Main

The discovery that mesenchymal stem cells (MSCs) contribute to tissue regeneration by modulating inflammation has revolutionized stem cell therapy for the treatment of inflammatory diseases. The mechanisms by which MSCs and inflammation interact during various pathological processes to regulate the immune response are under intense study. Despite the paucity of known MSC functions in vivo, their in vitro properties and potential clinical application are under investigation1. Although most studies focus on the differentiation potential and immunomodulatory ability of MSC populations expanded in vitro, it is generally believed that MSC-based therapy not only provides a source of cells with which to reconstitute a tissue but also regulates inflammation and 'empowers' other cells to facilitate tissue repair. Although cell replacement is an essential component of MSC therapy for some diseases, published investigations have shown that the therapeutic effect of MSCs is mainly a result of immunomodulation and that this function is 'licensed' by inflammation2. Thus, in response to inflammatory mediators, MSCs produce ample amounts of immunoregulatory factors, cell-mobilization factors and growth factors and thereby facilitate tissue repair by tissue-resident stem cells2.

MSCs affect cells of both innate immunity and adaptive immunity1,3, but their immunosuppressive ability is not constitutive; instead, it is induced by inflammatory cytokines, such as those in the inflammatory microenvironment4. The amounts and kinds of inflammatory cytokines vary substantially during the initiation and progression of inflammatory diseases and thus critically influence the activation of immunoregulation by MSCs, thereby determining the immunoregulatory effects of these cells. Such plasticity in MSC immunoregulatory function is the subject of intense investigation. Here we present an overview of the latest findings and their ramifications for preclinical studies and clinical applications.

Stem cells and tissue regeneration

The majority of metazoan tissues are believed to be capable of some degree of regeneration. Lower species such as planarians can regenerate an entire animal from a small piece of tissue, but mammalian tissues have minimal capacity for organ regeneration5. Nevertheless, mammalian tissues contain not only specific unipotent stem cells, such as skeletal muscle stem cells and spermatagonial stem cells, but also multipotent stem cells, such as neural stem cells and hematopoietic stem cells6. One such type of multipotent stem cell, originally identified as the 'colony-forming unit fibroblast', was discovered in bone marrow stroma in the 1960s. These cells were defined by their ability to form new bone when transplanted to an ectopic site7,8. They were later renamed 'mesenchymal stem cells' (MSCs)9. The 'stemness' of MSCs, although loosely defined, is often demonstrated by their potential for tri-lineage differentiation into adipocytes, osteoblasts or chondrocytes. Interestingly, studies have shown that MSCs can also trans-differentiate into cells of ectodermal and endodermal tissues, although the biological relevance of this is unknown10,11. Whether they originate from bone marrow cells or primitive cells in tissues, MSCs have been successfully isolated and expanded in number from diverse tissues, including adipose tissue, skin, tooth pulp and the umbilical cord. Given the lack of clonal analysis in most studies, however, these MSCs should be considered 'mesenchymal stem and progenitor cells'12. The phenotypic profiles of both human and mouse MSCs are that they express CD29, CD51, CD73, CD90 and CD105 but not CD31, CD45 or markers of the hematopoietic lineage13,14. While these cell-surface proteins are routinely used to define MSC populations expanded in vitro, no marker has been identified for specific tracking of MSCs in vivo8. Some markers, including nestin, leptin receptor, Gli1 and FAP, have been used to investigate the function of certain MSC subpopulations in vivo and to study the physiological relevance of the characteristics and functions of these cells after population expansion in vitro15,16,17,18,19. For example, nestin-expressing cells found close to endosteal niches have characteristic MSC functions, as defined by their activity as colony-forming unit fibroblasts in the bone marrow and their potential to differentiate into osteoblasts, chondrocytes or adipocytes, and they constitute a necessary niche component for the support of hematopoietic stem cell function17. Consistent with the function of nestin-expressing MSCs, it has been shown that MSC populations expanded in vitro can support hematopoiesis by acting as critical components of the endosteal hematopoietic stem cell niche17. While the lack of a well-defined specific marker has impeded determination of the physiological function of MSCs, adoptive transfer of MSC populations expanded in vitro has provided some important clues about their migration and engraftment in vivo, especially in disease situations.

To fulfill their roles in tissue regeneration, MSCs must be functionally equipped and properly recruited to the site of tissue damage. Much has been learned about this process. Following intravenous infusion of MSC populations expanded in vitro, despite the large number that become trapped in the lungs, some MSCs subsequently do home to damaged tissue, such as infarcted myocardium, traumatic brain injury, fibrotic liver and chemically damaged lungs, and to various types of tumors20. In experiments tracing MSCs expressing green fluorescent protein–tagged nestin, endogenous MSC-like cells have been observed to migrate from bone marrow to lung tissue following the induction of asthma21. Upon arriving in damaged tissue, MSCs are believed to exert their therapeutic effects in two ways: by cell replacement and by cell 'empowerment'. Many studies have attempted to exploit the potential of MSCs to differentiate and thus replace damaged resident cells, such as endothelial cells, smooth muscle cells, cardiomyocytes or hepatocytes, and thereby promote tissue regeneration in various organs such as the heart, kidneys and liver22,23,24,25. Many preclinical and clinical studies have provided growing evidence of the efficacy of MSC-based treatments. In most cases, however, the rate of MSC engraftment is poor, and engrafted MSCs tend to be short-lived, which indicates that there must be other mechanisms by which MSCs exert their therapeutic effects26,27. Also, in experimental animal models, such as liver cirrhosis, myocardial infarction, renal failure and neural degeneration, the success of MSC therapy does not correlate with the efficiency of cell engraftment and replacement28,29,30,31. Furthermore, inflammatory diseases have been effectively treated with only the culture supernatants of MSCs (the 'MSC secretome') containing growth factors, such as HGF or TSG6 ('tumor-necrosis factor (TNF)-stimulated gene 6')2,30,31. In addition to MSCs of autologous origin, those from allogeneic and xenogeneic sources are also efficacious in treating tissue injuries in immunocompetent people, which indicates that extended engraftment of MSCs is unnecessary for such therapies26,32. Therefore, the therapeutic effects of MSCs may depend largely on the capacity of MSCs to regulate inflammation and tissue homeostasis via an array of immunosuppressive factors, cytokines, growth factors and differentiation factors. These include interleukin 6 (IL-6), transforming growth factor-β (TGF-β), prostaglandin E2, HGF, epidermal growth factor, fibroblast growth factor, platelet-derived growth factor, vascular endothelial growth factor, insulin growth factor, stromal cell–derived factor 1 and, as discussed in more detail below, the tryptophan-catabolic enzyme IDO and nitric oxide (NO), a product of inducible nitric oxide synthase (iNOS)31,33,34. Together these secreted factors may inhibit inflammatory responses, promote endothelial and fibroblast activities, and facilitate the proliferation and differentiation of progenitor cells in tissues in situ. Figure 1 shows a hypothetical model of MSC-mediated tissue regeneration.

Figure 1: Modes of MSC-based therapy: cell replacement versus cell 'empowerment'.
figure 1

When tissue is damaged, inflammation occurs and MSCs are mobilized to the site. Since they have multipotent differentiation potential, the MSCs recruited are believed to differentiate into functional cells to replace damaged cells. However, in response to inflammatory cytokines, MSCs also help prepare the microenvironment by producing immunoregulatory factors that modulate the progression of inflammation. MSCs also produce large amounts of growth factors, which subsequently stimulate endothelial cells, fibroblasts and, most importantly, tissue progenitor cells in situ. The concerted action of these factors and cells facilitates tissue repair through angiogenesis, remodeling of the extracellular matrix (ECM) and the differentiation of tissue progenitor cells.

Full size image

MSCs and inflammation: reciprocity matters

The discovery that bone marrow–derived MSCs might suppress T cell proliferation led investigators to examine their immunomodulatory properties35,36,37. Many studies have demonstrated that MSCs can suppress the activation and function of various cells of the innate and adaptive immune systems, including macrophages, neutrophils, natural killer cells, dendritic cells, T lymphocytes and B lymphocytes. MSCs are now known to act on such cells at several key steps1,3,38,39. Many studies have demonstrated the inhibitory effects of MSCs on T cell proliferation generally, and it has been shown more specifically that MSCs inhibit differentiation into the TH1 and TH17 subsets of helper T cells and promote the generation of regulatory T cells (Treg cells)40. MSCs also induce dendritic cells to acquire a tolerogenic phenotype, which subsequently elicits the generation of Treg cells41. T cell apoptosis that results from inhibition by MSCs has been found to enable macrophages to produce TGF-β, which in turn promotes the induction of Treg cells42. In addition, switching of macrophages from a proinflammatory type 1 to an anti-inflammatory type 2 phenotype is believed to underlie the therapeutic effects of MSCs in treating sepsis43,44. MSCs also downregulate the activation of IL-2- or IL-15-driven natural killer cells45,46. These extraordinary immunomodulatory properties imbue MSCs with great potential for treating various inflammatory disorders. For full realization of this potential, the exact mechanisms by which MSCs mediate immunomodulation during various disease processes must be determined and are now under vigorous investigation.

Several factors and molecules secreted by MSCs have been linked to the immunoregulatory function of these cells. These include TGF-β, NO, IDO, TSG6, prostaglandin E2, IL-1 receptor antagonist, IL-10 and an antagonistic variant of the chemokine CCL2 (refs. 30,38,43,47,48,49,50,51). The broad variety of potential mediators and mechanisms might be a consequence of differences in the tissue types, microenvironments and species from which the MSCs were derived. Indeed, MSCs are benign and cannot suppress immune reactions unless they are first activated by certain combinations of inflammatory cytokines. Thus, MSCs are 'licensed' to exert their immunomodulatory effects after stimulation with interferon-γ (IFN-γ) in the presence of one (or more) other cytokine(s), including TNF, IL-1α or IL-1β. The critical role in this process of IFN-γ and its receptor IFN-γR has been demonstrated in experiments with antibodies to IFN-γ or to IFN-γR, as well as MSCs deficient in IFN-γR14,52. Interestingly, MSCs obtained from a patient with mutant IFN-γR1 (with deletion of Thr523) exhibit immunosuppression, although less intense53. However, dissemination of a typical mycobacteriosis in this patient may also have had a confounding effect on MSC function. Typically, upon stimulation with the appropriate pair of inflammatory cytokines mentioned above, MSCs express large amounts of immunomodulatory mediators, such as IDO and iNOS, as well as chemokines, such as ligands of the chemokine receptors CXCR3 and CCR5, which are critical for the chemotaxis of various T cells4,54. By this mechanism, chemokines recruit T cells into close proximity to MSCs, whereupon the locally active immunosuppressive factors produced by MSCs influence the inflammatory process in reciprocal fashion4. Thus, MSCs can potently influence the inflammatory process. Therefore, these seemingly minor findings have profound ramifications for the proper clinical application of MSCs in disease treatment and for understanding of the pathological processes involved.

A key role for inflammatory cytokine–induced NO in immunosuppression by mouse MSCs has been demonstrated. MSCs derived from several other species showed a clear dichotomy among mammals in the main immunosuppressive mediator produced by 'cytokine-licensed' MSCs. In mammals belonging to the phylogenetic clade Glires (which includes rodents and rabbits), MSCs use NO generated by iNOS. In contrast, in most other mammalian species, including humans, monkeys and pigs, MSCs use IDO55. This species difference probably accounts for much of the variation in molecules linked to MSC-mediated immunosuppression by different studies. Given the importance of this species difference for the relevance of rodent-based studies as models of human disease, a novel humanized mouse model of MSC-mediated immunomodulation has been created: mouse iNOS-deficient (Nos2−/−) MSCs transfected to express human IDO under the control of the mouse Nos2 promoter. These humanized IDO-expressing MSCs exhibit, in the mouse, immunomodulatory effects similar to those of normal human MSCs56 and thus represent a valuable mouse model with which to replicate the human system.

Other factors beyond inflammatory cytokines might also participate in 'licensing' the immunomodulatory ability of MSCs. MSCs can be activated by microbial molecular patterns through Toll-like receptors (TLRs), such as TLR3 and TLR4 (refs. 39,57). MSCs can respond differently to different inflammatory stimuli; MSCs acquire distinct immunophenotypes and activate different signaling pathways that may regulate immune responses differently. Such findings bring new insight to understanding of the crosstalk between MSCs and the inflammatory niche and provide practical information for improving the therapeutic potential of MSCs in the treatment of various diseases, especially conditions related to dysfunction of the immune system.

Plasticity of MSCs in immunomodulation

The immunomodulatory ability of MSCs is dependent upon the kinds and concentrations of inflammatory mediators present in their microenvironment. In fact, different states of inflammation can result in markedly different responses to MSC treatment, which indicates the plasticity of immunomodulation by MSCs. It has been reported that graft-versus-host disease (GvHD) can be successfully treated by MSCs administrated when vigorous inflammation is in progress but is less effectively treated when MSCs are infused on the same day as bone marrow transfusion, before inflammation has begun4,58. Likewise, the therapeutic effects of MSCs on experimental autoimmune encephalomyelitis are diminished when cells are given during disease remission51,59. Thus, the inflammatory condition seems to critically influence the immunosuppressive effect of MSCs. Inflammatory status changes throughout the course of an immune response, being affected by time, activators of the immune system and many other factors. A series of studies on the crosstalk between inflammation and MSC-mediated immunosuppression strongly support the notion that the immunoregulatory function of MSCs is highly plastic. MSCs can both promote an immune response and inhibit it, in accordance with the dynamics of inflammation and depending on the strength of activation of the immune system, the types of inflammatory cytokines present and the effects of immunosuppressants. Essentially, inflammation status determines the immunoregulatory fate of MSCs.

During the pathogenesis of inflammatory diseases, high concentrations of inflammatory cytokines in the acute phase promote disease progression, while low concentrations of these factors in the chronic or relapse phase prepare the microenvironment to facilitate tissue repair (although not always successfully). The plasticity of MSC-mediated immunomodulation in response to fluctuations in inflammation levels was formally demonstrated in a 'chessboard titration' of IFN-γ and TNF to assess how cytokine concentration affects the elicited immunosuppressive function of MSCs60. Interestingly, low concentrations of cytokines were sufficient to upregulate chemokine secretion but were not enough to induce substantial expression of iNOS or IDO. In this scenario, the absence of immunosuppression by NO or IDO allowed the chemokine-recruited lymphocytes to accumulate in the vicinity of MSCs without being suppressed, which led to greater inflammation. In essence, NO and IDO are the 'on-off' switch that controls the immunoregulatory function of MSCs. A similar phenomenon was observed after a low dose of concanavalin A was used to stimulate T cells in coculture with MSCs61. Moreover, antigen-pulsed MSCs stimulated with a low dose of IFN-γ have been found to act as antigen-presenting cells and can thus activate antigen-specific cytotoxic CD8+T cells62,63. Such findings support the hypothesis that while MSCs can be rendered immunosuppressive in the presence of strong inflammation, weak inflammation paradoxically causes MSCs to enhance the immune response.

The inflammatory cytokines and chemokines and participating cells of the immune system change during an inflammatory response. Thus, various cells may respond differently depending on the existing inflammatory status. For example, effector T cells and Treg cells are critical in inflammatory processes. Among the effector T cells, TH1 and TH17 cells are proinflammatory. Their production of IFN-γ, TNF and IL-17 dominates the pathogenesis of various autoimmune and infectious diseases64. During the pathogenesis of these diseases, MSCs are also recruited to the inflammatory sites, where they modulate inflammatory processes and prepare tissues for repair or regeneration. Cytokines present in the inflammatory environment elicit the immunosuppressive function of MSCs4. Among the cytokines linked to the stimulation of MSCs, IFN-γ is essential, and it acts synergistically with TNF. Moreover, the concurrent addition of another prominent inflammatory cytokine, IL-17, further boosts MSC-mediated immunosuppression both in vitro and in vivo, a result of enhancement of the stability of iNOS-encoding mRNA mediated by the RNA-binding protein AUF165. Likewise, IL-1 also acts in synergy with IFN-γ. Therefore, it is apparent that the types and amounts of inflammatory cytokines present in the inflammatory niche dictate how MSCs will regulate inflammation.

Immunosuppressive cytokines such as TGF-β often exist in inflammatory environments, where they are believed to serve a counterbalancing role66. Interestingly, MSCs express receptors I and II for TGF-β, which are believed to modulate the differentiation and regenerative capacities of these cells67,68. Surprisingly, when TGF-β1 or TGF-β2 is provided together with IFN-γ and TNF, the resulting MSCs are less immunosuppressive. The effects of TGF-β are a result of downregulation of the expression of iNOS (or IDO) in MSCs mediated by the signal transducer Smad3 (ref. 69). Ironically, MSCs themselves can produce abundant TGF-β, which probably acts as a feedback loop to partially sustain inflammation, in addition to modulating the regeneration process. Despite that MSC-generated TGF-β, the immunosuppressive ability of MSCs can be inhibited by the addition of IL-10 (ref. 61), which often works together with TGF-β66. Such data reveal the other side of the coin: normally immunosuppressive cytokines can become immune enhancing through their effects on MSCs.

A switch in MSC-mediated immunoregulation has also been demonstrated by ligation of TLRs expressed on MSCs. The activation of TLRs is believed to provide positive feedback to the inflammatory process70. Proinflammatory signals can be delivered to MSCs through TLR4 and can result in differentiation into the MSC1 phenotype, whereas anti-inflammatory signals can be supplied via TLR3 and result in an MSC2 phenotype39,57. These two distinct MSC phenotypes have opposite effects on inflammation. Although the mechanistic network that shapes the activity of MSCs in different inflammatory environments remains to be fully delineated, the plasticity of their immunoregulatory function may explain the disparate findings about their role in immunoregulation.

Since the nature of immunoregulation by MSCs is influenced by many factors in the inflammatory environment and immunosuppressant drugs are often given to treat diseases mediated by the immune system, an important question arises: could immunosuppressants also influence the immunoregulatory function of MSCs? The outcomes of treating patients with immunosuppressants are similar to those produced with MSCs; by suppressing effector T cells, both immunosuppressants and MSCs inhibit the inflammatory response. Studies to determine how they might interact in an inflammatory environment examined the outcome of the administration of MSCs concurrently with the common immunosuppressants cyclosporin A or dexamethasone. Cyclosporin A prevents the activation of T cells and is conventionally used to prevent the rejection of transplanted organs and to treat autoimmune conditions. In experiments, MSCs induce tolerance in mice that had received organ transplants, as expected, but this effect was reversed by cyclosporin A71. Similarly, dexamethasone, another widely used immunosuppressant, also reverts MSC-mediated immunosuppression both in vitro and in vivo72. Such molecular studies have shown that dexamethasone blocks both iNOS expression in mouse MSCs and IDO expression in human MSCs and thus allows inflammation to proceed unchecked72. Therefore, like TGF-β, immunosuppressants can disable the immunosuppressive function of MSCs in an inflammatory environment, a finding with important implications for the proper application of MSCs in therapy.

Several studies of immunomodulation by MSCs have examined their direct effects on T cells by determining how MSCs affect proliferation, apoptosis, differentiation and regulatory mechanisms in T cells, including the induction of Treg cells33. Several groups have reported that MSCs can induce the generation of Treg cells42,73. Since iNOS or IDO and chemokines are key in MSC-mediated immunomodulation, the effects of the catabolites of NO and IDO on the generation Treg cells deserves further investigation, especially in the context of other factors in an inflammatory environment. Strategies targeting NO or IDO and chemokines may be able to regulate the plasticity of MSCs during inflammatory disorders and thus diminish disease progression by inhibiting or enhancing the immune response (Fig. 2).

Figure 2: Plasticity of MSCs in immunomodulation.
figure 2

Proinflammatory cytokines, such as IFN-γ, TNF, IL-1 and IL-17, that are secreted by effector T cells exist in acute or relapsing phases of inflammatory diseases. In response to these cytokines, MSCs produce large amount of NO (in mice) or IDO (in humans) and chemokines, which are key participants in MSC-mediated immunomodulation. The chemokines attract T cells into close proximity to activated MSCs, where high concentrations of the catabolites of NO or IDO suppress the immune response by direct inhibition of T cells. Compared to acute inflammation, chronic inflammation or inflammation during remission is accompanied by suboptimal concentrations of proinflammatory cytokines and high concentrations of anti-inflammatory cytokines, such as TGF-β. In these latter two conditions, as during treatment with immunosuppressants such as dexamethasone or cyclosporin A (CsA), the production of iNOS (or IDO) is below the immunosuppressive threshold. Since chemokines are still expressed, although at lower levels, the recruited T cells are left unchecked and are free to promote the immune response. Essentially, NO (or IDO) is an 'on-off' switch that determines the outcome of immunomodulation by MSCs, and targeting it allows manipulation of the plasticity of MSC-mediated immunomodulation.

Full size image

Since MSCs are widely distributed in almost all tissues and are highly immunosuppressive in the context of inflammation, why do endogenous tissue-resident MSCs not help remedy immunological disorders to the same extent that MSCs provided exogenously do? One possibility is that the immunomodulatory properties of MSCs are acquired during culture in vitro. Another possible explanation is that relatively large size of MSC populations expanded in vitro are often administered; 1 × 106 to 2 × 106 cells per mouse and 1 × 106 to 2 × 106 cells per kg of body weight for humans should be regarded as high doses, especially given the tendency of MSCs to accumulate at sites of damaged tissue. Whichever possibility holds true, the lack of specific markers with which to monitor MSCs in vivo has slowed elucidation of how MSCs respond to different inflammatory conditions. Nevertheless, learning how to control the plasticity of immunomodulation by MSCs, both endogenous and exogenous, may provide an important new modality for better therapeutic application of MSCs, as well as improved understanding of the role of MSCs in different stages of disease.

Immunological plasticity and clinical application of MSCs

Since the first successful clinical application of MSCs to treat severe steroid- and cyclosporine-resistant GvHD in a 9-year-old boy in 2004 (ref. 74), many animal and clinical studies have demonstrated the extraordinary potential of MSCs to ameliorate various debilitating diseases3,20 (Fig. 3). The clinical trial registry of the US National Institutes of Health lists over 390 active MSC-based trials worldwide. Many studies have shown that treatment with MSCs improves the clinical condition of patients suffering from cardiovascular diseases75, liver diseases76, GvHD77 and autoimmune diseases78,79. The crosstalk between MSCs and inflammation is considered to be critical in the therapeutic efficacy of these cells. In fact, it has been suggested that the immunomodulatory ability of MSCs in response to inflammatory cytokines should be used in the standardization of MSC products80. It is conceivable that among the diseases that have been treated with MSCs, the inflammatory tissue microenvironment varies and therefore so too does the influence on MSCs, to the extent that the fate of the MSCs administered differs and results in different outcomes4,60,81.

Figure 3: Timeline for major events in studies of the immunosuppressive effects of MSCs.
figure 3

Purple boxes summarize findings from preclinical studies; orange boxes show results of clinical applications.

Full size image

The plasticity of the immunomodulation by MSCs should be recognized as a consideration in guiding their clinical application and also presents the possibility of manipulating its outcome. A very interesting observation from clinical studies is that MSCs are most effective in treating immunosuppressant-refractory acute inflammatory disease77,78. In fact, the first successful use of MSC therapy has been infusion of third-party haploidentical bone marrow–derived MSCs into a patient with cyclosporine- and steroid-resistant grade IV acute GvHD74. Patients with other severe, non–drug-responsive inflammatory diseases, such as systemic lupus erythematosus and Crohn's disease, have also been successfully treated with MSCs78,82. In such immunosuppressant-resistant patients, persistent inflammation probably primes MSCs for immunosuppression. On the other hand, in immunosuppressant-responsive patients, if MSCs are administered together with steroids or cyclosporine, they have no therapeutic effect. In a clinical trial of Prochymal, the first MSC-based clinical trial approved by the US Food and Drug Administration, treating GvHD with a combination therapy of Prochymal and steroids resulted in no significant clinical improvement83. Such clinical findings support the hypothesis that MSCs require proinflammatory cytokines to induce their immunosuppressive function and that insufficient concentrations of these cytokines can cause MSCs to become immune enhancing. Thus, the outcome of the clinical application of MSC therapy depends on the availability of proinflammatory cytokines. Accordingly, pretreating MSCs with inflammatory cytokines before infusion is proposed as a way to reinforce their immunosuppressive efficacy. Indeed, in animal models of colitis and acute myocardial ischemia injury, such pretreatment has improved the therapeutic effect of MSCs84,85. In addition, the use of MSCs pretreated with IFN-γ results in significantly enhanced survival rates for mice suffering from GvHD86. Similarly, MSCs pretreated with IFN-γ and TNF have better therapeutic effects on concanavalin A–induced mouse hepatitis, especially when IL-17 is included in the pretreatment cytokine 'cocktail', than do MSCs that are not pretreated in the same way65. Such pretreatment strategies may emulate the pathological inflammatory environment present in disease states, and their successful use indicates the importance of inflammation in activating the anti-inflammatory function of MSCs. While these findings are largely limited to preclinical studies, further delineation of the mechanisms that underlie the plasticity of immunomodulation by MSCs will aid the development of improved MSC-based clinical therapies.

A wealth of data indicates that inflammatory conditions determine the biological fate of MSCs. The lack of uniformity in responsiveness to MSC therapy is probably due in part to variability in inflammation, since a minimal amount of inflammation is needed to enable immunosuppression by MSCs. Thus, it is likely that conventional clinical anti-inflammation strategies could alter the inflammatory cytokine profile in the tissue microenvironment and thus modulate the effect of MSCs. In studies investigating the mechanisms underlying failures in MSC-based therapy in the presence of immunosuppressants, dexamethasone inhibits the expression of iNOS and IDO, while chemokine expression was unaffected72. Moreover, in the treatment of advanced liver fibrosis in mice, concurrent administration of steroids reversed MSC-mediated immunosuppression and eliminated their therapeutic effects72. Such data suggest that concomitant application of immunosuppressants with MSCs should be avoided.

Although their pathophysiological roles have yet to be fully elucidated, MSCs are emerging as a novel strategy for treating various diseases, a proposal supported by both animal studies and clinical studies. Interestingly, their effects on most diseases are most probably exerted through their potent immunoregulatory ability. While it is becoming clear that appropriate inflammatory stimulation is needed to elicit the immunosuppressive function of MSCs, further investigation into the underlying molecular mechanisms is needed to allow more effective manipulation of MSC function for clinical applications (Fig. 4). In addition, feasible approaches are needed for monitoring the inflammatory status of patients at the time MSCs are infused to help optimize MSC-based therapy, and relevant biomarkers that indicate potential efficacy must be developed.

Figure 4: Correlation between inflammatory status and efficacy of MSC therapy.
figure 4

The physiological response to tissue damage can be divided into three phases: inflammatory, reparative and remodeling. During this process, inflammatory status (defined as the types and concentrations of cytokines and cells of the immune system present) changes considerably: proinflammatory influences (red dashed line) are dominant in the inflammatory, infection-fighting phase and diminish in the reparative and remodeling phases that follow, which allows wound healing. In the context of the intensity of the immune response (right vertical axis), the inflammatory response (red dashed line) fluctuates during the wound-healing process. Such changes in inflammation substantially alter the effects of MSC-meditated immunomodulation, which results in variable correlation between the intensity of inflammation and efficacy of MSC treatment (solid black line). Only in the presence of strong inflammation (left end of horizontal axis) can efficacious MSC treatment (top of left vertical axis) be achieved. Therefore, it may be possible to optimize the efficacy of MSC therapy by evaluating the inflammatory status of patients to determine the ideal time at which to administer MSCs. In addition, pretreating MSCs with the appropriate inflammatory cytokines before administration or administering drugs to block other cytokines in vivo would also be expected to improve the effectiveness of MSC-based treatments.

Full size image

Of course, there are important concerns about the therapeutic use of these unique cells, such as the origins of the tissues from which MSCs are extracted, the use of autologous or allogeneic MSCs, the optimal passage time in culture before use, patient-selection parameters, clinical 'readouts', ethical issues and governmental regulations. However, these will probably be resolved over time as further study reveals the details of the immunoregulatory effects of MSCs and additional experience in their clinical application is gained.

Perspectives

Investigations in the past few years have provided new insight into the function of MSCs during inflammation and tissue repair. In addition to their well-studied differentiation potentials, the immunoregulatory properties of MSCs are appealing for their ability to reduce inflammation and facilitate tissue repair in treating immunological disorders. Although the relationship between MSCs engrafted long term in damaged tissues and their therapeutic effects has not been formally established, exogenous MSCs indeed home to damaged tissue and integrate into the microenvironment regardless of their duration of residence. The reparative capacity of MSCs depends on strong inflammation, which induces the MSC functions of immunosuppression and growth-factor production. The importance of the plasticity of their immunomodulatory function is underscored by their strikingly different responses to different levels of inflammation and types of cytokines and the presence or absence of immunosuppressants. Understanding the plasticity of MSC-mediated immunoregulation will help to guide the appropriate application of MSCs.

Another important issue is the pathophysiological role of MSCs that reside in normal and inflammatory tissue. Although current data clearly demonstrate the dramatic efficacy of MSCs in halting the progression of various immunological disorders, the physiological roles of tissue-resident MSCs in immunomodulation have yet to be investigated owing to the lack of specific markers with which to monitor MSCs in vivo. Undoubtedly, along with the identification of specific MSC markers, research into the mystery of tissue-resident MSCs will certainly bring new insight into the role of these unique cells in various pathophysiological conditions. Furthermore, it may be possible to exploit the plasticity of the immunoregulatory function of MSCs to tailor clinical treatments for specific inflammatory conditions. This will allow MSC-based clinical protocols to be optimized to achieve appropriate modulation of inflammatory responses at different stages of disease progression.