Modelling the structural pathways for transcapillary exchange.

The ultrastructural pathways and mechanisms whereby endothelial cells and the clefts between the cells modulate capillary permeability to water and solutes have been a central unresolved question in microvessel transport since the early 1950s. Freeze-fracture studies and ultrathin serial sections have demonstrated that endothelial cells are joined by an array of junctional strands which are interrupted at intervals, allowing for the passage of water and solutes, whereas cytochemical studies have indicated that the endothelial surface and portions of the wide part of the cleft contain matrix components. Neither constricted slit models based on the classic pore theory nor fiber matrix models are able to explain the large body of existing permeability measurements. In this review, we shall describe new three-dimensional modelling approaches which have resulted in a major revision of current ideas about the pathways for water and solutes through the junction strand and the structures that determine the molecular filter. For frog mesentery capillaries, these models predict (i) that the primary pathway for small ions is a previously unrecognized family of 2nm small pores that are distributed along the length of the junction strand; (ii) that the primary pathway for water and intermediate-sized solutes (1-3.5 nm radius) is an infrequent 150 nm long orifice-like pore whose height is the same as that of the wide part of the cleft; (iii) that the sieving structure for these solutes is a fiber layer, typically 100 nm thick, which extends from the surface into the entrance region of the cleft and (i.v.) that the interpretation of low molecular weight tracer studies to define the permeability pathways depends on the time-dependent filling of the extravascular space.