Article Text

Download PDFPDF

Thrombospondin in the eye
  1. J M Stewart
  1. Correspondence to: Dr Jay M Stewart University of California San Francisco, Department of Ophthalmology, 10 Koret Way, K301, San Francisco, CA 94143-0730, USA; ne62{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

A regulator of angiogenesis

Research into the pathophysiology of age related macular degeneration (AMD) has advanced at a rapid rate in recent years. To see the pace of progress, one need only pick up any issue of a major ophthalmic journal or attend a poster session at an ophthalmic society meeting. Efforts are under way to learn more about the ageing of Bruch’s membrane, drusen formation, and angiogenesis in choroidal neovascularisation (CNV). And it’s beginning to pay off: our understanding of these mechanisms has led to some promising new treatments, particularly in the area of angiogenesis.

In the case of CNV, much of the focus lately has been on pro-angiogenic proteins such as vascular endothelial growth factor (VEGF). Treatment strategies that target pathologically elevated levels of VEGF are easy to understand: they try to block or reduce a known stimulus for the growth of CNV. Some early successes have been reported with anti-VEGF therapies.1,2

The waters are still muddy, though, when it comes to more fundamental, or at least earlier, steps in the process that leads an eye to develop AMD. What factors cause an ageing Bruch’s membrane to become susceptible to fissure and invasion by CNV? Why do excess lipids accumulate to form drusen in some patients but not others? What disrupts the balance of pro-angiogenic and anti-angiogenic factors in the retinochoroid layers of the macula and promotes new vessel growth? These are only a few of the questions that have not yet been answered fully.

Enter thrombospondin. This glycoprotein was first described in 1971 and isolated in 1978 by researchers studying the mechanisms and regulation of the blood clotting process.3,4 It earned its name because it was released by platelets in response to treatment with thrombin. The protein was found to be an endogenous platelet participant in the haemagglutination process through its interactions with platelet bound fibrinogen.5 A multivalent molecule, it was found to have binding sites for many different molecules including collagen, fibronectin, fibrinogen, plasminogen, and calcium.6 Thrombospondin was also shown to be produced by other cell types including endothelial cells.7

Thrombospondin was thus understood early on to have an important role in the interactions of cells with other cells and with extracellular matrix. Immunostaining showed that thrombospondin was present in the interstices around multiple tissues in the body. Atherosclerotic lesions were found to have strong staining with anti-thrombospondin antibodies.8 This is especially relevant to studies of AMD because of its similarities to atherosclerosis in lipid accumulation and basement membrane damage.

Work on this molecule in the past few decades has progressed in many directions, since the molecule’s precise roles have not yet been completely defined. One investigator described thrombospondin as a “protein in search of a function” and asked “what is this thrombospondin doing to the cells with which it interacts?”9 Clotting, wound healing (in which thrombospondin is seen in healing but not healed wounds), and embryonic development (during which more thrombospondin staining is detected than in adults) are all areas that have pointed to a role for this protein in cell migration and adhesion.10

Studies suggest that thrombospondin is a critical regulator of angiogenesis in the eye—and much work remains to be done to understand its role fully

Of especial relevance to AMD researchers was the discovery that thrombospondin has a regulatory role in angiogenesis. Initial work showed that the molecule was a homologue of an anti-angiogenic factor in hamsters, gp140. Thrombospondin was able to inhibit angiogenesis in vivo and prevent endothelial cell migration in vitro.11 These studies provided the first clue that thrombospondin released from a cell into the extracellular matrix might block neovascularisation by preventing endothelial cells from attaching to target structures in that space. Its role in angiogenesis is not as a simple suppressor, though, since thrombospondin can also induce migration and spreading of endothelial cells.12 Thus, thrombospondin has come to be known as a “modulator” of angiogenesis.

How might thrombospondin participate in the pathogenesis of AMD? The molecule is produced and secreted by retinal pigment epithelium (RPE) cells in culture and can be identified in the cytoplasm of RPE cells in a human eye section.13 These findings raised the possibility that RPE cells could contribute to a Bruch’s membrane barrier against neovascular invasion from the choroid (CNV) by elaborating thrombospondin into the extracellular matrix. The next order of business was to begin to study thrombospondin in eyes both with and without AMD in order to understand how the balance of power changes in this disease.

In this issue of BJO (p 48) Uno and colleagues have taken this logical next step. Their simple yet crucial study contributes another piece to the angiogenesis in AMD puzzle. It introduces AMD to the thrombospondin literature and allows us to begin to apply all the knowledge about this protein that we have accumulated from other systems of the body.

In their study, Uno et al performed immunostaining of human eye sections with an anti-thrombospondin antibody. They compared 12 aged control eyes with 12 eyes with AMD. Their results support the notion that thrombospondin might have an anti-angiogenic function at Bruch’s membrane: they found less thrombospondin in eyes with AMD than in matched controls. In addition, eyes with late AMD had less staining at Bruch’s membrane than eyes with early AMD.

These investigators also sought an explanation for the lack of CNV in extramacular locations and therefore looked at the staining levels in the far periphery in addition to the macula. The finding that peripheral levels were lower than macular levels in all eyes, regardless of disease state, may merely be a function of Bruch’s membrane being thinner in the periphery, as the authors suggest. It might also hint at a more complex relation between thrombospondin and angiogenesis in the eye than the main finding of the study, that there is less thrombospondin staining in eyes with AMD, might suggest.

Indeed, this complexity (or confusion) is highlighted by the observation that thrombospondin added to cultured RPE cells can actually increase the amount of VEGF released by these cells.14 On the other hand, thrombospondin-1 knockout mice develop retinal and choroidal neovascularisation in the setting of an insult such as uveitis, unlike normal mice.15 Taken together with the findings of Uno et al reported here, these and other studies suggest that thrombospondin is a critical regulator of angiogenesis in the eye—and that much work remains to be done to understand its role fully. One certainty is that the extracellular matrix and Bruch’s membrane, where thrombospondin localises, are at centre stage for the action in the pathophysiology of AMD. Readers are certain to see more studies involving these important sites in the future.

Note in Proof

A regulator of angiogenesis


Linked Articles