RNA Interference Silencing of the Adaptor Molecules ShcC and Fe65 Differentially Affect Amyloid Precursor Protein Processing and Aβ Generation*

The amyloid precursor protein (APP) and its pathogenic by-product amyloid-β protein (Aβ) play central roles in Alzheimer disease (AD) neuropathogenesis. APP can be cleaved by β-secretase (BACE) and α-secretase to produce APP-C99 and APP-C83. These C-terminal fragments can then be cleaved by γ-secretase to produce Aβ and p3, respectively. p3 has been reported to promote apoptosis, and Aβ is the key component of senile plaques in AD brain. APP adaptor proteins with phosphotyrosine-binding domains, including ShcA (SHC1), ShcC (SHC3), and Fe65 (APBB1), can bind to and interact with the conserved YENPTY motif in the APP-C terminus. Here we have described for the first time the effects of RNA interference (RNAi) silencing of ShcA, ShcC, and Fe65 expression on APP processing and Aβ production. RNAi silencing of ShcC led to reductions in the levels of APP-C-terminal fragments (APP-CTFs) and Aβ in H4 human neuroglioma cells stably overexpressing full-length APP (H4-FL-APP cells) but not in those expressing APP-C99 (H4-APP-C99 cells). RNAi silencing of ShcC also led to reductions in BACE levels in H4-FL-APP cells. In contrast, RNAi silencing of the homologue ShcA had no effect on APP processing or Aβ levels. RNAi silencing of Fe65 increased APP-CTF levels, although also decreasing Aβ levels in H4-FL-APP cells. These findings suggest that pharmacologically blocking interaction of APP with ShcC and Fe65 may provide novel therapeutic strategies against AD.

The current treatments for Alzheimer disease (AD), 3 the most prevalent cause of dementia in the elderly, provide only modest, temporary, and palliative benefits (1). This is because current treatments do not target disease progression and, particularly, cerebral accumulation of the amyloid-␤ protein (A␤), the key component of senile plaques in AD neuropathology. A␤ is generated via serial proteolytic cleavage of the amyloid precursor protein (APP) by ␤-secretase (BACE) and ␥-secretase. Specifically, full-length (FL) APP is first hydrolyzed by BACE to generate a 99-residue membrane-associated C-terminal fragment (CTF) (APP-C99) (2)(3)(4)(5). APP-C99 is further cleaved to release an ϳ4-kDa peptide, A␤, and the amyloid precursor protein intracellular domain. This cleavage is achieved by an unusual form of proteolysis in which the protein is cleaved within the transmembrane domain (at residue ϩ40 or ϩ42) by ␥-secretase (6 -8). The majority of APP is cleaved by ␣-secretase in the middle of the A␤ region of APP, precluding A␤ generation and leading to the release of a large ectodomain (␣-APPs), leaving behind a carboxyl-terminal fragment of 83 amino acids (APP-C83) in the membrane. Although proteolysis of APP-C99 by ␥-secretase produces A␤, proteolysis of APP-C83 by ␥-secretase produces p3, an amino-terminally truncated form of A␤ (9 -11), which has been shown to induce apoptosis mediated by activation of c-Jun N-terminal kinase, caspase 8, and caspase 3 (12). The cleavage of the APP cytoplasmic tail by ␥-secretase generates the amyloid precursor protein intracellular domain, which contains the strongly conserved YENPTY motif, which is also present in the cytodomains of several tyrosine-kinase receptors and in nonreceptor tyrosine kinase. The YENPTY sequence is a consensus motif for the binding of adaptor proteins that possess a phosphotyrosinebinding domain present in several APP adaptor proteins, such as the X11, Fe65, Shc, and JIP families (13). We have previously reported that RNA interference (RNAi) silencing of X11␣ in H4 human neuroglioma cells increases levels of the APP-CTFs and lowers A␤ levels by attenuating ␥-secretase-mediated APP cleavage (14).
ShcA (encoded by gene SHC1) and ShcC (encoded by gene SHC3), the phosphotyrosine-binding domain-containing adaptor proteins that signal to cellular differentiation and survival pathways, are other types of APP adaptor proteins that also bind to and interact with the YENPTY motif of APP (15,16).
Activation of Shc signal transduction pathways in vitro has been reported to trigger APP-Shc interaction in glial cells but not in neurons (15). To date, the effects of reduced expression of ShcA and ShcC on APP processing and A␤ production, the key components of AD neuropathogenesis, have not been assessed. For this purpose, we established RNAi for ShcA and ShcC in H4 human neuroglioma cells overexpressing FL-APP (H4-FL-APP cells) and APP-C99 (H4-APP-C99 cells) and evaluated the effects of RNAi-mediated silencing of ShcA and ShcC on APP processing and A␤ production.
Fe65, along with Fe65L1 and Fe65L2, are APP adaptor proteins, which bind to the YENPTY motif of APP via their phosphotyrosine-binding domains (13). Overexpression in Fe65 or Fe65L accelerates secretory processing and maturation of APP and promotes APPs and A␤ secretion in H4-FL-APP cells and in Madin-Darby canine kidney cells (17,18). However, the effects of RNAi knockdown of APBB1 or Fe65 on APP processing and A␤ production have not been assessed. Therefore, we established RNAi for Fe65 in H4-FL-APP cells and evaluated the effects of RNAi-mediated silencing of Fe65 on APP processing and A␤ production.

EXPERIMENTAL PROCEDURES
Cell Lines-We employed H4 human neuroglioma cells stably transfected to express FL-APP (H4-FL-APP cells) or APP-C99 (H4-APP-C99 cells) in the experiments. Peptide APP-C99 is the product of BACE, which therefore contains ␣-and ␥-but not ␤-cleavage sites. This cell line provides a valid system to assess whether any effects on APP processing is dependent on BACE-mediated APP processing. The cells were cultured in Dulbecco's modified Eagle's medium (high glucose) containing 9% heat-inactivated fetal calf serum, 100 units/ml penicillin, 100 g/ml streptomycin, 2 mM L-glutamine, and 200 g/ml G418.
Western Blot Analysis of APP Processing-Western blot analysis was performed as described by Xie et al. (19). Briefly, 40 g of total protein of each sample was subjected to SDS-PAGE using 4 -20% gradient Tris/glycine gels under reducing conditions (Invitrogen). Next, proteins were transferred to a polyvinylidene difluoride membrane (Bio-Rad) using a semidry electrotransfer system (Amersham Biosciences). Nonspecific proteins were blocked using 5% nonfat dry milk in TBST (Trisbuffered saline/Tween) for 1.5 h. Blots were then incubated with a primary antibody followed by a secondary antibody (horseradish peroxidase-conjugated anti-rabbit antibody 1:10,000; Pierce). Blots were washed once with 1ϫ TBST for 30 min between steps. Polyclonal antibody ShcA at a ratio of 1:2000 (catalog number 610081; Transduction Laboratories) was used to recognize ShcA (65 and 50 kDa), monoclonal antibody ShcC at a ratio of 1:1000 (catalog number 610642; BD Transduction Laboratories) was used to detect ShcC (55 kDa), and polyclonal antibody PA1-752 at a ratio of 1:1000 (ABR, Golden, CO) was used to detect Fe65 (90 kDa). Polyclonal antibody ab2077 at a ratio of 1:1000 (Abcam, Cambridge, MA) was used to detect BACE (70 kDa). Antibody A8717 at a ratio of 1:1000 (Sigma) was used to visualize FL-APP (110 kDa), APP-C83 (12 kDa), and APP-C99 (10 kDa) in the Western blot analysis. The intensity of signals was analyzed and quantified using an image program (NIH Image version 1.62). We used the levels of ␤-actin to normalize the levels of ShcA, ShcC, Fe65, FL-APP, APP-C99, and APP-C83 (i.e. determining the ratio of ShcA amount to ␤-actin amount) to control for loading differences in total protein amounts. We have presented the changes in the protein levels of ShcA, ShcC, Fe65, FL-APP, APP-C99, and APP-C83 in the cells treated with ShcA, ShcC, or Fe65 siRNAs as the percentage of those in the cells treated with control siRNA.
Quantitation of A␤ Using Sandwich Enzyme-linked Immunosorbent Assay-Following the treatment with saline, control, Fe65, ShcA, or ShcC siRNA-conditioned medium were collected, and secreted A␤ was measured by a sandwich enzymelinked immunosorbent assay as described by Xie et al. (19).
Specifically, 96-well plates were coated with mouse monoclonal antibodies specific to A␤40 (2G3) or A␤42 (21F12). Following blocking with Block Ace, the wells were incubated overnight at 4°C with test samples of conditioned cell culture medium, and then an anti-A␤ (␣-A␤-HR1) conjugated to horseradish peroxidase was added. Plates were then developed with tetramethylbenzidine reagent, and well absorbance was measured at 450 nm. A␤ levels in test samples were determined by comparison with the signal from unconditioned medium spiked with known quantities of A␤40 and A␤42.
Statistics-Analysis of variance with repeated measurements was employed to compare the difference from the control group. p values Ͻ0.05 were considered statistically significant.

ShcC RNAi Decreased the Levels of APP-CTF and A␤ in H4-FL-APP
Cells-We first established conditions under which RNAi silencing of ShcC significantly reduced ShcC levels in H4-FL-APP cells. The cells were harvested 48 h after being transfected with either control siRNA or ShcC siRNA and were subjected to Western blot analyses in which antibody ShcC was used to visualize the ShcC levels. Immunoblotting for ShcC revealed a visible reduction in the ShcC levels following ShcC siRNA treatment as compared with control siRNA treatment (Fig. 1A). ShcC siRNA treatment significantly reduced ShcC levels by 41% (normalized to ␤-actin) as compared with control siRNA treatment (Fig. 1B).
Next, we assessed the effects of RNAi-mediated silencing of ShcC on APP processing in H4-FL-APP cells by measuring protein levels of FL-APP, APP-C99, and APP-C83 following ShcC siRNA treatment. 48 h after transfection of ShcC siRNA or control siRNA, the cells were harvested and subjected to Western blot analyses in which antibody A8717 was used to detect FL-APP, APP-C99, and APP-C83. Immunoblot analysis of APP-CTFs revealed decreases in the levels of APP-C99 and APP-C83 in the cells treated with ShcC siRNA as compared with control siRNA (Fig. 2A). As a positive control, the ␥-secretase inhibitor DAPT was employed to induce the accumulation of APP-C99 and APP-C83. Meanwhile, no significant differences in the FL-APP levels were observed between ShcC siRNA-, control siRNA-and DAPT-treated cells. Quantifica- tion of FL-APP, APP-C99, and APP-C83 (normalized to ␤-actin) revealed that ShcC siRNA treatment led to a 47% decrease in the ratio of APP-C83 to FL-APP (*, p Ͻ 0.05) (Fig. 2B) and a similar (50%) decrease in the ratio of APP-C99 to FL-APP (*, p Ͻ 0.05) (Fig. 2C) as compared with control siRNA treatment.
Next, we assessed the effects of ShcC siRNA on A␤ levels in the conditioned medium. 48 h after treatment with control siRNA or ShcC siRNA, we measured A␤40 and A␤42 levels, normalizing to total protein levels. ShcC siRNA treatment significantly decreased both A␤40 and A␤42 levels as compared with control siRNA treatment (*, p Ͻ 0.05) (Fig. 2D) as with A␤40 12 pg/ml/protein (ShcC siRNA) versus 20 pg/ml/protein (control siRNA) and A␤42 2.4 pg/ml/protein (ShcC siRNA) versus 4 pg/ml/protein (control siRNA). Collectively, these data indicate that RNAi silencing of ShcC reduces levels of APP-C83 and APP-C99 as well as secreted A␤ in H4-FL-APP cells.
RNAi Silencing of ShcC Did Not Affect APP Processing and A␤ Levels in H4-APP-C99 Cells-To determine whether the decreases in the levels of APP-C83, APP-C99, and A␤ following RNAi silencing of ShcC in H4-FL-APP cells were due to alter-ations in BACE, we employed H4-APP-C99 cells. APP-C99 is the product of BACE and harbors ␣and ␥-cleavage but not ␤-cleavage sites. 48 h after transfection with either ShcC siRNA or control siRNA in H4-APP-C99 cells, the cells were harvested and subjected to Western blot analyses as described for ShcC in H4-FL-APP cells above. Immunoblotting for APP with antibody A8717 revealed no significant difference in the FL-APP levels in ShcC siRNAversus control siRNA-treated cells (Fig.  3A). Likewise, ShcC siRNA treatment did not alter levels of APP-C99 or APP-C83 as compared with control siRNA treatment (Fig. 3, A-C). Next, we measured A␤ levels in the conditioned medium 48 h following treatments with control siRNA or ShcC siRNA. ShcC siRNA treatment decreased levels of neither A␤40 with 27 pg/ml/protein (ShcC siRNA) versus 30 pg/ml/protein (control siRNA) nor A␤42 with 6.2 pg/ml/protein (ShcC siRNA) versus 6 pg/ml/protein (control siRNA), as compared with control siRNA normalized to total protein amount (Fig. 3D). Taken together, these findings suggest that the effects of RNAi silencing of ShcC on APP processing and A␤ generation are at least partially dependent on BACE cleavage of APP.
RNAi Silencing of ShcC Decreased the Levels of BACE in H4-FL-APP Cells-To further determine whether RNAi silencing of ShcC can decrease levels of APP-CTFs and A␤ by reducing levels of BACE, we next assessed the effects of RNAi silencing of ShcC on BACE levels in H4-FL-APP cells. 48 h after transfection of H4-FL-APP cells with ShcC siRNA or control siRNA, the cells were harvested and subjected to Western blot analyses in which antibody ab2077 was used to detect BACE levels. Immunoblotting revealed a visible decrease in the BACE levels in the cells treated with ShcC siRNA as compared with control siRNA (Fig. 4A). Quantification of the Western blots (normalized to ␤-actin) revealed that ShcC siRNA treatment decreased BACE levels by 38% (*, p Ͻ 0.05) (Fig. 4B) as compared with control siRNA treatment.
RNAi Silencing of ShcA Did Not Affect APP Processing and A␤ Levels in H4-FL-APP Cells-We next asked whether the other Shc family protein, ShcA (encoded by gene SHC1), another APP adaptor protein, can similarly affect APP processing and A␤ production in H4-FL-APP cells. For this purpose, we established ShcA RNAi in H4-FL-APP cells. 48 h after transfection of H4-FL-APP cells with ShcA siRNA or control siRNA, the cells  2). FL-APP immunoblotting revealed that there was no significant difference in the levels of FL-APP in the cells treated with control siRNA or ShcC siRNA. There was no significant difference in the amounts of ␤-actin in the control siRNA-and ShcC siRNA-treated H4-APP-C99 cells. B, APP processing assessed by quantifying the ratio of APP-C83 to FL-APP in the Western blot. ShcC siRNA treatment (black bar) did not alter the ratio of APP-C83 to FL-APP as compared with control siRNA treatment (white bar) normalized to ␤-actin. C, APP processing assessed by quantifying the ratio of APP-C99 to FL-APP in the Western blot. ShcC siRNA treatment (black bar) did not alter the ratio of APP-C99 to FL-APP as compared with control siRNA treatment (white bar) normalized to ␤-actin. D, effects of ShcC RNAi on A␤ levels in H4-APP-C99 cells. ShcC siRNA treatment (black bar) decreased neither A␤40 nor A␤42 levels as compared with control siRNA treatment (white bar) normalized to total protein amount. FEBRUARY 16, 2007 • VOLUME 282 • NUMBER 7 were harvested and subjected to Western blot analyses in which antibody ShcA was used to detect ShcA levels. ShcA immunoblotting revealed a clear decrease in the ShcA levels in the cells treated with ShcA siRNA as compared with control siRNA or saline (Fig. 5A). Quantification of the Western blots (normalized to ␤-actin) revealed that ShcA siRNA treatment decreased the ShcA levels by 80% (*, p Ͻ 0.05) (Fig. 5B) as compared with control siRNA treatment.

ShcC, Fe65, and APP Metabolism
We then assessed the effects of RNAi-mediated silencing of ShcA on APP processing in H4-FL-APP cells. 48 h after transfection with either ShcA siRNA or control siRNA, the cells were harvested and subjected to Western blot analyses as described for ShcC above. Immunoblotting for APP with antibody A8717 revealed no significant difference in the FL-APP levels in ShcA siRNAversus control siRNA-treated cells (Fig. 6A). Likewise, ShcA siRNA treatment did not alter levels of APP-C99 or APP-C83 as compared with control siRNA treatment (Fig. 6, A-C).
Next, we measured A␤ levels in the conditioned medium. 48 h following treatments with control siRNA or ShcA siRNA, ShcA siRNA treatment did not significantly alter A␤40 levels with 23 pg/ml/protein (ShcA siRNA) versus 25 pg/ml/protein (control siRNA) or A␤42 levels 5.9 pg/ml/protein (ShcA siRNA) versus 6.6 pg/ml/protein (control siRNA), as compared with control siRNA normalized to the total protein amount (Fig. 6D). Taken together, these findings indicate that, in contrast to ShcC siRNA, ShcA siRNA treatment did not affect APP processing or A␤ generation in H4-FL-APP cells.
RNAi Silencing of Fe65 Increased APP-CTF Levels and Decreased A␤ Levels in H4-FL-APP Cells-We next assessed the effects of RNAi silencing of Fe65, another APP adaptor molecule, on APP processing and A␤ levels in H4-FL-APP cells. We first established conditions under which Fe65 siRNA treatment reduces Fe65 levels in H4-FL-APP cells. The cells were harvested 48 h after being transfected with either control siRNA or Fe65 siRNA and were subjected to Western blot analyses with antibody PA-751 to measure levels of Fe65. Fe65 immunoblotting revealed decreased levels (63% reduction) of Fe65 following Fe65 siRNA treatment as compared with control siRNA treatment (*, p Ͻ 0.05) (Fig. 7).
We then assessed the effects of RNAi-mediated silencing of Fe65 on APP processing in H4-FL-APP cells by measuring levels of FL-APP, APP-C99, and APP-C83 following Fe65 siRNA or control siRNA treatments. 48 h after transfection of Fe65 siRNA or control siRNA, the cells were harvested and subjected to Western blot analyses with antibody A8717 to detect FL-APP, APP-C99, and APP-C83, as described above for ShcA and ShcC. Levels of APP-C99 and APP-C83 were increased in the cells treated with Fe65 siRNA versus control siRNA or saline (Fig. 8A). Meanwhile, no significant differences in the FL-APP levels were observed for Fe65  siRNAversus control siRNA-and saline-treated cells (Fig. 8A). Fe65 siRNA treatment led to a 252% increase in the ratio of APP-C83 to FL-APP (*, p Ͻ 0.05) (Fig. 8B) and a 262% increase in the ratio of APP-C99 to FL-APP (*, p Ͻ 0.05) (Fig. 8C) normalized to ␤-actin, as compared with control siRNA treatment.
Next, we measured A␤ levels in the conditioned medium normalized to the total protein amount. 48 h after treatment with control siRNA or Fe65 siRNA in H4-FL-APP cells, Fe65 siRNA decreased A␤40 levels (27 pg/ml/protein for Fe65 siRNA versus 50 pg/ml/protein for control siRNA) and A␤42 levels (7 pg/ml/protein for Fe65 siRNA versus 15 pg/ml/protein for control siRNA) (*, p Ͻ 0.05) (Fig.  8D). Collectively, these data indicate that RNAi silencing of Fe65 affects APP processing and A␤ production in a manner similar to that of ␥-secretase inhibitor treatment in H4-FL-APP cells.

DISCUSSION
AD neuropathogenesis is profoundly affected by the balance between A␤ generation and clearance in the brain (20,21). A␤ is produced via serial proteolysis of APP by two proteases, BACE and ␥-secretase (2)(3)(4)(5). Cleavage by BACE first generates APP-C99, which is further cleaved by ␥-secretase to release A␤ and the ␤-amyloid precursor protein intracellular domain (6 -8). APP is also cleaved by ␣-secretase to release a large ectodomain (␣-APPs) and APP-C83; APP-C83 is sequentially cleaved by ␥-secretase to produce p3 and amyloid precursor protein intracellular domain (9 -11). Several APP adaptor proteins (13) have previously been shown to affect APP processing and A␤ production following overexpression (22)(23)(24)(25)(26)(27)(28)(29). Recent studies (14,30) show RNAi-mediated silencing of X11␣ and X11␤, and autosomal recessive hypercholesterolemia can also affect APP processing and A␤ levels. However, the effects of RNAi knockdown of other APP adaptor proteins, including Fe65, ShcA, and ShcC, on APP processing and A␤ production have not been previously reported. Here we show for the first time that RNAi silencing of ShcC and Fe65 (but not ShcA) significantly affects APP processing and A␤ levels.
RNAi silencing of ShcC decreased levels of APP-C99, APP-C83, and secreted A␤ in the absence of alterations in FL-APP levels in H4-FL-APP cells. These changes could be explained by either decreases in the activities of BACE and/or ␣-secretase or to the increases in the proteolytic breakdown of APP-C99 and APP-C83. Moreover, RNAi silencing of ShcC did not affect APP processing and A␤ gen-   1 and 2). B, the Fe65 levels assessed by quantifying Fe65 in the Western blot. Fe65 siRNA treatment (black bar) significantly decreased the Fe65 levels as compared with control siRNA treatment (white bar) (*, p Ͻ 0.05) normalized to ␤-actin. eration in H4-APP-C99 cells, suggesting that the effects of RNAi for ShcC on APP processing and A␤ generation are dependent on BACE. Finally, RNAi silencing of ShcC decreased the levels of BACE. Collectively, these findings suggest that RNAi silencing of the APP adaptor molecule ShcC may reduce levels of APP-CTFs and A␤, at least partially, by decreasing BACE level and activity.
Alternatively, RNAi silencing of ShcC could simply decrease the amount of APP traveling through the secretory pathway to the plasma membrane wherein proteolysis by BACE and/or ␣-secretase occurs. Further investigation will be necessary to assess this possibility. In contrast, RNAi silencing of ShcA had no effect on the levels of APP-CTFs or secreted A␤. Taken together, these findings indicate a selective effect of ShcC versus ShcA on APP processing and A␤ generation.
We also tested the effects of knockdown of another APP adaptor molecule, Fe65, encoded by the APBB1 gene. RNAi silencing of this gene led to increased levels of APP-C99 and APP-C83 and decreased levels of secreted A␤. Overexpression of Fe65 has previously been shown to accelerate trafficking and maturation of APP, thereby enhancing secreted levels of APP and A␤ in H4-FL-APP cells and Madin-Darby canine kidney cells (13,18,22).
Recent studies (31) have shown decreases in A␤x-42 levels in brains of male mice lacking Fe65 as well as those lacking both Fe65 and Fe65L1 when compared with wild-type mice. These results are consistent with our current findings. Interestingly, no significant alterations in A␤x-40 levels were reported in the brains of mice lacking Fe65, Fe65L1, or both, as compared with wild-type mice (31). These findings suggest that there could be differences in the effects on specific A␤ isoforms when measured in the medium of cultured cells versus mouse brain. However, in contrast, overexpression of Fe65 has also been reported to stabilize immature APP and inhibit secretion APPs and A␤ in human embryonic kidney 293 cells (32). Future studies will be needed to further assess the effects of Fe65 on APP processing and A␤ generation.
One possible explanation for our current results is that Fe65 interaction with APP potentiates A␤ generation, perhaps by facilitating ␥-secretase cleavage of APP. On the other hand, p65Fe65, an endoproteolytic cleavage product of Fe65, can suppress the ␣-secretase cleavage of APP (33). Thus, another possible explanation for our results is that knockdown of Fe65 could increase APP-C83 levels by preventing Fe65-mediated regulation of ␣-secretase cleavage of APP. However, because RNAi for Fe65 also increased APP-C99 levels while concurrently decreasing secreted A␤ levels, it would appear equally or more likely that knockdown of Fe65 inhibits ␥-secretase cleavage of APP to reduce A␤ generation. In this latter scenario, RNAi silencing of Fe65 could either inhibit ␥-secretase cleavage (of APP), regulate APP trafficking to cellular compartments where ␥-secretase can cleave APP, or be required for interaction of APP (APP-C99) with ␥-secretase. Further studies will be needed to assess these possible explanations for the observed results.
Collectively, our findings demonstrate for the first time that RNAi-mediated knockdown of ShcC (but not ShcA) and Fe65 decreases secretion of A␤, although by different molecular mechanisms regarding APP processing. These data, together with those of previous studies, imply that pharmacologically blocking interaction of APP with either Fe65 or ShcC and perhaps other APP adaptor proteins may provide a novel means for treating and/or preventing AD by lowering A␤ generation. In H4-FL-APP cells, Fe65 siRNA treatment increased the levels of APP-C83 and APP-C99 and decreased A␤ levels. A, APP processing in Western blot analyses. FL-APP immunoblotting revealed that there was no significant difference in the FL-APP levels in the cells treated with saline (column 1), control siRNA (columns 2 and 3) or Fe65 siRNA (columns 4 and 5). APP-CTF immunoblotting showed increases in the levels of APP-C99 and APP-C83 in the cells treated with Fe65 siRNA (columns 4 and 5) as compared with control siRNA (columns 2 and 3) or saline (column 1). There was no significant difference in the amounts of ␤-actin in the saline-, control siRNA-or Fe65 siRNAtreated H4-FL-APP cells. B, APP processing assessed by quantifying the ratio of APP-C83 to FL-APP in the Western blot. Fe65 siRNA treatment (black bar) significantly increased the ratio of APP-C83 to FL-APP as compared with control siRNA treatment (white bar) (*, p Ͻ 0.05) normalized to ␤-actin. C, APP processing assessed by quantifying the ratio of APP-C99 to FL-APP in the Western blot. Fe65 siRNA treatment (black bar) significantly increased the ratio of APP-C99 to FL-APP as compared with control siRNA treatment (white bar) (*, p Ͻ 0.05) normalized to ␤-actin. D, effects of RNAi silencing of Fe65 on A␤ levels in H4-FL-APP cells. Fe65 siRNA treatment (black bar) decreased both A␤40 and A␤42 levels as compared with control siRNA treatment (white bar) (*, p Ͻ 0.05) normalized to total protein amount.