VLA Integrin Expression in Normal and Diseased Human Kidney: How Should We Interpret Immunohistochemical Data?

Graham Hillis, MRCP, Department of Medicine and Therapeutics, University of Aberdeen, Foresterhill, Aberdeen AB9 2ZD (UK) Dear Sir, We were interested to read the article by Kanahara et al. [1] in the January issue of Nephron, relating to the distribution of extracellular matrix proteins and integrins in IgA nephropathy and agree that this is an area worthy of further investigation. Undoubtedly, the interactions between glomer-ular cells and their matrix are of fundamental importance to the pathogenesis of IgA disease and many other glomerulopathies [2]. Similarly, adhesion molecules play a pivotal role in these interactions [3, 4]. However, we feel that there are several methodological and interpretive issues in the paper which require clarification. The antibody referred to as targeting the ‘fibronectin receptor’ is, in fact, directed against the ßi-chain of the VLA integrins. As such it will detect the presence of all VLA integrins (VLA 1-6 and αvßι) which act as receptors for a variety of ligands (see table 1) in addition to fibronectin. Although this is noted in the penultimate paragraph the use of ‘fibronectin receptor’ is misleading, particularly as this term was originally used to refer specifically to the o1⁄8ßi integrin [5] -though later work confirms that other VLA integrins share this property [6]. In a similar way the ‘vitronectin receptor’ referred to is the αvß3 integrin which, again as noted, serves as a receptor for a variety of ligands not solely vitronectin. This may partially explain the lack of correlation between the presence ofvitronectin and the expression of this molecule though it may of course merely indicate that the vitronectin present has ocTable 1. Summary of molecular weights, ligands and main cell reactivity for the VLA integrins