Neurodegeneration-Like Pathological and Behavioral Changes in an AAV9-Mediated p25 Overexpression Mouse Model

Abstract

Background: The transgenic mice models overexpressing human p25 contribute greatly to the in vivo neurotoxic mechanism of p25 in neurodegenerative diseases. However, it is time-consuming to manipulate existing transgenic mice models.

Objective: Here we aim to establish a novel mouse model of neurodegeneration by overexpressing p25 mediated by recombinant adeno-associated virus serotype 9 (rAAV9).

Methods: AAV9-GFP-p25 encoding GFP-fused p25 driven by synapsin promoter, and the control, AAV9-GFP, were delivered in mice by tail-vein injection. Assessments of p25 expression, neurodegenerative pathology, and behavioral changes were performed.

Results: GFP expression was detected by in vivo imaging as early as one week after virus injection. Notably, widespread expression of p25 was obviously found in cortex, hippocampus, and cerebellum in AAV9-GFP-p25 mice. Moreover, decreased hippocampus volumes in AAV9-GFP-p25 mice were detected by 7T MRI examination about one month after injection. Further, these AAV9-GFP-p25 mice exhibited progressive memory impairment from three-month to six-month after virus injection. At last, hyperphosphorylated tau, neurofibrillary tangles, activated astrocytes and microglia cells were elevated in these p25 mice at about six months after virus delivery. However, amyloid-β plaques, overt neuronal loss, and apoptosis in the hippocampus and cortex were not significantly induced by AAV9-mediated p25 overexpression.

Conclusion: The AAV9-mediated p25 overexpression mouse model, which is a practical model exhibiting neurodegeneration-like pathological and behavioral changes, provides an easier and time-saving method to explore the functions of p25 in vivo, as well as an alternative tool for development of drugs against neurotoxic of p25.

Keywords

Adeno-associated virus 9 Alzheimer’s disease cyclin-dependent kinase 5 disease models magnetic resonance imaging p25 transgenic mice

INTRODUCTION

Cyclin-dependent kinase 5 (Cdk5) is a member of Ser/Thr kinases family [1, 2], which requires binding to noncyclin proteins p35 or p39 to play pivotal roles in neurogenesis, neuronal migration, axonal support, neurite outgrowth, synaptic plasticity, learning, and memory [3 –5]. Under neurotoxic stimuli, membrane-bound p35 is cleaved into p10 and p25 by calpain [6, 7]. The important finding, which firstly linked the significance of Cdk5/p25 to human degenerative diseases, is that p25, rather than p35, presents in the neurons containing neurofibrillary tangles (NFTs) in Alzheimer’s disease (AD) patients [6]. Neurotoxin of Cdk5/p25 has been further supported by the deregulation of Cdk5 activity in postmodern or animal disease models in the development of a growing number of diseases, such as Parkinson’s disease [8, 9], amyotrophic lateral sclerosis [10], Niemann-Pick Type C [11], ischemia, and stroke [12]. Although there is dispute on the higher activity of Cdk5/p25 compared with Cdk5/p35 in vitro [13], it is rational to believe that p25, with longer half-life and different subcellular location compared to p35, substantially extends the activation period of Cdk5, hyperphosphorylates some downstream targets, and induces neurotoxicity [8 –12].

However, the mechanism of Cdk5/p25 in the pathogenesis of neurodegenerative diseases remains unresolved. p25 transgenic (p25-Tg) mice models are the primary methods to explore the mechanism of neurotoxin of p25 in AD development [14, 15]. Transient overexpression of p25 in other models can be induced by neurotoxin factors, including ischemia and exposure to 1-methyl- 4-phenyl-1,2,3,6-tetrahydropyridine (MPTP/MPP+) or amyloid-β (Aβ) peptides [16 –18]. p25-Tg mice exhibit AD-like neuropathology with apparently hyperphosphorylated tau 4 weeks after induction and Aβ accumulations 8 weeks after induction [19]. Neuroinflammation occurs one week after induction of p25 [20]. However, it is not only technically required but also time-consuming to handle transgenic mice models.

Recombinant adeno-associated virus (rAAV) with excellent transduction profile and non-pathogenic nature belongs to the family Parvoviridae. Among central nervous system (CNS) expressing serotypes, AAV9 can pass through blood-brain barriers and provide a good level of transgenic expression, thus is a candidate vector for systemically deliver expression cassettes into the CNS of experimental animals [23]. Consequently they are the alternative vehicle systems for modeling genetic disorders of the CNS, and have potential advantages over transgenic models [21 –23].

In this study, we developed a p25 overexpression mouse model mediated by rAAV-9 virus, in which p25 was C-terminally fused to GFP and under control of synapsin promoter. One week after administration of AAV-9-p25 virus by tail-vein injection in C57BL/6 mice, green fluorescence was detected in vivo. The p25 overexpression mice demonstrated degeneration-like pathological changes, such as extensive phosphorylated tau and neuroinflammation, independently of neuronal apoptosis and Aβ deposit. Moreover, they displayed progressive mild memory impairment, compared to AAV-GFP control mice. Hence, a novel p25-overexpression mouse model mediated by AAV was successfully established.

MATERIALS AND METHODS

Preparation of AAV9-GFP-p25 virus

AAV cis-plasmid pAAV-hsynapsin-eGFP (hereafter pAAV-GFP) carrying an eGFP reporter gene driven by the human synapsin-1 promoter was purchased from the University of Pennsylvania Vector Core facility (Philadelphia, PA). Myc-tagged-p25 was generated by PCR using plasmid pcDNA3.1-C-p35 (NIH, Bethesda, MD) as template and was inserted in pAAV-GFP. The resulting plasmid was named as pAAV-GFP-p25, in which p25 was in frame with GFP on its c-terminus. Expression of GFP-p25 fusion protein was verified in human embryonic kidney 293 (HEK293) cells, which were cultured in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% fetal bovine serum (Gibco™, Thermo Fisher Scientific, Waltham, MA).

Large-scale preparations of AAV vectors were performed according to references [24, 25]. HEK293 cells were transfected by Polyethylenimine (PEI, Polysciences, Warrington, PA) method with three plasmids: pAAV-GFP or pAAV-GFP-p25, pAAV2/9 containing AAV2 rep and AAV9 cap genes, and adenovirus helper plasmid, pDF6. Both pAAV2/9 and pDF6 were from the University of Pennsylvania Vector Core facility. The culture medium containing the vector particles was collected. Virus purification was performed with PEG8000/(NH4) ₂SO4 aqueous two-phase extraction methods as reported [26]. The titer of rAAV virion was determined by measuring the copy numbers of vector genome (vg) with quantitative PCR. All AAV vectors were suspended in sterile phosphate-buffered saline (PBS) as the final concentration to 10¹³ vg/ml.

Experimental animals and injection of rAAVs

Animal protocols were approved by the Southern Medical University Committee on Animal Care (China) and conducted according to the Guideline for the Care and Use of Laboratory Animals. C57BL/6J mice (male, 4-5 weeks old) were purchased from Southern Medical University (Guangzhou, China). Male APP/PS1 (APPswe, PSEN1dE9) 85Dbo/MmJNju mice, also known as APP/PS1, expressing human amyloid-β protein precursor (AβPP) and a mutant human presenilin 1 under the control of mouse prion promoter, were purchased from Model Animal Research Center of Nanjing University (Nanjing, China). All mice were housed on a 12 h light/dark cycle with free access to water and food. Equal numbers of C57BL/6 J mice were randomly separated into three groups and received tail-vein injection with 200 μl PBS or viral solution containing 3 × 10¹¹ vg AAV9-GFP-p25 or AAV9-GFP in PBS, respectively. Animals were euthanized at the different time points as required after the injection.

Tissue collection and preparation

Mice were deeply anesthetized with sodium pentobarbital (120 mg/kg, intraperitoneal injection). After transcardiac perfusion with ice-cold saline, brain, spinal cord, heart, lung, liver, spleen, pancreas, kidney, and testis from each mouse were obtained by dissection, washed in saline and then were prepared for fluorescence imaging. For immunohistochemistry (IHC), brains of mice (7-month-old for APP/PS1-Tg mice, about 6 months post-injection for AAV9 mice) were obtained and bisected along the midline, cut into 20 μm sections and then placed in 4% paraformaldehyde solution at 4°C for a minimum 24 h, immersed in 15% and 30% sucrose for cryoprotection, paraffin-embedded according to the reported reference [20]. The 4 μm thickness slices were cut with a Leica RM2016 vibratome (Leica, Heidelberg, Germany). For future western blotting analysis, brains of mice (3 months or 6 months after injection of AAV9 virus) were used to obtain cortex and hippocampus tissues, which were immediately frozen in liquid nitrogen and stored in –80°C.

In vivo (whole animal) and ex vivo (organ) imaging

For in vivo imaging, mice were anesthetized and then shaved from the head to the lower torso to avoid interference from fur. For ex vivo imaging, fresh organs were isolated. Shaved animals or dissected organ were placed in the Kodak in-vivo Imaging System F (Kodak Molecular imaging, Rochester, NY) to analyze fluorescence intensity according to the manufacturer’s instruction. GFP was excited at 465 nm and detected at 535 nm. Data was collected as photons/sec/mm² using Kodak Molecular Imaging Software (Kodak), as described [27].

Histology and immunohistochemistry studies

Hematoxylin and eosin (H&E) staining, Nissl staining, and modified Bielschowsky silver staining were performed according to the published literature [14 , 29]. Terminal deoxynucleotide transferase-mediated dUTP-biotin nick-end labeling (TUNEL) staining was performed using the In Situ Cell Death Detection Kit (TMR Red, Roche, Mannheim, Germany), according to the manufacturer’s instructions.

For IHC studies, the tissue sections were deparaffinized in xylene, and rehydrated in a graded series of alcohols. Then, antigen retrieval was performed, endogenous peroxidase activity was blocked, sections were then incubated overnight at 4°C with the following antibodies (Abs): anti-GFAP monoclonal Ab (1 : 500 dilution, Abcam, Cambridge, UK), anti-p35/25 (C64B10) Ab (1 : 100 dilution, Cell Signaling Technology, Beverly, MA), anti-human PHF-tau monoclonal Ab (AT8, 1 : 100 dilution, Thermo scientific, Rockford, IL), GFP-tag polyclonal Ab (1 : 200 dilution, Life Technologies, Grand Island, NY), anti-ionized calcium-binding adapter molecule 1 Ab (Iba1, 1 : 500 dilution, Wako Chemicals, Richmond, VA), anti-neuronal nuclei Ab (NeuN, 1 : 500 dilution, Chemicon International Inc., Temecula, CA), and anti-beta Amyloid 1-42 Ab (Aβ_1 - 42, 1 : 500 dilution, Abcam), followed by incubation with goat anti-mouse or anti-rabbit horseradish peroxidase-conjugated secondary Abs (DAKO, Carpinteria, CA). All slices were further incubated with horseradish peroxidase/diaminobenzidine according to the manufacturer’s protocols. The stained slices were then dehydrated, cover-slipped and images were taken at 10 or 40 magnification under microscope (Olympus, Tokyo, Japan). The staining was reported as the relative intensity of staining (percent) of the area by using Image J (1.49v, NIH, Bethesda, MD).

Western blotting analyses

Western blotting analysis was performed routinely. Briefly, denatured protein samples from cortex, hippocampus, and/or cerebellum of each mouse were separated on 12% SDS-PAGE and electro-transferred to PVDF (Millipore). Incubation primary Abs were: myc-tag monoclonal Ab (1 : 1000 dilution, Proteintech, Rosemont, IL), AT180 mouse anti-tau monoclonal Ab (1 : 500 dilution, thermo scientific), anti-Tau phospho S199 polyclonal Ab (1 : 500 dilution, Abcam), anti-p35/25 (C64B10) Ab (1 : 1000 dilution, Cell Signaling Technology), and anti-GAPDH Ab (1 : 3000 dilution, Santa Cruz Biotechnology, Santa Cruz, CA). Bound primary Abs were detected with the anti-mouse or anti-rabbit horseradish peroxidase–conjugated secondary Ab (1 : 5000, CWBIO, Beijing, China), visualized by the enhanced chemiluminescence method (Millipore). The densities of protein blots were quantified by using Image J and normalized to the level of GAPDH.

Magnetic resonance imaging

About one month after the administration of AAV vectors, mice MRI image was carried out at the Mental Health Center, Shantou University Medical College, China, and methodologies was according to the published literature [30]. Mice was anesthetized with isoflurane (3% for induction, 1–2% for maintenance, Abbott Labratories Ltd, U.K.) mixed with oxygen (1 liter/min) and delivered through a nasal mask, and then was placed in a supine position with the head fixed on a palate holder equipped with an adjustable nose cone. Body temperature was maintained around 37°C using a heating blanket. A 7.0 T horizontal DriveDrive 2 MR system (Agilent, 7T/160/AS) with a 160-mm bore and 400 mT/m actively shielded gradient coil and a dedicated animal brain surface coil (Varian Medical Systems, Inc., Palo Alto, CA) were used to acquire in vivo MR imaging. The hippocampal formation in each image was demarcated manually, and the volume of the hippocampus in each brain was measured and calculated using Image J software.

Animal behavioral study

Mice (3-, or 6-month post injection) were housed and habituated for 2 h in the behavior testing room before the behavioral tests. All the tests were performed on consecutive days in a dimly illuminated room with standard conditions of temperature and free from any stray noise, according to reference [19]. All the behavioral equipment was cleaned with 75% ethanol and dried between each animal. Behavioral assessment was carried out by two trained assistants who were blind to the information of eachmouse.

Open field test

The open field (81 cm × 81 cm × 30 cm) was equally divided into 16 areas (20 × 20 cm). Each mouse was put into the field with the face to the corner, and the number of crossings in center, the total number of crossings, and rearing (standing on the hind legs) were measured within 5 min.

Rotarod test

The Rotarod test was performed in the TSE behavioral experiment machine (TSE technology, Guangdong, China). The accelerating rotating rate was set from 4 rpm to 40 rpm in 5 min. The score was 0 when mice fall down three times in less than 10 seconds. Latency to fall was recorded within 5 min.

Elevated Plus-Maze test

An elevated plus maze apparatus consisted of two opposite open arms (45 × 42 cm), two opposite closed arms (45 × 42 × 145 cm) and one central square (42 × 42 cm) in the TSE behavioral experiment machine. Each mouse was placed in the central square of the maze facing one closed arm. The time spent and number of entries in the open arms and the closed arms were recorded within 5 min. Anxiety-like behavior was measured by the percentages of time and entries in the open-arms out of total time and entries, respectively.

Radial maze test

The radial arm maze was performed as reported [19] with mild modifications. The maze consisted of eight arms (42 × 14 × 30 cm), numbered from 1 to 8, extended radially from the central area (TSE Radial maze, Guangdong, China). On training, each mouse was freely to explore all the arms within 10 min and eat a small piece from a fresh chocolate cake placed in only one arm in the order of 2, 4, 6, or 8 each day. Novel food was also placed outside of all the arms to avoid interference from smell. When test was performed, each mouse was placed in the center of the maze, with four arms having baits (arm 2, 4, 6, and 8). Time was recorded when the cakes were fully eaten or stopped when 5 min had elapsed. Test was performed for 10 constitute days. The reference memory errors represented the number of entries to an arm without bait. The working memory error was the number that each mouse repeatedly went to an arm with bait. The total time to finish the task wasrecorded.

Statistics

Data were presented as means ± standard errors. Continuous data were analyzed with unpaired t tests or one-way analysis of variance (ANOVA) followed by Tukey’s test by using SPSS 22.0 (IBM, Armonk, NY). p value less than 0.05 was considered significant statistically.

RESULTS

Green fluorescence was detected in mice after tail-vein injection of AAV vectors

To establish a p25-overexpressed mouse model, single strand AAV serotype 9, which is able to pass through the blood-brain barrier [23], was used to deliver p25 to mice brains via tail-vein injection. The pAAV-eGFP expression plasmid containing human synapsin-1 promoter was assured to targeting p25 to neuron cells. Myc-tagged-p25 was fused with GFP protein for monitoring of virus distribution. After transfected pAAV-GFP-p25 plasmid in HEK293 cells, an approximately 54 kDa protein band was detected by c-myc monoclonal Ab in the cell lysate (middle lane in Fig. 1A), which indicated successful expression of GFP-p25 fusion protein.

To assess the efficiency of AAV vectors by intravenous injected on juvenile mice (4 to 5 weeks old), AAV9-GFP, AAV9-GFP-p25 or PBS was administrated in mice. Green fluorescence in mice one week or one month after injection was monitored under in vivo imaging machine. Mice were anesthetized and shaved to avoid fur absorbs light noise [27]. As shown in Fig. 1B, GFP fluorescence was detected in mice as early as one week after injection with AAV9-GFP-p25, and continued to be detected in mice about four weeks after virus injection, compared with wild-type animal. Fluorescence in the brain was not seen due to blockage of the skull.

To check the bio-distribution of AAV vectors, organs of euthanized mice (one month after injection) were immediately dissected and fluorescence intensity was measured under in vivo imaging machine. The expression of GFP was visualized by the color of the organs (Fig. 1C), and the quantity of GFP intensity was represented in Fig. 1D. The data indicated that GFP expression was expressed higher in several organs in AAV9-GFP or AAV9-GFP-p25 mice, especially in brains, compared to PBS control mice (p < 0.01). These data suggest that successful delivery of targeting gene in mouse brain by systemic administration of AAV vectors.

p25 overexpression in brain reduced hippocampus volumes about one month after AAV injection

It is reported that neurodegeneration occurs as early as one month in p25 overexpressed transgenic mice [15]. Thus, we measured hippocampus volumes in PBS control mice, AAV9-GFP mice or AAV9-GFP-p25 mice by high-field MRI examination about one-month after virus injection. As shown in Fig. 2A, the hippocampus volumes decreased significantly in AAV9-GFP-p25 mice, compared with AAV9-GFP (p < 0.05).

After MRI experiment, mice were deeply euthanized and mice brains were harvested for Western blotting, H&E or Nissl staining analysis. As indicated in Fig. 2B, expression of GFP-p25 protein was measured in cortex, hippocampus and cerebellum samples by Western blotting analysis. Obvious GFP-p25 fusion protein was detected as an approximately 54 kDa band by c-myc monoclonal Ab. The ratios of overexpressed p25 to endogenous p35 in the AAV9-GFP-p25 mice were increased compared to the AAV9-GFP mice (Fig. 2C, p < 0.05). H&E staining revealed a smaller hippocampus in AAV9-GFP-p25 mice, and no difference was found by Nissl staining in the neurons in cornu ammonis 1 (CA1) region of hippocampus in AAV9-GFP-p25 mice compared with AAV9-GFP mice, as shown in Fig. 2D. Together, these data showed that overexpression of p25 in neuron cells negatively regulated endogenous p35 expression, reduced hippocampus volumes as early as about one month after AAVinjection.

p25 overexpression increased tau phosphorylation and caused inflammation without neuron apoptosis and neuron loss 6 months after AAV administration

It is reported that pathological changes are observed in 4–8 weeks in p25-Tg mouse model [15, 20] and 6-7 months in APP/PS1-Tg mouse model [31]. Here, we performed histological analysis in brain sections from 7-month-old AAV mice (6 months post-administration of AAV) and APP/PS1-Tg mice of the same age. As shown in Fig. 3A and F, after 6-month virus injection, GFP was still detectable in the cortex and hippocampus neurons in both AAV9-GFP mice and AAV9-GFP-p25 mice. As indicated in Fig. 3B and G, increased p25/p35 was found in the cortex and hippocampus of AAV9-GFP-p25 mice compared to that of AAV9-GFP mice (p < 0.01) or the hippocampus of APP/PS1-Tg mice (p < 0.05). Although the anti-p35/p25 Ab recognized the c-terminus of p35/p25 and GFP-p25 fusion proteins, we might still speculate that p25 was overexpressed in AAV9-GFP-p25 mice compared to AAV9-GFP mice.

Hyperphosphorylation of tau protein is considered to be one of the most important events in AD pathogenesis. As shown in Fig. 3C and H, the level of phosphorylation tau was higher in AAV9-GFP-p25 mice and APP/PS1-Tg mice, compared with AAV-GFP mice (p < 0.01).

Inflammation also plays an important role in neurodegenerative diseases. Previous studies reveal that neuroinflammation occurred as early as one week after induction of p25 in p25-Tg mice [20]. In our study, elevated expression of GFAP-positive astrocytes was found in the hippocampus of the p25-overexpressed mice compared with AAV9-GFP mice (p < 0.01), and in the cortex of APP/PS1-Tg mice but not in that of AAV9-GFP-p25 mice or AAV9-GFP mice (p < 0.01), as shown in Fig. 3D and I. Meanwhile, prominent activation of Iba1-positive microglia cells was detected in the cortex and hippocampus of AAV9-GFP-p25 mice compared with APP/PS1-Tg mice (p < 0.05) and the hippocampus of AAV9-GFP mice (p < 0.01), as shown in Fig. 3D and I. The results suggest that robust inflammation is associated with p25 overexpression in brains of AAV9-GFP-p25 mice.

Finally, we wanted to check whether overexpression of p25 would cause neuron apoptosis and neuron loss. Indeed, significant neuron loss and neuron apoptosis was observed in the cortex of APP/PS1-Tg mice but not AAV9-GFP-p25 mice, (p < 0.01, Fig. 4A–D). Although there is a trend of neuron loss in the hippocampus of AAV9-GFP-p25 mice (Fig. 4C), neuron apoptosis was not detected by TUNEL (Fig. 4D).

Collectively, p25 overexpression in our model increased phosphorylated tau and caused robust inflammation, but did not induce significantly neuron apoptosis and neuron loss 6 months after AAV administration.

The hallmarks of NFTs but no Aβ deposit appeared in the hippocampus and cortex of p25 overexpressed mice

Next, we wanted to check whether hyperphosphorylation of tau increased NFTs and Aβ deposit in p25 overexpression mice. Accumulation of two pathologic hyperphosphorylated tau isoforms, AT180 and S199, increased significantly in the hippocampus and cortex of the AAV9-GFP-p25 mice compared with AAV9-GFP mice about 6 months post injection of virus (p < 0.05, except S199 in the cortex: p < 0.01; Fig. 5A and B).

Further, modified Bielschowsky silver staining was performed to detect NFTs and Aβ deposit. As in Fig. 5C and D, results revealed that NFTs appeared obviously in 7-month-old p25-overexpressed mice (6-month post-injection) and APP/PS1-Tg mice but not in 7-month-old control mice with AAV9-GFP injection. Moreover, lots of argyrophilic structures including Aβ plaques and NFTs were seen in the frontal lobes and hippocampus in the 7-month-old APP/PS1-Tg mice, but Aβ plaques did not appear in p25-overexpressed mice of the same age.

Anti-Aβ_1–42 antibody staining further revealed that the amount of Aβ_1–42 was significantly higher in the hippocampus and cortex of APP/PS1-Tg mice compared with the p25-overexpressed model (p < 0.001, Fig. 5E and F). These data demonstrated that 6-month p25 overexpression caused formation of NFTs but no Aβ deposit in the AAV-based model.

Progressive memory impairment was observed in AAV9-GFP-p25 mouse model

Behavioral changes are important elements to evaluate disease animal models. No significant difference in the performance of open-field, Rotarod test, and Elevated Plus-Maze was detected among AAV9-GFP-p25, AAV9-GFP and PBS mice 3-months post-injection (data not shown) and 6-months post-injection (p > 0.05, Fig. 6 A–C). Notably, AAV9-GFP-p25 mice showed significant impairment of working memory compared with AAV9-GFP mice and PBS mice 3-months post-injection (p < 0.05) and 6-months post-injection (p < 0.01), shown in Fig. 6D. Moreover, 3 months after infection of virus, the reference memory of the p25-overexpressed mice appeared to be abnormal (p < 0.05) but returned to a nearly normal level after 10-day repetitive tests compared to control groups (p > 0.05), shown in Fig. 6E. After the infection duration extended to 6 months, AAV9-GFP-p25 mice revealed irreversible memory impairment (p < 0.05, Fig. 6E). The results suggested that the p25-overexpressed mice might experience mild reference memory impairment 3-months post-injection, and finally reached the level of dementia 6-months post-injection. Together, p25 overexpression mainly caused memory impairment. We did not observed obviously motor function impairment or anxiety in the AAV based p25 overexpression mouse model.

DISCUSSION

Here we successfully established an AAV9-mediated p25-overexpressed mouse model. We found that tail-vein-injected AAV9-GFP-p25 virus passed the blood-brain barrier and overexpressed GFP-p25 in brains lasted for over 6 months. Notably, overexpression of GFP-p25 in neurons negatively regulated endogenous p35 expression. Importantly, the novel AAV9-GFP-p25 mice had neurodegeneration-like pathologic changes, including hyperphosphorylated tau, NFT formation, and neuroinflammation, but without Aβ deposit, neuron apoptosis, and neuron loss. Moreover, p25 overexpression induced hippocampus atrophy and progressive memory impairment without any effect on locomotor activity. Therefore, our results demonstrate that neurodegeneration can be induced by AAV delivered neurotoxic GFP-p25 in relatively old age mice (4-5 weeks old). This AAV-based model of p25 overexpression provides the capability to explore the relationship between p25, inflammation and tau pathology in vivo, and an alternative tool for the therapeutic study on neurodegenerative diseases.

Compared to p25-Tg mice models, AAV9-GFP-p25 model had moderate pathologic changes. Several p25-Tg mice models using different promoters exhibit different pathologic or behavioral severity. Enolase-promotor-drived p25-Tg mice [28] exhibit increased locomotor activity such as whole-body exertion tremors, but less anxious as tested by the elevated-plus maze. Dramatic hyperphosphorylated tau and NFTs were detected in amygdala, hypothalamus, and cortex. Driven by cytomegalovirus or platelet-derived growth factor promoter, two p25-Tg mice models present neither tau pathology nor neuronal apoptosis [32, 33]. In a CaMKII-promoter-driven p25-Tg tTA inducible mouse model [14], characteristic hallmarks of AD-like pathological changes present in the cortex and hippocampus areas, including progressive neuronal loss, astrogliosis, NFTs, and Aβ formation. However, similar CaMKII-promoter-driven-p25-Tg tTA inducible mouse model constructed by an independent laboratory observed progressive neurodegeneration in hippocampus and activated microglia but no tau phosphorylation and apoptosis were trigged [15]. Here in our study, we observed neuroinflammation and tau pathology (Figs. 3 & 4) shared by all three phenotype positive models [14 , 28], however, we did not observe Aβ formation and obvious neuronal apoptosis and loss.

Memory impairment is one of the most important characteristics for neurodegenerative disease model. The AAV-based mice model had impaired working memory and mild reference memory impairment 3 months post-injection, which was much later than p25-Tg mice models [14, 19]. The phenomena were likely to be early events in the progression of neurodegenerative diseases, and might be due to the slightly increased level of p25. Learning itself can induce p25, and small amount of p25 is specifically required for memory extinction [34]. Thus, slight increased p25 affected mild memory impairment in the earlier adulthood in our model but long-term accumulation of p25 increased neurotoxicity and finally induced severe irreversible memory impairments in our model. Thereby, our AAV-based p25-overexpressed mouse model recapitulates mild memory impairment in the early stage and severe memory impairment in the late stage of dementia.

The absent Aβ formation, no obvious neuronal loss, and milder behavioral changes in our study, compared to the two inducible Tg mice models [14], are consistent with an AAV-based model of tau pathology (injection of AAV1 just after born) [35]. Three reasons may account for the milder phenotypes in our AAV-mediated p25 mice. First, virus may be eliminated by the immunity system and thus reduced the expression of p25. Second, elevation of exogenous protein may trigger inner mechanism of p25 elimination to return to the previous homeostasis. Third, the earlier the p25 overexpression is induced, the more mortality rate is presented. Thus the induction of p25 in adulthood like in our case usually had mild symptoms and different phenotypes [15].

Future study will focus on several aspects. The mechanism of decreased volume of hippocampus needs to be elucidated. One possibility might result from downregulation of p35, which is essential for neuron functions. p35 downregulation is not observed in p25-Tg mice. In addition, it will be interesting to know whether overexpression of p25 in aging mice or in neonatal mice will cause more severe phenotype. Third, in this experiment, for the purpose to distinguish overexpressed p25 from endogenous p35 or p25 with tagged GFP, we did not test whether GFP-tagged p25 would have the same fashion as untagged p25, even we had GFP control group. Cruz et al. [14] found that GFP-tagged p25, Cmyc-tagged p25, and untagged p25 had similar fashion. Thus, AAV9-p25 mice and AAV9-GFP-p25 mice will be compared to assure similar behavior in the coming treatment studies. Finally, CDK5/p25 specific inhibition, for example with TFP5 or TP5 [18, 36] will be examined in the AAV9-based-p25 mouse model to evaluate the therapeutic potential for neurodegeneration disease.

In conclusion, the AAV9-based p25 overexpression mouse model provides an easy and time-saving method to explore the functions of p25 in vivo, as well as an alternative degenerative mouse model for treatment strategies.

Footnotes

ACKNOWLEDGMENTS

The work was supported by the National Nature Science Fund of China (#81271430), Guangdong provincial nature science fund (1414050000990), Guangdong Provincial Universities fund for Experts Recruitment Program (#2012-328) to YH, and President Fund of Nanfang Hospital (2013C010) to XZ.

Authors’ disclosures available online ().

References

Lew

, Beaudette

, Litwin

, Wang

(1992) Purification and characterization of a novel proline-directed protein kinase from bovine brain. J Biol Chem 267, 13383–13390.

Meyerson

, Enders

, Wu

, Su

, Gorka

, Nelson

, Harlow

, Tsai

(1992) A family of human cdc2-related protein kinases. EMBO J 11, 2909–2917.

Tsai

, Delalle

, Caviness

Jr , Chae

, Harlow

(1994) p35 is a neural-specific regulatory subunit of cyclin-dependent kinase 5. Nature 371, 419–423.

, Tsai

(2011) Cyclin-dependent kinases in brain development and disease. Annu Rev Cell Dev Biol 27, 465–491.

Sharma

, Sharma

, Amin

, Albers

, Pant

(1999) Regulation of cyclin-dependent kinase 5 catalytic activity by phosphorylation. Proc Natl Acad Sci U S A 96, 11156–11160.

Patrick

, Zukerberg

, Nikolic

, de la Monte

, Dikkes

, Tsai

(1999) Conversion of p35 to p25 deregulates Cdk5 activity and promotes neurodegeneration. Nature 402, 615–622.

Rosales

, Nodwell

, Johnston

, Lee

(2000) Cdk5/p25(nck5a) interaction with synaptic proteins in bovine brain. J Cell Biochem 78, 151–159.

Smith

, Schlichtenbrede

, Tschernutter

, Bainbridge

, Thrasher

, Ali

(2003) AAV-Mediated gene transfer slows photoreceptor loss in the RCS rat model of retinitis pigmentosa. Mol Ther 8, 188–195.

Smith

, Jenkinson

, Johansen-Berg

, Rueckert

, Nichols

, Mackay

, Watkins

, Ciccarelli

, Cader

, Matthews

, Behrens

(2006) Tract-based spatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage 31, 1487–1505.

10.

Nguyen

, Sanchez-Pernaute

, Cunningham

, Bankiewicz

(2001) Convection-enhanced delivery of AAV-2 combined with heparin increases TK gene transfer in the rat brain. Neuroreport 12, 1961–1964.

11.

, Li

, Davies

, Vincent

(2002) Deregulation of cdk5, hyperphosphorylation, and cytoskeletal pathology in the Niemann-Pick type C murine model. J Neurosci 22, 6515–6525.

12.

Rashidian

, Rousseaux

, Venderova

, Qu

, Callaghan

, Phillips

, Bland

, During

, Mao

, Slack

, Park

(2009) Essential role of cytoplasmic cdk5 and Prx2 in multiple ischemic injury models, in vivo. J Neurosci 29, 12497–12505.

13.

Zheng

, Amin

, Hu

, Rudrabhatla

, Shukla

, Kanungo

, Kesavapany

, Grant

, Albers

, Pant

(2010) A 24-residue peptide (p5), derived from p35, the Cdk5 neuronal activator, specifically inhibits Cdk5-p25 hyperactivity and tau hyperphosphorylation. J Biol Chem 285, 34202–34212.

14.

Cruz

, Tseng

, Goldman

, Shih

, Tsai

(2003) Aberrant Cdk5 activation by p25 triggers pathological events leading to neurodegeneration and neurofibrillary tangles. Neuron 40, 471–483.

15.

Muyllaert

, Terwel

, Kremer

, Sennvik

, Borghgraef

, Devijver

, Dewachter

, Van Leuven

(2008) Neurodegeneration and neuroinflammation in cdk5/p25-inducible mice: A model for hippocampal sclerosis and neocortical degeneration. Am J Pathol 172, 470–485.

16.

Kim

, Frank

, Dobbin

, Tsunemoto

, Tu

, Peng

, Guan

, Lee

, Moy

, Giusti

, Broodie

, Mazitschek

, Delalle

, Haggarty

, Neve

, Lu

, Tsai

(2008) Deregulation of HDAC1 by p25/Cdk5 in neurotoxicity. Neuron 60, 803–817.

17.

Binukumar

, Zheng

, Shukla

, Amin

, Grant

, Pant

(2014) TFP5, a peptide derived from p35, a Cdk5 neuronal activator, rescues cortical neurons from glucose toxicity. J Alzheimers Dis 39, 899–909.

18.

, Shukla

, Amin

, Grant

, M

, Kuntz

, Steiner

, Pant

(2015) Peptide (TFP5/TP5), derived from Cdk5 activator P35, provides neuroprotection in the MPTP model of Parkinson’s disease. Mol Biol Cell 26, 4478–4491.

19.

Sundaram

, Poore

, Sulaimee

, Pareek

, Asad

, Rajkumar

, Cheong

, Wenk

, Dawe

, Chuang

, Pant

, Kesavapany

(2013) Specific inhibition of p25/Cdk5 activity by the Cdk5 inhibitory peptide reduces neurodegeneration in vivo. J Neurosci 33, 334–343.

20.

Sundaram

, Chan

, Poore

, Pareek

, Cheong

, Shui

, Tang

, Low

, Wenk

, Kesavapany

(2012) Cdk5/p25-induced cytosolic PLA2-mediated lysophosphatidylcholine production regulates neuroinflammation and triggers neurodegeneration. J Neurosci 32, 1020–1034.

21.

Janson

, McPhee

, Bilaniuk

, Haselgrove

, Testaiuti

, Freese

, Wang

, Shera

, Hurh

, Rupin

, Saslow

, Goldfarb

, Goldberg

, Larijani

, Sharrar

, Liouterman

, Camp

, Kolodny

, Samulski

, Leone

(2002) Clinical protocol. Gene therapy of Canavan disease: AAV-2 vector for neurosurgical delivery of aspartoacylase gene (ASPA) to the human brain. Hum Gene Ther 13, 1391–1412.

22.

Kirik

, Bjorklund

(2003) Modeling CNS neurodegeneration by overexpression of disease-causing proteins using viral vectors. Trends Neurosci 26, 386–392.

23.

Foust

, Nurre

, Montgomery

, Hernandez

, Chan

, Kaspar

(2009) Intravascular AAV9 preferentially targets neonatal neurons and adult astrocytes. Nat Biotechnol 27, 59–65.

24.

Huang

, Hartley

, Yin

, Herskowitz

, Lah

, Ressler

(2013) AAV2 production with optimized N/P ratio and PEI-mediated transfection results in low toxicity and high titer for in vitro and in vivo applications. J Virol Methods 193, 270–277.

25.

Grieger

, Choi

, Samulski

(2006) Production and characterization of adeno-associated viral vectors. Nat Protoc 1, 1412–1428.

26.

Guo

, El-Gohary

, Prasadan

, Shiota

, Xiao

, Wiersch

, Paredes

, Tulachan

, Gittes

(2012) Rapid andsimplified purification of recombinant adeno-associated virus. J Virol Methods 183, 139–146.

27.

Fujiki

, Tao

, Bianchi

, Giel-Moloney

, Leiter

, Johnson

(2008) Quantification of green fluorescent protein by in vivo imaging, PCR, and flow cytometry: Comparison of transgenic strains and relevance for fetal cell microchimerism. Cytometry A 73, 11–118.

28.

Ahlijanian

, Barrezueta

, Williams

, Jakowski

, Kowsz

, McCarthy

, Coskran

, Carlo

, Seymour

, Burkhardt

, Nelson

, McNeish

(2000) Hyperphosphorylated tau and neurofilament and cytoskeletal disruptions in mice overexpressing human p25, an activator of cdk5. Proc Natl Acad Sci U S A 97, 2910–2915.

29.

Litchfield

, Nagy

(2001) New temperature modification makes the Bielschowsky silver stain reproducible. Acta Neuropathol 101, 17–21.

30.

Shao

, Yan

, Xuan

, Peng

, Huang

, Wu

, Xu

(2015) Chronic social isolation decreases glutamate and glutamine levels and induces oxidative stress in the rat hippocampus. Behav Brain Res 282, 201–208.

31.

Webster

, Bachstetter

, Nelson

, Schmitt

, Van Eldik

(2014) Using mice to model Alzheimer’s dementia: An overview of the clinical disease and the preclinical behavioral changes in 10 mouse models. Front Genet 5, 88.

32.

Bian

, Nath

, Sobocinski

, Booher

, Lipinski

, Calla-han

, Pack

, Wang

, Walker

(2002) Axonopathy, tau abnormalities, and dyskinesia, but no neurofibrillary tangles in p25-transgenic mice. J Comp Neurol 446, 257–266.

33.

Takashima

, Murayama

, Yasutake

, Takahashi

, Yokoyama

, Ishiguro

(2001) Involvement of cyclin dependent kinase5 activator p25 on tau phosphorylation in mouse brain. Neurosci Lett 306, 37–40.

34.

Seo

, Giusti-Rodriguez

, Zhou

, Rudenko

, Cho

, Ota

, Park

, Patzke

, Madabhushi

, Pan

, Mungenast

, Guan

, Delalle

, Tsai

(2014) Activity-dependent p25 generation regulates synaptic plasticity and Abeta-induced cognitive impairment. Cell 157, 486–498.

35.

Cook

, Kang

, Carlomagno

, Lin

, Yue

, Kurti

, Shinohara

, Jansen-West

, Perkerson

, Castanedes-Casey

, Rousseau

, Phillips

, Bu

, Dickson

, Petrucelli

, Fryer

(2015) Tau deposition drives neuropathological, inflammatory and behavioral abnormalities independently of neuronal loss in a novel mouse model. Hum Mol Genet 24, 6198–6212.

36.

Shukla

, Zheng

, Mishra

, Amin

, Steiner

, Grant

, Kesavapany

, Pant

(2013) A truncated peptide from p35, a Cdk5 activator, prevents Alzheimer’s disease phenotypes in model mice. FASEB J 27, 174–186.