Human Mesenchymal Stem Cells Exert Immunosuppressive Capacities by Induction of Regulatory T Cells

It is widely accepted that Mesenchymal Stem Cells (MSCs) exhibit immunosuppressive capacities in vitro and in vivo. In an effort to understand the mechanism of suppression, co-cultures of MSCs with several types of immune cells have been studied. MSCs inhibit B- and T cell proliferation and inhibit the generation, maturation and function of dendritic cells. In the present study, we evaluated the effect of MSCs on the expansion and function of regulatory T cells (T regs). Human bone-marrow derived MSCs were cultured and expanded in low-glucose DMEM containing 10% FCS and frozen until used in experiments. Freshly isolated, MACS selected human CD4⁺CD25⁺ T regs were cultured in IMDM, supplemented with 10% pooled human serum and 300 units/ml IL-2, in the presence or absence of irradiated MSCs (60 Gy) (ratio MSCs: T regs = 1:5). Results are derived from 3 to 15 independent experiments. In the presence of IL-2 and MSCs, the percentage of FOXP3⁺ CD4⁺CD25⁺ T cells increased from 26.8% ±2.2 (no MSCs) to 42.3% ±2.4 (with MSCs) over a period of 5 days, representing a 1.6 (±0.1)-fold induction. Moreover, a distinct CD4⁺CD25⁺ population with high FOXP3 expression appeared after 5 days of culture in the presence of MSCs (23.3% ±2.5 in the presence of MSCs versus 7.2% ±1.0 in the absence of MSCs). This CD4⁺CD25⁺FOXP3^hi population was not observed after co-culturing MSCs and CD4⁺CD25⁻ T cells (1.7% ±0.6). To show that the MSC-induced T regs were functionally suppressive, freshly isolated CFSE-labeled CD4⁺ T cells were stimulated with PHA (0.8 μg/ml). After 3 days, about 40% of the T cells had undergone one or more cell divisions as measured by CFSE dilution. Addition of MSC-induced T regs (in a 1:1 ratio) resulted in a 50% reduction of the proliferation of CFSE-labeled T cells (down to 15–20% of cells undergoing one or more cell divisions). Control CD4⁺CD25⁺ T cells cultured in the absence of MSCs did not suppress T cell proliferation. These results indicate that MSC-induced CD4⁺CD25⁺ FOXP3^hi cells exert regulatory function. To study whether the induction of T regs by MSCs was dependent on cell-cell contact, co-culture experiments were performed in transwells where MSCs were physically separated from T cells. CD4⁺CD25⁺ T cells co-cultured with MSCs in transwells showed a 1.4 (±0.1)-fold increase in the percentage of FOXP3^hi cells, in comparison with T cells cultured in the absence of MSCs. Similarly, direct co-cultures of MSCs and T cells resulted in a 1.8 (±0.2)-fold increase in the percentage of FOXP3^hi T regs. In addition, medium derived from co-cultures of MSCs and CD4⁺CD25⁺ T cells and added to freshly isolated CD4⁺CD25⁺ T cells resulted in a 2.6 (±0.6) fold increase in the percentage of CD4⁺CD25⁺FOXP3^hi T regs, implicating that FOXP3^hi induction by MSCs was mediated by soluble factors. Since T Cell Receptor-stimulated naïve T cells (as well as natural T regs, generated in the thymus) require both IL-2 and TGF-β to become induced T regs, we analyzed the involvement of TGF-β. Addition of the pharmacological inhibitor of the TGF-β receptor (SB431542) only marginally reduced FOXP3 induction in the presence of MSCs, suggesting that MSC-mediated expression of high FOXP3 levels requires alternate or additional cytokines. In conclusion, we show that

• MSCs promote the induction of CD4⁺CD25⁺ T cells that express high levels of FOXP3 and

• these MSC-induced T regs suppress proliferation of PHA-stimulated CD4⁺ T cells.

These effects are mediated by soluble factors produced during the co-culture of MSCs and T cells. The cytokines involved are presently unknown, but likely do not involve TGF-β.