Regulation of proBACEl by Glycosaminoglycans

The !3-secretase (BACE1) is initially synthesized as a partially act ive zymogen contain ing a prodomain which can be furthe r activated th rough proteolytic cleavage of the prodomain by a fu rin-Iike protease.The active site of BACEl is large and although a number of high-affinity active-site inh ibitors of BACEl have been described, most of these compounds are large, polar and do not cross the blood-brain barrie r. However, it may be possible to target other regions of the protein wh ich regulate BACEl allosterically. We have found that proBACEl can be stimulated by relatively low concentrations (e.g. 1 u .q/ rnl) of heparin . Heparin initially increases proBACEl activity, probably by binding to the prodomain, which decreases steric inhibiti on at the active site. However, the heparin-activated zymogen also unde rgoes autocatalysis, wh ich ultimately leads to a loss of enzyme activ ity. We speculate that proBACEl can be regul ated by endogenous heparan sulfate proteoglycans and t hat drugs wh ich target this interact ion may have value in the t reatment of Alzheimer's disease. Copyrig ht e 200 8 S. Karger AG, Base l The f3secretase (BACEl) is an important target for Alzheimer's disease (AD) drug development because the enzyme cata lyzes the first step in the production of 13amy loid, and because the enzy me may have on ly a limited range of substrates in vivo [1]. BACE1 belongs to the family of aspartic pro teases, but it differs from many family members by being mem brane anchored. Fulllength BACEl contains 501 amino-acid residues including a signa l pep tide sequence and a 24-resid ue prodoma in. During its maturation in the secretory pathway, BACE1 undergoes complex N-linked glycosylation [2] . Evidence suggests that proBACEl becomes fully glycosylated first and then the prodomain, which is important for the proper folding of the protease domain, is removed by a furin-like protease, possibly in the trans-Golgi network (TGN) [3]. However, the exact subcell ular compartments in which the pro domain is cleaved off have not yet been established. Unlike many aspartic proteases, the prodomain does not completely suppress activity of the enzyme [4]. Therefore, proBACEl may have an enzymic function in vivo. BACE1 may be a suitable target for drug development because BACE1 knockout mice are viable [5]. Apart from a hypomyelination phenotype, caused by the fact that neuregulin 1 (a BACEl substrate) is important for oligodend rocyte development [6], BACEl knockout mice have no major neuropathological abnormalities. However, the development of drugs which inhibit BACE1 at the active site has been difficult. The active site of BACEl is large Prof. David H. Small Menzies Resear ch Instit ute, University of Tasmania Private Bag 23, Hobart TAS 7001 (Aust ralia) Tel. +61 3 62267700. Fax +61 36226 7704, E-Mail d .h.sm [email protected] .au Accessible on line at: www.karger.com /n dd e 2008 S. Karger AG, Basel 16602854/08/00540206524.5010 Fax +41 61 306 1234 E-Mail [email protected] www.karger.com KARGE R