Characterization of macrophage-derived apoVs and apoV uptake by MSCs.
STS was used to induce apoptosis of macrophages. Apoptotic and normal macrophages were observed under fluorescence microscope. A larger number of TUNEL positive stained cells (red) were observed in the STS group, while control group had seldom stained cells (Fig. 1A). Apoptotic macrophages could generate lots of apoVs, the output of apoVs was much higher than exosomes (Fig.S1). TEM showed that apoVs had a cup-shaped morphology and the diameter was about 200 nm (Fig. 1B). Nanoparticle tracking analysis showed that the diameter distribution of apoVs was 240.6 ± 115 nm (Fig. 1C). To investigate whether apoVs could be ingested by MSCs, MSCs were cultured with PKH-26-labeled apoVs (red) for 4h and 8h, respectively. The nuclei of MSCs were stained with DAPI (blue). The F-actin of MSCs was stained with phalloidin (green). Confocal laser microscopy showed that red stained particles appeared around the MSCs nucleus after 4 hours, and the number of red stained particles increased after 8 hours (Fig. 1D).
Macrophage-derived apoVs promoted adipogenesis of MSCs in vitro.
We set up different concentration gradients (0.5, 1, 2, 4 µg / mL) to clarify the optimal concentration of apoVs, and cultured MSCs in OM and AM respectively for 7 days. In AM culture, 2 µg / ml apoVs significantly increased the PPARγ expression; in OM culture, 2 and 4 µg / ml apoVs inhibited the RUNX2 expression levels more significantly (Fig.S2). Therefore, we used 2 µg / mL apoVs in the subsequent experiments. Next, to further investigate the role of macrophage-derived apoVs in MSC adipogenesis, we treated MSCs under AM with or without apoVs. After 14 days, the cells were stained with oil red O. Fig. 2A showed a great number of red-stained lipid droplets were formed when cells were cultured in AM, while the lipid droplets were significantly increased in MSCs treated with apoVs. In AM, oil red O quantitative analysis (Fig. 2B) of the group under stimulation by apoVs was higher than group without apoVs (P < 0.05). In addition, PPARγ and C/EBPα expression levels of the group treated with apoVs were significantly higher than group without apoVs (P < 0.001) (Fig. 2C). Moreover, the protein expression of PPARγ was up-regulated during adipogenesis in the group treated with apoVs compared with the group without apoVs (Fig. 2D, 2E).
Macrophage-derived apoVs inhibited osteogenesis of MSCs in vitro.
To further explore the effect of macrophage-derived apoVs on osteogenesis of MSCs, we treated MSCs under OM with or without apoVs for 7 days, the cells were examined by ALP staining (Fig. 3A) and ALP quantification (Fig. 3B), the results showed that apoVs significantly inhibited osteogenic differentiation of MSCs. The expression of ALP and RUNX2 was significantly decreased by apoVs (Fig. 3C). After treated MSCs under OM with or without apoVs for 14 days, the ARS staining and quantification showed similar results that apoVs could inhibit osteogenic differentiation of MSCs (Fig. 3D, 3E). Moreover, RUNX2 and BGLAP expression was significantly decreased by treatment with apoVs for 14 days in OM (Fig. 3F). In addition, the protein expression of RUNX2 was down-regulated during osteogenesis by apoVs (Fig. 3G, 3H).
Macrophage-derived apoVs promoted adipogenesis of MSCs in vivo.
To examine the role of macrophage-derived apoVs in MSC adipogenic differentiation in vivo, we combined MSCs (PM, AM, and AM + apoVs) with collagen sponges and implanted them into nude mice. H&E staining showed that the AM + apoVs group had more adipose tissue-like structures than the AM group, whereas the PM group showed a large amount of collagen membrane scaffold and no adipose tissue-like structures (Fig. 4A). Oil red O staining showed that the AM + apoVs group had more red-stained adipose tissue-like structures than the AM group (Fig. 4B). Therefore, macrophage-derived apoVs could promote adipogenesis of MSCs in vivo.
Macrophage-derived apoVs inhibited osteogenesis of MSCs in vivo.
In order to determine the role of macrophage-derived apoVs in MSC osteogenesis in vivo, we mixed groups (PM and PM + apoVs) with β-TCP and cell-free β-TCP, then implanted them into nude mice. H&E staining (Fig. 5A) showed that the PM group had more new, strongly eosinophilic tissue compared to the β-TCP group, whereas the PM + apoVs group had less bone tissue-like structures than PM group. Masson staining (Fig. 5B) showed more blue-green collagen fibers in the PM group than the β-TCP group, whereas the PM + apoVs group had less blue-green collagen fibers than the PM group. Therefore, macrophage-derived apoVs inhibited osteogenesis of MSCs in vivo.
MiR155 promoted adipogenesis of MSCs cultured with macrophage-derived apoVs.
Several studies have shown the contents of EVs included mRNAs, miRNAs, ncRNAs, protein and lipids. MiRNAs account for about half of the total RNAs of EVs, and play a key role in the transfer of biomolecules to recipient cells and cell-to-cell communication(32–34). The content of apoVs was higher than that of exosomes, and the content of miRNA in apoptotic vesicles was much higher than that of exosomes (Fig.S3). The change of miR155 in apoVs was highly significant, indicating that miR155 was highly enriched in apoVs (Fig. 6A). Next, we transfected macrophages with inhibitor-negative control (inhi-NC), inhibitor-miR155 (inhi-miR155), mimics-negative control (miR-NC) and mimics-miR155 (miR155), the expression levels of miR155 could be significantly decreased by (inhi-miR155) or increased by (miR155) in macrophages, and miR155 expression levels decreased or increased more significantly in corresponding apoVs: apoVs (inhi-NC), apoVs (inhi-miR155), apoVs (miR-NC) and apoVs (miR155), respectively (Fig. 6B). Subsequently, apoVs (inhi-NC), apoVs (inhi-miR155), apoVs (miR-NC) and apoVs (miR155) were added to MSCs for adipogenic induction. The results showed that the adipogenesis of MSCs was decreased in apoVs (inhi-miR155) group compared to the apoVs (inhi-NC) group. In addition, the adipogenesis ability of MSCs was improved in the apoVs (miR155) group compared to the apoVs (miR-NC) group (Fig. 6C, 6D). Collectively, these results showed that miR155 promoted adipogenic differentiation of MSCs cultured with macrophage-derived apoVs.
MiR155 inhibited osteogenesis of MSCs cultured with macrophage-derived apoVs.
In order to further clarify the role of miR155 in macrophage-derived apoVs in regulating the osteogenesis of MSCs, we cultured MSCs with apoVs (inhi-NC), apoVs (inhi-miR155), apoVs (miR-NC) and apoVs (miR155) in OM. The ALP staining and quantification showed that the osteogenesis of MSCs was up-regulated in the apoVs (inhi-miR155) group compared with apoVs (inhi-NC) group, while the osteogenesis of MSCs was decreased in apoVs (miR155) group than the apoVs (miR-NC) group (Fig. 7A, 7B). ARS staining and quantification confirmed these results (Fig. 7C, 7D). Therefore, our results suggested that miR155 could regulate osteogenesis and adipogenesis of MSCs in the presence of macrophage-derived apoVs.
MiR155 regulated adipogenic and osteogenic differentiation of MSCs cultured with macrophage-derived apoVs via the SMAD2 signaling pathway.
SMAD2 was a target gene of miR155(35, 36), and enhancement of the SMAD2 signaling pathway increased osteogenic differentiation and inhibited adipogenic differentiation(37–39). We cultured MSCs with apoVs (inhi-NC), apoVs (inhi-miR155), apoVs (miR-NC) and apoVs (miR155) in AM and OM, then used western blot analysis to clarify the changes of SMAD2 pathway proteins (Fig. 8A, 8B). In the apoVs (inhi-miR155) group, the SMAD2 signaling pathway was up-regulated compared to the apoVs (inhi-NC) group. Moreover, the SMAD2 signaling pathway was down-regulated in the apoVs (miR155) group compared to the apoVs (miR-NC) group. Our data suggested that miR155 regulated adipogenic and osteogenic differentiation of MSCs cultured with macrophage-derived apoVs via the SMAD2 signaling pathway.