Gene Expression Profiling in the APP/PS1KI Mouse Model of Familial Alzheimer’s Disease

Abstract

Alzheimer’s disease (AD) is a devastating neurodegenerative disorder characterized by early intraneuronal amyloid-β (Aβ) accumulation, extracellular deposition of Aβ peptides, and intracellular hyperphosphorylated tau aggregates. These lesions cause dendritic and synaptic alterations and induce an inflammatory response in the diseased brain. Although the neuropathological characteristics of AD have been known for decades, the molecular mechanisms causing the disease are still under investigation. Studying gene expression changes in postmortem AD brain tissue can yield new insights into the molecular disease mechanisms. To that end, one can employ transgenic AD mouse models and the next-generation sequencing technology. In this study, a whole-brain transcriptome analysis was carried out using the well-characterized APP/PS1KI mouse model for AD. These mice display a robust phenotype reflected by working memory deficits at 6 months of age, a significant neuron loss in a variety of brain areas including the CA1 region of the hippocampus and a severe amyloid pathology. Based on deep sequencing, differentially expressed genes (DEGs) between 6-month-old WT or PS1KI and APP/PS1KI were identified and verified by qRT-PCR. Compared to WT mice, 250 DEGs were found in APP/PS1KI mice, while 186 DEGs could be found compared to PS1KI control mice. Most of the DEGs were upregulated in APP/PS1KI mice and belong to either inflammation-associated pathways or lysosomal activation, which is likely due to the robust intraneuronal accumulation of Aβ in this mouse model. Our comprehensive brain transcriptome study further highlights APP/PS1KI mice as a valuable model for AD, covering molecular inflammatory and immune responses.

Keywords

Alzheimer’s disease amyloid differential gene expression next-generation sequencing transcriptome transgenic mice

INTRODUCTION

Alzheimer’s disease (AD) is the most prominent form of dementia and is associated with a progressive accumulation of extracellular amyloid plaques and intracellular neurofibrillary tangles. While plaques mainly consist of amyloid-β (Aβ) peptides derived from proteolytic processing of the amyloid-β protein precursor (AβPP), neurofibrillary tangles are composed of hyperphosphorylated tau protein [1]. In recent years, a variety of transgenic AD mouse models have been described which differ in the extent of pathology and the presence or severity of behavioral deficits. Many neuropathological features like amyloid deposition, inflammatory changes, or tau phosphorylation have been successfully modeled; however, overt neuron loss is still mostly lacking [2]. The APP/PS1KI mouse model used in the present study has been shown to harbor significant neuron loss in the CA1 region of the hippocampus [3, 4], in the frontal cortex [5], as well as in the cholinergic system [6] already at 6 months of age. In addition, it also suffers from robust deficits in working memory [7] and disturbed long-term potentiation [3].

In recent years, several studies on the expression profile of postmortem AD brain tissue and transgenic AD mouse models were carried out, applying diverse technological approaches [8 –17]. Earlier studies applied hybridization-based approaches using micro-arrays. Such studies mainly use fluorescently labeled cDNA, are relatively easy to perform, and rather inexpensive. On the other hand, they have several limitations including the risk of cross-hybridization and a relatively narrow dynamic range due to background noise or signal saturation. In contrast, sequencing-based techniques, e.g., Sanger sequencing of cDNA libraries, are laborious, rather expensive, and generally not quantitative [18]. Next-generation sequencing technology (“RNA-Seq”) bears several advantages, as it allows for high-throughput analysis of transcription profiles at single base resolution [18]. We already successfully applied this technique in expression profiling studies in AD mouse models [9].

Here, we carried out a whole-brain transcriptome study using the APP/PS1KI mouse model showing robust and widespread pathology in comparison to age-matched PS1KI and wildtype (WT) mice.

MATERIAL AND METHODS

Transgenic mice

APP/PS1KI mice overexpress human amyloid precursor protein (APP751) carrying the Swedish and London mutations in addition to murine presenilin-1 (PSEN-1) carrying the M233T/L235P mutations on a homozygous knock-in background [4, 19]. Mutant murine PSEN1 is expressed under the control of the endogenous promoter, leading to a lack of endogenous WT PSEN-1. AβPP is expressed under the control of the neuronal Thy1.2-promoter. Only female mice were used in this study.

Tissue harvesting

Mice were sacrificed via CO₂ anesthetization followed by cervical dislocation. Brain hemispheres were carefully dissected (the cerebellum was removed), frozen on dry-ice, and stored at –80°C for subsequent use.

RNA expression profiling

Expression profiling for 6-month-old APP/PS1KI, PSKI, and WT mice was performed by next-generation sequencing on a SOLiD 5500xl Genetic Analyzer (Life Technologies, Carlsbad, CA, USA). RNA extraction from individual mouse brain hemispheres was carried out as follows. The tissue was homogenized using a Polytron (VWR) homogenizer and subsequently treated with TRIzol (Life Technologies). Next, 5 μg of each total RNA sample were spiked with ERCC spike-in control mixes (Life Technologies), followed by removal of rRNA by use of a RiboZero Kit(Epicentre, Madison, WI, USA). The RNA was prepared for sequencing following the protocol provided by the manufacturer of the sequencer. In brief, the rRNA depleted RNA samples were fragmented by chemical hydrolysis, phosphorylated, and purified. Prior reverse transcriptions into cDNA, adaptors were ligated to the RNA fragments. The cDNA was purified and size-selected using two rounds of Agencourt AMPure XP bead purification (Beckman Coulters Genomics, Danvers, USA) and released from the beads. The sample was amplified by 12 PCR cycles in the presence of primers that contained unique sequences (barcoding). The size distribution and concentration of the fragments were determined using an Agilent 2100 Bioanalyzer and the corresponding chemicals (Agilent Technologies, Santa Clara, USA).

The cDNA fragments were pooled in equimolar amounts and diluted to 76 pg/ μl corresponding to a concentration of 500 pM. A 50 μl aliquot of this dilution was mixed with a freshly prepared oil emulsion, P1 and P2 reagents, and P1 beads in a SOLiD EZ Bead Emulsifier prepared according to the E80 scale protocol (Life Technologies). Emulsion PCR was carried out in a SOLiD EZ Bead Amplifier (Life Technologies) using the E80 setting. To enrich for beads carrying amplified template DNA, the beads were purified on a SOLiD EZ Bead Enricher using the recommended chemicals and software (Life Technologies). The purified beads were loaded onto a SOLiD 6-lane Flowchip and incubated upside down for 1 h at 37°C. The Flowchip was positioned in the 5500xl SOLiD System and the DNA was sequenced using the settings and recommended chemicals for sequencing 75 nucleotides in the forward direction and 35 nucleotides in the reverse direction (Life Technologies).

Sequence reads were mapped to the mouse genome reference sequence mm10 (ftp://hgdownload.cse.ucsc.edu/goldenPath/mm10/) using the workflow ‘whole.transcriptome.pe’ LifeScope-v2.5.1-r0 (Life Technologies). Reads mapping to RefSeq coding exons (http://hgdownload.cse.ucsc.edu/goldenPath/mm10/database/refGene.txt.gz, accessed 2012-06-27) and matching the coding strand were considered coding RNAs. All other mapping reads were considered non-coding.

Differential expression analysis

To normalize for GC-content within and between lanes, full-quantile normalization as implemented in the EDASeq R package [20] was performed. Dispersion was estimated with the help of the DESeq R package, treating all samples as replicates of a single condition. Following analyses were based on fitted dispersion values. Two conditions were compared with a negative binomial test. Differential expression was deemed significant if the Benjamini-Hochberg corrected p-value was smaller than 0.05. One hemisphere per time point and condition was used. The following conditions were compared: APP/PS1KI versus PS1KI, APP/PS1KI versus WT, and PS1KI versus WT.

Overlap in the up- and downregulated genes between the different groups was visualized using BioVenn [24].

Real-time quantitative PCR (qRT-PCR) confirmation

RNA for qRT-PCR verification was isolated from additional cohorts of female 6-month-old APP/PS1KI, PS1KI, and WT mice (n = 5 each) as described previously [20]. Briefly, frozen right brain hemispheres were homogenized with 10 strokes of a R50D homogenizer (CAT) set at 800 rpm in 1.5 ml ice-cold Trifast^® (Peqlab, Erlangen, Germany). 300 μl Chloroform (Merck) was added to each sample. After 10 min incubation, the samples were centrifuged at 12,000 g for 15 min at 4°C to separate the RNA. The upper RNA-containing aqueous phase was transferred into a new tube, vigorously mixed with 500 μl Isopropanol and incubated for 20 min on ice. After centrifugation at 12,000 g for 10 min at 4°C, the supernatant was discarded. RNA pellets were washed twice with 500 μl 75 % ethanol. After the pellet was air-dried, the RNA was dissolved in 30 μl of RNAse free water. RNA was stored at –80°C until further use. RNA purity and yields were determined by a Biophotometer (Eppendorf, Hamburg, Germany).

Total RNA (1 μg) was subjected to reverse transcription to synthesize cDNA using the First Strand cDNA Synthesis Kit (Fermentas, St. Leon-Rot, Germany) according to the manufacturer’s instructions. Prior to reverse transcription, RNA was subjected to digestion by DNase using a DNase Digestion Kit (Fermentas). Generated cDNA was diluted 1:10 in ddH₂O and used as the sample template for qRT-PCR. The obtained cDNA was stored at –20°C until use.

Deep sequencing results were validated by qRT-PCR. Several genes were selected and primers were purchased from Eurofins (Ebersberg, Germany) as intron-spanning validated primer pairs. The diluted first-strand cDNA was used for qRT-PCR using the SYBR-green Fast Start Universal SYBR Green qPCR Kit (Roche, Penzberg, Germany) containing ROX as an internal reference dye. Samples were normalized to the housekeeping genes β-Actin and GAPDH.

Analysis of brain transgene expression in APP/PS1KI, PS1KI, and WT animals was performed in the MX3000P Real-Time Cycler (Stratagene, Santa Clara, CA, USA) and data collected using the MxPro Mx3000P software (Stratagene). Statistical analysis of qRT-PCR measurements was done using the Relative Expression Software Tool V1.9.6 (REST, Qiagen, Hilden, Germany) [21]. The expression ratio results of the studied transcripts were tested for significance by Pair Wise Fixed Reallocation Randomization Test.

Overrepresentation analysis of gene annotation

Functional annotation clustering was performed using the DAVID software tool [26] using all genes that were up- or downregulated in 6-month-old APP/PS1KI mice compared to age-matched PS1KI and WT littermate controls with a p-value <0.05. Annotations from the Gene Ontology (GO) [22] and the Kyoto Encyclopedia of Genes and Genomes (KEGG) [28] were tested for enrichment.

RESULTS

Deep sequencing of mouse brains

In total, deep sequencing identified 15,352,395 reads for APP/PS1KI, 20,057,929 reads for PS1KI, and 11,134,008 reads for WT mice. For the 6-month-old APP/PS1KI mice, 8,142,864 reads (53.04% ), for age-matched PS1KI 10,833,034 reads (54.01% ), and for WT mice 4,436,061 reads (39.84% ) were mapped to exons.

Deep sequencing identified differentially expressed transgenic sequences

AβPP is expressed under the control of the neuronal Thy1.2-promoter, while mutant murine PSEN1 is expressed at endogenous levels. As expected, sequence reads pertaining to AβPP and a Thy1.2-promoter sequence were over-represented in APP/PS1KI brains (log2-fold-change (l2FC) = 0.9, p = 7.9E-03 and l2FC = 2.56, p = 3.5E-24 respectively, Supplementary Table 1) and therefore served as positive and internal controls for RNA-sequencing efficiency.

Differential gene expression in 6-month-old APP/PS1KI, PS1KI, and WT mice

In 6-month-old APP/PS1KI mice, 185 genes were differentially expressed compared to PS1KI mice, with 150 genes upregulated and 35 downregulated (Fig. 1A and Supplementary Table 1). In comparison to WT mice, 250 genes were found to be differentially expressed in 6-month-old APP/PS1KI mice, with 175 being upregulated and 75 being downregulated (Fig. 1B and Supplementary Table 2). In APP/PS1KI mice, 126 genes were overexpressed compared to both PS1KI and WT, while 49 genes showed an upregulation only compared to the WT situation and 24 genes were only found upregulated compared to the PS1KI line (Fig. 2). The top-20 genes, with regard to fold-change, upregulated in APP/PS1KI compared to PS1KI and WT are shown in Table 1. APP/PS1KI mice showed downregulation of 29 genes compared to both WT and PS1KI, while only 6 genes showed a downregulation compared to PS1KI only. A larger pool of 46 genes was exclusively downregulated in APP/PS1KI mice compared to age-matched WT mice (Fig. 2). Interestingly, only 6 genes showed a differential expression pattern in PS1KI compared to WT mice, with 4 genes being overexpressed and 2 genes being downregulated (Fig. 1C and Supplementary Table 3).

Overrepresentation analysis

GO enrichment analysis of genes upregulated in APP/PS1KI compared to PS1KI, WT and both was performed using DAVID [23, 24]. Genes upregulated in 6-month-old APP/PS1KI versus PS1KI mice were significantly (Benjamini-Hochberg corrected p-value <0.05) enriched for 145 GO classes (Biological Process (BP): 127, Cellular Component (CC): 13, Molecular Function (MF): 5). Genes upregulated in APP/PS1KI versus WT mice were significantly enriched for 145 GO classes (BP: 131, CC: 12, MF: 2), as well. The 126 genes that were significantly upregulated in APP/PS1KI compared to both PS1KI and WT are involved in diverse biological mechanisms, with a large group of differentially expressed genes (DEGs) being involved in several processes related to the immune system and inflammatory response. According to the BP annotation, these include, e.g., “immune response”, “positive regulation of immune system process” or “inflammatory response” among others. In terms of the CC annotation, “lysosome” or “external side of plasma membrane” showed the strongest enrichment. Regarding the MF annotation, “protein complex binding” or “MHC protein binding” showed the highest relevance (Table 2).

No enriched functional annotations could be found for downregulated genes in 6-month-old APP/PS1KI mice.

Genes that were differentially regulated in APP/PS1KI mice compared to PS1KI were further subjected to an overrepresentation analysis of canonical pathways defined by KEGG [25]. A Benjamini-Hochberg corrected p-value of <0.05 was deemed significant. The most striking pathways were “lysosome” with 16 genes “complement and coagulation cascades” with 8.1-fold enrichment each. The other pathways showing significant gene enrichment were “systhemic lupus erythematosus”, “hematopoetic cell lineage”, “antigen processing and presentation”, “cell adhesion molecules”, “glycosaminoglycan degradation”, and “leukocyte transendothelial migration”For APP/PS1KI versus WT the following pathways were identified based on a significant Benjamini-Hochberg corrected p-value <0.05: “lysosome” (18 genes, 7.4-fold enrichment), “antigen processing and presentation”, “systemic lupus erythematosus”, and “complement and coagulation cascades”.

Real-Time PCR validation of differentially expressed genes identified by deep sequencing

Several DEGs with either high or low fold-change values were selected for qRT-PCR validation using APP/PS1KI versus PS1KI mice. The selection contained some genes with high l2FC like Cystatin F ((Cst7), 8.57, p = 9.0E-37), or medium to low values like glial fibrillary acidic protein ((GFAP), 3.26, p = 1.0E-35) or transforming growth factor, beta receptor II (Tgfbr2), 1.7, p = 3.8E-06). Some genes like Cathepsin D ((Ctsd), 1.44, 2.5E-07) or S100 calcium-binding protein A6 ((S100A6), 1.51, p = 1.3E-02) had already been identified to be upregulated in an earlier study [26] and were confirmed in the present analysis. The same holds true for genes that were downregulated in APP/PS1KI versus PS1KI mice. Here, also genes with higher l2FC like dopamine transporter ((Slc6a3), –4.32, p = 4.2E-17) or lower values like Apelin ((Apln), –1.56, p = 2.9E-04) or kinesin family member 1A ((Kif1a), –0.9, p = 7.9E-03) were selected and could be confirmed by qRT-PCR analysis, thereby validating the deep sequencing approach (Fig. 3).

DISCUSSION

At the age of 6 months, APP/PS1KI mice show a lot of different pathological alterations in a variety of different brain regions which make it one of the most “complete” amyloid-based models available so far. This time point of 6 months comprises a variety of pathological hallmarks including robust extracellular amyloid pathology [5, 27], loss of CA1 neurons and hippocampal atrophy [3, 28], loss of neurons in frontal cortex and in cholinergic brain stem nuclei [5, 6], as well as robust behavioral deficits in working memory and motor performance [7] or Morris water maze [28]. In order to reproduce the complexity of neuropathological and behavioral alterations at this time point, we decided to carry out a whole brain approach, although this might bear the risk that more subtle changes in local gene expression levels might not be detected. However, we cannot exclude that the genes identified in the APP/PS1KI mice are mimicking the situation in patients with Down syndrome, having a triplication of the APP-carrying chromosome 21, or that of AD patients with a microduplication of the APP gene.

Upregulated genes in APP/PS1KI mice

Inflammatory response

APP/PS1KI mice represent an aggressive model mainly reflecting familial AD showing a very robust and early-onset amyloidosis [4 , 27]. In line with previous results and data from other mouse models like 5XFAD, the majority of the genes identified to be upregulated in 6-month-old APP/PS1KI mice are implicated in the regulation of an inflammatory response [13, 26]. An upregulation of inflammation-associated transcripts has been reported in most APP-transgenic mouse models harboring extracellular Aβ deposition (e.g., [13 , 29–35]) and both astrocyte and microglial activation are intimately linked to human AD pathology (e.g., [36 –38]). Deep sequencing revealed an upregulation of GFAP, which could be confirmed by qRT-PCR, corroborating earlier studies in this [26] and other APP transgenic models [13 , 39–41]. Immunohistochemical stainings against GFAP also validated the upregulation on the protein level in the APP/PS1KI model [26]. The astrocytic calcium/zinc binding protein S100A6 has been previously reported to be upregulated in human AD patients, as well as in transgenic AD models [26, 42], which supports the assumption of astrocytes being actively involved in Aβ clearance. Members of the complement system (e.g., C1qa, C1qb, C1qc, C4a, C4b) are also strongly upregulated in 6-month-old APP/PS1KI mice. Complement activation in AD is regarded as an attempt to initiate clearance of toxic protein aggregates. It has been shown that Aβ represents a strong complement activator by binding to C1q, which leads to the classical complement activation pathway [43, 44]. SerpinA3 (coding for alpha-1-antichymotrypsin, ACT) showed a strong upregulation in both deep sequencing and qRT-PCR verification. ACT has been shown to promote Aβ fibrillization, correlates with progression of dementia in AD patients, and has been attributed to induce tau phosphorylation in disease-affected neurons [45].

Other upregulated transcripts include microglia-associated cytokines like colony-stimulating factor 1 (Csf1) and related receptors like Csf1r and Csf3r that represent key regulators of the monocyte/macrophage lineage. It has been demonstrated that Csf1 provides powerful neuroprotective and survival signals in brain injury and neurodegeneration and that systemic administration of human recombinant Csf1 ameliorates memory deficits in a transgenic AD mouse model [46]. A related study demonstrated that injection of CSF1 to APP/PS1 mice prior to the appearance of learning and memory deficits prevented cognitive loss while it increased the number of microglia in the parenchyma and decreased the number of Aβ deposits [47]. APP/PS1KI mice at 6 months of age exhibit profound neuron loss in the hippocampus [3] and frontal cortex [5], as well as massive extracellular amyloid pathology. It has been also shown that CSF1 is capable of lowering lysosomal pH, thereby facilitating the degradation of fibrillar Aβ by activated microglia [48]. Therefore, the observed upregulation of Csf1 and its receptor in the current setting might represent a compensatory mechanism.

Increased levels of the toll-like receptor family member Tlr2 were found as previously reported in the same model [26]. This receptor has been shown to mediate an inflammatory response towards aggregated Aβ₄₂ peptides [49]. Cultured microglial cells incubated with Tlr2 agonists showed a markedly boosted ingestion of Aβ in vitro, supporting a major role in Aβ clearance [50]. Transmembrane glycoprotein NMB (Gpnmb) is a type I transmembrane protein implicated in cell differentiation, inflammation, tissue regeneration, and tumor progression. So far it has not been implicated in AD; however, it has been shown that Gpnmb is expressed in microglial cells [51] and that it is able to reduce the secretion of proinflammatory cytokines from macrophages in vitro. This suggests that it might act on immune effector cells and might alleviate excessive proinflammatory responses in the central nervous system.

Lipoprotein lipase (Lpl) is a member of a lipase family known to hydrolyze triglyceride molecules found in lipoprotein particles. It has been assumed that Lpl plays a role in cholesterol and lipid recycling from degenerating terminals [53], which might explain its upregulation at 6 months of age, a time-point with profound neuron and synapse loss [3 , 6]. Recent data further indicates that Lpl binds to Aβ and promotes cell-surface association and uptake of Aβ peptides in mouse primary astrocytes, suggesting that Lpl is an important Aβ-binding protein promoting cellular uptake and subsequent Aβ degradation [54].

Lysosomal activation

Macrophage expressed gene 1 (Mpeg1) and Cyba, encoding for the light chain of cytochrome b-245, have been previously detected to be upregulated in mouse models of lysosomal storage disorders [55], as well as in progranulin-deficient mice following traumatic brain injury [56].

One of the strongest upregulated genes in both RNA sequencing and qRT-PCR verification was Cst7 (coding for Cystatin F). It represents a papain-like lysosomal cysteine proteinase inhibitor selectively expressed in immune cells [57] that was identified as upregulated after lipopolysaccharide (LPS) stimulation of monocyte-derived dendritic cells [58]. After its synthesis, Cystatin F is translocated to the endosomal/lysosomal system and is able to regulate cathepsin activity in these vesicles [59]. In good agreement, a variety of different cathepsins, including Cathepsin B (Ctsb), Cathepsin E (Ctse), Cathepsin S (Ctss), Cathepsin Z (Ctsz), Cathepsin H (Ctsh), or Cathepsin D (Ctsd) was also found to be upregulated in APP/PS1KI compared to PS1KI mice. In comparison to WT mice, Cathepsin A (Ctsa), Cathepsin C (Ctsc), and Cathepsin L (Ctsl) were upregulated in addition to the aforementioned cathepsins. All found cathepsins represent endoproteases that are expressed in lysosomal compartments. Cathepsins also play an important role in autophagic processes in AD [60]. It has been demonstrated that, for example, Cathepsin B and D rise in the early stages of AD in Rab-positive endosomes [61]. In the brain, Aβ generated in autophagic vacuoles is delivered to lysosomes and degraded by Cathepsins. A deletion of Cathepsin B leads to elevated brain Aβ levels while increasing Cathepsin B expression has the opposite effect [62]. Ctss is particular interesting, as it has been shown recently that Cathepsin S is also expressed by cortical microglia in a circadian fashion. A genetic deletion of Ctss causes mice to exhibit hyperlocomotor activity and removes diurnal variations in the synaptic activity and spine density of cortical neurons. This suggests that Ctss secreted by microglia during the dark-phase decreases the spine density of the cortical neurons by modifying the perisynaptic environment [63]. Thus, an overexpression of Ctss might contribute to reduced synaptic spine density, which is a common phenotype in AD mouse models [64]. Cathepsin E has been previously demonstrated by immunohistochemistry to be abundant in microvessels, microglia and senile plaques in AD patients while it was only detectable in low amounts in normal brain [65]. Other upregulated genes like lysozyme 2 (Lyz2), Hexosaminidase A and B (HexA, HexB), Glucuronidase B (GusB), lysosomal-associated protein transmembrane 5 (Laptm5) or N-Acetylglucosaminidase alpha (Naglu) underscore the increased lysosomal activity.

Downregulated genes in APP/PS1KI

In comparison to PS1KI mice, 35 downregulated genes were identified in APP/PS1KI mice, while 75 genes were downregulated compared to WT mice. Among these genes, a variety of ion transporters like the dopamine transporter DAT (Slc6a3) or the vesicular monoamine transporter 2 (VMAT2, Slc18a2), as well as sodium/potassium cation transport ATPases, like Atp1a1 and Atp1a2 were detected. One of the most strongly downregulated genes identified in APP/PS1KI mice compared to both PS1KI as well as WT is endothelial lipase (Lipg). It has been suggested that lysophosphatidylcholine (lysoPC) is a preferred carrier of docosahexaenoic acid (DHA), an essential fatty acid needed for normal brain function. Endothelial lipase is secreted by cells of the blood-brain barrier and thought to play an important role in the delivery of DHA phospholipid carriers to the brain [66]. The downregulated neuropeptide ligand apelin (Apln) has been implicated in the promotion of neuronal survival by activating pro-survival signaling in addition to inhibition of NMDA receptor-mediated excitotoxicity signaling cascades [67]. The decreased expression level of the immediate early gene Arc might reflect the neuron loss in the CA1 region in this model to a certain extent and it correlates with working memory deficits [7] and impaired long-term potentiation [3]. Reduced levels of kinesin family member 1A (Kif1a), the primary anterograde motor protein required for axonal transport, have been previously described in older APP/PS1KI mice [68] and correlate with axonal degeneration in this model [69].

In summary, we provide comprehensive insight into the transcriptome of 6-month-old APP/PS1KI mice. Due to the presence of early amyloid pathology, overt neuron loss and behavioral deficits, these mice represent one of the most complete AD models. However, it should be considered that due to the combination of several APP and PSEN1 mutations, APP/PS1KI mice do not entirely reflect the pathophysiological changes in sporadic AD, but more closely mimic the situation in familial AD. The transcriptional profiling data presented in this study further emphasizes APP/PS1KI mice being a valuable model to understand the involvement of inflammatory and immune pathways in AD.

Footnotes

ACKNOWLEDGMENTS

TK’s position was funded by the Federal Ministry of Cultural Affairs of the Federal State of Mecklenburg-West Pomerania through the GANI_MED (Greifswald Approach to Individualized Medicine) project (03IS2061A). This work was supported by the Alzheimer Forschung Initiative e.V. (grant number #12802, to O.W.).

Authors’ disclosures available online ().

Appendix

References

Querfurth

, LaFerla

(2010) Alzheimer’s disease. New Engl J Med 362, 329–344.

Wirths

, Bayer

(2012) Intraneuronal Abeta accumulation and neurodegeneration: Lessons from transgenic models. Life Sci 91, 1148–1152.

Breyhan

, Wirths

, Duan

, Marcello

, Rettig

, Bayer

(2009) APP/PS1KI bigenic mice develop early synaptic deficits and hippocampus atrophy. Acta Neuropathol 117, 677–685.

Casas

, Sergeant

, Itier

, Blanchard

, Wirths

, van der Kolk

, Vingtdeux

, van de Steeg

, Ret

, Canton

, Drobecq

, Clark

, Bonici

, Delacourte

, Benavides

, Schmitz

, Tremp

, Bayer

, Benoit

, Pradier

(2004) Massive CA1/2 neuronal loss with intraneuronal and N-terminal truncated Abeta42 accumulation in a novel Alzheimer transgenic model. Am J Pathol 165, 1289–1300.

Christensen

, Kraus

, Flohr

, Cotel

, Wirths

, Bayer

(2008) Transient intraneuronal Abeta rather than extracellular plaque pathology correlates with neuron loss in the frontal cortex of APP/PS1KI mice. Acta Neuropathol 116, 647–655.

Christensen

, Bayer

, Wirths

(2010) Intracellular Abeta triggers neuron loss in the cholinergic system of the APP/PS1KI mouse model of Alzheimer’s disease. Neurobiol Aging 31, 1153–1163.

Wirths

, Breyhan

, Schafer

, Roth

, Bayer

(2008) Deficits in working memory and motor performance in the APP/PS1ki mouse model for Alzheimer’s disease. Neurobiol Aging 29, 891–901.

Bossers

, Wirz

, Meerhoff

, Essing

, van Dongen

, Houba

, Kruse

, Verhaagen

, Swaab

(2010) Concerted changes in transcripts in the prefrontal cortex precede neuropathology in Alzheimer’s disease. Brain 133, 3699–3723.

Bouter

, Kacprowski

, Weissmann

, Dietrich

, Borgers

, Brauss

, Sperling

, Wirths

, Albrecht

, Jensen

, Kuss

, Bayer

(2014) Deciphering the molecular profile of plaques, memory decline and neuron loss in two mouse models for Alzheimer’s disease by deep sequencing. Front Aging Neurosci 6, 75.

10.

Chen

, Tian

, Liang

, Sun

, Dai

, Lee

, LaFerla

, Grundke-Iqbal

, Iqbal

, Liu

, Gong

(2012) Brain gene expression of a sporadic (icv-STZ Mouse) and a familial mouse model (3xTg-AD mouse) of Alzheimer’s disease. PLoS One 7, e51432.

11.

Gatta

, D’Aurora

, Granzotto

, Stuppia

, Sensi

(2014) Early and sustained altered expression of aging-related genes in young 3xTg-AD mice. Cell Death Dis 5, e1054.

12.

Kim

, Moon

, Yu

, Mook-Jung

, Kim

(2012) RNA-Seq analysis of frontal cortex and cerebellum from 5XFAD mice at early stage of disease pathology. J Alzheimers Dis 29, 793–808.

13.

Landel

, Baranger

, Virard

, Loriod

, Khrestchatisky

, Rivera

, Benech

, Feron

(2014) Temporal gene profiling of the 5XFAD transgenic mouse model highlights the importance of microglial activation in Alzheimer’s disease. Mol Neurodegener 9, 33.

14.

Ray

, Ruan

, Zhang

(2008) Variations in the transcriptome of Alzheimer’s disease reveal molecular networks involved in cardiovascular diseases. Genome Biol 9, R148.

15.

Reddy

, McWeeney

, Park

, Manczak

, Gutala

, Partovi

, Jung

, Yau

, Searles

, Mori

, Quinn

(2004) Gene expression profiles of transcripts in amyloid precursor protein transgenic mice: Up-regulation of mitochondrial metabolism and apoptotic genes is an early cellular change in Alzheimer’s disease. Hum Mol Genet 13, 1225–1240.

16.

Twine

, Janitz

, Wilkins

, Janitz

(2011) Whole transcriptome sequencing reveals gene expression and splicing differences in brain regions affected by Alzheimer’s disease. PLoS One 6, e16266.

17.

Wirz

KTS

, Bossers

, Stargardt

, Kamphuis

, Swaab

, Hol

, Verhaagen

(2013) Cortical beta amyloid protein triggers an immune response, but no synaptic changes in the APPswe/PS1dE9 Alzheimer’s disease mouse model. Neurobiol Aging 34, 1328–1342.

18.

Wang

, Gerstein

, Snyder

(2009) RNA-Seq: A revolutionary tool for transcriptomics. Nat Rev Genet 10, 57–63.

19.

Bayer

, Wirths

(2008) Review on the APP/PS1KI mouse model: Intraneuronal Abeta accumulation triggers axonopathy, neuron loss and working memory impairment. Genes Brain Behav 7(Suppl 1), 6–11.

20.

Huttenrauch

, Baches

, Gerth

, Bayer

, Weggen

, Wirths

(2015) Neprilysin deficiency alters the neuropathological and behavioral phenotype in the 5XFAD mouse model of Alzheimer’s disease. J Alzheimers Dis 44, 1291–1302.

21.

Pfaffl

, Horgan

, Dempfle

(2002) Relative expression software tool (REST) for group-wise comparison and statistical analysis of relative expression results in real-time PCR. Nucleic Acids Res 30, e36.

22.

Ashburner

, Ball

, Blake

, Botstein

, Butler

, Cherry

, Davis

, Dolinski

, Dwight

, Eppig

, Harris

, Hill

, Issel-Tarver

, Kasarskis

, Lewis

, Matese

, Richardson

, Ringwald

, Rubin

, Sherlock

(2000) Gene Ontology: Tool for the unification of biology. Nat Genet 25, 25–29.

23.

Huang

, Sherman

, Lempicki

(2009) Bioinformatics enrichment tools: Paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 37, 1–13.

24.

Huang

, Sherman

, Lempicki

(2009) Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 4, 44–57.

25.

Kanehisa

, Goto

(2000) KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res 28, 27–30.

26.

Wirths

, Breyhan

, Marcello

, Cotel

, Bruck

, Bayer

(2010) Inflammatory changes are tightly associated with neurodegeneration in the brain and spinal cord of the APP/PS1KI mouse model of Alzheimer’s disease. Neurobiol Aging 31, 747–757.

27.

Wirths

, Bethge

, Marcello

, Harmeier

, Jawhar

, Lucassen

, Multhaup

, Brody

, Esparza

, Ingelsson

, Kalimo

, Lannfelt

, Bayer

(2010) Pyroglutamate Abeta pathology in APP/PS1KI mice, sporadic and familial Alzheimer’s disease cases. J Neural Transm 117, 85–96.

28.

Faure

, Verret

, Bozon

, El Tannir El Tayara

, Ly

, Kober

, Dhenain

, Rampon

, Delatour

(2011) Impaired neurogenesis, neuronal loss, and brain functional deficits in the APPxPS1-Ki mouse model of Alzheimer’s disease. Neurobiol Aging 95, 92–101.

29.

Apelt

, Schliebs

(2001) Beta-amyloid-induced glial expression of both pro- and anti-inflammatory cytokines in cerebral cortex of aged transgenic Tg2576 mice with Alzheimer plaque pathology. Brain Res 894, 21–30.

30.

Bornemann

, Wiederhold

, Pauli

, Ermini

, Stalder

, Schnell

, Sommer

, Jucker

, Staufenbiel

(2001) Abeta-induced inflammatory processes in microglia cells of APP23 transgenic mice. Am J Pathol 158, 63–73.

31.

Dudal

, Krzywkowski

, Paquette

, Morissette

, Lacombe

, Tremblay

, Gervais

(2004) Inflammation occurs early during the Abeta deposition process in TgCRND8 mice. Neurobiol Aging 25, 861–871.

32.

Frautschy

, Yang

, Irrizarry

, Hyman

, Saido

, Hsiao

, Cole

(1998) Microglial response to amyloid plaques in APPsw transgenic mice. Am J Pathol 152, 307–317.

33.

Jimenez

, Baglietto-Vargas

, Caballero

, Moreno-Gonzalez

, Torres

, Sanchez-Varo

, Ruano

, Vizuete

, Gutierrez

, Vitorica

(2008) Inflammatory response in the hippocampus of PS1M146L/APP751SL mouse model of Alzheimer’s disease: Age-dependent switch in the microglial phenotype from alternative to classic. J Neurosci 28, 11650–11661.

34.

Matsuoka

, Picciano

, Malester

, LaFrancois

, Zehr

, Daeschner

, Olschowka

, Fonseca

, O’Banion

, Tenner

, Lemere

, Duff

(2001) Inflammatory responses to amyloidosis in a transgenic mouse model of Alzheimer’s disease. Am J Pathol 158, 1345–1354.

35.

Schwab

, Klegeris

, McGeer

(2010) Inflammation in transgenic mouse models of neurodegenerative disorders. Biochim Biophys Acta 1802, 889–902.

36.

Dickson

(1997) The pathogenesis of senile plaques. J Neuropathol Exp Neurol 56, 321–339.

37.

Griffin

, Sheng

, Roberts

, Mrak

(1995) Interleukin-1 expression in different plaque types in Alzheimer’s disease: Significance in plaque evolution. J Neuropathol Exp Neurol 54, 276–281.

38.

Rogers

, Luber-Narod

, Styren

, Civin

(1988) Expression of immune system-associated antigens by cells of the human central nervous system: Relationship to the pathology of Alzheimer’s disease. Neurobiol Aging 9, 339–349.

39.

Benzing

, Wujek

, Ward

, Shaffer

, Ashe

, Younkin

, Brunden

(1999) Evidence for glial-mediated inflammation in aged APP(SW) transgenic mice. Neurobiol Aging 20, 581–589.

40.

Gordon

, Holcomb

, Jantzen

, DiCarlo

, Wilcock

, Boyett

, Connor

, Melachrino

, O’Callaghan

, Morgan

(2002) Time course of the development of Alzheimer-like pathology in the doubly transgenic PS1+APP mouse. Exp Neurol 173, 183–195.

41.

Richards

, Higgins

, Ouagazzal

, Ozmen

, Kew

, Bohrmann

, Malherbe

, Brockhaus

, Loetscher

, Czech

, Huber

, Bluethmann

, Jacobsen

, Kemp

(2003) PS2APP transgenic mice, coexpressing hPS2mut and hAPPswe, show age-related cognitive deficits associated with discrete brain amyloid deposition and inflammation. J Neurosci 23, 8989–9003.

42.

Boom

, Pochet

, Authelet

, Pradier

, Borghgraef

, Van Leuven

, Heizmann

, Brion

(2004) Astrocytic calcium/zinc binding protein S100A6 over expression in Alzheimer’s disease and in PS1/APP transgenic mice models. Biochim Biophys Acta 1742, 161–168.

43.

Jiang

, Burdick

, Glabe

, Cotman

, Tenner

(1994) beta-Amyloid activates complement by binding to a specific region of the collagen-like domain of the C1q A chain. J Immunol 152, 5050–5059.

44.

Rogers

, Cooper

, Webster

, Schultz

, McGeer

, Styren

, Civin

, Brachova

, Bradt

, Ward

(1992) Complement activation by beta-amyloid in Alzheimer disease. Proc Natl Acad Sci U S A 89, 10016–10020.

45.

Padmanabhan

, Levy

, Dickson

, Potter

(2006) Alpha1-antichymotrypsin, an inflammatory protein overexpressed in Alzheimer’s disease brain, induces tau phosphorylation in neurons. Brain 129, 3020–3034.

46.

Luo

, Elwood

, Britschgi

, Villeda

, Zhang

, Ding

, Zhu

, Alabsi

, Getachew

, Narasimhan

, Wabl

, Fainberg

, James

, Wong

, Relton

, Gambhir

, Pollard

, Wyss-Coray

(2013) Colony-stimulating factor 1 receptor (CSF1R) signaling in injured neurons facilitates protection and survival. J Exp Med 210, 157–172.

47.

Boissonneault

, Filali

, Lessard

, Relton

, Wong

, Rivest

(2009) Powerful beneficial effects of macrophage colony-stimulating factor on beta-amyloid deposition and cognitive impairment in Alzheimer’s disease. Brain 132, 1078–1092.

48.

Majumdar

, Cruz

, Asamoah

, Buxbaum

, Sohar

, Lobel

, Maxfield

(2007) Activation of microglia acidifies lysosomes and leads to degradation of Alzheimer amyloid fibrils. Mol Biol Cell 18, 1490–1496.

49.

Udan

, Ajit

, Crouse

, Nichols

(2008) Toll-like receptors 2 and 4 mediate Abeta(1-42) activation of the innate immune response in a human monocytic cell line. J Neurochem 104, 524–533.

50.

Tahara

, Kim

, Jin

, Maxwell

, Li

, Fukuchi

(2006) Role of toll-like receptor signalling in Abeta uptake and clearance. Brain 129, 3006–3019.

51.

Huang

J-J

, Ma

W-J

, Yokoyama

(2012) Expression and immunolocalization of Gpnmb, a glioma-associated glycoprotein, in normal and inflamed central nervous systems of adult rats. Brain Behav 2, 85–96.

52.

Ripoll

, Irvine

, Ravasi

, Sweet

, Hume

(2007) Gpnmb is induced in macrophages by IFN-γ and lipopolysaccharide and acts as a feedback regulator of proinflammatory responses. J Immunol 178, 6557–6566.

53.

Blain

J-F

, Paradis

, Gaudreault

, Champagne

, Richard

, Poirier

(2004) A role for lipoprotein lipase during synaptic remodeling in the adult mouse brain. Neurobiol Dis 15, 510–519.

54.

Nishitsuji

, Hosono

, Uchimura

, Michikawa

(2011) Lipoprotein lipase is a novel amyloid β (Aβ)-binding protein that promotes glycosaminoglycan-dependent cellular uptake of Aβ in astrocytes. J Biol Chem 286, 6393–6401.

55.

Ohmi

, Greenberg

, Rajavel

, Ryazantsev

, Li

, Neufeld

(2003) Activated microglia in cortex of mouse models of mucopolysaccharidoses I and IIIB. Proc Natl Acad Sci U S A 100, 1902–1907.

56.

Tanaka

, Matsuwaki

, Yamanouchi

, Nishihara

(2013) Increased lysosomal biogenesis in activated microglia and exacerbated neuronal damage after traumatic brain injury in progranulin-deficient mice. Neuroscience 250, 8–19.

57.

Magister

, Kos

(2013) Cystatins in immune system. J Cancer 4, 45–56.

58.

Hashimoto

, Suzuki

, Nagai

, Yamashita

, Toyoda

, Matsushima

(2000) Identification of genes specifically expressed in human activated and mature dendritic cells through serial analysis of gene expression. Blood 96, 2206–2214.

59.

Langerholc

, Zavašnik-Bergant

, Turk

, Abrahamson

, Kos

(2005) Inhibitory properties of cystatin F and its localization in U937 promonocyte cells. FEBS J 272, 1535–1545.

60.

Nixon

(2007) Autophagy, amyloidogenesis and Alzheimer disease. J Cell Sci 120, 4081–4091.

61.

Grbovic

, Mathews

, Jiang

, Schmidt

, Dinakar

, Summers-Terio

, Ceresa

, Nixon

, Cataldo

(2003) Rab5-stimulated up-regulation of the endocytic pathway increases intracellular β-cleaved amyloid precursor protein carboxyl-terminal fragment levels and Aβ production. J Biol Chem 278, 31261–31268.

62.

Mueller-Steiner

, Zhou

, Arai

, Roberson

, Sun

, Chen

, Wang

, Yu

, Esposito

, Mucke

, Gan

(2006) Antiamyloidogenic and neuroprotective functions of cathepsin B: Implications for Alzheimer’s disease. Neuron 51, 703–714.

63.

Hayashi

, Koyanagi

, Kusunose

, Okada

, Wu

, Tozaki-Saitoh

, Ukai

, Kohsaka

, Inoue

, Ohdo

, Nakanishi

(2013) The intrinsic microglial molecular clock controls synaptic strength via the circadian expression of cathepsin. Sci Rep 3, S2744.

64.

Perez-Cruz

, Nolte

, van Gaalen

, Rustay

, Termont

, Tanghe

, Kirchhoff

, Ebert

(2011) Reduced spine density in specific regions of CA1 pyramidal neurons in two transgenic mouse models of Alzheimer’s disease. J Neurosci 31, 3926–3934.

65.

Bernstein

, Wiederanders

(1994) An immunohistochemical study of cathepsin E in Alzheimer-type dementia brains. Brain Res 667, 287–290.

66.

Chen

, Subbaiah

(2007) Phospholipid and fatty acid specificity of endothelial lipase: Potential role of the enzyme in the delivery of docosahexaenoic acid (DHA) to tissues. Biochim Biophys Acta 1771, 1319–1328.

67.

Cook

, Gleichman

, Cross

, Doshi

, Ho

, Jordan-Sciutto

, Lynch

, Kolson

(2011) NMDA receptor modulation by the neuropeptide apelin: Imlications for excitotoxic injury. J Neurochem 118, 1113–1123.

68.

Christensen

, Hüttenrauch

, Mitkovski

, Pradier

, Wirths

(2014) Axonal degeneration in an Alzheimer mouse model is PS1 gene dose dependent and linked to intraneuronal Aβ accumulation. Front Aging Neurosci 6, 139.

69.

Wirths

, Weis

, Kayed

, Saido

, Bayer

(2007) Age-dependent axonal degeneration in an Alzheimer mouse model. Neurobiol Aging 28, 1689–1699.