Building a good lung, capable of efficient gas exchange between air and blood, poses a number of conflicting problems for physiology and cell biology. One of these is the complex architecture that allows dynamic functional integration of the lung’s structural elements (1); another is the role of cells in making a very thin and yet robust air–blood barrier, the topic of this essay. From a physiological perspective, the air–blood barrier must not only be very thin, measuring no more than a few tenths of a micrometer, but also of very large expanse, over 100 m2 in the human lung (2, 3). However, it must be built of biologically active cells, capable of selfmaintenance and self-renewal, as well as serving several functions beyond being a passive diffusion barrier. This holds particularly for the alveolar epithelium that forms the external sheet of the barrier (4). In the age of molecular cell biology, fascinating discoveries have advanced our understanding of the biology of these cells, obtained in sophisticated studies partly in vivo, but for a large part in vitro (4–13). It is, however, important to also consider the exceptional conditions under which these cells develop and are sustained in the complex architecture of the integral functioning lung (1). In that view, the alveolar epithelium deserves particular attention, because it must form and secure the very large surface of the air spaces.

The epithelial covering of the alveolar capillaries is so extraordinarily thin that it was long believed by many prominent histologists that the capillaries were “naked like a wound”(14, 15), as light microscopy could not reveal a continuous lining by epithelial cells. It was the pioneer in