INTERACTION OF POLYMORPHONUCLEAR LEUKOCYTES AND ENDOTHELIAL CELLS : FUNCTIONAL CONSEQUENCES

The mechanisms involved in the accumulation of polymorphonuclear leukocytes (PMNs) in an inflammatory reaction are complex. A key phase in this process is the attachment of the PMN to the microvascular (venular in most tissues) endothelial cell, initiated by the extravascular generation of a chemical mediator. Experiments in vitro suggest that mediators, such as C5a, may act in vivo by stimulating the increased expression of the CD18 complex on the surface of the PMN within the venule lumen (1), whereas IL-1 may act by causing the expression of an adhesive molecule on the endothelial cell (2). In vitro the former process is rapid whereas the latter is slow in onset. We have measured the local accumulation of intravenously-injected Ulln-PMNs in response to intradermally-injected mediators in the rabbit, in order to investigate possible mechanisms in vivo. PMN accumulation was found to be rapid in onset in response to C5a, the rate of accumulation falling progressively to low levels by 4 hours. In contrast PMN accumulation in response to IL-1 was slow in onset, reaching a peak rate at 3-4 hours. Intradermal injection of the vasodilator prostaglandins PGI₂; PGE₂ and the neuropeptides VIP and CGRP caused a marked potentiation of the rate of leukocyte accumulation. PMN accumulation induced by C5a was associated with increased microvascular permeability, as indicated by the leakage of intravenously-injected ¹²⁵I-albumin with a time-course in parallel with the rate of PMN accumulation enhanced by intradermally-injected vasodilators. Depletion of circulating PMNs abolishes these responses to C5a (3). In contrast, leukocyte accumulation induced by IL-1 was associated with little plasma protein leakage, even in the presence of intradermal vasodilators. This observation indicates that PMN emigration itself does not lead to increased microvascular permeability. C5a, but not IL-1, may stimulate emigrating PMNs to secrete an endogenous factor that increases permeability by an action on endothelial cells (3). This factor does not appear to be the phospholipid PAF (4). In contrast to the enhancing effects of local PGI₂, intravenously-infused PGI₂ inhibited PMN accumulation induced by C5a and IL-1, and plasma protein leakage induced by C5a (5). This effect is probably mediated by elevation of cyclic AMP in intravascular PMNs. We have shown that C5a stimulation of PMNs in contact with endothelial cells in vitro induces endothelial cell PGI₂ secretion (6). Thus, PGI₂ may be part of a negative feedback system in vivo to control interactions between PMNs and endothelial cells.

These observations provide some clues to the intricacies of mechanisms of leukocyte accumulation in vivo.