Pretreated Mesenchymal Stem Cells and Their Secretome: Enhanced Immunotherapeutic Strategies

Mesenchymal stem cells (MSCs) with self-renewing, multilineage differentiation and immunomodulatory properties, have been extensively studied in the field of regenerative medicine and proved to have significant therapeutic potential in many different pathological conditions. The role of MSCs mainly depends on their paracrine components, namely secretome. However, the components of MSC-derived secretome are not constant and are affected by the stimulation MSCs are exposed to. Therefore, the content and composition of secretome can be regulated by the pretreatment of MSCs. We summarize the effects of different pretreatments on MSCs and their secretome, focusing on their immunomodulatory properties, in order to provide new insights for the therapeutic application of MSCs and their secretome in inflammatory immune diseases.


Introduction
Mesenchymal stem cells (MSCs) play important roles in regenerative medicine due to their potential for self-renewal and multi-directional differentiation [1]. MSCs were also found to have strong immunoregulation capabilities caused by the secretion of various signal factors or cell to cell contact [2].
According to research, the immunosuppressive properties of MSCs are necessary for restoring immune homeostasis in inflammatory diseases; however, in an active disease state, inflammation frequently changes dynamically, which can alter the immune properties [12]. MSCs can become immunosuppressive under strong inflammation, while weak inflammation enhances their immune responses [13].

Hypoxia
MSCs are generally located in a microenvironment of low oxygen (≤ 2-8% O2) in vivo, whereas the normoxic oxygen tension used in conventional cells cultures is the atmospheric pressure (21% O2) [25], and this variation may have an effect on cells activities.
Hypoxia-pretreated MSCs could cause peripheral blood mononuclear cells (PBMCs) to undergo apoptosis by boosting the synthesis of the Fas ligand in vitro and reduce local skin inflammation by decreasing tumor necrosis factor-α (TNF-α) while increasing interleukin (IL)-10 [27]. Hypoxia and calcium ions exhibited enhanced immunomodulatory functions of MSCs used to treat GVHD via the monocyte chemoattractant protein-1 (MCP-1) and p53/p21 cascade [28]. The expression of indoleamine 2, 3-dioxygenase (IDO) [29] and inducible nitric oxide synthase (iNOS) [30], mediators of MSCs immunomodulation, increased after hypoxia pretreatment, and then inhibited CD8 + T-cell activation and promoted Treg proliferation in vitro. Under 1 and 20% oxygen levels, MSCs were capable of strong upregulation of IDO upon stimulation with interferon-γ (IFN-γ) and TNF-α to inhibit T-cell proliferation in vitro [15]. Interestingly, there was no difference between hypoxia and normoxia in the presence of inflammatory factors. Another in vitro study indicated that a combination of IFN-γ-and hypoxia-pretreated MSCs displayed higher induction of IDO over IFN-γ alone, which was related to the metabolic shift to glycolysis of MSCs caused by hypoxia [31]. The CM containing increased growth factors, IL-6 and IL-8 derived from hypoxia-pretreated MSCs promoted skin wound healing [32]. CM from MSCs cultured in near anoxic (0.1% O 2 ) conditions exhibited a significant increased production of vascular endothelial growth factor (VEGF), IL-8, MCP-1 and RANTES while decreased inflammatory cytokines, which indicated its regenerative potential [33]. Intravenous injection of the CM improved neuroinflammation after brain injury by increasing M2 microglia while decreasing M1 phenotypes [34]. Consecutive culture in 1% O 2 promoted MSCs to secrete immunomodulatory cytokines in vitro, and in acute lung injury (ALI) of mice, their CM could limit lung inflammation by modulating anti-inflammatory polarization of macrophages [35]. Furthermore, a hypoxic environment could promote MSCs to release more EVs, which significantly ameliorated the eosinophils and pro-inflammatory mediators in chronic asthmatic mice [36]. Meanwhile, the EVs could alleviate inflammatory bone loss by transferring miR-210-3p, which directly targets NF-κB1 to induce M2 macrophages polarization and inhibit osteoclastogenesis [37]. Exosomes derived from MSCs under hypoxia promoted functional behavioral recovery of spinal cord injury by shifting microglial polarization from M1 to M2 through miR-216a-5p/TLR4/NF-κB/PI3K/AKT signaling cascades [25]. Proteomics analysis of MSCs and their EVs found the enrichment of biological processes and pathways related to glycolysis, the immune system and extracellular matrix organization in hypoxic condition [38].
To sum up, by mimicking low oxygen levels in the inflammatory sites, hypoxic culture pretreatment of the MSC-derived secretome improved its immunomodulatory effects, despite differences in the hypoxic conditions. Beyond that, different tissues and organs in the human body have different levels of oxygen, so different hypoxic pretreatment conditions should be designed according to the specific application in order to better simulate the niche in vivo conditions required for MSCs.

Inflammatory Factors
Aside from hypoxia, the production of inflammatory factors is another simultaneous and related condition in inflammatory diseases, which can actively communicate with the resident MSCs in the tissue microenvironment. Many studies have demonstrated that the pretreatment of MSCs with inflammatory factors could improve their immunomodulatory functions (Table 2).

IFN-γ
IFN-γ is a key pro-inflammatory factor that plays an important role in regulating the immunomodulatory function of MSCs. In 2006, researchers first proposed IFN-γ as a key player in activating the immunosuppressive ability of MSCs, at least in part, by enhancing IDO activity [39]. Then, another study found that IFN-γ enhanced the immunosuppressive activity of MSCs against NK cell activity and protected MSCs from NK-cell-mediated cytotoxicity by promoting the production of IDO and prostaglandin E2 (PGE2) [40]. In addition, subsequent studies found that IDO expression was conducted through the JAK/STAT1 pathway [41] and glucose metabolic reprogramming [89]. Meanwhile, in experimental colitis, mucosal damage was reduced because the pretreated MSCs significantly inhibited type 1 T helper (Th1) inflammatory responses by decreasing T-cell activation and proinflammatory cytokine secretion [42]. Another study indicated that the pretreated MSCs inhibited T-cell proliferation, Th1 cytokine secretion and T-cell degranulation through B7H1 and B7DC/PD1 pathways [43]. Although IFN-γ pretreatment increased IDO activity of MSCs, IDO was unnecessary for inhibition of T-cell function [43]. In addition, the pretreatment significantly increased the factor H secretion of MSCs, and then inhibited the complement activation, suggesting another mechanism of MSCs' broad immunosuppressive effects [46]. Exosomes derived from pretreated MSCs increased miR-125a and miR-125b, which inhibited the differentiation of Th17 cells and then had therapeutic efficacy for colitis [44]. In addition, pretreated MSC-derived EVs, containing TNF-stimulated gene 6 protein (TSG-6) and A20, inhibited T-cell proliferation [45].

TNF-α
TNF-α, another important pro-inflammatory cytokine, was undoubtedly involved in the immunoregulation of MSCs. CM from TNF-α-pretreated MSCs eased the symptoms of experimental allergic conjunctivitis via cyclooxygenase 2 (COX2)/PGE2 signaling [47]. Furthermore, TNF-α enhanced CD73 expression of MSC-derived exosomes, leading to the polarization of M2 macrophages. Local injection of exosomes significantly reduced the resorption of periodontal bone in mouse periodontitis models and exosomal miR -1260b played an important role in inhibiting the osteoclast activity via the Wnt5a/RANKL pathway [48]. Meanwhile, miRNA-299-3p was upregulated in the pretreated MSC-derived exosomes, which attenuated inflammatory damage of acute liver failure and promoted liver tissue repair [49]. In addition, TNF-α pretreatment could also exert immunosuppressive function by increasing the factor H secretion of MSCs [46].

IL-1β
In inflammatory diseases, IL-1β also acts as a pathological mediator. MSCs pretreated with IL-1β could treat colitis by increasing COX2, IL-6 and IL-8 to regulate the polarization and differentiation of immune cells [50]. Pretreatment of IL-1β resulted in a strong upregulation of miR-146a [51] and miR-21 [52] in MSC-derived exosomes, and this transferred to macrophages, resulting in M2 polarization. Meanwhile, the exosomes significantly inhibited astrogliosis and inflammatory responses via the Nrf-2 signaling pathway [53], and enhanced anti-inflammatory activity via miR-147b mediated inhibition of NF-κB pathway [54]. By upregulating the expression of transforming growth factor-β1 (TGF-β1) and matrix metalloproteinases (MMPs), CM from pretreated MSCs promoted wound healing in vivo [55]. MSCs pretreated with IL-1, including IL-1α and β, increased expression of granulocyte-colony-stimulating factor via IL-1 receptor 1, and the CM decreased the production of inflammatory factors in activated microglial cells [56]. Additionally, by secreting pro-inflammatory cytokines, the secretome from pretreated MSCs regulated the inflammatory response and aggrecan deposition in degenerative intervertebral discs [57].

Lipopolysaccharide (LPS)
LPS is the main cell wall component of Gram-negative bacteria and play a crucial role in the interaction between pathogens and the host immune system [90]. It is also a commonly inflammatory inducer in the study of inflammatory diseases.
It was found that a low concentration of LPS-pretreated MSCs inhibited activated T-cell by up-regulating nitric oxide (NO) [58]. Pretreated MSCs displayed high levels of IL-6, IL-8 and more effectively promoted diabetic skin-flap survival [59]. The CM of pretreated MSCs defended against V. cholerae infection by inducing vibriocidal antibodies and regulating inflammatory and anti-inflammatory effects [60]. Meanwhile, the CM promoted the recruitment and phagocytosis of polymorphonuclear neutrophil granulocytes (PMNs) by upregulating IL-6, IL-8 and macrophage migration inhibitory factor (MIF) of MSCs [64]. Furthermore, LPS pretreatment of MSC-derived exosomes significantly increased anti-inflammatory cytokines while decreased pro-inflammatory cytokines of monocytes and macrophages, promoted their polarization to M2 macrophages and reduced the inflammatory response of diabetic skin wounds and MI [61,62]. The miRlet-7/TLR4/NF-B/STAT3/ATK pathway may be responsible for this effect [61]. Under inflammatory conditions, LPS-pretreated MSC-derived EVs could decrease the RANKL/OPG of PDLSCs by inhibiting the ROS/JNK pathway and promote M2 macrophages polarization through the ROS/ERK pathway, which enhanced the therapeutic efficacy for periodontitis [63].
Some studies have suggested that LPS promotes MSCs to generate a pro-inflammatory phenotype and reverse their therapeutic immunosuppressive effect. Brief LPS pretreatment of MSCs increased IL-6 and IL-8 and showed a decreased capacity to inhibit T cells. It also reversed the protective effect of MSCs against experimental autoimmune encephalomyelitis (EAE) and colitis [65,66]. Meanwhile, pretreated MSC-derived exosomes promoted M1 macrophages polarization, IL-6 and TNF-α production and Th17 cells differentiation [67]. Furthermore, LPS promoted the expression of thrombospondin-1 of MSCs, which was involved in LPS proinflammatory regulation [67].
This immunomodulatory difference of LPS-pretreated MSCs may be caused by cell types (mouse versus human), tissue origin, concentration of LPS, in vivo versus in vitro studies and cells' pretreatment time. A study indicated that different LPS pretreatment time could change the phenotype of MSCs, playing different immunomodulatory and therapeutic roles [68].
Previous research has observed diverse effects in vitro. TLR4-activated MSCs mostly expressed pro-inflammatory mediators, and TLR3-activated MSCs mostly expressed immunosuppressive factors [92]. This is consistent with the above results of Poly (I:C) pretreated MSCs, while there is still uncertainty related to LPS pretreatment.

Combined Pretreatment of Inflammatory Factors
Combining TNF-α and IFN-γ could also enhance the immunosuppressive activity of MSCs. In normoxic conditions, hypoxia-inducible factor 1α (HIF1α) was induced and maintained by the expression of ROS of activated MSCs [71]. Then, HIF1α enhanced the immunosuppressive activity of MSCs by modulating their metabolic status [71]. The pro-inflammatory effects caused by high levels of palmitate were fully reversed by the pretreated MSCs, which provided a robust and durable prelicensing regimen for protecting MSCs' immunosuppressive function in relation to type 2 diabetes [72]. Meanwhile, the pretreatment increased IDO production of MSCs and promoted M2 macrophages polarization to inhibit PBMCs [73]. In acute myeloid leukemia therapy, the pretreated MSCs increased tumor necrosis factor-α-related apoptosis-inducing ligand and IDO, enhancing the pro-apoptotic effect of leukemic cells [80]. The CM derived from combined pretreated MSCs efficiently accelerated wound closure via IL-6-dependent M2 polarization [74,75]. Related mechanism research showed that the combined pretreatment increased the production of IDO and TSG-6 by PI3K/AKT pathway, thereby enhanced the immunosuppressive ability of MSCs [76]. Interestingly, another study indicated that MSCs can be induced to generate a pro-inflammatory phenotype by the combined pretreatment, but these MSCs also produced anti-inflammatory factors, which concurred with the classical conception that the two microenvironments inhibiting and promoting tumor growth coexist [77]. The pro-inflammatory chemokines induced aggregation of immune cells closely associated with MSCs, creating a microenvironment in which the local effects produced by MSCs lead to strong immunosuppression. In addition, chaperone-mediated autophagy activation was inhibited in pretreated MSCs, leading to immunosuppressive function [78]. Another study indicated the pretreatment induced miR-155 expression of MSCs and inhibited the immunosuppressive capacity by reducing iNOS expression which might protect MSCs from NO in inflammatory situations [79].
IL-β combined with IFN-γ significantly improved the immunoregulatory effects of MSCs by increasing the production of IDO and PGE2 [81]. Meanwhile, pretreatment of MSCs promoted the secretion of IL-6, inhibited M1 macrophages polarization under inflammation and inhibited M2b polarization under anti-inflammation via the IL-6 pathway [82]. This finding supports the implication that inflammatory condition was a critical determinant of the immunomodulatory characteristics of MSCs described above.
IL-17 could dramatically enhance the immunosuppressive effect of MSCs induced by IFN-γ and TNF-α through the promoted expression of iNOS in liver injury [83] and tumor progression [84].
Compared to the combination of IFN-γ and TNF-α or single treatment, MSCs pretreated by LPS and TNF-α promoted M2 macrophages polarization and enhanced osteogenic differentiation by increasing PGE2 and Arginase-1 [85].
IFN-γ and poly(I:C) pretreated MSCs were more effective at improving experimental colitis. They increased Treg, decreased inflammation and stimulated intestinal epithelial regeneration by promoting the expression of IDO1 [86].
MSCs pretreated by IL-1β, IL-6, TNF-α and IFN-γ increased the production of inactive chemerin, which could then be activated by inflammation-induced tissue proteases, thereby promoting the migration of ChemR23-expressing immune cells with immunomodulatory effects [88].
To summarize, many inflammatory factors have been used to pretreat MSCs, and in general, pro-inflammatory pretreatments increased the immunosuppressive properties of MSCs by inducing the anti-inflammatory phenotype of immune cells and stimulating the expression of anti-inflammatory factors. However, due to the fact that there are numerous inflammatory factors and more choices for combined applications, it is difficult to reach unified conclusions or conduct in-depth research in a certain direction. Meanwhile, some studies suggest a combination of different inflammatory factors is superior to a single one, but the other studies directly applied combinations of inflammatory factors and compared them with untreated MSCs. So, it is impossible to verify whether the combination is better than the single factor, which needs more research to confirm this. There is one more point that, so far, there is no clear standard for the selection of combined inflammatory factors.
It is worth noting that MSCs, like immune cells, can retain information from environmental stimuli for a period of time. When pretreated with LPS or TNF-α, MSCs displayed increased secretion of IL-6, IL-8 or MCP-1 and retained high expression of these cytokines for over 7 days without the influence of cell division. These cytokines were increased further after secondary stimulation [59]. Following LPS pretreatment, MSCs increased the production of iNOS, IL-1β and IL-6 after IFN-γ and TNF-α retreatment, and decreased expression of proinflammatory genes after LPS retreatment [93]. Thus, the inflammatory stimuli treated MSC-derived-secretome can contain more abundant regulatory factors. This also suggests that the secondary stimulation of MSC-derived secretome could lead to more effective in immunomodulation, which is worth further exploration.

D Cell Culture
MSCs are typically grown in two-dimensional (2D) culture plates. However, removing them from their endogenous 3D niche, as well as enzymatic passaging, can result in a loss of multipotency, accumulation of chromosomal aberrations within the MSCs genome, dysfunction in DNA repair, replicative senescence, decreased surface markers (e.g., CD105, CD90, CD73) associated with the MSCs' undifferentiated phenotype [94][95][96][97][98]. Three-dimensional cell culture systems' ability to bridge the gap between cell-based methods and animal models for studying the repair and replacement of tissues and organs is becoming increasingly apparent [99].
Three-dimensional cell culture methods have been used to improve the immunomodulatory properties of MSCs (Table 3). A study indicated that the loss of T-cell-suppressive abilities in MSC spheroids was caused by increasing PGE2 and decreasing IDO and CD73 [100]. MSC spheroids promoted the paracrine of TSG-6 and ameliorated renal function of acute kidney injury (AKI) compared with 2D-cultured cells [101]. After transplantation into mouse, MSC spheroids promoted the production of IFN-γ and IL-6 while suppressing TNF-α, thereby alleviating liver necrosis and promoting regeneration [102]. MSC spheroids increased the production of immunomodulatory paracrine factors and the immunosuppressive effect on macrophages functional activity, which was further strengthened by pretreatment with IFN-γ and TNF-α [103]. The CM from MSC spheroids efficiently inhibited the expression of inflammatory cytokine, increased the expression of anti-inflammatory cytokines in LPS activated macrophages, and promoted the transformation of M1 into M2 macrophages [104][105][106]. It also had the potential to accelerate wound healing and mature, fully functional tissue regeneration [107]. In addition, the predominant expression of anti-inflammatory cytokines in CM from 3D cultured MSCs demonstrated its superior regenerative and anti-inflammatory potential in arthritis [108] and ear wound model [109]. Compared with 2D culture, spheroid MSC-derived exosomes indicated enhanced immunomodulatory effects by restoring the reactive Th17/Treg balance in inflamed periodontal tissues and experimental colitis [110]. MSC spheroids in combination with hypoxia or hypoxic-mimetic condition provided by dimethyloxalylglycine synergistically increased TSG-6, MMP-2 and VEGF involved in tissue repair processes [111].
MSCs cultured on electrospun gelatine/polycaprolactone fibers scaffolds in advance promoted the expression of IL-6 and chemokines, and the secretome significantly improved corneal wound healing [112]. Synthetic polyisocyanide hydrogel created a 3D culture environment and directed the secretome composition of MSCs, which had therapeutic potential in wound healing [14].
In summation, 3D cultures of MSCs efficiently exerted immunosuppressive effects. However, there was still some debate regarding the MSC spheroids. When they reached critical sizes, their nutrient and oxygen supply might be non-homogenous and could affect cells proliferation and viability [113]. Compared with spherical culture, the secretome produced by 3D hydrogel-cultured MSCs had enhanced immunomodulatory capabilities and regeneration potential [114]. Research has found that the matrix stiffness [115] and composition [14] of hydrogel, and enhanced cell-cell contacts in spheroids [103] could regulate the immunomodulatory function of MSCs. At present, most studies focus on the immune regulation and mechanism of 3D-cultured MSCs, while the interaction between cells has not been thoroughly and comprehensively explored. Studies showed that the microenvironment established within the spheroids acted in an autocrine process to increase the cell-cell interaction [107]. In addition, the matrix stiffness, viscoelasticity, porosity, and degradation of 3D hydrogel and scaffold directly affect MSCs' aggregation and promote their secretion by increasing cell-cell contacts [116].
Therefore, hydrogel could be designed with suitable structures and properties to ultimately generate an optimized secretome. Meanwhile, as simulated ECM, hydrogels could be used as a tool to collect the secretome of MSCs, which is worthy of further investigation.
Broad-spectrum antibacterial fusion peptide BPI21 and LL-37 transfected MSCs not only enhanced antibacterial and toxin-neutralizing activities but displayed significant immunomodulatory effects [128].
MSCs, overexpressed by HIF1α, exhibited increased resistance to NK-cell-mediated lysis, impaired DCs differentiation and induced higher attraction of monocytes which acquired immunosuppressive properties efficiently [129]. HIF1α-overexpressed MSCs with or without pro-inflammatory treatment released EVs with potent immunomodulatory activity [130,131].
Hepatocyte growth factor overexpressed MSCs could enhance the treatment effect on psoriasis by regulating T-cell differentiation and inflammatory factors expression [133].
Compared with MSC spheroids pretreated with IFN-γ, heparin microparticle-mediated release of IFN-γ in MSC spheroids could increase sustained immunomodulatory effects by inducing continuous IDO expression throughout 7 days of culture and anti-inflammatory cytokines secretion of monocytes [134]. Similarly, to achieve sustained IDO production of MSCs, the internalization of budesonide -loaded poly (lactic-co-glycolic acid) microparticles into MSCs could improve their immunosuppressive properties [135].
Engineered MSCs have advantages in maximizing and maintaining immunomodulatory potential, and they enable the enhanced expression of specific factors without consideration for which tissues or donors MSCs should derive. Engineered MSCs are mostly directly applied at present, and given that the effect of engineered MSCs is superior to the transfected factor itself, it is speculated that the secretome derived from engineered MSCs is more useful. Subsequent studies should concentrate on the effect of secretome to make the most of its advantages. Meanwhile, the complicated process and the influence of transfection efficiency and stability continue to restrict development. The addition of drug/cytokine-loaded microparticles provides an effective method to enhance and maintain immunomodulatory activity of MSCs. But drug/cytokine selection, sustained release time and dosage are still in need of further study.

Pharmacological or Chemical Agents
Pharmacological or chemical agents, with simple, cheap and efficient features, also represent common pretreatments for MSCs in immunomodulation (Table 5).
Tetrandrine is a potent calcium channel blocker and anti-inflammatory antioxidant agent. It can increase PGE2 secretion of MSCs through the NF-κB/COX2 signaling pathway and significantly reduced pro-inflammatory factors production [137].
Glucocorticoids have seen decades of clinical use as potent anti-inflammatory and immunosuppressive agents. Budesonide increased IDO expression of MSCs following IFN-γ treatment and restored IDO expression in over-passaged MSCs via glucocorticoid-induced expression of transcriptional enhancer Forkhead box O3 (FOXO3) [135]. Pretreatment of MSCs with dexamethasone enhanced immunosuppression in mixed lymphocyte reaction [138]. Dexamethasone and JWH-133, a cannabinoid receptor 2 selective agonist, could restore the anti-inflammatory and immunoregulatory properties of immune thrombocytopenia patient-derived-MSCs via Bcl2 signaling [139].
Pretreated MSCs with chlorzoxazone, a drug currently used for muscle relaxation, were able to significantly suppress T cells, increase IDO and other mediators, and more effectively reduce inflammatory infiltration and tissue damage associated with AKI by modulating the phosphorylation of FOXO3 [144].
Metformin, commonly used to treat type 2 diabetes, optimized the immunoregulatory properties of MSCs by enhancing the production of IDO, IL-10 and TGF-β through the STAT1 pathway, resulting in significant disease activity improvement in lupus [145].
Neuropeptide Substance P (SP) was involved in inflammation regulation and wound healing. By promoting the release of TGF-β1, SP pretreated MSCs might restore the reduced immunosuppressive function of late-passage MSCs, which could improve the therapeutic effectiveness of expanded MSCs in long-term culture [146].
Rapamycin was a macrolide used clinically for anti-fungal treatment and immunoregulation. Rapamycin pretreatment of MSCs mediated T-cell proliferation, migration and differentiation by increasing the secretion of TGF-β1 and CXC-chemokine ligand 8 of MSCs [147,148], and improved the pathological changes, inflammatory cytokine levels and hepatic function in the liver damage model [149]. All of the above effects were achieved by inducing autophagy.
Cytochalasin B (CB) had a reversible effect on the interaction of cytoskeleton and membrane. CB pretreated MSC-derived microvesicles could suppress the production of anti-sheep red blood cell antibody [150], the proliferation of PBMCs and the activation of Th cells, B cells and T-cytotoxic lymphocytes [151].
An IDO metabolite, kynurenic acid, specifically regulated TSG-6 production. Kynurenic acid pretreated MSCs could further promote TSG-6 production and thus enhance the therapeutic capacity of ALI [152].
Glutamine is a nonessential amino acid that modulates immune responses. Glutamine in high doses used to pretreat MSCs modulated immune responses by regulating anti-and pro-inflammatory cytokines, while the CM had immunosuppressive effects on lymphocytes and macrophages [153].
Magnesium was able to modulate the immunoregulatory properties of MSCs by decreasing pro-inflammatory cytokines and increasing anti-inflammatory cytokines of MSCs stimulated with LPS or TNF-α, and the CM modulated the production profile of inflammatory cytokines in macrophages [155].
To sum up, these studies demonstrated the potentially beneficial effects of MSCs pretreated with pharmacological or chemical agents. This kind of pretreatment can effectively avoid the side effects of drug administration caused by in vivo application, control the drug dosage and improve the efficiency of treatment. The study of drug selection, specific pretreatment conditions and related mechanisms should be carried out in the future.

Immunomodulatory Mechanisms of Pretreated MSCs and Their Secretome
Pretreated MSCs and their secretome modulate inflammatory and immune responses by promoting/inhibiting multiple immune cells and cytokines ( Figure 1) and play active roles in different inflammatory disease models ( Figure 2). Obviously, several factors play important roles in this process.
The immunosuppressive effects of IDO are related to its characteristic tryptophan catabolism: IDO mediated tryptophan depletion and production of kynurenine, which was involved in the inhibition of immune cells [158]. Tryptophan is an essential amino acid and its reduction would affect the survival of immune cells [159]. Kynurenine could induce T-cell arrest or apoptosis, and its accumulation may induce the differentiation of Treg [160].

Immunomodulatory Mechanisms of Pretreated MSCs and Their Secretome
Pretreated MSCs and their secretome modulate inflammatory and immune responses by promoting/inhibiting multiple immune cells and cytokines ( Figure 1) and play active roles in different inflammatory disease models ( Figure 2). Obviously, several factors play important roles in this process.  Inflammatory factors and engineering pretreatments promoted iNOS expression in the case of rodent MSCs. iNOS catalyzes arginine, resulting in the production of NO, which mediates a variety of biological processes, including immune regulation [161]. NO regulated the immune cells, particularly the proliferation and responsiveness [16,30,42,65,68,83], apoptosis [58], Th1 and Th17 cells differentiation [65,68] in vitro and T-cell infiltration [83], the percentage of Th1 and Th17 subsets [65,68] and inflammatory cytokines production [42] in inflammatory tissue, then consequently regulated the immune response.
The inhibition of T cells may be attributed to the process of NO production depleting L-arginine, which is an essential nutrient for T cells [162]. Meanwhile, NO caused T-cell cycle arrest by affecting the JAK signal transductor and STAT signal pathway. In addition, NO could also regulate MAPK and NF-κB [163]. Furthermore, NO induced Treg which inhibited T cells and Th17 cells in vitro and in vivo [164], interfered with the production of pro-inflammatory cytokines in macrophages [165], and attenuated the cytotoxicity and capacity of NK cells to produce inflammatory cytokines [166]. Additionally, iNOS attenuated inflammation by reducing the influx of DCs and T cells and their ability to produce inflammatory cytokines [167].
Researchers also identified a common downstream mechanism of T-cell inhibition in both IDO and iNOS as endoplasmic-reticulum stress [168].

COX2/PGE2
PGE2, a mediator of many physiological and pathological functions, and COX2, a key enzyme that catalyzed the rate-limiting steps in the conversion of AAs to prostaglandins, both play crucial roles in the inflammatory process [169]. Consecutive hypoxia, inflammatory factors, spheroid culture and pharmacological pretreatment effectively promote COX2/PGE2 expression. The COX2/PGE2 axis effectively promoted M2 macrophages polarization [35,50,85,105,106,170]. A study indicated that COX2-dependent production of PGE2 was a major paracrine mediator of MSCs in regulating M1 to M2 macrophages polarization by changing their metabolic status [171]. In addition, PGE2 reprogramed macrophages by acting on their EP2 and EP4 receptors of them and stimulated the expression of IL-10 [172].

MSC-Derived EVs
MSC-derived EVs regulated the biological function of immune cells by transferring multiple types of bioactive cargos, such as nucleic acids and proteins to targeting cells. Local and systemic application efficiently inhibited detrimental immune responses and promoted regeneration in inflamed tissues [181].

IL-6
IL-6 is a pleiotropic cytokine involved in inflammation, immune response and hematopoiesis [182]. Pretreated MSCs exhibited different expression of IL-6 and different immunomodulatory properties, because IL-6 had context-dependent pro-and anti-inflammatory properties [183]. It has been suggested that the regenerative or antiinflammatory effects of IL-6 were mediated by classic signaling whereas its pro-inflammatory effects were mediated by trans-signaling [184].
IL-6 increased following several different pretreatment methods. On one hand, high expression of IL-6 might play a positive role in the recruitment and migration of monocytes, macrophages and MSCs, and then promote wound healing [32,59,64,102,107,112]; on the other hand, it might have anti-inflammatory effects [50,57,66,71,75,103,[129][130][131]170,185]. However, due to the multiple simultaneously occurring factors after pretreatment, the actual function of IL-6 was not clarified.
Here, we try to focus on its anti-inflammatory mechanisms. IL-6, to a lesser extent, could modulate the in vitro differentiation of macrophages and DCs toward an antiinflammatory and reparative profile [176]. IL-6 could enhance macrophages differentiation into M2 subtype by activating JAK2/STAT3 signaling pathway [186], upregulate IL-4 receptor expression, promote STAT6 phosphorylation of macrophages, and promote M2 macrophages polarization [75,187]. Additionally, IL-6 promoted the production of IL-10 of T cells, which would restrict many inflammatory processes [188].
As a variable component of the secretome, IL-6 may be a keystone cytokine in regulating inflammatory states and deserves more attention.

Problems and Prospects
Earlier studies primarily attributed the therapeutic effect of MSCs to their ability to locally transplant and differentiate into multiple tissues, but the implanted cells are unable to survive for a long period of time [189]. Additionally, the direct application of MSCs is constrained in a few ways, including by the limited source of cells, immunocompatibility, tumorigenicity, emboli formation, phenotype changes in differentiation and proliferation, and problems related to cells' treatment, storage and transportation, as well as the time-consuming process of cell culture, senescence, sensitivity to toxic environments, and heterogeneity in isolation and culture methods between different laboratories [21,25,[190][191][192][193][194][195].
By contrast, using the MSC-derived secretome as a cell-free therapy strategy can help to effectively avoid the abovementioned problems and provides several advantages over MSC-based applications which have described in "Introduction".
Studies have shown that pretreating MSC-derived secretome reveals immunomodulatory potential more efficiently than in untreated MSCs. Compared to engineered methods, pretreating MSCs with drugs, cytokines, hypoxia or 3D culture may be faster and simpler approaches. Additionally, these types of induction techniques more closely the in situ stimulations of inflammatory pathological locations in the host. However, the accompanying issues should be tackled before putting the secretome to clinical use:

1.
Selection of pretreatment methods: it is essential to understand how each pretreatment method affects MSCs' paracrine behavior. Some studies suggest that the combination of different pretreatment methods is superior to one single method, but further studies are still required to confirm this. Meanwhile, pretreatment strategies can be categorized as selective or non-selective: selective approaches, such as engineered methods, target a single pathway or a small number of related pathways to achieve a desired secretome, whereas non-selective strategies, such as hypoxia or inflammatory factors, activate multiple signaling pathways that collectively increase expression of downstream signal factors or receptors.

2.
Heterogeneity: it is extremely difficult to understand the mechanism underlying the reported effects on pretreatment of MSCs due to their tissue origin, the health and age of the donors, cells separation and culture techniques and the animal models.

3.
Purification: the current method for collecting and purifying secretome is centrifugation to remove the cell debris in CM [112], and some studies concentrated the secretome after centrifugation [108,143]. However, additional culture media components may be collected simultaneously and may influence the purity of secretome. Currently, there is no standard purification method of secretome as there is for exosomes. 4.
Components selection: the comparative effect of the specific components (soluble factors or EVs) and the overall application of secretome remains to be verified.

5.
Standardization and optimization: this is problematic because secretome is a combination of various molecules and deserves further study to evaluate its potency and determine a safe dosage.
Collectively, in the subsequent studies, optimal design of pretreating and reliable isolation methods needs to be developed to purify the secretome; cell culture and cell line selection need to be unified. A simple and reliable method of standardizing secretome features is also required.

Conclusions
MSCs exhibit a regulatory phenotype and react quickly to the environmental stimulus that influence their secretome. Pretreatment of MSCs with hypoxia, inflammatory factors, 3D culture, engineering methods and pharmaceutical stimuli, or a combination of the above methods prior to application is a novel strategy to enhance the immunomodulatory effects of MSCs and their secretome in the local or systemic immune response. These models showed that it was possible to improve the regulation of innate and adaptive immune responses. The variable approaches and combinations employed in MSC pretreatment need to be optimized, and new techniques need to be developed to better characterize and standardize the secretome. In the near future, we believe this strategy may be studied further and potentially applied to treat individuals with immune system-related disorders.