Reduction of Blood Amyloid-β Oligomers in Alzheimer’s Disease Transgenic Mice by c-Abl Kinase Inhibition

Abstract

One of the pathological hallmarks of Alzheimer’s disease (AD) is the presence of amyloid plaques, which are deposits of misfolded and aggregated amyloid-beta peptide (Aβ). The role of the c-Abl tyrosine kinase in Aβ-mediated neurodegeneration has been previously reported. Here, we investigated the therapeutic potential of inhibiting c-Abl using imatinib. We developed a novel method, based on a technique used to detect prions (PMCA), to measure minute amounts of misfolded-Aβ in the blood of AD transgenic mice. We found that imatinib reduces Aβ-oligomers in plasma, which correlates with a reduction of AD brain features such as plaques and oligomers accumulation, neuroinflammation, and cognitive deficits. Cells exposed to imatinib and c-Abl KO mice display decreased levels of β-CTF fragments, suggesting that an altered processing of the amyloid-beta protein precursor is the most probable mechanism behind imatinib effects. Our findings support the role of c-Abl in Aβ accumulation and AD, and propose AD-PMCA as a new tool to evaluate AD progression and screening for drug candidates.

Keywords

Alzheimer’s disease amyloid-beta peptide amyloid-beta protein precursor c-Abl tyrosine kinase imatinib oligomers PMCA

INTRODUCTION

In this aging society, neurodegenerative diseases are a growing problem. Alzheimer’s disease (AD) the most common form of dementia, accounts for nearly 60–90% of all cases [1]. It is characterized by the loss of short-term memory, disorientation, and impairment in judgment and reasoning [2]. The neuropathological hallmarks of the AD brain are the senile plaques, resulting from the extracellular accumulation of amyloid-beta peptide (Aβ); and the presence of neurofibrillary tangles, intracellular aggregates of hyperphosphorylated tau protein [3, 4]. Other AD features include a generalized neuroinflammatory response, characterized by activated microglia and astrocytes, synaptic impairment, and neuronal loss in specific brain areas [5, 6]. The cognitive impairment in AD correlates strongly with the loss of synaptic density, which is accompanied by Aβ accumulation. Compelling evidence suggest that the misfolding and aggregation of Aβ could be the triggering event that causes the subsequent brain abnormalities [7]. Several therapies aiming at lowering Aβ accumulation are currently under investigation and others have been already examined in clinical trials. However, most of them have had modest success at best [1 , 9].

We have previously shown that c-Abl kinase plays a central role in Aβ-induced neurodegeneration [10, 11]. c-Abl is a member of the Abl family of non-receptor tyrosine kinases that plays an important role in cytoskeleton remodeling, neuronal development, neurogenesis, neuronal migration, axonal extension, and synaptic plasticity [12 –15]. Our laboratory demonstrated that c-Abl regulates synaptic structure and function [16] and that c-Abl kinase activity is required for synaptic loss induced by Aβ-oligomers [17]. We have also reported that c-Abl is activated in AD animal models, where the administration of imatinib, a c-Abl inhibitor, produces a significant improvement in memory and cognition [11]. Unexpectedly, imatinib also causes the reduction of senile plaques. However, the role of c-Abl on the accumulation of Aβ oligomers has not been evaluated yet. In this study, we aimed to address this question, evaluating the effects of c-Abl inhibition on Aβ oligomers levels and its consequences on AD pathology.

Several studies show that Aβ oligomers are present in biological fluids of AD patients and AD models [18 –21]. However, their detection is limited by the small amounts of these species in tissues other than the brain. To detect Aβ oligomers in the blood of AD transgenic mice we modified the protein misfolding cyclic amplification technology (PMCA), previously used to detect prions [22]. This technology is based on the seeding-nucleation properties of misfolded proteins [23]. Using this novel technique, we decided to assess the effect of imatinib outside the CNS, evaluating the levels of circulating Aβ-oligomers in the blood of AD mice models. Interestingly, we observed a significant reduction of Aβ-oligomers in the plasma and brain of AD transgenic mice that had been peripherally injected with the c-Abl inhibitor imatinib. These findings correlate with a significant reduction of AD features, such as size and number of Aβ plaques and active astrocytes in AD-transgenic mice treated with imatinib. In addition, we found that imatinib-treated mice show better performance in water maze behavioral testing. Although it has been reported that imatinib crosses the blood-brain barrier poorly [24], our findings suggest that peripheral administration of imatinib effectively inhibits activated c-Abl, decreasing the levels of phosphorylated c-Abl at the brain.

Overall, our data support the role of c-Abl in AD and the use of c-Abl inhibitors, such as imatinib, for effective AD treatment.

MATERIALS AND METHODS

Animal model and imatinib treatment

APPswe/PSEN1ΔE9 mice (Formally B6C3-Tg APPswe, PSEN1dE9 85Dbo/J, The Jackson Laboratory) were used as models of Aβ-accumulation. c-Abl KO mice were bred from c–Abl^loxp/c–Abl^loxp and Nestin–Cre⁺ mice (kindly donated by Dr. A.J Koleske, Yale School of Medicine). Animals were housed in groups of up to 5 in individually ventilated cages under standard conditions (22°C, 12-h light–dark cycle) receiving food and water ad libitum. All animal experiments were carried out in accordance to international regulations and approved by the committee of Bioethics and Biosafety of the P. Catholic University of Chile. APPswe/PSEN1ΔE9 transgenic mice received intraperitoneal (i.p.) injections every other day (25 mg/kg) during a two-week period. Control animals received i.p. injections of vehicle (Saline).

Aβ peptide and oligomers detection in plasma

After euthanasia, blood samples were obtained by cardiac puncture from deeply anaesthetized imatinib treated and non-treated mice. Blood fractions were obtained after centrifugation using a Ficoll gradient, and plasma was stored at –80°C until use. To detect high-molecular-weight oligomers, 500μL of plasma were filtered using 10 KDa cut off filters following the manufacture’s indications (Amicon, Millipore). Briefly, 500μL of plasma were centrifuged in Amicon filters at 14,000 rpm for 15 min. The amount of small Aβ species was detected by dot blot, adding 10μL of the filtrate in a dry nitrocellulose membrane. After blocking with PBS 5% milk, the membrane was incubated overnight with a 1:1,000 dilution of the anti-Aβ antibody 4G8 (Covance). The membrane was washed three-times with PBS 0.05% tween and incubated with a 1:5,000 dilution of anti-mouse HRP conjugated antibody (Pierce). The secondary antibody was detected through enhanced chemiluminescence using the ECL Plus Western blotting substrate (Thermoscientific). Densitometric analyses were performed using the NIH ImageJ software.

Conventional ELISA using 4G8 (Covance) antibody was used to measure the levels of Aβ high-molecular weight species (>10 KDa). After filtering 500μL of plasma using Amicon filters (as described above), 10μL of the concentrated retentate fraction were placed on an ELISA plate (Nunc) and incubated overnight at 4°C. The plate was blocked using PBS 3% horse serum (Pierce). The anti-Aβ antibody (4G8) was incubated at 37°C for 4 h. The secondary antibody and the ABC amplification system were used following the manufacture’s indications (Pierce).

The dot blot using conformational antibody A11 (Millipore) was performed following the method described previously [25], adding 10μL of plasma samples to dry nitrocellulose membranes. After blocking with TBS 10% milk, the membrane was incubated overnight with a 1:1,000 dilution of the anti-oligomer antibody A11. The membrane was washed three-times with TBS 0.01% tween and then incubated with a 1:5,000 dilution of anti-rabbit HRP conjugated antibody (Pierce). The secondary antibody was detected through enhanced chemiluminescence using the ECL Plus Western blotting substrate (Thermoscientific). Densitometric analyses were performed using the NIH ImageJ software.

AD-PMCA

AD-PMCA is based on the seeding properties of the amyloids. Given that the limiting step during the seeded-nucleated polymerization is the formation of stable oligomers (seeds), the addition of oligomers into a solution containing the monomeric protein accelerates protein misfolding and aggregation [23, 24]. Therefore, we can detect the presence of oligomers in peripheral blood by measuring the seeding activity of a plasma sample over a monomeric Aβ substrate. For the preparation of the monomeric Aβ-susbtrate, Aβ_1-42 peptide was synthesized by the Protein Core at Yale University using solid-phase chemistry. The Aβ-substrate at 2μM concentration was incubated at 25°C with plasma samples (10μL of 10 KDa Filter concentrated sample). Aliquots of the mix were incubated for the indicated times and centrifuged at 14,000 rpm for 10 min, at room temperature. To follow aggregation of the samples over time, we detected the remaining Aβ-substrate (soluble Aβ) in the supernatants by ELISA using the 4G8 antibody (Covance). 10μL of each supernatant sample were placed on an ELISA plate (Nunc) and incubated overnight at 4°C. The plate was blocked using PBS 3% horse serum (Pierce). The anti-Aβ antibody (4G8) was incubated at 37°C for 4 h. The secondary antibody and the ABC amplification system were used following the manufacture’s suggestions (Pierce). The value obtained in this way was subtracted from the initial concentration of the Aβ-substrate to calculate the amount of aggregated Aβ-substrate at each time point.

Immunohistochemistry and immunofluorescence

Mice were sacrificed by CO₂ inhalation following the proceedings of the committee of Bioethics and Biosafety of the P. Catholic University of Chile. Brains were removed, post-fixed into fixative solution (4% paraformaldehyde) followed by 20% and 30% sucrose in PBS at 4°C overnight. Brains were cut in 30μm sections with a cryostat at –20°C. Three to four sections per experimental animal were analyzed. Immunolabeling was performed with anti-Aβ antibody 4G8 (1:500, Covance) or WO2 (1:500, Millipore), and GFAP (1:500, Cell Signaling). Anti-mouse-Alexa Fluor-488 or anti-mouse-Alexa Fluor 555 (1:1,000, Molecular Probes) were used as secondary antibodies. Thioflavin S (ThS) staining (Sigma-Aldrich) was performed as previously described [26]. BSB (trans, trans)-1-bromo- 2,5-bis-(3-hydroxycarbonyl-4-hydroxy) styrylbenzene (AnaSpec) staining for oligomers was used following standard protocols [27].

Cell culture treatments and analysis

HT22 hippocampal-like cells or primary neurons were treated with the c-Abl inhibitor imatinib (Gleevec, Novartis, Basel, Switzerland) overnight. After treatment, cell extracts were prepared in ice-cold RIPA buffer (50 mM Tris–HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% NP-40, 0.25% sodium deoxycholate, 0.1% SDS) including phosphatase (1 mM NaF and 1 mM Na₃VO₄) and protease inhibitors (1μg/mL aprotinin, 1μg/mL leupeptin, 1μg/mL pepstatin, 1 mM PMSF). Proteins were quantified using the BCA-Kit (Pierce, Rockford, IL), and 40μg of samples were analyzed by western blot.

Western blot

For cell culture experiments, equal amounts of proteins (40μg) were separated by SDS-PAGE (12% gel) and then transferred to nitrocellulose membranes. The following antibodies were used: anti-APP 22c11 (1:1,000; Millipore), anti-Aβ WO2 (1:1,000; Millipore), anti-Aβ 4G8 (1:1,000, Covance) and anti-β-tubulin (5H1, 1:10,000). The secondary horseradish peroxidase (HRP)-conjugated antibodies (1:5,000) were obtained from Pierce. Brain homogenates were obtained by lysing tissue in RIPA buffer at 4°C (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP40, 0.5% Sodium deoxycholate, SDS 0.1%, 1 mM EDTA, 1 mM EGTA) plus protease inhibitors (1μg/μl aprotinin, 1μg/μl leupeptin, 1μg/μl pepstatin, 1 mM DTT, 1 mM PMSF) in a ratio of 10 mL of RIPA buffer per mg of mouse brain. The samples were centrifuged at 10,000 rpm for 10 min at 4°C. Proteins were quantified by BCA-Kit (Pierce), and 30μg of samples were separated by standard SDS-PAGE (12% gel), with the exception of brain Aβ-oligomers (NuPAGE^® Novex^® 4–12% Bis-Tris Gels, Invitrogen). After transfer to nitrocellulose membranes the following antibodies were used: anti-Aβ 4G8 (1:1,000; Covance), anti-Aβ WO2 (1:1,000; Millipore), anti-c-Abl (1:500; Santa Cruz, CA), anti-BACE (1:1,000, Millipore), anti-β-tubulin (5H1, 1:10,000; Sigma), and anti-GAPDH (6C5, 1:10,000; Santa Cruz). Horseradish peroxidase (HRP)-conjugated antibodies (1:5,000) were obtained from Pierce. The secondary antibodies were detected through enhanced chemiluminescence using the ECL Plus Western blotting substrate (Thermoscientific). Densitometric analysis was performed using the NIH ImageJ software.

Water maze behavioral testing

From three days before the start of the training protocol and during training, the mice received an i.p. injection of 12.5 mg/kg imatinib or saline every 2 days. Mice were trained in a circular water maze (1.2 m in diameter and 30 cm deep) using a four-trial-per-day regimen for five consecutive days, followed by 2 days off, and then for an additional 2 days.

Statistical analysis

Graphs are expressed as means±standard error (s.e.m.). T-test or two-way analysis of variance (ANOVA) followed by a post-hoc comparisons test were used to analyze differences among groups. Statistical analyzes were performed using the GraphPad Prism 5.0 software (GraphPad Software Inc).

RESULTS

Plasma circulating Aβ-oligomers are reduced in AD transgenic mice treated with imatinib

Here, we tried to dissect the role of c-Abl on Aβ accumulation and its consequences on AD pathology. To evaluate the role of c-Abl on the accumulation of Aβ oligomers, APPswe/PSEN1ΔE9 mice were treated every-other-day with i.p. injections of 25 mg/kg of imatinib during a two week period. We used 7-month-old mice because amyloid plaques begin to appear when the APPsw/PSEN1ΔE9 animals are 6 months old [28]. First, we evaluated the Aβ levels at the periphery. In order to quantitatively assess the reduction of Aβ species, we analyzed mice plasma samples by ELISA using the 4G8 antibody (which recognizes the amino acid residues 17–24 of Aβ). We ruled-out the contribution of non-oligomeric Aβ peptide by filtrating the plasma samples with a 10 KDa cut-off microcentrifuge filter. First, we measured the amount of Aβ in the filtrated plasma by dot blot, which should contain species smaller than 10 KDa (monomers/dimers), using an anti-Aβ antibody (4G8). A slight reduction of low molecular weight Aβ species was detected in plasma from imatinib treated mice, although it was not statistically significant (Fig. 1A). In contrast, analysis by ELISA of the fraction containing high-molecular-weight Aβ-oligomers (retentate) showed that plasma from transgenic animals that received imatinib contained significantly less Aβ-oligomers than the blood of control mice, with an approximate reduction of 50% (Fig. 1B). In addition, we were able to detect Aβ-oligomers in non-filtered plasma samples from AD transgenic mice using the conformational antibody A11 [25] (Fig. 1C). Interestingly, animals treated with imatinib displayed a significant decreased in A11-immunoreactivity compared with control mice, indicating that imatinib reduces dodecamers of Aβ (Fig. 1C). Finally, we evaluated whether imatinib was able to reduce Aβ-oligomers that are capable of catalyzing misfolding and aggregation (Aβ seeds), to rule out that imatinib decreases only by-stander oligomers that do not contribute to the pathology. To answer this question, we standardized an ultra-sensitive in vitro Aβ-seeding assay (AD-PMCA), slightly modifying the existing PMCA technique previously used to detect misfolding proteins in biological samples [22, 29]. This methodology allows detection of the polymerization of a soluble Aβ-substrate over time in response to the presence of Aβ oligomers (seeds). Samples containing higher amounts of oligomers will “seed”, accelerating the polymerization of the substrate, reducing the lag phase and changing its aggregation kinetics (Fig. 1F). We observed that while the blood of transgenic mice accelerated the misfolding of the Aβ-substrate, blood samples from mice treated with imatinib showed an increase in the length of the lag phase (lagT), indicating that there are significantly fewer Aβ seeds circulating in the blood (Fig. 1D). In addition, we analyzed the time required to achieve 50% of substrate-aggregation (IC₅₀), as samples containing higher amounts of Aβ seeds accelerate the polymerization of the substrate and reduce the time needed for aggregation. Blood samples from mice receiving imatinib had, in average, significantly longer IC₅₀, indicating a lower concentration of Aβ seeds (Fig. 1E). Altogether, these results suggest that imatinib significantly impacts the level of Aβ-oligomers in peripheral blood, and propose the use of AD-PMCA as a new parameter to assess AD progression in mice models.

Imatinib induced oligomer reduction in plasma correlated with decreased Aβ deposition in the brain

The extracellular accumulation of Aβ in the form of senile plaques is one of the main hallmarks of AD. To determine whether the decrease of plasma Aβ by c-Abl inhibition is accompanied by amyloid load reduction in the brain, we analyzed the extent of Aβ accumulation in APPswe/PSEN1ΔE9 transgenic animals treated with imatinib. In order to evaluate the effect of imatinib in the formation of fibrillar amyloid deposits, Thioflavin-S (ThS) staining was performed in sections of cerebral cortex and hippocampus, as these regions are the most affected by amyloid deposition in both AD patients and transgenic animals. Control APPswe/PSEN1ΔE9 mice have extensive amyloid deposition in both cortex and hippocampus, whereas transgenic mice treated with imatinib exhibited reduced ThS-positive plaques (Fig. 2A). In addition, histological analysis using the 4G8 antibody against Aβ, showed a reduction of amyloid aggregates in imatinib-treated mice compared with controls (Fig. 2B). In order to quantitatively evaluate the effect of imatinib in AD transgenic mice, we quantified the density of amyloid plaques, expressed as number of plaques per area in fluorescent ThS images normalized by the number of plaques in vehicle-treated mice. We found significant differences in plaque density in animals that received imatinib compared with the control group (p < 0.05, T-test) (Fig. 2C). Similarly, the amyloid burden expressed as the percentage of cortex area where ThS-positive plaques are present, showed a significant decrease (>30%) in imatinib-treated animals compared with mice injected with vehicle alone (p < 0.01, T-test) (Fig. 2D). These results indicate that c-Abl inhibition reduced the extent of mature fibrillar deposits and the number of amyloid plaques. To evaluate whether imatinib affected other Aβ species besides mature fibrillar deposits, we performed immunofluorescence labeling using an anti-Aβ antibody (W0-2) in order to measure the total Aβ levels. In agreement with the ThS-staining results, imatinib significantly decreased the percentage of cortex area with Aβ-immunoreactive plaques (p < 0.01, T-test) (Fig. 2E). In addition, we evaluated Aβ oligomer burden using a fluorescent dye that preferentially binds to oligomeric species (BSB). BSB fluorescent-staining shows decreased oligomer burden in the cortex and hippocampus of transgenic mice treated with imatinib compared to the age-matched mice that received vehicle (Fig. 2F). The quantification using image analysis showed a significant reduction of total oligomeric species in the animal model treated with imatinib (p < 0.001, T-test) (Fig. 2H). Thus, imatinib affected mature Aβ deposits and overall levels of Aβ, decreasing Aβ oligomers, the most toxic species in the Aβ aggregation pathway. We corroborated these findings by biochemical analysis of the brain homogenates using the 4G8 anti-Aβ antibody (Fig. 2G). This analysis revealed that among several Aβ species, the Aβ-dodecamer (Aβ^*56) was the main oligomer being decreased under c-Abl inhibition (p < 0.05, T-test) (Fig. 2I). Aβ^*56 is a specific assembly of Aβ peptide that causes memory impairment long before amyloid plaques and neurodegeneration appear in animal models of AD [30]. These findings correlated with the pharmacokinetics of imatinib in wild type mice, which indicates that in spite of imatinib reported low penetrance, it reached concentrations close to the IC₅₀ in the brain (0.4μM) (Supplementary Figure 2). In addition, we evaluated the presence of imatinib in brain extracts of i.p injected mice using an in vitro kinase inhibition assay. This assay is based on the sensitivity of K562 to the inhibition of the Bcr-Abl tyrosine kinase. When these cells are exposed to c-Abl inhibitors they stop proliferating. We showed that brain extracts from mice injected with imatinib significantly reduced K562 cell proliferation compared to vehicle-injected mice (Supplementary Figure 3). To assess whether imatinib is reaching the brain we also measured the levels of phosphorylated c-Abl (p-cAbl) in the brain of AD-transgenic mice injected with imatinib. We found that the levels of phosphorylated c-Abl are reduced in the cortex of transgenic mice treated with imatinib. A similar trend, although not statistically significant, was observed for the c-Abl substrate pCRKII (Supplementary Figure 3).

Imatinib decreased AD features in transgenic mice

In AD, amyloid deposition is associated with a chronic inflammatory response in the brain, in which amyloid plaques are intimately associated with activated astrocytes and microglia. Thus, we decided to evaluate the effect of imatinib in brain astrogliosis. Astrocytes were immunolabeled with an anti-GFAP antibody, which recognizes the glial fibrillary acidic protein. APPswe/PSEN1ΔE9 mice displayed high GFAP immunolabeling in cortex and hippocampus, brain areas where amyloid deposition is especially abundant. Transgenic mice that had received imatinib showed less GFAP immunoreactivity in cortex and hippocampus than the control group (Fig. 3A). Quantitative analysis showed a significant decreased of GFAP-positive cells in imatinib-treated animals compared to animals receiving vehicle alone (p < 0.01, T-test) (Fig. 3B). Since memory loss constitutes the main feature of AD, we decided to evaluate the cognitive status of transgenic mice using a Morris water maze test as another indicator of pathology progression. Transgenic mice treated with imatinib exhibited significantly lower escape latencies than vehicle-injected animals, taking shorter times to find the hidden platform at days 4, 5, 8, and 9 of testing (p < 0.05, Two-way ANOVA) (Fig. 3C). These results indicate that c-Abl inhibition significantly impacted AD features associated with Aβ accumulation such as cognitive impairment and neuroinflammation.

Imatinib and the absence of the c-Abl protein favored the non-amyloidogenic processing of the amyloid-β protein precursor

As our results indicated that Aβ-oligomers are reduced in transgenic mice treated with imatinib, we took an in vitro approach to study whether this inhibitor was altering the amyloidogenic processing of the amyloid-β protein precursor (AβPP). Interestingly, we found that neuron-like cells (HT22) treated with imatinib had increased levels of total AβPP (Fig. 4A). Concordantly, we observed that neuronal primary cultures derived from rat embryos also showed increased levels of AβPP when treated with imatinib (Supplementary Figure 1A). To further study whether imatinib was affecting AβPP processing, we analyzed the proteolytic fragments resulted from secretase activity. Proteolytic cleavage of AβPP at the β-site by the cleaving enzyme BACE1 resulted in the formation of the soluble AβPP beta fragment (sAβPPβ) and of an AβPP carboxi-terminal fragment-β (βCTF). This last fragment is a substrate for the subsequent cleavage by γ-secretase to generate Aβ (amyloidogenic pathway). Alternatively, AβPP cleavage by the α-secretase enzyme generates an AβPP carboxi-terminal fragment-α (αCTF) and a soluble AβPP alpha fragment (sAβPPα), which results in a non-amyloidogenic pathway. The analysis of cell lysates treated with imatinib showed increased levels of total AβPP and/or sAβPPα, as these two forms cannot be set apart with the WO-2 antibody. However, this antibody does not detect the sAβPPβ fragment, as the WO-2 epitope is contained in the βCTF fragment (Fig. 4C). Concordantly, we found that imatinib reduced the levels of βCTF in cell cultures (Fig. 4C). In addition, when we analyzed conditioned media from cells treated with imatinib, we observed an increase of sAβPPα, suggesting that imatinib boosted the non-amyloidogenic pathway (Fig. 4C). We corroborated these findings by inhibiting the enzyme BACE1 in order to assess only the non-amyloidogenic pathway. sAβPPα was increased when imatinib is added to the culture (Supplementary Figure 1B). In contrast, when we inhibited the α-secretase, an expected increase in sAβPPβ was observed, a phenomenon that was reversed byimatinib treatment (Supplementary Figure 1C). Finally, to confirm the effect of imatinib over the amyloidogenic pathway, we evaluated the levels of secreted Aβ in HT22 cells. We found that Aβ peptide levels were significantly reduced in conditioned media from cells treated with imatinib in a concentration dependent manner, confirming that imatinib treatment overrided the amyloid-pathway in neuron-like cells (Fig. 4D). To rule-out the possibility that another target of imatinib, besides c-Abl, is behind the amyloid reduction observed, we analyzed the levels of βCTF in mice null for brain c-Abl (cAbl-KO mice) (c–Abl^loxp/c–Abl^loxp Nestin–Cre⁺ as a way to evaluate the impact of c-Abl over the amyloidogenic pathway. As c-Abl-KO mice do not produce Aβ oligomers because they carry the wild type AβPP and Presenilin genes, it was not possible to evaluate the presence of Aβ oligomers in plasma or brain. However, we detected a significant reduction in the levels of βCTF, indicating a decrease in the amyloidogenic processing of AβPP (Fig. 4D). Yet, we did not detect any significant reduction in the levels of BACE in the brain of c-Abl-KO mice compared with control mice (Fig. 4D). Altogether, our results suggest that the absence of c-Abl or its inhibition by imatinib, decreases the processing of AβPP through the amylodogenic pathway, which may explain the decrease of Aβ oligomers observed in vivo.

DISCUSSION

In recent years, the activation of c-Abl has been demonstrated in human neurodegenerative diseases [10 , 31–33]. The role of c-Abl has been broadly studied in AD, where its activation is one of the mechanisms underlying neuronal apoptosis in hippocampal cultures exposed to Aβ fibrils [10]. Besides AD, c-Abl is involved in other neurodegenerative diseases such as Parkinson’s disease (PD). C-Abl is increased in the striatum of PD patients. These patients also show an increment of tyrosine phosphorylated parkin [32, 33]. Treatment with a novel c-Abl inhibitor prevented the progression of neuronal damage in a mice model of PD [34], while the c-Abl inhibitor Nilotinib decreased neuronal loss and behavioral deficits in PD patients [35].

Imatinib is commercially available as Gleevec (Novartis) and approved by the FDA for chronic myelogenous leukemia and gastrointestinal stromal tumors. Imatinib is not associated with major adverse symptoms [36, 37], but one of its limitations as a treatment for AD is its poor penetration into the CNS [24]. Mice exhibit cerebrospinal fluid concentrations 155-fold lower than in plasma [24], and a leukemia patient treated with imatinib had a concentration of the drug 92-fold lower in the cerebrospinal fluid than in blood [38]. However, our pharmacokinetics studies indicate that imatinib reached IC₅₀ concentrations in the brain (Supplementary Figure 2), which could account for its described effects preventing neuronal death and memory impairments [11]. Moreover, we found that brain homogenates from imatinib-injected mice induced apoptosis of K562 cells, a cell line that is sensitive to the inhibition of the Bcr-Abl tyrosine kinase (Supplementary Figure 3). In addition, we observed a significant decrease in the levels of phosphorylated c-Abl in the cortex of AD-transgenic mice treated with imatinib, and levels of the c-Abl substrate pCRKII also tend to decrease after imatinib treatment (Supplementary Figure 3). These data suggests that imatinib crosses the blood-brain barrier, at least in concentrations close to the IC₅₀.

Interestingly, we found that imatinib remained in plasma for several hours after treatment (Supplementary Figure 2). Since Aβ is the major component of amyloid plaques, we evaluated plasmatic Aβ levels in the APPswePSEN1ΔE9 mice model injected with imatinib. Using conformational antibodies, we established that plasmatic levels of Aβ oligomers were significantly reduced. Interestingly, using a novel technique (AD-PMCA seeding-assay), we found a significant decrease in misfolded Aβ-oligomers with catalyzing capability in plasma samples from mice receiving imatinib. Since compelling evidence points to Aβ misfolding as the primary cause of the disease, the reduction of oligomers able to catalyze Aβ-misfolding could greatly impact the progression of AD. In fact, the brain of transgenic mice treated with imatinib shows a significant decrease in senile plaques, brain inflammation and Aβ accumulation, plus improvement in cognitive deficits. Two non-excluding pathways could explain these findings: i) Although imatinib has low brain penetration, it reaches concentrations close to the IC₅₀, inhibiting brain c-Abl and decreasing Aβ levels, thus diminishing formation of senile plaques. ii) Imatinib that remains in the periphery could still function slowing the levels of circulating Aβ-oligomers and displacing the Aβ equilibrium from the brain to the periphery, in a similar fashion to what has been proposed to explain the antibody-mediated clearance of Aβ during anti-Aβ immunotherapy (reviewed in [39]). This “peripheral sink hypothesis” speculates that anti-Aβ antibodies do not exert their effect in the brain but rather in the periphery, where they bind to circulating Aβ, reducing its concentration in the blood. Indeed, Aβ appears to be efficiently transported across the blood brain barrier and peripheral administration of labeled Aβ can be found bound to amyloid plaques in the brain, indicating an equilibrium between brain and circulating Aβ [40, 41].

On the other hand, Sutcliffe and coworkers proposed the liver as the origin of brain Aβ deposits, as the activity of the enzyme Psen2 was heritable in the liver but not the brain [42]. Our in vitro data indicate that the absence of c-Abl or its inhibition decreases the amyloidogenic processing of AβPP, reducing Aβ levels and oligomerization. The mechanism underlying this effect could involve AβPP phosphorylation. In cell cultures expressing the active form of c-Abl AβPP is phosphorylated on Tyr682 [43, 44], and the AβPP mutation Y682G shifts the processing of AβPP toward a non-amyloidogenic pathway [45]. Thus, we propose that AβPP phosphorylation could increase the interaction of AβPP with BACE. Imatinib might prevent the altered trafficking of Tyr682-phosphorylated AβPP to early endosomes and lysosomes, compartments where BACE is located. Alternatively, imatinib could counteract the increased interaction that occurs after Tyr682-phosphorylation between AβPP and BACE, through the modulation of adaptor proteins. Unpublished data from our group suggest that c-Abl interacts with AβPP in Niemann-Pick type C disease, interaction that depends on AβPP phosphorylation on Tyr682 (manuscript submitted).

In addition, a new target for imatinib has been described. The gamma-secretase activating protein (GSAP) is inhibited by imatinib, preventing the interaction of this protein with the gamma-secretase substrate thus affecting Aβ production [46]. However, no evidence of interaction between GSAP and the C-terminal fragment of AβPP was found in co-immunoprecipitation studies [47]. In addition, we did not observe significant changes in the processing of the holo AβPP protein in cell cultures when we transiently knock-down GSAP with a shRNA (Supplementary Figure 1D). Moreover, herein we showed that brain homogenates from c-Abl-KO mice has altered AβPP processing compared with control animals carrying wild type c-Abl, supporting a role for c-Abl on Aβ production.

Here, we show that c-Abl inhibition by imatinib had a dual effect on AD pathology, decreasing not only circulating-Aβ oligomers but also lowering Aβ deposits in the brain. Our results suggest that imatinib may boost the non-amyloidogenic processing of AβPP and supports the use of imatinib-like compounds in AD clinical studies. In addition, we propose a novel method for the detection of Aβ oligomers in blood, and to evaluate the effect of potential drugs for AD treatment.

Footnotes

ACKNOWLEDGMENTS

This work was supported by FONDECYT Postdoctoral Grant N° 3110052 and FONDECYT Initiation into Research Grant N° 11130561 to L.D.E. FONDEF Grant N° D10I1077 and FONDECYT Grant N° 1120512 and N° 1161065 to A.R.A. Basal Funding CONICYT-PFB-12/2007 (CARE) to A.R.A. and N.I.

Authors’ disclosures available online ().

Appendix

References

Alzheimer’s Association (2013) 2013 Alzheimer’s disease facts and figures. Alzheimers Dement 9, 208–245.

Citron

(2010) Alzheimer’s disease: Strategies for disease modification. Nat Rev Drug Discov 9, 387–398.

Hardy

, Selkoe

(2002) The amyloid hypothesis of Alzheimer’s disease: Progress and problems on the road to therapeutics. Science 297, 353–356.

Iqbal

, Grundke-Iqbal

(2006) Discoveries of tau, abnormally hyperphosphorylated tau and others of neurofibrillary degeneration: A personal historical perspective. J Alzheimers Dis 9, 219–242.

Querfurth

, LaFerla

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

Holtzman

, Goate

, Kelly

, Sperling

(2011) Mapping the road forward in Alzheimer’s disease. Sci Transl Med 3, 114ps148.

Masters

, Selkoe

(2012) Biochemistry of amyloid beta-protein and amyloid deposits in Alzheimer disease. Cold Spring Harb Perspect Med 2, a006262.

Salloway

, Mintzer

, Weiner

, Cummings

(2008) Disease-modifying therapies in Alzheimer’s disease. Alzheimers Dement 4, 65–79.

Delrieu

, Ousset

, Caillaud

, Vellas

(2012) ‘Clinical trials in Alzheimer’s disease’: Immunotherapy approaches. J Neurochem 120(Suppl 1), 186–193.

10.

Alvarez

, Sandoval

, Leal

, Castro

, Kosik

(2004) Activation of the neuronal c-Abl tyrosine kinase by amyloid-beta-peptide and reactive oxygen species. Neurobiol Dis 17, 326–336.

11.

Cancino

, Toledo

, Leal

, Hernandez

, Yevenes

, Inestrosa

, Alvarez

(2008) STI571 prevents apoptosis, tau phosphorylation and behavioural impairments induced by Alzheimer’s beta-amyloid deposits. Brain 131, 2425–2442.

12.

Moresco

, Koleske

(2003) Regulation of neuronal morphogenesis and synaptic function by Abl family kinases. Curr Opin Neurobiol 13, 535–544.

13.

Schlatterer

, Acker

, Davies

(2011) c-Abl in neurodegenerative disease. J Mol Neurosci 45, 445–452.

14.

Bradley

, Koleske

(2009) Regulation of cell migration and morphogenesis by Abl-family kinases: Emerging mechanisms and physiological contexts. J Cell Sci 122, 3441–3454.

15.

Wang

(2014) The capable ABL: What is its biological function? Mol Cell Biol 34, 1188–1197.

16.

Perez de Arce

, Varela-Nallar

, Farias

, Cifuentes

, Bull

, Couch

, Koleske

, Inestrosa

, Alvarez

(2010) Synaptic clustering of PSD-95 is regulated by c-Abl through tyrosine phosphorylation. J Neurosci 30, 3728–3738.

17.

Vargas

, Leal

, Estrada

, Gonzalez

, Serrano

, Araya

, Gysling

, Inestrosa

, Pasquale

, Alvarez

(2014) EphA4 activation of c-Abl mediates synaptic loss and LTP blockade caused by amyloid-beta oligomers. PLoS One 9, e92309.

18.

Head

, Pop

, Sarsoza

, Kayed

, Beckett

, Studzinski

, Tomic

, Glabe

, Murphy

(2010) Amyloid-beta peptide and oligomers in the brain and cerebrospinal fluid of aged canines. J Alzheimers Dis 20, 637–646.

19.

Georganopoulou

, Chang

, Nam

, Thaxton

, Mufson

, Klein

, Mirkin

(2005) Nanoparticle-based detection in cerebral spinal fluid of a soluble pathogenic biomarker for Alzheimer’s disease. Proc Natl Acad Sci U S A 102, 2273–2276.

20.

Fukumoto

, Tokuda

, Kasai

, Ishigami

, Hidaka

, Kondo

, Allsop

, Nakagawa

(2010) High-molecular-weight beta-amyloid oligomers are elevated in cerebrospinal fluid of Alzheimer patients. FASEB J 24, 2716–2726.

21.

Gao

, Yam

, Wang

, Magdangal

, Salisbury

, Peretz

, Zuckermann

, Connolly

, Hansson

, Minthon

, Zetterberg

, Blennow

, Fedynyshyn

, Allauzen

(2010) Abeta40 oligomers identified as a potential biomarker for the diagnosis of Alzheimer’s disease. PLoS One 5, e15725.

22.

Saborio

, Permanne

, Soto

(2001) Sensitive detection of pathological prion protein by cyclic amplification of protein misfolding. Nature 411, 810–813.

23.

Soto

, Estrada

, Castilla

(2006) Amyloids, prions and the inherent infectious nature of misfolded protein aggregates. Trends Biochem Sci 31, 150–155.

24.

Wolff

, Richardson

, Egorin

, Ilaria

Jr (2003) The CNS is a sanctuary for leukemic cells in mice receiving imatinib mesylate for Bcr/Abl-induced leukemia. Blood 101, 5010–5013.

25.

Kayed

, Head

, Sarsoza

, Saing

, Cotman

, Necula

, Margol

, Wu

, Breydo

, Thompson

, Rasool

, Gurlo

, Butler

, Glabe

(2007) Fibril specific, conformation dependent antibodies recognize a generic epitope common to amyloid fibrils and fibrillar oligomers that is absent in prefibrillar oligomers. Mol Neurodegener 2, 18.

26.

Chacon

, Barria

, Soto

, Inestrosa

(2004) [beta]-sheet breaker peptide prevents A[beta]-induced spatial memory impairments with partial reduction of amyloid deposits. Mol Psychiatry 9, 953–961.

27.

Maezawa

, Hong

H-S

, Liu

, Wu

C-Y

, Cheng

, Kung

M-P

, Kung

, Lam

, Oddo

, LaFerla

, Jin

L-W

(2008) Congo red and thioflavin-T analogs detect Aβ oligomers. J Neurochem 104, 457–468.

28.

Jankowsky

, Slunt

, Ratovitski

, Jenkins

, Copeland

, Borchelt

(2001) Co-expression of multiple transgenes in mouse CNS: A comparison of strategies. Biomol Eng 17, 157–165.

29.

Salvadores

, Shahnawaz

, Scarpini

, Tagliavini

, Soto

(2014) Detection of misfolded Abeta oligomers for sensitive biochemical diagnosis of Alzheimer’s disease. Cell Rep 7, 261–268.

30.

Lesne

, Koh

, Kotilinek

, Kayed

, Glabe

, Yang

, Gallagher

, Ashe

(2006) A specific amyloid-beta protein assembly in the brain impairs memory. Nature 440, 352–357.

31.

Jing

, Caltagarone

, Bowser

(2009) Altered subcellular distribution of c-Abl in Alzheimer’s disease. J Alzheimers Dis 17, 409–422.

32.

Imam

, Zhou

, Yamamoto

, Valente

, Ali

, Bains

, Roberts

, Kahle

, Clark

, Li

(2011) Novel regulation of parkin function through c-Abl-mediated tyrosine phosphorylation: Implications for Parkinson’s disease. J Neurosci 31, 157–163.

33.

, Lee

, Shin

J-H

, Karuppagounder

, Gadad

, Koleske

, Pletnikova

, Troncoso

, Dawson

(2010) Phosphorylation by the c-Abl protein tyrosine kinase inhibits parkin’s ubiquitination and protective function. Proc Natl Acad Sci U S A 107, 16691–16696.

34.

Imam

, Trickler

, Kimura

, Binienda

, Paule

, Slikker

, Li

, Clark

, Ali

(2013) Neuroprotective efficacy of a new brain-penetrating C-Abl inhibitor in a murine Parkinson’s disease model. PLoS One 8, e65129.

35.

Karuppagounder

, Brahmachari

, Lee

, Dawson

, Ko

(2014) The c-Abl inhibitor, Nilotinib, protects dopaminergic neurons in a preclinical animal model of Parkinson’s disease. Sci Rep 4, 4874.

36.

Druker

, Tamura

, Buchdunger

, Ohno

, Segal

, Fanning

, Zimmermann

, Lydon

(1996) Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat Med 2, 561–566.

37.

Capdeville

, Buchdunger

, Zimmermann

, Matter

(2002) Glivec (STI571, imatinib), a rationally developed, targeted anticancer drug. Nat Rev Drug Discov 1, 493–502.

38.

Takayama

, Sato

, O’Brien

, Ikeda

, Okamoto

(2002) Imatinib mesylate has limited activity against the central nervous system involvement of Philadelphia chromosome-positive acute lymphoblastic leukaemia due to poor penetration into cerebrospinal fluid. Br J Haematol 119, 106–108.

39.

von Bernhardi

(2010) Immunotherapy in Alzheimer’s disease: Where do we stand? Where should we go? J Alzheimer Dis 19, 405–421.

40.

Ghilardi

, Catton

, Stimson

, Rogers

, Walker

, Maggio

, Mantyh

(1996) Intra-arterial infusion of [125I]A beta 1-40 labels amyloid deposits in the aged primate brain in vivo. Neuroreport 7, 2607–2611.

41.

Mackic

, Bading

, Ghiso

, Walker

, Wisniewski

, Frangione

, Zlokovic

(2002) Circulating amyloid-beta peptide crosses the blood-brain barrier in aged monkeys and contributes to Alzheimer’s disease lesions. Vascul Pharmacol 38, 303–313.

42.

Sutcliffe

, Hedlund

, Thomas

, Bloom

, Hilbush

(2011) Peripheral reduction of beta-amyloid is sufficient to reduce brain beta-amyloid: Implications for Alzheimer’s disease. J Neurosci Res 89, 808–814.

43.

Zambrano

, Bruni

, Minopoli

, Mosca

, Molino

, Russo

, Schettini

, Sudol

, Russo

(2001) The beta-amyloid precursor protein APP is tyrosine-phosphorylated in cells expressing a constitutively active form of the Abl protoncogene. J Biol Chem 276, 19787–19792.

44.

Perkinton

, Standen

, Lau

, Kesavapany

, Byers

, Ward

, McLoughlin

, Miller

(2004) The c-Abl tyrosine kinase phosphorylates the Fe65 adaptor protein to stimulate Fe65/amyloid precursor protein nuclear signaling. J Biol Chem 279, 22084–22091.

45.

Barbagallo

, Weldon

, Tamayev

, Zhou

, Giliberto

, Foreman

, D’Adamio

(2010) Tyr(682) in the intracellular domain of APP regulates amyloidogenic APP processing in vivo. PLoS One 5, e15503.

46.

, Luo

, Li

, Remmers

, Netzer

, Hendrick

, Bettayeb

, Flajolet

, Gorelick

, Wennogle

, Greengard

(2010) Gamma-secretase activating protein is a therapeutic target for Alzheimer’s disease. Nature 467, 95–98.

47.

Hussain

, Fabregue

, Anderes

, Ousson

, Borlat

, Eligert

, Berger

, Dimitrov

, Alattia

, Fraering

, Beher

(2013) The role of gamma-secretase activating protein (GSAP) and imatinib in the regulation of gamma-secretase activity and amyloid-beta generation. J Biol Chem 288, 2521–2531.