News & Views | Published:

Developmental biology

The hole picture

Naturevolume 442pages363364 (2006) | Download Citation


What's the best way to make a tube? Roll up a sheet? Hollow out a solid rod? Some innovative movies show how the problem is tackled during the development of blood vessels in embryos.

In simple organisms, only one or a few cells thick, diffusion brings nutrients and oxygen to individual cells and removes their waste. But in larger animals, this process is no longer sufficient, so transport tubes evolved to allow materials to move quickly throughout the organism. Such transport systems include blood vessels and the tubes that make up the digestive, respiratory and other organ systems. Generally these tubes are lined by a single layer of cells known as epithelial cells, but the inner surface of blood vessels are coated with modified epithelial cells called endothelial cells. Tubes must have a continuous hollow space, or lumen, in the centre to allow free passage of materials. A century-old question is how lumina form in tubes during development. Kamei et al. (page 453 of this issue)1 have watched living embryos to discover how lumina are made during the development of fish.

Lumen formation has been examined pre-viously using static images of endothelial and epithelial cells in culture. A commonly observed feature in these pictures is the formation of large2 vacuoles — membrane-bounded compartments — in the cell (Fig. 1, overleaf). These are thought to be formed by a process called endocytosis, in which cells suck in a small portion of their outer, plasma membrane to form a pocket and then pinch it off to create an internal 'bubble' containing some of the extracellular fluid. It had been imagined that these vacuoles fuse together and with the plasma membrane to produce lumina between cells, although the evidence, based on static images, was not definitive.

Figure 1: Making spaces.
Figure 1

Kamei et al.1 show that lumina in vascular tubes are formed by the fusion of intracellular vacuoles. Vacuoles form inside the endothelial cells, apparently by endocytosis. These vacuoles may fuse with each other to create larger vacuoles. The vacuoles then fuse with the plasma membrane, to create lumina between the cells.

Now Kamei et al. have filmed this process, first in an endothelial culture system and then in the developing blood vessels of fish embryos — which fortunately are transparent. In both cases the movies show actual fusion of the vacuoles with the plasma membrane to create lumina between cells (Fig. 1). In essence, the cells first create tiny intracellular lumina by sipping up extracellular fluid into endocytic vacuoles. These then fuse with the plasma membrane, thereby combining many small intracellular lumina into a single extracellular lumen between the cells that will line the nascent tube.

These movies of lumen creation are astounding: if a picture is worth a thousand words, the power of these films to answer a long-standing question is certainly worth a million. Kamei et al. exploit this power to show how developing lumina extend from one cell to another to lengthen the tube. To do so, they injected nanometre-sized fluorescent beads (quantum dots) into the blood circulation of the fish embryo. The quantum dots appeared first in a large blood vessel, the dorsal aorta. Then, as a new, smaller vessel began to bud off, the quantum dots emerged in the lumen of a previously unlabelled vacuole in an endothelial cell adjoining the aorta. Quantum dots showed up successively in a second and then a third neighbouring cell, as each cell's intracellular vacuole fused with its plasma membrane to extend the lumen of the nascent endothelial tube.

Many other epithelial organs face the problem of how to create a lumen. Superficially, at least, there is a multitude of ways by which such tubes form in different organs, but there should be common mechanisms underlying this diversity3,4. In some cases, especially larger tubes, cells in the middle may undergo programmed cell death to hollow out a lumen5. Large vacuoles are not usually seen inside epithelial cells in the process of forming lumina and tubes, although they do occur in certain situations6. It may be that in most circumstances, much smaller vacuoles or related membrane-bounded 'vesicles' are used to build the luminal surface of the epithelial cell7. These vesicles may be part of the endocytic pathway or of the pathway by which newly made membrane is normally delivered to the cell surface, or both.

Understanding the morphogenesis of multi-cellular structures, such as tubes, therefore requires analysis of the molecular mechanisms behind the movement and formation of membranes. Such membrane trafficking depends on a special class of fats or lipids, known as phosphatidylinositides, which are components of the greasy bilayer that make up all cellular membranes. One such lipid, phosphatidylinositol 3-phosphate, is characteristically found mainly on the membrane-bounded compartments of the endocytic system. Related phosphatidylinositides tend to occupy specific locations in the cell, leading to the idea that various phosphatidylinositides could determine the identity of different membrane compartments of the cell8. Perhaps the phosphatidylinositide composition of the plasma membrane facing the lumen of tubes holds the key to its assembly and identity.


  1. 1

    Kamei, M. et al. Nature 442, 453–456 (2006).

  2. 2

    Davis, G. E. & Bayless, K. J. Microcirculation 10, 27–44 (2003).

  3. 3

    Lubarsky, B. & Krasnow, M. A. Cell 112, 19–28 (2003).

  4. 4

    Hogan, B. L. & Kolodziej, P. A. Nature Rev. Genet. 3, 513–523 (2002).

  5. 5

    Debnath, J. & Brugge, J. S. Nature Rev. Cancer 5, 675–688 (2005).

  6. 6

    Vega-Salas, D. E., Salas, P. J. & Rodriguez-Boulan, E. J. Cell Biol. 104, 1249–1259 (1987).

  7. 7

    O'Brien, L. E., Zegers, M. M. P. & Mostov, K. E. Nature Rev. Mol. Cell Biol. 3, 531–537 (2002).

  8. 8

    De Matteis, M., Godi, A. & Corda, D. Curr. Opin. Cell Biol. 14, 434–447 (2002).

Download references

Author information


  1. Departments of Anatomy, and Biochemistry and Biophysics, University of California, San Francisco, 94143-2140, California, USA

    • Keith Mostov
    •  & Fernando Martin-Belmonte


  1. Search for Keith Mostov in:

  2. Search for Fernando Martin-Belmonte in:

About this article

Publication history


Issue Date



By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing