Morphological and cytochemical aspects of capillary permeability.

Transport of plasma soluble constituents across the capillary wall is of primordial importance in cardiovascular physiology. While physiological experiments have concluded with the existence of two sets of pores, a large one responsible for the transport of proteins and a small one designed for the diffusion of small solutes, the morphological counterparts have yet to get general agreement. In this review, we present the different proposed paths within and between the endothelial cells that do allow passage of plasma constituents and may respond to the definitions established by physiological means. The vesicular system existing in endothelial cells has been the first transendothelial path to be proposed. Several data have demonstrated the involvement of this system in transport, although others have systematically brought controversy. One alternative to the vesicles has been the demonstration of membrane-bound tubules creating, in certain cases, transendothelial channels that would allow diffusion of plasma proteins and other constituents across the capillary wall. Access to this tubulo-vesicular system could be restrained by the stomatal diaphragm and facilitated by specific membrane receptors. Further, we have demonstrated for the first time with morpho-cytochemical tools, that the intercellular clefts are the site of diffusion for small molecules such as peptides having a molecular weight inferior to 3,000. For the fenestrated capillary bed, we have shown that fenestrae are the site through which plasma constituents cross the capillary wall. However, and in spite of the existence of these large open pores, the endothelial cells still display the tubulo-vesicular system involved in transport of large molecules and their intercellular clefts are also the site of diffusion of small molecules. Making consensus on the existence of an intracellular tubulo-vesicular system in non-fenestrated capillaries, responsible for the transport of large molecules by the endothelial cells, and understanding the rational for the fenestrated capillary to have three paths for transport--the fenestrae, the tubulo-vesicular system, and the inter-endothelial clefts--require further investigation.