The RPMI-8226 human multiple myeloma cell line and human umbilical vein endothelial cells (HUVECs) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). The cells were cultured in medium containing 10% fetal bovine serum (FBS, Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) and 1% penicillin-streptomycin (Solarbio, Beijing, China) at 37°C with 5% CO₂.

MVs were isolated by continuous differential centrifugation as previously described (24) from the different conditioned media of myeloma cells. Briefly, MM cells growing in log phase were seeded at 5×10⁵ cells/ml and were randomly divided into 5 groups depending on treatment: control group (PBS), 10 or 100 nM bortezomib (Selleck Chemicals, Houston, TX, USA), 1 or 10 µM lenalidomide (MedChem Express, Monmouth Junction, NJ, USA). MVs derived from corresponding RPMI-8226 cells from the 5 groups above were named: N-MV (PBS), B10-MV (10 nM bortezomib), B100-MV (100 nM bortezomib), L1-MV (1 µM lenalidomide), and L10-MV (10 µM lenalidomide). After 24 h of culture, the conditioned media were centrifuged at 750 × g for 5 min and then at 1500 × g for 20 min to remove cells and debris. After further centrifugation at 16000 × g for 45 min at 4°C, the MVs were pelleted. After washing twice with PBS, the pelleted MVs were resuspended in PBS, stored at 4°C and used for experiments within a few days.

The number of MVs was determined by flow cytometry (CytoFLEX, Beckman Coulter, Inc., Brea, CA, USA) following the manufacturer's protocols. Standard microbeads (1.1 µm, Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) were used for size calibration, and the number of MVs were calculated using 3-µm standard microbeads (Sigma-Aldrich; Merck KGaA) as previously described (24). Briefly, 1.1-µm standard microbeads were used to set the gate, 3-µm standard microbeads of an equal concentration were added to the sample as an internal standard, and the counting endpoint was set at 100,000 events for the 3-µm standard bead population.

To evaluate the expression levels of pro-angiogenic factors, using qPCR, we analyzed the mRNA expression levels of VEGF, IL-6 and basic fibroblast growth factor (bFGF) in different groups of MM MVs and HUVECs, and by ELISA, we analyzed the protein expression levels of these factors in the conditioned medium of HUVECs. Additionally, we measured the mRNA expression of NF-κB in HUVECs. MM MVs from the 5 different groups were isolated and subjected to RNA preparation using TRIzol (Invitrogen; Thermo Fisher Scientific, Inc.) and to cDNA synthesis with a cDNA Synthesis kit (Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. The total RNA of HUVECs (5×10⁴ cells/ml) co-cultured for 24 h with MM MVs (10 µg/ml) from the 5 different groups above or without MVs was isolated, and cDNA was synthesized. Each qPCR reaction was prepared in triplicate. PCR was performed using SuperReal PreMix Plus (SYBR Green) (Tiangen, Beijing, China) with a real-time PCR system (Rotor-Gene Q, Qiagen, Hilden, Germany). The following primer sequences (Invitrogen; Thermo Fisher Scientific, Inc.) were used: VEGF, 5′-AATGCTTTCTCCGCTCTGAA-3′(F) and 5′-GCTTCCTACAGCACAGCAGA-3′ (R) (96 bp); IL-6, 5′-ATGAGGAGACTTGCCTGGTGAAAAT-3′ (F) and 5′-TCTGGCTTGTTCCTCACTACT-3′ (R) (104 bp); bFGF, 5′-TTCATAGCAAGGTACCGGTTG-3′ (F) and 5′-AGAAGAGCGACCCACACG-3′ (R) (97 bp); NF-κB, 5′-GGGGACTACGACCTGAATG-3′ (F) and 5′-GGGCACGATTGTCAAAGAT-3′ (R) (118 bp); and β-actin, 5′-GAGCTACGAGCTGCCTGAC-3′ (F) and 5′-GGTAGTTTCGTGGATGCCACAG-3′ (R) (121 bp). The qPCR conditions were 2 min at 50°C, 2 min at 95°C and then 40 cycles of 15 sec at 95°C and 30 sec at 60°C. A comparative ∆Ct method was used to determine gene expression and β-actin served as the normalization gene.

The expression levels of human VEGF, IL-6 and bFGF in the supernatants of HUVECs were measured by ELISA kits (NeoBioscience, Shenzhen, China) according to the manufacturer's instructions. Experiments were performed in triplicate.

Cell migration was determined using a scratch adhesion test. Briefly, the outside undersurface of a 6-well plate was scored with a marker pen; HUVECs (5×10⁵ cells/well) were seeded onto a 6-well plate and cultured overnight to confluence. Then, at time zero, the cell monolayer was wounded artificially across the marked line using a 200-µl pipette tip. The scratched area was captured using an inverted microscope (Axio Observer D1, Zeiss, Germany) in low-power fields (×40). Then, the cells were treated with medium (DMEM without FBS) and the 5 groups of MVs (10 µg/ml) for 24 h. Negative control cells were treated with PBS. The wounded area was captured again at 24 h and measured by ImageJ software (National Institutes of Health, Bethesda, MD, USA). The percentage of recovery was assessed. The experiment was repeated in triplicate, and each sample was tested in quadruplicate.

Endothelial cell invasiveness was investigated using Transwell Boyden chambers (6.5 mm, Costar, USA). The polycarbonate membrane (8-µm pore size) was coated with a layer of Matrigel matrix (Corning Inc., Corning, NY, USA). HUVECs (2×10⁵ cells per well) were added to the upper chamber and starved; 750 µl of complete medium with the indicated group of MVs was added at the indicated concentration (50 µg/ml) to the lower compartment of the chamber as a chemoattractant. PBS was used as the negative control. After incubation at 37°C for 6 h, the cells on the bottom surface of the membrane were fixed with methanol, stained with Diff-Quik (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) and counted in five representative high-power fields (×400).

Matrigel (50 µl, Corning Inc.) was added to each well of a 96-well plate and allowed to polymerize at 37°C for 30 min. Then, HUVECs (2×10⁴ cells per well) were cultured in DMEM with 6% FBS in the presence of the 5 groups of MVs (10 µg/ml). PBS was used as the negative control. Each treatment was repeated in three wells. After incubation for 6 and 24 h, images of the tubular structure that formed in each well were captured with an inverted microscope (Axio Observer D1, Zeiss, Germany) with an attached digital camera (Axiocam 506 color, Zeiss, Germany). The closed tube formation was quantified in five low-power fields (×40).

HUVECs were seeded on climbing slices in 24-well plates and co-cultured with the 5 groups of MM MVs or PBS for 24 h. To measure the effect of NF-κB activation, an NF-κB Activation Nuclear Translocation assay kit (Beyotime Biotech, Haimen, China) was used according to the manufacturer's protocol and a previous study (25). Nuclei (blue) and NF-κB (red) were viewed with a fluorescence-equipped microscope (BX51, Olympus, Tokyo, Japan). Co-localization of NF-κB with the nucleus appeared purple.

After co-culture with one of the 5 different MVs at 10 µg/ml or with PBS for 24 h, nuclear protein or total protein were extracted from HUVECs using a Nuclear Protein Extraction kit (BestBio, Shanghai, China) or a Total Protein Extraction kit (BestBio), respectively, and then quantified by the BCA method. Equal amounts (50 µg) of protein extracts were loaded and separated by SDS-PAGE and then transferred electrophoretically to a polyvinylidene difluoride (PVDF) membrane. Nonspecific sites were blocked by incubation with 5% nonfat dry milk in TBST at room temperature for 1 h. The membranes were incubated with the indicated primary antibodies [P65 (1:50000, 7R200963-9, Abcam, Cambridge, MA, USA, rabbit monoclonal), p-p65 (1:1000, 3033S, Cell Signaling Technology, Inc., Danvers, MA, USA; rabbit monoclonal), H3 (1:2000, KM9005T, Tianjin Sungene Biotech, Tianjin, China; mouse monoclonal), and β-actin (1:1000, sc-130657, Santa Cruz Biotechnology, Santa Cruz, CA, USA; rabbit polyclonal)] in blocking buffer overnight at 4°C. Anti-rabbit or mouse IgG (1:10000, 21032, Rockland, Limerick, PA, USA) was used as a secondary antibody and incubated with the membrane for 2 h at room temperature. After washing, reactive bands were identified by scanning infrared laser microscopy (LI-COR Biosciences, Lincoln, NE, USA).

All data are expressed as the means ± SEM of at least three experiments. Differences among groups were analyzed with one-way ANOVA, and differences were considered statistically significant at P<0.05.

MM cells had a high rate of MV shedding; significantly more MVs were observed in the bortezomib groups (B10-MV and B100-MV) than in the control group, but the number of MVs was not significantly altered by lenalidomide treatment (Fig. 1).

MM cells in the log phase of growth were seeded at 5×10⁵ cells/ml and were randomly divided into 5 groups according to treatment: PBS (control), 10 nM bortezomib, 100 nM bortezomib, 1 µM lenalidomide, or 10 µM lenalidomide. After 24 h, MVs were isolated from the various conditioned media, and RNA was extracted from MVs. The PCR results showed that both bortezomib and lenalidomide treatment significantly decreased the gene expression of VEGF, IL-6 and bFGF in MVs compared with those treated with PBS, but no difference was observed between the bortezomib and lenalidomide groups (Fig. 2A).

HUVECs were randomly divided into 6 groups and co-cultured with the different MM MVs (as above) or with PBS (control group) for 24 h. qPCR and ELISA assays were used to examine the expression levels of VEGF, IL-6 and bFGF in HUVECs and in the conditioned medium, respectively. Both qPCR (Fig. 2B) and ELISA (Fig. 2C) results showed that these cytokines were significantly higher in the normal MV-treated HUVECs than in the PBS group. Moreover, VEGF, IL-6 and bFGF expression levels in all drug-treated-MV groups were significantly decreased compared with those in the normal-MV group (Fig. 2B and C).

The angiogenic properties of MM MVs were evaluated as their ability to induce HUVEC migration, invasion and tubulogenesis. Cell migration was tested using a wound healing assay; after treatment with N-MVs for 24 h, HUVEC migration towards the scratched area was significantly improved. As shown in Fig. 3A, N-MVs (10 µg/ml) induced 55.71% (P<0.05) of wound recovery compared with 44.95% in the PBS group. Compared with N-MVs, the DMVs decreased cell motility, although the B100-MV and L1-MV groups were not statistically significantly different (Fig. 3A). Next, the invasive ability of HUVECs was demonstrated by the Transwell invasion system. In this experiment, different MM MVs were used as chemoattractants. N-MVs promoted cell invasion through Matrigel, but cell invasion was significantly inhibited by the 4 groups of DMVs compared with that of the N-MV group (Fig. 3B). Finally, the angiogenic activity of MVs was investigated by in vitro tube formation assay. Tubulogenesis was qualified by counting the number of closed tubes observed at 6 and 24 h. At 6 h, no significant effects were observed in the N-MVs group compared with the medium control, but all DMV groups significantly suppressed the tube formation activity of HUVECs compared with that of the N-MV group (Fig. 3C). At 24 h, the results showed that N-MVs significantly promoted the tube formation of HUVECs, but tube formation activity was significantly suppressed by the presence of DMVs (Fig. 3C).

The expression of the NF-κB p65 subunit was determined using qPCR. The expression of NF-κB p65 was increased in the N-MV group compared with the control group and was significantly decreased in all DMVs groups compared with the N-MV group (Fig. 4A). Then, the expression levels of p65 in the total protein and p-p65 in the nuclear protein of HUVECs were analyzed by western blot analysis, and the ratio of p-p65 to p65 was calculated. As shown in Fig. 4B, the protein expression level and the mean optical density of p-p65 were higher in the N-MV group than in the control group. However, all DMV treatments caused a reduction in the protein expression levels and the mean densities of p-p65 in nuclear proteins compared with those in the N-MV-treated group. However, p65 exhibited no significant differences among the MV groups. As shown in Fig. 4C, the same tendency was found in the immunofluorescence microscopy assay; p65 translocation was greatly increased in the N-MV group and reduced in the DMV groups. Our data suggested that N-MVs significantly increased NF-κB activation in HUVECs, whereas it was decreased by DMVs. These data supported the regulation of NF-κB activity in target HUVECs by different MVs.