Developmental biology

One cell, two fates

Article metrics

The two types of cell that make up blood vessels can develop from the same precursor. This discovery might improve our understanding of the role of blood vessels in disorders ranging from cancer to heart disease.

It seems odd, considering how important blood vessels are, how little we know about their development. We do know that they are generally composed of two cell types: endothelial cells line the inside, forming channels that conduct blood, while smooth muscle cells cover the outside, protecting the fragile channels from rupture and controlling blood flow. Conventional wisdom holds that endothelial and smooth muscle cells arise from separate 'precursor' cells through rounds of cell division and specialization.

But conventional wisdom is now being overturned. On page 92 of this issue1, Yamashita and colleagues describe a type of blood-vessel precursor from which both endothelial and smooth muscle cells develop (Fig. 1), in the culture dish and in mice. This precursor might contribute to the formation of both 'naked' endothelial capillaries and smooth-muscle-coated vessels — processes that are essential to development and reproduction and which, when derailed, contribute to cancer and some types of heart disease2.

Figure 1: A common origin for the two types of blood-vessel cell.

Centre, endothelial cells are those that make up the interior walls of a blood vessel. Smooth muscle cells or pericytes encase the endothelial channels. These basic cell types have been thought to arise from separate types of precursor. Endothelial cells arise from precursors called angioblasts or haemangioblasts in the embryo (top), or from circulating endothelial progenitors in the adult (not shown). Smooth muscle cells and pericytes, by contrast, can form from a variety of progenitors (bottom). These include mesenchymal cells (those that make up connective tissue) in embryo and adult, embryonic neural crest cells, and progenitors in the embryonic epicardium (the outer layer of the heart). Smooth muscle cells can also form from endothelial cells. But Yamashita et al.1 have discovered a new precursor cell (right) that gives rise to both types of blood-vessel cell. Vascular endothelial growth factor (VEGF) promotes the development of endothelial cells from this precursor. Platelet-derived growth factor-BB (PDGF-BB) stimulates their development into smooth muscle cells and pericytes.

It was already known that, in the developing embryo, endothelial cells arise either from precursors — better known as 'progenitors' or 'stem cells' — that can produce only endothelial cells (angioblasts), or from progenitors that give rise to both endothelial and blood cells3 (haemangioblasts; Fig. 1). Endothelial precursors have also been found in adults4,5. Most of these endothelial precursors divide and differentiate in response to a protein called vascular endothelial growth factor (VEGF), which is produced close to forming vessels4,5.

The cells that surround the endothelial channels vary throughout the blood-vessel system6. For instance, several layers of smooth muscle cells surround the large vessels close to the heart, whereas single cells (not joined into layers) called pericytes cover smaller, more distant vessels.

These cells also differ in their origins (Fig. 1). The first smooth muscle cells in the embryo6 and the smooth-muscle-like cells in the prospective cardiac valves7 originate from the endothelium in a process that — at least in the heart — requires 'transforming growth factor' (TGF) proteins, such as TGF-β3 (ref. 7). Another such protein (TGF-β1) may be involved in the differentiation of the loose connective tissue (mesenchyme) around endothelial channels to form smooth muscle progenitors8,9. Then, when these channels branch out, they produce a 'recruitment factor' — platelet-derived growth factor-BB (PDGF-BB). This instructs smooth muscle progenitors to migrate alongside the new branches8,9. Endothelium-surrounding cells in the vessels that nourish the heart are derived from a putative progenitor in the external layer of the heart. By contrast, smooth muscle cells in the large vessels close to the heart and pericytes in the forebrain are derived from neural cells6. In adults, smooth muscle cells can form from many types of precursor10.

Yamashita et al.1 further complicate this picture with their discovery that one type of embryonic precursor in the mouse can give rise to either endothelial cells or smooth muscle cells, depending on the growth factors to which it is exposed. In response to PDGF-BB, this precursor cell differentiates to form smooth muscle cells. These cells express a set of 'marker' molecules that distinguish smooth muscle cells from other cells3, and surround endothelial channels in culture and in vivo.

VEGF, by contrast, sends the precursor along the developmental pathway to becoming an endothelial cell. But it seems that VEGF may only favour (and not fix) cell fate — which is not too surprising, given that endothelial cells still form in embryos lacking VEGF11. Further molecules may fix endothelial fate. These molecules might resemble VEGF in binding one particular cellular receptor — VEGF receptor-2 — as the differentiation of endothelial cells is completely blocked when this receptor is lost12. It will be important to identify these VEGF-like molecules, and to find out whether they affect Yamashita et al.'s versatile precursor cell. Another question is whether this vascular progenitor also specifies whether the endothelial cells it forms contribute to both arteries and veins (Fig. 1).

What about PDGF-BB — does it fix, or only favour, smooth-muscle-cell fate? Studies of mice engineered to lack PDGF-BB indicated that the mesenchymal cells mentioned above still develop into smooth muscle progenitors in the absence of PDGF-BB8,13, as (to some extent) do Yamashita et al.'s common progenitor cells1. But, in PDGF-BB-deficient mice, new branches from endothelial channels lack smooth muscle cells8,13. It will be interesting to see whether this defect boils down to a problem with the growth and migration of the common vascular progenitors.

It is rather surprising that a common progenitor for endothelial and smooth muscle cells has only now been discovered. Precursors that give rise to several distinct types of neural cell, or to the different types of blood cell, have already been reported. And endothelial cells that line the inner surface of the heart share a common origin with their surrounding cardiac muscle fibres14. Moreover, endothelial cells can form skeletal muscle. Might there be other progenitor cells that have diverse fates in the blood- vessel system? And do haemangioblasts and Yamashita et al.'s vascular progenitors — which both express VEGF receptor-2 (ref. 3) — themselves arise from a single precursor?

Other unanswered questions relate to the possible medical implications of these results. Do the newly discovered progenitor cells, for instance, contribute to the growth of smooth muscle cells that occurs in atherosclerosis (the accumulation of lipids in, and thickening of, arterial walls)6? Are they involved in the formation of new blood vessels that accompanies tumour development? Blindness can result when pericytes are lost from blood vessels in the retina — might it become possible to selectively shift these progenitors to become pericytes?

These vascular progenitors might also be useful to treat heart disease that is characterized by a lack of oxygen supply, as this requires growth of both endothelial and smooth muscle cells3. In fact, signals from cancer cells, oxygen-depleted heart muscle and healing wounds in adults do recruit other endothelial progenitors4,5. The next step is to learn more about how these versatile vascular precursors are induced to grow and differentiate.


  1. 1

    Yamashita, J. et al. Nature 408, 92–96 (2000).

  2. 2

    Carmeliet, P. & Jain, R. K. Nature 407, 249–257 (2000).

  3. 3

    Carmeliet, P. Nature Med. 6, 389–395 ( 2000).

  4. 4

    Rafii, S. J. Clin. Invest. 105, 17–19 (2000).

  5. 5

    Isner, J. M. & Asahara, T. J. Clin. Invest. 103 , 1231–1236 (1999).

  6. 6

    Gittenberger-de Groot, A. C., DeRuiter, M. C., Bergwerff, M. & Poelmann, R. E. Arterioscler. Thromb. Vasc. Biol. 19, 1589– 1594 (1999).

  7. 7

    Nakajima, Y., Mironov, V., Yamagishi, T., Nakamura, H. & Markwald, R. R. Dev. Dyn. 209, 296–309 (1997).

  8. 8

    Hellstrom, M. et al. Development 126, 3047– 3055 (1999).

  9. 9

    Hirschi, K. K., Rohovsky, S. A. & D'Amore, P. A. J. Cell Biol. 141, 805– 814 (1998).

  10. 10

    Campbell, J. H., Efendy, J. L., Han, C. L., Girjes, A. A. & Campbell, G. R. J. Vasc. Res. 37 , 364–371 (2000).

  11. 11

    Carmeliet, P. et al. Nature 380, 435–439 (1996).

  12. 12

    Shalaby, F. et al. Cell 89, 981–990 (1997).

  13. 13

    Lindahl, P., Hellstrom, M., Kalen, M. & Betsholtz, C. Curr. Opin. Lipidol. 9, 407–411 ( 1998).

  14. 14

    Eisenberg, L. M. & Markwald, R. R. Circ. Res. 77, 1–6 (1995 ).

Download references

Author information

Correspondence to Peter Carmeliet.

Rights and permissions

Reprints and Permissions

About this article

Further reading


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.