Abstract
A previous epidemiological study in Northern Europe showed that the A673T mutation (Icelandic mutation) in the amyloid precursor protein gene (APP) can protect against Alzheimer's disease (AD). While the effect of the A673T mutation on APP processing has been investigated primarily in vitro, its in vivo impact has not been evaluated. This is mainly because most existing AD mouse models carry the Swedish mutation. The Swedish and Icelandic mutations are both located near the β-cleavage site, and each mutation is presumed to have the opposite effect on β-cleavage. Therefore, in the AD mouse models with the Swedish mutation, its effects could compete with the effects of the Icelandic mutation. Here, we introduced the A673T mutation into App knock-in mice devoid of the Swedish mutation (AppG-F mice) to avoid potential deleterious effects of the Swedish mutation and generated AppG-F-A673T mice. APP-A673T significantly downregulated β-cleavage and attenuated the production of Aβ and amyloid pathology in the brains of these animals. The Icelandic mutation also reduced neuroinflammation and neuritic alterations. Both sexes were studied. This is the first successful demonstration of the protective effect of the Icelandic mutation on amyloid pathology in vivo. Our findings indicate that specific inhibition of the APP-BACE1 interaction could be a promising therapeutic approach. Alternatively, the introduction of the disease-protective mutation such as APP-A673T using in vivo genome editing technology could be a novel treatment for individuals at high risk for AD, such as familial AD gene mutation carriers and APOE ε4 carriers.
- Alzheimer’s disease
- amyloid pathology
- amyloid precursor protein
- App knock-in mice
- Icelandic mutation
- protective mutation
Significance Statement
The A673T mutation (Icelandic mutation) in the amyloid precursor protein (APP) gene can protect against AD. The effect of the A673T mutation on amyloid pathology has not been evaluated in vivo. Utilizing a new AD mouse model that we have recently developed, we show that the APP-A673T attenuates amyloid pathology in vivo. We demonstrate that its protective effects are exerted by inhibiting β-cleavage and reducing the production of Aβ in the brain. Furthermore, we reveal that the Icelandic mutation also reduced neuroinflammation and neuritic alterations. Our findings indicate that specific inhibition of the APP-BACE1 interaction or introduction of protective variants via in vivo genome editing could be a promising therapeutic approach.
Introduction
Alzheimer’s disease (AD)—the most common cause of dementia—has been extensively investigated for >100 years since it was first reported (Alzheimer et al., 1995). Amyloid pathology is reported to precede the clinical onset of dementia by >20 years and is considered the trigger of AD (amyloid cascade hypothesis; Selkoe and Hardy, 2016). Amyloid precursor protein (APP) is first cleaved either by α-secretase or β-secretase (BACE1), leading to nonamyloidogenic and amyloidogenic pathways, respectively (Muller et al., 2017). In the amyloidogenic pathway, the C-terminal fragment after β-cleavage (CTF-β) is further processed by γ-secretase to produce APP intracellular domain (AICD) and β-amyloid peptide (Aβ) species of various lengths. Longer forms of Aβ such as Aβ42 are considered to be more aggregation-prone and toxic. More than 30 mutations in the gene encoding APP (APP) were reported to cause familial forms of AD (FAD; Hunter and Brayne, 2018). One of the most well-studied mutations associated with FAD is the Swedish mutation (Haass et al., 1995). This is a double mutation (KM670/671NL) located close to the β-cleavage site of APP protein, increasing the susceptibility of APP protein to β-cleavage, and thus favoring Aβ production (Citron et al., 1995; Haass et al., 1995). On the other hand, the A673T mutation of the APP gene (the Icelandic mutation), which also is located close to the β-cleavage site of the APP protein (Fig. 1A), was reported to be protective not only against AD but also against age-related cognitive decline in a Scandinavian population (Jonsson et al., 2012). The past report showed decreased Aβ levels of plasma in AD APP-A673T mutation carriers (Martiskainen et al., 2017).
Generation of AppG-F-A673T mice. A, Sequences of wild-type APP and APP with the Swedish or Icelandic mutations around Aβ. The red characters represent the Swedish (KM670/671NL) and Icelandic (A673T) mutations, respectively. B, Exact sequences showing sgRNA (orange line) with the PAM site (blue box) in the mouse App gene. The red characters represent the Icelandic (A673T), Arctic (E693G), and Beyreuther/Iberian (I716F) mutations, respectively. The green characters represent humanized sequences in mouse Aβ (G676R, F681Y, and R684H). C, Schematic illustration of CRISPR-Cas9–mediated genome editing in AppG-F-KI mouse zygotes by microinjection. D, Sanger sequencing results determined the AppG-F/G-F (top panel) and AppG-F-A673T/G-F-A673T (bottom panel) genotype, respectively. The boxes indicate the codons encode alanine (GCA, black) and threonine (ACA, magenta), respectively. E, Regional information of potential off-target sites that were identified using Cas-OFFinder (www.rgenome.net/cas-offinder/) and COSMID (https://crispr.bme.gatech.edu/).
The molecular mechanism underlying the protective effect of the APP-A673T mutation against AD has been primarily investigated in cellular models (Jonsson et al., 2012; Maloney et al., 2014; Kokawa et al., 2015; Wittrahm et al., 2023). In vitro analyses revealed that several mechanisms could be involved in the attenuation of amyloid pathology by APP-A673T: suppression of β-cleavage, reduction of aggregation properties of Aβ, and/or inhibition of the interaction between Aβ and γ-secretase (Xia et al., 2021). A rat model harboring the APP-A673T mutation in the endogenous rat App gene showed reduced β-cleavage and increased α-processing of APP (Tambini et al., 2020). However, the in vivo protective effect of the Icelandic mutation against amyloid pathology is not known because it was difficult to precisely assess its effect in the AD animal model. This is mainly because most of the AD mouse models harbor the Swedish mutation (Sasaguri et al., 2017) and possibly compete with the protective effect of the Icelandic mutation (Guyon et al., 2020; Fig. 1A).
Recently, we developed AppG-F knock-in (KI) mice harboring the Arctic (E693G) and Beyreuther/Iberian (I716F) mutations but devoid of the Swedish mutation (Watamura et al., 2022). The AppG-F mice showed an age-dependent amyloid pathology, neuroinflammation, and synaptic alteration similar to that of other App-KI mice that harbor the Swedish mutation (AppNL-F and AppNL-G-F lines; Saito et al., 2014). AppG-F-KI mice are thus considered a suitable model to evaluate the effects of the Icelandic mutation. Here, we generated AppG-F-A673T mice by knocking in the A673T mutation into AppG-F mice. We found that Aβ accumulation was attenuated in AppG-F-A673T mice compared with that seen in AppG-F mice. By evaluating APP processing, we demonstrate that a protective effect of APP-A673T was exerted by suppressing β-cleavage. This is the first successful demonstration of the beneficial effect of the protective APP mutation on amyloid pathology in vivo. Although AD treatment strategies utilizing BACE1 inhibitors have failed due to side effects from other substrates (Egan et al., 2018), our findings indicate that a more specific inhibition of the APP-BACE1 interaction could still be a promising therapeutic approach. Alternatively, introducing APP-A673T using in vivo genome editing technology could be a novel prophylactic treatment for individuals at high risk for AD, such as familial AD gene mutation carriers and APOE ε4 carriers.
Materials and Methods
Mice
All mice were bred and maintained in accordance with the regulations for animal experiments promulgated by the RIKEN Center for Brain Science. AppG-F mice expressing two familial AD mutations [the Arctic (E693G) and Beyreuther/Iberian (I716F) mutations] driven by the endogenous promoter, as well as the humanized Aβ sequence, were generated as described previously (Watamura et al., 2022). C57BL/6J mice were prepared as controls. APP-KO (B6.129S7-Apptm1Dbo/J) mice were bought from Jackson Laboratory. All mutant mice used in this study were homozygous for the expressed mutations. Both male and female mice were used in our experiments.
Generation of AppG-F-A673T mice
single-guide RNA (sgRNA)-targeting mouse App exon 16 was designed near the Icelandic mutation in silico using the CRISPR design tool (CRISPR direct, https://crispr.dbcls.jp/) according to the properties of SaCas9 (Ran et al., 2015). To reduce the possibility of off-target events, SaCas9 that recognizes NNGRRT as the protospacer adjacent motif (PAM) site was selected to introduce double-stranded breaks. Single-stranded oligodeoxynucleotide (ssODN) was designed to introduce the Icelandic mutation (A673T) overlapping the sgRNA-targeting region so that the oligonucleotide did not include silent mutations, thus preventing rebinding and recutting after the desired genome modification via homology-directed repair. A plasmid vector (Addgene, no. 61591) was used for in vitro transcription of SaCas9 mRNA, and sgRNA was synthesized as described previously (Yang et al., 2014). In vitro synthesis of CRISPR tools was performed with the primers as follows: sgRNA-App-A673T-forward, TGTAATACGACTCACTATAGGCTCGGAAGTGAAGATGGATGGTTTTAGTACTCTGGAAACAGAATC; sgRNA-App-A673T-reverse, AAAAATCTCGCCAACAAGTTGACGAGATAAACACGGCATTTTGCCTTGTTTTAGTAGATTCTGTTTCCAGAGTACTAAAAC; SaCas9 mRNA-forward, TAATACGACTCACTATAGGGCCCCATTGACGCAAAT; and SaCas9 mRNA-reverse, GGCAACTAGAAGGCACAGTCGA. The prepared SaCas9 mRNA (100 ng/μl) and sgRNA (100 ng/μl) along with ssODN (100 ng/μl) were coinjected into the cytoplasm of AppG-F/wt zygotes. Founder mice were identified by polymerase chain reaction (PCR) and sequencing analysis of the targeted site and crossed with WT mice to obtain heterozygous F1 mice.
Off-target effect analysis
Candidate sequences were identified in silico using COSMID (https://crispr.bme.gatech.edu/; Cradick et al., 2014) and Cas-OFFinder (www.rgenome.net/cas-offinder/; Bae et al., 2014), allowing up to 3 bp mismatches and 1 bp DNA and/or RNA bulge. Genomic DNA extracted from mouse tails was amplified by PCR with the primers as follows: OFF1-F, CCCTATAGTTTAACTTGGATCCTCTTACCC; OFF1-R, TGTTGACAGTCACGGCTCTGC; OFF2-F, GGTGGATGGGTAATGTACTGTGC; OFF2-R, CCTCAAACCCACCTTCAGTAACAC; OFF3-F, GGGCCTCTACAATTCAAGTTGGCT; OFF3-R, CCCGACTCTCAATGGTTCGAC; OFF4-F, CCTTGAGAACGGTTGGCGTC; OFF4-R, CAGGCGCGAGCATCATCACT; OFF5-F, CATGTGTGGGCAGCTTTCAGTAG; OFF5-R, CTCCCAACCCAGCAGTTTATAGTC; OFF6-F, GTTGGCAGAGACTGAAGAGGC; OFF6-R, CCAGAAACTTACCCACACAGGC; OFF7-F, TGGAGCTGCAAAATACTGCACGG; and OFF7-R, CCAGAAGAGGTTATGTGTGGTGGT. All genomic sequences of the amplicons were analyzed by Sanger sequencing using a DNA sequencer (ABI 3730xl).
Genotyping
Genomic DNA was extracted from mouse tails in lysis buffer [10 mM Tris-HCl (pH 8.5), 5 mM EDTA (pH 8.0), 0.2% SDS, 200 mM NaCl, and proteinase K (20 μg/ml)] through a process of ethanol precipitation. To distinguish App-KI lines, including AppG-F and AppG-F-A673T mice from WT mice, purified DNA was subjected to PCR and followed by Sanger sequencing analysis. Genotyping primers for App-KI mice are as follows: moAPP-Ex16-hs-F2, ACAGGCATTACATATTCAGCGT and moAPP-Ex16-hs-R2, ACTATCAACAGAGCCCCACT.
Western blotting
Mouse brain tissues were homogenized in lysis buffer containing 50 mM Tris (pH 7.6), 0.15 M NaCl, 1% Triton X-100, and cOmplete protease inhibitor cocktail (Roche Diagnostics) using a Multi-Beads Shocker (Yasui Kikai). Homogenates were incubated at 4°C for 1 h and centrifuged at 15,000 rpm for 30 min, and the supernatants were collected as loading samples. Concentrations of protein samples were measured with the aid of a bicinchoninic acid protein assay kit (Thermo Fisher Scientific). Equal amounts of proteins were subjected to SDS–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes. For detection of APP-CTFs, delipidated samples were loaded onto membranes and boiled for 5 min in phosphate-buffered saline (PBS) before blocking with enhanced chemiluminescence (ECL) primer blocking buffer (GE HealthCare). Membranes were incubated at 4°C with primary antibodies as follows: anti-APP A4 antibody (22C11; Sigma-Aldrich, MAB348), anti-amyloid precursor protein antibody (Y188; Abcam, ab32136), anti-sAPPα antibody (2B3; IBL, 11088), anti-sAPPβ antibody (Poly8134; BioLegend, 813401), anti-insulin–degrading enzyme (IDE; Abcam, ab32216), anti-mouse neprilysin/CD10 (R&D Systems, AF1126), anti-BACE1 antibody (EPR19523; Abcam, ab183612), anti-ADAM10 antibody (EPR5622; Abcam, ab124695), glyceraldehyde-3-phosphate dehydrogenase (HRP-60004, Proteintech); and anti-β-actin (Sigma-Aldrich, A5441). Targeted proteins were visualized with ECL select (GE HealthCare) and a LAS-4000 Mini Lumino image analyzer (Fujifilm).
Immunohistochemistry
Paraffin-embedded mouse brains were sectioned (thickness, 4 μm) and subjected to deparaffinization processing; antigen retrieval was then performed by autoclaving at 121°C for 5 min. Brain sections were treated with 0.3% H2O2 in a methanol solution for 30 min to inactivate endogenous peroxidases. Sections were rinsed with buffer containing 0.1 M Tris (pH 7.5), 0.15 M NaCl, and 0.05% Tween 20 (TNT buffer), blocked using a TSA Biotin System kit (PerkinElmer), and incubated at 4°C overnight with primary antibodies diluted in buffer containing 0.1 M Tris (pH 7.5), 0.15 M NaCl, and 0.5% Blocking Reagent (PerkinElmer; TNB buffer). Primary antibodies used are as follows: anti-β-amyloid, 1–16 (6E10; BioLegend, 803001), anti-human amyloid β (1–42; IBL, #18582), anti-glial fibrillary acidic protein antibody (clone GA5; Sigma-Aldrich, MAB3402), anti-Iba1 (Fujifilm, 013-27691), anti-human amyloid β (N; 82E1; IBL, #10323), anti PHF-TAU (AT8; Fujirebio, 90206), anti-LAMP1 (Abcam, EPR21026), and anti-human Aβ (N1D; Saido et al., 1995). Sections were stained for 15 min with Hoechst 33342 (Thermo Fisher Scientific) diluted in PBS and then mounted with PermaFluor (Thermo Fisher Scientific). Section images were obtained using a confocal laser scanning microscope FV-1000 (Olympus) and a NanoZoomer Digital Pathology C9600 (Hamamatsu Photonics). Quantification of immunoreactive signals was performed using MetaMorph Imaging Software (Molecular Devices) and Definiens Tissue Studio (Definiens).
Enzyme-linked immunosorbent assay
Mouse brain samples were homogenized in lysis buffer [50 mM Tris-HCl (pH 7.6), 150 mM NaCl, and protease inhibitor cocktail] using a Multi-Beads Shocker (Yasui Kikai). The homogenates were centrifuged at 70,000 rpm at 4°C for 20 min, and the supernatant was collected as a Tris-soluble (TS) fraction to which 1/11 (v/v) of 6 M GuHCl in 50 mM Tris and protease inhibitors were added. The pellet was loosened in lysis buffer with a Pellet Pestle (KIMBLE), dissolved in 6 M GuHCl buffer, and sonicated at 25°C for 1 min. The sample was incubated for 1 h at room temperature and then subjected to centrifugation at 70,000 rpm at 25°C for 20 min. The supernatant was collected as a GuHCl-soluble fraction. TS and GuHCl-soluble fractions were applied to 96-well plates using an Aβ ELISA kit (Wako) according to the manufacturer’s instructions. Mouse blood samples were collected and centrifuged at 1,200 rpm at 4°C for 20 min. The supernatant was collected, diluted fourfold, and subjected to ELISA using an Aβ ELISA kit (Wako). For the detection of Arctic Aβ produced from the brains of AppG-F and AppG-F-A673T mice, standard curves were drawn using human Aβ peptides carrying the Arctic mutation.
Quantification and statistical analysis
All data are shown as the means ± SEM within each figure. For comparisons between the two groups, data were analyzed by Student’s t test. For comparisons among three groups, we used a one-way analysis of variance (ANOVA) followed by Tukey’s post hoc analysis. All statistical analyses were performed using GraphPad Prism 7 software (GraphPad Software). The levels of statistical significance were presented as p values: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.
Results
Generation of AppG-F-A673T mice by CRISPR-Cas9
To demonstrate the effects of the Icelandic mutation in vivo, we introduced the APP-A673T mutation into the endogenous mouse App gene of an AD mouse model devoid of the Swedish mutation (AppG-F mice; Watamura et al., 2022). The AppG-F line harbors the Arctic (E693G) and Iberian (I716F) mutations together with humanized Aβ sequences (G676R, F681Y, R684H) in the mouse endogenous App gene and exhibits moderate brain amyloid pathology compared with other App-KI lines (Saito et al., 2014) starting from ∼4 months of age (Watamura et al., 2022). To introduce the APP-A673T mutation, single-guide RNA (sgRNA)-App-A673T and single-stranded oligodeoxynucleotide (ssODN) containing the A673T mutation (GCA→ACA) together with Staphylococcus aureus Cas9 (SaCas9) mRNA, where the protospacer adjacent motif (PAM) sequence is required as NNGRRT, were injected into the cytoplasm of AppG-F mouse zygotes (Fig. 1B,C). Sanger sequencing analysis revealed that the desired substitution via homology-directed repair occurred successfully in the AppG-F allele of the founder mice with an efficiency of 11.1% (Fig. 1D). Using the online COSMID (Cradick et al., 2014) and Cas-OFFinder (Bae et al., 2014) tools, we identified seven candidate off-target sites for the sgRNA used to generate AppG-F-A673T mice (Fig. 1E). Targeted sequencing analysis focusing on the candidate genomic regions revealed that no off-target modifications took place in the AppG-F-A673T mice. In subsequent experiments, all the mutant mice were homozygous for the mutations.
Effects of the Icelandic mutation on amyloid pathology of App-KI mice
We next analyzed the effects of APP-A673T on amyloid pathology in the AppG-F mice. We first quantified the amount of Aβ40 and Aβ42 in the brains of 3-month-old AppG-F-A673T mice before the mice exhibited amyloid pathology in the brain (Fig. 2A and Extended Data Fig. 2-1). Aβ40 and Aβ42 levels in the Tris-HCl–soluble (TS) fraction and Aβ40 in the guanidine HCl (GuHCl)-soluble fraction were significantly reduced compared with those of age-matched AppG-F mice. Interestingly, the amount of Aβ42 in the GuHCl-soluble fraction was not altered in AppG-F-A673T mice, possibly due to the presence of the Arctic mutation in Aβ that alters its characteristics of Aβ and makes it more aggregate-prone (Tsubuki et al., 2003; Lord et al., 2006).
Amyloid pathology of AppG-F-A673T mice. A, Aβ content detected by enzyme-linked immunosorbent assay (ELISA) using the Tris-HCl–soluble fraction (TS) and GuHCl-soluble fraction (GuHCl) of the cortices of AppG-F and AppG-F-A673T mice at 3 months (3 males and 3 females, respectively). # indicates that the value was below the detection limit. Each bar represents the mean ± SEM. B, C, Immunohistochemistry images showing Aβ deposition as indicated by immunostaining with 6E10 antibody against Aβ at 8 months of AppG-F/G-F (B) and AppG-F-A673T/G-F-A673T (C) mice. Scale bar, 1 mm. D, E, Quantification of amyloid plaque load in the cortex (D) and hippocampus (E) of AppG-F and AppG-F-A673T mice at 8 months. Scale bar, 1 mm. *p < 0.05, **p < 0.01, and ****p < 0.0001. n = 6 (3 males and 3 females) for each genotype, Student's t test. Please see Extended Data Figures 2-1 and 2-2 for more details.
Figure 2-1
Aβ plaque formation in the brains of 3- and 8-month-old AppG-Fand AppG-F-A673T mice. (A and B) Immunohistochemical analysis of Aβ plaques detected by 6E10 of AppG-F (A) and AppG-F-A673T(B) mice at 3 months (upper panels) and 8 months (lower panels) revealed that there was no Aβ plaque formation at 3 months in both mouse lines. Scale bar, 1 mm. Download Figure 2-1, TIF file.
Figure 2-2
Accumulation of Aβx-42 in the brains of 8-month-old AppG-F and AppG-F-A673T mice. Immunohistochemistry images showed Aβx-42 deposition in the brain of AppG-F (A) and AppG-F-A673T(B) mice at 8 months. The degree of Aβx-42 deposition was significantly higher in AppG-F than AppG-F-A673T mice in the cortex (C) and hippocampus (D). Plasma levels of Aβ42 were measured with an Aβ ELISA kit (Wako) (E). Scale bar, 1 mm. ****P < 0.0001. n = 6 (3 males and 3 females) for each genotype, Student’s t-test. Download Figure 2-2, TIF file.
Next, we performed immunohistochemical analyses to evaluate amyloid pathology in the brains of 8-month-old AppG-F-A673T mice (Fig. 2B–E). The 6E10 antibody was used to detect amyloid plaques because this antibody has been previously reported to capture CTF-β and Aβ regardless of the presence of the Icelandic mutation (Kokawa et al., 2015). Compared to AppG-F mice, we found that amyloid pathology was significantly attenuated by the Icelandic mutation both in the cortex (Fig. 2D) and hippocampus (Fig. 2E). In addition, we analyzed the Aβ species constituting amyloid plaques in the AppG-F-A673T mice using C-terminal (Aβx-42)-specific antibodies and found that accumulation of Aβx-42 was also attenuated by the presence of the APP-A673T mutation (Extended Data Fig. 2-2A–D). Aβ42 levels in plasma tended to be lower in AppG-F-A673T mice than in AppG-F mice (Extended Data Fig. 2-2E). These results indicate that APP-A673T exerts protective effects against AD by reducing the production of Aβ, thereby reducing the subsequent amyloid pathology in the App-KI mice.
Effects of APP-A673T on APP processing in vivo
Next, we evaluated APP processing in the AppG-F-A673T mice to clarify the molecular mechanism underlying protection against AD pathology. Expression of APP protein was not altered by the presence of APP-A673T (Fig. 3A,H). APP protein is cleaved either by α-secretase or β-secretase (BACE1), leading to nonamyloidogenic and amyloidogenic pathways, respectively (Muller et al., 2017). In the amyloidogenic pathway, the CTF-β is processed by γ-secretase to produce AICD and Aβ. In the nonamyloidogenic pathway, α-secretase cleaves APP protein within the Aβ sequence, with the resultant CTF-α processed by γ-secretase to produce AICD and a smaller P3 fragment than Aβ (Muller et al., 2017). Amounts of CTF-β were reduced in AppG-F-A673T mice compared with AppG-F mice (Fig. 3B,C). CTF-α also showed a trend to be decreased in AppG-F-A673T mice, although this was not statistically significant (Fig. 3B,D). The CTF-β/CTF-α ratio was significantly decreased in AppG-F-A673T mice compared with AppG-F mice (Fig. 3E), indicating that β-cleavage was suppressed by the presence of APP-A673T. Amounts of secreted APPβ (sAPPβ) were also reduced in AppG-F-A673T mice compared with AppG-F mice (Extended Data Fig. 3-1A,C). sAPPα showed a trend to be increased in AppG-F-A673T mice (Extended Data Fig. 3-1A,B). The discrepancy between the amounts of CTF-α and sAPPα is inferred to be due to the low stability of CTF-α. AICD is another fragment when CTF-α and CTF-β undergo γ-cleavage. AICD levels were not significantly different between AppG-F-A673T and AppG-F mice (Fig. 3B,F).
APP processing in AppG-F-A673T mice. A, Full-length APP expression in the cortices of WT, AppG-F, and AppG-F-A673T mice at 3 months (3 females). B, Western blot analysis using antibody targeting the C-terminus of APP in the cortex of WT, AppG-F, and AppG-F-A673T mice at 3 months (3 females). mAPP, mature APP; imAPP, immature APP. C–F, Quantification of CTF-β (C), CTF-α (D), CTF-β/CTF-α ratio (E), and AICD (F) of the blot shown in (B). G, Expression of enzymes involved in APP processing. Expression levels of full-length APP protein (A) and enzymes that are involved in APP processing (G) in the brain of AppG-F and AppG-F-A673T mice at 3 months old are quantified in the Western blot. Values of each band were normalized by expression levels of β-actin for APP and GAPDH for enzymes (H–L). **p < 0.01 and ***p < 0.001. One-way ANOVA followed by Tukey's multiple-comparisons test. Please see Extended Data Figure 3-1 for more details.
Figure 3-1
Quantification of sAPPα and sAPPβ. Western blot analysis using antibody targeting sAPPα and sAPPβ in the cortex of AppG-F and AppG-F-A673Tmice at 3 months (3females) (A). Values of each band were normalized by expression levels of β-actin (B, C). *P < 0.05, Student’s t-test. Download Figure 3-1, TIF file.
We further quantified expression levels of enzymes that are involved in APP processing (Fig. 3G, I–L). Amounts of α-secretase, ADAM10, and β-secretase (BACE1) were not altered in AppG-F-A673T mice (Fig. 3G, top two rows, I,J). Furthermore, amounts of the major Aβ-degrading enzymes neprilysin (NEP; Iwata et al., 2001) and insulin-degrading enzyme (IDE; Farris et al., 2003) were not changed in the AppG-F-A673T mice (Fig. 3G, third and fourth rows, K,L). These results indicate that the Icelandic mutation decreases the susceptibility of APP to BACE1 cleavage and thus reduces the amount of Aβ produced in vivo.
Effects of APP-A673T on neuroinflammation and neuritic alterations
Next, we assessed the pathological changes associated with amyloid pathology. To achieve this, we used 12-month-old AppG-F and AppG-F-A673T mice because amyloid pathology-related changes are relatively mild in 8-month-old AppG-F-A673T mice. We first evaluated amyloid pathology in these mice at 12 months old. We found that amyloid pathology was significantly attenuated by the Icelandic mutation both in the cortex and hippocampus in AppG-F-A673T mice at 12 months compared with AppG-F mice (Extended Data Fig. 4-1A–D). Moreover, we quantified and compared plaque size between AppG-F and AppG-F-A673T mice. Only the number of plaques with a size of above 20 μm was significantly decreased in AppG-F-A673T mice (Extended Data Fig. 4-1E–H). Neuroinflammation caused by activated microglia and astrocytes is frequently observed in the brain of AD patients as well as AD mouse models in association with amyloid pathology. We investigated the status of glial cells in AppG-F-A673T mice and found that the numbers of reactive astrocytes and activated microglia were significantly reduced in AppG-F-A673T mice compared with AppG-F mice (Fig. 4A–D). The formation of dystrophic neurites (DNs) is another pathological feature of AD (Tsering and Prokop, 2023). Amyloid plaques are often associated with phosphorylated tau-positive DNs and are termed neuritic plaques (NPs). NPs are also observed in App-KI mice as phosphorylated tau-positive dots located in close proximity to amyloid plaques (Extended Data Fig. 5-1). We examined the formation of DNs and found that the number of phosphorylated tau or LAMP1-positive signals was also attenuated in AppG-F-A673T mice compared with AppG-F mice (Fig. 5A–F). These results indicate that the Icelandic mutation attenuates neuroinflammation and neuritic alterations in relation to reduced amyloid pathology.
Neuroinflammation in AppG-F-A673T mice. A, B, Inflammatory responses in the cortices of AppG-F (A) and AppG-F-A673T mice (B) at 12 months. Antibodies against Aβ (82E1, gray), ionized calcium-binding adapter molecule 1 (Iba1, magenta), and glial fibrillary acidic protein (GFAP, yellow-green) were used, respectively. Scale bar, 500 μm. C, D, Quantification of Iba1 (C) and GFAP (D) signals in the cortex of AppG-F and AppG-F-A673T mice. Those units of the y-axis indicate the area fraction of the cortex. *p < 0.05 and **p < 0.01. n = 10 (5males and 5 females) for AppG-F mice, n = 7 (4males and 3 females) for AppG-F-A673T mice, Student's t test. Please see Extended Data Figure 4-1 for more details.
Figure 4-1
Quantification of amyloid plaque and plaque size in the brain of App-KI mice. Immunohistochemistry images showed formation of amyloid plaques in the brain of AppG-F (A) and AppG-F-A673T (B) mice at 12 months. The degree of Aβ deposition was significantly higher in AppG-F than AppG-F-A673T mice in the cortex (C) and hippocampus (D). Plaque size was also compared, with a threshold of 20 μm. The plaque size is altered by the A673 T mutation (E-H). Amyloid plaques were detected by anti-amyloid antibody (6E10, green). Scale bar, 500 μm.*P < 0.05, **P < 0.01, and ****P < 0.0001. n = 10 (5males and 5 females) for AppG-Fmice, n = 7 (4males and 3 females) for AppG-F-A673Tmice, Student’s t-test. Download Figure 4-1, TIF file.
Neuritic alterations in AppG-F-A673T mice. Formation of dystrophic neurites (DNs) in the brain of 12-month-old AppG-F (A, D) and AppG-F-A673T (B, E) mice. DNs were detected by antibodies against phosphorylated tau (AT8; A, B) and LAMP1 (D, E); the magenta arrows indicate DNs. Note that most of the AT8 (A, B) and LAMP1 (D, E) signals (arrows) appeared in close proximity to amyloid plaques (green, detected by 6E10 antibody). Quantification of DNs in the cortex (C, F) of AppG-F and AppG-F-A673T mice at 12 months. Scale bar, 100 μm. ***p < 0.001. n = 10 (5males and 5 females) for AppG-F mice, n = 7 (4males and 3 females) for AppG-F-A673T mice, Student's t test. Please see Extended Data Figure 5-1 for more details.
Figure 5-1
Formation of dystrophic neurites (DNs) in the brain of App-KI mice. Immunohistochemistry images showed formation of DNs in the brain of AppG-F (A) and AppNL-G-F (B) mice at 12 months. DNs were detected by anti-phosphorylated tau antibody (AT8, magenta). Note that most of AT8 signals (arrows) appeared in close proximity to amyloid plaques (yellow-green, detected by N1D antibody). Scale bar, 100 μm. Download Figure 5-1, TIF file.
Discussion
Taking advantage of the properties of a new AD mouse line, AppG-F (Watamura et al., 2022), we believe we are the first group to successfully demonstrate the protective effects of the Icelandic mutation (APP-A673T) on amyloid pathology in vivo. The AppG-F mouse line is the only AD model that recapitulates amyloid pathology in the mouse brain but does not harbor the Swedish mutation, is not dependent on APP overexpression, and is suitable for evaluating BACE1 activity (Sasaguri et al., 2022; Watamura et al., 2022). We showed here that the Icelandic mutation attenuated amyloid pathology and related neuroinflammation and synaptic alterations by suppressing the β-cleavage of APP and thus reducing the production of Aβ, consistent with previous in vitro and in vivo studies (Jonsson et al., 2012; Maloney et al., 2014; Tambini et al., 2020).
To date, the protective effects and molecular mechanisms of APP-A673T have been verified mainly in cell culture systems (Xia et al., 2021). In addition to the Swedish mutation (KM670/671NL), the homozygous APP-A673V mutation causes an inherited form of AD by altering APP processing and increasing β-CTF and Aβ production (Di Fede et al., 2009). While a residue at the Ala-673 site of the APP protein is a critical determinant of the amyloidogenic processing of APP (Zhang et al., 2017), the APP-A673V and A673T mutations seem to exhibit their effects in an opposing manner (Tambini et al., 2020). In a rat model that harbors the A673T mutation in the endogenous App gene and which expresses the humanized Aβ sequence without any FAD-causing mutations, α-cleavage is favored over β-processing. On the other hand, a rat model with the Swedish mutation showed an opposite APP processing shift (Tambini et al., 2020). While APP-A673T reduces the catalytic turnover of APP by BACE1 (Zhang et al., 2017), it has been shown that the binding affinity of APP for BACE1 is not affected by the APP-A673T mutation in cellular models (Maloney et al., 2014). Although the rat model described above harbors the A673T mutation and showed increased α-cleavage (Tambini et al., 2020), this increase in α-cleavage was not apparent in our AppG-F-A673T mice as assessed from the amount of CTF-α produced (Fig. 3D). This discrepancy may be due to the fact that AppG-F mice have FAD-causing mutations that may affect the structure and susceptibility of APP protein to α-cleavage.
Maloney et al. also reported that Aβ with A673T prepared in a 1:9 ratio mixture of Aβ1– 42/Aβ1– 40 displayed slightly slower aggregation kinetics compared with wild-type peptides in vitro, an observation that may contribute in part to the genetic protection afforded by the A673T variant (Maloney et al., 2014). Interestingly, the amount of Aβ42 in the GuHCl-soluble fraction was not altered in AppG-F-A673T mice, probably due to the presence of the Arctic mutation in the Aβ sequence. The Arctic mutation alters the characteristics of Aβ and makes it more aggregate-prone and resistant to Aβ-degrading enzymes such as NEP (Tsubuki et al., 2003; Lord et al., 2006). This result suggests that the APP-A673T mutation has a relatively weak effect on the structure and/or solubility of Aβ in our model, although this mutation needs to be introduced into a mouse model without any other mutation in the Aβ sequence before an accurate assessment can be made. Previously it was suggested that the Icelandic mutation also affects the susceptibility of APP fragments to γ-secretase as indicated by in vitro assay studies (Kokawa et al., 2015). This hypothesis was not tested in the current study because AppG-F mice harbor the Beyreuther/Iberian (I716F) mutation that affects γ-secretase activity, thus making it difficult to precisely assess γ cleavage. Further study is needed on mouse lines without the Iberian mutation to clarify the effect of APP-A673T on γ-cleavage.
Recently, the anti-Aβ antibody lecanemab was reported to significantly delay the progression of cognitive dysfunction and to decrease amyloid burden in mild cognitive impairment and AD patients at an early stage (van Dyck et al., 2023). This resulted in its approval by the Food and Drug Administration in the USA and attests to the important proof-of-concept that amyloid pathology is a druggable target and that amyloid-targeting therapies can modify the disease course even after the clinical onset of dementia. However, lecanemab failed to halt disease progression or improve cognitive function. This is thought to be because other physiological cascades, such as tau pathology and neuronal degeneration, had already begun by the time lecanemab was administered. To prevent AD onset, amyloid-targeting therapies need to be initiated at a much earlier phase of the disease. Furthermore, we cannot overlook amyloid-related imaging abnormalities (ARIAs) which present as edema (ARIA-E) or hemorrhages (ARIA-H), or both, which have been reported as side effects of anti-Aβ human monoclonal antibody therapies (Hampel et al., 2023). The development of new treatments and prevention methods for AD is an urgent issue. Although BACE1 inhibitors that directly inhibit APP cleavage reduce amyloid pathology, they have failed in clinical studies due to side effects associated with other BACE1 substrates than APP (Egan et al., 2018). Our findings indicate that more specific inhibition of the APP-BACE1 interaction could still be a promising therapeutic approach. Alternatively, it may be possible to introduce the Icelandic mutation into the APP gene through genome editing technologies such as base editing (Guyon et al., 2021) or prime editing (Tremblay et al., 2022), both of which are modified versions of the CRISPR-Cas9 system. Recently, in vivo genome editing utilizing these tools in mice has been actively studied and demonstrated to be effective in various tissues and cell types including the brain and neurons (Arbab et al., 2023; Ling et al., 2023). If the Icelandic mutation can be introduced into individuals at a young age who are at high risk of developing AD, such as carriers of the gene mutation responsible for familial AD or homozygous carriers of APOE ε4, it could be effective in preventing the disease. Moreover, the introduction of protective mutations using genome editing may also be applicable to the treatment of Down’s syndrome, in which overexpression of APP leads to dementia symptoms in adulthood and shows AD-like pathological lesions such as Aβ deposition.
It has also been suggested that the Icelandic mutation may be protective against age-related cognitive decline (Jonsson et al., 2012), although it is unclear whether the protective mechanism depends on amyloid pathology. As indicated in a previous in vivo study employing a rat model, many APP metabolites other than Aβ can be altered by the Icelandic mutation (Tambini et al., 2020) and may affect neuronal functions. For example, secretory APPα is reported to be neuroprotective (Rossner et al., 1998). In future studies, it would be interesting to introduce the Icelandic mutation into wild-type mice and evaluate different cognitive functions and anatomical and pathological changes in the brains of aged mice.
Footnotes
We thank Y. Nagai-Watanabe for secretarial work. This work was supported by JSPS KAKENHI Grant Number JP24KJ1951 (S.S.), JP18K07402 (H.S.) and the Takeda Science Foundation (H.S.).
The authors declare no competing financial interests.
- Correspondence should be addressed to Hiroki Sasaguri at hiroki.sasaguri{at}riken.jp.
