Calculations relating to the passage of fluid and protein out of arterial-limb fenestrae, through basement membranes and connective tissue channels, and into venous-limb fenestrae and lymphatics

Abstract

An equation is deduced showing the total passage of protein through a pore, in relation to the relative concentrations and hydrostatic pressures on both sides of the pore. This is a mixture of the filtration and diffusion equations, which allows for nonlinearity in the profile of the protein concentrations across the pore. It is shown that tracers obey a similar equation. This equation and the values obtained by direct measurement for fenestrae, and indirectly for the basement membranes and for the channels through the connective tissue in general, are used in calculations of the concentration profiles of protein across the tissue. The resultant colloidal osmotic pressure profiles and hydrostatic pressure profiles are also obtained. These results indicate that proteins pass out of arterial-limb fenestrae and most of them can reenter the blood vessel via the fenestrae on the venous limbs; smaller but qualitatively vital amounts also pass to the lymphatic system. It is these latter which appear to keep the mean protein concentrations in the tissues low enough so that the balancing hydrostatic pressures in the tissues remain negative. Thus oedema is avoided. If the lymphatics are obstructed, this situation does not occur and the calculated mean connective tissue pressures become positive, indicating that oedema is likely. It is shown that fenestrae with diaphragms are unlikely to contribute significantly to vascular permeability and that the basement membranes, while they appear insignificant for the regulation of the passage of water, are extremely important in regulating the passage of protein. If they did not have this effect, these calculations indicate that the concentration of protein in the tissues would be extremely high, with high positive connective tissue hydrostatic pressures and inevitable oedema. Three models (single, hierarchical, and random-net) for the connective tissue channels are examined. It is found that only the latter gives reasonable results.