synthesized by Biotech (Shanghai, China) as described in Supplementary Table 1. ACTB mRNA was used as control. The relative expression of the target gene was determined utilizing the 2^−ΔΔct method.

Western blot

Proteins were extracted from cells using the RIPA lysis buffer (Thermo Fisher Scientific, USA) and then separated with SDS-PAGE before being transferred to PVDF membranes (Millipore, USA). Next, the proteins on the PVDF membranes were incubated with primary antibodies for LC3B (CST, #3868), SQSTM1/p62 (CST, #23214), Beclin-1 (CST, #3738), YY1 (CST, #46395), ATG4B (CST, #13507), ALKBH5 (CST, #80283), and YTHDF1 (CST, #57530) before visualization using the Bio-Rad chemiluminescence system.

CRISPR/dCas9-mediated inhibition and activation

To activate or inhibit gene expression, the CRISPR/dCas9 approach was employed [29, 30]. Independent gRNAs targeting YY1, ALKBH5, YTHDF1 and ATG4B (Supplementary Table 2) were inserted into engineered CRISPRi and CRISPRa plasmids. Briefly, the wild type Cas9 coding sequence (Cas9 CDS) in the lentiCRISPR v2 plasmid was mutated to yield the dead Cas9 (dCas9) protein with D10A and H840A mutations and, hence, no endonuclease activity. Next, the BFP-KRAB domain obtained from pHR-SFFV-dCas9-BFP-KRAB (Addgene, #46911) and the VPR domain copied from SP-dCas9-VPR (Addgene, #63798) were subcloned to the C-terminus of dCas9 to produce the CRISPRi (Ci) and CRISPRa (Ca) constructions. Briefly, in this dCas9 system, infectious lenti-viruses supernatants were harvested from HEK-293T cells about 60 hours later after co-transfection of the cells with CRISPRa/i, psPAX2 and pMD2G plasmids at the same time. The virus supernatants were filtered with the PVDF filter (0.22 μm Millipore, USA), centrifuged at 12000 g for 10 minutes, and then stocked in a final volume of 100–200 μl. For the CRISPRi or CRISPRa screen, cancer cells were infected with the single or pooled lenti-viruses. 72 hours later, cancer cells were treated with puromycin (Invitrogen, USA), and after 2–3 weeks of puromycin selection, cells were collected and western blot was applied to verify efficiency of gene overexpression or knockdown, thus stable cell lines were obtained.

Data mining and analysis

The autophagy-related genes (222 of them) were downloaded from the Human Autophagy Database (Supplementary Table 3). The R package “Rcis Target” and the BRAT website (http://bartweb.org/) were used to identify the key transcription factors that regulate autophagy (Supplementary Table 3). The RNA sequencing data and clinical survival data of 375 gastric tumor patients and 32 precancerous tissues were downloaded from Xena (http://xena.ucsc.edu). The correlations between YY1 and other genes were determined using the Pearson correlation analysis.

Cell viability and proliferation assays

The CCK-8 solution was added directly to each well of all the tumor cells (2 × 10³ per well) seeded in 96-well plates, and the mixture was incubated for 2 h. OD values were measured at 450 nm using Enzyme Markers. For the colony formation assay, 2 × 10³ infected cells were maintained in each well, and colonies were fixed with 4% paraformaldehyde and then stained with 0.1% crystalline violet. Colonies containing more than 50 cells were counted.

Migration assays

Cell migratory capacity was gauged employing the transwell assay. Briefly, 2 × 10⁴ cells were inoculated into the upper chamber after resuspension with serum-free medium, and 20% FBS medium was added to the lower chamber before incubation was initiated. After 24 h of incubation, the cells were fixed and then stained with 0.1% crystalline violet (Sigma, USA). Migrated cells were counted and imaged.

MeRIP (Methylated RNA immunoprecipitation)

Cells were mixed with the RIP lysis buffer and immunoprecipitated by co-incubation with an m6A antibody using a Magna methylated RNA immunoprecipitation m6A Kit (Millipore, USA). The enrichment of mRNA containing m6A was then analyzed utilizing RT-qPCR.

Immunohistochemistry

Paraffin-embedded tumor tissues were cut into 4-mm sections, flattened in water at 45°C, and baked at 70°C for 30 min. Antibodies specific for Ki-67 (Cell Signaling Technology, #9449; 1:200 dilution) were then introduced, and incubation was initiated. In the end, the intensity was imaged and measured.

Confocal microscopy

Tumor cells were inoculated on glass-bottom culture dishes, infected with RFP-GFP-LC3 lentivirus for 48 h, fixed in 4% paraformaldehyde, and then stained with DAPI for nuclei scrutiny. Ultimately, the autophagic flux was monitored using confocal laser scanning microscopy (STELLARIS 5, Leica, USA).

Dual-luciferase reporter assay

The promoter of ATG4B was subcloned into a pGL3-basic vector, and the 3′-UTR of YY1 mRNA was inserted into the psiCHECK2 plasmid. Dual-luciferase assay was performed following instructions (Promega, USA), and the promoter activity of ATG4B was normalized to the Firefly/Renilla ratio, while the 3′-UTR activity of YY1 was normalized to the Renilla/Firefly ratio. Primers used here are listed in Supplementary Table 2.

ChIP assay

The ChIP assay was conducted using the ChIP assay kit (Beyotime Institute Biotechnology, P2078, China) per the manufacturer’s protocol. YY1-immunoprecipitated DNA was PCR amplified using specific primers, and the YY1 binding site of the putative ATG4B promoter was analyzed. Primers used here are listed in Supplementary Table 1.

RNA stability assays

Cells were incubated with actinomycin D (HY-17,559, MCE) for 0 h, 2 h, 4 h or 6 h before RNA collection. qRT-PCR was employed using the primers listed in Supplementary Table 1 to detect the half-life of YY1 mRNA.

Xenografts in mice

Five-week-old male BALB/c nude mice were used for tumor growth studies in vivo. Briefly, AGS cells were subcutaneously injected into the dorsal side of mice blindly and randomly (n = 5 per group). Tumor volume was measured twice a week and calculated as follows: V = W² × L/2. All animal experiments were performed in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals and approved by the ethics committee of the First Affiliated Hospital of Nanchang University.

Statistical analysis

GraphPad Prism 8.0.2 (USA) was used for data analysis, with the resulting data presented as the mean ± standard error of the mean (SEM). Statistical tests were two sided. A value of P < 0.05 was considered statistically significant.

Availability of data and materials

The data supporting the conclusions of this article are included in this article and its additional files.

YY1 is up regulated and promotes autophagy in GC

To identify the key transcription factors (TFs) that regulate autophagy, the 222 autophagy-related genes downloaded from the Human Autophagy Database (HADb, http://www.autophagy.lu) were subjected to upstream TF analysis using the R package “Rcis Target” and BRAT website (http://bartweb.org), and 14 TFs were identified (Figure 1A and Supplementary Table 1). Next, the expression levels of the obtained 14 autophagic genes-regulating TFs were evaluated in the acquired tumor and normal sample RNA sequencing data and clinical survival data of 375 GC patients and 32 precancerous tissues, respectively, downloaded from Xena (http://xena.ucsc.edu) (Figure 1A and Supplementary Figure 1). The Cancer Genome Atlas (TCGA) public database was used for analyzing the survival values of gene expression in GC patients. Through overlapping analysis, YY1 was identified as the only TF associated with both poor prognosis and enhanced autophagy activity in GC (Figure 1B).

Figure 1. YY1 is upregulated and promotes autophagy in cancer. (A) Venn diagram indicating overlapped genes from two publicly available bioinformatics analyses. The right panel shows gene expression features in cancer and adjacent tissues from the Xena platform. (B) Overlapping analysis (Venn diagram) revealing the 12 TFs with survival-related genes. (C) qRT-PCR analysis showing the expression of YY1 in GC cell lines. (D) Western blotting displaying the expression of autophagy-associated proteins in AGS and MGC803 cells treated with the empty vector (Ca-NC) or Ca-YY1. (E, F) Immunofluorescence staining and TEM scanning demonstrating fluorescence intensity and autophagic vacuoles in AGS and MGC803 cells treated with the empty vector (Ca-NC) or Ca-YY1, and those treated with DMSO, CQ (Chloroquine, 20 μmol/L) or RAPA (Rapamycin, 1 μmol/L) as positive controls for inactivated or activated autophagy. Scale bar: 10 μm in E and 1 μm in F. ^*P < 0.05 vs. GES-1, Ca-NC, DMSO.

To assess the expression of YY1 in GC cells, RT-qPCR was performed, with the results revealing significantly higher presence of YY1 in the cancer cells than in the normal GES-1 cells (Figure 1C). To further validate the regulatory role of YY1 on autophagy, the impact of YY1 overexpression or knockdown was scrutinized using the dCas9-based Clustered regularly interspaced short palindromic repeats (CRISPR) interference or activation (CRISPRi, Ci; CRISPRa, Ca) methods [11, 29]. Two independent gRNAs against YY1 (Ca-YY1, Ci-YY1) designed from the ZHANG LAB website (https://zlab.bio/guide-design-resources) and introduced into GC cell lines were used to examine the efficiency, and the superior Ca-YY1 #2- and Ci-YY1 #1-mediated performances of promotion or inhibition were selected for the subsequent gain-and-loss evaluation (Supplementary Figure 2A, 2B). To observe the regulatory effects of YY1 on autophagy in GC cells, western blot, RFP-GFP-LC3 double fluorescence with confocal microscopy and transmission electron microscopy (TEM) were applied. As shown in the results, there was an increase in the autophagy-specific markers of LC3B and BECN1 and a decrease in the expression of P62 in MGC-803, AGS, MKN-45 and BGC823 cells with YY1 overexpression (Figure 1D and Supplementary Figure 2C). According to the autophagy flux assay, YY1 promoted both the autophagosome (yellow fluorescence, RFP+GFP+) and autolysosome (red fluorescence, RFP+GFP-) formation in MGC-803, AGS, MKN-45 and BGC823 cells (Figure 1E and Supplementary Figure 2D). TEM analysis validated the elevated number of autophagic vacuoles in MGC-803 and AGS cells with stable YY1 overexpression transfection (Figure 1F). These results suggested that YY1 facilitated autophagy in GC.

YY1 promotes cancer progression through autophagy enhancement

The possible association between the oncogenic function of YY1 and autophagy was investigated. For this purpose, the autophagy inhibitor of 3-methyladenine (3-MA) and the agonists of adenosine triphosphate (ATP) were added to detect the cellular function of YY1 in GC. The CCK-8 (Figure 2A, 2B), colony formation (Figure 2C, 2D) and transwell assays (Figure 2E, 2F) all showed that YY1 presence promoted the viability, proliferation and migration activity of cancer cells, while the inhibition of YY1 resulted in the opposite outcome. Notably, 3-MA or ATP incubation-mediated autophagy inhibition or activation abolished these cellular effects, respectively. These findings suggested that YY1 promoted GC progression in an autophagy-dependent manner in vitro.

Figure 2. YY1 promotes cancer progression through autophagy enhancement. (A, B) The in vitro CCK-8 assay of AGS and MGC803 cells stably transfected with Ca-NC, Ca-YY1, Ci-NC or Ci-YY1 and those treated with DMSO, 3-MA (1.0 μM), PBS or ATP (0.1 mM). (C–F) Representative images (left) and the quantification (right) of colony formation (C, D) and transwell (E, F) assays showing the growth and migration of GC cells transfected with Ca-NC, Ca-YY1, Ci-NC or Ci-YY1 and those treated with 3-MA (1.0 μM) or ATP (0.1 mM). ^*P < 0.05 vs. Ca-NC+DMSO, Ci-NC+PBS.

YY1 facilitates ATG4B expression

Comprehensive analyses to elucidate the underlying mechanisms of YY1-induced autophagy in GC were conducted next. Through the overlapping analysis of 26 YY1 target genes from the Human Autophagy Database and 1044 YY1-related genes from the UCSC Xena platform (data source: TCGA database), ATG4B (autophagy-related 4B cysteine peptidase), ARNT (aryl hydrocarbon receptor nuclear translocator), GOPC (golgi-associated PDZ and coiled-coil motif containing) and RAF1 (Raf-1 proto-oncogene) were identified as the potential targets of YY1-mediated autophagy activation (Figure 3A and Supplementary Table 4). The expression features of those four genes in GC and the adjacent normal tissues were, therefore, evaluated, and ATG4B was found to be the most significantly differentially expressed gene (Figure 3B). Subsequently, a positive correlation was obtained between YY1 and ATG4B expression (n = 375, r = 0.72, P = 2.31e-61) in GC patient data from the UCSC Xena platform (Figure 3C).

Figure 3. YY1 facilitates ATG4B expression. (A) The analysis of the overlap (Venn diagram) between YY1 target genes and YY1-related genes revealing the four targets involved in the YY1 regulation of autophagy-related genes. (B) Heatmap showing the expression features of those four genes in GC and adjacent normal tissues from the Xena platform. (C) A correlation test validating the association between ATG4B and YY1 on the Xena platform. (D) qRT-PCR assay displaying the expression levels of ATG4B in GC and GES-1 cells. (E) ChIP assay unveiling YY1 enrichment on the ATG4B promoter in MGC803 and AGS cells stably transfected with Ca-NC, Ca-YY1, Ci-NC or Ci-YY1. (F) Schematic illustration showing the YY1 binding sites on the ATG4B promoter at -135/-121 bases from UCSC Genome Browser, and the wild-type (WT) or mutations (Mut) of ATG4B promoter reporters were designed. (G) Dual-luciferase assay disclosing the luciferase activity of ATG4B promoter in AGS and MGC803 cells stably transfected with Ca-NC, Ca-YY1, Ci-NC or Ci-YY1. (H, I) qRT-PCR and western blot assays showing the transcript and protein expression levels of ATG4B in MGC803 and AGS with YY1 overexpression or knockdown. ^*P < 0.05 vs. GES-1, Ca-NC, Ci-NC.

Studies have revealed that ATG4B promotes autophagy, and the elevated expression of ATG4B is associated with advanced metastasis and poor prognosis in GC [31]. Additionally, ATG4B-targeted inhibition results in reduced autophagy flux and tumorigenicity in glioblastoma cells [32]. The expression of ATG4B in GC cells was assessed using RT-qPCR, resulting in significantly highly expressed ATG4B in cancer cells than in normal GES-1 cells (Figure 3D). The ChIP (chromatin immunoprecipitation) assay established binding between YY1 and the ATG4B promoter (Figure 3E). Next, the specific DNA binding sequence of YY1 on the ATG4B promoter (AAAATGGCGGCTCCC, -135/-121 bases from its transcription start site) obtained from the UCSC database (http://genome.ucsc.edu/) was identified to design the mutated reporter of ATG4B promoter (Figure 3F). Notably, the luciferase activity of the mutated ATG4B promoter was not affected by YY1 in GC cells (Figure 3G). Then, stable activation or inhibition of YY1 using dCas9 applications facilitated or attenuated the promoter activity, mRNA transcripts and protein expression levels of ATG4B in MG803 and AGS cells (Figure 3G–3I). These results indicated that YY1 facilitated the transcription of ATG4B to promote autophagy in cancer.

ALKBH5 weakens the m6A levels of YY1 mRNA

According to previous studies, m6A modifications are the most prevalent internal regulations on eukaryotic mRNAs, playing key roles in cancer progression [21]. Critically, the m6A methyltransferase of METTL3 enhances YY1 mRNA stability and tumorigenesis in an m6A-dependent manner [28]. Nevertheless, the m6A regulation mechanisms of YY1 and its m6A regulator in GC are rarely reported. Herein, to probe the outcome of the m6A modification of YY1, the MeRIP-qPCR (methylated RNA immunoprecipitation-qPCR) assay was deployed, revealing that the m6A levels of YY1 were significantly higher in GC cells than in normal GES-1 cells (Figure 4A). Particularly, the m6A eraser that induced YY1 mRNA demethylation was not clear, and ALKBH5 (alkB homolog 5, RNA demethylase) and FTO (FTO alpha-ketoglutarate dependent dioxygenase) were the most reported and strongest m6A demethylases. As a result, the regulatory functions of these two m6A regulators on YY1 were detected. Notably, the overexpression or knockdown of ALKBH5 decreased or increased the mRNA and m6A levels of YY1 according to RT-qPCR and MeRIP-qPCR assays, respectively, but there were no changes in the FTO groups (Figure 4B, 4C and Supplementary Figure 3A, 3B). As the m6A modifications are mainly enriched in the 3′-UTR near the mRNA stop codon [33], and the SRAMP Browser (http://www.cuilab.cn/sramp/) revealed the m6A modification site with very high confidence on the YY1 3′-UTR, the wild-type or mutations of YY1 3′-UTR reporters were designed (Figure 4D). Per dual-luciferase and western blot assays, the stable activation or inhibition of ALKBH5 respectively attenuated or facilitated the 3′-UTR activity, protein expression levels and binding activity of YY1 in MGC803 and AGS cells (Figure 4E–4G). Notably, the luciferase activity of the mutated YY1 3′-UTR was not affected by ALKBH5 in GC cells (Figure 4E). To examine the regulation of mRNA stability, the RNA polymerase II inhibitor of actinomycin D was used, with the results revealing that ALKBH5 reduced the mRNA half-life of YY1 in GC cells (Figure 4H). In summary, ALKBH5 altered the stability of YY1 mRNA by reducing its m6A levels, leading to YY1 inhibition in GC cells.

Figure 4. ALKBH5 weakens the m6A levels of YY1 mRNA. (A) MeRIP assay showing the YY1 m6A levels in human GC cell lines compared to GES-1 cells. (B, C) qRT-PCR and MeRIP assays displaying the mRNA and m6A levels in MGC803 and AGS cells induced for the overexpression or knockdown of ALKBH5. (D) Schematic illustration showing the m6A modification site (AAGGGA, AAAAGA) at position of 130 and 290 base on the YY1 3’-UTR from SRAMP Browser, and the wild-type (WT) or mutation (Mut) of YY1 3’-UTR reporters were designed (m6A was replaced by T). (E, F) Dual-luciferase and Western blot assays revealing the 3’-UTR activity and protein levels of YY1 in GC cells stably transfected with Ca-NC, Ca-ALKBH5, Ci-NC or Ci-ALKBH5. (G, H) Dual-luciferase and qRT-PCR assays showing the binding activity and mRNA half-life of YY1 in GC cells stably transfected with Ca-NC, Ca-ALKBH5, Ci-NC or Ci-ALKBH5. ^*P < 0.05 vs. GES-1, Ca-NC, Ci-NC.

The M6A-dependent promotion of YY1 is positively associated with YTHDF1

While ALKBH5-mediated mRNA demethylation suppresses YY1 expression, the specific m6A reader that recognizes this m6A-modified YY1 mRNA and affects its function is yet unidentified. It is widely accepted that m6A modification effects are mediated via interactions between m6A sites and m6A reader proteins, including YTHDF1-3, IGF2BP1-3 and YTHDC1-2 [34], with YTHDF1 reportedly promoting YY1 mRNA and protein levels through binding to its m6A-methylated transcripts [35]. Importantly, previous study [36] has identified YTHDF1 as a potential m6A reader on the YY1 3′-UTR in an iCLIP (individual-nucleotide resolution UV crosslinking and immunoprecipitation) data (GSE78030). The YTHDF1 binding region analyzed from the m6A-Atlas v2.0 platform (http://rnamd.org/m6a/index.php, source data: GSE78030) is from chr14:100277763 to 100277903. Obviously, the enrichment peak region and binding sites of YTHDF1 on YY1 mRNA reveled from SRAMP and m6A-Atlas database are highly consistent. To further clarify this specific potential molecular mechanism, the YTHDF1-regulated YY1 m6A levels and expression features were assessed. MeRIP-qPCR experiments confirmed that YTHDF1 activation and its interference with the dCas9 methods (Ca and Ci) had no impact on YY1 m6A levels; however, ALKBH5 regulated YY1 m6A levels (Figure 5A). Notably, as the RT-qPCR dual-luciferase and western blot assay results show in Figure 5B, 5D, the overexpression and knockdown of YTHDF1 promoted and suppressed, respectively, the mRNA levels, 3′-UTR activity and protein expression of YY1; however, these effects were rescued after co-transfection with Ca-ALKBH5 or Ci-ALKBH5. Notably, the luciferase activity of the mutated YY1 3′-UTR was not affected by ALKBH5 and YTHDF1 in GC cells (Figure 5C and Supplementary Figure 3C). The effects of YTHDF1 on YY1 mRNA stability were also examined. As shown in Figure 5E, the stable overexpression of ALKBH5 attenuated the YTHDF1-induced increase in the half-life of YY1 mRNA, while the stable knockdown of ALKBH5 prolonged the reduced degradation time of YY1 mRNA in AGS cells through YTHDF1 inhibition (Figure 5F). Overall, these results suggest that the m6A-methylated YY1 mRNA was directly identified and stabilized by YTHDF1, and YY1 was regulated in GC via the m6A-ALKBH5-YTHDF1 axis.

Figure 5. The m6A-dependent promotion of YY1 is positively associated with YTHDF1. (A–D) MeRIP, qRT-PCR, dual luciferase and western blot assays displaying the m6A and mRNA levels, as well as the 3'-UTR activity and protein expression of YY1 in MGC803 and AGS cells transfected with Ca-YTHDF1 or Ci-YTHDF1 and co-transfected with Ca-ALKBH5 or Ci-ALKBH5. (E, F) qRT-PCR assay showing the half-life levels of YY1 mRNA in AGS cells treated with actinomycin D (1 μg/ml) at the indicated periods. ^*P < 0.05 vs. Ca-NC, Ci-NC.

YY1 drives cancer progression and autophagy in an m6A-dependent manner

To further reinforce the findings here, a rescue assay was performed to detect the role of YY1 in regulating the autophagic pathway and cancer progression in GC cells. YY1 overexpression facilitated the growth and migration capacity of MGC803 and AGS cells; however, these activities were rescued by ALKBH5 activation or YTHDF1 and ATG4B inhibition (Figure 6A, 6B and Supplementary Figure 4A, 4B). Next, the regulatory effects of m6A-YY1-ATG4B on autophagy in GC cells were tested. As shown in the results, YY1 enhanced the expression of ATG4B (Supplementary Figure 4C) and LC3B associated with the decreased expression of P62 in MGC-803 and AGS cells, while ALKBH5 activation or YTHDF1 and ATG4B inhibition abolished these effects (Supplementary Figure 4D). The autophagy flux assay and TEM analysis findings showed that the overexpression of YY1 accelerated the autophagosome and autolysosome formation of AGS cells, which were rescued by ALKBH5 activation or YTHDF1 and ATG4B inhibition (Figure 6C–6E). Taken together, these results suggest that YY1 promoted autophagy and GC progression in an m6A dependent manner.

Figure 6. YY1 drives cancer progression and autophagy in an m6A-dependent manner. (A, B) Representative images (left) and the quantification (right) of colony formation (A) and transwell (B) assays showing the growth and migration of AGS cells transfected with Ca-NC, Ca-YY1 or co-transfected with Ca-ALKBH5, Ci-YTHDF1 or Ci-ATG4B. (C, D) Representative images (C) and quantification (D) revealing the immunofluorescence staining intensity with mRFP-GFP-LC3 in AGS cells stably transfected with Ca-NC, Ca-YY1 or co-transfected with Ca-ALKBH5, Ci-YTHDF1 or Ci-ATG4B. Red (RFP+GFP-) puncta represent autolysosomes, and yellow (RFP+GFP+) puncta represent autophagosomes. Scale bar: 10 μm. (E) Representative images from TEM scanning (left panel) and quantification (right panel) exhibiting autophagic vacuoles in AGS cells stably transfected with Ca-NC, Ca-YY1 or co-transfected with Ca-ALKBH5, Ci-YTHDF1 or Ci-ATG4B. Scale bar: 1 μm. ^*P < 0.05 vs. Ca-NC+Ctrl.

The inhibition of YY1 suppresses tumorigenesis in vivo

Additionally, the therapeutic values of YY1 inhibition were explored via dCas9 application on tumor growth in vivo. The stable knockdown of YY1 suppressed tumor growth and weight, as well as the Ki-67 expression (marker of proliferation) of xenografts in AGS cell-engineered athymic nude mice (Figure 7A, 7B), suggesting that the inhibition of YY1 suppresses tumorigenesis in vivo. Overall, these results point to YY1 being regulated by ALKBH5-mediated m6A modifications. They also suggest that YY1 promoted autophagy and GC progression through targeting ATG4B (Figure 7C).

Figure 7. YY1 inhibition suppresses tumorigenesis in vivo. (A) Representative images, in vivo growth curve, and tumor weights of xenografts in nude mice established by the subcutaneous injection of AGS cells stably transfected with Ci-NC, Ci-YY1. (B) Representative images and IHC quantification staining showing Ki-67 protein expression in xenograft tumors. (C) Mechanisms underlying YY1-promoted autophagy and GC progression. ^*P < 0.05 vs. Ci-NC.