Neutrophils use an ordered sequence of known adhesive interactions to bind to cytokine-stimulated endothelium. However, the mechanisms regulating migration through and under the endothelial cell monolayer are uncertain. We hypothesised that constituents of the basement membrane might influence neutrophil migration, for instance by providing a reservoir for chemotactic agents or a haptotactic adhesive substrate.

To test this hypothesis, confluent monolayers of first passage human umbilical vein endothelial cells (HUVEC; isolated from umbilical cords obtained with informed consent) were cultured for 1, 3, 5, 10, 15 or 20 days, and then stimulated with 1, 10 or 100 U/ml of the inflammatory cytokine tumour necrosis factor-α (TNF) for 4 h prior to the experiment. Neutrophils were allowed to bind to the endothelial surface, and were directly observed by phase-contrast video-microscopy as they migrated over and under it. Several differences were noted in the responses when, for instance, day 1 and day 20 cultures were compared (all data are means ± S.E.M. of 3 to 5 experiments): (i) day 20 cultures induced higher levels of neutrophil adhesion at 1 U/ml TNF (67 ± 4 % of neutrophils adherent vs. 20 ± 5 % for day 1 cultures) or 10 U/ml TNF (61 ± 3 % of neutrophils adherent vs. 35 ± 3 % for day 1 cultures). ANOVA showed significant effect of culture time on adhesion (P < 0.01); (ii) the percentage of adherent neutrophils transmigrating was higher for day 20 cultures than day 1 (35 ± 3 % of adherent neutrophils transmigrated at 1 U/ml vs. 7 ± 2 % for day 1 cultures). ANOVA showed significant effect of culture time on transmigration (P < 0.01); (iii) at 100 U/ml TNF, adhesion and efficiency of transmigration were similar for day 1 and day 20 cultures, but the transmigrated neutrophils moved more slowly for the day 20 cultures (7 ± 0.3 µm/min vs. 12 ± 0.1 µm/min for day 1 cultures; P < 0.05 by Student’s t test). Using similar cultures, enzyme-linked-immuno-sorbent-assay (ELISA) showed increasing deposition of basement membrane components (e.g., collagen type IV and laminin) with culture time. In separate experiments, varying concentrations of purified basement membrane proteins (collagen type IV, laminin and fibronectin) were adsorbed to plastic dishes, and their ability to bind activated neutrophils was tested. Interestingly, at surface concentrations comparable to those found under cultures (verified by ELISA), neutrophil adhesion was markedly augmented by these proteins (e.g., ratio to control adhesion = 4.0 ± 1.0 for laminin; mean ± S.E.M. from 3 comparisons).

These data support the concept that substances laid down in the basement membrane influence neutrophil behaviour on endothelium, either by directly interacting with migrating cells or indirectly modifying the responses of the endothelial cells growing on them.