The mouse breast cancer cell line 4T1 was purchased from ATCC and cultured in RPMI 1640 medium supplemented with 10% FBS (cat. no. SV30087.02; HyClone; Cytiva), 1% penicillin-streptomycin solution (cat. no. 15140122; Gibco; Thermo Fisher Scientific, Inc.) and 1% nonessential amino acid solution (cat. no. 11140050; Gibco; Thermo Fisher Scientific, Inc.). The human breast cancer cell line MCF-7 was purchased from ATCC and cultured in Dulbecco's modified Eagle's medium (DMEM; cat. no. 11995065; Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS and 1% penicillin-streptomycin solution (cat. no. 15140122; Gibco; Thermo Fisher Scientific, Inc.). HUVECs (passage 3–4), human placenta-derived MSCs (hPMSCs; passage 4–6) and human umbilical cord-derived MSCs (hUC-MSCs; passage 4–6) were donated by AmCellGene Co. Ltd. and preserved at Nankai University (18). For the tracking of transplanted tumor cells in vivo, 4T1 cells were transduced with a self-activating lentiviral vector that carried a 5′LTR promoter-driven double fusion (DF) reporter gene that contained firefly luciferase and enhanced green fluorescence protein (Fluc-eGFP) as previously reported (19,20).

EVs were isolated by ultracentrifugation as previously described (11,18). The supernatant of MSCs was successively centrifuged at 500 × g for 10 min, 2,000 × g for 30 min and 10,000 × g for 30 min at 4°C to exclude cell debris and apoptotic bodies. After that, the EV pellet was collected by ultracentrifugation at 100,000 × g for 70 min and washed by a second ultracentrifugation at 100,000 × g for 2 h in phosphate-buffered saline (PBS). All the above ultracentrifugation steps were performed at 4°C. Finally, the purified EV pellet was resuspended in 200 µl PBS and stored at −80°C for further experiments. Transmission electron microscopy (TEM; HT7700; Hitachi, Ltd.) was used to observe the morphology of the isolated EVs (18). Briefly, 2 µg/µl EVs were loaded onto a carbon film (Zhongjingkeyi Technology), incubated for 5 min at room temperature and was subsequently stained with 2% phosphotungstic acid for 2 min at room temperature. Excess liquid was removed by filter paper. The samples were allowed to air dry for at least 10 min at room temperature and then observed using TEM at an acceleration voltage of 120 kV. The size distribution of EVs was determined using a Malvern Particle Size Analyzer (Zeta sizer Nano ZS; Malvern Instruments, Ltd.). The protein concentration was quantified using a BCA protein assay kit (Promega Corporation).

In the present study, the proteins expressed in MSCs and MSC-EVs, including tumor susceptibility gene 101 (TSG101), ALG-2 interacting protein X (ALIX) and CD63, were identified as previously reported (21). In brief, 30 µg of EVs lysates in 100 µl radioimmunoprecipitation assay (RIPA) buffer (cat. no. R0010; Beijing Solarbio Science & Technology Co., Ltd.), determined by BCA protein assay kit (Promega Corporation), were subjected to western blotting analysis. Proteins were heat-denaturized in 5X SDS-PAGE loading buffer (cat. no. CW0027S; CWBIO), fractionated on 10% SDS-polyacrylamide gels, and electro-transferred onto polyvinylidene fluoride membranes (MilliporeSigma). Blocked with 5% non-fat milk (cat. no. A600669; Beyotime Institute of Biotechnology) for 2 h at room temperature, proteins were respectively immunoblotted with the indicated antibodies. Briefly, they were incubated with primary antibodies at 4°C overnight and subsequently identified by second antibodies through a second incubation step of 2 h at room temperature. The following antibodies were used according to the manufacturer's instructions: ALIX (1:2,000; cat. no. WL03063; Wanleibio Co., Ltd.), TSG101 (1:2,000; cat. no. EPR7131; Abcam) and CD63 (1:2,000; cat. no. WL02549; Wanleibio Co., Ltd.). The secondary antibody applied in this study was HRP-labeled Goat Anti-Rabbit IgG (H+L; 1:1,000; cat. no. A0208; Beyotime Institute of Biotechnology). The blot bands were visualized with enhanced chemiluminescence detection reagent (cat. no. WP20005; Thermo Fisher Scientific, Inc.) and analyzed by ImageJ software (version 1.54j; National Institutes of Health).

To determine whether hPMSC-EVs could induce apoptosis in 4T1 cells, Annexin V/propidium iodide (PI) fluorescence-activated cell sorting (FACS) analysis was performed with a FACS instrument (FACSCalibur; BD Biosciences). Single-cell suspension of 4T1 cells was harvested after 24 h of EV treatment and stained using an Annexin V/PI kit (cat. no. CA1040; Beijing Solarbio Science & Technology Co., Ltd.) to assess the apoptotic proportion of 4T1 cells. According to the manufacturer's protocol, 100 µl of 4T1 cell suspension at a concentration of 1–5×10⁶/ml was firstly stained with 5 µl Annexin V for 5 min at room temperature. Then, 5 µl PI (20 µg/ml) and 400 µl phosphate-buffered saline (PBS) were added and detected immediately. The PBS-treated 4T1 cells were used as control, and the Annexin V⁻/PI⁻ cells were identified as live cells. The Annexin V⁺/PI⁻ cells were identified as early apoptotic cells, Annexin V⁺/PI⁺ cells were identified as late apoptotic cells, and Annexin V⁻/PI⁺ cells were scored as necrotic cells. Data were analyzed by FlowJo 10 (FlowJo LLC). To investigate whether hPMSC-EVs affect cell proliferation, 4T1 cells were treated with hPMSC-EVs and evaluated by bioluminescence imaging.

For the collection of 4T1 CM, 4T1 cells were primitively cultured in 1640 medium. When the confluence reached 40%, the old medium was replaced with a fresh complete medium supplemented with 5 mg EVs. After 48 h, the supernatants were harvested and stored at −80°C. For the collection of HUVEC-CM, which served as a control in the present study, the complete HUVEC medium was replaced with 7 ml of EGM-2 per T25 flask (Corning, Inc.), with HUVECs also reaching 40% confluence. After 48 h, the supernatants were collected and stored at −80°C.

Transwell assays were performed using Transwell chambers through a filter membrane with the pore size of 8 µm (MilliporeSigma). Briefly, 1×10⁵ Fluc-GFP-labeled 4T1 cells immersed in 200 µl of RPMI 1640 medium supplemented with 2% FBS were seeded in the upper inserts. RPMI 1640 supplemented with 10% FBS was added to the lower chamber. The low concentration of FBS in the upper inserts prevents cell proliferation, consistent with previous reports (22). After 24 h, non-migrated cells on the upper side of the filter were removed with a cotton swab. The migrated cells were counted under a confocal microscope at a magnification of 200× and five randomly chosen fields were selected for each insert. The experiments were repeated at least three times. For the wound healing assay, cells were seeded in 6-well plates at a density of 2×10⁵ per well and grown to 90% confluence. Sterile 10 µl tips (Corning Life Sciences) were used to create 2–3 straight wounds in each well, followed by medium replacement and treatment with 100 µg/ml EV. In this assay, similar to the Transwell assay, 10% FBS from RPMI 1640 was replaced with 2% low-concentration FBS to suppress 4T1 cell proliferation. The images were captured and reported temporally, with an interval of 12 h. The number of cells that migrated to the wound area was calculated using ImageJ software (Version 1.54j; National Institutes of Health).

HUVECs treated with hPMSC-EVs were subjected to a tube formation assay. In brief, precooled Matrigel (BD Biosciences) was pipetted into a 48-well plate, which was maintained at 4°C overnight. Before the HUVECs were seeded in Matrigel-treated 48-well plates, they were placed in an incubator and rewarmed to 37°C. Afterwards, the HUVECs were seeded into 48-well plates at a density of 5×10⁴ cells/well, which were randomly divided into two groups and cultured with EGM-2 medium. In particular, CM from 4T1 cells treated with hPMSC-EVs was added to the medium of HUVECs in the EV group. After culturing for 6 h, tube formation was observed under a microscope (ECLIPSE Ti-U; Nikon Corporation); five fields were chosen randomly and analyzed using ImageJ software (Version 1.54j; National Institutes of Health).

4T1 cells were seeded into a six-well plate at 1,000 cells per well and subsequently treated with hPMSC-EVs at a concentration of 100 µg/ml and an equal volume of PBS was also repeatedly added to the control group every other day. After two weeks, the cells were washed with PBS and then stained with crystal violet and the colonies with more than 50 cells were counted and analyzed.

In the present study, 12 BALB/c mice (female, 8–12 weeks old) were purchased from Biocytogen. To monitor the angiogenic effects of EVs in vivo in real time, transgenic mice expressing firefly luciferase under the promoter of VEGFR2 (VEGFR2-luc) were obtained from Xenogen Corporation. All experimental procedures were conducted according to institutional guidelines for the Care and Use of Laboratory Animals at Nankai University, Tianjin, China (23) and in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals, 8th edition (24). The mice were housed in a specific pathogen-free environment, the temperature and humidity of which were maintained at 21–23°C and 45–50%, respectively. With a 12-h light/dark cycle (lights on 07:00-19:00 h), the housing facility supplied the mice with sufficient food and water. The 4T1-derived xenograft model was established according to the following protocol. Mice were anesthetized with isoflurane, 2.0% induction for 4 min, 1.5% maintenance and 1l/min oxygen inhalation (19). For the animal model of the 4T1-derived xenograft model, BALB/c mice were inoculated with 1×10⁶ 4T1 (Fluc-eGFP) cells in the fourth pair of mammary fat pads. Once the tumors were palpable, which could be judged after ~4 days of tumor inoculation (day 0), the mice were randomly divided into two groups (six mice per group).

In the experimental group, 100 µg of MSC-EVs were peri-injected into the tumors on days 0, 2 and 4. In the control group, a peritumoral injection of PBS was administered on days 0, 2 and 4. All mice were observed for tumor progression and treatment effects by BLI for up to 14 days prior to sacrifice and tumor tissues from the two groups were collected for histological analysis. Under the same conditions, the VEGFR2-luc mice were inoculated with unlabeled 4T1 cells (15,25) and BLI was also performed on days 0, 4, 7, 10 and 14. On day 14, mice were sacrificed by cervical dislocation. The tumors were removed for further histological analysis. The humane endpoints (e.g., severe distress, inability to eat or drink, significant weight loss, behavioral abnormalities indicating suffering) were used to determine whether animals should be sacrificed before the end of the study.

BLI was performed in all experimental mice at specific time points using the IVIS Lumina imaging system (Xenogen Corporation). Fluc imaging of DF-labeled 4T1 cells was performed to observe tumor progression and metastasis and assess the therapeutic effects of hPMSC-EVs. For angiogenesis experiments in VEGFR2 transgenic mice, Fluc imaging was used to visualize VEGFR2 expression. Mice were anesthetized with isoflurane, 2.0% induction for 4 min, 1.5% maintenance and 1l/min oxygen inhalation (19). After intraperitoneal injections of reporter probe D-Luciferin (150 mg of luciferin/kg), animals were imaged for 2 sec to 2 min. The BLI signal was quantified by the average radiance of the region of interest over the xenograft tumor using Living Image Software (version 4.7; PerkinElmer, Inc.) (15,25).

Immunofluorescence staining was performed on frozen sections of 4T1 ×enograft tumor samples to detect neovasculature in 4T1 tumor tissue. Sections were incubated with primary antibodies against CD31 (1:200; cat. no. 550274; BD Biosciences), Ki67 (1:200; cat. no. 571599; BD Biosciences) for 2 h at room temperature, followed by incubation with Alexa Fluor 594-labeled secondary antibodies (1:500; cat. no. A21209; Invitrogen; Thermo Fisher Scientific, Inc.) for 30 min at room temperature. The nuclei were stained with DAPI for 5 min at room temperature. Fluorescence images (magnification, ×200) were acquired with a microscope (FV1000; Olympus Corporation) and analyzed with ImageJ software (Version 1.54j; National Institutes of Health).

To assess the mRNA expression levels of the analyzed genes, total RNA was extracted from HUVECs at a confluence of 80–90% and tumor tissues treated with TRIzol^® reagent (cat. no. 15596018CN; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. First-strand cDNA was synthesized using reverse transcriptase-mediated oligo-dT primers (cat. no. 11151ES60; Yeasen Biotech). Subsequently, RT-qPCR was performed on an Opticon System (Bio-Rad Laboratories, Inc.) in 20 µl reaction volumes. The level of mRNA expression was quantified using the TransStart Top Green qPCR SuperMix Kit (cat. no. AQ131-01; TransGen Biotech Co., Ltd.). The 2^−ΔΔCq method was used to determine the relative changes in mRNA folding (26). The primers and amplification conditions are listed in the Tables SI and SII.

All experiments were performed in triplicate for each condition and repeated at least three times. The data are presented as the means ± standard deviations (SDs). For comparisons between two groups, two-tailed paired Student's t-tests were performed. For comparisons among multiple groups, one-way ANOVA followed by the least significant difference post hoc test was used. P<0.05 was considered to indicate a statistically significant difference.

The phenotype of hPMSCs was confirmed by surface marker expression via FACS analysis (Fig. S1). EVs were isolated from the supernatant of hPMSCs by ultracentrifugation and then characterized by TEM, nanoparticle tracking analysis (NTA) and western blotting. The enriched particles exhibited a typical cup-shaped morphology under TEM (Fig. 1A) and relatively heterogeneous sizes of ~120 nm in diameter (Fig. 1B). Western blot analysis confirmed the expression of classical EV markers (TSG101, ALIX and CD63) in hPMSC-EVs (Figs. 1C and S2), which is consistent with a previous report (20).

A recent study reported that MSCs and their EVs can be tumorigenic or antitumorigenic by affecting multiple steps of tumor progression (6). To determine the effects of hPMSC-EVs on breast cancer, 4T1 cells were labeled with an Fluc-GFP double fusion reporter gene (Fig. S3) and apoptosis and proliferation were assessed after treatment with MSC-EVs. Annexin V and PI FACS analysis was performed (Fig. 2A), which revealed that hPMSC-EV treatment did not induce apoptosis in cultured 4T1 cells; however, changes in cell morphology, the elongation of some cells with filamentous ends and a decrease in cell number, were detected via brightfield microscopy, indicating that cell status was adversely affected by EV treatment (Fig. 2B). The morphology of 4T1 cells induced by EVs needs to be clarified to exclude the possibility that it is induced by cell-cell contact. Under identical experimental conditions, control group exhibited rapid 4T1 cell proliferation and increased cell-cell contact. According to the results, the effect of cell-cell contacts during proliferation tended to make the cells more tightly packed. The proportion of cells with elongated, filamentous ends was significantly higher in the EV group compared with the control group. The EV group exhibited significantly lower cell density and confluence compared with the control group, suggesting a diminished role of cell-cell contact in influencing 4T1 cell morphology. In summary, the morphological changes in 4T1 cells may be the result of both cell-cell contact and EV treatment and the role of EVs may be the primary factor involved. Concurrently, the hPMSC-EV-mediated suppression of the proliferation of other breast cancer cells, such as MCF-7 cells, was also validated. Consistent results were revealed using EVs from human umbilical cord MSCs (hUCMSCs), which validated the universality of the present study (Figs. S4 and S5). Furthermore, under BLI, EV treatment led to a significantly reduced Fluc signal, indicating that cell viability was decreased and that cellular proliferation was inhibited (Fig. 2C and D).

Angiogenesis is the basis for solid tumor growth and metastasis. Blocking the formation of new blood vessels in solid tumors has proven to be a highly efficient anticancer strategy. The present study explored the effects of hPMSC-EVs on tumor angiogenesis using HUVECs, a widely used model system for researches related to tumor vascular biology. First, the supernatant of 4T1 cells that had been pretreated with EV for 24 h was collected as CM. Then, a wound healing assay was performed in HUVECs and the results showed that 4T1-CM led to decreased cell motility in HUVECs (Fig. 3A and B). Similarly, the Ki-67 immunofluorescence assay showed that 4T1-CM reduced the number of proliferating endothelial cells (Fig. 3C and D). In addition, Matrigel was used to detect the tube formation of endothelial cells. By counting nodes and junctions, it was found that 4T1-CM treatment markedly reduced HUVEC tube formation (Fig. 3E and F). Furthermore, RT-qPCR revealed that the expression of angiogenesis-related genes (VEGFA, VEGFR2, Ang-1, Ang-2, b-FGF, HIF-1α and PDGF) was significantly downregulated in HUVECs following 4T1-CM treatment (Fig. 3G). These results indicated that hPMSC-EVs can indirectly reduce the neovasculature of 4T1 tumors.

Tumor metastasis occurs when tumor cells migrate from one site to another and stably proliferate. The migration and colony formation abilities of tumor cells indicate a metastatic phenotype, which is closely related to a poor prognosis. The results of the present study revealed that, according to the wound healing assay, 4T1 cell migration was significantly suppressed by EV treatment (Fig. 4A and B). Consistent results were obtained in the Transwell assay, showing a decrease in the migration of GFP-labeled 4T1 cells (Fig. 4C and D). The colony formation assay also revealed decreased colony formation after EV treatment (Fig. 4E and F). These findings suggested that hPMSC-derived EVs may suppress the migration of 4T1 breast cancer cells.

As the aforementioned assays revealed that hPMSCs-EVs could inhibit the growth and migration of 4T1 cells in vitro, the present study next performed in vivo studies for further verification by BLI. A murine model of breast cancer was established by subcutaneously transplanting 10⁶ Fluc/GFP-labeled 4T1 cells into BALB/c mice. BLI showed that the number of tumor cells was reduced and that tumor progression was inhibited by the local administration of hPMSC-derived EVs (Fig. 5A and B). The mice were sacrificed 14 days following hPMSC-EV treatment and the tumors were harvested. Morphologically, the tumors in the EV treatment group were smaller than those in the control group (Fig. 5C). This finding was verified by further statistical analysis of tumor volume and weight (Fig. 5D and E). Furthermore, immunofluorescence staining of the Ki67 protein, a classical tumor cell proliferation marker, revealed a significant decrease in the number of cells in the proliferating state and the quantification analysis also showed an identical trend (Fig. 5F and G). No toxicity of hPMSCs-EVs was observed and all animals reached endpoints of the present study.

VEGFR2-Fluc transgenic mice were subjected to angiogenesis observation via BLI at tumor sites. After hPMSC-EVs were administered, there was a dramatic decrease in VEGFR2 expression (Fig. 6A and B). CD31 immunofluorescence staining was performed after the mice were sacrificed and the tumors were removed, showing that EV treatment led to a reduction in the number of new blood vessels at the tumor sites (Fig. 6C). Then, RNA was extracted from tumor tissues and RT-qPCR revealed that the expression of the angiogenesis-related genes, VEGFA, Ang-1, Ang-2, b-FGF, and PDGF, was significantly downregulated at tumor sites (Fig. 6D). These results indicated that MSC-EVs suppressed angiogenesis in 4T1 breast tumors.