By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Autosomal-dominant Alzheimer's disease (ADAD) is a genetic disorder caused by mutations in Amyloid Precursor Protein (APP) or Presenilin (PSEN) genes. Studying the mechanisms underlying these mutations can provide insight into the pathways that lead to AD pathology. The majority of biochemical studies on APP mutations to-date have focused on comparing mechanisms between mutations at different codons. It has been assumed that amino acid position is a major determinant of protein dysfunction and clinical phenotype. However, the differential effect of mutations at the same codon has not been sufficiently addressed. In the present study we compared the effects of the aggressive ADAD-associated APP I716F mutation with I716V and I716T on APP processing in human neuroglioma and CHO-K1 cells. All APP I716 mutations increased the ratio of Aβ42/40 and changed the product line preference of γ-secretase towards Aβ38 production. In addition, the APP I716F mutation impaired the ε-cleavage and the fourth cleavage of γ-secretase and led to abnormal APP β-CTF accumulation at the plasma membrane. Taken together, these data indicate that APP mutations at the same codon can induce diverse abnormalities in APP processing, some resembling PSEN1 mutations. These differential effects could explain the clinical differences observed among ADAD patients bearing different APP mutations at the same position.
The amyloid precursor protein (APP) I716F mutation is associated with autosomal dominant Alzheimer's disease with the youngest age-at-onset for the APP locus. Here, we describe that this mutation, when compared to two other familial Alzheimer's disease mutations at the same codon (I716V and I716T), interfered distinctly with γ-secretase cleavage. While all three mutations direct γ-secretase cleavage towards the 48→38 production line, the APP I716F mutation also impaired the ε-cleavage and the fourth cleavage of γ-secretase, resembling a PSEN1 mutation. These features may contribute to the aggressiveness of this mutation.
Amyloid plaques are one of the main neuropathological hallmarks of Alzheimer's disease (AD) and are composed principally of extracellular aggregates of β-amyloid (Aβ) peptide. The Aβ peptide is generated by sequential proteolytic processing of the amyloid precursor protein (APP) by β-secretase (mainly β-site APP-cleaving enzyme 1 or BACE1) to produce a C-terminal fragment (β-CTF), which is subsequently cleaved by γ-secretase thereby liberating the Aβ peptide (Lleo and Saura 2011). The catalytic component of the γ-secretase complex is provided by presenilins, which are encoded by two paralogous genes, PSEN1 and PSEN2, although PSEN1/γ- secretase-containing complexes are the major contributors to Aβ production in the CNS (De Strooper et al. 1998; Franberg et al. 2011; Borgegard et al. 2012). According to the model proposed by Qi-Takahara et al. (Sato et al. 2003; Qi-Takahara et al. 2005), the γ-secretase cleavage of the β-CTF is also sequential; firstly the ε-cleavage generates Aβ48 or Aβ49 and the intracellular fragment of APP (AICD). Aβ48 and Aβ49 are then cleaved by consecutive γ-secretase cleavages every three or four amino acids in two production lines: Aβ49-Aβ46-Aβ43-Aβ40 and Aβ48-Aβ45-Aβ42-Aβ38.
Autosomal-dominant AD (ADAD) is a genetic disorder that accounts for less than 1% of all AD cases (Bateman et al. 2011). ADAD has been associated with mutations in three different genes: the APP, PSEN1 or PSEN2 genes. PSEN1 mutations account for the majority of cases of ADAD, whereas PSEN2 and APP mutations are relatively rare. Mutations in PSEN1 have been described across the entire protein albeit that the majority are clustered within the transmembrane domains (Lleo et al. 2004). Recently, it has been shown that these mutations produce variable effects on kinetic activity but consistently increase the ratio of Aβ42/40 (Chavez-Gutierrez et al. 2012). Mutations in APP that cause ADAD are clustered around three regions. A group of mutations found near the β-secretase cleavage site (e.g. the double Swedish mutation KM670/671/NL) have been shown to direct APP processing towards the amyloidogenic pathway, resulting in increased β-secretase cleavage and total Aβ production (Haass et al. 1995). Mutations found in the central fragment of the Aβ peptide sequence (e.g. Flemish, Italian, Dutch, Arctic and Iowa) result in an Aβ peptide which either has a greater tendency to aggregate (Fraser et al. 1992; Nilsberth et al. 2001) or equally, a greater resistance to degradation (Morelli et al. 2003; Tsubuki et al. 2003). Finally, the postulated mechanism for mutations located near the γ-secretase cleavage site (e.g. London) is an increase in the ratio of Aβ42/40, generating a higher proportion of the Aβ42 residues (Goate et al. 1991). Within this region, we recently described the novel APP I716F mutation that was associated with the youngest age-at-onset for the APP locus reported to-date (Guerreiro et al. 2010). We and others (Lichtenthaler et al. 1999; Tan et al. 2008; Herl et al. 2009; Guardia-Laguarta et al. 2010; Page et al. 2010) showed that this mutation leads to a prominent increase in the ratio of Aβ42/40, and our data also supported the observation that mutant β-CTF is inefficiently processed by γ-secretase (Guardia-Laguarta et al. 2010).
Most studies have assessed the pathogenic mechanism underlying APP mutations comparing mutations in different codons of APP (Campion et al. 1999; Janssen et al. 2003; Zekanowski et al. 2003; Raux et al. 2005; Guerreiro et al. 2010). As a result, the amino acid position has been assumed to be a major factor to predict the pathogenicity. In contrast, the influence of the amino acid change at the same position has not been sufficiently addressed. In the present study we compared the effects of the ADAD APP I716F mutation on APP processing to the other reported mutations at the same codon: I716V and I716T (Table 1). We found that although all mutations decrease the ratio of Aβ40/38 by directing γ-secretase cleavage towards the Aβ42 production line, the I716F mutation also interfered with γ-secretase cleavage at various points along that production line, consequently preventing Aβ42 from being processed further. This effect, in combination with other alterations, such as reduced AICD and APP β-CTF accumulation, could explain the earlier age-at-onset associated with the I716F mutation.
Table 1. Clinical characteristics of the APP mutations at codon 716
According to the AD Mutation database (Cruts et al. 2012).
Chinese hamster ovary (CHO-K1; ATCC® CCL-61™) and human H4 neuroglioma cells (ATCC® HTB-148™) were cultured in Dulbecco's Modified Eagle Medium and OPTI-MEM Medium, respectively, supplemented with 10% fetal bovine serum (FBS), 2 mM glutamine and 1000 units/mL penicillin/streptomycin at 37°C with 5% CO2 in a tissue culture incubator. Cells were transfected using X-tremeGENE 9 DNA Transfection Reagent (Roche, Indianapolis, IN, USA) according to the manufacturer's instructions. Treatment with 2 μM N-[(3,5-difluorophenyl)acetyl]-L-alanyl-2-phenyl] e-glycine-1,1-dimethylethyl ester (DAPT), a potent inhibitor of γ-secretase, was used as a control in wild-type (wt) APP or APP-Dendra2/C99-Dendra2 transfected cells.
The constructs used in this study are shown in Fig. 1. For Aβ ELISA and cell-free AICD assays, we used a cDNA construct containing full-length APP695. For imaging studies, we used cDNA constructs containing full-length APP695 or the C-terminal fragment of APP (C99; Asp597-Asn695) fused to the green fluorescent protein Dendra2 at the C-terminus in a pTurbo construct (Clontech, Mountain View, CA, USA). Mutations (I716F, I716V, I716T) were introduced into APP695 and C99 vectors by site-directed mutagenesis (Stratagene, Cedar Creek, TX, USA) using the following primers: I716F (5′TTG TCA TAG CGA CAG TGT TCG TCA TCA CCT TGG TG 3′ and 5′ CAC CAA GGT GAT GAC GAA CAC TGT CGC TAT GAC AA 3′), I716V (5′TTG TCA TAG CGA CAG TGG TCG TCA TCA CCT TGG TG 3′ and 5′ CAC CAA GGT GAT GAC GAC CAC TGT CGC TAT GAC AA 3′), I716T (5′CGG TGT TGT CAT AGC GAC AGT GAC CGT CAT CAC CTT 3′ and 5′ AAG GTG ATG ACG GTC ACT GTC GCT ATG ACA ACA CCG 3′). All constructs were verified by sequencing.
Conditioned media from CHO-K1 cells transfected with wt or mutant APP695 were collected 24 h after transfection, and levels of Aβ were detected using sensitive commercially available ELISA kits: Aβ1-40 and Aβ1-42 kits were purchased from Wako, Osaka, Japan and Aβ1-38, Aβ1-43 and Aβ1-x (as a measure of total Aβ) kits were purchased from IBL, Hamburg, Germany. Aβ42/40, Aβ40/38, Aβ38/42 ratios were calculated for wt and mutant APP-transfected wells. Results were normalized to total Aβ levels and expressed as a percentage of wtAPP levels.
Membrane preparations and cell-free AICD generation assay
AICD was generated in vitro from membrane preparations of CHO-K1 cells transfected with wt or mutant APP, as previously described (Guardia-Laguarta et al. 2010). The samples were electrophoresed and detected by immunoblotting with a rabbit anti-APP C-terminal (Sigma-Aldrich, St. Louis, MO, USA) antibody (Games et al. 1995; Vom Berg et al. 2012). To confirm that the Dendra2 tag does not interfere with physiological APP processing, we also analyzed AICD from CHO-K1 cells transfected with wt or mutant APP-Dendra2. The resulting products were detected using a rabbit polyclonal anti-Dendra2 (Evrogen JSC, Moscow, Russia) antibody (Figure S1).
Densitometric quantification of AICD and CTF fragments
Densitometric quantification of the AICD and CTF fragments was performed using ImageJ v1.48b (Schneider et al. 2012). The intensity of each lane was plotted for three independent immunoblots (8-bit images). The area under the plot for the AICD versus CTF was calculated for each construct and expressed relative to wtAPP.
Imaging of Dendra2-tagged vectors
H4 neuroglioma cells were seeded on glass coverslips and transiently transfected with 1 μg full-length APP695- or C99-Dendra2-tagged constructs in serum-free media. FBS (10%) was added to the media 4 h after transfection. Cells were fixed using 4% paraformaldehyde 4, 24 or 48 h after transfection, and nuclei were stained by incubating fixed cells with 2 μg/mL Hoechst 33258 pentahydrate bis-benzimide (Invitrogen, Carlsbad, CA, USA) for 7 min at 37°C. Cells were washed three times with phosphate-buffered saline and coverslips were mounted on glass microscope slides. Confocal images were taken with a SP5 confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany) using a 488 nm wavelength (63x objective; 4x zoom). Laser intensities were kept low to prevent spontaneous photo-conversion of Dendra, as described (Magrane et al. 2012). Confocal images were taken in multiple z planes (1 micron apart) in order to capture the whole cell. The z plane at the midpoint of the cell was used for the confocal images presented here. Cell surface accumulation of APP or its derivatives was determined by counting the number of cells at 24 h transfected with wt or mutant APP-Dendra2 (n = 20 per condition), in which the plasma membrane was clearly fluorescent. Counts performed by three investigators (independently and blinded) were then used to calculate the mean %. Membrane fluorescence was only considered positive when all three investigators were in agreement.
Total internal reflection fluorescence (TIRF) microscopy
For TIRF microscopy, cells were imaged in the dark using an Olympus XCellence microscope (Olympus Europe, Hamburg, Germany) using a 100x objective at 2.6x zoom in the 488 nm wavelength at 25% laser power to avoid photo-conversion.
Statistical analysis was performed using Graph Pad Prism 5.0 software (Graphpad Software Inc., La Jolla, CA, USA). The mean AICD/CTF relative to wtAPP or mean Aβ levels ± standard error of the mean (SEM) were compared using one-way anova followed by Dunnett's multiple-comparison. Level of significance was set at p < 0.05.
The APP I716F mutation impairs γ-secretase ε-cleavage
According to the sequential processing model proposed by Qi-Takahara et al. (Sato et al. 2003; Qi-Takahara et al. 2005), the first step in γ-secretase processing is achieved by the endopeptidase activity (ε-cleavage) generating Aβ48 or Aβ49 peptides as well as the AICD. We first investigated the effects of I716 mutations on AICD generation, using a previously reported cell-free assay (Sastre et al. 2001; Guardia-Laguarta et al. 2010). As shown in Fig. 2a, we found that production of AICD was decreased in cells transfected with APP-I716F (Fig. 2a). As shown in Fig. 2b, this was confirmed by densitometric quantification of three independent immunoblots which showed that the AICD/CTF ratio for the I716F mutation was 40.3% that of wtAPP (anova with post hoc Dunnett's test I716F vs. wtAPP p = 0.01). To confirm our results, we developed a protocol to visualize the β-CTF and AICD by fluorescence microscopy. We used a β-CTF construct (C99) fused to the green fluorescent protein Dendra2 (Fig. 1) at its C-terminus. The Dendra2 tag did not impair β-CTF processing, as demonstrated by the detection of AICD-Dendra2 by western blot in CHO-K1 transfected cells (Figure S1), and the reduction in AICD-Dendra in cells transfected with the I716F mutation was confirmed (Figure S1). No reduction was observed in I716V or I716T-transfected cells. Cells were then transfected with wtC99-Dendra2, and fluorescence was imaged at several time points (4, 24 and 48 h), which revealed a typical secretory/vesicular expression pattern by 24 h (Fig. 2c and Figure S2). Moreover, substantial accumulation of fluorescence was detected in the nucleus. The nuclear signal was completely abolished by DAPT treatment indicating that this nuclear signal corresponded to the AICD-Dendra2 fragment (Fig. 2c and Figure S2). This was supported by the fact that transfection of pDendra2 alone showed nuclear staining that was not modified by treatment with DAPT (Figure S2). We then transfected cells with C99-Dendra2 containing the different I716 mutations, and observed that the I716F-C99 mutation had reduced nuclear fluorescence when compared to wtC99-Dendra2 or to I716V or I716T C99-Dendra2 transfected cells (Fig. 2c and Figure S2). Taken together, our biochemical and imaging data confirm that the ε-cleavage that generates AICD is impaired in cells expressing APP-I716F but not in those expressing APP-I716V or APP-I716T.
APP I716 mutations change the production line preference of γ-secretase
Following the ε-cleavage by γ-secretase, the generated Aβ product is further processed in one of two production lines (Aβ49→Aβ40 or Aβ48→Aβ38) involving consecutive cleavages ultimately producing either Aβ40 or Aβ38 (via Aβ42). To investigate the effect of the I716 mutations on this process, we measured Aβ40, Aβ42 and Aβ38 levels in the conditioned media of CHO-K1 cells transiently transfected with wtAPP or the I716F, I716V or I716T mutations (Constructs used in this study are described in Fig. 1). The amount of each Aβ product was normalized to total Aβ (Aβ1-x) in order to control for inter-experimental variation, and was expressed as a percentage of Aβ levels in wtAPP-transfected cells. There were no differences in total Aβ levels between wtAPP and mutant transfected cells although a trend towards decreased total Aβ levels was observed for the I716F mutation compared to wtAPP (25% decrease; p = 0.07). We found that all three mutations increased Aβ42 in the conditioned medium compared to wtAPP-expressing cells (Fig. 3a); APP-I716F: 2.45 ± 0.34-fold (p < 0.01), APP-I716V 1.91 ± 0.18-fold (p < 0.05) and APP-I716T 2.69 ± 0.30-fold (p < 0.001). Conversely, Aβ40 secretion was decreased in APP-I716F (0.12 ± 0.03-fold; p < 0.001) and in APP-I716T (0.22 ± 0.03-fold; p < 0.001) but not in APP-I716V transfected cells (1.14 ± 0.16-fold; p = 0.53). Aβ38 was significantly increased in the APP-I716T (3.44 ± 0.59-fold; p < 0.01) and the APP-I716V expressing cells (3.13 ± 0.43-fold; p < 0.01). As a result, APP-I716F produced a dramatic increase in the Aβ42/40 ratio (21.91 ± 1.85-fold; p < 0.001, Fig. 3b), which was also considerably augmented for the I716T mutation (12.35 ± 0.51-fold; p < 0.001). In contrast, I716V showed a non-significant increase in Aβ42/40 ratio (1.73 ± 0.20-fold; p = 0.9). These results are in general agreement with previous studies from our group and others (Lichtenthaler et al. 1999; Tan et al. 2008; Herl et al. 2009) that show that the I716F mutation markedly increases the Aβ42/40 ratio (range 18-100-fold), whereas the I716T and I716V (De Jonghe et al. 2001; Eckman et al. 2001) mutations produce a milder effect (2.7 and 1.3-2-fold, respectively). Next, we compared the levels of Aβ40 and Aβ38 as the end product of each line (Fig. 3c). We observed that all three mutations significantly decreased the Aβ40/Aβ38 ratio compared to wtAPP: I716F 0.17 ± 0.02-fold (p < 0.001), I716V 0.39 ± 0.12-fold (p < 0.001) and I716T 0.06 ± 0.02-fold (p < 0.001). Finally, we also measured Aβ43 using a commercially available ELISA kit but was found to be undetectable (data not shown). Taken together, in agreement with a previous study (Chavez-Gutierrez et al. 2012), these results show that all three mutations increase Aβ42/40 ratio through at least one common mechanism, which is the change in the product line preference of γ-secretase from the Aβ40 to the Aβ38 line.
The APP I716F mutation impairs the final γ-secretase cleavage step
Since many PSEN mutations impair the fourth γ-secretase cleavage (Chavez-Gutierrez et al. 2012), we next investigated the effect of I716 mutations on this cleavage (Aβ42→Aβ38). We measured the Aβ38/42 ratio in cells transfected with the I716 mutations and found that this ratio was decreased compared to wt only in cells expressing the I716F mutation (Fig. 3d, 0.28 ± 0.04-fold; p < 0.05). These data demonstrate less efficient cleavage of the I716F APP protein at both the ε- and fourth γ-secretase sites.
The APP I716F mutation leads to abnormal APP accumulation in the cell membrane
We next analyzed whether I716 APP mutations affected the distribution of APP by transfecting the full-length APP-Dendra2 construct (Fig. 1) into H4 neuroglioma cells. In contrast to C99-Dendra2, we did not observe nuclear fluorescence accumulation of APP-Dendra2-transfected cells even at 48 h (Figure S3). This is likely because of a reduced kinetics in AICD generation when there is a need of prior shedding by BACE1. The expression patterns of APP-Dendra2 containing I716V and I716T mutations were similar to those transfected with wtAPP-Dendra2 (Fig. 4a). However, cells transfected with the I716F-APP-Dendra2 showed progressive accumulation of fluorescence at the cell surface. This pattern occurred in approximately 70% of the cells imaged, a value comparable to the 65% of wtAPP-Dendra2 expressing cells treated with DAPT. In contrast, we observed cell surface fluorescence in only 15%, 0% and 5% of wtAPP-Dendra2, APP-I716V-Dendra2 and APP-I716T-Dendra2 expressing cells respectively. To confirm the membrane localization we used TIRF microscopy, which is especially suited to analyze the outer 100–200 nm of the cell surface (Groves et al. 2008). We found that the I716F-APP mutation led to prominent accumulation of fluorescence at the cell surface similar to DAPT-treated cells (Fig. 4b). Because of the similarities between the patterns of the I716F-APP mutation and DAPT treatment, in which γ-secretase inhibition causes β-CTF accumulation, we postulate that the majority of fluorescence is accounted for by β-CTF (the direct γ-secretase substrate) accumulation at the cell membrane (a major site of γ-secretase processing). This finding is in agreement with our previous observation that I716F induces β-CTF accumulation in brain homogenates (Guardia-Laguarta et al. 2010).
Here, we report that ADAD-associated APP mutations at the 716 codon alter APP processing and Aβ generation by different mechanisms. While all I716 mutations change the product line preference of γ-secretase (Fig. 5) we found that the APP I716F mutation was unique in that it caused further disturbances in APP processing, such as impaired ε-cleavage and fourth γ-secretase cleavage, as well as abnormal β-CTF accumulation near the cell membrane.
The APP I716F mutation had been widely used in the past as an artificial mutation in cell culture experiments (Lichtenthaler et al. 1999; Tan et al. 2008; Herl et al. 2009; Page et al. 2010) because of its extreme effects on the Aβ42/40 ratio. Subsequently, we found this mutation in an ADAD family with the youngest mean age-at-onset (33 years) for the APP locus (Guerreiro et al. 2010); mean age-at-onset for APP mutation carriers is 51 years (Cruts et al. 2012). This finding reinforces the pathogenic role of the Aβ42/40 ratio in ADAD and the usefulness of this measure as an indicator of the aggressiveness of the mutation. Incidentally, we observed an inverse correlation between the Aβ42/40 ratio and age-at-onset of the disease in carriers of the APP716 mutations assessed in this study (data not shown). Subsequent functional studies by our group found that, in addition to the considerable increase in the Aβ42/40 ratio, the APP I716F mutation was associated with reduced ε-cleavage and APP β-CTF accumulation in transfected cells as well as in the brain homogenate of the only available mutation carrier (Guardia-Laguarta et al. 2010; Pera et al. 2012). We also observed a trend towards a decrease in total Aβ observed for the I716F mutation. Taken together these data support our hypothesis that the I716F mutation impairs ε-cleavage.
The current study confirms and extends our previous observations. We found that all APP716 mutations change the product line preference of γ-secretase towards the Aβ38 line, which is in general agreement with previous studies (Chavez-Gutierrez et al. 2012). We also report here that the APP I716F mutation also impairs the fourth γ-secretase cleavage (Aβ42→38). Interestingly, this is similar to the effect observed for all PSEN mutations (Chavez-Gutierrez et al. 2012). Therefore, it is possible that the basic common mechanism by which APP mutations cause AD is by changing the product line preference of γ-secretase towards the Aβ38 line, and that further disturbances in APP processing such as impaired ε- or fourth γ-secretase cleavage may act as aggravating factors associated with a younger age-at-onset.
One discrepancy with previous studies (Chavez-Gutierrez et al. 2012) is the absence of impairment of ε-cleavage by the I716T mutation as assessed by both a cell-free assay and confocal microscopy in our study. Neither did we observe an increase in the Aβ38/42 ratio in cells transfected with the I716T mutation (Fig. 3d) as previously reported (Chavez-Gutierrez et al. 2012). Overall, in our study the I716T mutation behaved qualitatively similar to the I716V mutation in all tested assays. These discrepancies may be because of the use of different cell types, affinity of antibodies and assays among studies.
Another relevant finding of our study is the abnormal accumulation of fluorescence observed at the cell membrane in cells transfected with Dendra2-tagged APP-I716F. Under normal conditions only about 10% of APP molecules reach the plasma membrane (Haass et al. 2012). APP is then rapidly internalized, delivered to endosomes and then either recycled to the cell surface or targeted for degradation in lysosomes (Haass et al. 1992). Here, we found that wtAPP or its derivatives can be detected in small amounts at the plasma membrane by TIRF microscopy. However, the amount of fluorescence at the cell surface was markedly increased in cells expressing the I716F mutation or cells treated with DAPT. The similarities between the effect in cells transfected with the I716F mutation and cells transfected with wtAPP treated with DAPT suggests that the fragment accumulated at the plasma membrane corresponds to β-CTF. This is consistent with the observation that γ-secretase cleaves APP on the cell surface and endosomes/lysosomes (Haass et al. 2012) and our previous data on the I716F mutation in cell culture and in human brain samples (Guardia-Laguarta et al. 2010; Pera et al. 2012). There is growing evidence indicating that accumulation of APP β-CTFs may be neurotoxic by itself and is able to impair synaptic plasticity and long-term memory in mouse animal models of AD (Yankner et al. 1989; Oster-Granite et al. 1996; McPhie et al. 1997; Kim et al. 2003; Saura et al. 2005; Jiang et al. 2010; Lauritzen et al. 2012; Mitani et al. 2012), therefore, it likely represents an active component of ADAD (Pera et al. 2012).
To our knowledge the only other APP mutation reported to have such a young age-at-onset is the Austrian APP T714I mutation, which has an age-at-onset in the mid thirties in three different families (Kumar-Singh et al. 2000; Edwards-Lee et al. 2005; Raux et al. 2005). Biochemical studies in human embryonic kidney (HEK) 293T cells (Kumar-Singh et al. 2000) showed similarities with the findings of the APP I716F mutation studied herein. In particular, the Austrian APP T714I mutation led to a ~ 11-fold increase in the ratio Aβ42/40 and 68% decrease in Aβ40 in the conditioned medium of APP T714I-transfected HEK cells (Kumar-Singh et al. 2000). Chávez-Gutiérrez et al. found that the APP T714I mutation also impaired AICD generation and shifted the product line preference of γ-secretase towards the Aβ38 line. Unlike the APP I716F mutation, however, the APP T714I mutation did not affect the fourth γ-secretase cleavage (Chavez-Gutierrez et al. 2012). The neuropathological study of individuals with the Austrian APP T714I mutation showed massive deposition of non-fibrillar diffuse plaques that were mainly immunoreactive for Aβ42-specific monoclonal antibodies, similar to the findings of the APP I716F mutation (Guardia-Laguarta et al. 2010). Taken together, this supports the role of Aβ40 in the formation of cored plaques, and that mutations that drastically reduce Aβ40 and elevate the ratio Aβ42/40, would lead to predominant Aβ42-immunoreactive diffuse plaques (Iwatsubo et al. 1994). Further research comparing these aggressive APP and PSEN1 mutations are still required to fully understand the mechanism by which they cause changes in Aβ profiles and such early age-at-onset.
In summary, our study demonstrates that different mutations at the same codon of APP may lead to changes in APP processing by various mechanisms. Although the common mechanism seems to be a change in the product line preference of γ-secretase towards the Aβ38 line, other abnormalities that resemble those of PSEN1 mutations may act as active components in the physiopathological disease process. In particular, the reduced ε-cleavage and fourth γ-secretase cleavage observed for the APP I716F mutation may contribute to the young age-at-onset of this mutation.
This work was supported by the Instituto de Salud Carlos III (FIS/PI100018 to A.Lleó) and CIBERNED. No authors declare a conflict of interest in the publication of this manuscript. Author Contributions: Conception and design: MS-C, OB, JM, JC, AL; Acquisition, analysis or interpretation of data; MS-C, OB, MP, NB, CG-L, LM, MC, AL; Drafting/Revision of Article: MS-C, OB, MP, NB, JM, CG-L, MC, JC, AL, Final approval of the version to be published; MS-C, OB, MP, NB, JM, CG-L, LM, MC, JC, AL.