Molecular medicine

The writing is on the vessel wall

Using an integrated approach, it is possible to identify the molecular addresses on blood-vessel walls that are read by blood cells — and, at least in rats, to deliver imaging agents and drugs to specific tissues.

Watching blood cells move through the blood vessels of living tissues under the microscope, researchers have learned that most of the time the cells move rapidly with the blood flow without stopping. Sometimes, however, they contact the walls of the vessel, slow down and roll along — perhaps even stopping and sticking in specific areas1,2,3,4. This and other indirect evidence has suggested that the endothelial cells that line the blood-vessel wall display different proteins on their surfaces according to where they are in the body and what events are occurring in a given organ. These proteins constitute the writing on the vessel walls — molecular ‘addresses’ that could tell circulating blood cells where they are and what to do in specific circumstances5.

It follows that scientists and physicians might also be able to take advantage of these addresses, using them to deliver imaging or therapeutic agents to specific organs, to sites of tissue damage, or to a developing tumour. Towards this end, on page 629 of this issue Oh et al.6 describe an integrated approach — involving proteomics, bioinformatics and molecular imaging — by which to identify and characterize the proteins on the blood-exposed surfaces of endothelial cells in lung and lung-tumour tissues. The authors also demonstrate that these proteins can be effectively targeted for imaging and therapy.

A number of proteins, for example members of the integrin family, have already been found to be expressed differentially on the blood-vessel walls of different tissues, and have been identified as promising targets to which to selectively deliver molecules for tumour imaging and treatment7,8,9. Nevertheless, researchers would benefit greatly from a complete map of the proteins on the blood-exposed surfaces of endothelial cells. A major problem is that these surfaces represent only a small fraction of any given tissue — so the proteins presented there are the proverbial needles in the haystack. The identification of subtle or under-represented blood-vessel markers therefore requires a focused approach to reduce the overall complexity of tissue proteins and yet to preserve the diversity of markers in the blood-vessel wall.

Oh and co-workers6,10 have come up with a clever way of doing just that. Analogous to using a magnet to extract needles from a haystack, these authors used a rapid, affinity-based isolation procedure to enrich and purify, from organs including rat lung and lung tumours, the parts of blood-vessel endothelial cells that contact the blood. They accomplished this by infusing colloidal silica particles into the bloodstream of rats, where these particles attached to the endothelial cells. Subsequent centrifugation of tissue homogenates allowed endothelial-cell membranes and attached organelles called caveolae10 to be separated from the remainder of the cells. Caveolae are small invaginations of the cell surface that carry out transport and signalling functions in cells. They are lined with an indicator protein called caveolin as well as other cell-surface proteins. For the final purification step, an antibody that recognizes caveolin, coupled to magnetic beads, was used to isolate caveolae and their associated proteins.

In the current work6, the authors analysed the components of this enriched cellular subset with mass spectrometry, database searches and antibodies for proteins that are found only in the wall of the blood vessels of given tissues, and then focused on the lungs and lung tumours. Next, they subjected those proteins that they had found exclusively in a given tissue vasculature to structural analyses, to ascertain which were most probably present on the endothelial cell surfaces and exposed to the circulation. Out of potentially hundreds of thousands of cellular proteins, Oh et al. extracted 37 that were present only in the endothelial membrane; 11 of these were probably in a configuration that includes an extracellular portion that could be presented to blood cells.

To validate this approach in an animal model, the authors used radiolabelled antibodies that recognize 2 of the 11 identified proteins — aminopeptidase-P and annexin A1 — to show that these proteins are found specifically in the blood vessels of lungs and lung tumours, respectively, in rats (Fig. 1). This brought the study full circle, from intact organs, to proteomic and bioinformatic analyses of selected proteins, to the imaging of specific signatures on the walls of blood vessels in defined organs of a living subject. As a coup de grâce, Oh et al. found that the radiolabelled antibodies against annexin A1 also prolonged survival, and may have caused significant remission of solid lung tumours in rats.

Figure 1: Molecular mapping.

Oh et al.6 have used an approach involving proteomics, bioinformatics and molecular imaging to identify proteins — two of which are shown here — that are expressed exclusively on the blood-exposed surface of endothelial cells in the lungs and in lung tumours of rats. (Endothelial cells, surrounded by a sheath of vascular smooth muscle, make up the inner wall of blood vessels.) This approach might therefore provide a way of directing imaging and therapeutic compounds to specific healthy or diseased tissues.

So, by learning what blood cells already know, we might be able to use the writing on the vessel wall both to design unique molecular tools by which to recognize specific diseases, and to direct therapies to specific organs9 — thereby avoiding sending drugs as unwanted ‘spam’ to all other organs in the body. For instance, proteins in the vasculature of tumours could enable imaging agents to locate growing malignancies, and drugs to specifically inhibit the tumour's endothelial cells and so prevent further tumour growth9.

It is likely that blood vessels in diseases other than cancer will have unique signatures as well; for example, the proteins expressed in and around atherosclerotic plaques might be used to identify (by imaging) and treat certain cardiovascular diseases. In many instances, therapies will need to be delivered to tissue cells beyond the blood-vessel wall. But Oh and colleagues' strategy might actually provide a solution to this problem, too. Given that the molecular targets they identified exist in the caveolae of endothelial cells, perhaps the role of these organelles in taking up molecules could be exploited to transport drugs across the endothelial barrier to reach cells beyond the endothelium.

To take this to the next level, perhaps it might even be possible to rebuild damaged tissues within the body through this type of tissue targeting. Imagine directing appropriate stem cells, known to promote tissue repair11, to sites of tissue damage in a patient. Despite the current controversy over the potential of adult stem cells from one tissue to form other types of cells (see, for example, refs 12, 13), advances in understanding stem-cell biology, cell-to-cell signalling and tissue remodelling, coupled with effective tissue targeting, might provide the necessary tools. If so, the writing would indeed be on the wall for all sorts of diseases.


  1. 1

    Miller, M. J., Wei, S. H., Cahalan, M. D. & Parker, I. Proc. Natl Acad. Sci. USA 100, 2604–2609 (2003).

    ADS  CAS  Article  Google Scholar 

  2. 2

    M'Rini, C. et al. J. Exp. Med. 198, 1301–1312 (2003).

    CAS  Article  Google Scholar 

  3. 3

    Condeelis, J. & Segall, J. E. Nature Rev. Cancer 3, 921–930 (2003).

    CAS  Article  Google Scholar 

  4. 4

    Cahalan, M. D., Parker, I., Wei, S. H. & Miller, M. J. Curr. Opin. Immunol. 15, 372–377 (2003).

    CAS  Article  Google Scholar 

  5. 5

    Folkman, J. Cancer Cell 1, 113–115 (2002).

    CAS  Article  Google Scholar 

  6. 6

    Oh, P. et al. Nature 429, 629–635 (2004).

    ADS  CAS  Article  Google Scholar 

  7. 7

    Pasqualini, R. & Ruoslahti, E. Nature 380, 364–366 (1996).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Pasqualini, R., Arap, W. & McDonald, D. M. Trends Mol. Med. 8, 563–571 (2002).

    CAS  Article  Google Scholar 

  9. 9

    Hood, J. D. et al. Science 296, 2404–2407 (2002).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Oh, P. & Schnitzer, J. E. J. Biol. Chem. 274, 23144–23154 (1999).

    CAS  Article  Google Scholar 

  11. 11

    Cowan, C. M. et al. Nature Biotechnol. 22, 560–567 (2004).

    CAS  Article  Google Scholar 

  12. 12

    Murry, C. E. et al. Nature 428, 664–668 (2004).

    ADS  CAS  Article  Google Scholar 

  13. 13

    Balsam, L. B. et al. Nature 428, 668–673 (2004).

    ADS  CAS  Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Contag, C., Bachmann, M. The writing is on the vessel wall. Nature 429, 618–619 (2004).

Download citation

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.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

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