Stomach-specific c-Myc overexpression drives gastric adenoma in mice through AKT/mammalian target of rapamycin signaling

Gastric cancer (GC) is one of the most common malignant cancers in the world. c-Myc, a well-known oncogene, is commonly amplified in many cancers, including GC. However, it is still not completely understood how c-Myc functions in GC. Here, we generated a stomach-specific c-Myc transgenic mouse model to investigate its role in GC. We found that overexpression of c-Myc in Atp4b+ gastric parietal cells could induce gastric adenoma in mice. Mechanistically, c-Myc promoted tumorigenesis through the AKT/ mammalian target of rapamycin (mTOR) pathway. Furthermore, AKT inhibitor (MK-2206) or mTOR inhibitor (rapamycin) inhibited the proliferation of c-Myc overexpressing GC cell lines and the initiation of gastric tumorigenesis in c-Myc transgenic mice. Thus, our findings highlight that gastric tumorigenesis can be induced by c-Myc overexpression through activation of the AKT/mTOR pathway.


INTRODUCTION
Gastric cancer (GC) is one of the most common malignant cancers in the world. It was reported that in 2018, the incidence of GC ranked the fifth in the world, while the mortality rate ranked the third [1]. East Asian countries, including China, have a high incidence of GC, partly due to the high-salt diet. Although the diagnosis means have improved in recent years, the diagnosis of early GC is still poor because of the lack of apparent symptoms. Besides, patients receiving a conventional treatment have a recurrence rate of 50% and a 5-year survival rate of 20% [2]. According to the Lauren classification [3], GC can be classified into two types: Intestinal type and diffuse type, of which the intestinal-type gastric cancer accounts for 60%-75% [4]. Gastric carcinogenesis follows a series of precancerous phases, which is called Correa' s cascade [5], including chronic gastritis, atrophic gastritis, intestinal metaplasia (IM), dysplasia, and eventually GC. As the precancerous lesions of GC can last for a long period, it is important to identify the causal drivers for the development of early GC.
Mouse model is commonly used to investigate the pathogenesis of various cancers since the protein-coding genes of mice and human share high similarity [6]. Establishment of mouse models of GC has progressed from chemically induced random mutagenesis, to bacterial-induced dysplasia and to genetically engineered mouse models (GEMMs) [7]. These models have revolutionized our understanding of the effects of diet, bacteria, and genes on gastric carcinogenesis. Of these, GEMMs are proven to be the most useful tool for dissecting the roles that individual genes and signaling pathways play in GC. These models include the introduction of mutations in oncogenes and tumor suppressor gene loci, as well as abnormal expression of signaling factors.
The c-Myc is a well-known oncogene involved in various cancers, including GC. Amplification of c-Myc in GC has been reported in several studies [8][9][10][11]. Gain of c-Myc copies (≥3) is linked with late on-set, intestinal-type, advanced tumor stage, and distant metastasis, while c-Myc hypomethylation is associated with diffuse-type GC [12]. It is reported that c-Myc overexpression is more frequently observed in GC than gene amplification [13,14]. c-Myc overexpression was described in over 40% of GC [15]. De Souza et al. observed that 77% of the gastric tumors present significantly increased c-Myc mRNA expression, which was associated with deeper tumor 435 www.bjbms.org from the Jackson Laboratory and were also described previously [38]. Atp4b-cre; Myc fl/+ , referred as Atp4b-cre; Myc OE mice were generated by crossing Atp4b-cre mice with Myc fl/fl mice. Myc fl/fl mice were used as control. Both male and female mice were used for experiments since no difference of sex has been observed. All the mouse strains were generated in a C57BL/6 background and were born and maintained in a specific-pathogen-free (SPF) facility and all experimental procedures were approved by the Animal Ethics Committee of School of Biomedical Engineering and Med-X Research Institute, Shanghai Jiao Tong University. All institutional and national guidelines for the care and use of laboratory animals were followed. Primers for genotyping are shown in Table S1.

Western blotting
Tissue and cell lysates were prepared by strong radioimmunoprecipitation assay buffer (Beyotime, P0013B) containing protease inhibitors and supplemented with protein phosphatase inhibitors (mammalian cell entry [MCE]). The proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels and then transferred to polyvinylidene fluoride membranes (Millipore). The membranes were blocked with 6% skim milk in Tris-Buffered Saline Tween-20 for 1.5 h at room temperature and subsequently incubated with specific primary antibodies overnight at 4°C followed by incubation with secondary antibodies for 1 h. The primary antibodies used in this study were as follows: Anti-flag-tag (cell signaling technology [CST], 14793, 1:1000), anti-c-MYC (Abcam, ab32072, 1:1000), anti-β-Actin (Abmart, extension and metastasis [12,16]. Notably, overexpression of c-Myc is more frequently observed in intestinal-type GC than diffuse-type GC [12,17,18] and is associated with malignant progress and poor survival in GC patients [18][19][20]. MYC protein expression increased progressively from chronic gastritis, IM, dysplasia, early GC to progressive GC [17,18,21,22]. However, whether overexpression of c-Myc is sufficient to cause GC remains unclear. The c-Myc plays a crucial role in several cellular functions, such as cell proliferation, differentiation, and cell cycle progression [23]. It was reported that c-Myc transcriptionally regulates the expression of TRAP1, which controls primary and metastatic tumor growth [24]. In renal cell carcinoma (RCC), c-Myc induces RCC in a glutamine-addicted way [25]. Significant upregulation of c-MYC proteins, which is resulted from alterations of the Wnt and Ras pathways, is often seen in 70% colorectal cancer [26][27][28]. While in GC, it was reported that BRD4 could promote GC progression through positive regulation of c-Myc in transcription and epigenetic levels [29] and knockdown of c-Myc could inhibit the growth and proliferation of GC cell lines [30,31]. Liu et al. reported that USP22 promoted GC progression through the activation of c-Myc/NAMPT/SIRT1dependent FOXO1 and YAP signaling pathways [32]. Xu et al. found that KLF5 and MYC could transcriptionally enhance the expression of LINC00346, which was a GC inducer in vitro and in vivo [33]. Choi et al. reported that YAP/TAZ activation could initiate gastric carcinogenesis through transcriptionally upregulating MYC in the knockout mice for Lats1 and Lats2 [34]. There have been several Myc-driven mouse models of cancer, including prostate cancer [35] and renal cell carcinoma [25], but not GC, to the best of our knowledge. It was reported that MYC inactivation could induce sustained regression of invasive liver cancer in a MYC transgenic mouse model [36]. Thus, investigating the direct impact of c-Myc on GC would be of great interest to uncover new therapies for GC.
In this study, we generated a novel gastric tumor model in which human c-Myc is highly expressed in gastric parietal cells to investigate the definite role of c-Myc in GC. We present data indicating that these mice developed the phenotypic features of the gastric adenoma, with a step-wise tumorigenic progression from hyperplasia to metaplasia, dysplasia, and finally adenoma in gastric mucosa. Importantly, our findings highlight a mechanism by which gastric adenoma can be induced by stomach-specific c-Myc overexpression through activation of the AKT/mammalian target of rapamycin (mTOR) pathway.

Mice
Atp4b-cre mice were gifted from Dr. Xiao Yang and described previously [37]. Myc fl/fl mice were purchased www.bjbms.org Cell culture AGS cell line was obtained from the American Type Culture Collection (ATCC) and cultured in RPMI-1640 supplemented with 10% fetal bovine serum (Thermo), 100 U/mL penicillin, and 0.1 mg/mL streptomycin (Thermo) at 37°C in a humidified 5% CO 2 atmosphere.

Plasmids and transfection
Human c-Myc cDNA was generated by PCR and cloned into pCMV6-Entry vector with Myc-tag and flag-tag. The constructs generated were confirmed by DNA sequencing. For transient transfection, AGS cells were transfected with the jetPRIME ® transfection reagent (Polyplus) according to the manufacturer' s instruction. Primers used for amplification of human c-Myc cDNA were as follows: Sense: 5'-AGTAAA GCTTATGGATTTTTTTCGGGTAGTGGAA-3' and antisense: 5'-ATATACGCGTCGCACAAGAGTTCCGTAG-3' .

CCK8 assays
Cell counting kit-8 (CCK-8, Dojindo, CK04), being non-radioactive, allows sensitive colorimetric assays for the determination of the number of viable cells in cell proliferation and cytotoxicity assays. Cells that transfected for 24 hours were seeded in 96-well plates at a density of 1 × 10 4 cells/well. After RNA extraction, reverse transcription, and real-time polymerase chain reaction (PCR) Total RNA was extracted from tissues using RNA extraction kit (Bioteke) following the manufacturer' s protocol. RNA was then reverse-transcribed with RT reagent kit (Takara, Japan). The cDNAs were subsequently subjected to SYBR Green-based real-time PCR analysis. GAPDH was used for normalization. Data were shown as average values ± standard error of the mean (SEM). The p value was calculated using the Student' s t-test. The primers used in qPCR were listed in Table S1.

RNA sequencing
Gastric mRNA was obtained from 12-week-old Atp4b-cre; Myc OE and wild type (WT) mice. Differential gene expression was analyzed using the DESeq2 package. The list of significance was determined by setting a false discovery rate (FDR) threshold at a level of <0.05 and |log 2 FC| of more than 0.585. All differentially expressed genes were subsequently analyzed for gene ontology (GO) and pathway analysis. Primers

Statistical analysis
All experiments were repeated at least three times. Unless otherwise indicated, data were presented as mean ± SEM and analyzed for statistical significance by Kruskal-Wallis or Mann-Whitney using GraphPad Prism 6 software or SPSS 19.0 software. p < 0.05 was considered to be statistically significant. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Data availability RNA-Seq raw data have been deposited in the Gene Expression Omnibus (GEO) under accession number GEO: GSE145583.

Overexpression of c-Myc in mouse gastric parietal cells induces gastric tumorigenesis
To investigate the role of c-Myc in GC, we first generated a mouse model with Cre-dependent targeted overexpression of c-Myc. c-Myc-floxed mice were crossed with Atp4b-cre to obtain Atp4b-cre; Myc OE mice ( Figure 1A). In Atp4b-cre; Myc OE mice, c-Myc was specifically overexpressed in Atp4b + gastric parietal cell lineages, which are the most abundant cells in gastric mucosa. Myc fl/fl littermates of mice were used as control, referred to here as WT. After genotyping, Atp4b-cre; Myc OE mice were selected ( Figure 1B). RT-qPCR and western blotting confirmed remarkable overexpression of c-Myc in the mouse stomach ( Figure 1C). Compared with WT mice, c-Myc transgenic mice were viable and showed no significant alterations in weight. However, visible tumors were observed in the stomachs of Atp4b-cre; Myc OE mice ( Figure 1D). Tumors were mainly occurred in the corpus of the mouse stomach. H&E staining showed abnormal hyperplasia in c-Myc transgenic mice ( Figure 1E). By immunostaining, significant-high expression of c-MYC was confirmed in the gastric mucosa ( Figure 1F). Collectively, these results demonstrate that stomach-specific overexpression of c-Myc can induce tumors in the mouse stomach.
c-Myc transgenic mice display an age-dependent progressive gastric histopathology To analyze tumorigenesis driven by c-Myc in the stomach, mice were sacrificed at sequential time points. At gross examination, there was obvious mass protrusion in the corpus region. Tumor area increased as the mice grew older (Figure 2A). In histology sections, we found microscopic changes in the gastric mucosal epithelium of Atp4b-cre; Myc OE mice ( Figure 2B and Table S2). At 12 weeks, there was a slight loss of parietal cells and chief cells, indicating atrophy. We also observed a moderate elongation of the surface-type epithelium, which was hyperplasia. At 25 weeks, these lesions became more severe, characterized by tubule branching and infolding, cell piling up, and increased nuclear-cytoplasmic (N-C) ratio (dysplasia). At 35 weeks, tumors progressed to adenoma. Cells in the lesions displayed hyperchromatic nuclei and loss of polarity with almost a total loss of parietal cells and chief cells. Inflammatory cells extended into submucosa and mucosa. Besides, we observed different degrees of IM in mice of three ages, with the most severe IM in mice at 25 weeks ( Figure 2B). The intense Ki67 staining of these lesions suggested that they were highly proliferative ( Figure 2C). Taken together, our results indicate that c-Myc overexpression in the mouse gastric mucosa triggers a stepwise progression from hyperplasia to adenoma. Besides, gastric tumorigenesis due to c-Myc overexpression shows characteristics of an early onset and a long precancerous stage.

Atp4b-cre; Myc OE transgenic mice exhibit increased intestinal characteristics and decreased gastric mucins
To further investigate the tumor characteristics in c-Myc transgenic mice, we performed immunostaining for intestinal and gastric markers in gastric tissues. It is well-known that IM is a precancerous lesion of GC [40]. IM is a process of gastric epithelial cells that undergo trans-differentiation to intestinal cells, which mainly express acid mucins [40]. As shown in Figure 3A, AB/PAS staining revealed AB + cells (indicating acid mucins) in Atp4b-cre; Myc OE mice and PAS + cells (indicating neutral mucins) in WT mice ( Figure 3A). The normal gastric mucosa specifically expresses MUC5AC, which is mainly found in the superficial epithelium [41]. MUC2 is intestinal mucin and cannot be detected in the normal gastric mucosa [41]. Notably, www.bjbms.org Transcriptome analysis reveals that c-Myc promotes tumorigenesis in mice by impacting PI3K/AKT signaling To explore the potential mechanism underlying c-Myc-mediated tumor growth, we performed RNA sequencing with gastric tissues from WT and Atp4b-cre; Myc OE mice. By analyzing and comparing transcriptome data from WT     and c-Myc transgenic mice, we identified 14930 differentially expressed genes, including 6718 upregulated genes and 8212 downregulated genes ( Figure 4A and B). A subsequent gene ontology (GO) analysis of biological process terms revealed a significant enrichment of genes related to cell cycle and other cellular functions, which could be attributed to high expression of c-Myc and subsequent tumorigenesis ( Figure 4C). RT-qPCR of several genes from GO data was performed to validate the results of RNA-seq ( Figure 4D). As expected, RNA expression levels of c-Myc target genes (Mcm2/Mcm5/eIF4E) were upregulated compared with WT mice. Notably, the RNA expression of Smad2/3/4, which is related to regulation of transcription, was significantly downregulated in c-Myc transgenic mice compared with WT mice. Furthermore, the RNA expression of Mcm2/Mcm5/E2f2, associated with cell cycle, was significantly upregulated in Atp4b-cre; Myc OE mice.
A subsequent pathway analysis revealed a significant enrichment of genes related to phosphatidylinositol signaling system, inositol phosphate metabolism, PI3K-Akt pathway, and mTOR pathway ( Figure 5A). Gene set enrichment analysis (GSEA) data showed that c-Myc overexpression enriched genes correlated with the PI3K-Akt pathway and mTOR signaling ( Figure 5B). To further validate the change of this pathway, RT-qPCR was performed to examine the RNA expression level of several key genes. Akt1 and mTOR, which are main factors of PI3K-Akt pathway, exhibited higher RNA expression levels in Atp4bcre; Myc OE mice compared with WT mice, while Pten, an inhibitory factor of this pathway, was slightly downregulated ( Figure 5C). The western blotting analysis confirmed that Atp4b-cre; Myc OE mice exhibited profound increases of p-mTOR/mTOR and a slight increase of p-AKT/ AKT (p > 0.05) compared with WT mice (Figure 5D), while p-PI3K/PI3K did not show a significant difference between two groups. Taken together, these results indicate that c-Myc may promote tumorigenesis of mice through AKT/mTOR signaling. To further elucidate the causal link between c-Myc and the AKT/mTOR pathway in GC, we transfected c-Myc plasmids into AGS cells and then treated these cells with an AKT inhibitor (MK-2206) or an mTOR inhibitor (rapamycin). As shown by CCK8 assays, overexpression of c-Myc in AGS cells ( Figure 6C) promoted cell proliferation, while treatment of MK-2206 or rapamycin significantly suppressed cell proliferation ( Figure 6A and B). Moreover, we also treated Atp4bcre; Myc OE mice with MK-2206 and rapamycin to test in vivo effect of inhibiting AKT/mTOR pathway in gastric tumorigenesis. Compared with the control group, Atp4b-cre; Myc OE mice treated with MK-2206 and rapamycin did not present any abnormal changes and typical features of hyperplasia or intestinal metaplasia ( Figure 6D). Thus, blocking AKT/mTOR signaling may inhibit the initiation of gastric tumorigenesis. In addition, analysis of TCGA datasets indicated that the c-Myc expression was positively correlated with Akt1 and Mtor expression in GC, respectively ( Figure 6E). Collectively, these results suggest that the oncogenic role of c-Myc is mediated through the activation of AKT/mTOR signaling and blocking AKT/mTOR signaling may be helpful to inhibit or postpone the onset of gastric tumors.

DISCUSSION
Although the incidence of GC has decreased in recent years attributed to the improvement of sanitary conditions and eating habits, there were over one million new cases of GC and more than 780,000 deaths due to GC worldwide in 2018 [1]. GC has a long and asymptomatic precancerous phase, which mainly includes intestinal metaplasia and dysplasia. Therefore, the study of precancerous stages of GC and identification of the drivers for this process are of great significance to prevent and to diagnose GC earlier. Previous studies suggest that overexpression of c-Myc is associated with malignant progress and poor survival in GC patients [19,20]. It is also reported that significantly higher MYC expression was observed in IM samples than gastritis samples from cancer-free individuals and this may facilitate tumor initiation [42]. However, the causal role of c-Myc in induction of GC has been unknown. Our work presents a definite answer regarding the sufficient function of c-Myc in causing the gastric epithelial cells to undergo serial steps of tumorigenesis from an early precancerous phase, including IM and dysplasia to the formation of www.bjbms.org adenoma. Our findings add c-Myc as a causal oncogene to the existing list of GC drivers, which includes Notch, hedgehog, CDH1, and TP53 [7,[43][44][45].
By the establishment of a conditional transgenic mouse model, we show that gastric adenoma induced by c-Myc overexpression is achieved through activation of AKT/mTOR signaling. Our findings are in agreement with many other studies. The AKT/mTOR pathway is a canonical pathway involved in the regulation of multiple cellular functions, including cell proliferation, apoptosis, and metabolism. Aberrations in the PI3K/AKT/mTOR pathway in head and neck squamous cell carcinoma (HNSCC) were associated with malignant  [46]. Activation of the AKT/mTOR pathway is often seen in oral squamous cell carcinoma [47], skin cancer [48,49], and RCC [50]. It has also been reported that the AKT/ mTOR pathway plays a crucial role in the development of GC [51,52]. Our results showed that expression levels of AKT and mTOR are significantly increased in c-Myc transgenic mice, and inhibition of AKT and mTOR can significantly decrease cell proliferation in AGS cells overexpressing c-Myc and inhibit or postpone the onset of gastric tumors in vivo. Importantly, our experiments demonstrate not only that c-Myc can be a driver for gastric adenoma but also that the AKT/mTOR pathway could be the underlying mechanism of gastric tumorigenesis caused by c-Myc overexpression.
It is worth pointing out that our study also shows an increased copy number of c-Myc gene can prompt the gastric tumorigenesis of transgenic mice toward a faster and more severe way. Based on our observation, 14-week-old Atp4bcre; Myc fl/fl mice exhibit submucosal invasion, while Atp4bcre; Myc OE (Atp4b-cre; Myc fl/+ ) mice do not at the same age. Similarly, in human GC, it is reported that increased Myc copy number is associated with a late-onset, intestinal-type cancer and the presence of distant metastasis [12]. Whether and when Atp4b-cre; Myc fl/fl mice exhibit distant metastasis needs further investigation.
Taken together, we generated a novel autochthonous transgenic mouse model of gastric adenoma that is generally useful for studying the initiation and progression of GC. It provides a new platform to further study the roles of more genes involved in GC through combining with mutations in other genes. It will facilitate our better understanding of the development of early GC and shed light on the molecular mechanisms by which c-Myc affects the development and progression of GC. More importantly, it will aid the clinical detection and therapeutic strategies for intervention at precancerous stages of GC so to improve patient survival.