Immunomodulatory Role of Mesenchymal Stem Cells in Liver Transplantation: Status and Prospects

Liver transplantation (LT) is a procedure that restores liver function by surgically implanting a healthy liver into a patient with end-stage liver disease. It has been utilized for treating a variety of progressive, irreversible, and fatal end-stage liver diseases including liver cancer, hepatitis, and chronic liver failure [1]. Recently, great strides have been made in LT with the advances of the imaging techniques such as contrast-enhanced ultrasonography and computed tomography, together with interventional therapy [2, 3]. Nevertheless, a large number of patients show immune rejection after LT and have to take a long-term administration of immunosuppressive agents, such as calcineurin inhibitors (e.g., tacrolimus), antimetabolite inhibitors (e.g., azathioprine and mycophenolate mofetil), and corticosteroids [4‒6]. These agents often lead to various adverse events such as nephrotoxicity, metabolic disorders, and primary malignancies [7]. Therefore, it is urgent to find alternatives for immunosuppressants, with an aim to attenuate the immunosuppressant-induced adverse events.

Immune tolerance, including spontaneous immune tolerance and induced immune tolerance, refers to functional unresponsiveness of the immune system toward cells or tissues, in which a recipient with a mature immune system will not attack the donor organ graft even without immunosuppressants [8]. It is considered the ultimate goal of organ transplantation [9]. Spontaneous immune tolerance is an immune tolerance state formed spontaneously without induction, which rarely occurs clinically. Meanwhile, the induced immune tolerance is a state induced by intervention before or after LT, which is a common form of immune tolerance in clinical practice.

At present, the major way for achieving immune tolerance is to reduce the dosage of immunosuppressants. Based on immunology, laboratory test results, and clinical data, Jhun et al. [10] screened eligible subjects that may produce immune tolerance after LT, which were featured by reduction or even termination of the immunosuppressants. In addition, Shamsaeefar et al. [11] reported that the dose of immunosuppressants could be reduced in the recipients with a relatively low risk of immune rejection, whom were treated with gradually terminating the steroids and mycophenolate mofetil at first, followed by giving a small dose of tacrolimus. Despite immune tolerance that can be induced by adjusting immunosuppressant dose, it exhibits a higher risk due to a lack of qualified patients. Meanwhile, these patients may be exposed to a high risk of immune rejection, which may finally affect the outcome of implantation. Interestingly, in addition to the reduction of immunosuppressants, immune cells also can induce immune tolerance in animal studies on LT. For example, Chen et al. [12] reported that interleukin-10 (IL-10)- and FasL-overexpressing dendritic cells could induce immune tolerance. Additionally, Zhao et al. [13] reported their experiences on immune tolerance after treating Kupffer cells (KCs) with IL-34, which activated the PI3K/AKT pathway and triggered the transformation of KCs phenotype from M1 to M2. However, great efforts should be given to improve the therapeutic effects of immune cells as shown to partial regulation of immune responses. Besides, the long-term efficacy cannot be guaranteed because of the short survival of the reinfused immune cells.

Mesenchymal stromal cells (MSCs), with an extraordinary capacity for regulating the phenotype of various immune cells, possess broad immunoregulatory properties to achieve dynamic interaction with innate and adaptive immune systems [14]. MSCs can induce immune tolerance via multiple pathways and were reported to involve in immunomodulatory responses through cell-cell contact and paracrine effects. Even if MSCs are cleared away by the host, their secreted cytokines could still exert curative effects in vivo. Therefore, MSCs may be an effective option to induce immune tolerance after LT. In this review, we summarized the roles of MSCs monotherapy, the combination of MSCs and immunosuppressants, MSCs genetic modification, and MSCs derivative therapy in LT, with an aim to enhance our understanding on the immunomodulatory roles of MSCs in modulating the immune rejection after LT.

MSCs are pluripotent stem cells with self-renewal ability and multi-directional differentiation potential, which have been reported to involve in several biological processes, such as anti-inflammation, anti-oxidation, anti-apoptosis, and immune regulation [15]. Currently, MSCs have been extensively utilized for treating various diseases such as autoimmune diseases and tissue injuries. In a meta-analysis of the randomized controlled trials, Zeng et al. [16] reported that MSCs were effective in treating autoimmune diseases such as rheumatoid arthritis, systemic lupus erythematosus, inflammatory bowel disease, multiple sclerosis, and ankylosing spondylitis. In a mouse model, Sun et al. [17] found that MSCs combined with exercise training could promote the recovery of motor function after spinal cord injury by activating PI3K/AKT/mTOR signaling pathway. In a randomized controlled trial enrolling 30 patients with chronic knee pain, Vega et al. [18] reported that allogeneic MSCs therapy was feasible for the treatment of knee osteoarthritis, which significantly improved the cartilage quality of patients.

MSCs could promote hepatic regeneration by differentiating into hepatocyte-like cells [19]. This would be promising in treating liver fibrosis, liver injury, cirrhosis, liver failure, and liver cancer. Two recent studies have shown that MSCs could hinder liver fibrosis progression in mice by inhibiting the activation of hepatic stellate cells [20, 21]. Meanwhile, splenic vein injection of MSCs contributed to the attenuation of acute liver injury in dogs [22]. Using a non-human primate model, Guo et al. [23] reported that MSC therapy could disrupt inflammatory cascade progression by inhibiting monocyte activation in monkeys with acute liver failure, despite it did not directly improve liver regeneration. Moreover, in a phase II clinical trial, MSCs safely improved histologic fibrosis and liver function in patients with alcoholic cirrhosis [24]. Furthermore, MSCs extracellular vesicles (MSCs-EVs) could be utilized for the delivery of doxorubicin in treating liver cancer as they could significantly increase the drug loads and improve the targeted inhibition of hepatocellular carcinoma [25].

Recent studies have indicated MSCs could secrete massive immunoregulatory factors such as transforming growth factor-β (TGF-β), hepatocyte growth factor (HGF), indoleamine 2,3-dioxygenase, and prostaglandin E2 (PGE2), showing immunomodulatory roles in almost all immune responses [26, 27]. Wu et al. [28] reported that MSCs affected the activation of NF-κB signaling pathway and then inhibited the maturation and differentiation of dendritic cells. In a mouse model of ischemia/reperfusion-induced liver sterile inflammatory injury, Li et al. [29] found that MSCs shifted macrophage polarization from M1 to M2 phenotype and diminished inflammatory mediators. Based on a chronic graft-versus-host disease mouse model, Guo et al. [30] found that MSCs-EV inhibited the activation of macrophages and B cells immune responses. In a model of lipopolysaccharide-induced acute lung injury, MSCs treatment induced down-regulation of immunoglobulin-related gene in B cells and attenuation of inflammatory responses [31]. In a pilot study of MSCs for acute liver allograft rejection, Shi et al. [32] reported a significant increase of TGF-β and PGE2 after MSCs infusion and confirmed that MSCs significantly increased regulatory T (Treg) cells and decreased Treg/T helper 17 (Th17) cell ratio, thereby regulating the balance between Treg cells and memory T cells. An in vitro study showed that the secretion of indoleamine 2,3-dioxygenase and PGE2 by MSCs not only inhibited natural killer (NK) cellular proliferation but also impaired the activity of NK cells [33]. Jiang et al. [34] suggested that under conditions of unrestricted neutrophil activation, MSCs could release superoxide dismutase 3 to remove superoxide anions, preventing neutrophil extracellular trap formation, neutrophil death, and matrix degradation. All these studies suggested that MSCs could secrete immunomodulatory factors that regulate immune responses both in the antigen-presentation phase and in the immune-killing phase. Consequently, MSCs have been well adopted as one of the most promising candidates to replace immunosuppressants.

MSCs can be isolated from a variety of tissues such as bone marrow (BM-MSCs), umbilical cord (UC-MSCs), and adipose tissues (AD-MSCs), and the most widely used tissue in LT is currently BM-MSCs. Zhang et al. [35] reported that infusion of hypoxia-pretreated BM-MSCs could improve liver function and suppress inflammatory responses in small-for-size LT (SFSLT) rats. Xia et al. [36] found that BM-MSCs infusion after LT could prevent graft-versus-host disease in rats. In a phase I/II trial, BM-MSCs and rituximab showed similar efficiency in preventing acute rejection after ABO-incompatible LT [37]. AD-MSCs and UC-MSCs have the advantages of easy access and less ethical controversy compared to BM-MSCs. Guo et al. [38] reported that AD-MSCs could promote liver regeneration and suppress immune responses in SFSLT models. A clinical trial by Shi et al. [32] showed that infusion of UC-MSCs could trigger increase of serum TGF-β and PGE-2 levels, together with attenuating the acute rejection. To compare the immunosuppressive effects of MSCs from different sources, Kim et al. [39] co-cultured MSCs with peripheral blood mononuclear cells, which indicated that periodontal ligament stem cells, UC-MSCs, and AD-MSCs inhibited peripheral blood mononuclear cellular proliferation in a dose-dependent manner. In the future, more studies should be performed to illustrate which type of MSCs show the best immunosuppressive capacity.

Some studies indicated that MSCs could attenuate the ischemia-reperfusion injury (IRI), which is a common complication of LT serving as a key limiting factor for a good outcome in the patients. Some studies revealed that MSCs could modulate oxidative stress, autophagy, and apoptosis that were closely associated with the pathogenesis of IRI [40, 41]. For instance, Wang et al. [42] indicated that BM-MSCs could protect the rats against IRI by enhancing the autophagy mediated by HO-1 (heme oxygenase-1). In addition, Li et al. [43] reported that BM-MSCs with overexpression of superoxide dismutase 2 could attenuate hepatic IRI through inhibiting oxidative stress and apoptosis of hepatocytes. Nowadays, there are two types of immune rejection, including host-versus-graft reaction and graft-versus-host reaction. The host-versus-graft reaction is common in LT and may lead to graft failure and the risk of hepatic retransplantation in those with severe conditions. Based on graft survival and histopathological data, Wang et al. [44] confirmed that BM-MSCs from donors, recipients, and third party could ameliorate acute rejection of rat liver allografts, and the immunoregulatory effects of MSCs were related to expansion of Treg cells. You et al. [45] reported a better allograft tolerance after LT in allograft recipients who received BM-MSCs, and such effect was associated with reprograming the phenotype of KCs from M1 polarization to M2 polarization. It is worth noting that although graft-versus-host reaction occurs less frequently in LT than host-versus-graft reaction, great attention should be paid as it will lead to transplant failure, together with severe adverse events to the recipients. Xia et al. [36] reported that BM-MSCs could prevent acute graft-versus-host disease. Animal studies suggested that in SFSLT models, BM-MSCs could reduce IRI and promote liver regeneration by sustaining early increased expressions of c-Jun N-terminal kinase, cyclin D1, and NF-κB [46]. Saidi et al. [47] reported that injection of human adipose-derived MSCs attenuated IRI and promoted liver regeneration in mice, with markedly decreased levels of serum interleukin-6 and alanine aminotransferase. Taken together, MSCs can significantly improve liver function and prolong graft survival, playing a considerable role in LT (Fig. 1; Table 1). Despite the good results of MSCs in animal experiments, there are still some problems. First, there is a lack of data on large animals as most of the studies are performed in rats. Second, there are very few data on the application of MSCs in the treatment of chronic rejection after LT as chronic rejection usually does not appear after LT attributed to the good immune tolerance of model animals. Third, there are still no reports on treatment with MSCs alone in LT clinical trials as the existing clinical trials report the use of MSCs in combination with conventional immunosuppressants.

Recently, the combination of MSCs and immunosuppressants has been proposed for treating immune rejection after LT as it could reduce doses of immunosuppressants and maintain a similar treatment outcome. In a randomized controlled trial, Casiraghi et al. [48] reported single injection of BM-MSCs (0.90–1.67 × 10⁶) to a LT recipient before transplantation was safe in the 5-year follow-up. Shi et al. [32] found that single or multiple peripheral intravenous infusions of UC-MSCs significantly increased serum TGF-β and PGE2 levels, reduced liver injury, and attenuated acute rejection in LT recipients treated with conventional immunosuppressive agents such as tacrolimus, corticosteroids, or mycophenolate mofetil. In a study focused on the safety and efficacy of human UC-MSCs in treating ischemic biliary lesions after LT, Zhang et al. [49] reported that, based on a 2-year follow-up, UC-MSCs peripheral intravenous infusions (6 doses) were clinically safe and beneficial in the short-term treatment of such lesions in the presence of conventional immunosuppressive agents (e.g., ursodeoxycholic acid, adenosine methionine, prostaglandin-E, and antibiotic). Another study by Zhang et al. [37] evaluated the safety and feasibility of UC-MSCs to replace rituximab in patients undergoing ABO-incompatible LT who received conventional immunosuppressive agents including steroids, basiliximab, tacrolimus, mycophenolate mofetil, and intravenous immunoglobulin. The results showed that although MSCs and rituximab showed comparable ability to prevent acute immune rejection, MSCs were more beneficial in preventing infection and biliary complications compared to rituximab. Furthermore, multiple injections of UC-MSCs could treat acute rejection caused by steroid resistance under conventional immunosuppression after LT [50]. These findings provide great support for the clinical application of MSCs as an alternative to immunosuppressive drugs. However, in a study involving LT recipients under standard immunosuppression, Detry et al. [51] reported that single infusion of third-party unrelated MSCs did not induce toxicity; however, it was not sufficient to allow withdrawal of immunosuppression (Table 2). These indicated that there were still disputes on the efficiency of MSCs in LT despite their favorable safety.

In fact, the use of MSCs in clinical practice still has a long way to go for the following reasons. First, MSCs may induce a high risk of recurrence as the majority of patients receive LT due to liver cancers. Second, there is still a lack of standardized protocols on cell source and culture of MSCs. Third, there are not adequate laws or regulations describing the ethical standards for MSC utilization in treating human diseases, which is one of the most important factors restricting the application of MSCs. Fourth, there are indeed antagonizing effects between MSCs and immunosuppressive agents, which may affect the treatment efficiency and safety.

To date, some scholars raise their concerns about the combination of MSCs and immunosuppressants. Will the immunosuppressants interfere to MSC function? Will the immunosuppressants inhibit the biological function of MSCs as they express the molecular targets of immunosuppressants? To their expectation, the combination of immunosuppressants and MSCs seems to show antagonistic effects. For example, Buron et al. [52] reported that both cyclosporine A and rapamycin antagonized the inhibitory effects of MSCs on mixed lymphocyte reactions. For the mechanisms, most scholars propose that the curative effects of MSCs are related to the immunosuppressant type and/or the concentration of the immunosuppressive agents [53]. Overall, there is no consensus on the effects of types and concentrations of immunosuppressants on MSCs. In the future, more reasonable experiments are required to investigate the potential causes.

Genetic modification is another way to improve the efficacy of MSCs. Genetically modified MSCs have been used for disease treatment as early as 2002 [54]. In 2007, Yu et al. [55] used MSCs overexpressing human hepatocyte growth factor (HGF) gene to treat IRI associated with SFSLT, which indicated that HGF-overexpressing MSCs could prolong graft survival by enhancing the regeneration of liver cells and migration of cells into liver grafts, which provided a novel approach for the application of MSCs in LT. To date, the most widely used MSCs are HO-1-modified MSCs. Wu et al. [56] reported that HO-1-transduced BM-MSCs could improve liver function and regulate serum inflammation-associated cytokines, thereby attenuating rejection in an acute cellular rejection model of rat LT. In an experimental study using male Lewis rats and BN rats, Li et al. [57] also indicated that HO-1 could enhance the inhibitory effects of BM-MSCs on acute rejection in orthotopic LT. Wang et al. [42] reported that infusion of HO-1-transduced BM-MSCs could promote autophagy and reduce liver injury in ischemia-reperfusion rats.

The other utilized genetic modification of MSCs involved TGF-β, programmed death ligand-1 immunoglobulin, forkhead box P3, SOD2 (superoxide dismutase 2), NaHS, IL-10, as well as C-X-C chemokine receptor type 4. In detail, the combination of TGF-β and MSCs, two effective immunosuppressive agents for organ transplantation, has been utilized in LT as it could inhibit the proliferation of immunocompetent cells and lymphocyte differentiation. In a rat model with acute rejection after LT, TGF-β-overexpressing BM-MSCs could promote Treg generation, improve immune rejection, and induce local immune tolerance [58]. Qi et al. [59] reported that portal infusion of FOXP3-overexpressing BM-MSCs could inhibit T-cell proliferation and induce donor-specific immune tolerance in LT model rats. Li et al. [60] found that modification of BM-MSCs with programmed death ligand-1 immunoglobulin could improve function and significantly prolong survival in acute rejection model rats. These modified MSCs are mainly used to treat immune rejection and IRI after LT. Among these genes utilized in the MSCs modification, antioxidant genes (e.g., HO-1, SOD2, and NaHS) were mainly used for IRI treatment after LT [43, 61], while anti-inflammatory and immunomodulatory genes (e.g., TGF-β, CXCR-4, and IL-10) were used to treat immune rejection [62, 63]. Although genetically modified MSCs have been reported to be promising in treating LT, the off-target effects of genetic modification pose potential biosafety risks. For instance, exogenous gene insertions would be integrated into the cellular genome when using some integrative vectors. This may cause mutations in key genes or activate proto-oncogenes, leading to an increased risk of malignancy. Therefore, full investigation should be given to the genetically modified MSCs before clinical promotion.

MSCs play immunomodulatory roles through cell-cell contact and paracrine secretion [64], and MSC-differentiated derivatives are foremost in exerting their functions [27]. In addition to the advantages of MSCs such as low immunogenicity, a wide source, and easy acquisition, MSCs derivatives often exhibit a higher safety characterized by the inability of proliferation and differentiation, long-term storage stability, and a low possibility of causing thrombosis after intravenous administration [65, 66]. Therefore, MSCs derivative therapy is considered the best alternative for MSCs therapy.

At present, MSCs derivatives mainly include exosomes (MSCs-exos), MSCs-EV, and conditioned medium (MSCs-CM) (Fig. 2). MSCs-exo showed similar immunomodulatory effects with parental MSCs, which can alternate immune cell phenotype and regulate cytokines secreted by immune cells such as T cells, NK cells, macrophages, and dendritic cells [67]. To date, several methods have been adopted to obtain MSC-exo, including ultracentrifugation, ultrafiltration, and extraction based on commercial kits. Michele et al. [68] recommended ultrafiltration for collecting MSC-exo as it yielded a higher concentration of MSCs-exo without aggregation of MSCs-EV, compared with the counterpart obtained after ultracentrifugation. MSCs-exos are also widely used in IRI treatment due to their strong ability to directly infiltrate injured target organs [66]. Mechanistically, they could attenuate IRI by increasing autophagy and ferroptosis [69, 70]. In addition, Zhang et al. [71] found that MSCs-exo could inhibit mitochondrial fission and promote mitochondrial biogenesis by modulating the expression of genes and proteins related to these processes, thereby maintaining mitochondrial homeostasis and alleviating IRI.

MSCs-EVs are nanoscale particles with a stable membrane structure and phospholipid bilayer, which transfer bioactive molecules into recipient cells mainly through membrane fusion, endocytosis, and specific receptor-ligand recognition [72, 73]. They are obtained from MSC-CM and often contain high amounts of soluble proteins that may mask the related function of MSCs-EV. Marta et al. [72] separated MSCs-EV from MSCs-CM using size exclusion chromatography and confirmed that MSCs-EV did not induce inflammatory responses, whereas the non-extracellular vesicle fractions might stimulate T cells and cause inflammatory responses. Additionally, MSCs-EVs have similar immunomodulatory functions as MSCs. Tsai et al. [74] reported that engineered FGL1/PD-L1-overexpressing MSCs-EV significantly inhibited activation and proliferation of T cells, together with alleviating the immune rejection in a heart allograft model.

MSCs-CM are whole ingredients (secretome) secreted by MSCs during culture; therefore, their efficacy is closer to MSCs than MSCs-exo and MSCs-EV. Jiao et al. [75] indicated that MSCs-CM therapy was a viable alternative to MSCs therapy. Daan et al. [76] found that MSCs-CM therapy could reduce hepatocyte apoptosis by up to 90%. In addition, Du et al. [77] reported that systemic infusion of MSC-CM could relieve liver injury and reduce the apoptosis of sinusoidal endothelial cells in reduced-size LT rats.

There are essentially no differences in three MSCs derivative therapies as they all play a therapeutic role through the paracrine secretion of MSCs. Indeed, there are some limitations in MSCs derivative therapies. First, the heterogeneity of maternal MSCs themselves and the difference in culture conditions may cause heterogeneity in contents and functions of MSCs-exo, MSCs-EV, and MSCs-CM produced by MSCs of different generations. Second, a common disadvantage of cell-free therapies is that they cannot be quantified accurately. To promote the development of MSCs cell-free therapies, it is recommended to add specific cytokines during MSCs culture to make MSCs grow toward the same phenotype, followed by screening out the desired specific subtype by flow cytometry.

Liver is an “immune-privileged” organ, and LT is more likely to develop immune tolerance than other organ transplants [78]. Once immune tolerance has been successfully induced, immunosuppressive drugs could be reduced or withdrawn to prevent transitional immunization. However, it is required to follow up immune-tolerant individuals to the impermanence of immune tolerance [79]. Pathology is the gold standard for disease detection. Needle biopsy is often used clinically to obtain target tissues, but it is traumatic with several contraindications. Therefore, it is a shared purpose for scholars to definite immune tolerance markers in blood, urine, or imaging, with an aim to detect immune tolerance through non-invasive means. The most studied indicators are hematological, with Treg and Th17 predominating. Zhou et al. [80] found that the levels of serum Th17 and Treg-related cytokines in the LT tolerance group were similar to those in the control group, which was significantly different from those in the rejection group. This suggested that serum Th17 and Treg may be used as immune tolerance markers. Based on immune indicators, Jhun et al. [10] reported that 14 patients were selected as candidates for immunosuppressive dose reduction after LT, in which 7 patients were considered immune tolerant as they did not develop immune rejection after dose reduction or even withdrawal. There was a significant increase in peripheral blood Treg/Th17, Th1/Th17, and CD8/Th17 in the immune-tolerant patients compared to the counterparts with immune rejection. In contrast, Adenugba et al. [81] proposed that changes in biomarkers should be monitored dynamically over time, rather than simply comparing biomarker levels at the same time point. They found that peripheral blood Treg in the immune-tolerant group was first enriched after LT followed by a remarkable decay, but no Treg enrichment was observed in the immune rejection patients. Additionally, Aldo et al. proposed to rule out the immune rejection based on ALT levels, serum donor-specific HLA antibodies, and liver stiffness by FibroScan [82]. Although many non-invasive biomarkers of immune tolerance have been proposed, future studies are required to validate their efficiency. Meanwhile, all these tolerance indicators have a certain time lag and cannot be predicted in advance.

Numerous studies show that MSCs are beneficial for LT. MSCs are available from a wide range of sources and all can exert immunomodulatory effects, but more studies are needed to investigate which is best source of immunomodulation. Most animal studies on MSCs therapy have focused on the single application of MSCs as they show good efficacy; therefore, very rare studies focus on the efficacy of MSCs combined with immunosuppressants. Clinical trials are very complex, requiring the selection of the corresponding immunosuppressant according to the actual situation. On this basis, the immunosuppressant dosing regimen may also be adjusted during the trials. Therefore, more attempts should be given to the optimal combination regimen for treating various diseases. Genetically modified MSCs have better immunomodulatory, antioxidant, and regenerative abilities in animal studies; however, the off-target effect severely limits its application in clinical LT. Meanwhile, there is a need to find a safe and effective means of genetic modification and verify its safety. Since most LT recipients are hepatocellular carcinoma patients with the possibility of tumor recurrence, MSCs derivatives have a better safety profile compared with MSCs treatment, but MSCs derivatives have a large heterogeneity and cannot be accurately quantified. In conclusion, MSCs have very great potential in LT therapy but also face many problems that need to be settled in depth to provide a solid foundation for its application in LT.

The authors have no conflicts of interest to declare.

This research was supported by the Natural Science Foundation of Fujian Province [Grant No. 2020J011163] and the Medical Innovation Project of Fujian Province [Grant No. 2020CXA054].

Haitao Li: research conception, research design, data acquisition, data analysis, manuscript draft, and critical revision. Saihua Yu and Lihong Chen: data acquisition, data analysis, and critical revision. Hongzhi Liu and Conglong Shen: data analysis and critical revision.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.