This study established an appropriate priming protocol to enhance the immunosuppressive function of BM-MSCs.
Although MSCs have been proposed for the treatment of inflammatory lung diseases (17), the sustainable translation of MSC therapies to clinical settings has been hampered by the heterogeneity of MSCs and nonstandardized in vitro culture
technologies (78). Therefore, the present study focused on enhancing the
immunoregulatory function of mouse BM-MSCs by using the Pam3CSK4-polarizing protocol.
Heterogeneous MSCs can mediate their effects by sensing fluctuating
proinflammatory cytokines and TLR ligands in the microenvironment (79). At the beginning of infectious challenges, TLR ligations link innate and adaptive immunity and trigger positive feedback of the immune system by activating proinflammatory cytokine and chemokine production. To maintain the health status of the host, proinflammation-prone MSCs help to eliminate pathogens by recruiting immune effector cells to the sites of infection. However, at the end of the challenges, proinflammatory damage and oxygen stress gradients are accumulated by resultant pathogen cell necrosis and tissue damage debris. To preserve host integrity, anti-inflammation-prone MSCs reduce effector immune cell infiltration but increase
regulatory immune cell induction (48). Hence, plasticity in immunomodulatory function causes MSCs to be both advantageous and disadvantageous in the context of
inflammation.
The polarization of BM-MSCs is apparently susceptible to the stimuli conditions.
This is why the discrepancies in the experimental settings both in vitro and in vivo led to several inconsistent and partially contradictory results related to TLR-polarized MSCs (35, 80). The differences in the source of MSCs, the concentration and timing of the stimuli used, and the culture conditions might explain the conflicting results (81). Given the uncertain effects of TLR polarization on MSCs, consistent therapeutic outcomes and immunoregulatory mechanism-based validations are required to establish the
effectiveness of modified MSCs in asthma therapy.
We demonstrated that the immunomodulatory BM-MSCs were polarized by a TLR2 agonist, Pam3CSK4. According to both our in vitro and in vivo functional assays conducted using [3H]-thymidine incorporation assays and acute OVA-induced asthmatic murine models, respectively, Pam3CSK4 augmented the immunosuppressive function of BM-MSCs. The results of our immunoregulation-prone BM-MSCs stimulated by Pam3CSK4 are consistent with the findings of Pevsner-Fischer et al. (82) but inconsistent with the findings of Lei et al (83). In those two studies, the mouse BM-MSCs were incubated with different concentrations (1 vs. 2 µg/mL) of Pam3CSK4 for the same period (48 h) before they were cocultured with different responder cells (T cell line specific to MOG p35-55 vs. splenocytes isolated from C57BL/6 and BALB/c spleens) in different MSC to responder cell ratios (1:20 vs. 1:10). The discrepant protocols might have caused the distinct results.
Based on our study, we observed that the immunosuppressive BM-MSCs should be administered in post-treatment instead of pretreatment settings. In other words, they should be cocultured with 24 h preactivated CD4+ T cells in vitro and infused after asthmatic mice were OVA-challenged in vivo. The suggested modification of
BM-asthma by using BM-MSCs after the disease had developed. The ideal BM-MSC to responder CD4+ T cell ratio was 1:2 to 1:200 for acting immunosuppressive function in vitro, and the ideal infused cell number of BM-MSCs was 5 × 105 cells per mouse. BM-MSCs treated with 1 µg/mL Pam3CSK4 exerted better immunosuppressive effects than those with 5 µg/mL Pam3CSK4 in vivo. Moreover, BM-MSCs treated with 1 µg/mL Pam3CSK4 for 96 h were more effective than those treated for 24 h in vitro.
According to our data, it was in contrast to the instinct that neither more BM-MSC administration (1 × 106 cells per mouse) nor BM-MSCs pretreated with higher
Pam3CSK4 dosage (5 µg/mL) had better therapeutic effects. However, this may not be as surprising if we take the double-edged properties of the immunosuppressive
molecules into account. The sustained production of NO endows macrophages with cytostatic or cytotoxic activities against pathogens and tumor cells (84). In contrast, suppression of iNOS activity by either chemical inhibitors or genetic ablation in MSCs largely reverses the therapeutic effect of MSCs in mouse models of GvHD and delayed-type hypersensitivity responses. A study has also shown that MSCs are the critical mediators of Mycobacterium tuberculosis-induced immunosuppression in an NO-dependent manner in vitro and in vivo, suggesting a function of the MSC-NO axis in endogenously induced immunosuppression (36). Hence, the discrepant outcomes are in fact discriminated by the concentration and source of NO, and by the exact stimulatory conditions which are used to induce NO production (85). The more NO production is therefore not necessary to guarantee the better immunosuppressive function at all time.
Additionally, Pam3CSK4 was able to upregulate the expression of some adhesive molecules and chemokines in our BM-MSCs. The chemoattractant secreted by MSCs may stimulate the migration of immune cells in proximity of MSCs. It may
consequently enable the local suppressive effect of MSCs through rapidly diffusing NO
at one time, but induce inflammatory burden through recruiting proinflammatory mediators at another (36).
On the other hand, it was compatible with the instinct that the long-term stimulation (96 h) was more effective than the short-term stimulation (24 h) of Pam3CSK4 to enhance the suppressive activity of BM-MSCs. The same idea as we describe above: acute stimuli are likely to point to pathogen burden while chronic stimuli are likely to refer to hyper-responsive inflammation. Therefore, priming protocols must be adjusted before different applications.
In our in vivo experiments, the symptoms of allergic asthma, including
proinflammatory cytokine production, airway resistance, and systemic IgG1/IgG2a levels on OVA-sensitized mice, were alleviated more significantly by BM-MSCs as long as they were coupled with Pam3CSK4 induction, whereas the serum IgE levels and histopathological changes were barely engaged. In our in vivo studies, when the Mpam treatment decreased the serum IgG1/IgG2a levels more effectively, it did not decrease the serum IgE levels at the same time. This result was possibly because serum IgE and IgG1 levels represent different stages of asthma. IgE was increased at the beginning of disease development in the murine asthma model. IgG1 was later increased while the Th2-mediated disease progressed while IgE was not increasing. Therefore, at certain time point, less amelioration in IgE levels, but more amelioration in IgG1/IgG2a levels would be observed in our Mpam-treated groups. On the other hand, the
histopathological amelioration of BM-MSCs might take more time to develop because the thickness of basement membrane, epithelium, and subepithelial smooth muscle layer, and the number of mast and goblet cells gradually decreased with the longer duration of BM-MSC treatment in a chronic asthma murine model (86). Overall, we
finally determined the unvarying protocol for the steady turnout of immunosuppressive BM-MSCs in treating the acute OVA-induced asthma murine model.
This study validated the Pam3CSK4-modified BM-MSCs with the understanding of their immunosuppressive mechanisms.
Increased NO secretion
The therapeutic advantages of Mpam were proved by its more effective asthma amelioration in the OVA-sensitized asthmatic mice. Next, to validate the enhanced immunosuppressive effects of the Pam3CSK4-stimulated BM-MSCs, we further studied the immunoregulatory mechanisms of the modified BM-MSCs.
In non-stimulated cells, STAT3, like other STAT proteins, is kept in an inactive cytoplasmic form. STAT3 activation is most commonly mediated by JAK1. Nuclear translocation of STAT3 is mediated by phosphorylation of a critical tyrosine residue (Tyr705). STAT3 transcriptional activity and DNA binding are further activated through serine (Ser727) phosphorylation (87). It is well-established that STAT3 signaling drives accumulation and activation of immunosuppressive cells. STAT3 is responsible for the activation of myeloid-derived immunosuppressive cells (MDSCs), differentiation of macrophages toward the M2 phenotype, and the absence of functional DCs (88, 89).
According to our [3H]-thymidine incorporation assays, the augmented
immunosuppressive functions of BM-MSCs induced by Pam3CSK4 were diminished with the incorporation of S3I-201, a STAT3 inhibitor. Evidence has shown that the constitutive activation of STAT3 in tumor cells is a crucial suppressive mechanism in
immune reactions (90). In addition, STAT3signaling not only upregulates
proangiogenic genes, such as vascular endothelial growth factor (VEGF), hypoxia-inducible factor 1-alpha (HIF-1a), and matrix metallopeptidase (MMP) 9, etc., but also mediates immunosuppressive function in regulatory B cells (91). Moreover, the
enhanced suppressive activity through STAT3 signals in Mpam is supported by the responses of both mouse and human myeloid-derived suppressor cells (MDSCs) to tumor-derived exosomes. Tumor-derived exosomes triggered STAT3 activation in MDSCs through a TLR2/MyD88-dependent pathway (66).
NO is a rapidly diffusing molecule and an unstable oxidative species. Therefore, it can only act in close proximity to the cells producing it. The biological activity of NO diminishes markedly over a distance of a few cell diameters (36). This might explain that Pam3CSK4, which enormously stimulated NO secretion in BM-MSCs, enhanced the suppressive function of BM-MSCs in the cell-cell contact-dependent setting, but not in the transwell system. In addition, the highly-secreted NO and the enhanced
suppressive function of BM-MSCs under Pam3CSK4 stimulation was diminished with STAT3 abrogation in a cell-cell contact-dependent manner. Further, NO production is catalyzed by NOS, nitric oxide synthase, for which there are three genes in humans and mice: iNOS, inducible primarily in macrophages; nNOS, in neurons; and eNOS, in endothelial cells. Among the three candidates, iNOS expression is inducible and plays a major role in immune regulation (36). Compatibly, other reports also demonstrate that iNOS transcription is downstream of the TLR and STAT signaling pathways in murine macrophages (92, 93). The unknown mechanism which regulates NO expression in the Pam3CSK4-induced BM-MSCs was thus hypothesized to be through the
TLR2/STAT3/iNOS signaling pathway. To support this idea, our results demonstrated
either the STAT3 or iNOS inhibitor. Diminished NO secretion because of STAT3 or iNOS inhibition in Mpam was therefore possibly related to the abrogation of the enhanced suppressive function of Mpam. Moreover, our Western blot results revealed that the STAT3 inhibitor impeded STAT3 phosphorylation, and subsequently
diminished iNOS expression in Mpam.
How then can such labile NO make MSCs immunosuppressive? One of the
characteristics of MSCs is their ability to migrate to sites of damaged tissue. MSCs have been shown to exert anti-inflammatory effects on almost all the cells of the innate and adaptive immune systems via a variety of mechanisms, notably cytokine and chemokine secretion. In addition, the chemoattractant profiles expressed by MSCs vary depending on the specific microenvironment which MSCs contact with. Through these chemokines and adhesion molecules, immune cells possibly accumulate in close proximity to the MSCs, where the high concentration of secreted NO can suppress the immune cells more efficiently (36, 94).
Our data suggested that Pam3CSK4 induced the expression of chemokines, such as CCL3, CCL5, ccl7, and mcp-1; and adhesion molecule, such as VCAM-1 in BM-MSCs.
CCL3 and CCL5 are CCR5 ligands, and CCL7 is one of the CCR3 ligands. CCR5 is a chemokine receptor that is preferentially expressed on Treg cells. CCR5+ T cells have been shown to play important roles in atherosclerosis, rheumatoid arthritis, and
autoimmune encephalomyelitis. CCR5 expression is associated with Treg’s homing to the damaged tissues and suppressing local inflammation (95). However, in lupus nephritis patients, CCR5 is one of the primary receptors in the mechanism of recruiting all T cells into the inflamed kidney (96). In addition, cancer cells can promote the recruitment of Treg cells into tumors and can facilitate their peripheral expansion and retention through secreting chemokines and cytokines. In 3-methylcholanthrene-induced
fibrosarcoma mice, the receptor-ligand interactions for CCR3 and CCR5 have no preference for Foxp3+ T cells over Foxp3- T cells (97). Therefore, it is uncertain whether the upregulated expression of CCL3, CCL5, and ccl7 in BM-MSCs by
Pam3CSK4 priming is associated with creating a better circumstance for the suppressive action of NO.
On the other hand, VCAM-1 is normally expressed on the surface of antigen presenting cells, mediating target cell binding by interaction with its receptor on
leukocytes and lymphocytes. In an in vitro study, a blocking antibody against VCAM-1 in MSCs reduces the number of adhesive T cells to MSCs significantly. Moreover, although the production of NO by MSCs is unaffected by blockade of VCAM-1 in MSCs, the VCAM-1 antibody significantly diminishes the suppressive function of MSCs observed using T cell [H3] incorporation assays (98). However, despite it is conceivable that the T lymphocytes may be held in place by VCAM-1 so that the effects of NO from MSCs can be attained in vitro, The interaction of T lymphocytes with VCAM-1 on MSCs has not been established in vivo.
MCP-1 is a CCR2 ligand. A report reveals that when MCP-1-/- BM-MSCs are transplanted into C57BL/6 mice, they fail to induce CD3+ T cell apoptosis and Treg upregulation in vivo. However, the MCP-1 secretion of BM-MSCs is suggested to be regulated by FASL in that article. FASL-induced MCP-1 secretion in BM-MSCs recruits T cells for FASL-mediated apoptosis. The apoptotic T cells subsequently triggered macrophages to produce high levels of TGF-b, which in turn leads to the upregulation of CD4+CD25+Foxp3+ Treg cells (99). In contrast, compared to Mctrl, Mpam neither expressed more FASL, nor triggered more caspase-3 in T cells in accordance with our results. Therefore, Mpam-induced mcp-1 expression was unlikely
Apparently, our results could only suggest that Pam3CSK4 enhanced the
immunosuppressive function of MSCs through stimulating NO secretion in BM-MSCs. However, more experiments are still needed to explore the mechanisms by which Mpam could bring T cells closer enough to enable higher level NO to act.
More CD4+CD25+Foxp3+ Treg cells
The peripheral tolerance mechanisms work naturally as a compensational strategy for the self-reactive immune cell eliminations (100). MSCs have been shown to inhibit macrophage and dendritic cell maturation, induce a shift in the T helper cell
polarization, and promote regulatory T cell (Treg) differentiation (101). Antigen-specificity and longevity make Treg a more potent therapy with prolonged protection against recurrence of disease. Several T cell subsets with regulatory properties have been described, such as the IL-10-producing Tr1 cells and CD8+CD28- T cells.
However, the most studied cell in recent years is notably the naturally occurring
CD4+CD25+ Tregs that express the transcription factor Foxp3 (8). IL-2R is composed of a-chain (CD25), b-chain (CD122), and the common cytokine receptor g chain (CD132).
Although all three chains contribute to IL-2 binding, only the b- and g-chains are responsible for signal transduction following IL-2 binding to the receptor. Furthermore, because inducing the production of CD4+CD25+Foxp3+ lymphocytes has been indicated as a potential mechanism of MSCs in ameliorating autoimmune disease and in
immunoprivileging cancer cells (102, 103), Foxp3 expressing CD4+CD25+ T cells are primarily under the scope of our investigation.
Our in vitro results demonstrate that the production of NO, an immunoregulatory molecule of BM-MSCs, was highly induced by Pam3CSK4. NO secreted from Mpam
might subsequently induce additional CD4+CD25+Foxp3+ regulatory T cells, as shown in other reports, in a cell-cell contact-dependent manner (104). The CD4+CD25+Foxp3+ cells were also detected in the lung tissues of the asthma murine model conducted in this report. In consistent with our in vitro study, the Mpam treatment induced more CD4+CD25+Foxp3+ cells in the lung tissues than the Mctrl treatment did.
Our data suggested that cocultured BM-MSCs upregulated Foxp3 expressing CD4+CD25+ T cells. The immunosuppressive BM-MSCs polarized by Pam3CSK4
additionally educated more CD4+CD25+Foxp3+ lymphocytes through iNOS activity in the downstream of STAT3 stimulation. However, NO has been implicated in the differentiation of various T cell subtypes, including the NO-dependent induction of Foxp3-- or Foxp3+ Treg. In addition, to help protect the organisms against exaggerated immune responses, iNOS-derived NO has also been shown to suppress T cells through versatile mechanisms other than Treg induction. For instance, NO promotes apoptotic signals while inhibits proliferative signals in T cells (105, 106). At high concentration, NO can inhibit TCR-induced T cell proliferation and cytokine production. Also, the high concentration of NO is found to suppress T cell proliferation through inhibiting STAT5 phosphorylation (36). Thereafter, inducing additional CD4+CD25+Foxp3+ T cells might be just one mechanism through which Mpam performed its enhanced immunosuppressive function. Furthermore, because MSCs have been demonstrated to reduce allergic sensitization in asthmatic mice either through weakening the antigen presenting properties of dendritic cells or through skewing the suppressive phenotype of macrophages (107, 108), the transplantation of Mpam might have multiple effector targets to suppress asthmatic signs in the murine model. Furthermore, the therapeutic advantages of MSCs in chronic asthmatic felines led us to envision the notion of
modification (109). The induction of regulatory T cells by MSCs is additionally suggested as a crucial point for the long-term effects of MSCs after infusion (110).
CD4+CD25+Foxp3+ Treg can be developed in the thymus as well as can be induced in the periphery. Thymus-derived and peripherally induced Treg cells are both critically involved in prevention of autoimmune diseases. However, despite thymic and
peripherally induced Treg are phenotypically and functionally similar, their differential mechanisms are distinct. Treg development and Foxp3 induction are highly dependent on CD28 costimulation in thymus, whereas only when TGF-b is present, do Foxp3+ Treg cells are generated in periphery (111).
In detail, at day 3 after birth, autoreactive Treg repertoire starts being shaped in the thymus through a two-step positive selection, combined with a reduced sensitivity to negative selection (69). The transgenic mice whose thymic cortical epithelial cells express a single MHC class II/peptide ligand are generated to clarify the differences between Treg and conventional T cells in positive selection process. Functional, single ligand-reactive Treg precursors, are preferentially selected by these transgenic mice. It indicates that positive selection of Treg precursors occurring in thymic cortical
epithelium requires high avidity interactions. By contrast, positive selection of
conventional T precursors is very MHC/peptide-specific and is mediated by nonagonist self-ligands (112). After high-avidity autoreactive Treg precursors are directly selected in thymic cortical epithelium, the second step of positive selection and the negative selection are taken place in thymic medullary epithelium. A transcription factor, autoimmune regulator (AIRE), is expressed in thymic medullary epithelium. AIRE promotes the expression of peripheral tissue antigens in thymic medullary epithelial cells. Autoimmune disorders which are developed either for AIRE deficiency or for MHC class II lack in thymic medullae demonstrate the idea that interactions of
developing Treg cells with MHC class II medullae are sufficient to support Treg.
Hence, after migration of developing Treg cells from the cortex to the medulla, interactions with cognate antigen on thymic medullary epithelium may then serve to selectively induce the survival and/or population expansion of otherwise short-lived Treg precursors (113).
The thymic medulla is packed with BM-derived antigen presenting cells and is permeable to circulating self-antigens entering from the blood stream. Thus, the medulla
The thymic medulla is packed with BM-derived antigen presenting cells and is permeable to circulating self-antigens entering from the blood stream. Thus, the medulla