Effects of CD2-associated protein deficiency on amyloid-β in neuroblastoma cells and in an APP transgenic mouse model

CD2-associated protein (CD2AP) is an SH3-containing scaffold adaptor protein which regulates the actin cytoskeleton. Recently, CD2AP was identified as a genetic risk factor for Alzheimer’s disease (AD) by several genome-wide association studies. One of the hallmarks of AD is the accumulation of aggregated forms of Amyloid-β (Aβ) in the brain. In humans, CD2AP AD susceptibility locus (rs9349407) is associated with an increased plaque burden. Aβ production is highly regulated by endocytosis and is influenced by lysosomal function. Lysosomal trafficking is influenced by CD2AP. In this study, we decreased CD2AP levels in N2a neuroblastoma cultures and PS1APP mice and analyzed Aβ levels and plaque burden. Our data show that suppressing CD2AP expression using shRNA in N2a-APP695 cells results in decreased cell membrane amyloid precursor protein, decreased Aβ release and a lower Aβ42/Aβ40 ratio. CD2AP protein is expressed in the brain as detected by western blot, and the expression level is dependent on gene dosage. In 1-month old PS1APP mice, complete loss of CD2AP in brain resulted in a decreased Aβ42/Aβ40 ratio in brain tissue lysates while there was no effect on Aβ deposition or accumulation in PS1APP mice expressing one copy of CD2AP. CD2-Associated Protein affects Aβ levels and Aβ42/Aβ40 ratio in vitro. The effect of CD2-Associated Protein on Aβ metabolism is subtle in vivo.

Background CD2-associated protein (CD2AP) was originally identified as a scaffold protein required for organization of the immunological synapse -the specialized interface between a T lymphocyte and an antigen-presenting cell [1]. It was later shown that CD2AP, by virtue of its multiple protein-protein binding modules, interacts with multiple proteins involved in diverse biological processes. These associations have implicated CD2AP in receptor tyrosine kinase internalization, actin cytoskeleton remodeling and vesicular trafficking [2].
Alzheimer's disease (AD) is a neurodegenerative disorder characterized by the impairment of memory and other cognitive functions as well as the presence of extracellular amyloid plaques and intracellular neurofibrillary tangles [3]. There is substantial evidence indicating that amyloid-β (Aβ) plays an essential role in the development of AD [4,5]. The deposition of Aβ into amyloid plaques is dependent on the concentration of brain interstitial fluid Aβ [6], which is regulated by endocytosis [7]. Recently CD2AP was detected as a risk factor for AD by several genomewide association studies [8][9][10]. In a yeast model for cellular toxicity elicited by Aβ, a functional homolog of CD2AP [11] was identified as a suppressor of Aβ toxicity [12]. In addition, a recent study in humans suggests that CD2AP AD susceptibility locus (rs9349407) is associated with increased plaque burden [13]. However, the relationship between CD2AP and Aβ has never been reported in mammalian cells or in mouse models expressing human amyloid precursor protein (APP)/Aβ.
Since the reported AD susceptibility locus (rs9349407) which has impact on plaque load in humans [13] is in the CD2AP gene, we asked whether manipulating CD2AP expression level affects Aβ levels. In the current study, we knocked down CD2AP expression in N2a-APP695 cells and observed a decrease in Aβ levels as well as the ratio of Aβ 42 /Aβ 40 in the cell culture medium. We crossed CD2AP knockout mice with PS1APP mice and observed a reduction of Aβ 42 /Aβ 40 in the brain tissue. Due to the fact that CD2AP knockout mice have glomerular disease and do not survive beyond a few months of age until plaque onset in PS1APP mice [14], we also measured the effects of CD2AP haploinsufficiency on amyloid plaque deposition. There was no effect of CD2AP haploinsufficiency on Aβ deposition up to 7 months of age.

Results and discussion
Effects of CD2AP deficiency in cultured N2a-APP695 cells APP processing is regulated by endocytosis. Given that CD2AP plays an important role in regulating endocytosis, we first tested whether CD2AP has an effect on Aβ synthesis or Aβ release in cultured cells. We used CD2AP shRNA to knockdown CD2AP levels (Figures 1, 2A) in neuroblastoma N2a-APP 695 cells and measured Aβ in the cell lysates and Aβ released into the cell culture medium. The results showed that CD2AP sh RNA1 (Sh1) significantly decreased both Aβ 40 and Aβ 42 secreted into the culture medium by about 20~30% (Figure 1A,B) while the Aβ 40 and Aβ 42 in cell lysates were not affected ( Figure 1D,E). CD2AP sh2 had greater effects on the Aβ 42 in the medium as compared to CD2AP sh1 ( Figure 1A,B). The Aβ 40 and Aβ 42 in cell lysates were increased about 30% by CD2AP sh2 (Figure 1D,E). Interestingly, both CD2AP sh1 and CD2AP sh2 decreased the Aβ 42 / Aβ 40 ratio in cell culture medium ( Figure 1C). However, the Aβ 42 / Aβ 40 ratios in cell lysates were unaltered ( Figure 1F). We further examined the total APP and APP on the cell surface in these N2a cells. The results showed that CD2AP sh1 and sh2 did not change total APP levels in N2a cells ( Figure 2B,D). However, membrane APP levels were decreased by CD2AP sh1 and sh2 ( Figure 2B,C). In cells, nascent APP is posttranslationally modified and transported from the endoplasmic reticulum to the plasma membrane [15]. To be proteolytically cleaved into Aβ, APP must be internalized from the cell surface into the cell and transported to endosomes where β-secretase and γ-secretase complexes cleave APP to produce Aβ [7,16]. In the current study, knocking down CD2AP in N2a-APP695 cells may decrease APP on cell surface which would result in less APP getting into endosomes and less Aβ being released into cell culture medium.

Expression of CD2AP in the brain
Before we studied the effects of CD2AP on Aβ pathology in vivo, we first determined whether CD2AP is expressed in the brain and whether the expression level correlates with CD2AP gene dosage. Using western blot, we detected CD2AP in CD2AP +/+ brains with CD2AP −/− brains serving as a negative control ( Figure 3A). As reported in the kidney [17], CD2AP protein level in the brains of mice with CD2AP haploinsufficiency (CD2AP +/− ) is about 50% of the level in CD2AP +/+ mice ( Figure 3B).
Effects of CD2AP knockout on brain Aβ levels in 1-month old PS1APP mice Next, we assessed whether CD2AP has similar effects on Aβ levels and Aβ 42 /Aβ 40 ratio in vivo. If CD2AP affects Aβ production or release, we would expect to see the changes in young mice before plaque deposition. Therefore, we generated 1-month old PS1APP/CD2AP +/+ (female, n = 6; male, n = 6) and PS1APP/CD2AP −/− (female, n = 7; male, n = 5) mice and measured Aβ 40 and Aβ 42 levels in the PBS soluble fraction of cortical tissue. The results showed that there were no significant changes in the absolute concentration of Aβ 40 and Aβ 42 ( Figure 4A,B). However, the ratio of Aβ 42 /Aβ 40 was lower in PS1APP/ CD2AP −/− mice as compared to PS1APP/CD2AP +/+ mice ( Figure 4C). This effect was similar in both females (0.37 ± 0.008 vs 0.29 ± 0.0025 for PS1APP/CD2AP +/+ vs PS1APP/CD2AP −/− , respectively, p < 0.05, Student's t-test) and males (0.35 ± 0.015 vs 0.30 ± 0.010 for PS1APP/ CD2AP Although inhibiting CD2AP expression levels in cultured cells resulted in decreased levels of Aβ in the cell culture medium, we did not observe changes in the absolute concentration of cortical Aβ 42 or Aβ 40 in CD2AP knockout  PS1APP mice as compared to PS1APP mice expressing two copies of CD2AP. While Aβ levels in cultured APP-expressing N2a-APP695 cells are mainly determined by APP processing, Aβ levels in vivo are regulated not only by Aβ production and release but also Aβ uptake and degradation by different cell types, Aβ clearance mediated by interstitial fluid (ISF) and cerebrospinal fluid (CSF) bulk flow and Aβ transport across the blood brain barrier. Among these mechanisms, many of them are regulated by vesicular trafficking, which can be influenced by manipulating CD2AP levels. For example, disruption of endocytosis in neurons inhibits APP processing and reduces Aβ levels [7]. On the other hand, disruption of endocytosis in the microglia or astrocytes could result in an increase of extracellular Aβ due to decreased Aβ uptake and degradation [18]. Therefore, these effects could theoretically cancel each other out when knockout of CD2AP occurs in all the cell types. This may explain why we did not observe the same change in Aβ concentration in vivo as we have seen in vitro.
In cultured N2a-APP695 cells, knocking down CD2AP resulted in a lower Aβ 42 /Aβ 40 ratio. Similar changes also occurred in the soluble Aβ 42 /Aβ 40 ratio in 1-month old PS1APP/CD2AP −/− mice as compared to PS1APP/ CD2AP +/+ mice. The ratio of Aβ 42 /Aβ 40 is determined by γ-secretase cleavage of APP. Mutations of presenilin, an active enzymatic component of the γ-secretase complex, lead to autosomal dominant familial AD [19,20] likely in large part due to an increased Aβ 42 /Aβ 40 ratio. On the other hand, mutations in APP [21] or gamma secretase modulators [22] can influence γ-secretase cleavage and alter the Aβ 42 /Aβ 40 ratio. Some molecules such as phosphatidylinositol clathrin assembly lymphoid-myeloid leukemia (PICALM) shift the Aβ 42 /Aβ 40 ratio through affecting internalization of γ-secretase [23]. In the current study, the change of Aβ 42 /Aβ 40 ratio occurred before plaque deposition, suggesting CD2AP knockout could affect the selective production of Aβ 40 and Aβ 42 . Since CD2AP binds membrane proteins, it could affect Aβ cleavage by interacting with γ-secretase complex, by modifying APP directly or through some intermediary molecules to shift the ratio of Aβ 42 /Aβ 40 . It is shown recently that a homolog of Nephrin, a protein interacting with CD2AP [14], is required for γ-secretase mediated Notch and APP-like cleavages in Drosophila [24].

Effects of CD2AP haploinsufficiency on amyloid deposition in 7-month old PS1APP mice
Shifting the ratio of Aβ 42 /Aβ 40 results in an altered time course of plaque deposition in both humans [19,20] and mice [25]. Since we observed changes in the Aβ 42 /Aβ 40 ratio in cultured cells and 1-month old PS1APP mice, we next asked whether CD2AP deficiency affects amyloid plaque load in older PS1APP mice. CD2AP −/− mice have a~6-week life-span due to renal failure [14] while the average plaque onset age in PS1APP mice is at~4-month. Therefore, we were not able to assess plaque deposition in PS1APP/CD2AP −/− mice. Mice with CD2AP haploinsufficiency (CD2AP +/− ) live a normal life span but express~50% less CD2AP in the brain compared to CD2AP +/+ mice ( Figure 3A,B). We therefore asked whether a~50% reduction of CD2AP levels affects Aβ pathology in PS1APP mice. We generated PS1APP/ CD2AP +/− and PS1APP/CD2AP +/+ mice and characterized their Aβ pathology at the age of 7-months. We first measured Aβ levels in the cortical tissue lysates. The results showed no difference in the absolute level of Aβ 40 and Aβ 42 in the PBS (soluble forms of Aβ), Triton or Guanidine (insoluble forms of Aβ) brain fractions between PS1APP/CD2AP +/− and PS1APP/CD2AP +/+ groups ( Figure 5). In the PBS soluble fraction, the Aβ 42 /Aβ 40 ratio in the female PS1APP/CD2AP +/− group was significantly lower than that in female PS1APP/ CD2AP +/+ group (p < 0.05, Student's t-test). However, the effects on males tended to trend in the opposite direction ( Figure 5). In the Triton and Guanidine fractions, there was no change in the Aβ 42 /Aβ 40 ratio associated with CD2AP gene status ( Figure 5). To determine whether CD2AP haploinsufficiency affects the morphology or other properties of the Aβ plaques, we stained the tissue with biotinylated anti-Aβ 1-13 monoclonal antibody HJ3.4B ( Figure 6A) or Thioflavin S ( Figure 6C) which stains fibrillar forms of Aβ plaques. We did not observe any significant changes in plaque distribution, individual plaque size or plaque morphology associated with CD2AP gene status. We further quantified the% area covered with plaques in the cortex. The results demonstrated that neither Aβ immunostained plaques ( Figure 6B) nor fibrillar plaques ( Figure 6D) were different in mice with CD2AP haploinsufficiency. Taken together, although CD2AP haploinsufficiency lowered the Aβ 42 /Aβ 40 ratio in PBS fraction of female PS1APP mice, the majority of our data demonstrate that CD2AP haploinsufficiency did not cause changes in Aβ accumulation.
For certain genes that have clear effects on Aβ metabolism and deposition such as apolipoprotein E or PICALM, a 50% reduction of expression has significant effects on Aβ pathology [26,27]. However, in our in vivo studies, lowering of CD2AP expression levels by 50% in the PS1APP/CD2AP +/− mice did not affect Aβ pathology as assessed by biochemical or histological methods as compared to that in control PS1APP/CD2AP +/+ mice at the age of 7-months. It is possible that expression of 50% the level of CD2AP is sufficient to maintain adequate CD2AP function in the brain. In the future, additional work on CD2AP conditional knockout mice which live a longer life-span can be done to verify the effects of CD2AP on amyloid deposition in mice expressing lower than levels found in C2DAP haploinsufficiency in the brain.
medium. CD2AP expression is readily detectable in the brain and so we extended out in vitro studies to an APP mouse model. In 1-month old PS1APP mice, a complete loss of CD2AP reduced the Aβ 42 /Aβ 40 ratio in the cortical tissue but the absolute levels of Aβ 40 and Aβ 42 were unaltered. In 7-month old PS1APP mice, CD2AP haploinsufficiency did not cause significant changes in Aβ pathology as analyzed by biochemical and histological assays. In the future, conditional knockout of CD2AP in different brain cells can be produced to further confirm or refute the effects of CD2AP on Aβ pathology. Primary neurons cultured from PS1APP/CD2AP −/− mice need to be used to confirm the in vitro effects of CD2AP on Aβ production and release. Besides Aβ plaques, another hallmark of AD is the accumulation of insoluble tau protein in structures such as intracellular neurofibrillary tangles. It was found that RNAi targeting Cindr, the fly ortholog of the human CD2AP, enhances tau toxicity in a Drosophila model in a recent study [28]. Therefore, CD2AP could also modify AD status by interacting with tau or potentially via other mechanisms. In the future, additional possible mechanisms should be explored to establish the mechanism(s) underlying the role of CD2AP in AD pathogenesis.

CD2AP knockdown in cultured cells
N2a-APP695 cells were grown in DMEM/Opti-MEM (50:50) supplemented with 5% FBS and 200 μg/ml of G418. Control (firefly luciferase target sequence, fLuc; GCTTACGCTGAGTACTTCGA) and two different CD2AP-specific shRNA duplexes (CD2AP target sequences, CD2AP sh1 and sh2: GTGGAACCCTGAACAATAAG and GGAACCAATGAAGATGAACTTACA, respectively) were cloned into the pFLRu lentivirus as previously described [30]. Viral supernatants were generated in 293 T cells by transfection of the lentiviral plasmids with Lipofectamine 2000 (Invitrogen) and the packaging plasmids as described. Pooled supernatants, harvested at 24 and 48 hours post-transfection, were applied to N2a cells with 8ug/ml polybrene and spun at room temperature for 2 hours at 2000 rpm. Supernatants were replaced with fresh medium immediately after centrifugation. To confirm the knock down of CD2AP, the level of CD2AP protein in RIPA cell lysates was assessed by western blot using a rabbit anti-CD2AP polyclonal antibody [1]. Actin was detected using a mouse monoclonal antibody (Sigma) which served as internal control. The cells were then cultured in a 12-well plate in serum free medium at a density of 80%. After 8 hrs, the cell culture medium was collected and the cells were homogenized in RIPA buffer. The Aβ 40 and Aβ 42 in the cell culture medium and cell lysates were measured by sandwich ELISA. All the values were normalized to the protein concentration in the cell lysate to correct for the differences caused by different cell numbers in each well. To assess total APP and membrane APP, N2a-fluc, N2a-sh1 and N2a sh2 cells were cultured in a 6-well plate at a density of 90%. The membrane proteins were biotinylated by incubating the cells with EZ-Link Sulfo-NHS-SS-Biotin (Thermo Scientific) on ice for 30 min. Then the cells were homogenized in lysate buffer containing 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% (v/v) Nonidet P-40, 0.5% (w/v) deoxycholate, and 1× protease inhibitor mixture [27]. Total APP in the cell lysates was assessed using western blot with a rabbit anti-APP polyclonal antibody (Zymed). The value was normalized to actin in the same lysates. Cell lysates containing same amount of total protein were incubated with Dynabeads® MyOne™ Streptavidin T1 (Invitrogen) to pull down biotinylated membrane protein. Western blot for APP was performed on the pull-down products to assess the membrane APP.

Western blot for CD2AP in brain lysates
Four-week old CD2AP +/+ , CD2AP +/− and CD2AP −/− mice were perfused with ice-cold PBS containing 0.3% heparin. The cortices were dissected and homogenized in RIPA buffer. CD2AP was detected by western blot using a rabbit anti-CD2AP polyclonal antibody [1]. The western blot bands were quantified using image J (National Institutes of Health). The levels of CD2AP were normalized using tubulin values from the same sample.

Immunohistochemistry
Serial coronal sections at 50-μm thickness were collected from the rostral to the caudal end of each brain hemisphere using a freezing sliding microtome (Leica). Aβ plaques were immunostained using biotinylated anti-Aβ 1-13 monoclonal antibody HJ3.4B [31].

Thioflavine S staining
For fibrillar plaques, brain sections were stained with 0.025% Thioflavin S (Sigma) in 50% ethanol for 10 min. Then the sections were washed with 50% ethanol twice followed by PBS [32].
Imaging Immunostained brain sections were scanned using a Nanozoomer slide scanner (Hamamatsu Photonics). Quantitative analysis of immunopositive staining was performed as previously described [33]. Briefly, images of immunostained sections were exported with NDP viewer (Hamamatsu Photonics), converted to 8-bit grayscale using ACDSee Pro 2 software (ACD Systems) and threshold was set to highlight positive staining and analyzed using ImageJ (National Institutes of Health). 3 sections per mouse (Bregma, −1.4 mm caudal to Bregma, −2.0 mm caudal to Bregma) were quantified (the cortex immediately dorsal to the hippocampus) and the average was used to represent each mouse.

Statistics
Two-tailed Student's t-test was used to determine if there were significant differences between two groups unless otherwise specified. One-way ANOVA was used to compare differences among 3 or more groups followed by Tukey post-test unless otherwise specified. Data in all the figures are expressed as mean ± S.E.M unless otherwise specified.