- Top of page
- Materials and methods
Alzheimer's disease is associated with abnormal accumulation of Aβ, which is produced from the β-amyloid precursor protein (APP) by the β-site APP-cleaving enzyme (BACE1) and γ-secretase. Our previous studies showed that heparin can decrease APP processing by decreasing the levels of BACE1 and ADAM10. In this study, we examined the effects of glycosaminoglycans (GAGs) on APP processing and Aβ production with the aim of understanding the specificity of the effects. Various GAG analogs were incubated with primary cortical cells derived from APP (SW)Tg2576 mice and the level of APP, proteolytic products of APP and APP-cleavage enzymes were measured. The effect of GAGs on APP processing was both size- and sulfation-dependent. 6-O-Sulfation was important for the effect on APP processing as heparin lacking 6-O sulfate were less potent than native heparin. However, deletion of carboxyl groups on heparin had no significant effect on APP processing. Our studies suggest that there is structural specificity to the effect of GAGs on APP processing and that certain GAGs have a greater effect on Aβ production than others. This suggests that it might be possible to alter the structure of GAGs to achieve more specific inhibitors of APP processing that can cross the blood–brain barrier.
Alzheimer's disease (AD) is an irreversible neurodegenerative disease (Ferri et al. 2005) that is characterized by the deposition of amyloid plaques and neurofibrillary tangles in the brain (Kidd 1964; Terry et al. 1964). The major component of the amyloid plaque is a 40–42 amino acid residue polypeptide (Glenner and Wong 1984; Masters et al. 1985) called the β-amyloid protein (Aβ). Aβ is generated from the β-amyloid precursor protein (APP) (Kang et al. 1987) by cleavage by the β-site APP-cleaving enzyme-1 (BACE1) (Sinha et al. 1999; Vassar et al. 1999; Yan et al. 1999; Cai et al. 2001) to liberate a C-terminally truncated fragment (C99), which is subsequently cleaved by γ-secretase (Haass et al. 1992). Both ADAM10 and ADAM17, also known as tumor necrosis factor-α-converting enzyme or tumor necrosis factor-α-converting enzyme (TACE), are α-secretases that can cleave APP within the Aβ sequence to produce an 83-amino acid fragment known as C83 (Sisodia et al. 1990; Sisodia 1992; Buxbaum et al. 1998). Cleavage by either of these α-secretases precludes formation of Aβ (Esch et al. 1990).
Oligomeric forms of Aβ are now thought to be the major toxic species in the initiation and development of AD (Lambert et al. 1998; Hartley et al. 1999; Kim et al. 2003; Haass and Selkoe 2007). Considering the central role of Aβ in the pathogenesis of AD, one of the main therapeutic strategies is to decrease the production of Aβ (Small et al. 2004).
Our studies (Cui et al. 2011) and those of other groups (Snow et al. 1988; Kisilevsky et al. 1995; Dudas et al. 2002; Scholefield et al. 2003; Bergamaschini et al. 2004) raise the possibility that glycosaminoglycans (GAGs) or GAG analogs may be effective in the treatment of AD. Peripheral administration of enoxaparin, a low molecular weight (LMW) heparin, has been reported to reduce Aβ load in the brain (Bergamaschini et al. 2004). Heparin can potentially influence Aβ production by disrupting β-secretase processing of APP. Leveugle et al. (1997) first reported that heparin stimulates β-secretase cleavage of APP in a cultured cell line. Heparin binds close to the prodomain of the BACE1 zymogen (proBACE1), and this binding stimulates proBACE1 activity (Beckman et al. 2006; Klaver et al. 2010). In contrast, Scholefield et al. (2003) reported that heparan sulfate and its more highly sulfated analog, heparin, can inhibit BACE1 activity and decrease Aβ production in cell culture. Our more recent studies have shown that treatment with heparin can lower Aβ secretion from primary cortical cells (Cui et al. 2011). However, although heparin can bind directly to BACE1, decreased secretion of Aβ in cell culture is because of a decrease in the level of BACE1, rather than direct inhibition of the enzyme.
The development of GAG analogs which can be used for the treatment of AD will require the identification of high-potency compounds that have the ability to cross the blood–brain barrier (BBB). Several reports indicate that low molecular weight GAGs can penetrate the BBB (Leveugle et al. 1998; Ma et al. 2002) and this idea is further supported by the observation that peripheral administration of enoxaparin can lower brain amyloid load (Bergamaschini et al. 2004). However, developing high-affinity compounds that can inhibit Aβ production may be problematic. Our studies showed that the most potent GAG heparin inhibits Aβ production in cell culture at micromolar concentrations (Cui et al. 2011). However, the development of GAG analogs which can be used for the treatment of AD may require high-potency compounds acting in the nanomolar concentration range.
The pattern of sulfation of heparan sulfate (HS) may provide specificity for binding to certain proteins. For example, studies by Nurcombe et al. (1993) have shown that the specificity of HS for binding to fibroblast growth factor receptors is controlled by the sulfation pattern. Similarly, the fine structure of HS may regulate syndecan-1 function (Sanderson et al. 1994), whereas a specific HS sulfation pattern regulates retinal axon targeting (Irie et al. 2002). Studies by Patey et al. (2006) show that specific sulfation patterns on heparin derivatives can result in high-affinity compounds with great selectivity for inhibition of BACE1 activity and reduced activity against Factor Xa and other proteases. Indeed, the different sulfation patterns of different HS species may reflect the need to bind specifically to different ligands. On this basis, then, it may be possible to alter the sulfate pattern of GAGs to achieve high affinity and specific effects on APP metabolism and Aβ secretion.
In view of the relationship of GAG size to BBB permeability and of sulfation pattern to binding specificity, the aim of this study was to examine the role of molecular size and sulfation of GAGs on APP processing and Aβ production in primary cell cultures. We tested the effects of various GAGs and sulfated polysaccharides on APP processing in cortical cells derived from transgenic mice expressing human APP695 with the Swedish familial AD mutant (Tg2576 mouse). These mice were used for the study because we wished to examine effects on human APP processing and because human APP and its fragments can be more easily detected by existing anti-human antibodies than rodent APP and Aβ. Our study shows that LMW heparin species can alter APP processing and that the effect of heparin on APP processing is dependent upon the degree of sulfation. Although no high-potency GAG analogs were identified in this study, the results demonstrate that there is structural specificity to the effect of GAGs on APP, raising the possibility that high-affinity BBB-permeable GAGs may eventually be identified.
- Top of page
- Materials and methods
In this study, the effects of different sulfated carbohydrates on APP processing and Aβ production was examined in primary cortical cells derived from APP (SW) Tg2576 mice. Our results showed that both size and structure of GAGs are important for effects on APP processing. Indeed, the sulfation of MH is essential for an effect on APP processing. However, the carboxyl group on MH was not important as deletion of the carboxyl group did not block the effect of MH on APP processing.
Lung heparin, which contains more highly sulfated disaccharide units than MH (Hileman et al. 1998), was similar to MH in its effects on APP processing. PPS and fucoidan, which are highly sulfated polysaccharides (Degenhardt et al. 2001; Li et al. 2008), also reduced APP proteolytic processing. As PPS and fucoidan have a similar degree of sulfation to MH and LH, the results support the view that sulfation is important for the effect on APP processing. The results also suggest that the backbone structure of the carbohydrate may not be as important for the effect on APP processing, as PPS and fucoidan, which do not belong to the GAG family, were as potent as MH and LH. The fact that polysulfated compounds that are not GAGs also altered APP processing suggests that other sulfated polysaccharides could be candidates for the screening of therapeutic agents for AD. Compared with the highly sulfated GAGs and polysaccharides, HS and chondroitin sulfate A, B, and C, which are less highly sulfated (Hileman et al. 1998), were less potent in lowering BACE1 and ADAM10. Furthermore, they did not lower Aβ levels. Taken together, these results indicate that sulfation is vital for the effect on the level of BACE1, ADAM10, and on APP processing.
To confirm the role of sulfation on APP processing further, individual sulfate groups were removed from MH and the effect of removal was determined. Our results indicated that the removal of any sulfate group decreased the potency of MH on APP processing. The 6-O sulfate was the most important of all the sulfate groups, as MH de 6S had no effect on the level of BACE1 and ADAM10. This finding is consistent with a previous study that reported that removal of 6-O sulfate of MH reduced the inhibitory potency of MH on BACE1 (Scholefield et al. 2003; Patey et al. 2006). Although our study indicated that the decreased β-secretase cleavage of APP is because of a decrease in the level of BACE1, rather than inhibition of the enzyme activity, these findings indicate the 6-O sulfate of MH is critical for the activity of MH. MH derivatives lacking all sulfate groups lacked the ability to disrupt APP processing. In contrast, removal of the carboxyl groups on MH did not attenuate the effects of MH on APP cleavage, suggesting that while sulfation of GAGs is important for inhibition of APP processing, the effect was not solely because of the negative charge on the carbohydrate. A concentration-dependence study indicated that MH lacking carboxyl groups had a similar potency to MH in its ability to decrease BACE1 (Fig. 7a and b). The results of this experiment suggest that while it may be possible to modify the structure of GAG to achieve more specific GAG derivatives for the treatment of AD, whether there are modified GAGs that possess a higher potency than MH for inhibiting APP processing is unclear.
These results provide evidence that it may be possible to selectively modify the structure of GAGs and reduce the unwanted side effects without changing the potency of GAGs on APP processing. For instance, the carboxyl group is essential for the anticoagulant and vasodilatory activity of heparin (Agarwal and Danishefsky 1986; Paredes-Gamero et al. 2012). Chemical removal of the carboxyl group of heparin could reduce these unwanted effects, but still potently inhibit APP processing.
Our results also show that the ability of GAGs to inhibit APP processing is dependent on chain length. A minimum GAG size of 6 kDa, which is equivalent to a length of 17 saccharide monomers, was necessary to achieve a similar effect on APP processing as that obtained with MH. While 6-kDa MH was similar to native 18-kDa MH in its effect on BACE1 and ADAM10, small heparin fragments (e.g. 3 kDa) only weakly reduced the level of BACE1 and ADAM10 and only weakly inhibited APP processing. Scholefield et al. (2003) previously reported that the inhibitory effect of MH on BACE1 activity is size dependent. Together with our results, this suggests that the size of MH is important for action on BACE1 activity.
Heparin fragment size may influence its activity in vivo. It has been reported that a chain length of 17 saccharides (approximately 6 kDa) is required for efficient thrombin inhibition (Petitou et al. 1999). The chain length of heparin is also important for its binding to fibroblast growth factor-2, as the tetrasaccharides or longer oligosaccharides of heparin are required to bind to fibroblast growth factor-2 and induce proliferation of chlorate-treated rat mammary fibroblasts (Delehedde et al. 2002). These data suggest that it may be possible to design GAGs, which can alter APP processing but which have fewer unwanted side effects.
It may also be possible to reduce the size of GAGs so that they can cross the BBB and still retain the ability to decrease Aβ production. Previous studies have shown that full-length MH cannot cross the BBB, whereas 3 kDa or smaller heparins can cross (Leveugle et al. 1998; Ma et al. 2002). In our study, the 3-kDa MH derivative decreased APP processing, albeit weakly, raising the possibility that small MH derivatives, without hemorrhagic side effects, could pass through the BBB and inhibit APP processing.
The mechanism by which polysulfated carbohydrates decrease BACE1 and ADAM10 and alter APP processing is unclear. The decreased level of BACE1 and ADAM10 was probably not caused by a decrease in BACE1 or ADAM10 mRNA, because BACE1 and ADAM10 mRNA levels were not significantly changed after MH treatment. This suggests that the effect was on some post-translational event such as BACE1 and ADAM10 turnover. However, further studies will be needed to delineate this mechanism.
Whether GAGs will have therapeutic value for the treatment of AD is still unclear. While it may be possible to design compounds which are more limited in their actions (i.e., affect APP processing and Aβ production without unwanted side effects) and which cross the BBB, it is still unclear whether high-potency compounds will be identified. In this regard, it was of particular interest to note that fucoidan was as potent as MH in lowering BACE1 and Aβ levels (Fig. 5a and b and Fig. 7a and b). Fucoidans are a group of polysaccharides derived from algae and seaweed. Because of their considerable structural diversity, it seems logical to investigate further the effect of other fucoidans on Aβ production.
Another consideration with regard to the suitability of GAGs as drugs for the treatment of AD is their action on sAPPα. GAGs also reduced the secretion of sAPPα in our study. However, some studies suggest that sAPPα may have neurotrophic actions (Saitoh et al. 1989; Milward et al. 1992; Mattson 1994; Smith-Swintosky et al. 1994; Meziane et al. 1998). To date, the side effects of decreasing sAPPα production are unknown. GAG derivatives which act specifically on the β-secretase cleavage pathway of APP may be needed. In our study, MH de 2S and MH de NS treatment decreased the secretion of Aβ but had no significant, or had only a small effect on the level of sAPPα and ADAM10. These findings suggest that it is possible to design GAG derivatives that selectively target the β-secretase cleavage of APP.
In summary, this study shows that there is structural specificity to the effects of GAG on APP processing. LMW heparins can cross the BBB (Leveugle et al. 1998; Ma et al. 2002) and potentially may attenuate Aβ-induced inflammation (Kisilevsky et al. 1995; Zhu et al. 2001), decrease Aβ aggregation (Kisilevsky et al. 1995), lower Aβ generation, and improve cognition (Bergamaschini et al. 2004; Timmer et al. 2010). Ultimately, it may be possible to design more potent GAG derivatives which act specifically to inhibit β-secretase cleavage of APP that can be used for the treatment of AD.