microRNA-10a-5p from gastric cancer cell-derived exosomes enhances viability and migration of human umbilical vein endothelial cells by targeting zinc finger MYND-type containing 11.

Tumor-derived exosomes (exo) could modulate the biological behaviors of human umbilical vein endothelial cells (HUVECs). Here, the role of microRNA (miR)-10a-5p-modified gastric cancer (GC) cells-derived exo for HUVECs was studied. GC tissue specimens were collected, and miR-10a-5p and zinc finger MYND-type containing 11 (ZMYND11) levels were determined. HUVECs interfered with ZMYND11 or miR-10a-5p-related oligonucleotides. Exo was extracted from GC cells (HGC-27 exo), and miR-10a-5p mimic-modified HGC-27 exo were co-cultured with HUVECs. HUVECs viability, migration and angiogenesis were evaluated, and miR-10a-5p/ZMYND11 crosstalk was explored. It was observed that GC patients had raised miR-10a-5p and reduced ZMYND11, and miR-10a-5p negatively mediated ZMYND11 expression. Suppression of miR-10a-5p or overexpression of ZMYND11 inhibited viability, migration and tube formation ability of HUVECs. Notably, miR-10a-5p mimic-modified HGC-27 exo enhanced the viability, migration and tube formation ability of HUVECs, but this effect was impaired after up-regulating ZMYND11. In summary, miR-10a-5p from GC cells-derived exo enhances viability and migration of HUVECs by suppressing ZMYND11.

Introduction

Gastric cancer (GC) is still a widespread disease [1]. East Asia, Eastern Europe and South America have hot spots in the incidence and mortality of GC and the disease incidence in men are twice that of women [2]. Helicobacter pylori infection accounts for the main cause of GC, and the prognosis of GC patients is dismal partially due to clinical silence in the early stage [3]. Endoscopic ultrasound, computerized tomography and laparoscopy are the effective modalities for the staging of GC [4]. Moreover, gratifying the signs of progress have been developed to treat GC, including neoadjuvant chemotherapy, radiotherapy, immunotherapy and molecular-targeted therapies [5]. Also, a thorough insight into GC-oriented mechanisms provides a niche for better control of GC.

Tumor cell-derived exosomes (exo) are extracellular vehicles (EVs) that have multiple effects on the tumorigenic progression. For example, tumor cells-derived exo could enhance angiogenesis in lung cancer [6], and oral squamous cell carcinoma-derived exo induce migration and tube formation of human umbilical vein endothelial cells (HUVECs) [7]. GC cells-derived exo could strengthen the migratory phenotype, induce autophagy and activation of neutrophils in GC [8], and remodel the pre-metastatic microenvironment in GC [9]. microRNA (miRs) are endogenous non-coding RNAs, possessing prognostic and therapeutic values in GC [10]. Synergism of miRs and GC cells-secreted exo has applicable potential to treat tumors. GC cells-derived exosomal miR-21-5p is pro-metastatic [11] and GC cells-secreted exosomal miR-130a is pro-angiogenic [12]. As a member of miRs, miR-10a-5p from cancer-associated fibroblast (CAF)-derived EVs could enhance angiogenesis and tumor growth in cervical squamous cell carcinoma (CSCC) [13], and CAF-derived exosomal miR-10a-5p stimulates the malignant phenotype of pancreatic cancer cells [14]. Through bioinformatics analysis, zinc finger MYND-type containing 11 (ZMYND11) was identified as a downstream effector of miR-10a-5p. ZMYND11 refers to an H3.3-specific reader of H3K36me3 [15] that links the control of transcription elongation with tumor suppression [16]. Yang JP et al. have established that deregulation of ZMYND11 acts reversely to augment the tumorigenicity in glioblastoma multiform (GBM) [17].

We hypothesize that miR-10a-5p-modified GC cells-derived exo aggrandize angiogenesis and tumor growth in GC by silencing ZMYND11, hoping to develop a potential treatment approach for GC.

Methods and Materials

Ethics statement

Written consents were obtained from all patients. This research was implemented with the approval of the Ethics Committee of The First Affiliated Hospital, and the College of Clinical Medicine of Henan University of Science and Technology.

Samples

GC specimens and normal gastric mucosal tissues were obtained from 61 patients admitted to The First Affiliated Hospital, and the College of Clinical Medicine of Henan University of Science and Technology and all GC patients were clinically confirmed by two pathologists via histology or pathology. All specimens were stored at −80°C.

Cell culture

Gastric cancer cell line SGC-7901, MKN45, SNU-520, HGC-27 and AGS and normal gastric epithelial cell line GES-1 (Yaji Biotechnology Co., Ltd., Shanghai, China) were maintained in the Roswell Park Memorial Institute (RPMI) 1640 medium that was supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin [18]. The media were provided by HyClone, Thermo Scientific (UT, USA). HUVECs (ATCC®CRL-1730™) were cultivated in 10% FBS-RPMI 1640 medium (GIBCO, NY, USA) [19].

Overexpression (oe)-ZMYND11, miR-10a-5p inhibitor, miR-10a-5p mimic and their negative controls (NC) (GenePharma, Shanghai, China) were transfected into HUVECs. Transfection was performed using lipofectamine 2000 reagent (Invitrogen, CA, USA).

Exo isolation and identification

HGC-27 cells at 80% confluence were placed in FBS-free RPMI-1640 for 48 h. Subsequently, the conditioned medium was continuously centrifuged at 300 × g, 2000 × g and 110,000 g. Then, the supernatant was filtered using a 0.22 μm filter and centrifuged at 110,000 × g again. The precipitate (exo) was detected by Micro bicinchoninic acid (BCA) kit (Thermo Fisher, MA, USA) to measure protein concentration. The purity of exo was identified by transmission electron microscopy (TEM) and nanoparticle size analysis (NTA) while the exosomal markers CD63 (1:1,000), Alix (1:1,000, both from Santa Cruz Biotechnology, CA, USA) and TSG101 (1:500, Abcam, MA, USA) were tested by Western blot [20].

Co-culture of HUVECs and GC cells

PKH26-labeled exo (100 μg/mL) were incubated with HUVECs for 24 h and photographed under a fluorescence microscope (AF6000, Leica, Wetzlar, Germany).

HUVECs were incubated with HGC-27 cells-derived exo and PBS, respectively, and named exo group and PBS group. The exo from miR-10a-5p mimic or its NC-modified HGC-27 cells were incubated with HUVECs and named exo-miR-10a-5p mimic group or exo-mimic NC group. HUVECs co-cultured with exo-miR-10a-5p mimic were further transfected with oe-ZMYND11 or its NC, named exo-mi-miR-10a-5p + oe-ZMYND11 group or exo-mi-miR-10a-5p + oe-NC group [21].

3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay

HUVECs dispersed in the 96-well plate (2 × 103 cells/well) were co-cultured with HGC-27 exo for 12 h, supplemented with MTT solution (20 nmol/L; Beyotime, Shanghai, China) for 4 h and dissolved by 150 μL dimethyl sulfoxide. Absorbance at 450 nm was read [22].

Cell counting kit (CCK)-8 test

CCK-8 kit (MAO218, Meilun Biomeilunbio® Company, Dalian, China) was used. After seeding in 96-well plates, HUVECs were added with 10% CCK-8 reagent on the 1, 2, 3, 4, and 5 d, and incubated for 2 h. A microplate reader (Stat Fax 2100; Awareness, USA) was employed to measure the optical density450 nm [23].

Colony formation assay

HUVECs were detached, made into cell suspension and cultured for 14 d. Then, HUVECs were fixed with 4% paraformaldehyde, performed GIMSA staining and air-dried [20].

Wound healing test

HUVECs on the 6-well plates were grown to 80% confluence and scraped with a 10 mL pipette tip to form two linear areas. Next, HUVECs were cultivated in an FBS-free medium and photographed at 0 and 24 h [24].

Transwell assay

An 8-μm transwell (BD Falcon, NJ, USA) was inserted into a 24-well plate. The transwell was coated with 50 μL Matrigel and 1 × 105 HUVECs were added in the upper chamber. The invading cells were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet. Images were captured using a stereomicroscope (Leica) [25].

Tube formation assay

Matrigel (BD Biosciences, NJ, USA) was polymerized in the 24-well plate (100 mL/well). HUVECs were resuspended in an FBS-free medium, transferred to the plate (1 × 105 cells/well) and examined under an optical microscope. Branch points of the formed tube were quantified in at least five fields [26].

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

Total RNA was isolated from tissues and cells with Trizol reagent (Invitrogen). To detect mRNA, cDNA was prepared using Applied Biosystems High-Capacity cDNA Reverse Transcription Kit and the qRT-PCR was performed using SuperScript III Platinum One-Step qRT-PCR Kit (Thermo Fisher Scientific, Schwerte, Germany). To detect miR-10a-5p, cDNA was prepared using the miScript SYBR Green PCR Kit (Qiagen). All reactions were performed on the ABI 7900 real-time PCR system (Applied Biosystems). All primers were listed in Supplementary Table 1. miRNA or mRNA was normalized to U6 or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and gene expression was measured using the 2−∆∆Ct method.

Western blot assay

The total protein was extracted by centrifugation at 15,000 × g. Next, the protein concentration was measured using a BCA kit (Pierce, IL, USA). The protein was then separated by electrophoresis and transferred to a polyvinylidene fluoride membrane, followed by blockade with 5% bovine serum albumin and incubation with primary antibodies: rabbit anti-human ZMYND11 (1:1000, Millipore, MA, USA), Vascular endothelial growth factor (VEGF, 1:1000), Matrix Metalloproteinase-9 (MMP-9, 1:1000), GAPDH (1:10,000, Abcam) and horseradish peroxidase-labeled goat anti-rabbit immunoglobulin G (ab205718, 1:20,000, Abcam). The bands were developed using enhanced chemiluminescence and quantified using ImageJ 1.48 u software (Wayne Rasband, National Institutes of Health, MD, USA) [27].

Dual-luciferase reporter gene assay

The binding sites of miR-10a-5p and ZMYND11 were searched on Starbase 2.0 (http://starbase.sysu.edu.cn/index.php). In the light of the prediction results, the mutant and wild sequences of ZMYND11 and miR-10a-5p binding sites were designed, respectively, which were then inserted into the luciferase reporter gene vector (pGL3-Basic), named as mutant type (MUT)-ZMYND11 and wild-type (WT)-ZMYND11, respectively. The recombinant vector was co-transfected with 30 nM miR-10a-5p mimic or its NC into HEK-293 T cells. After that, the activity of firefly luciferase and Renilla luciferase was evaluated [28].

Statistical analysis

GraphPad Prism 6.0 (GraphPad Software Inc., CA, USA) and SPSS 20.0 (IBM, NY, USA) were applied to data analysis. Data are presented as mean ± standard deviation. The two groups were compared by t-test, and multiple groups by one-way analysis of variance (ANOVA). P  < 0.05 was of statistical significance. All experiments were repeated in triplicate.

Results

miR-10a-5p is up-regulated in GC; down-regulating miR-10a-5p inhibits angiogenesis

miR-10a-5p can promote the migration and invasion of human pancreatic ductal adenocarcinoma cells [29]. RT-qPCR verified miR-10a-5p high expression in GC tissues (Figure 1a) also found that versus GES-1 cells, miR-10a-5p was up-regulated in GC cell line SGC-7901, MKN45, SNU-520, HGC-27 and AGS, and was mostly up-regulated in HGC-27 cells (Figure 1b).

Figure 1.

miR-10a-5p is up-regulated in GC; down-regulating miR-10a-5p inhibits angiogenesis. A-B. RT-qPCR detection of miR-10a-5p expression in tissues and cells; C. RT-qPCR detection of miR-10a-5p expression in cells; D. MTT detection of cell viability; E. CCK-8 test detection of cell proliferation; F. Colony formation analysis of cell viability; G. Wound healing test analysis of cell migration; H. Transwell analysis of cell invasion; I. Tube formation test analysis of tube-forming ability; J. Western blot analysis of VEGF and MMP-9 expression. # P < 0.05 vs. the GES-1; * P < 0.05 vs. the inhibitor NC group; the measurement data were expressed as the mean ± standard deviation, and the comparison was analyzed by the t-test or one-way ANOVA.

miR-10a-5p inhibitor-treated CAF-EVs could reduce tumor volume and blood vessels around the tumor [13]. To survey the effect of miR-10a-5p on angiogenesis, we used miR-10a-5p inhibitor to suppress miR-10a-5p expression in HUVECs (Figure 1c). In addition, we measured the viability, invasion, migration and angiogenesis capacity of HUVECs via MTT, CCK-8 assay, wound healing test, transwell and tube formation assays, as well as VEGF and MMP-9 protein expression via Western blot. Experimentally, we noticed that miR-10a-5p inhibitor-modified HUVECs manifested inhibited viability, migration and tube formation abilities (Figure 1d-j). It was indicated that down-regulating miR-10a-5p inhibited angiogenesis.

miR-10a-5p negatively regulates ZMYND11 expression

Then, we determined the low expression of ZMYND11 in GC tissues (Figure 2a) and GC cell lines (Figure 2b) through RT-qPCR. Next, we searched for genes mediated by miR-10a-5p in the Starbase and found that miR-10a-5p and ZMYND11 had binding sites (Figure 2c). Then, the luciferase reporter gene assay revealed that miR-10a-5p bound with ZMYND11 (Figure 2d).

Figure 2.

miR-10a-5p negatively regulates ZMYND11 expression. A-B. RT-qPCR detection of ZMYND11 expression in tissues and cells; C. The binding site of miR-10a-5p and ZMYND11 through Starbase; D. Dual-luciferase reporter gene determination of the binding of miR-10a-5p and ZMYND11; E-F. Western blot detection of ZMYND11 expression; G. Pearson correlation analysis of the correlation between miR-10a-5p and ZMYND11; $ P < 0.05 vs. GES-1; # P < 0.05 vs. the mimic NC group; & P < 0.05 vs. the inhibitor NC group; the measurement data were expressed as the mean ± standard deviation, and the comparison was analyzed by the t-test or one-way ANOVA.

Then, the regulation of miR-10a-5p on ZMYND11 expression in HGC-27 cells was explored. We discovered that miR-10a-5p mimic transfection reduced ZMYND11 while miR-10a-5p inhibitor transfection caused a reverse expression trend of ZMYND11 in HGC-27 cells (Figure 2e, f). Furthermore, Pearson correlation analysis figured out that miR-10a-5p was negatively correlated with ZMYND11 expression in GC tissues (Figure 2g). Our data validated that miR-10a-5p negatively regulated ZMYND11 expression.

ZMYND11 is down-regulated in GC; Overexpression of ZMYND11 promotes angiogenesis

ZMYND11 is lowly expressed in GBM [17] and breast cancer [30]. To study the effect of ZMYND11 in GC, we transfected oe-ZMYND11 with HUVECs and detected an increase in ZMYND11 expression by RT-qPCR and Western blot (Figure 3a, b). Afterward, we determined the role of ZMYND11 on the biological functions of HUVECs. It was revealed that oe-ZMYND11 transfection lessened the viability, migratory, invasive, tube-forming phenotypes and protein expression of VEGF and MMP-9 of HUVECs (Figure 3c-i). Thus, a conclusion was collected that overexpression of ZMYND11 suppressed angiogenesis.

Figure 3.

Overexpression of ZMYND11 promotes angiogenesis. A-B. RT-qPCR (a) and Western blot (b) detection of ZMYND11 expression in HUVECs; C. MTT detection of cell viability; D. CCK-8 test detection of cell proliferation; E. Colony formation analysis of cell viability; F. Wound healing test analysis of cell migration; G. Transwell analysis of cell invasion; H. Tube formation test analysis of tube-forming ability; I. Western blot analysis of VEGF and MMP-9 expression. * P < 0.05 vs. the oe-NC group; the measurement data were expressed as the mean ± standard deviation, and the comparison was analyzed by the t-test.

miR-10a-5p is packaged into HGC-27 exo

miR-10a-5p is overexpressed in plasma EVs from prostate cancer patients [31]. Herein, we speculated that miR-10a-5p may be involved in the tube formation through delivery by GC cells-derived exo. Exo extracted from GES-1 and HGC-27 cells were identified: TEM observation displayed that (Figure 4a) exo had a spherical structure formed by a lipid bilayer molecular membrane; NTA (Figure 4b) suggested that the diameter of exo was about 100 nm; Western blot analysis presented that CD63, Alix and TSG101 were highly expressed in HGC-27 exo (Figure 4c), confirming the successful extraction of exo.

Figure 4.

miR-10a-5p is packaged into HGC-27 exo. A. TEM observation of exo; B. Nanoparticle size analysis of the diameter of exo; C. Western blot analysis of exo surface markers (CD63, Alix and TSG101); D. RT-qPCR detection of miR-10a-5p level in GSE-1 exo and HGC-27 exo; E. RT-qPCR detection of ZMYND11 level in GSE-1 exo and HGC-27 exo; F. Immunofluorescence staining of exo; * P < 0.05 vs. the GES-1-exo group; the measurement data were expressed as the mean ± standard deviation, and the comparison was analyzed by the t-test.

Then, we detected the increased expression of miR-10a-5p and decreased expression of ZMYND11 in HGC-27 exo (Figure 4d, e). Additionally, PHK26-labeled miR-10a-5p was transfected into HGC-27 cells, and then HUVECs were co-cultured with HGC-27 exo. Immunofluorescence results implied that the fluorescence intensity of PHK26 in HUVECs was dominant (figure 4f).

GC cells-derived exo enhance the viability and migration of HUVECs by transferring miR-10a-5p

To evaluate the role of exo-miR-10a-5p in HUVECs, we transfected miR-10a-5p mimic into HGC-27 cells, then HGC-27 exo were extracted and incubated with HUVECs. After incubation, we examined that miR-10a-5p expression increased in HUVECs (Figure 5a), further proving that miR-10a-5p can be delivered from HGC-27 exo to HUVECs.

Figure 5.

GC cells-derived exo enhances the viability and migration of HUVECs by transferring miR-10a-5p. A. RT-qPCR detection of miR-10a-5p expression; B. MTT detection of cell viability; C. CCK-8 test detection of cell proliferation; D. Colony formation analysis of cell viability; E. Wound healing test analysis of cell migration; F. Transwell analysis of cell invasion; G. Tube formation test analysis of tube-forming ability; H. Western blot analysis of VEGF and MMP-9 expression. * P < 0.05 vs. the exo-mimic NC group; # P < 0.05 vs. the PBS group; the measurement data were expressed as the mean ± standard deviation, and the comparison was analyzed by the t-test.

Functionally, HGC-27 exo enhanced viability (Figure 5b-d), migration (Figure 5e), invasion (figure 5f) and tube formation ability (Figure 5g), as well as elevated VEGF and MMP-9 protein expression (Figure 5h) of HUVECs. Shortly, GC cells-derived exo enhanced viability and migration of HUVECs by transferring miR-10a-5p.

Up-regulation of ZMYND11 reverses the role of miR-10a-5p-modified GC cells-derived exo in HUVECs

We further investigated whether miR-10a-5p-treated HGC-27 exo affects HUVECs through ZMYND11. We observed that oe-ZMYND11 transfection reversed the exosomal miR-10a-5p-mediated reduction of ZMYND11 mRNA expression (Figure 6a), as well as promotion of cell viability (Figure 6b-d), migration (Figure 6e), invasion (figure 6f), tube formation ability (Figure 6g) and protein expression of VEGF and MMP-9 (Figure 6h). In summary, miR-10a-5p-modified GC cells-derived exo modulated the biological activities of HUVECs by regulating ZMYND11 expression.

Figure 6.

Up-regulation of ZMYND11 reverses the role of miR-10a-5p-modified GC cells-derived exo in HUVECs. A. RT-qPCR detection of ZMYND11 expression; B. MTT detection of cell viability; C. CCK-8 test detection of cell proliferation; D. Colony formation analysis of cell viability; E. Wound healing test analysis of cell migration; F. Transwell analysis of cell invasion; G. Tube formation test analysis of tube-forming ability; H. Western blot analysis of VEGF and MMP-9 expression.* P < 0.05 vs. the exo-mimic NC group; # P < 0.05 vs. the exo-mi-miR-10a-5p + oe-NC group. the measurement data were expressed as the mean ± standard deviation, and the comparison was analyzed by one-way ANOVA.

Discussion

GC is a multifactorial disease, ranking the 4th cause of cancer-related deaths [32]. Our research mainly investigated the action of miR-10a-5p from HGC-27 exo in the biological functions of HUVECs. Experimentally, we determined that miR-10a-5p was up-regulated in GC and silencing of miR-10a-5p impaired the viability, migration, invasion and tube-forming abilities of HUVECs. Next, we validated a miR-targeted relation between miR-10a-5p and ZMYND11 and further discovered that ZMYND11 was down-regulated in GC and restoration of ZMYND11 lessened the activities of HUVECs. Afterward, we further delved out miR-10a-5p could be delivered into HUVECs via HGC-27 exo, thus enhancing angiogenesis. Finally, we identified that up-regulation of ZMYND11 reversed the role of miR-10a-5p-modified HGC-27 exo in HUVECs. In short, miR-10a-5p from GC cells-derived exo augmented angiogenesis by targeting ZMYND11.

In regard to the functions of GC cells-derived exo, Guang Deng et al. have discussed that through disrupting the mesothelial barrier and peritoneal fibrosis, exo from GC cells have supportive effects on peritoneal metastasis [9]. Furthermore, Xiang Xia et al. have investigated a molecular mechanism by which hypoxic GC cells-secreted exo augment the proliferative and metastatic behaviors of GC cells through the delivery of miR-301a-3p [33]. Additionally, GC cells-produced exo could also transmit miR-130a into VECs, thus activating angiogenesis and tumorigenic development [12]. Intriguingly, a paper established by WeiHong Ren et al. has depicted a pro-tumor role of GC cells-released exo, which could augment expansion of myeloid-derived suppressor cells through a transfer of miR-107 [34]. GC cells-derived exo activates HUVECs to proliferate, migrate and form tubes, and such effects are enhanced via transfer of 26-nt-long ncRNA [35].

Not limited to the present study, other papers have identified the integral feedback of tumor-derived-exo and miR in cancers. Exampled by an article composed by Fanyang Kong et al., CAF-released EV could support the malignant behaviors of pancreatic cancer cells [14]. A novel research has unraveled that miR-10a-5p from CAF-EV increases angiogenesis and tumorigenicity in CSCC while down-regulating miR-10a-5p in CAF-EV exerts oppositely [13]. miR-10a-5p overexpression is accordingly examined in the plasma of patients with ovarian cancer [36], papillary thyroid carcinoma [37] and non-small cell lung cancer [38], having a significant value in the disease diagnosis. Focusing on the action of miR-10a-5p, Yaoyong Lu et al. have validated that miR-10a-5p expression suggests an increment in GC which augments cell migration [39]. Consistently, X Fei et al. have profiled that in pancreatic cancer, miR-10a-5p expression goes to raise which elicits promotion of cell proliferation and invasion [40]. On the other hand, Guangbing Xiong et al. have also investigated an increase in miR-10a-5p expression in pancreatic ductal adenocarcinoma cells which are resistant to gemcitabine, and high miR-10a-5p level encourages cell migration and invasion [29].

As to ZMYND11, it has been reviewed to differentially act depending on the cancer type [41]. Importantly, ZMYND11 could interact with miRs in the progression of cancer. Notably, a reduction of ZMYND11 expression is tested in GBM and restoration of ZMYND11 could impair miR-196a-5p-mediated support to proliferation and invasion [17]. Mechanistically, Zhang Z et al. have verified that ZMYND11 is down-regulated in ovarian cancer, and overexpression of ZMYND11 could reduce miR-196b-mediated pro-tumor effects [30]. Besides, in the course of glioma, elevating ZMYND11 expression facilitates the anti-tumor role of the Hsa_circ_0008225/miR-890 axis [42].

Conclusion

To summarize, our research for the first time studied the role of exosomal miR-10a-5p/ZMYND11 axis for HUVECs and identifies a mechanism by which GC cells-derived exosomal miR-10a-5p activates angiogenesis through repressing ZMYND11 expression. Since our findings are collected on a relevant small scale, many efforts are asked to further investigate the detailed mechanism of this signal axis.

Click here for additional data file.