Roles of brain proteoglycans and neuroglia in amyloid-beta (Abeta) deposition and clearance in Alzheimer's disease

Alzheimer's disease (AD) is the most common cause of dementia in the elderly and one of a host of diseases termed amyloidoses. In AD, the amyloidogenic protein, amyloid-beta (Abeta), a 39-43 amino acid cleavage product of the trans-membrane amyloid precursor protein (APP), accumulates in brain lesions along with proteoglycans, which have been shown to play a number of important roles in regulating disease progression. Both heparan-sulfate proteoglycan (HSPG) and chondroitin-sulfate proteoglycan (CSPG) have been shown to inhibit degradation of fibrillar Abeta by enzymes [39] and microglia. Therefore, we hypothesized that targeting Abeta-HSPG interactions may result in enhanced clearance of Abeta by microglia and/or astroctyes and provide a useful strategy for the treatment of AD. We began by characterizing the binding of different Abeta species to HSPG and found that fibrillar Abeta binds with high affinity to HSPG whereas non-fibrillar species, including dodecamer, bind only at pH<6. This is consistent with previous work by Gupta-Bansal et al. [39]. We then tested a number of compounds for the ability to inhibit Abeta binding to HSPG or promote release of Abeta from HSPG. We found that sulfated compounds such as low molecular weight heparin [LMWH], poly-(vinylsulfonate) [PVS], and 3-amino-1-propanesulfonic acid (Tramiprosate) were potent inhibitors of Abeta binding to HSPG and also released Abeta from HSPG at very low concentrations. Interestingly, recombinant human ApoE also inhibited binding of Abeta to HSPG in an isoform-specific manner consistent with its pattern as a genetic risk factor in AD [26]. We then showed that HSPG has a small but significant inhibitory effect on degradation of fibrillar Abeta by mouse neonatal microglia (MNM), but not by adult mouse astrocytes (AMA), and that this inhibition can be reversed by sulfated compounds. However, when PVS and LMWH were tested in an in situ assay for MNM-mediated removal of Abeta from PDAPP mouse brain sections, the compounds failed to promote clearance. These results provide some insight into the failure of the Tramiprosate clinical trial and suggest that HSPG may not play as significant a role in protecting Abeta against proteolysis and microglial degradation as originally suspected.;Previous work from our laboratory and others has demonstrated that both adult mouse astrocytes (AMA) and mouse neonatal microglia (MNM) have the capacity to degrade fibrillar Abeta in vitro [80, 93, 94] and from brain sections of PDAPP mice [80, 81, 95]. In the latter case, MNM required the presence of an anti-Abeta antibody to degrade Abeta [33], suggesting a fundamental difference in Abeta processing by AMA and MNM. Further studies demonstrated that AMA and MNM also differ in their ability to degrade fAbeta released from heparan-sulfate proteoglycan (HSPG) by sulfated compounds. Since only rapidly-sedimentable components of fAbeta were found to bind to HSPG, we extended our degradation studies to investigate the ability of AMA and MNM to process Abeta species of different sizes. We found that MNM degraded 40% of the rapidly-sedimentable Abeta species generated by centrifugation of fAbeta, whereas AMA degraded none. In contrast, AMA degraded a significantly larger percentage of the slower-sedimenting Abeta species (80%) than MNM (55%). We then used inhibitors of phagocytosis, lysosomal acidification, and various secreted proteases to characterize the mechanism of Abeta degradation by AMA and MNM in vitro. We found that MNM degraded Abeta by a mechanism which was inhibited by cytochalasin-D and NH4Cl, whereas AMA secreted protease(s) which degraded Abeta extracellularly. We performed similar studies in situ by plating AMA and MNM on Abeta-rich brain sections from PDAPP mice and found that the same clearance mechanisms were utilized in situ. These studies provide mechanistic insight into the role of anti-Abeta antibodies as modulators of Abeta clearance and represent the first evidence of preferential processing of Abeta species of different sizes by AMA and MNM. Furthermore, the understanding of microglial and astrocyte-mediated clearance mechanisms provided by these studies may be useful in the design of new treatment strategies for AD.