Introduction

Gastric cancer (GC), as one of the most common cancers worldwide, ranks the fourth in incidence and second in cancer-related mortality [1]. Currently, gastroscopy is still the criterion standard test for diagnosing GC [2]. However, this screening method is too expensive and invasive to be popularized in China. Moreover, non-invasive or minimally invasive markers such as carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) have been widely used in clinical, but these similar markers are neither sensitive nor specific enough to facilitate early detection of GC [3]. Due to the lack of early diagnosis and non-specific symptoms, a large number of GC patients were diagnosed at advanced stages [4, 5]. In spite of the development of surgery, chemotherapy, and other therapies in the management of GC patients, their prognosis remains poor [6]. Therefore, non-invasive and cost-effective biomarkers are urgently needed for early detection of GC.

MicroRNAs (miRNAs) are a class of endogenous non-coding RNAs (about 22 nt) that post-transcriptionally modulate genes expression by degrading target mRNAs or blocking their translation [7, 8]. Exosomes, a class of 30–150 nm vesicles secreted by a wide range of mammalian cell types, consist of a lipid bilayer membrane surrounding a small cytosol and are devoid of cellular organelles [9]. Exosomes contain a range of constituents derived from parent cells, such as proteins, adiposes, mRNAs as well as miRNAs [10]. It has been widely reported that cancer cells could release miRNAs through exosomes to promote proliferation and migration [11,12,13,14]. Exosomal miRNAs secreted into biofluids, such as serum, plasma, urine, tear, and gastric juice, could avoid the degradation by RNases [15]. Furthermore, circulating miRNA expression profile was more accurate than circulating mRNA expression profiles [16]. Up to date, emerging studies have declared the biomarker value of circulating miRNAs used for diagnosis and predicting prognosis of patients with various tumors [17,18,19].

In this study, after integrated analysis of two GEO datasets, six circulating miRNAs were founded to be consistently upregulated in GC patients among which miR-1246 was the one with the largest fold change verified by qRT-PCR. Subsequently, we demonstrated that elevated serum miR-1246 was tumor-derived by being packaged into exosomes with the help of ELAVL1 (HuR) for the first time. Then we discovered the diagnostic value of serum exosomal miR-1246 to discriminate GC patients with TNM stage I from healthy controls (HCs) and patients with benign diseases (BDs) was remarkably high. Besides, bioinformatics analysis exhibited that potential targets of miR-1246 mainly located in cellular membrane and were significantly associated with several pathways in GC, such as Hippo signaling pathway.

Materials and methods

Clinical patients

All patients with GC (n = 117), gender/age matched healthy controls (n = 82) and patients with benign gastric diseases (gastric ulcer, atrophic gastritis, gastric polyps, n = 30) were enrolled from The Second Affiliated Hospital of Nanjing Medical University. GC Patients were confirmed through histopathological analysis of surgical resected tumor. All the procedures were approved by the Research and Ethical Committee of The Second Affiliated Hospital of Nanjing Medical University, and the written informed consent was obtained from each participant.

Samples processing

Serum samples collected from venous blood were stored at – 80 °C after being centrifuged at 4000 rpm for 10 min. GC tissues and adjacent normal tissues (ANTs) were frozen in liquid nitrogen immediately after resection. Total RNA was isolated using Trizol LS reagent (Invitrogen, USA) according to the manufacturer’s protocol. Cel-miR-39-3p (Takara, Japan), acting as external reference, was added into each serum sample at a final concentration of 10−4 pmol/ul. U6 (GenePharma, Shanghai) was chosen as the internal control for miRNA expression.

Cell culture

GC cells (AGS) were cultured in RPMI1640 (KeyGEN BioTECH, China) containing 10% fetal bovine serum (FBS) and antibiotics (100 U/mL streptomycin and 100 U/mL penicillin) in 37 °C with 5% CO2. When growing at 80% confluency, AGS cells were treated with 10 uM GW4869, an exosomes inhibitor (Sigma, USA), for 2 h.

Exosomes purification

Exosomes were isolated from cultured media using RiboTM Exosome Isolation Reagent Kit according to the manufacturer’s instruction (RioBoBio, China). Briefly, after collecting media (serum), cellular debris were removed by centrifugation at 2000g for 30 min. Then media (serum) was transferred to a fresh tube and exosomes were precipitated using precipitation solution at 4 °C overnight (3 h). The mixture was centrifuged at 1500g for 30 min at 4 °C to pellet exosomes which were resuspended in PBS subsequently.

Transmission electron microscopy analysis

Isolated exosomes were re-suspended in 100 μl phosphate buffer (pH 7.4) and fixed in 50 μl glutaraldehyde. Subsequently, fixed exosomes were dropped onto a formvar-carbon coated grid and left to dry at room temperature for 5 min and stained with phosphotungstic acidoxalate for 1 min. Then, excess liquid was removed. The grid was dried at room temperature for 10 min and visualized on Tecnai G2 F20 transmission electron microscope (TEM) (FEI, United States) at 185 kV.

qRT-PCR for miRNA quantitation

Reverse transcription and qRT-PCR for miR-1246, external reference miR-39 and endogenous control U6 snRNA were performed using Hairpin-itTM microRNA RT-PCR Quantitation Kit (GenePharma, China) according the manufactures’ instructions. The reactions were initiated with denaturation at 95 °C for 3 min, followed by 40 cycles of 95 °C for 15 s and 62 °C for 34 s. The relative expression of miRNA was calculated by 2−△△Ct method.

Transfection

miR-1246 mimics labeled with biotin at 3′ end and negative control (NC) were commercially synthesized (RioBoBio, China). SiELAVL1, a small interfering RNA against ELAVL1, were synthesized by GenePharma (GenePharma, China). Transfection was carried out with a final concentration of 50 nM biotin-labeled-miR-1246 mimics, 100 nM siELAVL1, and corresponding NCs using the Lipofectamine 2000 reagent (Invitrogen, USA) following the manufacturer’s protocol.

RNA pull-down

After transfecting AGS cells with NC and 3′ biotin-labeled mimics at a final concentration of 50 nM, cells were scraped to prepare fresh lysates supplemented with protease inhibitor (PI) and RNase inhibitor in the lysis buffer. Then wash the beads three times with 1 ml lysis buffer and incubate the Streptavidin-Dyna beads (RioBoBio, China) with lysis buffer on a rotator at 4 °C for 2 h. Put the tube containing the beads on the magnetic stand, take out the supernatant and wash the beads for twice. Mix 500 μl lysates with the beads and incubate at 4 °C on the rotator overnight. 50 μl lysates left were saved at − 80 °C as input. Centrifuge the tube and put it on magnetic stand and remove the supernatant briefly. Wash the beads four times with 1 ml lysis buffer and eluted in PBS. The eluted protein complexes were denatured and prepared for western blot.

Western blot

Briefly, harvested and denatured proteins were separated by 10% SDS-PAGE, transferred to PVDF membranes. The membranes were blocked with TBST that contains 5% skimmed milk, then blocked with primary antibodies (mouse anti-CD63, 1:1000, ab59479; mouse anti-TSG101, 1:1000, ab83; Rabbit anti-AGO2, 1:1000, 10,686–1-AP; Rabbit anti-HuR, 1:2000, 11,910–1-AP; Rabbit anti-beta-actin, 1:5000, 20,536–1-AP) overnight at 4 °C. After being washed four times, the membranes were incubated with anti-IgG conjugated to horseradish peroxidase at room temperature for 1 h. Bands were visualized using the enhanced chemiluminescence system (ECL) reagent (KeyGEN BioTECH, China).

GEO datasets analysis

After searching GEO datasets using keywords “gastric cancer” and “miRNA”, GSE106817 and GSE113486 were enrolled. GSE106817 included 300 serum samples collected from 115 healthy controls and 115 GC patients. GSE113486 included 141 serum samples collected from 101 healthy controls and 40 GC patients. The platform of both GSE106817 and GSE113486 were 3D-Gene Human miRNA V21_1.0.0 (GPL21263).

Bioinformatics analysis

The significantly dysregulated miRNAs were calculated using GEO2R (https://www.ncbi.nlm.nih.gov/geo/geo2r/), an online analysis tool in GEO database. The targets of miR-1246 in GC were identified by integrated analysis of public bioinformatics algorithms TargetScan (https://www.targetscan.org/vert_72/) and a GEO dataset. GO, as a common bioinformatics tool, could provide comprehensive information on genes. GO contains three items: molecular functions (MF), biological processes (BP) and cellular components (CC) [20]. The KEGG is a reference resource for biological interpretation of high-throughput genes and protein pathways [21]. The GO and KEGG analysis were available in the online tool DAVID (https://david.ncifcrf.gov/), which is a bioinformatics database composing functional annotation of large quantities of genes [22]. Protein–protein interaction (PPI) network was constructed using online tool String [23] and Cytoscape.

Statistical analysis

Statistical analysis was performed using SPSS 19.0 (IBM, USA) and GraphPad Prism 8.0 (GraphPad Software, USA) using the Student’s paired or unpaired t test, and Chi-square test. The ROC was performed to evaluate the utility of miR-1246 as diagnostic biomarker of GC. Cut-off value of the relative expression of miRNAs were determined by Youden index from ROC curves. P value < 0.05 was considered to be statistically significant.

Results

Identifying circulating miR-1246 as candidate

After integrated analysis of GSE106817 and GSE113486, 1366 and 1080 miRNAs were found to be upregulated with logFC ≥ 1 and p < 0.05. As shown in Fig. 1a, among top 10 upregulated miRNAs, six circulating miRNAs (miR-29b-1-5p, miR-191-5p, miR-1246, miR-6131, miR-451a, and miR-103a-3p) were consistently upregulated in GC patients. qRT-PCR demonstrated miR-1246 was the one with the largest fold change (Fig. 1b). ROC analysis showed that serum miR-1246 had a remarkably high value to differentiate GC patients from HCs when compared to other miRNAs (Fig. 1c). Therefore, serum miR-1246 was selected as candidate to explore whether it could serve as a biomarker for the early diagnosis of GC patients.

Fig. 1
figure 1

Identifying circulating miR-1246 as candidate. a Six circulating miRNAs (miR-29b-1-5p, miR-191-5p, miR-1246, miR-6131, miR-451a, and miR-103a-3p) were consistently upregulated in GC patients. b The expressions of six circulating miRNAs analyzed by qRT-PCR. c ROC of six circulating miRNAs to discriminate GC patients from HCs

Full size image

Elevated serum miR-1246 was tumor-derived by being packaged into exosomes

Li et al. reported that breast cancer cells could secrete exosomal miR-1246 to promote proliferation, migration, and drug resistance [24]. We wondered whether elevated serum miR-1246 in GC patients was released by tumor in the form of exosomes. Exosomes isolated from serum and cultured media were verified by TEM and western blot (Fig. 2a). qRT-PCR showed that circulating exosomal miR-1246 expressions were significantly upregulated in GC patients (Fig. 2b) and could discriminate GC patients from HCs with an AUC value of 0.95 (Fig. 2c). In addition, serum exosomal miR-1246 expressions were significantly downregulated after tumor resection (Fig. 2d). Next, different amounts of AGS cells were cultured and exosomes in media were collected in different time points. As shown in Fig. 2e, we observed that exosomal miR-1246 expressions in the culture media increased with time and with increasing numbers of cells. Moreover, after AGS cells being treated with GW4860, exosomal miR-1246 expressions in the culture media were significantly downregulated (Fig. 2f). Interestingly, analysis results of GSE93415 showed that miR-1246 expressions were significantly downregulated in GC tissues (Fig. 2g). qRT-PCR verified that compared to ANTs, miR-1246 expressions were significantly downregulated in GC tissues (Fig. 2h). It has been reported that several miRNAs were downregulated in cells, but they can be sorted into exosomes through RNA binding proteins (RBPs), such as SYNCRIP and hnRNP A2/B1, and be released into circulating system [25, 26]. We wondered whether exosomal miR-1246 was mediated by specific RBP. Firstly, potential RBPs binding miR-1246 were searched in RBPDB (https://rbpdb.ccbr.utoronto.ca/index.php). SFRS1 and ELAVL1 were identified. Then biotin-labeled miRNA pull-down assay demonstrated that miR-1246 could bind ELAVL1 through AUUUU motif (Fig. 2i). Then we knocked down the expressions of ELAVL1 by transfecting cells with siELAVL1 (Fig. 2j) and found that exosomal miR-1246 was significantly downregulated in media of cells transfected with siELAVL1 (Fig. 2k).

Fig. 2
figure 2

Elevated serum miR-1246 was tumor-derived by being packaged into exosomes. a Exosomes verified by TEM and western blot. b Circulating exosomal miR-1246 expressions were significantly upregulated in GC patients. c ROC of circulating exosomal miR-1246 to discriminate GC patients from HCs. d Serum exosomal miR-1246 expressions were significantly downregulated after tumor resection. e Exosomal miR-1246 expressions in the culture media increased with time and with increasing numbers of cells. f GW4860 significantly inhibited exosomal miR-1246 expressions in the culture media. g GSE93415 and h qRT-PCR verified that miR-1246 expressions were significantly downregulated in GC tissues. i Biotin-labeled miRNA pull-down assay demonstrated that miR-1246 could bind ELAVL1. j SiELAVL1 significantly knocked down the expressions of ELAVL1. k Exosomal miR-1246 was significantly downregulated in media of cells transfected with siELAVL1

Full size image

High levels and diagnostic utility of exosomal miR-1246 in serum validated in an independent cohort

A total of 165 serum samples, including those from GC patients (n = 85), BD patients (n = 30) and HCs (n = 50), were tested to validate the diagnostic potential of exosomal miR-1246. The clinicopathological characteristics of participants are summarized in Table 1. As shown in Fig. 3a, the relative expressions of exosomal miR-1246 in serum were significantly overexpressed in GC patients compared to that in HCs. ROC analysis demonstrated that circulating exosomal miR-1246 expressions could differentiate GC patients from HCs with an AUC value of 0.911 (95% CI 0.86–0.96, p < 0.001) (Fig. 3d). The sensitivity and specificity were 82.35% and 86%, respectively. The cut-off value was 1.67. Moreover, the gradual increase of circulating exosomal miR-1246 levels was clearly discernible with significantly higher levels in stage III-IV patients compared with stage I–II patients (Fig. 3b). We further explored whether exosomal miR-1246 in serum could differentiate GC patients in TNM stage I from HCs working for early diagnosis of GC. ROC analysis demonstrated that circulating exosomal miR-1246 expressions could differentiate GC patients in TNM stage I from HCs with an AUC value of 0.843 (95% CI 0.75–0.93, p < 0.001) (Fig. 3e). The sensitivity and specificity were 85.71% and 74%, respectively. The cut-off value was 1.43. When compared to patients with BDs, circulating exosomal miR-1246 levels were also significantly higher in GC patients with TNM stage I (Fig. 3c). ROC analysis demonstrated that circulating exosomal miR-1246 expressions could differentiate GC patients in TNM stage I from BDs patients with an AUC value of 0.811 (95% CI 0.71–0.93, p < 0.001) (Fig. 3f). The sensitivity and specificity were 78.57% and 80%, respectively. The cut-off value was 1.53. Soon afterwards, 85 GC patients were divided into high miR-1246 group (n = 42) and low miR-1246 group (n = 43) according to the median value of exosomal miR-1246 expressions. Chi-square test showed that high exosomal miR-1246 expressions were significantly correlated with advanced TNM stage (p = 0.033) and distant metastasis (p = 0.035) (Table 2).

Table 1 The clinicopathological characteristics of participants
Full size table
Fig. 3
figure 3

High levels and diagnostic utility of exosomal miR-1246 in serum validated in an independent cohort. a, b The expression levels of exosomal miR-1246 in serum in GC patients and HCs. c The expression levels of exosomal miR-1246 in serum in GC patients with TNM stage I, patients with BDs, and HCs. d ROC of circulating exosomal miR-1246 expressions to differentiate GC patients from HCs. e–f ROC of circulating exosomal miR-1246 expressions to differentiate GC patients with TNM stage I from HCs and patients with BDs

Full size image
Table 2 Associations between exosomal miR-1246 expressions in serum and clinicopathological parameters of GC patients
Full size table

Bioinformatics analysis of miR-1246

Considering that miR-1246 was significantly downregulated in GC tissues and that the regulatory mechanisms of miR-1246 had not been reported in GC, we conducted a bioinformatics analysis of miR-1246 in GC. Firstly, 659 upregulated genes and 1818 downregulated genes with ∣logFC∣ > 1 were found in another GEO dataset (GSE112369) which means these genes may modulate the development and progression of GC (Fig. 4a). Potential targets of miR-1246 were searched by public bioinformatics algorithms (TargetScan). Among these predicted targets of miR-1246, 285 targets were significantly dysregulated in GC tissues (Fig. 4b). Thereafter, we carried out functional and pathway enrichment analysis of these significantly dysregulated targets. GO term enrichment analysis showed that in the BP category, targets of miR-1246 were significantly enriched in positive regulation of cell proliferation, positive regulation of transcription, positive regulation of gene expression, palate development, and negative regulation of apoptotic process (Fig. 4c). CC analysis showed that most targets were located in plasma membrane (Fig. 4d). According to the results of MF analysis, targets of miR-1246 were associated with protein binding, including transforming growth factor beta binding, sequence-specific DNA binding, and growth factor binding (Fig. 4e). Additionally, KEGG pathway analysis discovered that miR-1246 participated in several pathways in cancer, including Hippo signaling pathway, FoxO signaling pathway, PI3K-Akt signaling pathway, and GnRH signaling pathway (Fig. 4f). If there were more than ten terms enriched in these categories, the top ten terms based on p value were chosen. Furthermore, we carried out a PPI network analysis (Fig. 4g) and identified a molecular complex (Fig. 4h) using Cytoscape.

Fig. 4
figure 4

Bioinformatics analysis of miR-1246. a Differentially expressed genes in GSE112369. b Potential targets that were significantly dysregulated in GC tissues. c–e Functional enrichment analysis, f KEGG pathway analysis, and g PPI network analysis of dysregulated targets of miR-1246. h Molecular complex identified in PPI network

Full size image

Discussion

In this study, we discovered that circulating exosomal miR-1246 were significantly upregulated in GC patients and elevated exosomal miR-1246 in serum were tumor-derived. ROC analysis exhibited that circulating exosomal miR-1246 expressions could differentiate GC patients in TNM stage I from patients with BDs and HCs with remarkable value. Besides, based on bioinformatics analysis, we discovered targets of miR-1246 mainly located in cellular membrane and regulated protein binding to promote proliferation and transcription of GC cells through several signaling pathways.

Dysregulated miRNAs in peripheral blood acting as diagnostic biomarkers have been explored in various cancers. Liu et al. reported that circulating miR-182 could serve as diagnostic biomarker of colorectal cancer [27]; Zhang et al. discovered serum miR-100 was a potential biomarker for detection and outcome prediction of glioblastoma patients [28]; Orangi et al. evaluated that miRNA-9 and miRNA-34a may be potential biomarkers for diagnosis of breast cancer in Iranian women [29]. Circulating miR-1246 has been regarded as a novel biomarker for oesophageal squamous cell carcinoma [30], high-grade serous ovarian carcinoma [31], aggressive prostate cancer [32], and hepatocellular carcinoma [33]. Based on a systematic review and meta-analysis, Wei et al. declared exosomal miR-1246 in body fluids is a potential biomarker for gastrointestinal cancer [34]. However, gastrointestinal cancer enrolled in their meta-analysis only consists of pancreatic cancer [35], colon cancer [36], and pancreatobiliary tract cancer [37]. To the best of our knowledge, the diagnostic value of circulating exosomal miR-1246 in GC has not been reported till now. Our study demonstrated that elevated exosomal miR-1246 in serum were tumor-derived and remarkably higher in GC patients than HCs as well as patients with BDs which provided a novel potential biomarker for the early diagnosis of GC patients. Unexpectedly, compared to normal tissues, the expressions of miR-1246 in GC tissues were significantly downregulated. In addition to our finding that GC cells could forwardly sort miR-1246 into exosomes and release it into peripheral blood, a various of cancers have been verified to release exosomal miR-1246, including oral squamous cell carcinoma, breast cancer [24, 38,39,40]. For example, exosomal miR-1246 secreted from breast cancer cells could be transferred into non-malignant breast cells through direct uptake to promote cell proliferation, invasion and drug resistance by targeting CCNG2 [24]. Besides, several RNA binding proteins have been proved to control the sorting of specific miRNAs into exosomes, regardless of the downregulated expressions of miRNAs, such as hnRNPA2B1 [25, 26]. Xu et al. discovered exosome miR-1246 is derived from RNU2-1 degradation [41]. We speculated that in spite of downregulated expression in GC cells, miR-1246 was regarded as a kind of metabolite. Because the metabolism of tumorous cells was faster than normal ones [42], to keep homeostasis, tumors had to secret miR-1246 into extracellular environment. Therefore, it is meaningful to investigate the relationship between exosomal miR-1246 and tumor microenvironment in the future.

Several studies have reported the biological functions of miR-1246 in cancers. Lin et al. discovered that miR-1246 expressions were significantly upregulated in oral carcinomas and enhanced cancer stemness and chemoresistance by targeting CCNG2 [43]. In addition, Zhang et al. revealed miR-1246 was a tumor-initiating miRNA and promoted non-small cell lung cancer progression [44]. However, the regulatory mechanisms of miR-1246 in GC have not been investigated before. Our study demonstrated that miR-1246 expressions were significantly downregulated in GC tissues. Functional and pathway enrichment analysis revealed miR-1246 could regulate protein binding to promote proliferation and gene transcription through several vital signaling pathways, such as Hippo signaling pathway. PPI network of targets revealed a molecular complex which consisted of KRAS, SOX2, ABCG2, NT5E, KITLG, MME, PROM1, and KIT. Tian et al. reported SOX2 is a commonly activated tumor promoter that activate AKT signaling in GC [45]. Polom et al. declared KRAS mutation was significantly associated with the prognosis of GC patients [46]. Besides, Jiang et al. showed that high expression of NT5E was significantly associated with lymph node metastases and a promising prognostic factor in different types of cancer [47]. Our study suggested that targeting this molecular complex may be a novel therapeutic strategy of GC.

Taken together, we demonstrated that exosomal miR-1246 in serum was a novel potential biomarker for the early diagnosis of GC patients.