Editor—In response to the commentary by Professor Walters,1 we must first correct him in referring to our model for the normal alveolus2 as “dry” when, in fact, it is based on many classic morphological studies3 4demonstrating fluid largely confined to “pools” at the septal corners. In proposing that an oligolamellar lining of surface–active phospholipid (SAPL) adsorbed to epithelium “pushes water aside” from the gas–exchange surface, we agree that it is difficult to prove direct binding conclusively. However, any intervening aqueous layer is no more than that normally sandwiched between adjacent planes of polar groups in such structures (fig 1). Moreover, this physiological milieu contains mobile cations which can neutralise the negative phosphate ions in the SAPL molecules to render them cationic, facilitating their tight binding into a very effective molecular barrier.5Hence SAPL is pseudocationic.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Electron micrograph of the alveolar wall of a 3 month old infant displaying an oligolamellar lining of SAPL apparently bound to epithelium. Note how it forms a continuous barrier spanning an intercellular junction (arrowed). The bar represents 50 nm.

In citing evidence to support the conventional concept of a continuous liquid layer separating the surfactant lining from alveolar epithelium, Professor Walters refers to the recent study by Bastacky, Clements, and others,6 who do, indeed, demonstrate such a liquid layer. However, they have created a totally artefactual situation by pre-inflating the lungs to 15 cm H₂O—just enough to squeeze out all blood—before freezing and fixation. The resulting alveolar surface is totally concave with respect to air whereas, in normal air filled lungs, scanning electron microscopic photographs demonstrate how at least 60% of the alveolar surface is convex as red cells bulge their way through capillaries just beneath the septal walls.4 Convex interfaces tend to resolve fluid whereas concave surfaces accumulate fluid.7

We fully appreciate the classic studies of Professor Walters and his predecessors8 9 showing the role of ion–channel water pumps; although it is still a moot point whether β-adrenergic stimulation can increase pumping capacity to the level needed to account for such rapid water clearance during normal birth. The vital question seems to be why these pumps are so severely compromised in respiratory distress syndrome that it can take 2–6 days to clear the fluid even after administering exogenous surfactant. To be constructive, it is particularly interesting when they8 9find that “for a secretory organ to be capable of generating a chemical gradient, a barrier must be present to restrict molecular diffusion” and, in the fetal lung, at least, “this barrier resides in the pulmonary epithelium.”8 Surely, the oligolamellar SAPL lining shown in our paper by epifluorescence microscopy and by electron microscopy in fig 1 (for a normal infant) is ideal for this function. Even a monolayer of SAPL bound to a solid can decrease ion permeability by an order of magnitude.10 It would also seem reasonable that, by spanning intercellular junctions, as seen in fig 1, SAPL layers not only act as a “first line of defence”1 against airborne pathogens,2 but also provide a membrane of known semi-permeability2 for preventing protein leakage and allowing those proteins to pump water under the known gradients.8 Thus an adequate lining of epithelial bound SAPL could be vital to both ion–channel and oncotic water pumps, in addition to any physical action in “dewatering.”