2B). These trends are similar to those
we previously reported with the two filter model (McGettrick et al., selleck screening library 2010). Lymphocytes are known to take longer to migrate across these artificial filters (hours) than to transit through the EC (McGettrick et al., 2009a). Endothelial cells in these filter models are able to respond to cytokine treatment as expected, up-regulating surface expression of the adhesion receptors, E-selectin and VCAM-1 and producing chemokines, including CXCR3 ligands at the mRNA level (Supplemental Fig. 2). A combination of prolonged settling periods and non-specific delays in transit across the filters are likely to explain the similarities in migration observed between cultures treated with or without cytokines. We also analysed onward migration through the layer of fibroblasts. When fibroblasts were cultured alone, lymphocytes migrated through the monolayer quite readily, with a tendency for more to migrate when the fibroblasts
have been treated with Sirolimus cytokines (Fig. 2C). Interestingly, PBL migrated across fibroblasts in the co-cultures much less efficiently than in the mono-cultures (Fig. 2C). The above results raised the question whether fibroblasts would similarly increase the migration of PBL through endothelial cells when presented in a 3-D matrix, and/or influence progress of PBL through that matrix. To test this, we designed a construct in which we could visualise PBL migration through and away from the EC, and then through a collagen gel Liothyronine Sodium incorporating fibroblasts (Fig. 1B). For unstimulated cultures,
PBL were allowed to settle for 3 h on the EC to allow adequate levels of adhesion for migration analysis. Under these conditions, fibroblasts promoted PBL adhesion, but significantly reduced the efficiency of subsequent transendothelial migration of the adherent cells, and also tended to inhibit the entry of those cells that had crossed the endothelium into the gel (Fig. 3; clear bars). After treatment with cytokines, only 10 min settling was needed to obtain efficient adhesion to EC (as previously described; McGettrick et al., 2009a). However, in this case fibroblasts had little effect on the ability of PBL to adhere (Fig. 3A; filled bars). They retained a tendency to reduce migration through the endothelium and penetration into the underlying gel (Fig. 3B–C; filled bars), but neither effect was statistically significant. Thus in this model, fibroblasts failed to promote transendothelial migration as seen in the filter model, but did retain a tendency to hinder migration of cells after crossing that barrier. In some co-cultures on gels, we observed that the endothelial monolayer retracted and/or some cells detached during the culture (Supplemental Fig. 3).