Knock-in rats with homozygous PSEN1L435F Alzheimer mutation are viable and show selective γ-secretase activity loss causing low Aβ40/42 and high Aβ43.

Familial forms of Alzheimer's disease (FAD) are caused by mutations in the gene encoding amyloid precursor protein, whose processing can result in formation of β-amyloid (Aβ). FAD can also result from mutations in the presenilin 1/2 (PSEN1/2) genes, whose protein products partially compose the γ-secretase complex that cleaves Aβ from amyloid precursor protein fragments. Psen1 KO mice and knock-in (KI) mice with homozygous FAD-associated L435F mutations (Psen1LF/LF) are embryonic and perinatally lethal, precluding a more rigorous examination of the effect of Alzheimer's disease-causing Psen1 mutations on neurodegeneration. Given that the rat is a more suitable model organism with regard to surgical interventions and behavioral testing, we generated a rat KI model of the Psen1LF mutation. In this study, we focused on young Psen1LF rats to determine potential early pathogenic changes caused by this mutation. We found that, unlike Psen1LF/LF mice, Psen1LF/LF rats survive into adulthood despite loss of γ-secretase activity. Consistent with loss of γ-secretase function, Psen1LF/LF rats exhibited low levels of Aβ38, Aβ40, and Aβ42 peptides. In contrast, levels of Aβ43, a longer and potentially more amyloidogenic Aβ form, were significantly increased in Psen1LF/LF and Psen1LF/w rats. The longer survival of these KI rats affords the opportunity to examine the effect of homozygous Psen1 Alzheimer's disease-associated mutations on neurodegeneration in older animals.


INTRODUCTION
Familial Alzheimer's disease (AD) are caused by mutations in PSEN1 and PSEN2, with the majority occurring in PSEN1 (1). These genes encode Presenilin 1 (PS1) and Presenilin 2 (PS2), members of the γ-secretase complex (2,3). ADcausing mutations occur at hundreds of different loci over the span of PSEN1 (www.alzforum.org/mutations), and the biochemical effect of PSEN1 mutations is complex. In general, PSEN1 mutations result in a decrease of endopeptidase activity and altered γ-processivity, resulting in reduced amounts of γ-secretase products and a relative increase the longer forms of γ-secretase products. With regard to Amyloid Precursor Protein (APP) processing, PSEN1 mutations show reduced levels of amyloid beta(4) (Aβ) and some but not all mutations show a relative increase in longer forms of Aβ (5), the accumulation of which is seen in AD. In addition to the diverse effects on metabolite levels of a single substrate, γ-secretase has multiple substrates (6) whose function may impact neurodegenerative and neurodevelopmental processes in manner unrelated to the neurodegeneration caused by Aβ. Knock-in mouse (7) and in vitro models (8) of the AD-causing PSEN1 L435F mutation show a near complete abrogation of γ-secretase activity and a reduction in total amyloid production. There are reports of a relative increase in Aβ43, a longer and potentially more amyloidogenic form of Aβ, in PSEN1 L435F AD-brains (9) and cell lines (9,10) expressing PS1-L435F, though the absolute amount of Aβ43 produced is low, and in the case of KI-mouse models (7), undetectable. The PSEN1 L435F mutation has not been studied in homozygosis, as Psen1 L435F homozygote mice are perinatally lethal (7) in a manner that resembles the early embryonic lethality of Psen1 knockout (KO) mice (11), likely the result of the PS1 L435Fmediated disruption of Notch signaling. Given this lethality, the Psen1 L435F mutation was characterized in heterozygosis, on Psen2 KO background to eliminate compensation from PS2 (7). Analysis of heterozygote Psen1 L435F, Psen2-KO mice showed marked synaptic, memory deficits, and an age-dependent neurodegenerative phenotype (7). Here, we create a rat knock-in model of the Psen1 L435F mutation in a rat that expresses APP in which the Aβ region has been humanized (Psen1 LF rats). A CRISPR/Cas9mediated knock-in system was chosen to avoid the artifacts induced by the transgenic approach (i.e. non-physiological overexpression, the use of nonendogenous and/or non cell type-specific regulatory elements, and the disruption of endogenous genes at integration sites). The rats were placed on a humanized APP background (12) to accommodate the possibility of differences in pathogenicity of rodent and human Aβ. Consistent with the mouse KI model, we find loss of γsecretase function in Psen1 LF/LF rats, which show minimal levels of Ab38, Ab40 and Ab42 peptides: in contrast, concentrations of Aβ43 are significantly increased in both Psen1 LF/LF and Psen1 LF/w rats. Unexpectedly, we also find that homozygote Psen1 LF rats are born in Mendelian ratios, survive into adulthood, and have preserved neurodevelopment and Notch signaling, despite altered APP metabolism. Psen1 LF rats may therefore provide a useful model for the examination of neurodegenerative changes caused by PSEN1 L435F mutation.

Generation of Psen1 LF rats carrying humanized App alleles (App h/h ).
F0-Psen1 LF rats were crossed to Long-Evans rats to generate F1-Psen1 LF/w rats. These crossings were repeated four more times to obtain F5-Psen1 LF/w rats. The probability that F5 rats carry unidentified off-target mutations (except those, if present, on Chr. 6) is ~1.5625%. To generate Psen1 LF rats on a background in which rat App has a humanized Aβ region, F5-Psen1 LF/w and App h/h rats were crossed to generate F1-Psen1 LF/w ; App h/w rats. The App w allele was removed in subsequent crosses. For all data generated in this study, all rats are on the App h/h background, which produces human and not rodent Aβ species.
To verify that the Psen1 LF mutations were correctly inserted into Psen1 exon 12, we amplified by PCR the Psen1 gene exon 12 from Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats.
Sequencing of the PCR products shows that the mutations were correctly inserted in the Psen1 LF/w and Psen1 LF/LF genomes (Fig. 1A). RT-PCR analysis performed on RNA from p0 rat brain lysate from Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats confirmed no alterations in Psen1 or Psen2 expression is caused by the L435F mutation ( Fig.  1B (7) and developmental abnormalities consistent with Psen1-KO mice (11). To determine the lethality of the L435F mutation in KI rats, Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats were genotyped at birth, weighed at weaning, and followed for several weeks. Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats were born in Mendelian ratios (Fig. 1D). At the time of weaning (p27), male and female Psen1 LF/LF rats weighed significantly less than Psen1 w/w and Psen1 LF/w littermates (Fig. 1E) Figure]. Psen1 LF/LF rat survival declined to 65% by day 28, and then stabilized into adulthood, while Psen1 w/w and Psen1 LF/w littermates showed no significant postnatal lethality in the same period (Fig. 1F).
Presenilinase activity, a prerequisite for γsecretase function, is reduced in Psen1 LF rat brains.
Typically, γ-secretase mediates intramembranous cleavage of C-terminal fragments derived by prior processing by either aor b-secretase. As for APP, aand b-secretase generate two γ-secretase substrates, APP-αCTF and APP-βCTF, respectively. a-Secretase cleavage of N-cadherin yields the γ-secretase substrate N-cad-CTF. Thus, to assess γ-secretase function in Psen1 LF rats, solubilized brain lysate from p4 Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF pups was analyzed by western blotting for steady-state levels of these γsecretase substrates. Psen1 LF/LF rat brains showed a sex-independent increase in APP-αCTF, APP-βCTF and N-cad-CTF, while Psen1 LF/w rat brains were indistinguishable from wild-type controls ( Fig γ-Secretase activity and processivity is reduced in Psen1 LF rat brains, increasing the ratio of long Aβ peptides/short Aβ peptides. Cleavage of APP-βCTFs by γ-secretase generates Aβ peptides, which vary in length depending on the processivity of γ-secretase: reduced processivity increases the relative amounts of longer Aβ peptides as compared to shorter Aβ peptides. To complete the assessment of γ-secretase function in Psen1 LF rats, solubilized brain lysates from Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF pups were analyzed by ELISA. Solubilized lysate was chosen for analysis as the rats show no insoluble Aβ plaques by immunohistochemistry (Fig. 5E). Psen1 LF/LF rat brains had lower levels of Aβ38, Aβ40, and Aβ42, in a sex independent manner (  Figure]. Psen1 LF/w animals in general had similar amyloid levels as compared to wild-type rats, with the exception of lower levels of Aβ38 in female Psen1 LF/w rats (Fig. 3A). Notably, the Aβ42/Aβ40 ratio was also increased in Psen1 LF/LF rat brain lysates in a sex independent manner ( Overall, the decrease in γ-secretase products (Aβ peptides), increase in γ-secretase substrates (APP-βCTF, APP-αCTF, N-cad-CTF) and decrease in the autocatalysis of PS1 in Psen1 LF rats is indicative of a loss of γ-secretase function.
To assess for α-and β-secretase cleavage of APP, solubilized brain lysates were analyzed by ELISA for soluble APP ectodomain levels, sAPPα and sAPPβ, which are the other products of α-and β-cleavage of APP, respectively. No differences were seen in Psen1 LF/w rat brains as compared to wild-type rats (Fig. 3B). Surprisingly, Psen1 LF/LF rat brain lysates show an increase in sAPPα levels and a decrease in sAPPβ levels (

PS1-L435F forms a γ-secretase complex.
Given the overall trend towards Psen1 LF conferring a loss of function phenotype, Aβ43 levels notwithstanding, we wished to determine if the loss of function was the result of the inability of PS1-L435F to form an active γ-secretase complex. Anti-PS1 and anti-PS2 antibodies were used to immunoprecipitate γ-secretase complexes from 1%-CHAPSO solubilized brain lysate from Psen1 w/w and Psen1 LF/LF rats (Fig. 4). Both Psen1 w/w and Psen1 LF/LF samples co-immunoprecipitated Nicastrin and Pen2 when immunoprecipitated with anti-PS1, indicating that PS1-L435F forms a γsecretase complex. Anti-PS2 immunoprecipitated PS2-containing complexes, as seen by the presence of Nicastrin in the eluate, though Pen2 levels were below the level of detection. To determine if the biochemical changes caused by the Psen1 L435F mutation impact neurodevelopment or cause neuropathology, we used histology and immunohistochemical (IHC) analysis to characterize brains from p15 Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats. Regions of analysis included the frontal cortex, cingulate cortex, whole hippocampus, and entorhinal cortex. No gross morphological changes were evident by hematoxylin and eosin (H&E) staining in any of the rats analyzed (Fig. 6A). Qualitative inspection of NeuN staining showed no appreciable changes in neuronal density in any of the regions analyzed in Psen1 LF rats, though a quantitative analysis of total NeuN signal finds a small but statistically significant increase in male Psen1 LF/LF rat whole hippocampus and the CA2-3 region (Fig. 6B). No evidence of astrocytosis or microgliosis was seen by staining with GFAP and IBA1, respectively, in any of that rats tested (Fig. 6C-D). Amyloid plaques, as measured by staining the anti-Aβ antibody 6E10 were absent in all rats tested ( Fig.  6E-F). Overall, histological analysis of these rats shows no evidence of neurodevelopmental impairments or AD-like pathology at 15 days.

DISCUSSION
The choice of animal model and genetic approach have profound implications on the phenotypic expression of disease-associated mutations. Given the better suitability of rats for behavioral tests, surgical procedures, and the expression of tau isoforms that more closely reflects human tau splicing, we chose to model AD-related mutations using a Long Evans rat KI model (15). The Psen1 L435F mutation was selected given its profound alteration of APP metabolism (8) and age-dependent neurodegenerative changes seen in the KI-Psen1 LF mice (7). Here, we have studied young Psen1 LF rats with the purpose of determining potential early pathogenic mechanisms caused by this pathogenic mutation. Unexpectedly, we found that in contrast to Psen1 LF/LF mice, Psen1 LF/LF rats survive into adulthood. This survival is likely the result of the Notch-sparing phenotype seen in Psen1 LF/LF rats that is absent in Psen1 LF/LF KI-mice. Three nonmutually exclusive possibilities may underlie this Notch-sparing effect: 1. PS1-L435F can assemble in a g-secretase complex (Fig. 4). Given the profound decrease in processing of APP and Ncadherin by in Psen1 LF rats, it is unlikely though still formally possible that the mutant PS1 is catalytically active in rats in a substrate specific manner. This possibility would be in line with previous data showing that FAD mutant PS1 were able to rescue the Notch phenotype independent of the APP pathway (16). 2. It is also possible that, given the reduction in autocatalysis, PS1-L435F changes localization and is sequestered within the cell such that is exposed to different substrates than PS1-WT. 3. There may be a partial compensation of catalytically inactive PS1-L435F by PS2 (17). Rat Psen2 may be expressed earlier as compared to mouse Psen2 during embryonic development. At P0, this compensation would necessarily be qualitative change in localization/activity and not a difference in quantity, as Psen2 expression and PS2 levels are unchanged in Psen1 LF/LF rats.
Amyloid levels vary considerably between models. While Aβ40 and Aβ42 levels were undetectable in Psen1 LF/LF mice, both species were detected in our present study at about 9% and 40% of wild-type controls for Aβ40 and Aβ42, respectively. In general, cell culture and in vitro models show that PS1 L435F mediates a loss of Aβ40 and Aβ42 production. PS1 L435Freconstituted PS1/2-KO mouse embryonic fibroblasts (10) demonstrates undetectable Aβ40 levels and a >90% reduction in Aβ42, while stablytransfected PS1-L435F HEK cells (9) show >90% reduction in both species, though it must be considered that both these cell lines overexpress APP and PS1, and therefore no inference can be made on the absolute levels of Aβ production. In liposome-based in vitro assays of recombinant PS1 L435F activity (4), PS1 L435F γ-secretase activity, as measured by Aβ40 and Aβ42, was found to be nearly undetectable, at 0.007x the activity of wildtype PS1. Measurement of Aβ43 has similarly varied across models and groups, with in vitro overexpression models of PS1-L435F activity demonstrating an increase in the relative amounts of Aβ43 (9,10), a finding not recapitulated in Psen1 LF KI mice (7). Our models reveal an absolute increase in Aβ43 and in Aβ42/40 and Aβ43/40 ratios in KI rats, though these observations occur in the setting of decreased total Aβ and no apparent Aβ aggregation. How well these models relate to amyloid metabolism of PSEN1 L435F AD patients is unclear, as total amyloid levels have not been determined in autoptic brain tissue, though aggregated forms of Aβ42 and Aβ43 are present in histopathological analysis (9).
The use of KI system, in which endogenous APP is expressed, allows for a more complete analysis of APP metabolism beyond Aβ. APP-CTFs are the direct substrate of PS1/2, and β-CTFs are expectedly increased in Psen1 LF/LF rats concurrent with Aβ reduction. This increase in APP-CTFs may have per se a pathogenic effect (18)(19)(20)(21). In addition, Psen1 LF/LF rats show an effect on the metabolism of full-length APP as well. Specifically, there is a significant increase in sAPPα and a significant decrease in sAPPβ, indicative of a shift toward the α-processing of APP. A co-ordination between γ-and α/βprocessing is possible given the recent evidence that a fraction of γ-secretase exists in a tripartite macromolecular complex with APP and ADAM10 (22) or BACE1 (23). A stalled or otherwise inactivated PS1-L435F-containing γ-complex may differentially affect the complex's ability to bind ADAM10 or BACE1. Apart from the potential general impact that PS1-L435F may on α/βsecretase, the finding of increased sAPPα and decreased sAPPβ is significant by itself, as these and other non-Aβ metabolites of APP have been implicated as modulators of synaptic activity (12,(24)(25)(26) and neuronal survival (27).
While the IHC analysis of day 15 Psen1 LF rats is consistent with normal neurodevelopment, there is also no indication of Aβ plaques, astrogliosis, or microgliosis that occur in AD. The lack of amyloid pathology at day 15 is unsurprising considering that even in animal models in which APP with ADrelated mutations is overexpressed, plaques take at least 6 weeks to develop (28). Psen1 LF rats may require extensive aging or additional mutations to develop AD-related histopathological changes; however, given the survival of Psen1 LF/LF rats and the avoidance of a Notch-related phenotype, the Psen1 LF rat KI model presents a useful, physiologically appropriate model in which to study age-related neurodegeneration in AD.

EXPERIMENTAL PROCEDURES Rats and ethics statement.
Rats were handled according to the Ethical Guidelines for Treatment of Laboratory Animals of the NIH. The procedures were described approved by the Institutional Animal Care and Use Committee (IACUC) at Rutgers.
Generation of rats expressing the FAD Psen1 L435F mutation (Psen1 LF rats). The rat Psen1 gene (GenBank accession number: NM_019163; Ensembl: ENSRNOG00000009110) is located on rat chromosome 6. We created a Long-Evans rat model with point mutation CTT>TTT at rat Psen1 locus by CRISPR/Cas-mediated genome engineering. This mutation will create a rat that carries a Psen1 gene coding for PS1 with the FAD L 435 F mutation. The detailed procedures are reported in the Supporting Information file. Rat brain preparation. Rats were anesthetized with isoflurane and perfused via intracardiac catheterization with ice-cold PBS. Brains were extracted and homogenized using a glass-teflon homogenizer (w/v =100 mg tissue/1 ml buffer) in 250 mM Sucrose, 20 mM Tris-base pH 7.4, 1 mM EDTA, 1mM EGTA plus protease and phosphatase inhibitors (ThermoScientific), with all steps carried out on ice or at 4°C. Total lysate was solubilized with 0.1% SDS and 1% NP-40 for 30 min rotating. Solubilized lysate was spun at 20,000 g for 10 min, the supernatant was collected and analyzed by ELISA and Western blotting.
ELISA. For analysis of Aβ38, Aβ40, Aβ42, sAPPα and sAPPβ, the following Meso Scale Discovery kits were used: Aβ38, Aβ40, and Aβ42 were measured with V-PLEX Plus Aβ Peptide Panel 1 6E10 (K15200G) and V-PLEX Plus Aβ Peptide Panel 1; sAPPα and sAPPβ were measured with sAPPα/sAPPβ (K15120E). Measurements were performed according to the manufacturer's recommendations. Plates were read on a MESO QuickPlex SQ 120. For analysis of Aβ43, IBL Human Amyloidβ (1-43) (FL) Assay Kit (27710) was used according to the manufacturer's recommendations. Specificity of this Kit was validated by us using a rat App d7/d7 hypomorph control (12,29) (validation data are presented in supporting information, Fig. S2). Data were analyzed using Prism software and represented as mean ± SD.
Immunoprecipitation. Total brain lysate was diluted in IP buffer (50 mM Tris, 150mM NaCl, 1mM EGTA, 1mM EDTA, pH 8.0) with 1% CHAPSO, solubilized for 30 min at 4°C rotating, and spun at 20,000 g for 10 min. Solubilized lysate was used as input for immunoprecipitation with anti-GFP (Cell signaling, 2555), anti-PS1-CT, or anti-PS2-CT antibodies and protein A/G beads (Thermo, 20421) overnight at 4°C rotating. After several wash steps, bound protein was eluted by 5m incubation with 1X-LDS sample buffer at 55°C. Input and eluates were analyzed by western analysis.
RT-PCR. Total brain RNA was extracted with RNeasy RNA Isolation kit (Qiagen) and used to generate cDNA with a High-Capacity cDNA Reverse Transcription Kit (Thermo). 50ng cDNA, TaqMan™ Fast Advanced Master Mix (Thermo 4444556), and the appropriate TaqMan (Thermo) probes were used in the real time polymerase chain reaction. Samples were analyzed on a QuantStudio 6 Flex Real-Time PCR System (Thermo), and relative RNA amounts were quantified using LinRegPCR software (hartfaalcentrum.nl). The probe Rn00570673_m1 (exon junctions 11-12, 12-13, and 13-14) was used to detect rat Psen1 and samples were normalized to Gapdh levels, as detected with Rn01775763_g1 (exon junctions 2-3, and 7-8). Levels of Notch target gene transcripts was determined using the RT² Profiler™ PCR Array Rat Notch Signaling Pathway plate (Qiagen, 330231 PARN-059Z) according to the manufacturer's recommendations. Student's t-test was used for all analyses, with data presented as mean ± SD. Immunohistochemistry (IHC). Staining tissue preparation and sectioning. Rat brain tissue was prepared and stained as described previously. Briefly, intracardiac PFA-perfused rat brains were extracted and stored in 70% ethanol prior to cerebral coronal sectioning. Sections were dehydrated and paraffin embedded and the processed into 15 cross sections targeting the frontal cortex at the level of the isthmus of the corpus callosum, anterior and posterior hippocampus. IHC staining was performed in accordance with Biospective Standard Operating Procedure (SOP) # BSP-L-06. Slides were manually de-paraffinized and rehydrated prior to the automated IHC. Slides initially underwent antigen retrieval, either heat-induced epitope-retrieval or formic acid treatment. All IHC studies were performed at room temperature on a Lab Vision Autostainer using the REVEAL Polyvalent HRP-AEC detection system (Spring Bioscience). Briefly, slides were incubated sequentially with hydrogen peroxide for 5 minutes, to quench endogenous peroxidase, followed by 5 minutes in Protein Block, and then incubated with primary, antibodies as outlined in Table 1. Antibody binding was amplified using the Complement reagent (20 min), followed by an HRP-conjugate (20 min), and visualized using the AEC chromogen (20 minutes). All IHC sections were counterstained with Acid Blue 129 and mounted with aqueous mounting medium (30). Image analysis of IHC sections. The IHC and histology slides were digitized using an Axio Scan.Z1 digital whole-slide scanner (Carl Zeiss). The images underwent quality control (QC) review and final images transferred to the Biospective server for qualitative image analysis. All qualitative assessments were performed blinded to the tissue genotype.
Statistical analysis. Statistical significance was evaluated using Ordinary one-way ANOVA followed by Post-hoc Tukey's multiple comparisons test when applicable (i.e. when the Ordinary one-way ANOVA test showed statistical significance). Statistical analysis was performed with GraphPad Prism v8 for Mac. Significant differences were accepted at p < 0.05.
Data Availability Statement. The authors indicate that all of the data is contained within the manuscript  represented as mean ± SD. Data were analyzed by ordinary one-way ANOVA followed by post hoc Tukey's multiple comparisons test when ANOVA showed statistically significant differences. **p < .01; ****p < .0001. N's of each genotype: 12 female Psen1 w/w , 29 female Psen1 LF/w , 13 female Psen1 LF/LF , 20 male Psen1 w/w , 27 male Psen1 LF/w , and 9 male Psen1 LF/LF . (D) Survival curve of Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats from birth to day 28, n=110 total. (E) Levels of Psen1 RNA from brain lysate were measured in Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF p0 rats by quantitative RT-PCR, and normalized to Gapdh levels. No significant differences between Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats were evident. Data are represented as mean ± SD. Data were analyzed by ordinary one-way ANOVA. N≥5 rats per genotype. (F) Levels of Psen2 RNA from brain lysate were measured in Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF p0 rats by quantitative RT-PCR, and normalized to Gapdh levels. No significant differences between Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats were evident. Data are represented as mean ± SD. Data were analyzed by ordinary one-way ANOVA. N≥5 rats per genotype.  Data are represented as mean ± SD. Data were analyzed by ordinary one-way ANOVA followed by post hoc Tukey's multiple comparisons test when ANOVA showed statistically significant differences. N=4 rats per sex per genotype. *p < .05; **p < .01; ***p <0.001; ****p <0.0001. Figure 3. ELISA measurements of amyloid species and soluble APP species in Psen1 LF rats. (A) ELISA levels of Aβ38, Aβ40, and Aβ42 in male and female Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF p4 rat brain lysate. Ratio of Aβ42/ Aβ40 is also presented. We used the following numbers of samples: Psen1 w/w females n=6, Psen1 LF/w females n=7, and Psen1 LF/LF females n=7, Psen1 w/w males n=6, Psen1 LF/w males n=7, and Psen1 LF/LF males n=6. (B) ELISA levels of Aβ43 in Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF p4 rat brain lysate. Ratios of Aβ43/Aβ40 and Aβ43/Aβ42 are also presented. Samples used: n=4 per sex, per genotype. Aβ43 was quantified with the IBL Human Amyloidβ (1-43) (FL) Assay Kit (27710), validated by us using a rat App hypomorph control as shown in Fig. S2. (C) ELISA levels of sAPPα and sAPPβ in Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF p4 rat brain lysate. Samples used: Psen1 w/w females n=6, Psen1 LF/w females n=7, and Psen1 LF/LF females n=7, Psen1 w/w males n=6, Psen1 LF/w males n=7, and Psen1 LF/LF males n=6. Data are represented as mean ± SD. Data were analyzed by ordinary one-way ANOVA followed by post hoc Tukey's multiple comparisons test when ANOVA showed statistically significant differences. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001. Figure 4. Immunoprecipitation of γ-secretase complexes from Psen1 w/w and Psen1 LF/LF rats. 1% CHAPSOsolubilized membrane from Psen1 w/w and Psen1 LF/LF rat brains was immunoprecipitated and analyzed by western blot. γ-Secretase complexes were immunoprecipitated with anti-PS1 or anti-PS2, and anti-GFP was used as a control. Input (IN) and eluate (IP) are shown. Both PS1-WT and PS1-L435F specifically bind Nicastrin and Pen2. Figure 5. RT-PCR analysis of Notch intracellular domain target genes in Psen1 LF rats. Expression levels of several NICD target genes was evaluated in RNA derived from p0 rat brain lysate from Psen1 w/w and Psen1 LF/LF pups. Expression of (A) Cyclin Dependent Kinase Inhibitor 1A (Cdkn1a), (B) CASP8 and FADD-like apoptosis regulator (Cflar) (C-D) Hairy and enhancer of split-1 and 5 (Hes1 and Hes5) showed no significant differences between Psen1 w/w and Psen1 LF/LF rats. Data are represented as mean ± SD. Data were analyzed by ordinary oneway ANOVA. Samples used: for Cdkn1a, CASP8, Cfla and Hes1; Psen1 w/w n=3, Psen1 LF/LF n=4; for Hes5, Psen1 w/w n=6, Psen1 LF/LF n=5. Figure 6. Histopathological analysis of Psen1 LF rats. IHC analysis of p15 rat whole hippocampus (HC), hippocampal subregions (CA1, CA2-3, CA4-DG), or entorhinal cortex (EntCx). A.) H&E-stained sections from Psen1 w/w , Psen1 LF/w , and Psen1 LF/LF rats. B.) Evaluation of neuronal number by NeuN staining. Quantitation of total NeuN intensity on the right. N≥5 rats per sex, per genotype. Data are analyzed by one-way ANOVA with Kruskal-Wallis post-test, *p<0.05. Data are represented as mean ± SD. C.) Evaluation of astrogliosis by Gfap staining. Quantitation of total Gfap intensity on the right. Data are analyzed by one-way ANOVA. Data are represented as mean ± SD. D.) Evaluation of microgliosis by Iba1 staining. Quantitation of total Iba1 intensity on the right. N≥5 rats per sex, per genotype. Data are analyzed by one-way ANOVA. Data are represented as mean ± SD. E.) Evaluation of amyloid plaques by 6E10 staining. Scale bar 500 µm.