STRUCTURAL ANALYSIS OF A SKIN COLLAGEN-LINKED FLUOROPHORE THAT INCREASES IN DIABETES AND END-STAGE RENAL DISEASE

摘要

Protein-linked autofluorescence at 440 nm upon excitation at 370 nm; i.e., henceforth long-wavelength fluorescence or LW, was introduced many years ago by us as a putative and surrogate marker for the Maillard reaction in vivo. Ever since, this marker has been used in a large number of studies with the assumption that it is an advanced glycation endproduct (AGE) representative of the broader Maillard reaction in vivo. However, our recent data with EDIC patients together with studies by Smit's group, as well as the introduction onto the commercial market of diagnostic devices capable of noninvasively measuring LW in human forearm skin for screening individuals for diabetes and its complications strongly suggests that the biochemical origin needs to be precisely understood. Most noteworthy is that longwave fluorscence can be generated in vivo through the oral administration of methylglyoxal to rodents. Possible precursors for LW include glucose, ribose, glycolaldehyde, ascorbic acid, Ehrlich-positive Chromogen-related product , lipid peroxidation, but not lysyl oxidase-dependent hydroxypyridinium crosslinks. Undoubtedly, both oxidative and glycoxidative processes can generate LW. In short, there is a long list of potential precursors for LW, and while many authors have assumed that LW is a marker for AGEs, there is no evidence that LW formed in vivo is biochemically related to AGE formation. Therefore, much of the existing data regarding the origins of LW in tissue proteins is confusing and controversial. Thus, clarification of these structures is urgently needed. Below, we describe the isolation, purification, and partial molecular characterization of a single major protein-bound fluorophore from enzymatic digests of human skin collagen referred to as LW1 whose levels are elevated by aging, diabetes and end-stage renal disease (ESRD).