Vascular permeability and transvascular fluid and protein transport in the dog lung.

We used steady state lymph-to-plasma concentration ratios of six endogenous protein fractions (effective hydrodynamic radii of 37, 40, 44, 53, 100, and 120 A) to estimate pulmonary capillary permeability characteristics. Pulmonary lymph was collected from an afferent lymphatic to the left tracheobronchial lymph node, and left atrial pressure was elevated in steps until the lymphatic protein concentration obtained a constant value. Lymph flow and transvascular protein flux increased with each increase in left atrial pressure. Convective flux was the predominant mode of transport for the smaller fractions, and solvent-drag reflection coefficients increased as lymph flow increased. This dependency on lymph flow (capillary filtration) indicates a heteroporous membrane system. For lymph flows greater than five times control, the solvent-drag reflection coefficients were: 0.59 +/- 0.11, 0.52 +/- 0.06, 0.66 +/- 0.05, 0.70 +/- 0.05, 1.01 +/- 0.04, and 1.05 +/- 0.03 for the six fractions. Osmotic reflection coefficients estimated from the minimal lymph-to-plasma concentration ratios were: 0.50 +/- 0.03, 0.59 +/- 0.02, 0.67 +/- 0.04, 0.72 +/- 0.03, 0.94 +/- 0.01, and 0.96 +/- 0.01 for the six protein fractions. The osmotic reflection coefficients are consistent with a two- "pore" exchange model possessing equivalent pore radii of 80 and 200 A. Theoretical considerations indicate that only the two largest protein fractions (100 A and 120 A radii) achieved filtration-independent concentrations in pulmonary lymph, even at the highest filtration rates. This suggests that the reported osmotic reflection coefficients of the four small protein fractions underestimate their true values.