Influence of microbial products and inflammation on the function of mesenchymal stromal cells isolated from different sources

MSC can be isolated from different tissue sources including bone marrow (BM), adipose tissue (AT) and Wharton’s Jelly (WJ). Although fulfilling the ISCT criteria required to be recognized as MSC, MSC from these different sources could disclose some differences taking into account their different anatomical origin and ontogeny as well.

In the present work, we investigated the influence of MSC source on their immunosuppressive as well as differentiation properties. We further extended our study to the role of the microenvironment (infection and inflammation) on these features.

We show that BM-MSC express Toll-like receptors (TLR) from TLR1 to TLR6. In an inflammatory environment, TLR2, 3 and 4 are significantly upregulated. By upregulating TLR3 and TLR4 transcription, inflammation increases BM-MSC responsiveness to LPS (TLR4 ligand) and poly(I:C) (TLR3 ligand) leading to a pro-inflammatory shift of their cytokine profile. The effect of TLR ligation on BM-MSC osteogenic potential is donor dependent. Inflammation as well as stimulation with LPS and poly(I:C) result in a decrease of BM-MSC immunosuppressive capabilities.

We further observed that BM-, AT- and WJ-MSC do not have the same pattern of TLR expression and consequently do not respond the same way to bacterial or viral infection. WJ-MSC do not express TLR4 and although TLR3 is present at the protein level it is not functional as its ligation do not trigger cytokine expression. Inflammation modulates this TLR pattern expression by upregulating TLR3 in all three MSC types and TLR4 only in BM-MSC. TLR ligation increases the production of inflammatory cytokines in BM- and AT- but not in WJ-MSC and augments anti-inflammatory cytokines in AT-MSC. Although inflammation increases in all MSC types the secretion of inflammatory cytokines, additional TLR triggering does not further affect WJ-MSC. The immunosuppressive potential of WJ-MSC on mixed leucocytes reaction (MLR) is not affected either by inflammation or by TLR triggering.

On the differentiation side, WJ-MSC has the lower potential to differentiate into osteoblast as compared to BM- and AT-MSC, as revealed by alkaline-phosphatase (ALP) activity and by measuring extracellular Ca2+ deposits. However, inflammation is able to strongly increase the osteogenic differentiation of WJ-MSC as calcification and ALP activity appears as early as at day 7. However this latter enzymatic activity remains much lower than that disclosed by BM-MSC. TLR3 or TLR4 triggering does not affect the osteogenesis of WJ-MSC while it increases it in AT- and also, although to lesser extent, in BM-MSC.

Our work establishes that the source from which MSC is derived is of major importance for the design of MSC based immunointervention. WJ-MSC appear to be the most attractive cell type when an immunosuppressive action is required in an inflammatory or infectious context. Although WJ-MSC are poorly osteogenic, a complete osteogenic differentiation can be obtained under inflammatory conditions. Taking into account their easy accessibility as well as their huge proliferative potential, these data open an avenue for using these cells in regenerative medicine particularly in clinical settings where chronic inflammation or infection have to be considered.