Review
TRENDS in Immunology
Vol.28 No.5
Mesenchymal stem cells: a new strategy for immunosuppression? Antonio Uccelli1,3*, Vito Pistoia2* and Lorenzo Moretta2,3,4 1
Department of Neurosciences, Ophthalmology and Genetics, University of Genova, Genova, Italy G. Gaslini Institute, Genova, Department of Experimental Medicine (DIMES), University of Genova, Genova, Centre of Excellence for Biomedical Research, University of Genova, Genova, Italy 3 Department of Experimental Medicine (DIMES), University of Genova, Genova, Italy 4 Centre of Excellence for Biomedical Research, University of Genova, Genova, Italy 2
In vitro-generated mesenchymal stem cells (MSCs) initially attracted interest for their ability to undergo differentiation toward cells of different lineages. More recently, a major breakthrough was the discovery that MSCs exert a profound inhibitory effect on T cell proliferation in vitro and in vivo. Subsequently, MSCs were shown also to exert similar effects on B cells, dendritic cells and natural killer cells. These results suggested that MSCs could be used to dampen immune-mediated diseases and transplant rejection. It is possible that some of the beneficial effects of MSCs might reflect, in part, the trophic and protective activities they exert on injured cells and tissues, rather than resulting from a true transdifferentiation. In immune-mediated diseases, the protective effects might function in concert with the immunosuppressive and anti-inflammatory activities. Bone marrow-derived pluripotent stromal cells affect immune responses The hematopoietic stem cell (HSC) ‘niche’ is the site to which HSCs home in the bone marrow (BM), in close contact with osteoblasts, endothelial and stromal cells [1]. BM-derived stroma contains a subset of heterogeneous mesodermal progenitor cells named ‘multipotent mesenchymal stromal cells’ [2] or, more commonly, ‘mesenchymal stem cells’ (MSCs), which are difficult to isolate and characterize as unmanipulated cells ex vivo. By contrast, MSCs can be rapidly expanded in culture as adherent cells from the BM of humans and animals. Therefore, most studies addressing functional features in vitro, and following in vivo infusion, have been performed with cultured MSCs. In vitro expanded MSCs are capable of supporting hematopoiesis and differentiate into cell types of the mesodermal lineage but also, in some experimental conditions, into cells of other lineages, including neural cells [3]. These findings have been crucial in support of the use of MSCs to favor the engraftment of HSCs following allogeneic HSC transplantation (HSCT) [4] and for the treatment of osteogenesis imperfecta [5]. However, the rationale for the use of MSCs for tissue repair purposes stems from their capacity not only to differentiate into multiple cell types, but also to enhance the recruitment and proliferation of local *
Corresponding author: Moretta, L. (
[email protected]). These authors contributed equally to this paper. Available online 2 April 2007.
www.sciencedirect.com
endogenous progenitor and stem cells and possibly to foster survival of injured cells [6]. Unexpectedly, MSCs have recently been shown to modulate some T cell functions. Irrespective of the different experimental approaches attempted and the diverse immune functions analyzed in different studies, human and mouse T cells cultured in the presence of MSCs do not proliferate following mitogenic or antigenic triggering. These findings are supported by an arrest of cell division in the G0–G1 phase of the cell cycle associated with inhibition of cyclin D2 expression [7]. Similarly, inhibition of cell proliferation as a consequence of cell division arrest was also shown for B lymphocytes [8], natural killer (NK) cells [9] and dendritic cells (DCs) [10]. Together with this common mechanism of induction of ‘division arrest anergy’, MSCs have been shown to affect several other functions [11] of T cells, including cytokine secretion and cytotoxicity [12–15]; B cells, including maturation and antibody secretion [8]; NK cells, including cytokine production and cytotoxicity [9]; and DCs, including maturation, activation and antigen presentation [12,16,17]. The mechanisms underlying these effects are largely unknown but are likely to be mediated by direct cell-to-cell interactions and soluble factors. For T cells, several soluble factors, including transforming growth factor-b1 (TGF-b1), hepatocyte growth factor [18], indoleamine 2,3-dioxygenase (IDO) [19], nitric oxide (NO) [20] and prostaglandin E2 [12], have been proposed, mostly on the basis of in vitro experiments, but a clear consensus, based on in vivo experiments specifically targeting these molecules, is still lacking. Inhibition of lymphocyte proliferation has not been associated with the induction of apoptosis, with the exception of a single report claiming that MSCs induced apoptosis in proliferating but not in resting T cells, through an IDO-mediated mechanism [21]. These findings prompted the suggestion that MSCs might protect lymphocytes from the induction of apoptosis. Indeed, it was observed that MSCs can rescue thymocytes and centroblasts from spontaneous apoptosis and protect them from cell death induced by triggering of the Fas pathway [22]. MSCs have recently been demonstrated to support the survival of apoptotic neurons [23,24] and tumor cells [25] and of mammalian cells with nonfunctional mitochondria, through the transfer of mitochondria, thus rescuing aerobic respiration [26]. The main, well established features of MSCs are summarized in Box 1.
1471-4906/$ – see front matter ß 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.it.2007.03.001
220
Review
TRENDS in Immunology Vol.28 No.5
Box 1. MSCs: the facts MSCs are rare cells of non-hematopoietic stromal origin, residing in the BM [95]. However, MSCs have also been isolated from almost all other tissues [96]. MSCs are a heterogeneous population of multipotent stromal cells, which grow in culture as adherent cells able to differentiate into cells of the mesodermal lineage, including osteoblasts, adipocytes and chondrocytes [2]. BM-derived MSCs are part of the HSC niche, where they support hematopoiesis [97]. As a consequence, MSCs have been used to support allogeneic BM engraftment following transplantation [98]. MSCs inhibit the proliferation of T lymphocytes, B lymphocytes and NK cells, and impair DC maturation [11]. Because MSCs distribute to different organs [27], they could be used for gene delivery and tissue repair [5,99].
MSCs are influenced by the microenvironment Current evidence suggests that, following in vivo infusion, MSCs distribute with varying levels of efficiency to different organs [27]. Importantly, MSCs seem to migrate preferentially to sites of injury, suggesting that they can sense the local microenvironment, where they promote functional recovery [28]. Their capacity to cross the basement membrane is regulated by metalloproteases (MMPs) secreted by MSCs under the influence of inflammatory cytokines, such as tumor necrosis factor-a (TNF-a), TGF-b1 and interleukin (IL)-1b [29]. Moreover MSCs express on their surface a limited array of functional chemokine receptors, such as chemokine (C-X-C motif) receptor 4 (CXCR4), which have a pivotal role in their recruitment to the site of injury [30]. MSCs also express functional Toll-like receptors, which, following the appropriate triggering, modulate the capacity of MSCs to proliferate and differentiate [31]. In addition, in vitro studies have demonstrated that inhibition of lymphocyte proliferation requires crosstalk between MSCs and immune cells that can produce inflammatory cytokines, such as interferon (IFN)g [32] and IL-1b [33] following activation. These results suggest that, in vivo, the functional responses of transplanted MSCs recruited to the local microenvironment are probably dictated by danger signals, such as inflammation and hypoxia. MSCs in the treatment of immune-mediated and autoimmune diseases Based on the demonstration that MSCs could inhibit T cell proliferation, Bartholomew et al. [34] first demonstrated that MSC administration in vivo could prolong skin graft survival. This study, together with many others indicating the immunomodulatory activity of MSCs [11], paved the way for the clinical use of these cells in immune-mediated disorders. The most significant results so far have been obtained by Le Blanc and co-workers [35,36], whose pioneering work led to the successful treatment of severe, treatment-resistant, graft-versus-host disease (GvHD). The use of MSCs to prevent rejection of allogeneic grafts is still mostly limited to animal models, and the results obtained are conflicting. Indeed, successful prevention of allograft rejection might require administration of MSCs plus concomitant manipulation of the host immune system through NK cell depletion [37] or the induction of an appropriate myeloablative setting [38]. www.sciencedirect.com
Based on their capacity to modulate immune responses and promote tissue repair, MSCs have also been proposed as a treatment for autoimmune diseases [39]. Indeed, MSCs have been used for the treatment of experimental models of diabetes, rheumatoid arthritis (RA), systemic lupus erythematosus (SLE) and multiple sclerosis (MS). The intravenous injection of MSCs into diabetic NOD/scid mice (which are prone to diabetes and that have severe combined immunodeficiency) resulted in an increased number of pancreatic islets and insulin-producing b cells [40], and MSCs were also able to inhibit autoreactive T and B cells from BXSB mice, an experimental model of SLE [41]. The effects of MSCs in a murine model of RA are more contentious because the infusion of an MSC-derived cell line to treat collagen-induced arthritis was not effective [42]. However, a recent study in the same model demonstrated that allogeneic MSCs could prevent tissue destruction by dampening T cell attack on the cartilage [43]. Studies have been carried out on the effect of MSCs on experimental autoimmune encephalomyelitis (EAE), a murine model of human MS. It was observed that MSCs infused intravenously following the onset of disease can ameliorate EAE through the induction of T cell tolerance occurring at the level of the peripheral lymphoid organs [15]. More recently, these results were extended, when it was shown that the intravenous infusion of MSCs also suppresses pathogenic B cell responses in vivo [44]. Fluorescently labeled MSCs were detected inside the damaged nervous system, with no evidence of transdifferentiation into neural cells. Of relevance, a decrease in neuronal loss was observed in EAE mice treated with MSCs, suggesting a protective effect on damaged tissues [44] (Figure 1). Infusion of MSCs also seems to confer a protective effect on injured neurons in other neurological diseases, besides EAE. Indeed, MSCs have also been proposed for the treatment of spinal cord injury [45], stroke [46], amyotrophic lateral sclerosis [47] and neurometabolic diseases [48]. A common feature of these studies is that the therapeutic effect of MSCs does not seem to be associated with differentiation into neural cells but mainly to be the result of paracrine effects on the surrounding neural tissue. This indirect effect might also occur through the recruitment of local progenitors and their subsequent induction to differentiate into neural cells [6,49]. This interpretation is in line with other in vivo studies demonstrating that the beneficial effect of infused MSCs on disease course and pathology findings is mostly associated with anti-inflammatory activity coupled with a protective effect on damaged tissues, and is not due to MSC transdifferentiation into cells of the damaged organ. Administration of MSCs was found to enhance recovery of renal function in a model of acute renal failure [50], to increase insulin secretion in diabetic NOD/ scid mice [40] and to reduce inflammation and collagen deposition in a model of lung fibrosis [51]. In addition, a protective effect by MSCs on infarcted myocardial cells has been demonstrated, thus providing the rationale for their use in the treatment of ischemic heart disease [52]. MSCs in HSCT: crucial interactions with NK cells Because MSCs are the precursors of BM stromal cells, it was first envisaged that they could promote HSC
Review
TRENDS in Immunology
Vol.28 No.5
221
Figure 1. The proposed dual function of MSCs in the therapy of EAE. MS and its murine model EAE are characterized by progressive demyelination leading to progressive loss of neural axons. Different immune-mediated mechanisms involving T and B cells and macrophages are responsible for the progressive damage. Antibodies, T cell- or macrophage-derived cytokines (TNF-a; IFNg), or reactive oxygen intermediates (ROI) or nitrogen species (NO) released by activated macrophages are responsible for the effector mechanisms leading to nerve injury. Infusion of MSCs has been shown greatly to improve EAE. As proposed in this figure, in addition to their strong inhibitory effect (red inhibition bars) on T and B cell functions and on macrophage activity, MSCs exert a ‘trophic’ effect (green arrows) on both neurons and oligodendrocytes. The histology panel shows the striking effect of MSC infusion in EAE-affected mice (b,d,f) compared with untreated EAE controls (a,c,e) on demyelination (a,b), T cell infiltration (c,d) and neuronal damage, as revealed by a decrease in axonal loss (e,f). This positive, protective activity is likely to parallel the immunosuppressive effect and contribute greatly to the prevention of axonal loss and permanent disability [15,44,94]. These experimental data support the possible use of MSCs in the therapy of MS in humans.
engraftment. Indeed, this assumption proved to be correct, and MSC infusion was found to improve the success of HSCT and the clinical outcome [53], although, as discussed earlier, the underlying mechanisms are likely to be more complex than previously thought. In addition to the aforementioned facilitation of HSC engraftment following transplantation, the ability of MSCs to inhibit T cell functions, such as activation and proliferation in response to alloantigens, was a positive indicator for the use of MSCs in the treatment (or prevention) of GvHD [35]. In addition, results from Phase I trials support the feasibility and safety of the infusion of in vitroexpanded MSCs [54]. Thus, the use of MSCs could become a common approach in HSCT. Inhibition of T cell proliferation did not require MHC compatibility but was also mediated by allogeneic MSCs [55]. In addition, as a result of the low expression of MHC class I molecules, MSCs can escape detection by T cells [56]. However, it should be mentioned that recent in vivo data in mice challenge the concept of the immunoprivilege of MSCs because, at least in some instances, MSCs infused into allogeneic, MHC-mismatched mice were rejected [38,57]. Nonetheless, it is unlikely that this would occur in human HSCT settings because of the immunosuppressive regimen routinely delivered to prevent graft rejection. NK cells seem to have a major role in the successful clinical outcome of allogeneic HSCT, particularly in the haploidentical setting, in the treatment of acute myeloid leukemia (AML). NK cells are capable of killing various types of tumor, including AML [58,59]. The www.sciencedirect.com
susceptibility of AML cells to NK-mediated lysis in haploidentical HSCT was shown to correlate with a mismatch between the human MHC class I-specific inhibitory killer-Ig-like receptors (KIR) expressed by donor NK cells [58] and the human MHC class I alleles expressed by the patient (these NK cells are defined as ‘alloreactive’) [59–61] (Figure 2). Indeed, in the presence of KIR–human MHC class I mismatches (using appropriate conditioning protocols and high numbers of transplanted HSCs), >50% patients survived after five years, versus