The A673T mutation in the amyloid precursor protein reduces the production of β-amyloid protein from its β-carboxyl terminal fragment in cells

Abstract

Introduction: The A673T mutation in the amyloid precursor protein (APP) protects against Alzheimer's disease by reducing β-amyloid protein (Aβ) production. This mutation reduced the release of the soluble APP fragment (sAPPβ), which is processed by β-secretase, suggesting a concomitant decrease in the β-carboxyl fragment of APP (C99), which is a direct substrate of γ-secretase for Aβ production. However, it remains controversial whether the level of C99 is significantly reduced in cells expressing APP that carry A673T as the cause of reduced Aβ production. Here, we investigated the effect of the A673T mutation in C99 on γ-cleavage in cells.

Results: We found that the level of C99 in cells expressing APP A673T was indistinctive of that observed in cells expressing wild-type APP, although the release of sAPPβ was significantly reduced in the APP A673T cells. In addition, our reconstituted β-secretase assay demonstrated no significant difference in β-cleavage on an APP fragment carrying the A673T mutation compared with the wild-type fragment. Importantly, cells expressing C99 containing the A673T mutation (C99 A2T; in accordance with the Aβ numbering) produced roughly half the level of Aβ compared with the wild-type C99, suggesting that the C99 A2T is an insufficient substrate of γ-secretase in cells. A cell-free γ-secretase assay revealed that Aβ production from the microsomal fraction of cells expressing C99 A2T was diminished. A sucrose gradient centrifugation analysis indicated that the levels of the C99 A2T that was codistributed with γ-secretase components in the raft fractions were reduced significantly.

Conclusions: Our data indicate that the A673T mutation in APP alters the release of sAPPβ, but not the C99 level, and that the C99 A2T is an inefficient substrate for γ-secretase in cell-based assay.

Fig. 1

The A673T substitution in APP decreased sAPPβ secretion, but not C99. Constructs used in this study. The APP695- or C99-coding fragment was inserted prior to IRES-GFP in pMXIG (a). APP A673T cells released one third of the levels of sAPPβ compared with those of wild-type APP. However, C99 levels in APP A673T cells were statistically indistinguishable from those of APP WT cells (b). Levels of sAPPβ and C99 were normalized for amount of APP (c). C99 level in APP A673T cells was almost identical to that in wild-type APP cells, as well as from internal standard GFP expression levels (d). Data represent means ± SD of three independent experiments. NS, not significant; * P < 0.05 (unpaired t-test)

Fig. 2

Pulse chase analysis of APP processing. Schematic diagram of pulse chase analysis of APP processing (a). CHO cells were transfected with APP WT or APP A673T construct and treated with 1 μg/ml tetracycline for 4 h. After replacing media, cells and media were collected at the time indicated. Detection of APP, sAPPβ, C99, Aβ, C83 and sAPPα by Western Blotting (b). C99 was barely detected by using anti-FLAG antibody, although robust C83 bands were visualized. Quantitative analyses of APP, sAPPβ, C99, Aβ, C83 and sAPPα (c). Levels of sAPPβ and Aβ were reduced in APP A673T cells compared with those in APP WT cells, while levels of APP and C99 were almost similar to those of APP WT cells. It is interesting to note that generation rate of C83 is distinct from those of APP and C99. Data represent means ± SD of three independent experiments. Open circle, WT; closed circle, A673T. *P < 0.05; **P < 0.01 (unpaired t-test)

Fig. 3

β-Secretase-dependent cleavage of the APP A673T fragment. An APP751 fragment (633–685) fused with FLAG tag (APP633–686-FLAG) was incubated with β-secretase. The E50 antibody was used to visualize the equal amounts of Aβ33-FLAG that were produced by β-secretase from the WT of the APP633–686-FLAG substrate and its A673T mutant substrate. The reaction solution of the WT substrate was loaded onto a gel, as indicated by standard curve. Data are expressed as means ± SD of three independent experiments. An unpaired t-test detected no significant differences between substrates

Fig. 4

Conditioned media of C99 WT and C99 A2T cells were subjected to Western blotting to visualize and quantify the Aβ and p3 species. C99 A2T cells exhibited reduced levels of Aβ40 and Aβ42, but not of p3 40 (a). Lysates of transfectants were subjected to Western blotting, to visualize and quantify GFP, C99, and intracellular Aβ (b). The lysate of C99 WT cells was loaded into a gel, as indicated by standard curve. The levels of GFP and C99 in cells expressing C99 WT and C99 A2T were indistinctive, whereas intracellular Aβ levels were significantly reduced in C99 A2T cells. Data represent means ± SD of three independent experiments (c). NS, not significant. * P < 0.05 (unpaired t-test)

Fig. 5

CHAPSO-solubilized γ-secretase assay. Recombinant C99-FLAG substrates were incubated with solubilized γ-secretase and subjected to Western Blotting to visualize and quantify levels of Aβ production (a). Aβ production from C99 A2T was indistinctive to that from C99 WT. Data represent means ± SD of three independent experiments. NS, not significant (unpaired t-test). Substrates were immobilized on anti-FLAG magnetic beads and incubated with CHAPSO-solubilized γ-secretase (b). After sufficient wash, the beads were subjected to Western Blotting. γ-Secretase interacted with C99 A2T as well as C99 WT. -, DMSO vehicle control; D, 1 μM DAPT. Quantitative analyses of interacted γ-secretase components (c). Data represent means ± SD of three independent experiments in the absence of 1 μM DAPT. NS, not significant (ANOVA, Scheffe’s post hoc test compared with WT DMSO)

Fig. 6

The microsomal fraction of C99 A2T was incubated and subjected to Western Blotting analysis (a). Aβ production from the microsomal fraction of C99 A2T cells was significantly reduced compared with that detected in C99 WT cells. Aβ production was determined by subtracting the amount of Aβ at 0 min from that at 20 min. Data represent means ± SD of four independent experiments (b). NS, not significant. **P < 0.005 (unpaired t-test)

Fig. 7

C99 A2T cells were treated with CHAPSO and fractionated using sucrose gradient centrifugation (a). The presenilin 1 carboxyl terminal fragment (PS1CTF) was enriched in raft fractions (#4 and #5). C99 was also distributed in these fractions, in part. However, C99 A2T distribution was significantly shifted into denser fractions (#5 and #6) compared with C99 WT (#4, #5, and #6) (b). Data represent means ± SD of three independent experiments. NS, not significant. *P < 0.05 (unpaired t-test)