Endothelial gaps as sites for plasma leakage in inflammation.

OBJECTIVE: In 1961, Majno and Palade proposed that plasma leakage in acute inflammation caused by histamine, serotonin, or bradykinin results via gaps that form between endothelial cells of postcapillary venules. Now the relevance of endothelial gaps in plasma leakage is being questioned. The purpose of this review is to summarize experimental evidence from our studies showing that endothelial gaps participate in plasma leakage in inflammation.
METHODS: Using neurogenic inflammation as a model of plasma leakage in acute inflammation, we compared five methods to determine whether endothelial gaps form in the microvasculature of the rat trachea. 1) Endothelial cells borders and gaps were stained with silver nitrate and visualized by light, scanning, and transmission electron microscopy. 2) The luminal surface of endothelial cells was examined by scanning electron microscopy. 3) The luminal surface of endothelial cells was stained with a biotinylated lectin and avidin-biotin-peroxidase histochemistry, and then was examined by differential interference contrast microscopy. 4) Endothelial junctions were reconstructed from serial sections photographed by transmission electron microscopy. 5) Leakage was measured after perfusion of lectins or tracers through aldehyde-fixed vessels in situ.
RESULTS: The results from the five methods used in this system were consistent with the formation of gaps between endothelial cells. Endothelial gaps were rare or absent under baseline conditions, but appeared with the onset of plasma leakage and had a distribution that matched the distribution of leakage. Gaps had a complex morphology and were accompanied by fingerlike cell processes, which may anchor adjacent endothelial cells to one another and participate in gap closure. In contrast to normal vessels, vessels that were leaky in life continued to leak after aldehyde fixation, in evidence that, once formed, the leakage pathway did not require energy-dependent membrane movement or vesicle shuttling. Holes through endothelial cells were less than 1% as frequent as intercellular gaps.
CONCLUSIONS: Taken together, the results show that endothelial gaps are a consistent feature of leaky vessels in the model system we studied, and are not an artifact of a particular method. The morphological complexity of the openings and accompanying fingerlike cell processes and overlapping endothelial cell borders make gaps difficult to distinguish from transcellular holes in thin sections viewed by transmission electron microscopy. However, scanning electron microscopic observations show that most of the openings in leaky venules are intercellular gaps, not transcellular holes. The formation and closure of gaps are likely to be energy-dependent, but the process of plasma leakage is not, provided there is adequate driving force for extravasation. The cellular mechanisms of gap opening and closure remain to be elucidated.