Permeability of lung capillaries to macromolecules in foetal and new-born lambs and sheep.

1. The permeability of lung capillaries to macromolecules was investigated in immature and mature foetal lambs, new-born lambs and young sheep. The placental circulation of the foetal animals was maintained intact after delivery by Caesarian section. New-born lambs and sheep were mechanically ventilated. Samples of plasma and lymph that had drained from the lung via the thoracic duct were collected over a period of 1-5 hr.2. The proteins in plasma and lymph samples were separated by fractionation on columns of Sephadex G-200. Plasma yielded three peaks of protein concentration. The K(av) value of each peak was determined, and, by calibrating the columns with known proteins, the mean radius of equivalent sphere (a) of the proteins in peak I was shown to be similar to that of fibrinogen >/= 110 A, peak II to gamma-globulin approximately 54 A and peak III to albumin approximately 34 A. Lung lymph contained the same three constituent peaks as plasma but in lower concentration. In all four groups mean lymph/plasma concentration (L/P) ratio was significantly different for each of the three peaks, being lowest for the largest molecules (peak I) and highest for the smallest (peak III).3. In five mature foetal lambs polydisperse polyvinylpyrrolidone labelled with (125)I ([(125)I]PVP) was injected I.V. early in the experiment: count rates in fractionated samples showed for plasma a continuous decline with time after injection, and for lung lymph an increase to a maximum then a decline. Steady-state L/P ratios for eleven fractions of PVP of differing molecular size ranging from 110 to 17 A were derived by compartmental analysis. For a given molecular size PVP L/P ratios were similar to protein L/P values.4. The regression of PVP L/P ratio on K(av) was linear (correlation coefficient r = 0.99), and the slope of the regression of protein L/P ratio on K(av) was significantly steeper for new-born lambs than for mature foetuses (P < 0.025) and sheep (P < 0.005), and steeper for immature foetuses than sheep (P < 0.01).5. PVP and protein L/P ratios (mature foetuses) plotted against a showed a sigmoid relation with agreement between the two sets of L/P ratios. The goodness of fit between our experimental results and Landis & Pappenheimer's (1963) capillary pore theory (eqn. (1)) was examined: L/P ratios for the larger molecules (>/= 75 A) appeared to be too high. By recalculating ratios on the assumption that the largest molecules (110 A) escape unrestricted from the capillary via leaks, the discrepancy disappears.6. Values for pore radius (r), and pore area per unit path length (A/Deltax) have been calculated for each of the four groups; r ranged from 90 to 150 A, A/Deltax from 3.3 to 0.2 cm x 10(3).kg(-1). In new-born lambs the value of r was significantly smaller, and A/Deltax larger than that of any other group. The inferences to be drawn from these results are discussed.