Strategies to improve the efficiency of mesenchymal stem cell transplantation for reversal of liver fibrosis

Abstract End‐stage liver fibrosis frequently progresses to portal vein thrombosis, formation of oesophageal varices, hepatic encephalopathy, ascites, hepatocellular carcinoma and liver failure. Mesenchymal stem cells (MSCs), when transplanted in vivo, migrate into fibrogenic livers and then differentiate into hepatocyte‐like cells or fuse with hepatocytes to protect liver function. Moreover, they can produce various growth factors and cytokines with anti‐inflammatory effects to reverse the fibrotic state of the liver. In addition, only a small number of MSCs migrate to the injured tissue after cell transplantation; consequently, multiple studies have investigated effective strategies to improve the survival rate and activity of MSCs for the treatment of liver fibrosis. In this review, we intend to arrange and analyse the current evidence related to MSC transplantation in liver fibrosis, to summarize the detailed mechanisms of MSC transplantation for the reversal of liver fibrosis and to discuss new strategies for this treatment. Finally, and most importantly, we will identify the current problems with MSC‐based therapies to repair liver fibrosis that must be addressed in order to develop safer and more effective routes for MSC transplantation. In this way, it will soon be possible to significantly improve the therapeutic effects of MSC transplantation for liver regeneration, as well as enhance the quality of life and prolong the survival time of patients with liver fibrosis.


| INTRODUCTION
The liver is a digestive organ that stores glycogen, scavenges toxins and participates in protein synthesis for metabolic homeostasis.
Because it makes direct contact with external toxins, the liver is easily injured under stress conditions. In response to repeated and chronic liver injury induced by hepatitis B virus (HBV), hepatitis C virus (HCV), sustained alcohol consumption and fat deposition, [1][2][3][4] the liver can accumulate aberrant myofibroblasts and extracellular matrix, thus generating liver fibrosis with poor prognosis. The pathogeneses of primary biliary cirrhosis, primary sclerosing cholangitis and autoimmune hepatitis are very different from other types of chronic liver fibrosis, as portal fibroblasts are found around bile ducts. 5 Intriguingly, schistosomiasis induces liver fibrosis through accumulation of parasitic ova and periovular granulomas in portal veins. 6 Hepatolenticular degeneration, known as Wilson's disease, accounts for a small proportion of metabolic liver diseases. It is caused by a mutation in the Wilson disease protein (ATP7B) gene and frequently leads to liver fibrosis. 7 In addition, metabolic syndromes including obesity, insulin resistance and diabetes have recently been found to be closely related to end-stage liver fibrosis. 8 Although the mortality of liver cirrhosis varies substantially across different regions of the world, it has been universally acknowledged by multiple investigators that liver cirrhosis has gradually become an increasing health burden worldwide, as liver cirrhosis and other chronic liver diseases contributed to 2% of deaths worldwide in 2015, with a relative increase of 10.3% from 2005. 9 They also highlighted that the increasing mortality mainly attributed to viral hepatitis, alcoholic liver disease and non-alcoholic fatty liver disease in developed countries. 9 The long-term inflammatory response and fibrotic state induced by various factors leads to other complications, including hepatocellular carcinoma (HCC) and liver failure.
Although multiple drugs are available for recovering liver function in patients, there are almost no effective drugs for reversing the preexisting accumulation of myofibroblasts and extracellular matrix. Currently, the most effective treatment for end-stage liver fibrosis is liver transplantation, but it is limited by scarce donor grafts, immunologic rejection, complex surgery, high costs, etc. Although hepatocyte transplantation, which emerged as a substitution, is able to restore liver function and promote liver regeneration, this treatment is limited because hepatocytes easily lose their viability in vitro. Transplantation of stem cells, including mesenchymal stem cells (MSCs), haematopoietic stem cells and endothelial progenitor cells, has proven to be effective in eliminating chronic liver injury to repair fibrotic livers by promoting hepatocyte transdifferentiation and hepatocyte proliferation, inhibiting activated hepatic stellate cells (HSCs), up-regulating the activity of matrix metalloproteinases (MMPs) and promoting neovascularization in liver tissues. 10 However, it is hard to regress more significant liver fibrosis (cirrhosis), thus an intervention that targets the fibrosis is needed. Considering that MSCs have abundant resources, strong proliferative ability, multilineage potential and no ethical considerations for widespread application to repair various organ injuries, they are currently transplanted in vivo to reduce hepatocyte apoptosis and promote hepatocyte regeneration. 11 Before application, the isolated and purified MSCs must met three criteria according to the International Society for Cellular Therapy: adherence to plastic under standard culture conditions; expression of CD105, CD73 and CD90, and lack of expression of   haematopoietic and endothelial markers including CD11b, CD14,   CD31, CD34, CD45 and HLA-DR; differentiation into adipocytes, osteocytes and chondrocytes under specific in vitro culture conditions. 12 However, only a small number of MSCs migrate to injured tissues after cell transplantation, so multiple studies have tried to investigate effective strategies for improving the survival rate and activity of MSCs to treat liver fibrosis. Repairment of the injured tissues of liver fibrosis is influenced by multiple factors including the delivery route, the resources of transplanted cells, the number of infused cells, culture conditions, gene modification of MSCs and other potential factors. Hence, we herein arrange and analyse the current evidence related to MSC transplantation in liver fibrosis and summarize the detailed mechanisms and new strategies of MSC transplantation for promoting the regression of liver fibrosis. We anticipate the development of safer strategies to improve MSC activities in vivo to repair liver function and promote the regression of liver fibrosis in regenerative medicine.

| POTENTIAL MECHAN ISMS
Chronic liver injury induces liver fibrosis via up-regulating the accumulation of extracellular matrix in vivo, and then normal hepatic architecture is replaced by a nodular structure of fibrous septa. In general, myofibroblasts are the major source of extracellular matrix, HSCs are considered to be the principal precursor population for myofibroblasts. Because liver tissue consists of multiple cell types, in vitro studies do not completely mimic the complex situation of the liver, but animal models can be used as a gold standard for in vivo study. A majority of studies investigated the potential mechanisms of MSC and MSC derivative-based therapies in liver fibrosis for enhancing the therapeutic effects ( Figure 1).
Mesenchymal stem cells migrate into fibrogenic liver tissues after transplantation in vivo and then differentiate into hepatocyte-like cells or fuse with hepatocytes to protect liver function. 13 Interestingly, Park et al previously showed that human palatine tonsilderived MSCs responded only to disease tissue, as they were detected in damaged livers but not in healthy livers, and it was also shown that these implanted MSCs differentiated into hepatocyte-like cells to eliminate liver fibrosis via activation of autophagy and downregulation of the TGF-β signalling pathway. 14 In contrast, menstrual blood-derived stem cells were demonstrated to eliminate collagen deposition and inhibit proliferative HSCs via paracrine mediators, but few of the transplanted cells differentiated into functional hepatocyte-like cells despite migrating to the sites of injury. 15 Transplanted MSCs up-regulate the expression levels of key enzymes associated with glucose metabolism in carbon tetrachloride (CCl 4 )-induced liver cirrhotic rats, thus maintaining normal metabolic status in liver fibrotic rats. 16 Transplantation of MSCs consequently improved liver function and reduced liver histopathology and hepatobiliary fibrosis by inhibition of HSCs, down-regulation of collagen deposition and enhancement of extracellular matrix remodelling via the up-regulation of MMP-13 and down-regulation of tissue inhibitor of metalloproteinase (TIMP)-1. 17 Coculture of MSCs and HSCs inhibited the proliferation of HSCs and promoted cell apoptosis of HSCs via down-regulating the E3 ligase S-phase kinase-associated protein 2 (SKP2) level, attenuating the ubiquitination of p27 and increasing the stability of p27. 18 Moreover, MSCs are demonstrated to produce various growth factors and cytokines with anti-inflammatory effects in vitro and in vivo to reverse the liver fibrotic state, as transplantation of MSCs increases the serum levels of vascular endothelial growth factor (VEGF), hepatocyte growth factor (HGF), IL-10 and MMP-9 in injured livers. 19 MSCs in vivo attenuated hepatic fibrosis as shown by decreased serum levels of collagen I, type III procollagen, collagen IV, hyaluronic acid and laminin, and down-regulated liver collagen proportionate area, hepatic hydroxyproline and liver αsmooth muscle actin (α-SMA); this progress is accompanied by reduced expression of serum TGF-β1 and reduced hepatic levels of TGF-β1, Smad3 and Smad4 but increased Smad7 expression. 20,21 MSCs significantly ameliorated liver fibrosis in mice via stimulation of interferon (IFN)-γ and inhibition of lymphocyte proliferation, 22 and MSCs has been proved to enhance the release of serum interleukin (IL)-4 and IL-10 and improve the motility of Kupffer cells for elimination of liver fibrosis in rats. 23  MSCs directly or indirectly eliminate the deposition of extracellular matrix by hepatogenic differentiation, fusion with hepatocytes, paracrine effects and immunological regulation, contribute to degradation of scar tissue and promotion of myofibroblast apoptosis. All of these potential mechanisms cooperate to regulate the therapeutic effects of MSC transplantation in liver cirrhosis models.

| ROUTES OF TRANSPLANTATION
As MSC transplantation can be applied via different routes and acquire different efficiency, the optimal way to transplant MSCs may significantly improve the prognosis of liver fibrosis. However, the optimal route for MSC transplantation in the regression of liver fibrosis remains controversial. Truong et al showed that transplantation of MSCs by not only the tail vein but also the portal veins effectively improved liver function and protected the liver from F I G U R E 1 The potential mechanisms of MSC and MSC derivative-based therapies in liver fibrosis for enhancing the therapeutic effects continued development of fibrogenesis. 28 Intravenous and intrasplenic MSC transplantation demonstrated a comparable restoration of liver function, but the intravenous route significantly decreased the levels of IL-1β, IL-6 and INF-γ in liver tissues than the intrasplenic route. 29 Wang et al demonstrated that intraportal transplantation was a better route than tail vein transplantation for improving the therapeutic effects of MSCs on liver fibrosis. 30 MSC transplantation via portal vein significantly decreased hepatic arterial perfusion index, but increased portal vein perfusion and total liver perfusion as shown by computed tomography perfusion scan; moreover, liver functional test and histological findings in portal vein group were significantly improved but there seemed to be no significance in rats receiving MSCs from tail vein. 30 Zhong et al detected no signal change in the vena caudalis group by magnetic resonance imaging, but they showed that MSC migration was gradually increased after immediate transplantation and decreased gradually after 3-hour transplantation through portal vein route. Then, MSC signals disappeared in tissues around portal area while appeared in fibrous tuberculum at the edge of the liver at day 14 after transplantation. They concluded that portal vein route seemed to be more beneficial than the vena caudalis route on MSC migration to fibrotic liver. 31 Comparing intravenous, intrahepatic and intraperitoneal injection routes, Zhao et al demonstrated that intravenous injection was the most effective route for improving serum IL-10 levels and decreasing IL-1β, IL-6, tumour necrosis factor (TNF)-α and TNF-β in liver tissue to reverse liver fibrosis and restore liver function. 32 In addition to the transplantation route, the transplantation frequency can be changed for improving the transplantation efficiency. Repeated infusions of MSCs three times significantly improved survival, liver fibrosis and necrosis than infusions of the same number of MSCs in a single dose. The repairment was accompanied by up-regulation of the fibrogenic-related genes and improved homing of MSCs after longterm observation for 3 weeks. 33 According to current studies, the majority of investigators accept that portal vein route seems to be the best choice for MSC transplantation. However, the optimal route for transplantation in animal models still needs to be further investigated to achieve better effects in clinical trials.

| SOURCES OF MSCS
Considering that MSCs from different sources may have unique features, specialists have compared the therapeutic effects of different MSCs at the molecular level and pathological process. Sayyed et al demonstrated that umbilical cord blood-derived CD34 + MSCs were more efficient than bone marrow-derived MSCs in elevating albumin level and reducing alanine aminotransferase (ALT) level, and they concurrently showed that umbilical cord blood-derived CD34 + MSCs reduced the levels of COL1A1, TGF-β1 and α-SMA to a lower degree and increased the level of MMP-9 to a greater degree than bone marrow-derived MSCs. 34 Rengasamy et al suggested that bone marrow-derived MSCs were more effective in reducing liver fibrosis than Wharton's jelly-derived MSCs in CCl 4 -induced liver fibrotic rats, as shown by lower levels of α-SMA, higher levels of MMP-1 and greater activation of hepatic progenitor cells in rats treated with bone marrow-derived MSCs. 35 Hao et al found that adipose-derived MSCs achieved greater reductions of the proliferation and activation of HSCs and secreted higher levels of nerve growth factor and TGF-β1 in the cell culture medium than bone marrow-derived MSCs.
They also found that although adipose-derived MSCs improved antiinflammatory and anti-fibrotic effects than bone marrow-derived MSCs, the differences in inflammatory activity and fibrosis staging scores were not significant. Thus, they concluded that bone marrowderived MSCs and adipose-derived MSCs are similarly effective at attenuating liver fibrosis. 36 In contrast, Baligar et al demonstrated that bone marrow-derived CD45 + MSCs had better anti-fibrotic ability than adipose-derived MSCs because they expressed higher levels of MMP-9 and MMP-13 and inhibited HSC proliferation more effectively. 37 Most comparisons were conducted between bone marrowderived MSCs and other MSCs, while the optimal choice is still undetermined for treating liver fibrosis. To this end, we want to highlight that bone marrow-derived MSCs are applied as the main source of MSC transplantation, hence, we call for more studies to compare the advantages and disadvantages among MSCs from various resources. showing that these cells had higher homing rates and stronger effects on recovery from liver fibrosis than undifferentiated MSCs. 40 However, there is still controversy surrounding the beneficial effects of these two kinds of cell transplantation. MSCs are more effective in maintaining liver function compared with hepatogenic MSCs, although both of them can effectively reverse liver fibrosis in a rat model. 41 The use of secretome derived from MSCs or conditioned media to reduce liver fibrosis has gradually become a hot topic in current regenerative medicine. The secretome is a special set of factors (soluble proteins, free nucleic acids, lipids and extracellular vesicles) secreted into the extracellular space and changes in response to fluctuations in various conditions. 42 Cell-free secretome isolated from MSCs exerted antifibrotic effects by inhibiting activation of TGF-β/ Smad signalling and HSCs. 43 Exosomes, which are small (30-100 nm in diameter) membrane vesicles released by MSCs from various resources, are proved to have the same functions in vivo with their derived cells. 44 They are also able to reduce surface fibrous capsules, soften textures and alleviate inflammation and collagen deposition in fibrotic livers by inhibiting epithelial-to-mesenchymal transition. 45 Conditioned medium from MSCs induces apoptosis of HSCs, protects hepatocytes from apoptosis and down-regulates the number of infiltrating macrophages, thus exerting antifibrotic effects and healing fibrotic scarring in liver tissue. 46 (Table 1).

OF MSCS
Although hypoxia has been proved to activate STAT3/hypoxia-in-   MSC transplantation, and this strategy significantly enhanced the homing and repopulation of MSCs and improved liver function than the control group, for eliminating liver fibrosis. 61 Pretreatment with sodium nitroprusside on CCl 4 -injured mice improved the MSC homing rate, thus decreasing the expression of fibrotic markers including α-SMA, collagen 1α1, TIMP, nuclear factor κB (NFκB) and inducible nitric oxide synthase (iNOS) and liver fibrosis. 62

| GENE MODIFICATION
Gene modulation is generally applied to reprogram somatic cells into a stemness state, and it can effectively enhance the effects of MSC transplantation in liver fibrotic models as well (Table 2).
Considering the important role of TIMP-1 in liver fibrosis progression, lentiviral vector-mediated silencing of TIMP-1 in MSCs significantly reduced fibrotic area and collagen deposition in a rat model of liver fibrosis. 66 In addition, MSCs transfected with MMP-1 enhanced the reduction of liver fibrosis than MSC group by downregulating collagen content and inhibiting activation of HSCs. 67 Overexpression of IGF-I in MSCs significantly increased the levels of IGF-I and HGF in the livers of treated mice than MSC group, and multiple doses of modified MSCs dramatically suppressed inflammatory responses and reduced collagen deposition in fibrotic livers. 68 Overexpression of HGF in MSCs also enhances the migratory ability of MSCs, and these modified MSCs have stronger responses to SDF-1α than control MSCs. 69     repair liver dysfunction, the low efficacy of genetic modification and its potential for tumourigenicity may limit its application.

| CLINICAL TRIALS
Investigators focus on developing new method to improve MSC efficiency in vivo in animal models with liver cirrhosis, but the real effects should be estimated by clinical trials ( Table 3) After one year, MSC transfusion proved to be clinically safe and effectively reduced ascites in these patients. 79  decrease in the levels of TGF-β1, COL1A1 and α-SMA. 82 In addition, 72 patients with alcoholic cirrhosis were enrolled for MSC transplantation, and although one-time and two-time transplantation reduced collagen deposition by 25% and 37%, respectively, there was no significant difference in fibrosis quantification between the two groups. 83  In consideration to liver cirrhosis induced by various factors, MSC transplantation significantly increased the levels of serum albumin and prealbumin and improved MELD scores but exerted no significant influence on the levels of coagulation indicators or AFP in patients with decompensated cirrhosis (alcoholic liver disease, 37 with HBV infection and 2 with HCV infection). 88

| CONCLUSIONS
Long-term exposure to viral hepatitis, toxic chemicals, alcohol, lipid deposition, parasites or autoimmune elements in human and animals lead to liver fibrosis with a poor prognosis. The imbalance between MMPs and F I G U R E 2 Drugs, liver transplantation, hepatocyte transplantation and stem cell transplantation can be applied to liver fibrosis induced by multiple toxic factors TIMPs is a key contributor to the pathogenesis of liver fibrosis, and HSC activation also plays a vital role in its progression. Drugs, liver transplanta-