The function of caveolae — flask-shaped invaginations at the plasma membrane — has remained elusive for almost 50 years. Reporting in Science, Kurzchalia and colleagues now describe a new tool that might shed light on the functions of this mysterious organelle — a caveolin-1 knockout (cav-1−/−) mouse.

Caveolin-1 is the main protein constituent of caveolae; and it was shown years ago to be sufficient to generate caveolae in lymphocytes, which normally do not contain caveolae. Kurzchalia and colleagues now confirm that caveolin-1 is essential for the formation of these organelles, as cav-1−/− mice completely lack caveolae. So, by knocking out a single protein, the authors have in fact knocked out an organelle — a remarkable feat in itself.

Caveolae are often considered a specialized type of lipid raft — lateral aggregates of cholesterol and glycosphingolipids that concentrate certain proteins. But lipid-raft preparations from cav-1−/− mice did not show any gross defects in composition, which indicates that caveolin-1 and/or caveolae are not required for the organization of these membrane domains.

The authors then tested a few of the postulated functions of caveolae. They found that transcytosis — which is thought to be the mechanism by which albumin crosses the endothelium — is probably not dependent on caveolae, as the albumin concentration in the cerebrospinal fluid remained unchanged in cav-1−/− mice. Also, caveolae are probably not required for cholesterol transport, as the blood lipoprotein composition and the cholesterol content of high-density lipoprotein were unchanged in the knockout mice.

However, caveolin-1 and/or caveolae do function in signalling during several physiological processes. Isolated aortic rings of cav-1−/− mice did not establish a steady contractile tone, and their relaxation after acetylcholine stimulation was highly increased. Both processes are known to be mediated by nitric oxide (NO), and, indeed, the basal release of NO was 31% higher and cGMP levels were three times higher in cav-1−/− mice. So, caveolin-1 and/or caveolae are probably negative regulators of NO-mediated vascular relaxation. Loss of caveolin-1 and/or caveolae also resulted in a markedly decreased response to vasoconstrictors, such as angiotensin II, endothelin-1 or phorbol ester in vascular smooth muscle cells. Last, the myogenic tone of these cells was reduced.

In addition to the marked defects in vascular physiology, cav-1−/− mice also displayed uncontrolled hyperproliferation of angioblastic cells and fibrosis in lung alveolar septa, which implicates caveolin-1 and/or caveolae in the local control of cell proliferation.

Remarkably, despite the profound dysfunction of the vascular system and the pathomorphological defects in lung alveolar septa, cav-1−/− mice were viable, and had only relatively minor physical disabilities — for example, they could not swim as long as their wild-type littermates. To explain the non-lethal phenotype, the authors propose that lipid rafts, which seem to be normal in these mice, could carry out many of the functions that are proposed for caveolae.