Background

Gastric cancer (GC) ranks the fifth most common cancer and the third most common cause of cancer-related mortality worldwide [1]. Although early stage GC patients feasible for curative surgery retain 70 to 95% 5-year overall survival (OS) rate, more than two thirds of GC patients are diagnosed with advanced stages with unresectable diseases [2, 3]. For unresectable advanced or metastatic GC, the prognosis remains poor with median OS around 8 to 12 months after 1st-line chemotherapy and a 5-year OS at around 5% [4]. Since chemotherapy has been the major treatment option for advanced GC, development of chemoresistance has largely limited the effectiveness of chemotherapy and resulted in disease recurrence and grave prognosis [2]. The mechanisms of chemoresistance in GC are multifactorial that involve in drug efflux, drug interaction and dysregulation of cellular signaling as well as pathways that regulate cancer stemness [2, 5]. Hence, there has been an urgent need to explore new therapeutic targets to overcome chemoresistance in GC.

MET (Mesenchymal-epithelial transition factor) is a receptor tyrosine kinase that is proposed as a promising target in cancer therapy due to the predominant oncogenic signaling cascades that has been shown crucial to malignant progression and tumorigenesis of several cancer types [6,7,39,40,41], were identified to be significantly upregulated in GC tissues as compared to adjacent normal gastric tissues (Additional file 1: Fig. S1B). In addition to MCM2 to MCM7, six downstream oncogenic targets of c-MYC implicated in cancer survival and metabolism including PSMA1, PSMA6, PSMB2, HDAC2, LDHA and SPRING [45] were positively correlated with IFITM3 in 71 GC specimens using the Cho dataset [27] (Fig. 7B, and Additional file 1: Fig. S1C). Consistently, the gene expressions of MCM2, MEM3, MCM4, PSMA1, PSMA6, PSMB2, LDHA and SPRING were significantly elevated in IFITM3-overexpressing TMK-1 cells, while reduced expressions were observed in IFITM3-silenced TSGH cells (Fig. 7C and Additional file 1: Fig. S1D), further supporting IFITM3 as a positive regulator in MET/AKT/c-MYC signal axis in GC. To elaborate c-MYC induction is required during IFITM3-mediated tumor progression and chemoresistance acquisition, a c-MYC-silenced model was established using IFITM3-overexpressing GC cells (Fig. 7D). Our data demonstrated that both cellular proliferation and migration abilities of IFITM3 high-expressing cells were significantly compromised by the knockdown of c-MYC (Figs. 6F, 7E). Further, the diminished sphere formation by knockdown of c-MYC in IFITM3-overexpressing cells suggested c-MYC as a crucial target of IFITM3 (Fig. 7G). In line with these observations, c-MYC knockdown significantly lowered the 5’FU and cisplatin chemoresistance attributed by overexpression of IFITM3, confirming the crucial function of c-MYC in mediating the oncogenic effects of IFITM3 (Fig. 7H). Collectively, our present study revealed a novel mechanism modulated by IFITM3-associated HGF/MET/AKT signaling complexes that suppressed FOXO3, consequently leading to c-MYC upregulation to promote cell proliferation, metastasis, stemness and chemoresistance in GC (Fig. 8).

Fig. 7
figure 7

c-MYC and its target genes are regulated by IFITM3 in GC oncogenesis and chemoresistance. A Gene Set Enrichment Analysis (GSEA) was carried out to analyze the co-expression network and determine Pearson’s correlation coefficient between IFITM3 and other hallmarks of cancer in cBioPortal. The right panel represents an enrichment plot of IFITM3-correlated genes in c-MYC-regulated targets. B Spearman rank correlation coefficient of individual genes (including PSMA1, PSMA6, PSMB2, HDAC2, LDHA and SPRING) against IFITM3 in 71 GC tissues from Cho Gastric dataset (GSE138861). C Quantitative RT-PCR analyses were conducted on the six c-MYC target genes for their gene expression in TMK-1 IFITM3-overexpression or TSGH IFITM3-depletion models (*p < 0.05, **p < 0.01). D Western blot analyses of IFITM3 and c-MYC expression in control or IFITM3-overexpressing TMK-1 cells with or without c-MYC knockdown. Cellular growth curves (E) and migration rates (F) of the same experimental groups were determined by measuring viable and migrated cell numbers, respectively. G Sphere formation assays were performed to assess the influences c-MYC knockdown on IFITM3-overexpressing TMK-1 cells. (Left panel, scale bar, 500 μm; right panel is the statistical results, **p < 0.01) H Chemosensitivities of the same indicated groups of TMK-1 cells were determined based on the relative number of cells survived after treatment with 5’FU (20 μg/ml) or cisplatin (3 μM) for 72 h. *p < 0.05, **p < 0.01.

Full size image
Fig. 8
figure 8

Cartoon illustration that depicts a model of IFITM3-mediated MET/ AKT/FOXO3/c-MYC cascade signaling pathway. Data presented in this study reveals a new signaling model in which IFITM3 associates with MET and AKT complex, leading to enhanced HGF/MET signaling as well as suppression of FOXO3, consequently resulting in upregulation of c-MYC to promote proliferation, metastasis, cancer stemness and chemoresistance of GC cells

Full size image

Discussion

Consistent with literature reports that showed IFITM3 as a prognostic marker associated with progression of head and neck squamous cell carcinoma [Coimmunoprecipitation

Five mg of indicated cell extract was incubated with 2 mM dithiobis[succinimidylpropionate] (DSP) (22,586, Thermo Fisher) in PBS for 10 min and quenched by adding 1 M Tris (pH 7.5) and subsequently terminated by the addition of quenching buffer (20 mM Tris, (pH 7.5) and 100 mM KCl) for 15 min at room temperature. These cell lysates were then precleared using 2 μg of either mouse or goat IgG and Protein A/G Magnetic Beads (#15,752,442, Thermo Fisher Scientific) for 1 h at 4 °C on a rotator. The precleared lysates were then subjected to immunoprecipitation by incubating with indicated antibody at 4 °C for 3 h. Next, protein A/G beads were then added and mixed further by incubation on a rotator at 4 °C for 16 h. The protein complexes were further analyzed by immunoblotting with indicated antibodies.

Tumor sphere formation assay

GC cells were digested with 0.25% trypsin, washed twice with PBS and suspended in the sphere formation medium (+ 100 mg/ml EGF + 100 mg/ml bFGF + 1 ml B27 supplement /50 ml DMEM-F12) and seeded on an ultra-low attachment plate (2000 cells/ml), subsequently cultivated for 7 days, and the spheres counted under a microscope. The size of Oncospheres larger than 100 μm were isolated and counted. The experiments were repeated at least three times, and the results are presented as means ± SEM.

Xenograft tumor formation

For the tumor propagation assay, IFITM3-overexpressing GC cells were subcutaneously injected into 6–8-week-old male BALB/c nude mice randomly. The tumor volumes were measured twice per week. These groups of mice were sacrificed 28 days after implantation. The tumor formation was assessed and calculated. The sample sizes were based on literatures. All of the experiments were conducted in the observer-blinded and randomized manner. All procedures were performed in accordance with the Guide for Care and Use of Laboratory Animals issued by the Institutional Animal Care and Use Committee of Chang Gung University and the National Institutes of Health of United States (CGU106-142).

Real-time RT-PCR

To quantify IFITM3 transcripts in GC, total RNAs were extracted from GC cells and clinical tissues with RNeasy mini Kit (QIAGEN Inc., Dusseldorf, Germany). Then, these RNAs were converted into cDNA using MMLV reverse transcriptase (Thermo Fisher Scientific Inc., Kalamazoo, USA) after DNase (QIAGEN Inc., Dusseldorf, Germany) pre-treatment. Q-RT-PCR was performed using the SYBR Green system (Biotools Co., Ltd., Taipei, Taiwan). Fluorescence emitted by SYBR Green was observed using the ABI PRISM 7500 sequence detection system (Applied Biosystems, Werrington, UK). To further eliminate the interference of genomic DNA in qRT-PCR analysis, we designed the PCR primers that span exon-exon junctions of specific genes for detecting the expressions of IFITM3 in clinical GC specimens and MCM2, MCM3, MCM4, MCM5, MCM6, MCM7, PSMA1, PSMA6, PSMB2, HDAC2, LDHA and SPRING in IFITM3-overexpressing or silenced GC cells. 18S ribosomal RNA is used as control for qRT-PCR analyses. The Q-RT-PCR experiments in GC cell lines were performed in triplicates, and the results are presented as means ± SEM. The primer information is listed below: IFITM3: (F5’-GGCTTCATAGCATTCGCCTACT-3’, R5’-TCACGTCGCCAACCATCTT-3’), MCM2: F5'-GGCGGAGAGGATCGTGGTA, R5'-TGGATGCCATGGTGAAGGAT, MCM3: F5'-GCGGGAGGCTCAGAGAGAT, R5'-TCTGATAAATTCCCTGGTCTTCCT MCM4: F5'-CCCTCCCCAAATGCATTCT, R5'-CGTATGTCAGTGGTGAACTAACATCA, MCM5: F5'-AAGGAGTTCCTGCGGCAGTA, R5'-CGCTTGAGTTCATCCCTGTATTT,MCM6: F5'-GTTCCTGGACTTCTTGGAGGAGTT, R5'-AATCAGTTCCTCTGCTAATTGCAA, MCM7: F5'-GGTGGTGGCCACTTACACTTGT, R5'-GCACATGATCAGAGGCATGAAA, PSMA1 (F5’-ATGGGCCCTCACATTTTCC-3’, R5’-CATGGCTCTGCAGTCAAAATAGTT-3’), PSMA6: (F5’-CGAGGGTCGGCTCTACCAA-3’, R5’-TTTCCCTCTGACAGCTACTGATGT-3’, PSMB2: (F5’-TGTCCAGATGAAGGACGATCAT-3’, R5’-CCTCTCCAACACACAGGAGTAAT-3’), HDAC2 (F5’-CAAGGAGGCGGCAAAAAA-3’, R5’-GGGATGACCCTGTCCATAATAATAA-3’, LDHA (F: 5’-CCGCCCGACGTGCAT-3’, R: 5’-TCCTTTAGAGTTGCCATATTGGACTT-3’, SPRING (F: 5’-GTACAGCTGCGCAACATGGT-3’, R: 5’-AGACAGGCTTCAAACTCCACACT-3’), and 18S rRNA (F: 5’-GAGCCGCCTGGATACC-3’, R: 5’-CCTCAGTTCCGAAAACCAACAA-3’).

Statistical analysis

All the in vitro experiments were reproducible and repeated at least three times, and the results are presented as means ± SEM. The Statistical analysis was performed with the GraphPad Prism software (GraphPad Software, CA) using Mann–Whitney U or Fisher’s exact test for between-group comparisons. The two-tailed paired or unpaired t-tests were performed to determine the significance between the groups compared. P values < 0.05 were considered statistically significant.