Summary Mesenchymal stem cells (MSCs) have an immunosuppressive effect and can inhibit the proliferation of alloreactive T cells in vitro and in vivo. Cotransplantation of MSCs and hematopoietic stem cells (HSCs) from HLA‐identical siblings has been shown to reduce the incidence of acute graft‐vs.‐host disease. MSCs are heterogeneous and data on the inhibitory effects of different MSC subsets are lacking. The antigen Stro1 is a marker for a pure primitive MSC subset. We investigated whether Stro‐1‐enriched induce a more significant suppressive effect on lymphocytes in a mixed lymphocyte reaction (MLR), and whether this action is related to a specific gene expression profile in Stro‐1‐enriched compared to other MSCs. We demonstrated that the Stro‐1‐enriched population elicits a significantly more profound dose‐dependent inhibition of lymphocyte proliferation in a MLR than MSCs. One thousand expanded Stro‐1‐enriched induced an inhibitory effect comparable to that of 10 times as many MSCs. Inhibition by Stro‐1‐enriched was more significant in contact‐dependent cultures than in noncontact‐dependant cultures at higher ratio. The Stro‐1‐enriched inhibitory effect in both culture types was linked to increased gene expression for soluble inhibitory factors such as interleukin‐8 (IL‐8), leukemia inhibitory factor (LIF), indoleamine oxidase (IDO), human leukocyte antigen‐G (HLA‐G), and vascular cell adhesion molecule (VCAM1). However, tumor growth factor‐β1 (TGF‐β) and IL‐10 were only up‐regulated in contact‐dependant cultures. These results may support using a purified Stro‐1‐enriched population to augment the suppressive effect in allogeneic transplantation.

Mesenchymal stem cells (MSCs), have been shown to elicit immunosuppressive effect on allogeneic lymphocyte response. However, MSCs are heterogeneous and data on the inhibitory abilities of different MSC subsets are lacking. In the present study, we selected Stro-1+ cells from human bone marrow and evaluated the inhibitory capability of this MSC subset in mixed lymphocyte reactions (MLRs) or in mitogen stimulation asssays, in comparison to that of Stro-1− cells. To evaluate the two MSC subsets for immunomodulation in vitro, we added 1,000–30,000 Stro-1+ or Stro-1− cells to MLR at the beginning of the experiment. When comparing the inhibitory effects of the two subsets, PBLs proliferation was significantly more inhibited by Stro-1+MSCs (11.0%–63.7%) than by stro-1−MSCs (35.5%-106%) (P<0.01). Furthermore, as few as 1,000 Stro-1+ MSC could inhibit lymphocyte proliferation more effectively than 10 times more (10,000 cells) Stro-1−cells. As it was observed with the mixed lymphocyte reaction, suppression of the response to the mitogen also occurred in a dose dependent fashion, but to a lesser extent with the Stro-1−cells (25.5%–80.1% vs 7.5%–38.4% in Stro-1+cells) (P<0.05). To investigate whether the difference of suppressive effect that we observed between Stro-1+ and Stro-1− cells, still exist when MSC subsets are separated physically from PBL, we performed MLR in the upper chamber of a transwell and we seeded the lower chamber either with Stro-1+ or Stro-1− cells. The inhibitory effect of Stro-1+ cells was significantly more profound than the one observed when Stro-1− cells were used in the Transwell culture system (p<0.05) (Figure 3), demonstrating that one or several soluble factors was involved in production of different suppressive effects. Cytokine and chemokine genes, IL-10, TGF-β1, SDF-1, SCF and IL-6 expression were evaluated in both MSC subsets by quantitative RT-PCR. Low levels of IL-6, SCF, SDF-1 were observed in Stro-1+, which induced a fold increase around 1 (0,96 ± 0,32; 0,96 ± 0,24; 0,96 ± 0,24), indicating that there is no signifiant difference of these genes expression between the two MSC subsets. However, we observed in Stro-1+ a decreased gene expression for IL-10 (0,24 fold ± 0,59; p <0,05) and for TGF b1 (0,43 fold ± 0,32; p <0,05). This finding suggested that the candidate T-cell inhibitory factors TGF b1, IL-10, which are lower expressed in Stro-1+ cells, are not responsible for the more profound inhibition of immunoreactivity by Stro-1+ cells. We show here that significant differences do exist within these two subsets. Stro-1+ cells inhibit lymphocyte proliferation significantly more profoundly than Stro-1−cells. The difference is in part mediated by soluble factors, but not IL-10 and TGF-β1. These results point to the notion that Stro-1+ cells can elicit more powerful immunosuppressive ability and a pre-selection of Stro-1+MSC for clinical use may be advisable. These findings suggest that pre-selection of MSC before clinical use might produce more effective immunosuppression in different therapeutic applications, especially in clinics for the prevention of graft versus host disease (GVHD).