Management of Fibrosis: The Mesenchymal Stromal Cells Breakthrough

1Healthy tissues can be damaged under various conditions by acute or chronic stimuli such as mechanical or chemical injuries, infections, or autoimmune reactions. In most cases, the repair process consists of dead and damaged cells replacement, thus restoring the organ's unimpaired functionality. The first stage of this mechanism, known as the regenerative phase, corresponds to the replacement of damaged cells by cells of the same type, thus ensuring organ functionality. During the second phase, known as fibroplasia or fibrosis, connective tissue replaces degraded normal parenchymal tissue. Unchecked fibrosis leads to substantial remodeling of the ECM (extracellular matrix) with pathological features which results in the formation of permanent scar tissue. Fibrosis may ultimately lead to organ malfunction and death. It mainly originates from chronic inflammation, tissue ischemia, and imbalance in the ECM accumulation/degradation ratio.

2Most organs are susceptible to fibrotic diseases, generally as a consequence or feature of a preexisting pathology (Figure 1). Obesity, aging, and environmental aggressions are the main causes of fibrogenesis. Fibroproliferative diseases are believed to be responsible for around 45% of deaths in developed countries. Although considerable efforts are being devoted to the search for antifibrotic treatments, there are currently few effective therapies for fibrotic diseases that do not result in severe secondary effects. Anti-inflammatory drugs have been considered as the most promising candidates in clinical trials. A wide range of antioxidants have also been tested. Nevertheless, most drug therapy protocols have failed in achieving sufficient antifibrotic effect.

3Thus, cell therapy has recently been put forward as a possibility. In particular, mesenchymal stromal cell (MSC) therapy seems to be a promising treatment. Indeed, preclinical and clinical trials have shown MSCs' ability to improve outcomes in various diseases such as the consequences of radiotherapy, autoimmune pathologies, neurodegenerative disorders, and other etiological agents. Preclinical and clinical studies have also put forward the ability of MSCs to adapt to their environment. Indeed, the regulation of MSCs' secretome is highly influenced by the surrounding tissue. Therefore, MSC therapy yields different results with different pathologies. Consequently, these effects have led several laboratories to investigate the antifibrotic potential of MSCs.

4Tissue injuries induce damages to resident cells which secrete inflammatory mediators which initiate an antifibrinolytic-coagulation cascade associated with vascular congestion. A temporary ECM is formed to serve as a scaffold for dead cells replacement. Subsequent platelet activation causes the release of various mediators including vasoactive factors (vasodilatation, increased vascular permeability, and edema by plasma exudation), cytokines, and chemokines that enable the recruitment of leukocytes. The formation of a fibrin clot serves as a matrix for cell migration and platelet adhesion. Fibrinolysis is later activated and leads to the dissolution of the fibrin clot replaced by a granulation tissue. Plasmin is released from the fibrin clot and activates the complement system, triggering the release of chemotactic and vasoactive anaphylatoxins.

5Next, recruited leukocytes home by adhesion to molecules such as selectins, integrins, and immunoglobulins. Phagocytosis of tissue debris, dead cells, and any exogenous organisms is carried out by macrophages and neutrophils. They also produce cytokines and chemokines to recruit endothelial cells necessary for neovascularization. The interaction of fibroblasts, fibrocytes, or other resident cells, such as hepatic stellate cells (HSCs), with the microenvironment induces their differentiation into myofibroblasts which synthesize ECM and growth factors including profibrotic TGF-β1 (transforming growth factor-β1). The secretion of autocrine hormones enables the maturation of myofibroblasts. α-SMA (α-smooth muscle actin) and vimentin expression by myofibroblasts are responsible for their contractile activity. This contractibility is required for the closure of the wound. The formation of this so-called granulation tissue is characterized by the presence of many blood capillaries allowing the supply of nutrients, hormones, and respiratory gas.

6Finally, the migration and maturation of epithelial and endothelial cells then allow the formation of scar tissue and neovascularization. The provisional ECM is degraded by matrix metalloproteinases (MMPs) once complete tissue replacement is achieved. The subtle equilibrium between MMPs and their inhibitors, tissue inhibitors of metalloproteinases (TIMPs), controls ECM accumulation and degradation throughout the repair process. Thus, it guarantees proper ECM remodeling by inducing a shift in matrix composition. Next, myofibroblasts disappear by apoptosis, triggered by the establishment of a negative activation loop indicating regeneration of the injured tissue.

7Various fibroproliferative pathologies share common features. Fibrosis begins as a normal tissue regeneration process. Resident and recruited cells are activated to produce a provisional ECM facilitating repair. However, in the case of bacterial infection, ischemia, chronic inflammation, or other persistent stimuli, a constant loop of myofibroblast activation sets in, leading to excessive ECM accumulation. Activated myofibroblasts also produce chemokines to recruit cells from the immune system (macrophages, T- and B-cells, neutrophils, and eosinophils), thus perpetuating chronic inflammation. The pathologic matrix progressively invades the tissue, eventually ending in the presence of a permanent fibrotic scar. Histologically, fibrosis can be defined by two distinct stages. Development corresponds to the onset of matrix accumulation where only scattered fibrosis areas are seen in the tissue, whereas the endpoint is characterized by diffused spans of ECM possibly distributed through the entire tissue. The progressive replacement of dead cells by ECM suppresses organ function and induces stiffness. Ultimately, the best course of treatment for advanced fibrosis is often organ transplantation.

8Fibrosis is a complex pathology driven by numerous biological factors such as chronic inflammation and hypoxia. Ionizing radiation, for example, induces endothelial cell death and oxidative stress, resulting in prolonged inflammation and potentially fibrosis. The constant recruitment of inflammatory cells generates an activation loop of myofibroblasts and maintains a steady pool of profibrotic cells.

9One of the main molecular agents inducing fibrosis is TGF-β1, mainly synthesized by T-cells during the healing process. TGF-β1 is secreted in a latent form associated with LAP (latency associated peptide). LAP is cleaved to allow the activation of TGF-β1 which is able to bind its receptors TGF-βR1 (transforming growth factor receptor-β1) and TGF-βR2. Therefore, there is a large pool of inactive TGF-β1 in the extracellular environment. Various agents can induce TGF-β1 activation: MMPs, reactive oxygen and nitrogen species (ROS and RNS), cytokines, or other stimuli such as ionizing radiation. The binding of TGF-β1 to its receptors activates the Smad (small mothers against decapentaplegic homolog) signaling pathway which induces the transcription of various genes, including genes encoding members of the extracellular matrix (collagens mostly). It also activates the differentiation of fibrocytes toward functional fibroblasts.

10EMT (epithelial-to-mesenchymal transition) and EndMT (endothelial-to-mesenchymal transition) are also described as important sources of fibroblasts. Epithelial or endothelial cells assume a spindle shape, lose their cell markers, and express typical fibroblast markers such as FSP-1 (fibroblast specific protein-1), α-SMA, and vimentin. They also acquire the ability to produce collagen and fibronectin (extracellular matrix components). TGF-β1 has also been shown to decrease the expression and activity of MMPs and increase the expression of TIMPs. Thus, TGF-β1 is considered to be one of the major factors in fibrosis development.

11Other growth factors take part in prolonged fibrogenesis. CTGF (connective tissue growth factor) acts synergistically with TGF-β1 to stimulate the signal transduction pathway dependent on TGF-β1. CTGF can also stimulate the proliferation, migration, and adhesion of fibroblasts and the production of the extracellular matrix.

12Thus, fibrosis is a multicomponent pathology driven by multiple factors (Figure 2). One of the main issues in treating fibrosis lies in its self-maintenance. Hence, various therapies might be considered depending on the stage of fibrogenesis. Indeed, preventive or curative strategies should differ based on the ECM components and the mechanisms involved. Moreover, combined therapies should be used to simultaneously act on various profibrotic mechanisms and enhance treatment efficacy.

13Over the years, many models of fibrosis in animals have been developed. Mechanical or chemical procedures are used to mimic damage observed in patients.

14Heart. Cardiac fibrosis is characteristic of many heart diseases. Doxorubicin (DOX) or isoproterenol (ISO) is widely used to induce myocardial infarction (MI). It is hypothesized that DOX-induced cardiac damage increases the concentration of reactive oxygen species, thus causing injury to mitochondria, leading to apoptosis and fibrosis. ISO injection directly into the heart produces diffuse myocardial cell death and fibrosis, leading to progressive heart failure. Finally, ligation of the interventricular artery results in ischemia and eventually leads to fibrosis.

15Kidney. Interstitial fibrosis and glomerulosclerosis are common features of kidney pathologies such as chronic kidney disease (CKD), chronic allograft nephropathy (CAN), or ureteral obstruction. In the reversible unilateral ureteral obstruction (UUO), fibrosis is induced by oxidative stress. Atherosclerotic renal artery stenosis (ARAS) is found among 50% of atherosclerotic patients with other atherosclerotic diseases. In preclinical studies, ARAS is modeled by placing an irritant coil in one of the main renal arteries to induce chronic inflammation. Removal of one or both kidneys and kidney allograft can be performed to create a CAN model. “Nephrectomy + ischemia-reperfusion + cyclosporine” (NIRC) is a recent model mimicking CKD. Oxidative stress caused by ischemia, exacerbated by the immunosuppressive effect of cyclosporine, induces interstitial fibrosis following ischemia-reperfusion. Lastly, in the remnant kidney model (RKM), also called 5/6 nephrectomy (5/6 NX), interstitial fibrosis is induced by removing one kidney and two-thirds of the second. It is hypothesized that subsequent oxidative stress and inflammatory reaction generate fibrosis.

16Liver. Fibrosis in the liver, or cirrhosis, is the common endpoint of chronic liver diseases. It originates from not only numerous pathologies such as alcoholic liver disease and viral or autoimmune hepatitis but also hepatotoxic drugs and toxins. Carbon tetrachloride (CCl4) induces irreversible pathologies such as fatty liver, fibrosis, cirrhosis, and cancer and is mainly used in liver damage models.

17Lungs. Pulmonary fibrosis is an increasingly frequent pathology due to the growing number of smokers and the pollution resulting from current lifestyles. The onset of fibrosis in the bleomycin, mainly originating from DNA single and double strand breaks, is a major side effect of this drug which is now widely used in the development of animal models of pulmonary fibrosis. Exposure to silica also induces fibrotic responses. The resulting persistent toxic effect causes chronic inflammation resulting in fibrogenesis.

18Peritoneum. Peritoneal fibrosis can be initiated by toxins, infectious peritonitis, or incompatible dialysate products. Chlorhexidine gluconate (CG) was one of the first compounds believed to cause encapsulating peritoneal sclerosis (EPS) during dialysis. Peritoneal exposure to CG leads to an inflammatory reaction causing fibrosis in animal models.

19Skin. Skin fibrosis is part of a wide range of human disorders including keloids, hypertrophic scars, and scleroderma. Subcutaneous injections of bleomycin produce lesions mimicking scleroderma. Radiation exposure can lead to fibrosis in a number of different organs. Cutaneous radiation-induced fibrosis is caused by a strong inflammatory reaction, apoptosis, and oxidative stress and is a commonly used animal model. Another in vivo cutaneous fibrosis model has been developed in mice by producing full-thickness wounds which consequently lead to chronic inflammation.

20Pancreas. The incidence of chronic pancreatitis is approximately 30 per 100,000 and is increasing over time. Since existing treatments are limited, continuous efforts are being devoted to preclinical studies in animal models. Intravenous administration of dibutyltin dichloride (DBTC) induces damage to the bile duct epithelium. Subsequent inflammation causes fibrosis in the pancreas.

21Colon-Rectum. 5 to 10% of patients receiving pelvic radiotherapy develop chronic radiation proctopathy due to the high radiosensibility of organs in the radiation field (colon, rectum, and bladder). Radiation proctopathy is modeled in animals by delivering a high radiation dose to the rectum. Radiation-induced damage to the tissue as well as oxidative stress induces fibrosis in this model.

22Common features are characteristic of these animal models of fibrosis. Chemical compounds, physical agents, or surgery procedures are used to induce the initial injury. This protocol is often repeated periodically or maintained over a prolonged time. Subsequent damage to the tissue induce chronic inflammation, oxidative stress, and/or hypoxia necessary to activate resident and recruited cells toward a profibrotic phenotype. In most cases, fibrotic features appear weeks to months after the initial stimulus.

23MSCs are widely described for their immunoregulatory properties. Nevertheless MSCs' antifibrotic functions are poorly described. Syntheses of in vivo study outcomes are described in Table 1 (heart), Table 2 (liver), Table 3 (kidneys), Table 4 (lungs), Table 5 (peritoneum), Table 6 (pancreas), Table 7 (skin), and Table 8 (rectum). The synthesis of in vitro study outcomes is shown in Table 9.

24Pathogenic fibrosis results from chronic inflammatory reactions. Recent advances in the immunobiology of MSCs have led to increased interest in their use as a new therapeutic modality to address chronic inflammation associated with fibrosis (Figure 3). The immunosuppressive effect of MSCs has been extensively studied and documented, particularly because of its value in organ transplantation. MSCs operate on the T and B lymphocytes by blocking them in the G0/G1 phase of the cell cycle, inhibiting the production of immunoglobulins (IgA, IgG, and IgM) and the differentiation of B lymphocytes. MSCs induce a change in polarity in T lymphocytes from a proinflammatory Th1 state to an anti-inflammatory Th2 condition. They act in the differentiation and maturation of dendritic cells and make them tolerogenic. Furthermore, MSCs inhibit the cytotoxic activity of natural killer cells on HLA-1 (human leukocyte antigen-1) negative cells and reduce the production of cytokines: TNF-α (tumor necrosis factor-α), IFN-γ (interferon-γ), and IL-10 (interleukin-10). Therefore, MSCs are of value for the treatment of diseases with an inflammatory component.

25Numerous studies have highlighted the benefits of immunomodulation by MSCs in the treatment of fibrosis. MSC-induced decreased TLR (toll-like receptor) expression suggests their ability to limit chronic inflammation. After the transplantation of MSCs, a decreased infiltration of monocytes/macrophages, neutrophils, and lymphocytes in the tissue was observed in various models. This correlates with the decreased expression of MCP-1 (monocyte chemoattractant protein-1) in some cases. Additionally, underexpression of VCAM-1 (vascular cell adhesion molecule-1) and ICAM-1 (intercellular adhesion molecule-1), involved in leukocyte-endothelial cell interactions, suggests reduced inflammatory cell infiltration. In a model of radiation-induced skin fibrosis, MSCs induced macrophage transition toward a regulatory phenotype, thus limiting chronic inflammation causing fibrosis. Decreased iNOS (nitric oxide synthase) expression after MSC transplantation suggests a reduction of M1 macrophage activity. An increased proportion of anti-inflammatory M2 macrophages were reported after MSC transplantation in a heart fibrosis model and a radiation-induced proctitis model. Microvesicles purified from MSC-conditioned medium, while significantly decreasing the amount of inflammatory cells, produced lower effects compared to MSC transplantation in the lung.

26MSCs inhibit the expression of IFN-γ, which exerts a proinflammatory effect by inducing overexpression of IL-6 and TNF-α. The decrease in mRNA expression and protein concentration of TNF-α, a profibrotic cytokine, was detected in the tissue after MSC transplantation. IL-1α, IL-1β, and IL-6 are underexpressed in several fibrosis models following MSC injection. Increased expression of anti-inflammatory cytokines IL-4 and IL-10 after MSC transplantation was observed, suggesting the transition of T lymphocytes to a Th2 profile. Similarly, MSCs induced increased IL-10 expression and concentration in a model of cutaneous and rectal radiation-induced fibrosis.

27Antiapoptotic effects of MSC therapy can also be discussed, as fewer apoptotic events correlate with reduced inflammation. In fibrotic tissues following MSC transplantation, a decrease in apoptotic events was observed. Accordingly, MSCs may protect resident cells, increasing functionality and recovery.

28MSCs may induce regression in pathophysiological processes associated with fibrosis. These effects are in part mediated by a reduction in chronic inflammation. MSCs likely proceed by a change in immune cell function, an increase in anti-inflammatory cytokines, and a decrease in proinflammatory cytokines and cell apoptosis. These immune mechanisms contribute to a modification of the microenvironment, thus diminishing tissue fibrosis, increasing resident stem cell proliferation, and eventually leading to tissue regeneration.

29TGF-β1 has been described as one of the major players in fibrosis. Its binding to receptors induces the activation of a signaling cascade leading to the proliferation of phenotypically profibrotic cells such as myofibroblasts. In particular, it induces the EMT and EndMT in part responsible for the proliferation of cells synthesizing ECM. The TGF-β1 signaling pathway is one of the prime targets for antifibrotic therapies and its regulation has been abundantly studied in treatment trials with MSCs. Generally, MSC transplantation reduces the expression and concentration of TGF-β1. The same effect is induced by transplanting exosomes isolated from MSC-conditioned medium. In vitro, Ueno et al. showed the inhibition of TGF-β1 overexpression induced by glucose in a coculture model of MSCs and peritoneal mesothelial cells. This effect was associated with the decrease in the phosphorylation of Smad-2, as also shown in an exosome transplantation model. Reduced expression of α-SMA and the lower number of α-SMA positive cells suggest a decrease in the proliferation of myofibroblasts and, to a lesser extent, of TGF-β1-mediated EMT. In vitro, a reduced concentration of α-SMA in a coculture of MSCs and HK2 (human kidney 2) cells pretreated with TGF-β1 suggests a direct effect by MSCs on phenotypic changes leading to the accumulation of profibrotic cells. A decreased expression and concentration of CTGF in several models also participate in diminishing profibrotic cells proliferation.

30Interestingly, several studies have underlined the importance of HGF (hepatocyte growth factor) secreted by MSCs for their antifibrotic effects. MSCs transfected with an HGF expression plasmid yielded better results than nontransfected MSCs in a pulmonary fibrosis model. The use of recombinant HGF partially reproduced the effects of MSCs in a coculture model with albumin-treated proximal tubular epithelial cells (PTECs). The inhibition of TGF-β1 expression by HGF and its ability to ameliorate the degradation of collagen through the increase in MMP-1 concentration highlights the value of such therapy. Moreover, the increased expression of p-Met, which induces the phosphorylation of c-Met, the HGF membrane receptor, is also part of the action mechanisms of MSCs.

31Recently, Qi et al. highlighted the importance of TSG-6 (TNF-stimulated gene 6) in the antifibrotic effect of MSCs. In addition to suppressing the secretion of TNF-α by activated macrophages, this protein induces a change in the TGF-β1/TGF-β3 balance, from a profibrotic high ratio to an antifibrotic low ratio. These results are confirmed in a coculture model in which recombinant TSG-6 partially reproduced the effects of MSCs.

32Accumulation of ECM in the tissue, death of endothelial cells, and increased levels of reactive oxygen and nitrogen species (ROS and RNS, resp.) result in hypoxia and oxidative stress during fibrosis. These factors lead to increased apoptosis and activation of TGF-β1. The improved vascularization of tissue and a more effective neutralization of oxidizing radicals would therefore enhance the effectiveness of antifibrotic therapies.

33MSCs' ability to relieve oxidative stress has already been shown in several works. First, they seem to increase the expression and concentration of enzymes responsible for scavenging free radicals, such as NQO1 (NADPH quinone oxidoreductase 1), Gr (glutathione reductase), GPx (glutathione peroxidase), and HO-1 (heme oxygenase 1). Nrf2 (nuclear factor (erythroid-derived 2)-like 2) activation is protective against oxidative stress and induces SOD (superoxide dismutase) production which decreases ROS concentration in the liver. MSC treatment correlates with an increase in Nrf2 and SOD which might reduce ROS accumulation, thus decreasing oxidative stress. In a coculture model, an increased survival of cerebellar neurons is correlated with the secretion of SOD3 by MSCs.

34MSC-mediated angiogenesis has also been demonstrated. MSCs are able to secrete a large range of angiogenic factors such as VEGF (vascular endothelial growth factor), FGF-2 (fibroblast growth factor-2), and MCP-1. Some studies also suggest their ability to promote endothelial cell proliferation. The reduced expression of VEGF, associated with improved microcirculation in the tissue after MSC transplantation, was observed. Mias et al. showed a stimulation of angiogenesis following treatment with MSCs. The transplantation of MSC sheets into the scarred myocardium increased neovascularization in a myocardial infarction model. The authors also reported evidences of MSCs differentiating to participate in the formation of new vascular structures.

35Conversely, an increased expression of VEGF posttreatment, with the concomitant overexpression of HIF-1α, was shown in a renal fibrosis model, indicating elevated tissue hypoxia. HIF-1α (hypoxia-inducible factor-1α) stimulates the expression of VEGF under hypoxic conditions. In the same way, in a radiation-induced proctitis model, the overexpression of VEGF was accompanied by a reduction in angiopoietin and PDGF expression. It can be hypothesized that insufficient angiogenesis in these models induces these variations. This gene expression profile may reflect proangiogenic signals mediated by MSCs. The evaluation of tissue vascularization would give better insights into MSCs effect on angiogenesis in these models.

36MSCs may therefore act in different ways on hypoxia and oxidative stress by increasing angiogenesis in the tissue and by improving the inactivation of ROS and RNS. This feature, contributing to the inhibition of LAP cleavage from TGF-β1 and reduction of apoptosis, could contribute to MSCs' antifibrotic effects.

37Excess production of ECM and the failure to degrade it are the hallmark of fibrosis. Thus, the ultimate goal in case of fibrotic diseases is to restore a nonpathological healing process, by inhibiting ECM production and enabling the degradation of its various components. Indeed, the imbalance of MMPs, responsible for the degradation of ECM, and TIMPs, their inhibitor, results in improper ECM remodeling, hence preventing restoration to a nonpathological matrix.

38In different fibrosis models, a decreased expression and concentration of collagen, the main component of the ECM, were found after MSC transplantation. This effect is also obtained after transplanting microvesicles or exosomes secreted into an MSC culture, suggesting a paracrine control of MSCs on ECM degradation.

39Changes in the expression and concentration of MMPs and TIMPs have also been studied. After MSC transplantation, the increased expression of MMP-2, MMP-9, MMP-13, and MMP-14 has been observed in several fibrosis models. Following the addition of MSC-conditioned culture medium to a culture of heart fibroblasts, an increase in the activity of MMP-2 and MMP-9 was found. Conversely, several studies have shown reduced expression, concentration, or activity of MMPs. Accordingly, Alfrano et al. noted the decreased activity of MMP-2 after transplantation in the NIRC model. In some fibrosis models, MMP-2, MMP-9, and MMP-13 have a lower expression and concentration following treatment with MSCs. However, these variations suggest restoration to levels similar to untreated controls.

40MSCs seem to have a repressive effect on the expression of TIMPs such as TIMP-1. A reduction in the concentration of TIMP 1 to 4 was shown after MSC transplantation. In an in vitro model, a decrease in the expression of TIMP-2 was observed, suggesting that MSCs have a paracrine effect. Finally, Linard et al. demonstrated a tendency toward the resolution of fibrosis by calculating the collagen-to-MMP-to-TIMP ratio, a marker of fibrosis evolution.

41MMP and TIMP expression are impaired in fibrotic pathologies. In fact, lower TIMP expression is generally associated with fibrosis resolution. In cases of heart failure, an increased expression of MMPs has been observed in the initial and final phase. It has been shown that increased MMP-2 activity is associated with pathological ECM remodeling in the kidney. Thus, decreased activity following MSC therapy suggests a transition to a nonpathological state. On the contrary, it has been shown that MMP-2 is implicated in alveolar regeneration, which could explain its increased activity after transplantation in a pulmonary fibrosis model. Finally, as certain MMPs activate latent TGF-β1, a decrease in their concentration would result in a lesser activation of downstream effectors. Taken together with a decreased fibrotic area and ECM component (collagen, fibronectin, etc.) expression, these results indicate a change in ECM composition, close to that observed in nonpathological animals. Hence, MSCs seem to improve ECM quality, allowing the appearance of a microenvironment favorable to tissue regeneration.

42Various transplantation conditions have been assessed in the studies reported in this work including MSC activation and the optimization of MSC delivery. First, melatonin has been shown to improve MSC survival after transplantation, as well as having proangiogenic abilities. In both occurrences of this treatment, melatonin-treated MSCs exerted increased beneficial effects compared to nontreated cells, as evidenced by reduced ECM deposit and inflammation. Qiao et al. showed potentiation in predifferentiated MSCs treated with baicalin, which possesses anti-inflammatory and antioxidant properties. Cotreatment with atorvastatin increased the survival and efficacy of MSCs.

43Multiple transplantation timings have been compared to investigate their respective effect. Alfarano et al. showed that transplantation 7 days after ischemia-reperfusion was more effective on ECM deposition, myofibroblast proliferation, and MMP activity in their model compared to transplantation after 14 days. In bleomycin-induced lung fibrosis, Ortiz et al. also observed greater effectiveness from MSCs when transplanted earlier.

44Interestingly, Ishikane et al. demonstrated that the transplantation of fetal membrane or bone-marrow-derived MSCs yielded similar results on myocardial infarction.

45In two different studies, the value of MSC differentiation before transplantation was observed. In the rat model of CCl4-induced fibrosis, opposite effects were reported. Hardjo et al. showed a higher potential for nondifferentiated MSCs, compared to adipogenic and hepatogenic differentiation, on ECM accumulation and MMP expression. Conversely, in the exact same model, Qiao et al. found that hepatogenic predifferentiation had no significant influence on the effect of MSCs.

46Recently, new delivery procedures have been studied to improve MSCs engraftment in fibrotic tissues. MSCs grown in two-layered sheets and transplanted in a rat model of myocardial infarction were found in significant number 28 days after transplantation. Part of these cells showed evidences of differentiation, participating in neovascularization of the infarct. Indeed, MSC homing in the damaged tissue is generally transient, which could explain the decreased long-term benefit often observed. Embedding MSCs in scaffolds or biomaterials could improve their beneficial effects.

47In clinical settings, the transplantation of MSCs has been studied on numerous pathologies. A systematic review of clinical trials evaluated the safety of MSC injections. Thirty-six studies were included representing 1012 patients. The meta-analysis did not reflect any serious complications related to MSC injections. Only a transient fever was highlighted (reviewed in). Around 30 clinical trials are currently registered worldwide for evaluating MSC therapy for fibrosis (http://clinicaltrials.gov). Liver and pulmonary fibrosis are most widely represented, but some occurrences of renal and vocal fold treatment exist. MSCs engraft preferentially in the lungs and liver which is the reason for a higher number of clinical trials on these organs. In most of these studies, only organ functionality is evaluated but not fibrosis markers. Thus, it is not clear whether the improvement of the symptoms and quality of life is due to fibrosis reduction or the amelioration of other pathological features.

48Bone-marrow-derived MSCs improve liver function in patients with liver cirrhosis as evidenced by phase I clinical trials. The Model for End-stage Liver Disease (MELD) score is used to evaluate the mortality risk in patients with end-stage liver disease (reviewed in). The mean MELD score is significantly lower after MSC injection compared to placebo controls. In patients with decompensated liver fibrosis, MSCs significantly improved quality of life as evidenced by the increase in physical and mental component scales and through the SF-36 questionnaire. Inducing hepatic differentiation prior to MSC injection improved liver function in treated patients. Finally, fibrosis markers were measured on 30 patients during a phase I trial. Laminin, hyaluronic acid, and type IV collagen were significantly decreased 48 weeks after intervention. On the other hand, HGF, an antifibrotic growth factor, was increased after 48 weeks, as compared to nontreated patients. Based on these clinical trials, it appears that MSCs may exert an antifibrotic effect on liver cirrhosis.

49The results of a phase I study show the ability of MSCs to reduce allograft rejection after renal transplantation. MSCs decreased graft rejection by exerting immunosuppression and probably by preventing interstitial fibrosis. The absence of a placebo control in this trial did not permit the comparison and identification of the specific effect of MSCs. Thus, it is necessary to gather additional clinical data.

50MSC therapy has proven to be effective in patients suffering from complications following acute myocardial infarction. In the first trial, functional testing showed an improvement in both heart and lung functions. There was evidence that MSC treatment led to reverse remodeling, which could be correlated with fibrosis reduction. Six months after treatment, global symptom scores were significantly better in the MSC group versus the placebo group. In the second study, MSC treatment reduced symptoms associated with ischemic cardiomyopathy. There was also evidence of reverse remodeling concomitant with infarct size reduction, probably linked to reduced fibrosis.

51Pelvic radiation disease (PRD) is induced in 5 to 10% of patients within 10 years after abdominopelvic radiotherapy. Fibrosis to the colon and rectum is the main characteristic of late complications of radiotherapy. Since no satisfactory treatment exists for PRD and given the results of MSC therapy on radiation-induced burns, the curative potential of MSCs is being evaluated in clinical trials for PRD treatment. In particular, 4 patients suffering from serious intestinal radiation-induced lesions following overdosage of radiotherapy have been treated. The systemic administration of MSCs resulted in efficient analgesic and anti-inflammatory effects as well as hemorrhage reduction. These results indicate the potential of MSC to diminish the adverse effect of radiotherapy and possibly radiation-induced fibrosis.

52Based on these clinical trials, MSC therapy has proven to be safe and effective in patients suffering from diseases associated with fibrosis without the adverse effect of MSC transplantation. Nevertheless, there is a need for randomized trials (phase 3) to gather statistically significant data and to demonstrate MSCs' efficacy in limiting fibrosis.

53The future of MSC therapy for fibrotic diseases mostly relies on a comparison with current management strategies. Results from preclinical and clinical trials highlight the ability of MSCs to act on fibrosis through different mechanisms: (i) immunosuppression, (ii) inhibition of the TGF-β1 pathway, (iii) reduction of hypoxia and oxidative stress, and (iv) restoration of ECM degradation. Thus, the potential of MSC therapy lies in the ability to act simultaneously on various fibrogenesis parameters. There are currently several therapy protocols for fibrotic therapies under assessment in clinical trials. Most of those treatments are designed to act on a single pathway underlying fibrosis development and progression, unlike cell therapy.

54Presently, therapy protocols for fibroproliferative diseases mostly consist of symptomatic treatments. For example, patients with idiopathic pulmonary fibrosis (IPF) are often prescribed oxygen therapy and vaccination against viral and bacterial infections of the airways is recommended, if any exists. Likewise, antifibrotic strategies in the liver are most effective when they are able to cure the underlying disease. Many anti-inflammatories and antioxidants have been unsuccessful candidates for fibrosis treatment. Ultimately, organ transplantation is required to ensure the survival of patients with fibrosis.

55The first example of clinically used pharmacological antifibrotic agent is pirfenidone, which acts on TGF-β1 activity and inflammation and which has antioxidative properties. It has been approved for the treatment of IPF in Europe, Canada, South Korea, and Japan. Preclinical studies have shown its ability to suppress TGF-β1 gene expression and to significantly reduce its concentration in lavage fluid in models of pulmonary fibrosis. Pirfenidone is also effective in animal models of heart, kidney, liver, and radiation-induced fibrosis. The FDA has not yet approved pirfenidone for pulmonary fibrosis based on a lack of efficacy and survival benefit, especially in long-term clinical trials. Moreover, a meta-analysis of clinical trial results shows that pirfenidone induces adverse gastrointestinal, neurological, and dermatological adverse effects.

56Other antifibrotic drugs are currently being examined for clinical use (reviewed in). Those pharmacological agents are mainly anti-inflammatory drugs and inhibitors of the TGF-β1 signaling pathway acting on different molecular targets. Despite the fact that some of these drugs have been evidenced to exert antifibrotic effects in animal models, there is a lack of clinical data that may lead to their approval.

57Although some pharmacological compounds have proven to be effective, the necessity to use multiple drugs for the treatment of fibrosis is increasingly recognized. Furthermore, MSCs specifically home to damaged tissues and are able to behave depending on the surrounding environment, delivering transiently and locally specific molecules necessary for restoring tissue homeostasis. Conversely, drugs affect every organ, regardless of its pathological state. There is a need for more clinical data on MSC therapy to ascertain its effectiveness and safety. However, while inducing minor side effects, MSCs have shown promising antifibrotic effects, regardless of the organ, and should be considered as a major candidate.

58Altogether, the objective analysis of the literature supports the antifibrotic effect of MSCs. It is sometimes argued that MSCs could have profibrotic properties because they are likely to acquire a myofibroblastic phenotype in vitro or that the mesenchymal origin of myofibroblasts indicates profibrotic properties. Nevertheless, there is, to our knowledge, no example showing MSC transplantation to have a profibrotic effect on a developing or established disease.

59Since fibrosis is a very complex multicomponent process, it can be hypothesized that MSCs act through different secreted factors on multiple pathways (Figure 4). This assumption is supported by the fact that the role of transplanted MSC depends mainly on the surrounding environment.

60The principal mode of action of MSCs may be exerted mainly through inhibition of the TGF-β1 signaling pathway, mainly by blocking the passage from its latent form to its active form. The reduction of inflammation in the tissue, the improvement of angiogenesis, and the reduced oxidative stress seem to be responsible for this effect. The decrease in the concentration of activated TGF-β1 would lead to reduced EMT and myofibroblast proliferation, consequently shifting the balance between synthesis and degradation of the ECM. Furthermore, results suggest that MSCs possess the ability to inhibit TGF-β1 mRNA as well as protein synthesis. Thus, they would act on two different levels, preventing injury-triggered TGF-β1 overexpression and modifying the surrounding microenvironment to lessen the concentration of TGF-β1-activating factors.

61Another interesting and extensively studied feature of MSC therapy against fibrotic diseases is their immunomodulatory ability. In numerous studies reported here, MSCs seem to reduce immune cell homing in the damaged tissue. This could in part explain the decrease in proinflammatory cytokines mRNA expression and production. Most notably, TNF-α and IFN-γ, two major profibrotic cytokines, were underexpressed in several studies reported here. These observations are consistent with the implementation of an antifibrotic “virtuous circle” in which fewer immune cells migrate to damaged tissues, hence reducing proinflammatory cytokines production. By inhibiting the acute inflammatory reaction, it is conceivable that MSCs reduce the consequent chronic inflammation.

62Reduced hypoxia and oxidative stress are also an important effect of MSCs in this context. In fact, high ROS and RNS concentrations, combined with low oxygen intake, further increase TGF-β1 activation. It also induces apoptosis in resident cells, resulting in increasingly elevated inflammation. The ability of MSCs to improve the neutralization of free radicals, already described in other models, is supplemented by indications of improved angiogenesis. The resulting improvement in tissue vasculature reduces ischemia, allowing better regeneration of the injured organ.

63As expected, the inhibition of the TGF-β1 signaling pathway induces a substantial remodeling of the ECM toward a nonpathological state. The decreased expression and concentration of ECM components, associated with the restoration of the MMP/TIMP balance, improve the quality of the connective tissue. This can mostly be explained by a lower profibrotic cell population (myofibroblasts mainly). This allows for better homing of the cell types necessary for regeneration of the damaged tissue, suggesting the possibility of reversing fibrosis under the influence of MSCs.

64MSCs seem to have a paracrine effect highlighted by the results obtained in studies using MSC-conditioned medium. Several factors have been put forward as mediating this effect. First, HGF, an antifibrotic mediator which also has antiapoptotic properties, should be mentioned. MSC therapy combined with antibodies against HGF greatly reduces the effects of the treatment and recombinant HGF administration partially reproduces the effects of MSCs. The treatment of fibrosis by HGF has already been assessed in earlier studies and has shown great potential. Moreover, TSG-6, a recently discovered protein highlighted for its immunosuppressant effect, seems to play a major role in the antifibrotic action of MSCs. The use of antibodies or gene silencing methods significantly reduces MSCs' ability to alleviate fibrosis. Indeed, TSG-6 has been demonstrated to inhibit the secretion of TNF-α by macrophages and to alter the TGF-β1/TGF-β3 balance toward an antifibrotic ratio.

65The study of MSC transplantation conditions also needs extensive investigation. Data reported in this paper indicate that the pretreatment of MSCs to potentiate their effect may yield better outcomes. Equally, transplantation timing after injury is of great importance. In fact, results suggest that earlier therapies improve the efficacy of MSCs on fibrosis. This is to be expected, as inhibition of the acute inflammatory reaction by MSCs would prevent the onset of chronic inflammation. MSC source is also an important factor to be considered. It has been reported in this paper that fetal membrane and bone-marrow-derived MSCs were equally effective. The comparison between different sources of MSCs is an important matter considering the fact that some tissues, such as adipose tissue, are easier to harvest and/or contain higher numbers of stem cells. Moreover, the value of predifferentiation is to be further investigated as contrary data have been gathered. In any case, supplementary studies need to be conducted to confirm these effects. Finally, although preclinical data suggest the strong antifibrotic effect of MSCs, most studies were carried on the early stages of fibrosis development. Since fibrosis is often diagnosed in more advanced phases, assessment of the effects of MSCs on established fibrosis is required in order to consider the routine use of MSC therapy on such pathologies.

66These observations highlight the great potential of MSCs in the treatment of fibrotic diseases. Given these results, MSCs seem to act in the same way, regardless of the organ, and no occurrence of profibrotic effects has been reported. However, the mechanisms by which MSCs act on fibrosis have not yet been clearly elucidated and additional studies are needed. Besides, concerns about effects promoting certain pathologies, such as cancer, are still preventing their routine clinical use. Thus, emphasizing many pathways triggered by MSC homing is of great importance. Furthermore, the regulation of phenotypic changes in MSCs needs to be thoroughly evaluated. As described previously, exposing MSCs to profibrotic stimuli may trigger various changes in their secretome, probably leading to variable responses. Understanding the relative implication of the factors influencing MSC phenotype would provide valuable insight into potentiation and possible adverse effects. In addition, it has been shown that microvesicles or exosomes secreted by MSCs partially reproduce their effect. Describing their composition and elucidating the triggers influencing their content are essential. The importance of MSC homing to damaged tissues also needs to be addressed, mostly in terms of cell-to-cell contacts and microenvironment influence. Although few reports show the importance of engraftment and the differentiation of MSCs, these processes are likely to play a role in the beneficial effects of cell therapy. Also, optimal treatment protocols remain to be established. First, the timing of MSC transplantation surely influences the success of the therapy. The immunomodulatory effect of MSCs should in fact be most effective when transplantation is undergone during the acute inflammatory reaction to prevent the installation of chronic inflammation. Based on the results reported in this review, it is unclear whether MSCs could reverse fibrosis in its more advanced stage and fully restore tissue homeostasis. Nonetheless, MSC therapy for the treatment of fibrosis in any organ should be strongly considered and studied as it shows promising potential.