Statistics from Altmetric.com
In this issue of the BJO an interesting study is presented by Salzmann et al (p 1092) in which they analyse, by immunohistochemistry, the epiretinal membranes in proliferative diabetic retinopathy (PDR) for the presence of certain matrix metalloproteinases (MMPs) and their inhibitors. MMPs are a group of zinc dependent proteolytic enzymes that play an important part in the degradation of extracellular matrix components during developmental, physiological, and pathological processes. These enzymes are naturally inhibited in the extracellular matrix by the tissue inhibitors of metalloproteinases (TIMPs), and expression of MMPs and TIMPs is regulated by a variety of growth factors, cytokines, oncogenes, and tumour promoters. Currently, at least 18 MMPs and four TIMPs have been identified, and their potential roles in vitreoretinal disorders have been discussed recently in aBJO perspective.1
MMPs are very likely to have an important role in the pathological processes underlying PDR. However, with such a complex system of enzymes and enzyme regulation, a tenet that is central to the work of Salzmann et al is that certain MMPs or TIMPs play a specific part in PDR; if these could be identified, then it may be possible to selectively modulate their actions in order to improve the treatment of this condition. Salzmann et al looked for the presence of MMP-1, MMP-2, MMP-3, MMP-9, TIMP-1, TIMP-2, and TIMP-3 in PDR membranes and essentially demonstrated the presence of all of them in a large proportion of membranes. In addition, they obtained similar data when proliferative vitreoretinopathy (PVR) membranes were analysed. These findings raise a number of issues that will now be discussed.
If the various MMPs and TIMPs found in the membranes are synthesised by cells contained within the pathological membranes this would tend to suggest that they have specific roles in the underlying pathological processes. It is quite possible that some or all of the MMPs and TIMPs are synthesised by the cells within the membranes, but other possibilities also exist. Some of the MMPs and TIMPs maybe derived from the circulation as there is increased vascular permeability as a result of the neovascular process. It is also possible that they are derived from the vitreous as all of the MMPs and TIMPs identified in the PDR membranes by Salzmann et al have also been identified in the normal vitreous.2 3 The vitreous MMPs were generally found to be in the inactive proform. However, it has been demonstrated that a single vitreous gel contained sufficient (latent) MMP2 to cause considerable structural damage when activated and introduced in vitro into another vitreous gel.4Therefore, there is sufficient endogenous metalloproteinase in the normal vitreous to have a significant biological effect. The vitreous may therefore be acting as a source of MMPs and TIMPs that are used during PDR and PVR membrane formation. It is also possible that the MMPs and TIMPs that were identified as associated with PDR and PVR membranes were derived from the vitreous and may have been masking more specific and subtle expression patterns produced by cells within the membranes. These uncertainties can be addressed to some extent by looking at the proportion of active enzyme in the membrane and, indeed, a recent report showed that a significant proportion of the MMP-2 and MMP-9 in PDR membranes were in an active form.5 An alternative approach would be to look at the expression of specific mRNAs for MMPs and TIMPs within the pathological membranes.
A difficulty with this study is that, inevitably, late stage membranes were analysed where much of the active growth and remodelling had already taken place. This is particularly a problem because the development of PDR membranes is a multistage process. Initially, there is an angiogenic process in which basement membrane degradation, endothelial cell migration, capillary tube formation, and endothelial cell proliferation occur. Then there is degradation of the internal limiting lamina (ILL) allowing the neovascular tissue to enter the vitreous cavity. Once in the vitreous cavity the neovascular tissue grows along and into the cortical vitreous gel and there is a second wave of cellular proliferation and extracellular matrix deposition which results in the formation of contractile fibrovascular tissue. Finally, the new blood vessels and associated fibrous tissue undergo remodelling. It is likely that by the time that PDR membranes are surgically removed these underlying pathological processes will be at a late stage in this sequence of events, while an understanding of the early events, such as initiation of angiogenesis and degradation of the ILL, is more likely to point the way towards new therapeutic strategies.
Despite these difficulties one very interesting observation made by Salzmann et al was that MMP9 expression was specifically seen within perivascular matrix in PDR membranes, perhaps suggesting a specific role for MMP9 in the neovascular process. These findings resonate with other work which has shown increased levels of MMP-9 (predominantly in the latent proform) in the vitreous of diabetic patients.6 7 MMP-9 is one of the metalloproteinases that is capable of degrading type IV (basement membrane) collagen and as such may well be involved in the basement membrane degradation that occurs during angiogenesis and breakdown of the ILL. In addition, a further, as yet unidentified 75 kDa metalloproteinase has been identified in PDR vitreous, whereas this enzyme was not commonly seen in controls.3
The work by Salzmann et al and others makes an important contribution to our understanding of the role of MMPs and TIMPs in PDR. How can we further enhance our understanding and what are the most direct paths to allow us to move from basic biomolecular research to improved clinical management? Much has still to be learnt about the functions of MMPs and TIMPs and the regulation of their expression. It is likely that genetic experiments will contribute greatly to this area; for example, generating and studying the phenotype of “knockout” mice where a specific MMP or TIMP gene has been removed will allow us to understand more about the individual functions of these proteins. Much of the current research on the role of MMPs and TIMPs in proliferative disorders is driven by the pharmaceutical industry in the hope that drugs that modulate their function could be used in the treatment of cancer. Already a number of synthetic MMP inhibitors are undergoing clinical trials and a pragmatic approach for ophthalmic researchers may be to test these various pharmaceutical agents in experimental models of preretinal neovascularisation (and PVR) and to determine empirically whether these agents modify the disease process.