Conceived and designed the experiments: MLB. Performed the experiments: SL MS ML MLB. Analyzed the data: MLB ML. Contributed reagents/materials/analysis tools: MLB. Wrote the paper: ML MLB MS SL.
The authors have declared that no competing interests exist.
NADPH oxidase is implicated in neurotoxic microglial activation and the progressive nature of Alzheimer's Disease (AD). Here, we test the ability of two NADPH oxidase inhibitors, apocynin and dextromethorphan (DM), to reduce learning deficits and neuropathology in transgenic mice overexpressing human amyloid precursor protein with the Swedish and London mutations (hAPP(751)SL).
Four month old hAPP(751)SL mice were treated daily with saline, 15 mg/kg DM, 7.5 mg/kg DM, or 10 mg/kg apocynin by gavage for four months.
Only hAPP(751)SL mice treated with apocynin showed reduced plaque size and a reduction in the number of cortical microglia, when compared to the saline treated group. Analysis of whole brain homogenates from all treatments tested (saline, DM, and apocynin) demonstrated low levels of TNFα, protein nitration, lipid peroxidation, and NADPH oxidase activation, indicating a low level of neuroinflammation and oxidative stress in hAPP(751)SL mice at 8 months of age that was not significantly affected by any drug treatment. Despite
Together, this study suggests that while hAPP(751)SL mice show increases in microglial number and plaque load, they fail to exhibit elevated markers of neuroinflammation consistent with AD at 8 months of age, which may be a limitation of this animal model. Despite absence of clear neuroinflammation, apocynin was still able to reduce both plaque size and microglial number, suggesting that apocynin may have additional therapeutic effects independent of anti-inflammatory characteristics.
Alzheimer's disease (AD) is a devastating and progressive neurodegenerative disease that culminates in dementia, affecting over 5 million people in the United States alone. Current treatment is largely unable to halt disease progression. The hallmark neuropathology of AD consists of insoluble extracellular plaques containing β -amyloid (Aβ) and intraneuronal neurofibrillary tangles in the cortical region of the brain. Microglia, the resident immune cells in the brain, have been implicated in the progressive nature of numerous neurodegenerative diseases, particularly AD
NADPH oxidase is an enzyme complex in phagocytes, such as microglia, that is activated during host defense to catalyze the production of superoxide from oxygen
The premise of deleterious microglial activation in AD has been supported by analysis of post-mortem brains from AD patients
Several compounds have demonstrated the ability to inhibit microglial NADPH oxidase, including memantine
In the current study, we addressed whether chronic administration of known NADPH oxidase inhibitors (apocynin and DM), beginning at the time plaque deposition began to occur in hAPP(751)SL mice, could prevent neuroinflammation, neuron damage, and behavioral learning and memory deficits.
Lipopolysaccharide (LPS; strain O111:B4) was purchased from EMD Chemicals (Gibbstown, NJ). Cell culture reagents were obtained from Invitrogen (Carlsbad, CA). HALT protease inhibitor was obtained from Thermo Fisher Scientific (Rockford, IL). Fluorescent Aβ peptide was purchased from AnaSpec, Inc. (Fremont, CA), and non-fluorescent Aβ was purchased from American Peptide Company (Sunnyvale, CA). Dextromethorphan, apocynin, staurosporine, fMetLeuPhe and all other reagents were purchased from Sigma Aldrich Chemical Co. (St. Louis, MO).
A total of 53 male transgenic hAPP(751)SL mice with a C57BL/6xCBA background were used for the
For
The hAPP(751)SL mice were randomly assigned to one of 4 treatment groups: vehicle, dextromethorphan (DM) 15 mg/kg, DM 7.5 mg/kg, or apocynin 10 mg/kg. Starting at 4 months of age (±2 weeks), animals were either treated with saline (vehicle, n = 14), DM 15 mg/kg (n = 13), DM 7.5 mg/kg (n = 12), or apocynin 10 mg (n = 14) by oral gavage daily for 4 months.
At the end of the 4 month treatment period, mice were trained in the MWM. The MWM tests took place in a black circular pool with a diameter of 100 cm filled with water and divided into four virtual quadrants. A transparent platform (diameter of 8 cm) was placed in the southwest quadrant of the pool. The walls surrounding the pool were marked with bold geometric symbols for spatial orientation. During behavioral testing, mice were placed in the pool and allowed to find the hidden platform. Ifthe mouse was unable to locate the platform, the investigatorguided the mouse. After each trial, mice were allowed to rest on the platform for 10–15 seconds and orient themselves. Mice performed three swimming trials per day for four consecutive days. During the trials, motion within the pool was detected with a computerized tracking system. These data were used to quantify swimming speed, escape latency (time, in seconds, for the mouse to find the hidden platform and escape the water), pathway (length traveled, in meters, before reaching the target), and abidance in the target quadrant (measured in percentage of the total trial time). Following the final trial on the fourth day, mice completed a ‘probe trial’ where the platform was removed and the number of crossings over the former platform position and abidance in the target quadrant were measured.
Following behavioral testing, animals were sacrificed and brain tissue was collected for further study. All mice were sedated using Isofluran inhalation before tissue collection. Mice were transcardially perfused with 0.9% saline and the brains were removed and divided into the right and left hemisphere. The left hemisphere was immediately processed for histology, while the right hemisphere was frozen on dry ice and stored at −80°C until use.
Lipid peroxidation in tissue samples was determined by the thiobarbituric acid reactive substances (TBARS) assay. Brain tissue was homogenized in 2.5% SDS with 5 mM butylated hydroxytoluene. 400 µL of this homogenate was mixed with 375 µL of 20% acetic acid, pH 3.5, and 225 µL of thiobarbituric acid (1.33%). The resulting mixture was incubated for 1 hour at 95°C. After incubation, 1 mL of 15∶1 butanol:pyridine was added and the mixture was centrifuged for 10 minutes at 4000 g. The amount of TBARS were determined by measuring the optical density of the organic layer at 535 nm and comparing the absorbance to a malondialdehyde (MDA) standard.
Membrane fractions from both cell culture and tissue were isolated using differential centrifugation followed by lipid extraction
From tissue samples, protein was isolated by suspending frozen tissue in 10 volumes of lysis buffer (Cytobuster Protein Extraction Reagent; EMD Chemicals; Darmstadt, Germany) with 10 µL/mL HALT protease inhibitor and 10 µL/mL EDTA. Samples were homogenized using a motorized pellet mixer and then centrifuged for 5 minutes at 5000 g. The protein concentration of the resulting supernatant was determined using a BCA protein assay (ThermoScientific; Rockford, IL).
Protein samples were resolved by SDS-PAGE on 10% gels. Protein was then transferred to nitrocellulose membranes, blocked for 1 hour in 5% milk, and incubated overnight at 4°C in primary antibody (mouse anti-GAPDH, rabbit anti-p47-phox, or rabbit anti-p67-phox; Millipore; Temecula, CA). Blots were then probed with horseradish peroxidase-conjugated secondary antibodies and visualized using enhanced chemiluminescence (GE Healthcare; Piscataway, NJ).
The production and release of TNFα was measured using 100 µg/well of whole brain homogenate with a commercial enzyme-linked immunosorbent assay (ELISA) kit from R&D Systems (Minneapolis, MN), as described previously
The amount of nitrated proteins was measured using 100 µg/well of whole brain homogenate with a commercial enzyme-linked immunosorbent assay (ELISA) kit from Millipore (Temecula, CA), per the manufacturer protocol.
One hemisphere from each mouse brain was fixed by immersion in a solution of 4% paraformaldehyde in PBS (pH 7.4; freshly prepared) at 4°C for 24 hours. After fixation, brains were transferred to a 15% sucrose/PBS solution for 24 hours. Brains were then frozen in dry-ice cooled Isopentane and stored at −80°C until use. Frozen brains were sectioned into 15-10 µm thick sections per level (5 levels) starting at the level of the total appearance of the dentate gyrus and according to Paxinos and Franklin
The presence of amyloid depositions was visualized immunohistochemically using an anti-β-amyloid antibody directed against amino acids 1–17 of the human β-amyloid peptide (Signet Laboratories; Dedham, MA) with a Cy3 secondary antibody (Jackson Laboratories; Bar Harbor, ME). Additionally, tissue sections were stained with ThioflavinS to recognize beta-sheet structures. Briefly, sections were washed in H2O for 3 minutes and then placed in 1% ThioflavinS for 7 minutes. Sections were then washed in 80% ethanol and PBS before incubating in 1% hydrogen peroxide in methanol at room temperature for 15 minutes. Sections were then blocked using MOM-blocking reagent and MOM-diluent according to the manufacturer's protocol (MOM-Kit; Vector Labs; Burlingame, CA). After blocking, samples were incubated with 6E10 antibody (Signet Laboratories; Dedham, MA) for 30 minutes at room temperature, washed with PBS, and incubated in 10% non-immune goat normal serum for 60 minutes at room temperature. Sections were then washed and incubated with Cy3 goat anti-mouse antibody (Jackson Laboratories; Bar Harbor, ME) for 60 minutes in the dark at room temperature. Finally, the sections were washed in PBS and H2O before adding coverslips.
Measurement of 6E10 and ThioflavinS staining was done using Image-Pro Plus software (MediaCybernetics). Briefly, an area of interest (AOI) was measured encompassing both the hippocampus and cortex of each section. Within this AOI, stained objects were detected that were over a threshold level of intensity and a size of 8.75 µm2. A measurement of the area of each object, sum of stained area, and the number of objects was made in each AOI. Mean plaque size was calculated by dividing the sum area of plaques by the total number of plaques. The plaque area percentage was measured by dividing the sum area of plaques by the region area and multiplying the result by 100.
To determine microglial activation in brain slices, slices were stained with CD11b antibody. Synaptic density was visualized by staining with a synaptophysin antibody in separate brain slices. For both antibodies, frozen brain sections were washed for 10 minutes in PBS and then for 4 minutes in 1 mg/ml sodium-borohydrate in PBS. Sections were then washed and treated with 1% hydrogen peroxide in methanol at room temperature for 10 minutes. Non-specific binding was then blocked with 10% horse serum for 30 minutes and MOM-diluent (Mom-Kit; Vector Labs; Burlingame, CA) for 5 minutes. Sections were then incubated with anti-CD11b antibody (Serotec; Raleigh, NC) or anti-synaptophysin antibody (Thermo Fisher Scientific; Fremont, CA) for 1 hour at room temperature. Samples were incubated with blocking reagent (10% non-immuno goat-normal serum for CD11b and Vectastain Elite ABC Kit (Vector Labs; Burlingame, CA) for synaptophysin) for 20 minutes and room temperature and then washed with PBS. CD11b samples were then incubated with Cy 3 goat anti-rat antibody (Jackson Laboratories; Bar Harbor, ME), washed, and then stained with DAPI and methanol (Sigma Aldrich Chemical Co.; St. Louis, MO) for 15 minutes to stain cell nuclei. Sections were washed in 80% ethanol followed by H2O before adding coverslips. After primary antibody and blocking of synaptophysin-stained samples, samples were washed with PBS and incubated for 30 minutes with Vectastain ABC Reagent (Vector Labs; Burlingame, CA), washed, and developed for 18 minutes with HistoGreen (Linaris; Bettingen, Germany). Tissues were then washed in TBS and H2O and dehydrated with a graded alcohol series and xylol before adding coverslips.
The number of microglia in each section was measured similarly to the protocol for 6E10 and ThioflavinS staining, except that the count only concentrated on CD11b staining that co-stained with the nucleus of the cell. Sections were co-stained with CD11b and DAPI, and cells were only counted in the AOI if their nucleus was within the name 10 µm thick section.
Synaptic density was also measured using Image-Pro Plus software (MediaCybernetics). Synapse number was counted at 1000-fold magnification from three images per region (CA1, CA3, and GDmb regions of the hippocampus). The total number of synapses was divided by the measured area (µm2) and averaged between the three images analyzed for each region.
Rat cortical neuron-glia cultures were prepared using a previously described protocol
Primary enriched microglia cultures were prepared from the whole brains of day-old Fisher 344 rat pups, using the procedure described previously
The rat microglia HAPI cells were a generous gift from Dr James R. Connor
The ability of cells to phagocytose β-amyloid peptide was measured using a protocol modified from Floden and Combs
Extracellular superoxide (O2−) production from microglia was determined as reported previously
Levels of hydrogen peroxide production in cell culture were determined as previously described, with slight modifications
Cell survival was measured using thiazole blue (MTT) to evaluate metabolic viability of cells
Microglial cell number was measured by taking microglia cell counts from mixed neuron-glia cultures treated for 24 hours with 10 ng/mL LPS or 2 µM Aβ with or without 100 µM apocynin. After treatment, cells were fixed in 3.7% formaldehyde, washed once with PBS, and treated with 1% hydrogen peroxide. Cells were then washed three times with PBS and blocked for one hour in PBS with 1% bovine serum albumin, 0.4% Triton X-100 and 4% goat serum. Plates were then incubated overnight at 4°C in a 1∶1000 dilution of anti-IBA-1 antibody (Wako Pure Chemical Industries, Ltd., Richmond, VA) in Dako antibody diluent (DAKO, Capinteria, CA). After incubation with primary antibody, cells were washed three times and incubated with Vectastain ABC Kit reagents according to the manufacturer's instructions (Vector Laboratories, Burlingame, CA). Images were taken on an AxioCam MRc5 imaging system (Carl Zeiss MicroImaging, Thornwood, NY). Cell numbers were quantified by counting nine representative areas per well in a 24 well plate at 100X magnification (an average of numbers counted by at least 2 individuals is reported).
Group differences in the behavioral tests were calculated using a parametric ANOVA with a Bonferroni's multiple comparison post-hoc test or a non-parametric Kruskal Wallis ANOVA with a Dunn's multiple comparison test if Gaussian distribution was missing. For assessments of learning deficits, a two-way ANOVA was used followed by Bonferroni's multiple comparison test. For
Brain slices from each group (vehicle, 15 mg/kg DM, 7.5 mg/kg DM, 10 mg/kg apocynin) were stained for two markers of Aβ deposition: 6E10 (measuring all Aβ peptide) and thioflavin S (measuring β-sheets of Aβ). This allowed for the measurement of plaque number, mean plaque size, and the percentage of area occupied by plaques in both the cortex and hippocampus. Using 6E10 staining, both the cortex and hippocampus display reduced plaque size in apocynin treated animals, compared to vehicle-treated controls (p<0.05;
Mice were treated daily with 15 mg/kg dextromethorphan (DM), 7.5 mg/kg DM, or 10 mg/kg apocynin for four months. The size of β-amyloid plaques was measured for each group and compared to control, vehicle-treated animals. Representative images show 6E10 staining of β-amyloid protein for each group in the cortex (A) and hippocampus (B), respectively. Quantification of plaque size the cortex (C) and hippocampus (D) revealed that only apocynin significantly decreased the size of plaques, compared to vehicle. DM, at either dose, did not alter plaque size in the cortex or the hippocampus. Plaque size was determined as the absolute plaque area divided by the absolute plaque number. *p<0.05 vs. vehicle, 1-way ANOVA with Bonferroni post-hoc test.
The number of microglia in both the cortex and hippocampus of hAPP(751)SL mice with DM or apocynin treatment was counted using CD11b immunoreactivity. Decreases in the number of microglia in the cortex was observed in mice treated with 10 mg/kg apocynin (p<0.05;
Mice were treated daily with 15 mg/kg dextromethorphan (DM), 7.5 mg/kg DM, or 10 mg/kg apocynin for four months. The number of microglia was then counted for each group by staining with anti-CD11b antibody and each treatment group was compared to control. CD11b-stained microglia were only counted if they corresponded to a DAPI stained nuclei (data not shown). Representative images from each group of the stained microglia are shown in panel (A). Apocynin reduced the number of microglia in the cortex of hAPP(751)SL mice, whereas neither dose of DM reduced microglia number (B). *p<0.05 vs vehicle, 1-way ANOVA with Bonferroni post-hoc test.
Behavioral deficits were measured in hAPP(751)SL mice by performance in the Morris Water Maze (MWM) through 3 daily trials over 4 consecutive days after 4 months of treatment with vehicle (0.9% saline), 15 mg/kg DM, 7.5 mg/kg DM, or 10 mg/kg apocynin. Overall performance in the MWM was determined by escape latency (seconds) and swimming path (meters). A downward trend was observed within groups on subsequent days (data not shown), indicating that each treatment group was able to learn and improve overall performance. No significant changes were seen in escape latency or swimming path between groups on any of the days (
Synapse density was measured in the CA1, CA3, and GDmb regions of the hippocampus with synaptophysin immunoreactivity. Neither DM nor apocynin altered synapse density of any of the regions examined (data not shown). This is consistent with the lack of behavioral changes seen in hAPP(751)SL mice treated with apocynin and DM.
The ability of DM and apocynin to reduce NADPH oxidase activation in hAPP(751)SL
Brain homogenates from each group (vehicle, 15 mg/kg DM, 7.5 mg/kg DM, 10 mg/kg apocynin) were used to measure the levels of lipid peroxidation using a TBARS assay. Levels of malondialdehyde (MDA) from each group were approximately 2.5 µM, suggesting low levels of oxidative stress. This is particularly interesting, as lipid peroxidation has previously been reported to increase significantly in post-mortem analysis of preclinical
Thus, neuroinflammation and oxidative stress were not readily apparent at 8 months of age in the hAPP(751)SL mice tested, which may explain why the known NADPH oxidase inhibitors failed to reduce these parameters. Together, these findings also indicate that although apocynin inhibited microglial number and plaque formation, it is very likely that it did so through mechanisms that are independent of anti-inflammatory and antioxidant properties. In addition, these findings also indicate that Aβ plaque load, microglia number, and learning deficits may occur independently of neuroinflammation and oxidative stress.
To confirm that both apocynin and DM were capable of inhibiting NADPH oxidase at all, we next tested their ability to reduce the production of extracellular ROS and neurotoxicity in response to Aβ. Both apocynin and DM were able to reduce the production of Aβ-induced extracellular superoxide to nearly control levels in primary microglia cultures (
(A) Enriched microglia cultures were treated with media alone (Control), apocynin (10 µM), Dextromethorphan (DM, 10 µM), Aβ (2 µM), Apocynin + Aβ, and DM + Aβ. The production of extracellular superoxide was measured by the superoxide dismutase (SOD)-inhibitable reduction of tetrazolium salt, WST-1 at 30 minutes post-treatment. Results are mean ± SEM. Data are from four separate experiments. *p<0.05, compared with control cultures. (B) Apocynin and DM protect against Aβ-induced toxicity in cortical neuron-glia cultures.) Cortical neuron-glia cultures were treated with media alone (Control), Apocynin (10 µM), Dextromethorphan (DM, 10 µM), Aβ (2 µM), Apocynin + Aβ, and DM + Aβ. Toxicity was assessed by MTT 7 days later. Graphs show the results expressed as percentage of the control cultures and are the mean ± SEM from three independent experiments in triplicate. * p<0.05, control compared to treatment.
We confirmed that apocynin was working as predicted by demonstrating that 30 minute pretreatment with apocynin will attenuate LPS-induced H2O2 production (
(A) Apocynin attenuates LPS-induced hydrogen peroxide (H2O2), as predicted. Microglia-enriched cultures were treated with Hank's balanced salt solution (HBSS), or HBSS with LPS (10 ng/mL), apocynin (100 µM), or the combination of apocynin (100 µM) and LPS for 3 hours. The level of H2O2 was then measured in each group and compared to control levels. Apocynin does significantly reduce LPS-induced increases in H2O2, returning levels to control values. *p<0.05 vs. control; #p<0.05 vs. LPS, 1-way ANOVA with Bonferroni post-hoc test. (B) Pre-treatment with 2 µM Aβ significantly reduces phagocytosis of fluorescent Aβ, and apocynin does not act to reverse this decrease. Microglia-enriched cultures were treated with control media, or media with β-amyloid (Aβ; 2 µM), apocynin (100 µM), or the combination of apocynin (100 µM) and Aβ (2 µM) for 24 hours. Fluorescently labeled Aβ (final concentration 0.1 µM) was then added to each well, and incubated with the cells for 6 hours to allow for phagocytosis of the fluorescent protein. The amount of phagocytosis of fluorescent Aβ was measured for each group and compared to control levels. *p<0.05 vs control, 1-way ANOVA with Bonferroni post-hoc test.
We then focused
(A) Microglia-enriched cultures were treated with control media, lip polysaccharide (LPS; 10 ng/mL), and/or Apocynin (100 µM) for 24 hours. Tumor necrosis factor alpha (TNFα) levels in the supernatant were measured via ELISA. LPS (10 ng/mL) significantly increased levels of TNFα and pre-treatment with 100 µM Apocynin significantly reduced the amount of TNFα released by microglia. *p<0.05 vs. control; #p<0.05 vs. LPS, 1-way ANOVA with Bonferroni post-hoc test. (B) Apocynin does not protect against inflammation-induced cell death. Microglia-enriched cultures were treated with control media, 1000 ng/mL LPS, 100 µM Apocynin, or LPS and apocynin for 24 hours. After incubation, cell survival was measured with the MTT assay. 100 ng/mL LPS significantly reduced microglial cell survival (through inflammation-induced cell death), which is not rescued by apocynin). (C) Microglia-enriched cultures were treated with control media, 2 µM staurosporine (SS), 100 µM Apocynin, or SS and apocynin for 24 hours. After incubation, cell survival was measured with the MTT assay. Data show that 2 µM SS significantly reduces microglial cell survival (through apoptosis), and is not reversed by the addition of apocynin. *p<0.05 vs control,1-way ANOVA with Bonferroni post-hoc test. (D) Apocynin does not alter Aβ or LPS-induced increases in microglia number
We also considered that the reduction of microglial cell number in the brains of hAPP(751)SL mice could be a result of increases microglial cell death or a reduction in proliferation. To test microglial cell survival in response to a number of toxic stimuli, and the effect that apocynin has on this response, primary microglia-enriched cultures were treated with 2 µM Aβ, 1000 ng/mL LPS (to cause inflammation-induced cell death), or 2 µM staurosporine (to induce apoptotic cell death) in the presence and absence of 100 µM apocynin. Neither Aβ, apocynin, nor the combination reduced cell survival
To look at the effect of Aβ and/or apocynin on microglial cell proliferation (to possibly account for the reduction in microglial number observed
Accumulating evidence indicates that the ideal therapeutic window for anti-inflammatory treatment targeting neurotoxic microglial activation may be early in the neurodegenerative process
Importantly, this study also addressed the utility of the hAPP(751)SL transgenic mouse model for testing anti-inflammatory compounds. The hAPP(751)SL mice over-express human APP Swiss and London mutations, with elevated expression in neurons throughout the brain, pronounced expression in the hippocampus, and little expression the periphery
However, despite the lack of evidence for NADPH oxidase-induced pathology
The reduction in cortex and hippocampus plaque size conferred by apocynin could be the consequence of a number of processes, including plaque phagocytosis, deposition, degradation, or APP processing and transport. As loss of microglial phagocytic function has been implicated as a key component to the development of plaques and AD progression
Another interesting finding emphasized by this work is the disconnect between plaque size and memory deficits in the hAPP(751)SL mice. While apocynin was able to reduce plaque size (
The
In summary, apocynin treatment for 4 months in hAPP(751)SL mice reduced plaque size and microglial number, resulting in brains that resembled younger mice.
(DOC)
(DOC)
(DOC)