Heart capillary permeability to lipid-insoluble molecules.

1. A study has been made, in the isolated, beating dog heart perfused with blood, of the transcapillary exchange of the following substances: [(3)H]water, (22)Na, (86)Rb, (36)Cl, [(14)C]urea, [(3)H]glycerol, [(3)H]glucose, [(14)C]sucrose and [(3)H]inulin.2. The method used to study the exchange was the ;indicator diffusion' technique. It consists in a rapid arterial injection of a mixture containing a diffusible and a non-diffusible molecule, followed by a rapid split collection of the venous outflow, up to 30 sec. The fractional extraction, E, of the diffusible substance was obtained by comparing the relative concentrations of both tracers in injected medium and in each venous sample.3. E for [(3)H]water was the highest (0.90 +/- 0.3), and it did not vary with flow. All other molecules had values for E that decreased as flow increased.4. Capillary permeability constant, P, was estimated from PS = - F ln (1 - E), in which S is the surface area of exchange and F is the blood perfusion rate. To test the validity of the equation, E was measured at different blood perfusion rates. It was found that the equation did not apply at relatively low flows for the more diffusible substances.5. The average values of P estimated for inulin, sucrose, glucose, glycerol and urea were 0.27, 0.8, 1.0, 1.5 and 3.1 x 10(5) (cm/sec), respectively. The ratio P/D (in which D is the free diffusion in water constant) was the same for all these substances. This can be interpreted as showing that if pores exist in the capillary endothelium, they must be larger than 80-100 A diameter. It is concluded that the pores could actually be the intercellular slits as previously suggested by electron microscope studies. Endothelial cell participation in the exchange appears to be small except for [(3)H]water.