Background
Gastric cancer (GC) is a frequent tumor type across the globe [
1]. The treatment conditions for GC have been ameliorated, but patients still face a grim survival status [
2]. Therefore, early diagnosis and effective treatment are particularly significant. The study of the pathogenesis of GC can offer a rationale behind new treatment techniques.
Recently, the mechanism of cancer-associated fibroblasts (CAFs) in GC has become one of the research hotspots [
3]. In the tumor microenvironment, cancer cells activate non-carcinoma fibroblasts (NFs) by secreting some growth factors and induce them to transform into CAFs. In turn, CAFs also secrete more growth factors to promote the malignant progression of cancer cells [
4]. Microvesicles (MVs) are one of the important substances secreted by CAFs. MVs are some vesicles that are detached from cells during physiological or pathological processes. Their size is generally 100–1000 nm. They can carry nucleic acids, lipids and proteins, and deliver them to nearby or distal cells to affect the function and behavior of the recipient cells. MVs can promote cancer cell angiogenesis, drug resistance, proliferation and metastasis in different cancer types by transmitting growth factors, signaling molecules, DNA, coding and non-coding RNA, etc [
5‐
7]. Hence, it is of great meaning to study the specific regulatory effect of MVs in GC CAFs on cancer cells.
MicroRNAs (miRNAs) can bind to mRNAs to achieve post-transcriptional regulation in cells, thereby affecting a variety of cell biological functions [
8]. miR-223-3p can modulate solid tumor development. For example, miR-223-3p boosts the malignant progression of prostate cancer by targeting SEPT6 [
9]. miR-223-3p suppresses glioma cell proliferation via targeting inflammation-related cytokines [
10]. Meanwhile, one study revealed that miR-223-3p can also aggravate the malignant progression of GC [
11].
Sorbin and SH3 domain-containing protein 1, also CAP/ponsin (
SORBS1) is an adaptin in nature [
12] that interacts with cytoskeleton regulators to mediate cytoskeleton structural organization [
13], and cell spreading and movement [
14].
SORBS1 can also hamper tumor metastasis and enhance the sensitivity of cancer to chemotherapy drugs [
15]. Based on the existing research background, we designed a series of experiments to further relevant mechanisms. Through experiments, it was uncovered that CAFs-secreted MVs carried miR-223-3p in GC tissue targeted
SORBS1, and played a cancer promotor role. This study has practical implications for future GC treatment and drug development.
Materials and methods
GSE93415 dataset (cancer tissue (CT): 20; corresponding para-cancerous tissue (PT): 20) of GC tissue was offered by the Gene Expression Omnibus (GEO) database. R package limma was applied for differential analysis for miRNAs (padj < 0.05, |logFC|> 1.5). Combined with the literature, the target miRNA (miR-223-3p) was determined. Then, miR-223-3p expression in tissue, cells, tissue fluid, blood exosomes and MVs was predicted using EVmiRNA database. The targets of miR-223-3p were predicted using miRDB, mirDIP, and Targetscan tools. The Cancer Genome Atlas (TCGA)-STAD data were obtained, and edgeR package was utilized for differential analysis (|logFC|> 2, padj < 0.05) to acquire differentially expressed mRNAs (DEmRNAs). The predicted target genes were overlapped with the DEmRNAs to determine the mRNA targeted by miR-223-3p.
Clinical samples
In this study, 20 cases of CT samples and 20 corresponding PT samples (more than 10 cm away from the negative margin) were gathered from GC patients in The Second Affiliated Hospital of Zhejiang University School of Medicine from 2018 to 2020. Each CT and paired PT samples were from the same patient. Part of the tissue samples were collected for the isolation of fibroblasts, and the other part were stored at − 80 ℃ until use. All cases were assessed by pathologists based on histopathology, re-checked by pathologists and defined with GC (American Joint Committee on Cancer Version II, III, IV, VII). This study got approval from the Ethics Committee of The Second Affiliated Hospital of Zhejiang University School of Medicine with written informed consent from all participants.
Extraction of CAFs, NFs and tumor cells (TCs)
Fresh CT or PT samples were rinsed with Dulbecco’s Modified Eagle medium (DMEM) (Thermo Fisher Scientific, USA) without serum, and transferred to 0.15% collagenase IV (Thermo Fisher Scientific, USA) solution. Then at 37 °C, they were cultivated (Thermo Fisher Scientific, USA) with 5% CO2 for 40 min. The fully digested cells were filtered through a 40 mm cell filter (BD Biosciences, USA) and centrifuged for 10 min (1500 rpm). Cell suspension was cultured in fibroblast medium (Innoprot, Spain) for 24 h for adherent growth. After 24 h, the nonadherent cells were washed off and the adherent ones were subcultured further. The fibroblasts isolated from GC tissue were used as CAFs, and the fibroblasts isolated from adjacent tissue were used as NFs. TCs were isolated from fresh GC tissue using the Cancer Cell Isolation Kit (Thermo Fisher Scientific, USA).
Cell culture
Human GC cell line SGC7901 (GDC150) was acquired from China Center for Type Culture Collection (CCTCC). Human GC cell lines AGS (BNCC309318) and BGC-823 (BNCC337689), and normal gastric mucosal cell line GES-1 (BNCC337970) were offered by BeNa Culture Collection (China). The above cell lines were cultivated in Roswell Park Memorial Institute (RPMI) 1640 medium (Media, USA) containing 10% fetal bovine serum (FBS) (Thermo Fisher Scientific, USA) and cultured at 37 °C with 5% CO2 (Thermo Fisher Scientific, USA). CAFs and NFs isolated from tissue samples were cultured in fibroblast medium (Innoprot, SPIAN) containing 10% FBS (Thermo Fisher Scientific, USA) under routine conditions.
Cell transfection
miR-223-3p mimic, miR-223-3p inhibitor, si-SORBS1 and corresponding negative controls (NCs) were designed and provided by Shanghai GenePharma Co. Ltd. (China). Cell transfection was undertaken using Lipofectamine 3000 (Thermo Fisher Scientific, USA).
MVs were extracted from cell culture medium by hypervelocity centrifugation. Briefly, the supernatant of the cell medium was first taken, and the debris was removed by centrifugation at 3000×g for 10 min. Then a 100,000×g centrifuge was conducted at 4 °C for 2 h using an ultracentrifuge (Thermo Fisher Scientific, USA). Finally, the extracted MVs were resuspended in phosphate buffer saline (PBS) to prepare for further transmission electron microscope (TEM) observation. The concentration of MVs was evaluated by bicinchoninic acid assay (BCA, Thermo Scientific, USA).
TEM
The MV suspension was loaded into a carbon film-coated TEM copper grid, stained with uranyl acetate, and then dried. MVs were observed and photographed using Hitachi JEM-2100 TEM (Japanese Electronics Co., Ltd., Tokyo, Japan).
Co-culture of MVs and GC cells
5 × 105 GC cells were inoculated in 25-cm2 culture dish. 100 ug MVs were added to 5 ml complete medium. 24 h later, GC cells were gathered for the following steps. MVs co-culture was conducted as the MV treatment group in the cell experiment.
qRT-PCR
RNA was isolated using the miRNeasy mini kit (Qiagen, Germany). The purity and concentration of the extracted RNA were detected by NanoDrop 2000. The corresponding complementary DNA (cDNA) was obtained by reverse transcription with miScript II RT (Qiagen, Germany). The expression level was ascertained through qRT-PCR by using miScript SYBR Green PCR Master Mix (Qiagen, Germany). The qPCR detection was performed using Real-Time PCR on ABI7500 (Thermo Fisher Scientific, USA). U6 and GAPDH were selected as internal references. All PCR primers were designed by Shanghai Genepharma Co, Ltd. (China) as displayed in Table
1.
Table 1
qRT-PCR primer sequence
miR-223-3p | F: AGCTGGTGTTGTGAATCAGGCCG |
R: TGGTGTCGTGGAGTCG |
SORBS1 | F: ATTCCCAAGCCTTTCCATCAG |
R: TTTTGCTGTTCTCGATTGTGTTG |
U6 | F: CTCGCTTCGGCAGCACA |
R: AACGCTTCACGAATTTGCGT |
GAPDH | F: GAACGGGAAGCTCACTGG |
R: GCCTGCTTCACCACCTTCT |
Western blot
MVs and cells were cleaved in radioimmunoprecipitation assay (RIPA) cell lysis reagent containing the protease inhibitor. The equivalent protein samples (15 μg) were electrophoresed on 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis gel, and then mounted on polyvinylidene fluoride (Bio-Rad Laboratories, Inc., USA). After SDS-PAGE, the membrane was blocked with 5% skimmed milk at room temperature for 1 h and then incubated overnight with the primary antibodies at 4 °C. The primary antibodies were rabbit anti-human antibodies: anti-α-SMA antibody (ab5694, Abcam, UK), anti-FAP antibody (ab207178, Abcam, UK), anti-CEA antibody (ab207718, Abcam, UK), anti-CK-18 antibody (ab133263, Abcam, UK), anti-CD63 antibody (ab134045, Abcam, UK), anti-E-cadherin antibody (ab40772, Abcam, UK), N-cadherin antibody (ab76011, Abcam, UK), anti-vimentin antibody (ab92547, Abcam, UK), anti-SORBS1 antibody (ab224129, Abcam, UK), and anti-GAPDH antibody (ab6721, Abcam, UK). The secondary antibody was horseradish peroxidase-labeled goat antirabbit IgG antibody (ab6721, Abcam, UK).
CCK-8 assay
The pretreated GC cells were inoculated into 96-well plates (5 × 103 per well) and cultured for 0, 24, 48, and 72 h, respectively. Afterward, 10 μl CCK-8 solution was added at each period (MedChem Express, USA). The absorbance at 450 nm was tested with a microplate analyzer.
The pretreated GC cells (5 × 102 per well) were plated into 12-well plates and incubated in cell medium with 10% FBS for two weeks. After completion of culture, the cell colonies were fixed with 10% formaldehyde and stained using crystal violet dye. The staining results were photographed and then analyzed using Image J software.
Transwell assay
For invasion assay, 50 μl matrix gel U5GL (BD Biosciences, USA) was applied to the upper chamber. The pretreated GC cells (2 × 104 per well) were suspended in 100 μl medium without serum and inoculated in the upper chamber (BD Biosciences). The lower chamber was added with 500 μl medium containing 20% FBS. After 48 h incubation, the uninvaded cells were wiped with cotton swabs, and the remaining cells were fixed with 4% polyformaldehyde and stained using 0.1% crystal violet for 15 min. Then, 5 random visual fields per chamber were selected for counting cells under a microscope (100 ×). The cell migration experiment was similar to the above steps, except for the following differences: (1) no matrix gel coating was required at the upper surface of the insert; (2) 24 h of incubation.
Dual-luciferase reporter gene detection
Firstly, the targeting sequence of miR-223-3p in SORBS1 3’- untranslated region (UTR) was predicted by miRDB. The wild-type (WT) and mutant (MUT) SORBS1 3’-UTR were then amplified by PCR and introduced into the pMIR-REPORT vectors (AddGene, USA). pRL-TK plasmid was the internal reference luciferase reporter plasmid (AddGene, USA). Finally, miR-223-3p or corresponding NC and constructed reporter gene plasmid were co-transfected into GC cells. After 48 h, fluorescence activity was evaluated on the Dual-Luciferase® Reporter Assay System.
Cell uptake of MVs
MVs were stained using PHK-26 (Umibio, China). The PHK-26 labeled MVs were co-incubated with GC cells for 24 h under general conditions. A fluorescence microscope (Nikon, Japan) was employed to observe the initial imaging and imaging after 0, 6, and 24 h. Flow cytometer was utilized for analyses and the data acquired were subject to calculation of median fluorescence intensity (MFI) of GC cells (MVs carrying MVs).
Xenograft mouse models
GC cells (2 × 106 cells/ml) were injected into the lower left limb of 10 BALB/c nude mice (Laboratory Animal Resources, Chinese Academy of Sciences, China) aged 4–6 weeks. After 3 days, tumor mass formation was observed. On day 8, nude mice were divided into 3 groups randomly (PBS group; inhibitor NC group and inhibitor group, 5 mice in each group), and MVs (miR-223-3p inhibitor and its NC were respectively transfected into CAFs and then MVs were isolated. PBS was added to attenuate the MV suspension with a protein concentration of 0.5 ug/ul. 15 μg/mouse) or corresponding volume of PBS were injected into tumor mass of nude mice in different groups (marked as day 0). The injection was done every 3 days, before which MV suspension was blown and mixed evenly. The mass size was measured with a vernier caliper every 3 days according to formula 1/2 (length × width2). All the mice were euthanized by carbon dioxide asphyxiation on day 24, and death was confirmed after 5 min of observation. When performing euthanasia using CO2 asphyxiation, the mice were in the induction phase and provided with air. The CO2 concentration was constantly increased until the mice were confirmed by cardiac arrest and respiratory arrest. Besides, the CO2 was at a flow displacement rate of 30%-70% of the chamber volume per min, so as to ensure that the mice were unconscious prior to pain. The tumor masses were removed and weighed. The tumor masses were also used for subsequent qRT-PCR detection and immunohistochemical (IHC) assay.
IHC assay
The paraffin-embedded tissue fixed with formaldehyde was sectioned (4 um), and then the tissue was placed in an oven at 60 °C and heated for 30 min before being dewaxed with xylene. The endogenous peroxidase was then removed by incubation with 3% H2O2 for 10 min. After blocking, the sections were added with primary anti-Ki67 (ab15580, Abcam, UK) to incubate overnight at 4 ℃. On the next day, after being washed by PBS, sections were reacted with secondary antibody IgG H&L (HRP) at 37 ℃ for 1 h, and then colored by DAB (3,3’-diaminobenzidine). The staining was observed with an upright optical microscope.
RNA immunoprecipitation (RIP)
This step was taken using RNA binding protein immunoprecipitation kit (Millipore, Bedford, MA). miR-223-3p mimic-treated GC cells were lysed with RIPA buffer (Cell Signaling Technology) containing phosphatase and proteases inhibitors (Sigma-Aldrich). Magnetic beads (Invitrogen) were incubated with IgG antibody (Cell Signaling Technology) or Ago2 for 30 min. The lysate was subjected to immunoprecipitation and rotated at 4 °C. RNA was purified from RNA–protein complex binding beads for qRT-PCR analysis.
Flow cytometry
Pre-treated cells were rinsed with cold PBS and fixed with 80% ethanol. The cells were then centrifuged in a cold spinning machine and re-suspended in cold PBS. Being cultured at common temperature for 30 min, cells were added with propidium iodide (Sigma-Aldrich; 20 mg/ml) and bovine pancreatic RNAase (Sigma; 2 mg/ml) for 20 min of incubation. 2 × 104 cells were analyzed using BD FACSCanto, the data of which were assessed using FLOWJO software (Tree Star, Inc, Ashland, OR).
Cell apoptosis detection kit (Roche, USA) was utilized for TUNEL. Paraffin-embedded sections were subjected to gradient hydration, fixation with 4% formaldehyde, and incubation with protease K at common temperature for 15 min. 3% hydrogen peroxide was utilized to block endogenous peroxidase. Fresh TUNEL reaction solution containing rTdT was prepared. The sections were washed with PBS and counterstained with hematoxylin. Apoptotic cells were tested with a microscope (Nikon, Japan).
Data analysis
Data processing was performed using GraphPad Prism (GraphPad Software, USA). All assays were performed at least 3 times and the results were exhibited as mean ± standard deviation. One-way analysis of variance was used to compare the differences between groups, and t-test was used for post hoc test. P < 0.05 represented statistically remarkable, and * in the figure denotes p < 0.05.
Discussion
GC is caused by a variety of factors, including helicobacter pylori infection, smoking, dietary habits, and genetic factors [
1]. Though the molecular mechanism of GC is not yet clear, a recent study showed that CAFs is crucial in promoting the development of GC, and its main mechanism is to secrete various growth factors, cytokines and RNA to promote the malignant progression of GC tissue, such as angiogenesis, drug resistance, proliferation, migration and invasion [
3]. In the above processes, MVs often act as carriers to maintain the interaction between CAFs and cancer cells by transmitting signaling molecules [
18]. Herein, it was uncovered that the MVs secreted by CAFs carried miR-223-3p and delivered it into GC cells, and accelerated the malignant progression of GC cells by mediating
SORBS1.
miR-223-3p was highly expressed in GC in the previous bioinformatics analysis and the expression detection. This result is consistent with several published studies [
11,
19,
20]. For example, Yiping Zhu et al. [
11] found that compared with PT, miR-223-3p expression is high in GC tissue, and miR-223-3p level in GC tissue is positively correlated with lymph node metastasis and invasion depth. Similarly, it was discovered here that miR-223-3p was relatively highly expressed in GC tissue or GC cell lines.
At present, MVs have not been fully studied in GC, but some studies showed that specific MVs can influence GC or can be used as markers of GC [
21‐
23]. Malgorzata Stec et al
. [
23] demonstrated that MVs derived from tumors can deliver signaling molecules to GC cells and promote tumor growth. Similar experimental results were also obtained in our study, but the difference was that the MVs in our study were derived from CAFs. Hence, it could be seen that there may be different sources of MVs in GC tissue that simultaneously affected the behavior and function of GC cells.
In fact, there are few studies on
SORBS1 in cancer, and no studies have illustrated
SORBS1 expression in GC and its mechanism. However, studies related to breast cancer showed that
SORBS1 is lowly expressed in breast cancer patients [
24], and forced expression of
SORBS1 can suppress the metastasis of cancer cells and improve the sensitivity of cancer cells to chemotherapy drugs [
15]. Similarly, it was noted from our experimental results that
SORBS1 was less expressed in GC cell lines. miR-223-3p regulated the progression of GC cells by targeting
SORBS1. Hence, the role of
SORBS1 in GC is similar to that in breast cancer. Down-regulating
SORBS1 by miR-223-3p with high expression in CAFS-MVs may provide a mechanism for promoting the development of GC, but it is by no means the only one. miR-223-3p can accelerate GC cell processes by regulating multiple target genes and pathway proteins (Arid1a, NLRP3, and NDRG1). Whether other genes and pathways are implicated in GC biological functions requires further research.
Taken together, our study indicated that CAFs-derived MVs carried miR-223-3p and delivered it into GC, and targeted SORBS1 to boost the cell proliferation, migration, invasion, and EMT process, and modulate cell apoptosis and cell cycle in GC. The study of this mechanism also offers a novel theoretical basis for GC diagnosis and therapy.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.