Abstract
Bone marrow mesenchymal stem cells (MSC) have potent immunosuppressive properties and are being evaluated in human trials of graft versus host disease (GVHD). The nature of their suppressive capacity is not well understood but attributed to their stem cell function. Evidence that adult stromal cells such as fibroblasts (Fb) also modulate T cell functions has important implications for immunoregulation and cellular therapy. We have investigated the phenomenon of MSC-mediated immunosuppression by comparing MSC with Fb of different origins in in vitro assays of T cell suppression and modulation. We have then isolated RNA from paired samples of dermal Fb and MSC from 6 healthy volunteers for comparative gene expression studies. Adherent Fb were isolated from digested dermis, synovium and lung. MSC were obtained from BM aspirate. Fb from the dermis, lung and synovium possess potent immunomodulatory properties. Fb suppress allogeneic T cell activation by autologously derived cutaneous antigen presenting cells and other stimulators. Fb-mediated suppression through soluble factors is dependent on IFNγ from activated T cells. IFNγ induces indoleamine-2, 3, dioxygenase in Fb with accelerated tryptophan metabolism being partly responsible for suppressing T cell proliferation. T cell suppression is reversible and exposure to stromal cells during activation reprogrammes T cells, increasing secretion of interleukin-4 2.3 fold, interleukin-10 4.3 fold and interleukin-13 15 fold (means of 4 experiments) upon restimulation. Increased Th2 polarization by stromal cells is associated with amelioration of pathological changes in an in vitro human GVHD model. Our findings also show that Fb from different sources are indistinguishable from MSC with respect to morphology, phenotype, growth and differentiation capacity in vitro. Clonogenicity (ratio of CFU to CD73+CD45- cells) of Fb and MSC are similar (range 0.2 to 0.46 CFU/cell) proving that the immunosuppressive effects of Fb from adult tissues are not due to the expansion of rare ‘stem’ cells. Using paired isolates of dermal Fb and MSC to control for inter-individual variation, we were able to define consistent differences in gene expression. Microarray assays were performed using a Human Genome Focus Affymetrix array and analysed with GeneSpring GX. 143 of 9600 probesets showed reproducible differences in transcript levels between dermal Fb and MSC. Probesets upregulated in MSC include genes encoding immunomodulatory mediators: vascular endothelial growth factor (7 fold), hydroxysteroid 17β dehydrogenase (10 fold) and jagged1 (5 fold); extracellular adhesion molecules: proteoglycan1 (264 fold), vascular cell adhesion molecule (175 fold), transglutaminase (67 fold) and procollagen (8 fold); and developmental regulators in the Hedgehog and Wnt signalling pathways. Our findings are further evidence that immunosuppression is a generic property of Fb isolated from several sources and not restricted to MSC. We have for the first time identified a differential expression profile of MSC compared with Fb. These differences may not confer unique in vivo immunosuppressive properties and the potential of Fb as an alternative source of cellular therapy remains untested.
Author notes
Disclosure: No relevant conflicts of interest to declare.