Mesenchymal stromal cells: Are they all good for the liver?


  • M. Behnan Sahin

    Corresponding author
    1. Department of Medicine, University of Minnesota, Minneapolis, MN
    • Department of Medicine, University of Minnesota, 420 Delaware Street Southeast, Minneapolis, MN 55455
    Search for more papers by this author
    • Telephone: 612-626-4710; FAX: 612-625-6919

  • View this article online at

Mesenchymal stromal cells (MSCs), which are also known as mesenchymal stem cells, are multipotent cells with the ability to differentiate into cells of mesodermal lineage, such as osteocytes, chondrocytes, and adipocytes.1 Recent studies have shown that MSCs are also capable of differentiation into cells derived from the ectoderm and endoderm, including hepatocyte-like cells.2, 3 They are characterized by their ability to adhere to plastic surfaces in culture and to express certain phenotypic markers such as CD29, CD44, CD73, CD90, and CD105. MSCs, which were originally isolated from bone marrow, can now be isolated from virtually all tissues, including skin, adipose tissue, and umbilical cords.4

In addition to their potential to differentiate into multiple cell types, MSCs also possess significant anti-inflammatory activity, which has recently led to numerous studies exploring their therapeutic potential for a wide range of autoimmune and inflammatory diseases.5, 6 These studies suggest that MSCs exert their anti-inflammatory and immunosuppressive effects by their interactions with lymphocytes, which result in the reduced secretion of proinflammatory cytokines such as tumor necrosis factor α, interferon-γ, and interleukin-6 (IL-6) and in the increased secretion of anti-inflammatory cytokines such as IL-4 and IL-10. These effects appear to be mediated mainly by indoleamine-2,3-dioxygenase (IDO), heme oxidase, and programmed death ligand (PDL) pathways.7, 8 However, according to the inflammatory milieu, it seems that MSCs are also capable of expressing inflammatory mediators such as prostaglandin E2 and thus paradoxically perpetuating inflammation. Because inflammation is well known to play a major role in the progression of chronic liver disease to cirrhosis and liver failure, it is not surprising that MSCs have also been studied as a potential therapeutic modality in various models of chronic liver disease.9

In this issue of Liver Transplantation, Gómez-Aristizábal et al.10 report their investigation of the effects of 2 distinct populations of MSCs on hepatocytes and lymphocytes. They generated MSCs from adult bone marrow [ie, bone marrow–derived mesenchymal stromal cell (BM-MSCs)] and human umbilical cord perivascular cells (HUCPVCs) and tested their abilities to support hepatocyte function and inhibit the proliferation of phytohemagglutinin-stimulated peripheral blood mononuclear cells (phaPBMCs). First, they performed a messenger RNA microarray analysis to determine the differences in the gene expression profiles of BM-MSCs and HUCPVCs, and they found that even though BM-MSCs and HUCPVCs shared similar overall expression profiles, there were some significant differences, notably in IL-11, PDL1, and PDL2 expression. These anti-inflammatory factors were expressed at a higher level in HUCPVCs versus BM-MSCs. HUCPVCs also expressed higher levels of 4 chemokines [chemokine (C-X-C motif) ligand 1 (CXCL1), CXCL2, CXCL6, and CXCL8] involved in angiogenesis, whereas some hepatotrophic factors (c-kit ligand, jagged-1, and decorin) were found at lower levels in HUCPVCs versus BM-MSCs.

The investigators then proceeded with functional assays and tested these 2 populations of MSCs for their ability to support in vitro hepatocyte function with a coculture system. They found that even though both BM-MSCs and HUCPVCs were effective as hepatocyte stromal cells, they exhibited some differences. Although HUCPVCs were better at maintaining hepatic ureagenesis, hepatocytes supported by BM-MSCs had higher cytochrome P450 activity. In comparison with BM-MSCs, HUCPVCs also expressed higher levels of hepatotrophic factors (laminin, hepatocyte growth factor, IL-6, and connexin 43) according to quantitative reverse-transcription polymerase chain reaction. Next, the authors analyzed the ability of BM-MSCs and HUCPVCs to inhibit the proliferation of phaPBMCs with another coculture assay, and they demonstrated that BM-MSCs and HUCPVCs were equally effective in their inhibitory activity on lymphocytes and that this activity was mediated by IDO because it was blocked by a chemical inhibitor of IDO.

Overall, this is a well-designed study with a sound experimental approach and convincing results showing that there are significant differences between BM-MSCs and HUCPVCs in their hepatotrophic and anti-inflammatory properties. Nevertheless, the study has some limitations. The authors conclude that both BM-MSCs and HUCPVCs support hepatic function better than native hepatic stromal cells, but this is conceptually difficult to accept. Even though their results support this conclusion, it is likely to be an artifact of the culture system that they used because it is well known that primary hepatic stromal cells are notoriously difficult to culture and maintain in their biologically functional state in vitro. The other surprising finding is that the investigators did not observe any differences between the anti-inflammatory effects of BM-MSCs and HUCPVCS in the phaPBMC inhibition assay despite their microarray data showing some differences between them. This may again be a result of restrictions imposed by the in vitro culture conditions because inflammation is a very complex process involving interactions of many different types of cells in vivo. Therefore, it is crucial to determine how these 2 populations of MSCs would behave in an animal model of chronic liver disease. We certainly hope that the authors will perform follow-up in vivo studies to answer this important question.

In summary, Gómez-Aristizábal et al.10 have demonstrated that MSCs derived from different tissues, although they are phenotypically very similar, exhibit significant biological differences in their anti-inflammatory properties and in their abilities to support hepatic function. As such, this is an important and tantalizing study with major clinical implications. It strongly suggests that different populations of MSCs might have differential therapeutic effects and lead to different outcomes in the treatment of liver diseases.

Although the majority of preclinical studies published in the literature have shown that the effects of MSCs in the treatment of chronic liver diseases are beneficial (as reported by Tsai et al.11), there have also been some studies reporting a negative impact of MSCs on liver function.12-14 For instance, Russo et al.12 found that in a murine model of cirrhosis, BM-MSCs made a significant contribution to fibrosis by differentiating into stellate cells and myofibroblasts in the liver. Because negative studies tend not to be published, we should be concerned about the possibility of adverse outcomes associated with the transplantation of MSCs in humans with liver diseases. Nevertheless, buoyed by the positive results obtained from the majority of animal studies, researchers have started clinical trials involving the transplantation of MSCs for liver diseases (there are currently 2 clinical trials posted on the National Institutes of Health web site for clinical trials: one for acute liver failure and the other for cirrhosis), and a few studies have already been reported with encouraging results.15, 16 These are small, early-phase trials, and we currently have no data about the long-term outcomes and safety of transplanting MSCs in humans. We can address these issues only through well-designed clinical trials using MSCs that have undergone vigorous quality control and preclinical testing in appropriate animal models. MSCs are a promising and exciting new tool for the treatment of liver diseases. However, as scientist and clinicians, we have an ethical obligation to ensure the safety of our patients first.


BM-MSC, bone marrow–derived mesenchymal stromal cell; CXCL, chemokine (C-X-C motif) ligand; HUCPVC, human umbilical cord perivascular cell; IDO, indoleamine-2,3-dioxygenase; IL, interleukin; MSC, mesenchymal stromal cell; PDL, programmed death ligand; phaPBMC, phytohemagglutinin-stimulated peripheral blood mononuclear cell.