Specific antibody binding to the APP672–699 region shifts APP processing from α- to β-cleavage

Alzheimer's disease (AD), a progressive neurodegenerative disorder that is the most common cause of dementia in the elderly, is characterized by the accumulation of amyloid-β (Aβ) plaques and neurofibrillary tangles, as well as a progressive loss of synapses and neurons in the brain. The major pertinacious component of amyloid plaques is Aβ, a variably sized peptide derived from the integral membrane protein amyloid precursor protein (APP). The Aβ region of APP locates partly within its ecto- and trans-membrane domains. APP is cleaved by three proteases, designated as α-, β-, and γ-secretases. Processing by β- and γ-secretase cleaves the N- and C-terminal ends of the Aβ region, respectively, releasing Aβ, whereas α-secretase cleaves within the Aβ sequence, releasing soluble APPα (sAPPα). The γ-secretase cleaves at several adjacent sites to yield Aβ species containing 39–43 amino acid residues. Both α- and β-cleavage sites of human wild-type APP are located in APP672–699 region (ectodomain of β-C-terminal fragment, ED-β-CTF or ED-C99). Therefore, the amino acid residues within or near this region are definitely pivotal for human wild-type APP function and processing. Here, we report that one ED-C99-specific monoclonal antibody (mAbED-C99) blocks human wild-type APP endocytosis and shifts its processing from α- to β-cleavage, as evidenced by elevated accumulation of cell surface full-length APP and β-CTF together with reduced sAPPα and α-CTF levels. Moreover, mAbED-C99 enhances the interactions of APP with cholesterol. Consistently, intracerebroventricular injection of mAbED-C99 to human wild-type APP transgenic mice markedly increases membrane-associated β-CTF. All these findings suggest that APP672–699 region is critical for human wild-type APP processing and may provide new clues for the pathogenesis of sporadic AD.

APP 673 mutation (A-V) and APP 693 mutation (E-G) can enhance Ab production and accelerate formation of amyloid fibrils. [8][9][10] APP 682 mutation (E-K) blocks APP b 0 -site and shifts cleavage to b-site, thus increasing Ab 1-40/42 production. 6 Although sporadic AD (sAD), the more common type of AD comprising 90 to 95% of all AD cases, lacks mutations in the APP gene, region-specific protein modifications within the ED-C99 region may affect wild-type APP processing similarly to APP gene mutations. For example, phosphorylation of ED-C99 at the threonine 687 (of APP 770 isoform, or corresponding threonine 668 of APP 751 isoform; Supplementary Figure S1) facilitates APP processing by g-secretase. 11 Therefore, the elucidation of potential influences of region-specific modifications, induced by either endogenous or exogenous molecules, on wild-type APP processing would be especially critical for clarifying the mechanisms underlying the pathogenesis of sAD.
To confirm this hypothesis, we used one mouse monoclonal antibody specifically recognizing ED-C99 (mAb ED-C99 ) with its epitope at APP 674-679 (Supplementary Figure S1). The influences of mAb ED-C99 binding on human wild-type APP processing were evaluated in vitro using Chinese hamster ovary cells expressing human wild-type APP (CHO/APP wt cells) and cortical neurons derived from human wild-type APP transgenic (TgAPP wt ) mice. The in vitro effects of ED-C99 binding with mAb ED-C99 on wild-type APP processing were further evaluated and confirmed in vivo using TgAPP wt mice and 5 Â FAD transgenic mice (Tg6799 line).

Results
Specific binding of mAb ED-C99 inhibits a-but promotes b-cleavage of human wild-type APP. Western blotting (WB) analysis demonstrated that, compared with IgG 1 isotype, 2 h treatment of CHO/APP wt cells with mAb ED-C99 dose-dependently inhibits APP a-cleavage, as evidenced by the markedly decreased sAPPa and a-CTF levels ( Figure 1a). In contrast, neither mAb22C11 (anti-APP 66-81 antibody) nor mAb2B3 (sAPPa-specific antibody) inhibited the a-cleavage of APP ( Figure 1b). Most notably, mAb ED-C99 shifted APP processing from a-to b-cleavage, as indicated by decreased sAPP-a and a-CTF levels in combination Figure 1 Treatment with mouse monoclonal-specific anti-ED-C99 antibody (mAb ED-C99 ) markedly inhibits a-cleavage but promotes b-cleavage of human wild-type APP. (a) Human wild-type APP stably transfected CHO (CHO/APP wt ) cells were plated in 24-well plates at 5 Â 10 5 /well and treated with mAb ED-C99 at 0-1.25 mg/ml as indicated. (b) CHO/APP wt cells were treated with mAb ED-C99 , mAb22C11 (m22C11, recognizes APP 66-81 ), or mAb2B3 (specifically and structurally recognizes sAPPa, but not full-length APP) antibodies, or isotype IgG 1 control at 1.25 mg/ml. Immediately after 2 h treatment, cell supernatants were collected for western blotting (WB) analysis of sAPPa (using mAb2B3, upper panels) and Ab secretion (using a monoclonal Ab 1-12 antibody BAM10, mAbBAM10, middle panels); cell lysates were prepared for WB analysis of APP processing products (using a polyclonal anti-C-terminal APP 751/770 antibody, pAb751/770) and b-actin (internal control, lower panels). (c) Primary neuronal cells were cultured from cortical tissues of 1-day-old TgAPP wt mouse pups and seeded into 24-well-plates at 2 Â 10 5 /well for 18 h. These primary neuronal cells were treated with mAb ED-C99 or IgG 1 isotype control at 1.25 mg/ml for 2 h and then cell cultured media were collected for WB analysis of sAPPa (left upper panel) and secreted Ab (left lower panel) levels as indicated; cell lysates were prepared for WB analysis of full-length APP (holo APP), APP-CTFs (right upper panel), and b-actin (right lower panel). These WB data are representative of four independent experiments with similar results. (d) In addition, the cell culture media were collected from the separated primary cortical neuronal cells following 8 h treatment with mAb ED-C99 or IgG 1 isotype control at 1.25 mg/ml for Ab 1-40/42 and sAPPa-ELISA. The results were presented as ng of Ab 40/42 or sAPPa per mg of total intracellular proteins (mean ± S.D.; ***Po0.001). These ELISA data are representative of three independent experiments with similar results Binding on APP 672-699 promotes APP b-cleavage S Li et al with clearly elevated b-CTF level ( Figure 1b). However, mAb ED-C99 did not increase Ab production, as assessed using mAbBAM10 (recognizes Ab 1-12 ; Figures 1a and b, middle panels). Consistent with these findings in CHO/APP wt cells, 2 h treatment of primary cultured cortical neurons derived from TgAPP wt mice with mAb ED-C99 dramatically inhibited a-cleavage but enhanced b-cleavage of APP, as indicated by decreased levels of sAPPa and a-CTF as well as markedly increased b-CTF generation, compared with IgG 1 -treated control cells (Figure 1c). In addition, sAPPa levels were also significantly decreased with mAb ED-C99 treatment (Figure 1d, right panel). However, as seen with CHO/APP wt cells, mAb ED-C99 did not significantly alter Ab 1-40/42 productions (Figure 1c, left lower panel) and their levels ( Figure 1d, left panel). Altogether, these results indicate that specific binding of mAb ED-C99 to the ectodomain of b-CTF (ED-C99) reduces a-secretase processing while increasing b-secretase processing of APP.
F(ab 0 ) 2 fragment of mAb ED-C99 sufficiently inhibits a-but promotes b-cleavage of human wild-type APP. In living cells, the Fc fragment of antibody binds nonspecifically to cell surface Fc receptors. In order to determine whether nonspecific binding of mAb ED-C99 to the cell surface Fc receptors is necessary to inhibit APP a-processing, we generated the F(ab 0 ) 2 fragment of mAb ED-C99 that lacks the Fc fragment. Consistent with the results obtained with mAb ED-C99 , 2 h treatment of CHO/APP wt cells with mAb ED-C99 F(ab 0 ) 2 fragment dose-dependently reduced sAPPa expression  (Figure 2d). The increased level of b-CTF is further confirmed using mAbBAM10 that is specific for b-CTF but not a-CTF (Figure 2c, middle panels).
Most importantly, the monoclonal anti-b-actin antibody clearly detected not only b-actin but also IgG 1 heavy and light chains in the whole mAb ED-C99 -treated condition. In contrast, we only observed IgG 1 light chain in the mAb ED-C99 F(ab 0 ) 2 fragment-treated condition, confirming the purity of the prepared F(ab 0 ) 2 fragment. We did not observe IgG 1 heavy and light chains in the IgG 1 -treated condition, suggesting that both mAb ED-C99 F(ab 0 ) 2 fragment and whole mAb ED-C99 can specifically bind to membrane-associated full-length human wild-type APP, whereas control IgG 1 cannot (Figure 2c, lower panels).
Specific binding of mAb ED-C99 or its F(ab 0 ) 2 fragment directly inhibits ADAM10-mediated a-cleavage of human wild-type APP. We hypothesized that the decreased sAPPa and a-CTF levels elicited by mAb ED-C99 or its F(ab 0 ) 2 fragment might be because of the direct inhibition of ADAM10-mediated a-cleavage of human wild-type APP. As expected, compared with IgG 1 -treated control, WB analysis showed that both whole mAb ED-C99 and F(ab 0 ) 2 fragment significantly reduced WB band density ratio of a-CTF to full-length human wild-type APP, after incubation of full-length recombinant human wild-type APP protein with active ADAM10 in a cell-free system ( Figure 3). These results clearly suggest that the binding of mAb ED-C99 or its F(ab') 2 fragment to ED-C99 blocks ADAM10 from proteolysing human wild-type APP at a-cleavage sites.
Specific binding of mAb ED-C99 reduces APP endocytosis while increasing cell surface b-CTF. After reaching cell surface via secretory pathway, the matured human wild-type APP molecules could either be predominantly proteolyzed by a-secretases or rapidly endocytosed into the early endosomes through endocytosis pathway and then metabolized by b-and g-secretases to generate Ab. 12,13 As mAB ED-C99 did not alter levels of Ab, we hypothesized that specific binding to ED-C99 may reduce APP endocytosis. In order to confirm whether mAb ED-C99 binding to APP 672-699 region could modulate human wild-type APP endocytosis, we further assessed the impacts of mAb ED-C99 on full-length APP accumulations on the cell surface using the cell surface biotinylation technique. 14 As compared with IgG 1 isotype control, WB analysis of biotinylated proteins revealed that 2 h treatment of CHO/APP wt cells with mAb ED-C99 markedly increased cell surface full-length APP level (Figure 4a, right panel) as well as significantly enhanced WB band density ratio of cell surface full-length APP to pan cadherin (Figure 4b, right panel). In parallel with these findings in CHO/APP wt cells, 2 h treatment of primary cultured cortical cells derived from TgAPP wt mice with mAb ED-C99 also significantly increased cell surface fulllength APP level ( Figure 4c, right panel) as well as significantly increased WB band density ratio of cell surface full-length APP to pan cadherin ( Figure 4d, right panel). The accumulation of cell surface APP was confirmed by flow cytometry that revealed that 2 h treatment of CHO/APP wt cells with mAb ED-C99 led to a significant higher percentage of APP-positive cells when compared with IgG 1 isotype control (Figures 4e and f). These results suggest that specific antibody binding to ED-C99 indeed reduces APP endocytosis. Interestingly, mAb ED-C99 also dramatically enhanced cell surface b-CTF levels (Figures 4a and c, right panels) as well as significantly increased WB band density ratio of b-CTF to pan cadherin (Figures 4b and d, left panels) in both CHO/APP wt cells and primary cultured cortical cells derived from TgAPP wt mice. The markedly increased cell surface b-CTF level in both cell types following treatment with mAb ED-C99 were further confirmed by mAbBAM10, an antibody specifically recognizing b-CTF but not a-CTF (Supplementary Figure S2). Taken together, these findings suggest that the specific bindings of mAb ED-C99 to APP 672-699 region inhibits human wild-type APP a-cleavage, while also blocking human wild-type APP endocytosis and promoting its b-cleavage and/or cell surface accumulation of b-CTF.
Specific mAb ED-C99 binding enhances the colocalization of human wild-type APP with cholesterol. As growing evidence have suggested that cholesterol is of particular importance in regulating APP processing, 15-17 favoring b-cleavage and amyloidogenic processing, we further determined colocalization of human wild-type APP with cholesterol in CHO/APP wt cells following 2 h mAb ED-C99 treatment. Although 2 h treatment with mAb ED-C99 increased the cholesterol level on both cellular and subcellular plasma membranes (as indicated by dispersed fillipin staining, Figure 5a), this enhanced human wild-type APP with cholesterol colocalization was primarily observed on cell surface but rarely in intracellular compartments (Supplementary Figure S3), indicating that the mAb ED-C99 -induced human Figure 2 The mAb ED-C99 F(ab 0 ) 2 fragment binding sufficiently enhances human wild-type APP b-cleavage. CHO/APP wt cells were treated with F(ab 0 ) 2 fragment of mAb ED-C99 at 0-2.5 mg/ml. (a) Immediately following 2 h treatment, cell supernatants were collected for WB analysis of sAPPa (using mAb2B3, upper panel), Ab secretion (using mAbBAM10, middle panel), and APP processing (using pAb751/770, lower panel). (b) The secreted sAPPa and Ab 40/42 levels were analyzed by ELISA and presented as ng of Ab 1-40/42 or sAPPa per mg of total intracellular proteins (mean±S.D.; ***Po0.001). (c) CHO/APP wt cells were treated with mAb ED-C99 F(ab 0 ) 2 fragment, mAb ED-C99 , or isotype control IgG 1 at 1.25 mg/ml for 2 h, washed three times with PBS containing CaCl 2 and MgSO 4 (PBS-CM), and then cell lysate portions of these cells were directly subjected to WB analysis using pAb751/770 or mAbBAM10 (left panels). The remaining cells were biotinylated with Sulfo-NHS-LC-Biotin dissolved in ice-cold borate buffer, quenched with NH 4 Cl-PBS-CM, and lysed. These cells lysates were then immunoprecipitated using Neutravidin beads. The intracellular (intra) proteins obtained by IP/Neutravidin depletion (middle panels) and the cell surface (cell surf) proteins obtained by IP/Neutravidin precipitation (right panels) were subjected to WB analysis using pAb751/770 and mAbBAM10. (d) For WB quantitative analysis, the band density ratio of cell surface b-CTF to cell surface a-CTF was analyzed and presented as mean±S.D. As indicated, there is no statistically significant difference between mAb ED-C99 F(ab 0 ) 2 fragment and whole mAb ED-C99 treatments (P40.05). The data are representative of three independent experiments with similar results mAb ED-C99 promotes b-cleavage of human wild-type APP in vivo. Eight-month-old TgAPP wt female mice were treated with mAb ED-C99 or isotype IgG 1 as negative control via intracerebroventricular (i.c.v.) injection. At 24 h after treatment, we found that WB band ratio of membrane-bound b-CTF to total APP in the mAb ED-C99 -treated group was significantly higher than that in the IgG 1 -treated control group (Figure 7a, middle panels, and Figure 7b, upper panel). Most interestingly, consistent with our in vitro data, mAb ED-C99 did not alter the levels of Ab species (Figure 7a, lower panels). Ab ELISA analysis of brain homogenates also confirmed that Ab 1-40/42 species were not significantly changed in the mAb ED-C99 -treated group compared with IgG 1 control group (Figure 7b, lower panel). In addition, 8-month-old 5 Â FAD female transgenic mice (Tg6799 line) were treated with mAb ED-C99 or isotype IgG 1 as negative control via i.c.v. injection. At 24 h after treatment, we found that immunohistochemical staining also disclosed comparable amount of b-amyloid plaques in retrosplenial cortex (RSC), entorhinal cortex (EC), and hippocampus (H) regions of 5 Â FAD transgenic mouse brains (Supplementary Figure S4).

Discussion
Mutations in APP gene cause early onset of autosomaldominant AD. 3,[18][19][20] Relative to their wild-type homologs, the English (APP 677 H-R) and Tottori (APP 678 D-N) substitutions accelerate the kinetics of Ab secondary structure change from statistical coil-a/b-b and produce oligomer size distributions skewed to higher order that are more toxic to cultured neuronal cells than wild-type oligomers. 21 The Icelandic APP 673 mutation (A-V) affects APP processing, resulting in enhanced Ab production (quantity) and formation of amyloid fibrils in vitro (quality). 8 In contrast, alternative APP 673 mutation (A-T) results in an B40% reduction in the formation of amyloidogenic peptides and therefore protects against AD and cognitive decline in the elderly. 22 Thus, these pathogenic mutations located in the APP 672 to APP 699 region of APP ectodomain (also named ED-C99) encompassing APP a-and b-cleavage sites can alter b-cleavage and Abrelated AD pathology. However, unlike autosomal-dominant fAD, sAD patients generally lack mutations of APP gene. Therefore, the mechanisms underlying pathogenesis of sAD Cell surface a-cleavage is the predominant processing pathway for human wild-type APP. In the present study, we found that the specific binding of the APP ED-C99 region with mAb ED-C99 dose-dependently blocks a-but promotes b-cleavage of human wild-type APP (Figure 1). These effects were further confirmed by another specific antibody, 4G8, recognizing the ED-C99 domain (with epitope at APP 688-695 , data not shown). In contrast, N-terminal APP antibody 22C11 (epitope of APP 66-81 ) or sAPPa-specific antibody mAb2B3 (epitope of APP 672-688 , absence of affinity to full-length APP) showed no effect on human wild-type APP processing. This lack of impact of mAb22C11 and mAb2B3 further indicates that the effect of mAb ED-C99 on human wild-type APP processing is region specific. In addition, these modulations on human wild-type APP a/b-cleavage induced by mAb ED-C99 can be recapitulated by using the F(ab 0 ) 2 fragment of mAb ED-C99 ( Figure 2). As F(ab 0 ) 2 lacks the nonspecific Fc binding subunit, these results confirm that mAb ED-C99 -induced modification of APP processing is ED-C99 region specific. Our present study further reveals that this a-cleavage inhibiting activity is due, in part, to the direct physical blocking of ADAM10 (Figure 3).

Figure 5
Colocalization of human wild-type APP with cholesterol in CHO/APP wt cells after mAb ED-C99 treatment. (a) CHO/APP wt cells were plated to 8-well slide chamber and then, after overnight incubation, the cells were treated with mAb ED-C99 or IgG 1 isotype control for 2 h. These cells were stained by Filipin in strict accordance with the manufacturer's instructions of the cholesterol assay kit. (b) After Filipin staining and washing, some of these cells were permeabilized with 0.05% Triton X-100 for 5 min, washed, and stained with rabbit anti-APP-C-terminal antibody overnight at 41C. Alexa Fluor 594 Donkey anti-rabbit IgG was used to detect APP signals. Confocal images were taken by Olympus fluoview FV1000 laser scanning confocal microscope (Tokyo, Japan) Figure 4 Cell surface b-CTF and full-length APP are increased following treatment with mAb ED-C99 . (a) CHO/APP wt cells in 24-well plates (5 Â 10 5 /well) were treated with mAb ED-C99 or IgG 1 isotype control at 1.25 mg/ml for 2 h, washed three times with PBS-CM, and then cell lysate portions of these cells were directly subjected to WB analysis using pAb751/770 (left panels). The remaining cells were biotinylated with Sulfo-NHS-LC-Biotin, quenched with NH 4 Cl-PBS-CM, lysed, and immunoprecipitated (IP) using Neutravidin beads. The intracellular proteins obtained by IP/Neutravidin depletion (middle panels) and the cell surface (cell surf) proteins obtained by IP/Neutravidin precipitation (right panels) were subjected to WB analysis using pAb751/770. As shown below each panel, as an internal control, b-actin was analyzed for cell lysates and intracellular (intra) proteins and pan cadherin was analyzed for total cell surface proteins. (b) For WB quantitative analysis, band density ratios of cell surface b-CTF or fulllength APP to pan cadherin were analyzed and presented as mean±S.D. (**Po0.01, ***Po0.001). The WB data are representative of three independent experiments with similar results. (c) Primary neuronal cells were cultured from cortical tissues of 1-day-old TgAPP wt mouse pups and replated in 24-well plate at 2 Â 10 5 /well overnight. These primary neuronal cells were treated with mAb ED-C99 or IgG 1 isotype control at 1.25 mg/ml for 2 h, washed three times with PBS-CM and then cell lysates were directly subjected to WB analysis using pAb751/770 (left panels). The remaining cells were biotinylated, immunoprecipitated with Neutravidin beads, and then subjected to isolation of intracellular and cell surface proteins. The intracellular proteins obtained by IP/Neutravidin depletion (middle panels) and the cell surface (cell surf) proteins obtained by IP/ Neutravidin isolation (right panels) were subjected to WB analysis using pAb751/770.   Figure 7 The mAb ED-C99 promotes APP b-secretase processing in vivo. TgAPP wt female mice at 8 months of age were treated with mAb ED-C99 or control IgG 1 at 5 mg/ mouse by intracerebroventricular (i.c.v.) injection and killed 24 h after the treatment (n ¼ 6). (a) Cytosolic-(left panels) and membrane-associated proteins (right panels) prepared from mouse brain homogenates were subjected to WB analysis for APP processing. (b) For WB quantitative analysis, band density ratio of membrane-associated b-CTF to membrane-associated total APP was analyzed (upper panel). Ab 40/42 was also analyzed by ELISA (lower panel, n ¼ 6). The results are presented as pg of Ab 40/42 per mg of total intercellular proteins (mean ± S.D.; ***Po0.001) Previous studies have suggested that intracellular trafficking is pivotal for human wild-type APP proteolysis. 12,13 In contrast, non-amyloidogenic processing occurs mainly at the cell surface, where a-secretases are present. Amyloidogenic processing involves transit through the endocytic organelles, where APP encounters b-and g-secretases. 13,15 In our present study, mAb ED-C99 enhanced cell surface APP accumulation, which was confirmed by both WB and flow cytometry analyses, indicating that ED-C99 binding reduces the APP endocytic pathway. Blockage of human wild-type APP a-cleavage by mAb ED-C99 may consequently lead to an elevated b-cleavage that has been confirmed by dramatically increased b-CTF. This suggests that ED-C99 may bind and trap APP in the membrane (or endosomes in the process of fusing with the plasma membrane), making the proteolytic sites for ADAM10 inaccessible, and thus shifting the proteolysis of APP in the plasma membrane or endosomes to b-secretase cleavage.
This proposed blockage of human wild-type APP a-cleavage by mAb ED-C99 may consequently lead to an elevated b-cleavage that has been confirmed by the dramatically increased b-CTF levels on the cell surface ( Figure 4). Our findings suggest that this elevated b-cleavage can be because of the cell surface APP with cholesterol colocalization ( Figure 5). Indeed, growing evidence has demonstrated that cholesterol is of particular importance in regulating a-and b-cleavage of APP. [15][16][17] Cholesterol can bind to APP 696-699 , an N-loop structure at the end of the ED-C99 region, thereby inhibiting a-cleavage while favoring b-cleavage of APP. In contrast to b-secretase, previous studies have indicated that g-secretase cleavage of C99 appears to occur primarily in rafts located in the endosomes. 12,[23][24][25][26][27][28][29] Thus, the specific binding of mAb ED-C99 to APP inhibits APP a-cleavage and endocytosis, thereby locking APP on the cell surface to be cleaved by b-secretase while functionally blocking g-secretase from producing Ab. This would explain how the mAb ED-C99 promoted cell surface b-CTF accumulation with a concordant rise in Ab protein levels.
In addition to cholesterol, it has also been shown that LRP1 plays a crucial role in APP processing. [30][31][32] Multiple APP processing steps may be modulated by LRP1, most of which can be attributed to its ability to bind and endocytose APP via its LRP1-CT domain. In the absence of LRP1, APP internalization rates are reduced by B50%, while cell surface APP and APP-CTFs accumulate. 32 In order to verify the possibility that mAb ED-C99 alters APP processing by reducing endocytosis, CHO/APP wt /LRP1-CT cells were treated with mAb ED-C99 . Our results demonstrated a market restoration of the inhibited endocytosis, as shown by the decreased cell surface b-CTF level (Figure 6a) and increased Ab levels (Figure 6b). Surprisingly, the reduction of sAPPa by mAb ED-C99 was attenuated by LRP1-CT overexpression (Figure 6b), suggesting that LRP1 may block the mAb ED-C99 binding on APP 672-699 region and subsequently release APP from mAb ED-C99 -mediated inhibition of a-secretase.
To confirm whether the promotion of cell surface b-CTF accumulation induced by binding of mAb ED-C99 to ED-C99 in vitro could be repeated in vivo, TgAPP wt mice at 8 months of age were treated (i.c.v.) with mAb ED-C99 or IgG 1 as negative control. We found that mAb ED-C99 treatment significantly increased cell surface b-CTF, as indicated by elevated ratio of b-CTF/total APP (Figure 7b). Similar to our in vitro study, mAb ED-C99 treatment also yielded no significant changes of Ab production compared with the IgG 1 control. Although the amyloid cascade hypothesis is potentially viable in cases of genetic mutation-caused autosomal-dominant fAD, accumulating evidence suggests that it may not apply in the vast majority of patients with late-onset sAD lacking mutations of APP/presenilin genes.
These works suggest several possibilities in terms of defining a possible etiology and treatment target for both fAD and sAD. First, Ab-related plaques is one relatively common finding in the non-demented elderly. In fact, recent studies have demonstrated that accumulation of b-CTF may have direct deleterious effects on cognitive function. For example, inhibition of b-site amyloid cleaving enzyme (BACE) rescued synaptic/memory deficits in a mouse model of familial Danish dementia. 33 However, g-secretase inhibition worsened memory deficits in these mice that correlated with increased levels of APP a/b-CTFs in synaptic fractions of the hippocampus. 34 In another report, prolonged (8 days) treatment with g-secretase inhibitors (GSIs) produced no positive effects on memory deficits of older Swedish APP transgenic mice, but induced cognitive deficits in young Swedish APP transgenic mice or wild-type mice. 35 Indeed, a recent phase III clinical trial with the GSI Semagacestat was halted because it worsened clinical measures of cognition and the ability to perform activities of daily living. These results suggest that b-CTF rather than Ab may be more directly responsible for causing cognitive impairment associated with AD and that GSIs may worsen cognitive impairment by enhancing the accumulation of b-CTF. Our results also point to a need to identify, epidemiologically, the presence or absence of mAb ED-C99 -like antibodies or proteins that may increase with age and correlate with onset of AD-like signs and symptoms. In addition, this work suggests that such an antibody or protein could reduce a-secretase function and thus generation of sAPPa, a peptide that is neuroprotective. [36][37][38] In summary, the present study verifies our hypothesis that ED-C99 region (APP 672-699 ) is critical for human wild-type APP processing. Specific binding of this region will direct inhibit APP a-secretase activity and reduce APP endocytosis, thereby enhancing cell surface b-CTF accumulation. These effects may have deleterious effects of cognition, and this should be further explored in future studies. Modifications (physical of functional interactions) of this region with either exogenous or endogenous molecules will affect human wildtype APP processing, potentially enhancing or ameliorating the development of AD (Supplementary Figure S5).
Cell culture and treatment. CHO cells engineered to express wild-type human APP (CHO/APP wt ) were kindly provided by Dr. Steanie Hahn and Dr. Sascha Weggen (University of Heinrich Heine, Düsseldorf, Germany). Plasmid PLHCX-LRP1-CT was a generous gift from Dr. David Kang (University of South Florida, Tampa, FL, USA). LRP1-CT was subcloned to PCDNA vector by PCR and restrict enzyme HindIII and NotI digestion. The primers were used as follows: forward 5 0 -AGCTCGTTTAGTGAACCGTCAGATC-3 0 and reverse 5 0 -ATGC GGCCGCTCATGCCAAGGGGTCCCCTATC-3 0 . Stable cell lines were generated by transfection of PCDNA-LRP1-CT to CHO/APP wt cells and single colony was picked up after G418 administration. These cells were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum, 1 mM sodium pyruvate, and 100 U/ml penicillin/streptomycin. 39 For TgAPP wt mouse-derived cortical neurons, cerebral cortices isolated from 1-day-old TgAPP wt mice were mechanically dissociated in trypsin (0.25%) individually after incubation for 15 min at 371C. Cells were collected after centrifugation at 1200 Â g, suspended in DMEM supplemented with 10% fetal calf serum, 10% horse serum, uridine (33.6 mg/ml, Sigma-Aldrich), and fluorodeoxyuridine (13.6 mg/ml, Sigma-Aldrich) and seeded in 24-well collagen-coated culture plates at 2.5 Â 10 5 cells per well. After reaching confluence (B70-80%), cells were treated with mAb ED-C99 at 0-1.25 mg/ml for 2 h. In additional experiments, cells were treated with mAb22C11 or mAb2B3 or mAb ED-C99 F(ab 0 ) 2 fragment at 1.25 mg/ml. Transgenic APP wt mice and i.c.v. injection. Transgenic wild-type B6.Cg-Tg (PDGFB-APP) 5Lms/J strain (TgAPP wt ) female mice and 5 Â FAD transgenic female mice (Tg6799 line) were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). All mice were housed and maintained in the Animal Facility of College of Medicine at University of South Florida. At 8 months of age, both mice were anesthetized using isoflurane (chamber induction at 4-5% isoflurane, intubation and maintenance at 1-2%). After reflexes were checked to ensure that mice were unconscious, they were positioned on a stereotaxic frame (Stoelting Lab Standard, Wood Dale, IL, USA) with ear-bars positioned and jaws fixed to a biting plate. The ED-C99-specific antibody mAb ED-C99 and isotype control IgG 1 were dissolved in sterile distilled water at a concentration of 1 mg/ml. mAb ED-C99 and control IgG 1 (5 ml) were injected into the left lateral ventricle with a microsyringe at a rate of 1 ml/min with the coordinates (bregma: À 0.6 mm anterior/posterior, þ 1.2 mm medial/lateral, and À 3.0 mm dorsal/ventral) according to our previous methods. 36,40 The needle was left in place for 5 min after injection before being withdrawn. At 24 h after i.c.v. injections, animals were killed with isofluorane and brain tissues were collected. All dissected brain tissues were rapidly frozen for analysis. All experiments involving mice were performed in compliance with the US Department of Health and Human Services Guide for the Care and Use of Laboratory Animals and in accordance with the guidelines of the University of South Florida Institutional Animal Care and Use Committee.
Cell surface biotinylation. Confluent cell cultures in dishes were washed three times with phosphate-buffered saline (PBS)-CM. Cells were biotinylated with 0.5 mg/ml Sulfo-NHS-LC Biotin dissolved in ice-cold borate buffer for 30 min at 41C. Biotin was changed twice during the 30-min incubation period. Biotinylation was quenched with three 50 mM NH 4 Cl-PBS-CM washes followed by two PBS washes. Cells were harvested and lysed in NP-40 buffer and protein concentration was determined via BCA. Equal amounts of protein were immunoprecipitated using Neutravidinagarose beads overnight at 41C. Proteins were eluted from the Neutravidinagarose beads by heating at 951C for 10 min.
WB analysis. Briefly, for WB analysis, cultured cells were lysed in ice-cold lysis buffer. Proteins were separated using 10% gel, transferred to 0.2-mm nitrocellulose membranes (Bio-Rad, Hercules, CA, USA) and visualized using standard immunoblotting protocol. All antibodies were diluted in Tris-buffered saline (TBS) containing 5% (w/v) nonfat dry milk. Blots were developed using the Luminol reagent (Thermo Fisher Scientific) and densitometric analysis was performed as described previously, using a Fluor-S MultiImager with Quantity One software (Bio-Rad). 40,41 Flow cytometry analysis. CHO/APP wt cells were plated in six-well plate and treated overnight with mAb ED-C99 or IgG 1 isotype control. The cells were then detached by trypsin, stained with mAb ED-C99 and Alexa Fluor 488-conjugated donkey anti-mouse IgG (Invitrogen, Gaithersburg, MD, USA) for 30 min on ice, washed with Flow buffer, and analyzed by Accuri C6 Flow Cytometer (Rochester, MN, USA).
Filipin staining. After treatment, the cells were fixed and Filipin staining was performed following the manufacturer's instructions of the cholesterol assay kit (AbCam).
ELISA. Total Ab species, including Ab 40/42 , in cell conditioned media and brain homogenates were detected by Ab 1-40/42 ELISA kits (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. The sAPPa level in cell conditioned media was determined using sAPPa ELISA kit (IBL). Ab and sAPPa levels are represented as ng/mg of total cellular protein.
Statistical analysis. Data are expressed as mean±S.D. of n independent experiments. Comparison between two groups was performed by Student's t-test. Po0.05 was considered statistically significant.