APOEe4.Trem2*R47H Mice Show Changes in Alzheimer’s Disease-Relevant Processes in the Absence of Amyloid Plaques


 Late-onset Alzheimer’s disease (LOAD) is the most common human neurodegenerative disease. Legacy amyloidogenic mouse models have been useful for understanding disease progression, however in the face of failing human trials more focus on disease translation with new mouse strains that better model human Alzheimer’s disease (AD) is required. MODEL-AD (Model Organism Development and Evaluation for Late-onset AD) groups are identifying and integrating disease-relevant, humanized gene sequences from public databases to create more translatable mouse models for therapy development. Mice expressing strong genetic risk factors for LOAD, APOEe4 and Trem2*R47H, were extensively aged and assayed using a multi-disciplined phenotyping approach associated with and relative to human AD pathology. Behavioral, transcriptomic, metabolic, and neuropathological assays identified sex and age as the main sources of variation between genotypes including age-specific enrichment of AD-related processes in the absence of mouse amyloid plaque formation. These data provide an important, baseline understanding of the individual effects and interaction between two strong genetic risk factors for LOAD. These two alleles together form a sensitized, background strain (B6.APOE4.Trem2*R47H, which we have termed ‘LOAD1’) necessary to examine how important underlying risk factors interact with any subsequent genetic or environmental cues to drive pathology.


19
Alzheimer's disease (AD) is the most common form of dementia, currently affecting 5.8 million 20 patients in the United States (1), ~95% of which are late-onset (LOAD) (2). There are no approved 21 interventions to delay or reverse AD-related neurodegeneration (3). LOAD is a heterogeneous disease 22 defined by a few widely accepted hallmarks: extracellular amyloid plaques, intracellular tau tangles, 23 vascular disfunction, immune activation, synapse loss, neuron death, and cognitive decline (4). Dissecting 24 the etiology of these properties in animal models has provided key insights into our understanding of the 25 disease and propose strategies to treat AD, however human clinical trials of resultant therapeutics have not 26 yielded an approvable drug.

27
To decipher how these two strong risk factors drive AD-relevant phenotypes, we created a double 1 homozygous B6.APOE4.Trem2*R47H model, accompanied by single genotype controls, on a C57BL/6J 2 (B6) background (Table 1). In appreciation of sexual dimorphism observed in human aging and disease, 3 cohorts of males and females were established for phenotyping at 4-, 8-, 12-and 24-months using a cross-4 sectional design. To determine whether the LOAD risk variants APOEe4, Trem2*R47H, or the combination 5 produced in vivo phenotypes independent from normal healthy aging, a comprehensive cross-sectional 6 phenotyping battery was conducted and included in vivo frailty assessments, metabolic screening, 7 microbiome sampling, biomarker evaluation, behavioral phenotyping, and in vivo imaging. Postmortem 8 brain tissue was further examined for transcriptomic and neuropathological indications of disease. All 9 accumulated data sets and observations are disseminated for public availability (44).

11
Biometric profiles of APOE4.Trem2*R47H mice change with age 12 Comparison of biometric data from young (4 months) and aged mice (24 months), comprised of 13 both sexes from four genotypes, revealed the expected age-related accumulation in physical frailty 14 characteristics (Figure 1 A,B) with inverse correlations in body temperature (Supplemental Figure 1) and 15 age-related increases in body weights (Figure 1 C,D). However, effects due to genotype were not observed 16 overall (see also Supplemental Table 1 Table 2). Additional cohorts,

11
Compared to 4-month-old animals, we observed a decrease in activity during the active period (dark phase) 12 by 24 months (Supplemental Figure 4). At 24 months of age, home cage running wheel activity suggested 13 a correlation between increased activity during the active period and expression of the APOE 4 allele (both 14 B6.APOE4 and B6J.APOE4/Trem2*R47H mice) by way of total distance traveled and day-time activity in 15 B6.Trem2*R47H mice (Supplemental Figure 4 E,F). In addition to monitoring physical wellness and 16 behavior of otherwise healthy mice, we also wanted to track the effects of sex and genotype on animal 17 longevity, in the absence of amyloid-associated AD. For this analysis, mortality was defined as subjects 18 that were found dead with no obvious signs of infection, trauma, or intervention during daily monitoring.

19
Mortality risk for each allele was determined by comparing survival scores of cohorts aged to 24 months.

20
Overall, females had a greater risk than males, and survival probabilities were lowest in both sexes for 21 animals expressing both LOAD risk alleles (Supplemental Figure 5). 22

PET/MRI identified age-and genotype-dependent differences in glycolysis and tissue perfusion 24
In an effort to understand the role of risk alleles on regional glycolysis and tissue perfusion, 25 translationally relevant regional measures were acquired via 18F-FDG and 64Cu-PTSM PET/MRI and 26 autoradiography, respectively. By 12 months glycolysis was altered in key brain regions associated with 27 sensory integration, cognition, vision and motor function in B6.APOE4 and B6.APOE4.Trem2*R47H mice, 28 when compared with controls (Figure 3), and were confirmed via post-mortem autoradiography, which has 1 a 40 folder greater resolution than PET. As expected, these changes were greater in number of regions 2 and magnitude of change in female mice when compared to males (Figure 3 B,C). These changes were 3 similarly observed through time, where female mice showed significantly altered glycolysis at 4, 8 and 12 4 months, while male mice largely showed a hypoglycolytic phenotype at 8 months, that was virtually 5 mitigated by 12 months (Supplemental Figure 6). Since these risk alleles can alter metabolic functionality 6 and neuroinflammatory-driven tissue perfusion in an independent manner, we quantitatively measured 7 changes in regional tissue perfusion via 64Cu-PTSM PET/MRI and confirmed this via autoradiography.

10
Interestingly, these changes were manifested temporally, with the greatest reductions occurring across 11 genotypes and regions in both sexes at 4 months (Supplemental Figure 7). Unlike glycolysis, these 12 changes were largely resolved by 8 months in female mice, while males continued to show regional 13 reduction in perfusion at this same age.

15
Biochemical and neuropathological effects of APOEe4 and Trem2*R47H alleles 16 Confirmation of protein expression levels in brain tissue were confirmed for alleles encoding human 17 APOE4 and mouse TREM2 carrying the R47H mutation (Supplemental Figure 8). Similar to reports of 18 R47H variant-mediated reduction in Trem2 transcript levels (6), TREM2 protein levels in the brains of these 19 animals were also decreased. However, instead of a near knock-out of all TREM2 that has been reported 20 previously (6), levels fell by approximately 50% in Trem2*R47H animals compared to C57BL/6J 21 (Supplemental Figure 8 A,B). We have previously shown APOE4 protein levels are similar to endogenous 22 mouse APOE (14) and expression of APOE4 appeared similar between male and female 23 APOE4.Trem2*R47H mice (Supplemental Figure 8C) . Additional molecular characterization of these 24 animals showed both age-and genotype-driven differences in levels of cytokines present in the brain and 25 blood (Supplemental Figure 9). IL-6 and KC/GRO concentrations were highest in 26 B6.APOE4.Trem2*R47H brain tissue at 8 months, while blood plasma concentrations continued to increase 27 with age in those mice (Supplemental Figure 9 B,D,E). In multiple occasions a trend appeared to suggest 1 increased cytokine concentrations in mice expressing mutated allele Trem2.
2 Neuropathological features of AD were then investigated by hematoxylin and eosin (H&E, structure) 3 and luxol fast blue/cresyl violet (LFB/CV, myelin) staining but did not reveal any gross anatomical changes 4 to tissue architecture or myelin ( Figure 5A). Brain sections were also imaged via immunofluorescence and 5 included neuritic plaque-reactive-microglia (X34/Lamp1/Iba1), vascular leakage (CD31/Iba1/Fibrin), and 6 ThioflavinS (amyloid plaques and neurofibrillary tau tangles) ( Figure 5B). No gross abnormalities in cell 7 counts, nor additional neuropathological features were observed in 24 month B6.APOE4.Trem2*R47H 8 mice. We focused particularly on the cortex and hippocampus where episodic memory (hippocampus) and 9 memory behavior (cortex) are regulated. Amyloid plaques and hyperphosphorylated Tau were not observed 10 in mice of any genotype at any age ( Figure 5B).

12
Transcriptional profiling revealed Individual and synergistic effects of APOEe4, Trem2*R47H, and 13 age 14 Brain hemispheres from 4 and 24 month male and female B6.APOE4.Trem2*R47H mice and single 15 genotype and C57BL/6J controls were assessed using RNA-seq (see Methods). Transcriptomic analysis 16 measured the expression levels (log-transformed TPM counts) of mouse Apoe, Trem2, and human APOE 17 genes across all mouse models (Figure 6 A-C). We observed higher expression of human APOE gene in 18 mice carrying humanized APOEe4 (B6.APOE4 and B6.APOE4.Trem2*R47H mice), whereas mouse Apoe 19 gene was highly expressed in B6 and Trem2*R47H mice (Figure 6 A,B). As expected based on protein 20 levels (Supplemental Figure 7 A,B), expression of Trem2 was significantly reduced (p< 0.05) in 21 B6.Trem2*R47H and B6.APOE4.Trem2*R47H compared to age-matched B6 (Figure 6C), an effect likely 22 caused by a novel effector splice site and truncation introduced by the R47H mutation. Furthermore, 23 expression level of Trem2 increased with age across all mouse models, but no such patterns were observed 24 in the expression levels of mouse Apoe and human APOE genes (Figure 6 A,B). In addition, there was 25 lower expression of Trem2 in B6.APOE4.Trem2*R47H compared to B6.Trem2*R47H mice at advanced 26 age (24 months), suggesting expression of Trem2 might be suppressed by APOEe4. Next, principal 27 component analysis (PCA) identified two distinct clusters corresponding to male and female samples 28 separated along the first principal component (26% of total variance), suggesting sex-specific differences 1 are profound in mice ( Figure 6D). Analysis of samples from different age groups revealed a gradient of 2 discrimination along the second principal component (14% of total variance) ( Figure 6D), implying the 3 presence of age-dependent molecular changes in the brain transcriptomes.

13
In 8 months old B6.APOE4 mice, a total of 145 genes were significantly differentially expressed (42 14 upregulated, 103 downregulated) (p < 0.05) in male mice, whereas a total of 25 genes were differentially    At 24 months old, there were a total of 144 DEGs (15 upregulated, 129  8 downregulated) compared to at younger ages (4-12 months) (Supplemental Figure 10), suggesting that 9 the most dramatic transcriptional changes arise between 12 and 24 months. We also found substantial 10 overlap in downregulated DEGs between B6.Trem2*R47H and B6.APOE4.Trem2*R47H mice for both 11 sexes (Supplemental Figure 10). This suggests that the Trem2*R47H allele is the major driving force in    28 downregulated) were identified in the running B6.APOE4.Trem2*R47H mice compared to 24 months old 1 B6 male mice. Enrichment analysis identified multiple enriched KEGG pathways such as "oxidative 2 phosphorylation", "thermogenesis", and "retrograde endocannabinoid signaling" in the upregulated list of 3 genes, whereas immune system associated pathways were enriched in the downregulated list of genes 4 (Supplemental Figure 11A). To identify the effect of exercise, running B6.APOE4.Trem2*R47H male mice 5 were compared with sedentary 24 month old B6.APOE4.Trem2*R47H male mice and there was a total of 6 600 DEGs (312 upregulated, and 288 downregulated). Upregulated DEGs were enriched in pathways such 7 as "oxidative phosphorylation" and "Ribosome" (Supplemental Figure 11A). The expression of these 8 upregulated DEGs enriched for oxidative phosphorylation showed reduced expression in age-and sex-9 matched B6.APOE4 and B6.Trem2*R47H compared to running B6.APOE4.Trem2*R47H mice 10 (Supplemental Figure 11B). Finally, the expression of the upregulated DEGs associated with oxidative 11 phosphorylation pathway were assessed in transcriptional data from AMP-AD. Reduced expression of 12 these running signature genes was observed in AD cases compared to controls across multiple brain 13 regions such as parahippocampal gyrus (PHG) and frontal pole brain regions (FP) (Supplemental Figure   14   11C). This suggests that exercise induces beneficial effects on health by increasing the expression of  and tangles in the brain and correlating hyperactivity which is a confound of many cognitive behaviors.

28
However, LOAD is ~20x more prevalent than EOAD and further implicates aging, inflammation, 1 environmental, and many more genetic risk factors in disease development. Our APOE4.Trem2*R47H mice 2 were healthy late into life allowing a better understanding of the effect of AD risk factors in the context of 3 aging (Supplemental Figure 5). As the heterogeneity of this disease becomes more appreciated, so is the 4 importance of appropriate disease staging. Molecular targets of interest may only be available during 5 particular evolving disease stages: debris (cell fragments, plaques, tangles, etc.) accumulates over time 6 and inflammation, interruption/loss of neuronal function all also change with disease progression. In light of 7 the repeated short-comings of "fit-for-all" therapies, efforts may be better directed at targeted therapies (19, 8 49, 50). Faithfully modeling a complex, polygenic disease will be aided by the creation of platform strains 9 that carry multiple genetic risk factors to motivate with a scientific rationale rather than a grant-focused one.

10
Of the candidate risk variants identified, expression of the e4 allele of APOE and the R47H mutation in 11 TREM2 were identified as the strongest candidates for initial development of a novel LOAD mouse strain.

12
Introduction of the R47H mutation into Trem2 resulted in the creation of a novel murine splice site yielding 13 a decrease in approximately 50% of TREM2 protein (Supplemental Figure 8 A,B), and 20% decrease in Trem2 allele expressing the full-length R47H risk factor at levels similar to wild-type Trem2. APOEe4 is 18 strongly associated with disease development and severity (15,54) and at least one allele is present in 19 approximately 65% of AD patients (55). Unfortunately, endogenous Apoe in mice does not express the 20 isoform diversity seen from the APOE allele in humans. Insertion of humanized APOE alleles into mouse 21 genomes has been a successful strategy to dissect the biology of APOE isoforms in mice (17-20, 56). The

22
MODEL-AD APOEe3 and APOEe4 allelic series on a C57BL/6J background has subsequently be shown to 23 consistently reproduce the biology presented in other APOE mouse models and most importantly, human 24 patients (14, 56). Therefore, APOE4 formed the basis for multiple platform strains that include: 25 B6.APOE4.Trem2*R47H (for which we have provided the identifier 'LOAD1').

26
APOE binds to high-density lipoproteins to facilitate cholesterol and phospholipid transport to LDL  Table 2). We observed an age-dependent decrease 1 in glucose levels across both sexes and all genotypes ( Figure 1L), an indication of increased frailty and 2 aging that is common in aging-related-disease studies (57). The contributions of metabolic and vascular 3 factors are strongly implicated in disease progression, and how APOE or any other metabolic trait, via 4 systemic pathways, can influence CNS function should be a continued focus for intervention (58). The brain 5 has one of the richest networks of blood vessels and are especially vulnerable (block or reduce blood flow, 6 oxygen and nutrients). For example, the respective influence of APOEe4, with and without Trem2*R47H 7 expression, in regional changes in brain glycolytic metabolism (59) (Figure 3

11
Despite the heterogeneity of AD, age is the strongest risk factor in the human population. As an 12 aging disease, monitoring and evaluating mouse models in relation to age is crucial for understanding onset

3
In the absence of additional environmental or genetic risk factors, B6.APOE4/Trem2*R47H mice 4 did not display penetrant behavioral phenotypes beyond the expected aging-related changes but did exhibit 5 decreased survival probabilities by 24 months (Supplemental Figure 5). Very few C57BL/6J mice 6 succumbed during the 24-month aging process (<5%), whereas mortality was higher in mice expressing 7 both LOAD alleles in both males (~20%) and females (~35%). Male mice expressing either allele alone had 8 survival probabilities similar to C57BL/6J, whereas females with APOE 4 or Trem2*R47H showed a 9 mortality rate of ~20%. Therefore, it would seem that these two LOAD risk alleles show an equal and 10 additive risk when expressed together, but in females their interaction appears synergistic.

11
We investigated the molecular signatures in the brain transcriptomes of LOAD mouse models at 12 different ages in both sexes. We identified age-dependent molecular changes associated with LOAD 13 pathologies in mouse models. Introduction of the R47H mutation revealed a novel Trem2 isoform identical 17 expression, mouse models carrying R47H mutation in Trem2 gene did not exhibit any significant 18 transcriptional changes at young age, in contrast APOE4 mice exhibited significant changes only at 8 19 months of age. We further identified significant downregulation of genes associated with oxidative 20 phosphorylation pathway in the 12 months old B6.Trem2*R47H mice, suggesting that the oxidative 21 phosphorylation could be prominent early feature for the onset of neurodegeneration/inflammation process.

24
Interestingly, at 12 months of age we did not observe any significant transcriptional changes in 25 B6.APOE4.Trem2*R47H mice compared to control mice, suggesting that the effect of Trem2 gene is 26 suppressed due to the presence of APOEe4. Similarly, when mouse models were compared with human 27 co-expression modules, we observed strong negative correlation between the B6.Trem2*R47H mice and 28 immune-related human co-expression modules from multiple brain regions, and this inflammatory response 1 is dampened in the presence of APOEe4 in the B6.APOE4.Trem2*R47H mice. Distinct mouse models 2 showed concordance with distinct human co-expression modules reflecting a different transcriptional 3 response driven by the human APOEe4 and Trem2*R47H risk variants. We also observed age dependent 4 shift in co-expression patterns associated with LOAD pathologies. A strong negative correlation between 5 co-expression modules associated with cell cycle and DNA repair was observed in the early-aged mouse 6 B6.APOE4 model, whereas advanced-aged B6.APOE4 female mice showed strong positive correlation 7 with these co-expression modules. This overlap with human late-onset co-expression signatures early in 8 life was observed for a number of different brain regions and was absent in Trem2*R47H knock-in mice.

9
Furthermore, aged B6.Trem2*R47H mice showed a moderate overlap with several human neuronal co-10 expression modules enriched for genes that play an important role in synaptic signaling and myelination.

11
At advanced age, a strong correlation between the mouse models and immune related human co-

17
APOE 4 carriers don't have increased AD risk with Trem2*R47H and Trem2*R47H only increases risk on 18 APOE 3 carriers (65, 66). Additionally, we observe more differentially expressed genes at middle age than 19 at a later age supporting evidence of an earlier aging phenotype than C57BL/6J mice, with a realignment 20 of transcriptomes at later timepoints (58, 67). We employed a weighted gene co-expression network 21 analysis (WGCNA) used to identify modules of correlated genes. Each module was tested for differential 22 expression by strain, then compared with human postmortem brain modules from the Accelerating    animals. In subsequent studies utilizing new mouse strains, the utility of the APOE4.Trem2*R47H datasets 27 will grow. Ultimately, strains carrying combinations of risk factors that more closely align with human 28 disease will be incorporated into the pre-clinical testing core of MODEL-AD to assess the potential of 1 prioritized compounds to treat AD.

19
weight was recorded and upon production, fecal sample was collected with forceps to prevent 20 contamination. Sample was placed in a pre-marked 1.5mL tube and snap-frozen immediately on dry ice.

21
Container and forceps were cleaned with 70% ethanol before collecting from subsequent mice. Fecal

23
attached to a 1mL syringe, is inserted into the right atrium of the exposed heart and the plunger gently

13
end forceps, the two skull plates were removed to expose the brain. The brain was carefully removed from 14 the skull, weighed, and divided midsagitally, into left and right hemispheres, using a brain matrix.

10
To provide high contrast grey matter images, at least two days prior to PET imaging, mice were induced 11 with 5% isoflurane (balance medical oxygen), placed on the head coil, and anesthesia maintained with 1-

13
Siemens Prisma clinical MRI scanner outfitted with a dedicated 4 channel mouse head coil and bed system 14 (RapidMR, Columbus OH). Images were acquired using a SPACE3D sequence (80) using the following

15
Image Analysis

16
All PET and MRI images were co-registered using a ridged-body mutual information-based normalized 17 entropy algorithm (87) with 9 degrees of freedom, and mapped to stereotactic mouse brain coordinates   1 Gels were transferred to PVDF membranes for immunoblotting and imaging using an iBlot2 dry blotting 2 system (Thermo Fisher). Membranes were blocked in 5% non-fat dry milk in 1xPBS+0.1% Tween20 for 1 3 hour prior to incubating with primary antibodies diluted in 5% non-fat dry milk in 1xPBS+0.1% Tween20 for 4 1 hour at room temperature. Membranes were washed in 1xPBS+0.1% Tween20 before incubating with 5 secondary antibodies diluted in 5% non-fat dry milk in 1xPBS+0.1% Tween20. HRP-conjugated secondary 6 antibodies targeting primary antibody host IgG were incubated at 1 hour at room temperature. Membranes 7 were washed in 1xPBS+0.1% Tween20 before digital imaging with SuperSignal West Pico PLUS 8 chemiluminescent substrate (34579, Thermo Fisher). Images for immunoblot were quantified using ImageJ  Total RNA was extracted from snap frozen right brain hemispheres using Trizol (Invitrogen, Carlsbad, CA).

19
Subsequently, we added gene annotation of human APOE gene into mouse gene annotation file.

20
Additionally, we have also introduced annotation for novel Trem2 isoform in mouse gene annotation file

21
(GTF file), that is identical to primary transcript, but truncated exon2 by 119 bp from its start position.

22
Afterwards, a STAR index was built for this chimeric mouse genome sequence for alignment, then STAR 23 aligner output coordinate-sorted BAM files for each sample mapped to chimeric mouse genome using this    were selected based on a specific enrichment in LOAD cases when compared to controls 5 (10.7303/syn11914606). Finally, multiple co-expression module algorithms were used to identify a set of 6 30 aggregate modules that were replicated by the independent methods (6).

8
Mouse-Human correlation analysis 9 First, we performed differential gene expression analysis for each mouse model compared to age and sex-10 matched B6 control mice using the limma (95)                          Variations in regional glucose metabolism due to expression of APOEε4 and Trem2*R47H. Positron emission tomography (PET; red scale) of radioactive 18F-FDG marker was used to measure tissue glucose uptake, guided by magnetic resonance imaging (MRI) (black and white) mapping to brain regions of interest, indicated by bregma coordinates (far left) (A). Intensity of PET signal in brains regions, normalized to cerebellum, are quanti ed in B,C. Post-mortem autoradiography of coronal brain tissue is represented in A (rainbow; far right). Genotype-dependent differences determined by ANOVA: *p<0.05; **p<0.01; ***p<0.001. All alleles expressed were homozygous.

Figure 4
In vivo neuroimaging reveals differences in regional perfusion driven by expression of risk alleles. Positron emission tomography (PET; red scale) of radioactive 64Cu-PTSM marker was used to measure tissue perfusion, guided by magnetic resonance imaging (MRI) (black and white) mapping to brain regions of interest, indicated by bregma coordinates (far left) (A). Intensity of PET signal in brains regions, normalized to cerebellum, are quanti ed in B,C. Post-mortem autoradiography of coronal brain tissue is represented in A (rainbow; far right). Genotype-dependent differences determined by ANOVA: *p<0.05; **p<0.01; ***p<0.001. All alleles expressed were homozygous.   KEGG pathways enrichment analysis. Signi cantly enriched KEGG pathways (p < 0.05) in the downregulated and upregulated list of genes across mouse models at different ages for both sexes. All alleles expressed were homozygous.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.