Sialylation enhances the secretion of neurotoxic ... - Wiley Online Library

Journal of Neurochemistry, 2006, 96, 924–933

doi:10.1111/j.1471-4159.2005.03595.x

Sialylation enhances the secretion of neurotoxic amyloid-b peptides Kazuhiro Nakagawa,*,¶ Shinobu Kitazume,*,¶ Ritsuko Oka,* Kei Maruyama,à Takaomi C. Saido, Yuji Sato,§ Tamao Endo§ and Yasuhiro Hashimoto*,¶ *Glycochain Functions Laboratory, Suprabiomolecular System Group, Frontier Research System, RIKEN, Wako-shi, Saitama, Japan Proteolytic Neuroscience Laboratory, Brain Science Institute, RIKEN, Wako-shi, Saitama, Japan àDepartment of Pharmacology, Saitama Medical School, Moroyama, Saitama, Japan §Glycobiology Research Group, Tokyo Metropolitan Institute of Gerontology, Itabashi-ku, Tokyo 173–0015, Japan ¶CREST, Japan Science and Technology Agency, Kawaguchi, Saitama, Japan

Abstract Alzheimer’s disease (AD) is characterized by amyloid-b peptide (Ab) deposition in the brain. Ab is produced by sequential cleavage of amyloid precursor protein (APP) by b-secretase (BACE1: b-site APP-cleaving enzyme 1) and c-secretase. Previously, we demonstrated that BACE1 also cleaves b-galactoside a2,6-sialyltransferase (ST6Gal-I) and downregulates its transferase activity. Here, we report that overexpression of ST6Gal-I in Neuro2a cells enhanced a2,6sialylation of endogenous APP and increased the extracellular levels of its metabolites [Ab by two-fold, soluble APPb (sAPPb) by three-fold and sAPPa by 2.5-fold). Sialylationdeficient mutant (Lec-2) cells secreted half as much Ab as wild-type Chinese hamster ovary (CHO) cells. Furthermore, wild-type CHO cells showed enhanced secretion of the APP

metabolites upon ST6Gal-I overexpression, whereas Lec-2 cells did not, indicating that the secretion enhancement requires sialylation of cellular protein(s). Secretion of metabolites from a mutant APP (APP-Asn467,496Ala) that lacked N-glycosylation sites was not enhanced upon ST6Gal-I overexpression, suggesting that the N-glycans on APP itself are required for the enhanced secretion. In the mouse brain, the amount of a2,6-sialylated APP appeared to be correlated with the sAPPb level. These results suggest that sialylation of APP promotes its metabolic turnover and could affect the pathology of AD. Keywords: amyloid-beta, amyloid precursor protein, sialylation, b-galactoside a2, 6-sialyltransferase J. Neurochem. (2006) 96, 924–933.

The neuropathological hallmarks of Alzheimer’s disease (AD) include amyloid-b peptide (Ab) deposition in plaques, neurofibrillary tangle formation and neuronal cell loss in vulnerable brain regions (Selkoe 2001; Hardy and Selkoe 2002). The amyloid plaques contain an aggregated population of heterogeneous Ab peptides derived from amyloid precursor protein (APP), a type-I membrane glycoprotein. APP is initially cleaved at its b-site by a membraneanchored aspartyl protease called b-site APP-cleaving enzyme 1 (BACE1) (Hussain et al. 1999; Sinha et al. 1999; Vassar et al. 1999; Cai et al. 2001; Kitazume et al. 2001), thereby generating a soluble N-terminal fragment (sAPPb) and a membrane-bound C-terminal stub (Cterminal fragment b, CTFb). CTFb is subsequently cleaved by a second protease (c-secretase) to yield Ab and CTFc, and this pathway is known as the b-cleavage pathway (De

Strooper et al. 1998; Wolfe et al. 1999). In another pathway, APP is initially cleaved within the Ab sequence by a-secretase, generating a different soluble N-terminal

924

Received June 21, 2005; revised manuscript received September 11, 2005; accepted October 4, 2005. Address correspondence and reprint requests to Yasuhiro Hashimoto, Glycochain Functions Laboratory, Suprabiomolecular System Group, Frontier Research System, The Institute of Physical and Chemical Research (RIKEN), 2–1 Hirosawa, Wako-shi, Saitama 351-0198, Japan. E-mail: [email protected] Abbreviations used: Ab, amyloid-b peptide; APP, amyloid precursor protein; APPNL, Swedish-type mutant APP; BACE1, b-site APP-cleaving enzyme 1; CHO, Chinese hamster ovary; CTF, C-terminal fragment; MAM, Maackia amurensis; N2a, cells, Neuro2a cells; PNGase, peptide:N-glycanase; sAPP, soluble APP; SSA, Sambucus sieboldiana; ST6Gal-I, b-galactoside a2,6-sialyltransferase.

2006 International Society for Neurochemistry, J. Neurochem. (2006), 96, 924–933

Sialylation enhances amyloid-b secretion 925

fragment (sAPPa) and C-terminal stub (CTFa). CTFa is further cleaved by c-secretase to yield a non-pathogenic p3 peptide and CTFc, and this pathway is known as the a-cleavage pathway (Buxbaum et al. 1998; Koike et al. 1999; Lammich et al. 1999). Previously, we found that BACE1 is also involved in the cleavage and secretion of b-galactoside a2,6-sialyltransferase (ST6Gal-I) in cultured cells (Kitazume et al. 2001; Kitazume et al. 2003). ST6Gal-I secretion was increased in BACE1transgenic mice and decreased in BACE1-knockout mice (Kitazume et al. 2005), confirming the involvement of BACE1 in ST6Gal-I secretion in vivo. ST6Gal-I catalyzes a2,6-sialylation of Galb1,4-GlcNAc residues on the N-glycans (asparagine-linked oligosaccharides) of glycoproteins (Weinstein et al. 1987). Sialylation and glycosylation are known to affect many properties of glycoproteins, such as protein folding, structural stability, enzymatic activity and intracellular transport (Opdenakker et al. 1993). Sialic acids are often present at the non-reducing ends of glycans, conferring strong negative charges on the protein and affecting protein–protein interactions (Fujita-Yamaguchi et al. 1985; Watanabe et al. 2003). APP contains both O- and N-glycans, the latter of which are attached to Asn467 and/or Asn496 (Saito et al. 1993). Sialylation and glycosylation of APP appear to be important for its proteolytic processing and secretion (Pahlsson and Spitalnik 1996; Yazaki et al. 1996; McFarlane et al. 1999). In the present study, we investigated the effects of APP sialylation on its metabolism.

Experimental procedures Reagents The following materials were used in this study: protein molecular weight standards (Bio-Rad, Hercules, CA, USA); columns for DNA purification (Qiagen, Chatsworth, CA, USA); anti-APP (22C11) monoclonal antibody (Chemicon International, Temecula, CA); antiAPP (6E10) monoclonal antibody (Signet Laboratories, Dedham, MA, USA); and anti-Ab (1–40) polyclonal antibody (SigmaAldrich, St. Louis, MO, USA). PCR was performed using Pfu DNA Polymerase (Stratagene, La Jolla, CA). Expression plasmids Rat ST6Gal-I-FLAG-pSVL (Kitazume-Kawaguchi et al. 1999) and human APPNL-pcDNA3.1 (Kitazume et al. 2001) were constructed as described previously. Human ST3Gal-IV-FLAG-pSVL was constructed as described for ST6Gal-I-FLAG-pSVL (KitazumeKawaguchi et al. 1999). Human APPNL-N467,496A-pcDNA3.1 was generated by ligation-free PCR-mediated mutagenesis using human APPNL-pcDNA3.1 as a template and the following oligonucleotides as primers: 5¢-AATTATGAGCGCATGGCTCAGTCTCTCTCC-3¢ (for the N467A mutation of APP) and 5¢-CTTCAGAAAGAGCAGGCCTATTCAGATGACG-3¢ (for the N496A mutation of APP).

Transfection Transfection was performed with the FuGENE6 transfection reagent (Roche Diagnostics, Mannheim, Germany). To prepare Neuro2a (N2a) cells that stably expressed human APPNL or human APPNLN467,496A, the cells were transfected with APPNL-pcDNA3.1 or human APPNL-N467,496A-pcDNA3.1, respectively. After G418 selection, clones stably expressing APP were obtained by limiting dilution. Pulse-chase analysis N2a or Chinese hamster ovary (CHO) cells were transiently transfected with either rat ST6Gal-I-pSVL or vector alone. Next, the cells were labeled using an Express protein labeling mix (100 lCi/mL; PerkinElmer, Wellesley, MA, USA), in methionineand cysteine-free Dulbecco’s modified Eagle’s medium for 2 h, chased in non-radioactive serum-free medium, and then lysed in 1 · radioimmunoprecipitation assay buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 5 mM EDTA, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate) containing a protease inhibitor cocktail (Roche Diagnostics). APP in the cell lysates and extracellular Ab peptides in the medium were immunoprecipitated with the anti-APP (C15) antibody (Maruyama et al. 1990) and antiAb antibody (Benjannet et al. 2001), respectively. The precipitated proteins were analyzed by electrophoresis using a NuPAGE 12% Bis-Tris Gel (Invitrogen, Carlsbad, CA, USA), and the radiolabeled proteins were detected using a BAS 2000 radioimage analyzer (Fuji Film, Tokyo, Japan). Western blot analysis N2a or CHO cells were transiently transfected with a cDNA of ST6Gal-I. After 22 h in culture, the culture medium was replaced with fresh medium and the cells were incubated for an additional 8 h. Total sAPP (sAPPa + sAPPb) in the culture medium was precipitated with heparin-agarose (Pierce, Rockford, IL, USA). a2,6-sialylated sAPP was precipitated with Sambucus sieboldiana (SSA) lectin-agarose. For the analysis of membrane-bound (intact) APP, cells were lysed in lysis buffer (10 mM HEPES, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 5 mM EDTA, 10% glycerol) containing a protease inhibitor cocktail. An aliquot of the lysate was clarified by centrifugation at 17 000 g for 15 min and then diluted in 7 vol. of phosphate-buffered saline containing the protease inhibitor cocktail. a2,6-Sialylated APP in the lysate was precipitated with SSA lectin-agarose. Each sample was mixed with Laemmli buffer (Laemmli 1970) and subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis using a 4–20% gradient gel. Next, the separated proteins were transferred to a nitrocellulose membrane, and incubated with an appropriate primary antibody. The following primary antibodies were used: anti-APP (C15) antibody for endogenous membrane-bound (intact) APP; anti-APP (6E10) antibody for human intact APP and sAPPa; anti-APP (22C11) antibody for total sAPP and its a2,6-sialylated form; antibNL antibody for sAPPNLb; anti-sAPPa antibody for endogenous sAPPa; and anti-sAPPb antibody for endogenous sAPPb. The membranes were subsequently incubated with an appropriate secondary antibody, namely horseradish peroxidase (HRP)-conjugated goat anti-mouse or anti-rabbit IgG. Binding of each antibody to its antigen was detected using a chemiluminescent substrate (Pierce).

2006 International Society for Neurochemistry, J. Neurochem. (2006) 96, 924–933

926 K. Nakagawa et al.

Detection of amyloid precursor protein in the mouse brain Mouse brains were individually homogenized at 4C in 10 vol. of homogenization buffer (10 mM HEPES, pH 7.4, 1 mM EDTA, 0.25 M sucrose) containing a protease inhibitor cocktail, and the resulting homogenates were centrifuged at 100 000 g for 60 min. The a2,6- and a2,3-sialylated proteins in the supernatants were precipitated with SSA lectin-agarose and Maackia amurensis (MAM) lectin-agarose, respectively. An aliquot of each supernatant was incubated with 10 units of peptide:N-glycanase (PNGase; Roche) for 6 h at 37C prior to the precipitation and then subjected to western blot analysis using the anti-APP (22C11) antibody. The membrane fractions (100 000 g pellets) prepared from the mouse cerebrum were treated with the lysis buffer as described above. The a2,6- and a2,3-sialylated APP in the lysates were precipitated with SSA lectin-agarose and MAM lectin-agarose, respectively, and then detected with the anti-APP (C15) antibody.

Results

Effects of sialyltransferase overexpression on the extracellular levels of secreted amyloid precursor protein and amyloid-b Initially, we analyzed the effect of a2,6-sialyltransferase (ST6Gal-I) overexpression on endogenous APP sialylation in murine neuroblastoma N2a cells. The cell-associated (intact) form of APP and extracellular sAPP were detected as triple bands on western blots (Figs 1a and b), which possibly represent splice variants of APP and/or heterogeneous post-translational modifications, including differences in glycosylation. The a2,6-sialylated form of intact APP was precipitated with SSA lectin-agarose and then detected with the anti-APP (C15) antibody. Overexpression of ST6Gal-I increased the amount of the a2,6-sialylated form of intact APP in N2a cells, relative to the total amount of intact APP (Fig. 1a). Furthermore, ST6Gal-I overexpression significantly increased the amounts of both sAPPa (2.5-fold) and sAPPb (3.3-fold) in the culture medium (Figs 1b and c), indicating simultaneous increases in both the a- and b-cleavage products. Human neuroblastoma SK-N-SH cells, human embryonic kidney HEK293 cells and CHO cells demonstrated similar increases in sAPPa and sAPPb upon ST6Gal-I overexpression (CHO cell data are shown in the next section; SK-N-SH and HEK293 cell data are not shown). Overexpression of another sialyltransferase, ST3Gal-IV, which catalyzes a2,3-sialylation of Galb1,4-GlcNAc structures on N-glycans, produced similar increases in the amounts of sAPPa and sAPPb to ST6Gal-I overexpression (Figs 2b and c). Thus, overexpression of sialyltransferases tends to produce simultaneous increases in both the a- and b-cleavage products. In contrast, overexpression of BACE1 decreased the amount of sAPPa and concomitantly enhanced the amount of sAPPb (Fig. 1b), confirming the previous notion that sAPPa and sAPPb are generated from the common substrate APP by competitive cleavage at the

a- and b-sites (Skovronsky et al. 2000). Next, we examined whether ST6Gal-I overexpression enhanced the secretion of the pathogenic peptide Ab. A pulse-chase analysis using [35S]methionine and [35S]cysteine revealed that ST6Gal-I overexpression in N2a cells enhanced the generation of extracellular Ab peptides (Figs 2a and b), and the two Ab bands observed were tentatively assigned as Ab1–40 and Ab11–40 according to a previous report (Benjannet et al. 2001). These results suggest that ST6Gal-I overexpression promotes the proteolytic processing of APP, thereby enhancing the secretion of its metabolites. To investigate whether ST6Gal-I overexpression also increased the amounts of other secreted proteins, we analyzed the total extracellular proteins in the culture media. The majority of the secreted proteins were unaffected by overexpression of the sialyltransferases, whereas several showed slight decreases (Fig. 2c), suggesting that the increased secretion was limited to a small number of proteins, including the APP metabolites. Effects of sialylation deficiency on amyloid precursor protein metabolism Overexpression of ST6Gal-I or ST3Gal-IV enhanced the amount of sAPP, suggesting that sialylation of cellular protein(s) accelerates APP processing. To clarify this issue, we used sialylation-deficient CHO (Lec-2) cells, which lack a CMP-sialic acid transporter protein that translocates CMPsialic acid (a donor substrate for sialylation) from the cytosol to the Golgi lumen where sialylation of proteins occurs (Eckhardt et al. 1996). It should be noted that the soluble APP from Lec-2 cells showed a higher mobility on immunoblots due to the lack of sialyl residues on the molecules (Fig. 3c). A pulse-chase analysis revealed that the Ab peptide amounts secreted by Lec-2 cells were half those secreted by wild-type CHO (Pro-5) cells (Figs 3a and b). Overexpression of ST6Gal-I in sialylation-deficient Lec-2 cells did not increase the amounts of sAPPb and total sAPP (Figs 3c and d), whereas ST6Gal-I overexpression in wildtype Pro-5 cells significantly enhanced the amount of sAPPb. These results suggest that the increase in sAPP secretion requires sialylation of cellular protein(s), which is catalyzed by the overexpressed ST6Gal-I.

N-Glycans of the amyloid precursor protein molecule itself are required for the enhanced amyloid precursor protein secretion Next, we prepared N2a transformants that stably expressed Swedish-type mutant APP (APPNL), a preferred substrate for BACE1, and designated them N2a-APPNL cells. The sAPPb generated from APPNL has an Asn-Leu sequence at the amino-terminus and can be distinguished from the endogenous form using a cleavage-site specific antibody (antibNL). The sAPPb generated from APPNL appeared as a



Fig. 1 Effects of sialyltransferase expression on amyloid precursor protein (APP) metabolism in murine neuroblastoma Neuro2a (N2a) cells. (a) N2a cells were transiently transfected with a b-galactoside a2,6-sialyltransferase (ST6Gal-I) cDNA-expressing vector (+) or a control vector (–). Cell-associated (intact) APP in cell lysates was detected with an anti-APP (C15) antibody. The a2,6-sialylated form of intact APP was precipitated with Sambucus sieboldiana (SSA) lectinagarose and then detected with the anti-APP (C15) antibody. (b) After

overexpression of ST6Gal-I, ST3Gal-IV or BACE1, intact APP in the cells was detected with an anti-APP (C15) antibody. Extracellular soluble APP (sAPP) was precipitated with heparin-agarose and then detected with an anti-sAPPa or anti-sAPPb antibody. (c) The relative amounts of sAPPa and sAPPb to intact APP were calculated after transient expression of ST6Gal-I or ST3Gal-IV. The values represent means ± SEM (n ¼ 3; *p < 0.05).

single band on a western blot (Fig. 4a). The band shifted to a lower molecular weight after sialidase treatment, indicating that most of the sAPPb was sialylated. The presence of a2,6sialyl residues on the sAPP was confirmed by precipitation with SSA lectin-agarose (Fig. 4b). The band corresponding to the a2,6-sialylated form increased after ST6Gal-I overexpression. In addition, ST6Gal-I overexpression also increased

the amounts of sAPPb and sAPPa by 1.9-fold and 1.4-fold, respectively, relative to vector-only controls (Figs 4c and d). We hypothesized that the increase in sAPP production was due to sialylation of APP itself. To test this hypothesis, we prepared N2a transformants that stably expressed a glycosylation-deficient mutant of APPNL, in which the N-glycosylation sites were replaced with alanines, and designated them



Fig. 2 Effects of b-galactoside a2,6-sialyltransferase (ST6Gal-I) overexpression on amyloid-b peptide (Ab) secretion. (a) Neuro2a (N2a) cells were transiently transfected with an ST6Gal-I cDNAexpressing vector (+ ST6Gal-I) or a control vector. The cells were pulse-labeled with [35S]methionine and [35S]cysteine for 120 min and then chased for 0, 60 or 120 min. Intact amyloid precursor protein (APP) in the cells and Ab in the media were immunoprecipitated with an anti-APP (C15) and anti-Ab antibody, respectively. (b) Effect of the

chase time on the relative amount of radiolabeled extracellular Ab produced by cells overexpressing ST6Gal-I (open squares) or transfected with a control vector (closed circles). Values represent means ± SEM (n ¼ 3; *p < 0.05). (c) After overexpression of ST6Gal-I or ST3Gal-IV, total extracellular proteins in the media were precipitated with trichloroacetic acid, separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and stained with Coomassie Brilliant Blue.

N2a-APPNL-N467,496A cells. The glycosylation-deficient APP and its sAPP showed mobility shifts due to their lack of N-glycans (Fig. 4c). In contrast to the case for N2a-APPNL cells, ST6Gal-I overexpression did not increase the amount of sAPPb produced by N2a-APPNL-N467,496A cells (Figs 4c and d), indicating that the increase in this cleavage product required the presence of N-glycans on the APP molecule.

that some of the sAPP was a2,6- and/or a2,3-sialylated in vivo. These signals were not observed after pre-treatment of the samples with PNGase prior to the precipitation, indicating that the sialyl residues were mainly present on the N-glycans of sAPP in the mouse brain. We further analyzed the sialylation of the membrane-bound form of APP in individual brains from mice at various ages (2–30 months), and found that the sialylation levels varied among individual mice at different ages. The level of a2,6-sialylation (ratio of the a2,6-sialylated form to total APP) was positively correlated with the sAPPb level (r ¼ 0.47599, p ¼ 0.03388) (Fig. 5b). No correlation was observed between the level of a2,3-sialylation and the sAPPb level (data not shown). This result suggests that a2,6-sialylation could affect APP metabolism in vivo as well as in vitro.

Sialylation of amyloid precursor protein in the mouse brain Next, we analyzed the sialylation isoforms of sAPP in the mouse brain. A fraction of the sAPP (sAPPa + sAPPb) in the brain homogenate was precipitated with either SSA lectin-agarose or MAM lectin-agarose (Fig. 5a), indicating



Fig. 3 Effects of sialylation on the secretion of amyloid-b peptide (Ab) and soluble amyloid precursor protein (sAPP). (a) Wild-type (Pro-5) and sialylation-deficient (Lec-2) Chinese hamster ovary (CHO) cells were pulse-labeled for 120 min and chased for 0, 90 or 180 min. Intact APP and Ab peptides were detected as described in the legend for Fig. 2(a). (b) Effect of chase time on the relative amount of radiolabeled extracellular Ab secreted by Pro-5 (closed circles) or Lec-2 (open squares) cells. Values represent means ± SEM (n ¼ 3;

Discussion

APP b-cleavage is an initial process that triggers the pathogenesis of AD. Therefore, any factor that enhances b-cleavage and the subsequent production of Ab could be a potential risk for this disease (Selkoe 2001; Hardy and

*p < 0.05). (c) After overexpression of b-galactoside a2,6-sialyltransferase (+ ST6Gal-I) or a control vector in Pro-5 or Lec-2 cells, intact APP in the cells was detected with an anti-APP (22C11) antibody. Extracellular sAPPb and total sAPP were detected with an antisAPPb and anti-APP (22C11) antibody, respectively. (d) The relative amounts of sAPP to intact APP were calculated after transient expression of ST6Gal-I in Pro-5 or Lec-2 cells. Values represent means ± SEM (n ¼ 3; *p < 0.05).

Selkoe 2002). McFarlane et al. (1999) previously reported that overexpression of ST6Gal-I enhanced the secretion of sAPP, although they did not analyze the subspecies of sAPP or the production of Ab. In the present study, we have confirmed that sAPP secretion is enhanced upon ST6Gal-I overexpression and, more importantly, demonstrated



Fig. 4 Effects of amyloid precursor protein (APP) sialylation on soluble APP (sAPP) secretion. (a) Neuro2a (N2a) cells stably expressing APPNL were transiently transfected with a b-galactoside a2,6-sialyltransferase (ST6Gal-I) cDNA-expressing vector (+ ST6Gal-I) or a control vector. sAPP in the media was incubated with (+) or without (–) 0.01 units of sialidase. sAPPb derived from APPNL was detected with an anti-bNL antibody. (b) After overexpression of ST6Gal-I, the a2,6-sialylated form of sAPP in the medium was precipitated with Sambucus sieboldiana (SSA) lectin-agarose and then detected with

an anti-APP (22C11) antibody. (c) ST6Gal-I was transiently overexpressed in N2a cells stably expressing APPNL or glycosylationdeficient APPNL-N467,496A. Intact APP and sAPPa were detected with an anti-APP (6E10) antibody, and sAPPb was detected with an anti-bNL antibody. (d) The relative amounts of sAPP to intact APP were calculated after transient expression of ST6Gal-I in N2a cells stably expressing APPNL or APPNL-N467,496A. Values represent means ± SEM (n ¼ 3; *, p < 0.05).

concomitant increases in sAPPb and sAPPa together with the promotion of Ab production (Figs 1 and 2). We have also demonstrated that ST6Gal-I overexpression does not increase the secretion of metabolites from sialylation-deficient Lec-2 cells or glycosylation-deficient APP in N2a cells, suggesting

that sialylation of the N-glycans on APP itself is required for the enhanced secretion (Figs 3 and 4). Furthermore, enhanced secretion was not observed for the majority of the other secreted proteins (Fig. 2c), implying that a particular molecular mechanism is involved in the enhanced



Fig. 6 A possible model for BACE1-dependent cleavage of APP and ST6Gal-I, both of which regulate APP secretion. Ab, amyloid-b peptide; APP, amyloid precursor protein; BACE1, b-site APP-cleaving enzyme 1; ST6Gal-I, b-galactoside a2,6-sialyltransferase.

Fig. 5 Amyloid precursor protein (APP) sialylation in the mouse brain. (a) The a2,6- and a2,3-sialylated isoforms of soluble APP (sAPP) were precipitated with Sambucus sieboldiana (SSA) lectin-agarose and Maackia amurensis (MAM) lectin-agarose, respectively. sAPP was detected by western blot analysis using an anti-APP (22C11) antibody. An aliquot of each sample was treated with (+) or without (–) peptide:N-glycanase (PNGase) prior to the precipitation. (b) Correlation between the amount of sAPPb and the ratio of the a2,6-sialylated isoform to intact APP (a2,6-sialylated APP/total APP). sAPPb was precipitated using heparin-agarose and then detected with an antisAPPb antibody. The a2,6-sialylated form of APP was precipitated with SSA lectin-agarose. Intact APP and its a2,6-sialylated form were detected with an anti-APP (C15) antibody. The dashed line shows a positive correlation (r ¼ 0.47599, p ¼ 0.03388).

secretion of APP metabolites. One possible mechanism is that sialylation of APP affects its susceptibility to secretases, resulting in enhanced cleavage and secretion. Another possibility is that sialylation promotes APP transport to the compartments where the proteolytic processing occurs. Ho and Sudhof (2004) reported that an ’APP ligand protein’ inhibited the secretion of APP metabolites. It is tempting to speculate that the interaction of this ligand with APP is

attenuated by the transfer of sialic acid onto APP, thereby inducing an increase in the selective transport of APP. Alternatively, certain transport machinery may recognize the presence of sialic acid together with a part of the APP molecule. An example of this type of sorting machinery that recognizes both glycosyl residues and the core protein is the system involving the mannose-6-phosphate receptor that targets proteases and hydrolases to lysosomes (Dustin et al. 1995). The analysis of APP sialylation in the mouse brain revealed a positive correlation between the a2,6-sialylation of APP and the level of sAPPb (Fig. 5b). This result suggests that a2,6-sialylation of APP in vivo may play a role in enhancing b-cleavage. In the AD brain, a2,6-sialylation analysis of APP and its correlations with the levels of APP metabolites, including sAPPb, may provide a clue to the pathological relevance of sialylation to the development of this disease. BACE1 cleaves APP and enhances the secretion of sAPPb and Ab (Hussain et al. 1999; Sinha et al. 1999; Cai et al. 2001). We previously demonstrated that BACE1 also cleaves ST6Gal-I and down-regulates a2,6-sialylation of cellular proteins in HEK293 cells (Kitazume et al. 2001). Considering the previous and present results together, we speculate that BACE1 regulates APP metabolism not only by cleaving APP but also by down-regulating sialylation via ST6Gal-I cleavage (Fig. 6). Some types of familial AD are associated with specific mutations in APP. For example, the Swedishtype mutant APPNL is efficiently cleaved by BACE1, resulting in rapid accumulation of Ab in the brain. The APPNL mutant partly competes with ST6Gal-I cleavage (Kitazume et al. 2001), which would increase APP sialylation



to enhance the secretion of Ab and contribute to early onset of the disease. On the other hand, an increase in BACE1 activity has been proposed as a primary cause of sporadic AD (Fukumoto et al. 2002; Holsinger et al. 2002; Yang et al. 2003). Despite the enhanced BACE1 activity, the pathogenic processes of sporadic AD appear to develop very slowly. The enhancement of BACE1 activity may down-regulate APP sialylation to attenuate Ab production, thereby contributing to the prolonged course of disease development. In this context, any BACE1 inhibitors developed as AD therapeutics should be selective for APP cleavage by BACE1, and not inhibit the cleavage of ST6Gal-I. Such selectivity should also be important for preventing other adverse side-effects. Acknowledgements We thank Dr Tae-Wan Kim (Harvard Medical School) for supplying the human BACE1-pcDNA3.1 and Dr James C. Paulson (The Scripps Research Institute) for providing the ST3Gal-IV cDNA. This study was partly supported by the Frontier Research System Fund (YH), a Strategic Research Grant (YH) and an Industrial Collaboration Grant (YH) from the RIKEN Institute, partly by a Leading Project Fund and Grants-in-Aid for Scientific Research on Priority Areas, nos. 16015329 (YH), 17025047 (YH), 17046025 (SK) and 15770090 (SK), from the Ministry of Education, Science, Sports and Culture of Japan (MEXT), and partly by the CREST Fund of the Japan Science and Technology Agency (YH).

References Benjannet S., Elagoz A., Wickham L. et al. (2001) Post-translational processing of beta-secretase (beta-amyloid-converting enzyme) and its ectodomain shedding. The pro- and transmembrane/cytosolic domains affect its cellular activity and amyloid-beta production. J. Biol. Chem. 276, 10 879–10 887. Buxbaum J. D., Liu K. N., Luo Y., Slack J. L., Stocking K. L., Peschon J. J., Johnson R. S., Castner B. J., Cerretti D. P. and Black R. A. (1998) Evidence that tumor necrosis factor alpha converting enzyme is involved in regulated alpha-secretase cleavage of the Alzheimer amyloid protein precursor. J. Biol. Chem. 273, 27 765– 27 767. Cai H., Wang Y., McCarthy D., Wen H., Borchelt D. R., Price D. L. and Wong P. C. (2001) BACE1 is the major beta-secretase for generation of Abeta peptides by neurons. Nat. Neurosci. 4, 233–234. De Strooper B., Saftig P., Craessaerts K., Vanderstichele H., Guhde G., Annaert W., Von Figura K. and Van Leuven F. (1998) Deficiency of presenilin-1 inhibits the normal cleavage of amyloid precursor protein. Nature 391, 387–390. Dustin M. L., Baranski T. J., Sampath D. and Kornfeld S. (1995) A novel mutagenesis strategy identifies distantly spaced amino acid sequences that are required for the phosphorylation of both the oligosaccharides of procathepsin D by N-acetylglucosamine 1-phosphotransferase. J. Biol. Chem. 270, 170–179. Eckhardt M., Muhlenhoff M., Bethe A. and Gerardy-Schahn R. (1996) Expression cloning of the Golgi CMP-sialic acid transporter. Proc. Natl Acad. Sci. USA 93, 7572–7576. Fujita-Yamaguchi Y., Sato Y. and Kathuria S. (1985) Removal of sialic acids from the purified insulin receptor results in enhanced insulin-

binding and kinase activities. Biochem. Biophys. Res. Commun. 129, 739–745. Fukumoto H., Cheung B. S., Hyman B. T. and Irizarry M. C. (2002) Beta-secretase protein and activity are increased in the neocortex in Alzheimer disease. Arch. Neurol. 59, 1381–1389. Hardy J. and Selkoe D. J. (2002) The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 297, 353–356. Ho A. and Sudhof T. C. (2004) Binding of F-spondin to amyloid-beta precursor protein: a candidate amyloid-beta precursor protein ligand that modulates amyloid-beta precursor protein cleavage. Proc. Natl Acad. Sci. USA 101, 2548–2553. Holsinger R. M., McLean C. A., Beyreuther K., Masters C. L. and Evin G. (2002) Increased expression of the amyloid precursor betasecretase in Alzheimer’s disease. Ann. Neurol. 51, 783–786. Hussain I., Powell D., Howlett D. R. et al. (1999) Identification of a novel aspartic protease (Asp 2) as beta-secretase. Mol. Cell. Neurosci. 14, 419–427. Kitazume S., Tachida Y., Oka R., Shirotani K., Saido T. C. and Hashimoto Y. (2001) Alzheimer’s beta-secretase, beta-site amyloid precursor protein-cleaving enzyme, is responsible for cleavage secretion of a Golgi-resident sialyltransferase. Proc. Natl Acad. Sci. USA 98, 13 554–13 559. Kitazume S., Tachida Y., Oka R. Kotani N., Ogawa K., Suzuki M., Dohmae N., Takio K., Saido T. C. and Hashimoto Y. (2003) Characterization of alpha 2,6-sialyltransferase cleavage by Alzheimer’s beta-secretase (BACE1). J. Biol. Chem. 278, 14865– 14871. Kitazume S., Nakagawa K., Oka R. et al. (2005) In vivo cleavage of alpha2,6-sialyltransferase by Alzheimer beta-secretase. J. Biol. Chem. 280, 8589–8595. Kitazume-Kawaguchi S., Dohmae N., Takio K., Tsuji S. and Colley K. J. (1999) The relationship between ST6Gal I Golgi retention and its cleavage-secretion. Glycobiology 9, 1397–1406. Koike H., Tomioka S., Sorimachi H., Saido T. C., Maruyama K., Okuyama A., Fujisawa-Sehara A., Ohno S., Suzuki K. and Ishiura S. (1999) Membrane-anchored metalloprotease MDC9 has an alpha-secretase activity responsible for processing the amyloid precursor protein. Biochem. J. 343, 371–375. Laemmli U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680– 685. Lammich S., Kojro E., Postina R., Gilbert S., Pfeiffer R., Jasionowski M., Haass C. and Fahrenholz F. (1999) Constitutive and regulated alpha-secretase cleavage of Alzheimer’s amyloid precursor protein by a disintegrin metalloprotease. Proc. Natl Acad. Sci. USA 96, 3922–3927. Maruyama K., Terakado K., Usami M. and Yoshikawa K. (1990) Formation of amyloid-like fibrils in COS cells overexpressing part of the Alzheimer amyloid protein precursor. Nature 347, 566–569. McFarlane I., Georgopoulou N., Coughlan C. M., Gillian A. M. and Breen K. C. (1999) The role of the protein glycosylation state in the control of cellular transport of the amyloid beta precursor protein. Neuroscience 90, 15–25. Opdenakker G., Rudd P. M., Ponting C. P. and Dwek R. A. (1993) Concepts and principles of glycobiology. FASEB J. 7, 1330–1337. Pahlsson P. and Spitalnik S. L. (1996) The role of glycosylation in synthesis and secretion of beta-amyloid precursor protein by Chinese hamster ovary cells. Arch. Biochem. Biophys. 331, 177–186. Saito F., Yanagisawa K. and Miyatake T. (1993) Soluble derivatives of beta/A4 amyloid protein precursor in human cerebrospinal fluid are



both N- and O-glycosylated. Brain Res. Mol. Brain Res. 19, 171–174. Selkoe D. J. (2001) Alzheimer’s disease: genes, proteins, and therapy. Physiol. Rev. 81, 741–766. Sinha S., Anderson J. P., Barbour R. et al. (1999) Purification and cloning of amyloid precursor protein beta-secretase from human brain. Nature 402, 537–540. Skovronsky D. M., Moore D. B., Milla M. E., Doms R. W. and Lee V. M. (2000) Protein kinase C-dependent alpha-secretase competes with beta-secretase for cleavage of amyloid-beta precursor protein in the trans-golgi network. J. Biol. Chem. 275, 2568– 2575. Vassar R., Bennett B. D., Babu-Khan S. et al. (1999) Beta-secretase cleavage of Alzheimer’s amyloid precursor protein by the transmembrane aspartic protease BACE [see comments]. Science 286, 735–741. Watanabe I., Wang H. G., Sutachan J. J., Zhu J., Recio-Pinto E. and Thornhill W. B. (2003) Glycosylation affects rat Kv1.1 potassium

channel gating by a combined surface potential and cooperative subunit interaction mechanism. J. Physiol. 550, 51–66. Weinstein J., Lee E. U., McEntee K., Lai P. H. and Paulson J. C. (1987) Primary structure of beta-galactoside alpha 2,6-sialyltransferase. Conversion of membrane-bound enzyme to soluble forms by cleavage of the NH2-terminal signal anchor. J. Biol. Chem. 262, 17735–17743. Wolfe M. S., Xia W., Ostaszewski B. L., Diehl T. S., Kimberly W. T. and Selkoe D. J. (1999) Two transmembrane aspartates in presenilin-1 required for presenilin endoproteolysis and gamma-secretase activity. Nature 398, 513–517. Yang L. B., Lindholm K., Yan R. et al. (2003) Elevated beta-secretase expression and enzymatic activity detected in sporadic Alzheimer disease. Nat. Med. 9, 3–4. Yazaki M., Tagawa K., Maruyama K., Sorimachi H., Tsuchiya T., Ishiura S. and Suzuki K. (1996) Mutation of potential N-linked glycosylation sites in the Alzheimer’s disease amyloid precursor protein (APP). Neurosci. Lett. 221, 57–60.
