Mesenchymal Stromal Cells (MSCs) are being developed as a cellular therapeutic, and used clinically used for induction of immunomodulation and suppression of inflammation after application via the intravenous route. However, relatively little is known on how MSCs interact with the vessel wall to induce tissue-specific egress. To elucidate potentially underlying mechanisms, we analyzed human bone-marrow derived MSCs in parallel plate flow chambers and characterized their interaction with endothelial cells or immobilized endothelial ligands as to

  • tethering and rolling,

  • sudden arrest,

  • adhesion strengthening, and

  • transendothelial migration.

Flow cytometric analysis of MSCs confirmed the known expression predominantly of integrins α4, α5 and β1, and revealed only low detectable levels of a range of chemokine receptors including CCR1, 5, 6, 7 and CXCR1, 4 and 5. To investigate a potential role of chemokines in the adhesion process of MSCs, we first seeded MSCs without shear onto immobilized recombinant (r) VCAM-Fc fusion protein without or with co-immobilized chemokines in parallel plate flow chambers for 3 min, and subsequently started flow by stepwise increasing shear stress from 0.35 to 15 dynes/cmE2. Co-immobilization of various chemokines including CCL15/HCC-2, a known activator of CCR1 and 3, CCL20/LARC (of CCR6), CCL19/ELC (of CCR7), CXCL8/IL-8 (of CXCR1/2), CXCL12/SDF-1 (of CXCR4) and CXCL13/BCA-1 (of CXCR5) with rVCAM-1 increased shear-resistant binding of MSCs by up to four-fold compared to controls on rVCAM alone, with only moderate increases through CXCL8/IL-8 and CXCL12/SDF-1 and strongest effects through CCL19/ELC and CCL20/LARC. The reverse was observed in hematopoetic progenitor cells, which responded best to CXCL12/SDF-1. No or little adhesion of MSCs to rICAM-1 was observed under analogous conditions. We next chose to study CXCL12/SDF-1 and CCL19/ELC in more detail on intact endothelial cells. MSCs were flushed through parallel plate flow chambers on Human Umbilical Vein Endothelial Cells (HUVECs) at an initial shear stress of 0.35 or 0.5 dyn/cmE2 for 1 min followed by an elevated shear stress of 5 dyn/cmE2 for further 10 min. We found that rolling of MSCs was increased after pre-treatment of HUVECs with TNF-α but only few spontaneous arrests were obsereved under these conditions. However, when HUVECs were additionally overlaid with CXCL12/SDF-1 or CCL19/ELC, MSCs arrested firmly on the endothelium. Pretreatment of MSCs with function-blocking anti-CD44 or anti-VLA-4 antibodies, or pre-treatment of HUVECs with anti-E-selectin or anti-VCAM-1 antibodies partially (anti-CD44, -E-selectin) or completely (anti-VLA-4, -VCAM-1) suppressed arrest of MSCs. Analysis of adhesion strengthening on HUVECs as well as on immobilized rVCAM-1 confirmed the mediation of shear-resistant binding of MSCs by CXCL12/SDF-1 and CCL19/ELC. Pre-incubation of MSCs with pertussis toxin also suppressed both, sudden arrest and firm adhesion, indicating a role of signalling through G protein coupled receptor via Gαi. Finally, MSCs were able to transmigrate chemokine-coated HUVECs but not untreated. We conclude that MSCs respond to individual chemokines with different efficiencies to induce arrest and firm adhesion before their transendothelial migration. MSCs were found highly responsive to chemokines which physiologically act on lymphocytes to enter secondary lymphoid organs or sites of inflammation. This points to a preference of MSCs to follow lymphocyte egress pathways and reach sites of inflammation to induce immunomodulation and –suppression.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author