Abstract
In addition to their extensive proliferation and differentiation potential, adult bone-marrow derived mesenchymal stem cells (MSC) have been recently demonstrated to display non-HLA restricted immunosuppressive capacities in vitro. Accordingly, preliminary clinical trials have begun using MSC for prevention or treatment of acute graft-versus-host disease (GVHD) in allogeneic hematopoietic stem cell transplantation setting. However, very few conclusive data are currently available concerning i) the mechanisms of MSC-mediated immunomodulation, ii) the migration and engraftment of infused MSC, and iii) the influence of cellular environment and soluble factors on their outcome and immunological properties, especially in the context of GVHD-associated inflammation.
Using a specific inhibitor of indoleamine 2,3-dioxygenase (IDO), we first confirmed the major implication of this enzyme in the MSC-dependent inhibition of PBMC and purified T cell alloantigen-induced proliferation. Such results were extended to clinical-grade MSC generated in a large scale closed culture system in the presence of clinical-grade fetal calf serum and basic fibroblast growth factor. In addition, MSC exerted the same dose-response anti-proliferative effect either they were third-party or autologous to MLR-responder cells and irradiated or non-irradiated. We could not point out in this system a role for PGE2, despite the use of two well-known inhibitors, or for membrane and soluble HLA-G molecules. Moreover, we shown that IFN-γ was sufficient to trigger, in a dose-dependent manner, the simultaneous induction of IDO expression, at the mRNA and protein levels, and IDO activity, evaluated through the measurement of tryptophan/kynurenine ratio by HPLC. On the contrary, purified peripheral blood monocytes constitutively expressed an inactive form of IDO, and IFN-γ operated only at posttranscriptionnal level in these cells. Antagonist anti-IFN-γ MoAb blocked the induction of IDO activity induced on MSC by allogeneic MLR supernatant, suggesting that IFN-γ is absolutely required for induction of IDO in MSC during immune response. TNF-α and LPS also modulated MSC immune functions through induction of a complex set of genes, such as those coding for CXCL9, CXCL10, and CCL5 inflammatory chemokines involved in GVHD. However, they could not induce IDO activity. CD40 marker was always detected at the beginning of clinical-grade MSC culture. It was lost after several passages but remained inducible by IFN-γ and TNF-α. Treatment of MSC by trimeric CD40L induced expression of several genes involved in immune reaction, such as BAFF, but not CD80 and CD86 costimulatory molecules or IDO.
In conclusion, IDO is an essential factor for immunosuppressive properties of clinical-grade MSC. Inflammatory cytokines, LPS, and contact with activated CD40Lpos T cells could markedly alter immunological properties of MSC. Such modifications must be taken into account for their further use as immunomodulator treatments in GVHD context.
Author notes
Corresponding author
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal