3.1 Characterization of UMSCs and UMSCs-Exo
UMSCs displayed as long spindle-shaped morphology, and expressed with CD29, CD44 and CD73, and without CD31, CD45 and HLA-DR on the membrane by flow cytometry analysis (Fig. 1A). Meanwhile, UMSCs also could differentiate into multiple lineages including osteoblasts (positive staining of Alizarin Red), adipocytes (positive staining of Oil red O) and chondrocytes (positive staining of Alcian Blue) (Fig. 1B). UMSCs-Exo exhibited the characteristic saucer-shaped morphology by TEM (Fig. 1C) and showed the size was distributed about 70–150 nm by NTA (Fig. 1D). Western blot analysis confirmed the positive expression of CD63 and TSG101 (Fig. 1E). All data above indicated UMSCs-derived particles collected in our experiments were exosomes.
3.2 UMSCs-Exo preserved cardiac function and alleviated inflammation after MI
According to the data of echocardiographic parameters, we found with UMSCs-Exo administration, the extent of LV dysfunction was not different between PBS and Exo groups at 3 days after MI. However, after 7 days, especially 28 days, treatment with UMSCs-Exo, cardiac function was better preserved, infarct size and interstitial fibrosis fraction were reduced relative to PBS group (Fig. 2A-C; Supplemental Fig. 1A). Further investigations for cardiac pathology and function showed the levels of VEGFA and VEGFR2, which were closely related to neovascularization, were significantly higher in the Exo group in the border zone at 28 days after MI (Fig. 2D). Meanwhile, the expression of MMP2 and MMP9 in myocardium were also reduced in Exo group (Fig. 2D). The histological assessment of heart tissues also confirmed more angiogenesis (Supplemental Fig. 1B) in the Exo group. These results together demonstrated UMSCs-Exo facilitated cardiac function restore after MI as we previous reported [15].
The prognosis of myocardial infarction is intimately associated with the intensity and duration of inflammatory reaction. The residence of exosomes in tissues after transplantation is the basis of their function. The results of immunofluorescent staining indicated that DiI (red)-labelled UMSCs-Exo were still localized in the myocardium at 3 days after transplantation (Supplemental Fig. 2); HE staining showed there was relatively less inflammatory cells infiltration in the Exo group (Fig. 2E). Further we detected the expression of several injury-related cytokines to investigate the effects of UMSCs-Exo on myocardial repair. The results showed that the protein levels of NFκB-P65 and p-P65 (proinflammatory cytokines) were reduced, while heme oxygenase-1 (HO-1), an anti-inflammatory, anti-apoptosis, anti-oxidative cytokine, was significantly upregulated in the border zone with UMSCs-Exo transplantation (Fig. 2F). These data suggested that UMSCs‐Exo alleviated overactive cardiac inflammation following MI. It is well-known that the activation of NF-κB pathway favors M1 macrophage polarization whereas the upregulation of HO-1 promotes M2 macrophage polarization [16]. Thus, we speculated UMSC-Exo alleviates inflammatory action by regulating macrophage polarization.
3.3 Depletion of macrophages reduced the therapeutic effects of UMSCs-Exo after MI
Considerable evidence revealed macrophages play pivotal roles in tissue repair and regeneration [17]. Our aforementioned results have demonstrated UMSCs-Exo promote myocardial repair, however, these benefits whether through regulating macrophage polarization remains to be confirmed. Cl2MDP was administered to deplete macrophages at 1 day before and post MI respectively as previous described [13, 18]. As expected, macrophages (F4/80+) in hearts and spleens were reduced significantly by flow cytometry and immunofluorescence analysis (Fig. 3A, B). The result of echocardiography showed LV function was slightly improved in Exo + Cl2MDP group compared with PBS + Cl2MDP group, but there was no significant different (Fig. 3C, D). This implied that macrophage depletion weakened the therapeutic benefits of UMSCs-Exo, that indicated macrophages are essential for the cardioprotective effects of exosomes therapy.
3.4 UMSCs-Exo regulated macrophage polarization after MI
Subsequently, we aimed to figure out the role of macrophages in UMSCs-Exo mediated cardiac repair. Among macrophage subpopulations, M2 macrophages could promote injury repair. To further investigate the influence of UMSCs-Exo on macrophage behaviors, we detected the subpopulations of macrophages in the infarcted myocardium. As showed in Fig. 4A and Fig. 4B, the M2 (F4/80+CD206+) phenotype ratio of total macrophages (F4/80+) were increased in Exo group (PBS group: 46.7 ± 0.95%; Exo group: 58.8 ± 2.66%). And the ratio of M1 (F4/80+iNOS+) macrophages were reduced (PBS group: 63.0 ± 3.27%; Exo group: 46.0 ± 2.42%), as showed in Fig. 4A and Fig. 4C. Furthermore, western blotting showed that M2 makers (CD206 and Arg1) were also significantly increased and M1 maker (iNOS) was reduced in Exo group (Fig. 4D), which was consistent with immunofluorescence (Fig. 4E, F). All these data demonstrated that UMSCs-Exo could enhance macrophages polarize towards M2 phenotype after MI.
3.5 UMSCs-Exo promotes macrophages polarize towards M2 phenotype in vitro
As showed in Fig. 5A, the DiI (red)-labelled UMSCs-Exo were internalized into RAW264.7 macrophages within 24 hours. Flow cytometry analysis showed that with UMSCs-Exo pretreated, the ratio of M2 macrophages (CD206+) were significantly increased whereas M1 macrophages (iNOS+) were markedly reduced with LPS inducing (Fig. 5B-D). Further, qPCR results also showed the relative expression of CD206 and Arg1 (M2 phenotype markers) were increased while iNOS and TNFα (M1phenotype markers) were remarkably reduced in Exo + LPS group (Fig. 5E, F). Consistently, the results of western blot analysis revealed a uniform trend (Fig. 5G).
In addition, peritoneal macrophages were isolated from mice to explore the effects of UMSCS-Exo on macrophage polarization (Supplemental Fig. 3A, B). And as expectation, similar results of macrophage polarization were observed (Supplemental Fig. 3C-H). All data above indicated that UMSCs-Exo could promote macrophages to polarize towards M2 phenotype.
3.6 Exosomal RNAs played pivotal roles in macrophage polarization
Exosomes mediate intercellular communication by transferring different biomolecules, including proteins, RNAs, and lipids from one cell to another. To date, several important exosomal protein and nucleic acid cargos responsible for regulating macrophage polarization have been identified, but many remain undiscovered. To identify whether protein or RNA components are responsible for the UMSCs-Exo-induced macrophage polarization, we got exosomal proteins or RNAs as previous reported (Fig. 6A) [14]. The TEM images showed that intact exosomes were ruptured and aggregated following the freeze–thaw operation (Fig. 6B), and RNase A or proteinase would be enforced completely. Then we used exosomal proteins or nucleic acid compenent respectively to treat RAW264.7 macrophages and LPS induced them polarization. Flow cytometry and qPCR analysis indicated that exosomal nucleic acid play dominant roles in promoting macrophages polarize towards M2 phenotype, rather than exosomal proteins (Fig. 6C-E).
3.7 Exosomal miR-24-3p mediate M2 macrophage polarization by suppressing Plcb3 expression and NF-κB signaling activation
MiRNAs are main components of exosomal RNA, regulates the gene expression of recipient cells by pairing and binding to the mRNA and promoting mRNA degradation [19]. To explore the candidate genes and associated pathways for UMSCs-Exo mediated M2 polarization, RNA-sequencing analysis of macrophages was performed. Lots of differentially expressed genes were identified, and further analysis showed 14 genes were closely related to inflammation or macrophage polarization. Combined with our previous study about UMSC-Exo miRNA profile analysis, we speculated exosomes transfer of miR-24-3p to macrophages and targeted to Plcb3, which might be a key target gene concerned with macrophage polarization (Fig. 7A). MiRWalk revealed Plcb3 is a potential target of miR-24-3p (Fig. 7B), therefore, we cloned the Plcb3 binding sequence fragment into a luciferase reporter plasmid. Subsequently, transfected the reporter plasmid, and miR-24-3p mimics into HEK293T cells. Luciferase assays revealed that Plcb3 transcriptional activity decreased markedly in the presence of miR-24-3p mimic (Fig. 7C). Further, we transfected miR-24-3p mimic into RAW264.7 macrophages and western blotting analysis showed the protein level of Plcb3 were significantly reduced (Fig. 7D). The flow cytometry analysis also showed that miR-24-3p mimic obviously facilitated the macrophages to polarize towards M2 phenotype in vitro (Fig. 7E).
According to the loss of function strategy, we reduced miR-24-3p expression in exosomes as we previously reported and treated macrophages (miR-24 inhib Exo group) (Fig. 7F). The results showed the up-regulation of M2 marker (CD206) was induced as well as the down-regulation of M1 marker (iNOS) was partially negated (Fig. 7G).
NF-κB signal pathway is important for the inflammation of macrophages [12, 13, 20], previous findings showed Plcb3 was relevant to inflammatory reaction by activating NF-κB signal pathway [21, 22]. With miR-24-3p mimic transfection, Plcb3, iNOS, P65 and p-P65 were markedly down-regulated while CD206 and Arg1 were significantly up-regulated (Fig. 7H). Further, the result of the Exo group was similar with miR-24-3p group, and the trend was partially reversed after knocking down the miR-24-3p expression in UMSCs-Exo (Fig. 7I). These data confirmed that exosomal miR-24-3p participated in the UMSCs-Exo mediated macrophage polarization by targeting the Plcb3/NF-κB signal pathway in the inflammatory microenvironment.