Background
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that is the most common cause of dementia in the elderly [
1‐
3]. A major neuropathological feature of AD is the presence of neuritic plaques that are primarily composed of the 42-amino acid β-amyloid (Aβ) peptide, Aβ42. Aβ peptides are endoproteolytically processed from the amyloid-β precursor protein (APP) via sequential cleavages enacted by BACE1 (β-secretase) and γ-secretase [
4]. Most of the genetic factors known to increase susceptibility to Alzheimer’s disease increase the levels of Aβ in the brain (reviewed in [
3]). Though numerous studies have shown that cerebral accumulation of Aβ plays a critical early role in AD pathogenesis, the underlying mechanism by which Aβ causes neurodegeneration remains unclear.
Mitochondrial dysfunction has been implicated in age-related cognitive decline (reviewed in [
5]) and potentially plays a central role in the progression of Alzheimer’s disease. Pyramidal neurons, which require large amounts of cellular energy, are the most vulnerable to mitochondrial dysfunction [
6]. Compared to age-matched controls, AD patients have significantly higher levels of mitochondrial DNA (mtDNA) deletions in large vulnerable neurons of the hippocampus and neocortex [
7,
8]. Interestingly, Down syndrome patients with AD dementia also have an increase in mtDNA mutations [
9]. Aside from changes in mtDNA, AD mitochondria have decreased electron transport chain complex (ETC) IV, morphological changes in cristae, accumulation of osmiophilic material and decreased size [
8,
10].
There is evidence that Aβ-mediated toxicity may cause morphological, chemical and genetic changes in mitochondria.
In vitro studies show that the accumulation of Aβ impairs mitochondrial biogenesis, dynamics and axonal transport [
11‐
13]. Aβ also promotes oxidative stress, which can consequently introduce mutations in mtDNA and impair mitochondrial function [
14‐
16]. In an
in vivo study, severe toxicity was induced in the immediate vicinity of amyloid plaques, causing structural and functional abnormalities in mitochondria [
17].
In our study, we sought to determine how Aβ mediated toxicity and Aβ pathology interacts with mitochondrial dysfunction
in vivo. We modeled mitochondrial dysfunction
in vivo by utilizing a mouse that contains a knockin mutation that inactivates the proofreading function of mitochondrial DNA polymerase-γ (
PolgA D257A) [
18]. The proofreading activity of mitochondrial DNA polymerase-γ has been shown both in mice and human cells to be critical for preventing accumulation of mtDNA mutations with age (reviewed in [
19]). In humans, mutations in the
PolgA gene cause various central nervous system disorders including cognitive decline [
20]. In
PolgA D257A mice, previous reports show that the animals rapidly develop a myriad of mitochondrial bioenergetic defects in multiple tissues, including the brain, and physical phenotypes that mimic premature ageing, including increased mortality after 1 year of age [
18,
21‐
23].
We crossed the
PolgA D257A mice with a well-established transgenic AD mouse model carrying the APP familial London mutation (APPV717I; APP/Ld). APP/Ld mice do not exhibit neuron loss but develop amyloid plaques at ~1 year of age [
24]. Since age is the greatest risk factor in Alzheimer’s disease and
PolgA D257A mice exhibit a premature aging phenotype, we investigated whether
PolgA D257A; APP/Ld bigenic mice may model the interaction between mitochondrial dysfunction associated with aging and Aβ toxicity in the onset and progression of AD. We hypothesized that mitochondrial dysfunction may affect the balance between Aβ synthesis and clearance, thus contributing to amyloidogenesis and potentially triggering neurodegeneration. Here, we provide evidence that the
PolgA D257A mutation may increase β-amyloid accumulation by reducing IDE levels and thus impairing Aβ-clearance. In contrast, levels of the Aβ-generating enzymes BACE1 and PS1 or the APP C-terminal fragments (CTFs) C99 and C83 do not change in the presence of the
PolgA D257A mutation. We also provide morphological and biochemical evidence of neurodegeneration in mice expressing the
PolgA D257A mutation and APP/Ld transgene characterized by cortical and hippocampal atrophy and neuronal swelling and vacuolization. Our results suggest synergism between mitochondrial dysfunction and cerebral Aβ accumulation in some aspects of brain atrophy and neurodegeneration. These findings lend insights into the roles of mitochondrial dysfunction, amyloid pathology, and - more broadly - ageing in AD pathogenesis.
Discussion
Age is the primary risk factor for AD, yet the mechanism responsible for this association remains enigmatic. Reduced mitochondrial function is hypothesized to occur during aging and in AD (reviewed in [
5]) and may result from age-associated accumulation of point mutations and deletions in mtDNA (reviewed in [
49]). Neuronal populations are potentially more vulnerable to mitochondrial dysfunction due to their high bioenergetic demands, thus providing a possible mechanistic link between aging, mitochondrial dysfunction and AD.
Despite strong evidence that cerebral Aβ accumulation plays an early part in AD pathogenesis, the precise mechanism of Aβ neurotoxicity remains elusive. Several APP transgenic mouse models of AD successfully recapitulate key features of the disease such as amyloid plaque pathology, but fail in other aspects such as neurodegeneration (reviewed in [
50]). Interestingly, intracerebral injection of fibrillar Aβ in aged but not young primates resulted in neuronal loss, suggesting that Aβ neurotoxicity is a pathological response of the aging brain [
51]. In our study, we employed mitochondrial DNA mutator
PolgA D257A mice that accumulate mtDNA mutations with age to test the hypothesis that age-related mitochondrial dysfunction can exacerbate amyloid pathology and induce neurodegeneration in APP transgenic mice.
A major finding in our study was that both Aβ42 levels and amyloid plaque load were increased in the brains of D257A; APP/Ld mice when normalized to transgenic APP/Ld protein levels. Therefore, the D257A mutation exacerbated the cerebral accumulation of Aβ42 per given amount of APP in the bigenic mice. Although trends toward increased absolute Aβ42 levels and amyloid plaque density were observed in the bigenic mice, they became statistically significant when normalized to transgenic APP/Ld levels (Figures
1 and
2). Our immunoblot analysis for human APP revealed that there was close to 40% reduction in APP/Ld protein levels in D257A; APP/Ld bigenic mice compared to the parental APP/Ld monogenic mice. In contrast, endogenous mouse APP levels were unaffected in D257A monogenic mice, demonstrating that the
PolgA mutation alone does not alter APP expression levels. We speculate that the reduced levels of APP/Ld protein in brains of bigenic mice might relate to neurodegenerative processes that were occurring prior to neuron death. Bigenic neurons appeared to be in a severe state of degeneration (Figure
6) and showed caspase-3 activation and increased p25 levels (Figure
7). Given their state of pathology, degenerating D257A; APP/Ld neurons subject to Aβ neurotoxicity could well have reduced APP/Ld transgene transcription or translation. Alternatively, APP/Ld in D257A; APP/Ld mice could have undergone cleavage by increased levels of activated caspase-3 [
52], thus lowering total APP/Ld levels. Future studies will be necessary to clarify the mechanism of decreased APP/Ld levels in D257A; APP/Ld bigenic brains.
The exacerbated cerebral accumulation of Aβ in D257A; APP/Ld mice could have resulted from either increased production or decreased clearance/degradation of Aβ. Because we did not observe any changes in the levels of BACE1, PS1, C99, and C83 in D257A; APP/Ld bigenic versus APP/Ld mice (Figure
3), all evidence supporting increased Aβ production in the brains of the bigenic mice is lacking. It should be noted that BACE1 levels were significantly increased in the brains of D257A; APP/Ld and APP/Ld compared to D257A and wild-type mice, consistent with our previous work showing that BACE1 levels are elevated by amyloid pathology in human AD and APP transgenic mouse brains [
32,
33,
53]. Nevertheless, the lack of support for a D257A-associated increase in Aβ production suggested that impaired Aβ clearance/degradation might have led to the increase of Aβ accumulation in D257A; APP/Ld brain. To test this hypothesis, we analyzed cerebral levels of two major Aβ degrading enzymes, IDE and NEP. Consistent with findings from other APP transgenic mice [
38,
39], levels of IDE, but not NEP, were increased in response to Aβ accumulation in the brains of APP/Ld mice (Figure
4). This Aβ-induced increased IDE level might reflect a feed-back mechanism that attempts to counteract the build-up of Aβ in the brain. Additionally, IDE strongly prefers to localize within GFAP-positive astrocytes surrounding Aβ plaques [
39]. Most importantly, in the D257A; APP/Ld bigenic mice, we observed that the D257A mutation completely abrogated the Aβ-induced increase in IDE level. Our study is the first to show
in vivo that mitochondrial dysfunction, by the D257A mutation, can prevent the Aβ-induced IDE increase. Interestingly, a previous report indicated that the D257A mutation can affect metabolism by changing levels of major hormones such as leptin and ghrelin [
54]. Therefore, it is not unprecedented that the D257A mutation can affect enzymes in metabolic pathways, like IDE. Additionally, D257A; APP/Ld mice phenotypically mimic aspects of advanced aging and increased mortality at time of analysis at 12 months of age. Thus, the abrogated IDE increase in these mice might represent an aging-related effect [
55]. Because Aβ peptide levels have been shown to correlate inversely with IDE levels
in vivo [
37,
56], the blocked Aβ-induced IDE increase could be responsible for the exacerbated amyloid accumulation in the bigenic mice.
Previous studies of APP/Ld transgenic mice show absence of substantial brain atrophy [
24,
57], implying that Aβ accumulation on its own does not cause significant neurodegeneration. Indeed, our study supports a two-hit hypothesis whereby Aβ is insufficient to cause substantial neurodegeneration by amyloid buildup alone. An additional stress factor, such as mitochondrial dysfunction, is proposed to combine with Aβ to trigger the anticipated cascade of events leading to neurodegeneration.
A growing body of work from two independent knockin mouse lines has demonstrated that the
PolgA D257A mutation causes increased accumulation of mtDNA mutations that leads to progressive mitochondrial dysfunction with age [
18,
21,
22,
58‐
63]. Our work suggests the possibility that Aβ toxicity might exacerbate mitochondrial dysfunction associated with the D257A mutation, thus causing brain atrophy. Future studies should be undertaken to investigate whether Aβ can worsen pre-existing age-related mitochondrial deficits and their effects on neurodegeneration.
Hematoxylin staining of coronal brain sections revealed that only the bigenic D257A; APP/Ld mice exhibited a smaller overall brain size with shrinkage of the cortex and hippocampus, compared to the other genotypes (Figure
5). Additionally, cortical neurons in the bigenic mice revealed abnormal nuclear morphologies reminiscent of neurons undergoing apoptosis [
40,
64]. In contrast, cortical neurons of D257A and APP/Ld mice appeared morphologically similar to those of wild-type mice. A time-course study examining morphological changes in the brains of D257A; APP/Ld compared to the other genotypes will be important for determining when these abnormalities appear.
We also observed that the D257A; APP/Ld mice bigenic mice exhibited significantly increased levels of activated caspase-3 17 kDa fragment compared to the other genotypes (Figure
7). Activated caspase-3 levels were also increased in APP/Ld mice, but were less elevated compared to the bigenic mice. We corroborated our biochemical analysis with immunohistochemical staining of brain sections that revealed activated caspase-3-positive puncta in layer V/VI cortex in D257A; APP/Ld and APP/Ld mouse brains. Interestingly, in D257A mice post-mitotic tissues such as the brain have been shown to be more resistant to the induction of apoptosis by mtDNA mutations [
18]. However, the bigenic mice carrying the D257A mutation appeared to have reduced resistance to apoptosis. These results are significant because activated caspase-3 immunoreactivity has been observed in AD brain [
42] and in other APP transgenic mouse models [
43,
65]. Although the underlying mechanism of brain atrophy in the D257A; APP/Ld mice is not known, it might involve Aβ42-induced mitochondrial cytochrome c release and subsequent activation of caspase-3 to trigger apoptosis [
66]. Importantly, significant activation of caspase-3 can lead to eventual neuron loss (reviewed in [
40,
67,
68]). Additionally, the levels of p25, a marker of neurodegeneration, were the highest in bigenic mice compared to the other genotypes (Figure
7).
Although we did not observe frank neuron loss in D257A; APP/Ld mice by cell counting (Figure
8), we suspect that neurons were in the process of dying, given the brain atrophy, vacuolated neurons, activated caspase-3, and elevated p25 levels of the bigenic mice.
PolgA D257A knockin mice do not live long past one year [
18], and we observed a dramatic increase in mortality of D257A; APP/Ld mice at that age. Therefore, we speculate that we would have observed frank neuron loss in the bigenic mice had they been able to live longer. We hypothesize that the observed brain atrophy and increased markers of neurodegeneration are primarily the result of axon and dendrite degeneration, synaptic loss, white matter loss, or a combination of these factors, which precede neuron death.
Finally, oligomeric forms of Aβ are widely thought to underlie important aspects of synaptic dysfunction and neurodegeneration in AD (reviewed in [
69]). Although we did not examine Aβ oligomers in the current study, the possibility exists that the D257A mutation might affect the proportion or toxicity of Aβ oligomers in the brains of D257A; APP/Ld bigenic mice. It will be important to conduct future studies to address the potential influence of the D257A mutation on Aβ oligomer level, form, and neurotoxicity.
Methods
Mice
APPV717I (APP/Ld) mice were generated and characterized previously [
24]. These mice were crossbred to the
PolgAD257A/D257A mitochondrial DNA mutator mice [
18]. At 12 months of age, the mice were deeply anesthetized with ketamine/xylazine and transcardially perfused with PBS containing protease (Cocktail Set III, Animal-free, Cat#: 535140, Calbiochem) and phosphatase (Halt Phosphatase Cocktail Prod#: 1861277 Thermo Scientific) inhibitors. Brains were excised and one hemibrain was drop fixed in 4% PFA and cryopreserved in 30% sucrose, PBS for immunohistochemistry, while the other hemibrain was flash frozen in liquid nitrogen for biochemical analysis. All mice were maintained in microisolator cages in the Barrier Facilities of Northwestern University Center for Comparative Medicine. All animal procedures were in strict accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were Northwestern University Animal Care and Use Committee approved.
Tissue preparation for biochemical analysis
Hemibrains were flash frozen in liquid N2 and stored at -80°C. Frozen mouse hemibrains were homogenized in 1× PBS, 1% Triton X-100, 1× protease inhibitor cocktail (Calbiochem), and 1× Halt phosphatase inhibitor cocktail (Thermo Scientific). Total protein concentration was determined by the BCA method (Pierce).
Human Aβ42 ELISA
Total Aβ42 levels in brain homogenates were determined using a human Aβ (1-42) ELISA kit (Wako Pure Chemical Industries, Ltd.), according to manufacturer’s recommendations. Briefly, APP/Ld and D257A; APP/Ld brain homogenates (~1 mg total protein) was extracted for four hours in 5 M guanidine-HCl at room temperature (300 μl total volume). Homogenates were further diluted 10-fold in ice-cold casein buffer (0.25% casein/0.05% sodium azide/5 mmol/l EDTA, pH 8.0 in PBS with 1× protease inhibitor cocktail (Calbiochem), centrifuged at 16,000 g for 20 min at 4°C, and then diluted again 1:50 with Standard diluent. Samples were run in duplicates on Aβ42-specific ELISA (100 μl/well). Optical densities (450 nm) of each well were read on a Spectra Max 250 plate reader (Molecular Devices Corp., Sunnyvale, CA), and sample Aβ42 concentrations were determined by comparison with the Aβ42 standard curves. Each reading was conducted in the linear range of the assay. Aβ42 concentration values were normalized to total brain protein concentrations and were expressed as nanograms of Aβ42 per milligram total protein, and the average of the duplicates was defined as the Aβ42 concentration for a given mouse.
In order to normalize Aβ42 concentration to APP transgenic protein level for each mouse, the Aβ42 concentration was divided by a normalization factor (NF) calculated for a given mouse. To do so, immunoblots were performed on mouse brain homogenates to measure full-length human APP levels using the 6E10 antibody (see Immunoblotting Section). Following enhanced chemiluminescence detection, the raw sum intensities of APP immunoblot band signals for all mice were determined and normalized to Ponceau S staining. Next, the mean of the normalized APP immunoblot signal intensities for the APP/Ld mice as a group was determined. The APP immunoblot signal intensity for each APP/Ld and D257A; APP/Ld mouse was then divided by the APP/Ld mean intensity to derive an individual NF for every mouse. Finally, a given Aβ42 ng/mg total protein ELISA value was divided by the individual NF for each mouse to calculate the mouse’s Aβ42 concentration normalized to APP transgenic protein level, and then means and SEMs for each genotype were determined.
Immunoblotting
In general, 10-30 μg protein from whole-brain homogenates were heated at 95°C in sample boiling buffer (60 mM Tris HCl, 5% SDS, 10% glycerol, pH = 6.8) with 5× loading dye (90% BME, 10% of 5% Bromophenol blue dye) before SDS-PAGE separation on 4%–12% NuPAGE Bis-Tris gels in 1× MES or MOPS running buffer (Invitrogen, Carlsbad, CA). Alternatively, samples for PS1-NTF immunoblot were not boiled but were otherwise prepared similarly. For C99 and C83 immunoblots, samples were separated on 16% Tris-glycine gels using 1× Tris-glycine as a cathode buffer and 200 mM Tris-base (pH 8.8) as an anode buffer. Proteins were electrophoretically transferred onto Millipore Immobilon-P polyvinylidene difluoride (PVDF) membrane (Millipore, Billerica, MA). Multiple protein gels were run in parallel to accommodate the large number of samples. When conducting the protein transfer step, gels were aligned horizontally and transferred onto a single piece of PVDF membrane. This process ensured consistency as all samples then were developed on the same PVDF membrane. After completing protein transfer, PVDF membranes were stained with Ponceau S and imaged on a scanner. Blots were then blocked in 5% non-fat dry milk in Tris-buffered saline (TBS), 0.1% Tween 20 (TBST; Sigma) for 1 h at room temperature, then incubated in primary antibody (human APP: Aβ Mouse mAb 6E10, Chemicon Cat. # MAB1560, 1:1,000; human and mouse APP: Anti-Alzheimer Precursor Protein A4 Mouse mAb N-term specific, 22C11, Millipore Cat. # MAB348SP
OR Rabbit mAb APP C-terminus, Y188, Epitomics Cat. # 1565-1, 1:5000; BACE1: Mouse mAb, 3D5 [
32], 1:1000; C99 and C83: Rabbit mAb APP C-terminus, Y188, Epitomics Cat. # 1565-1, 1:5000; PS1: α-PS1NTF Rabbit pAb, a generous gift from Dr. Gopal Thinakaran (U. Chicago), 1:5000; Insulin Degrading Enzyme: α-IDE Rabbit pAb, Abcam Cat. # ab32216, 1:1000; Neprilysin: Anti-CD10 Rabbit mAb, EPR2997, Abcam Cat. # ab79423, 1:1000; Cleaved Caspase-3: Rabbit mAb, 5A1E, Cell Signaling Cat. # 9664, 1:1000; Caspase-3 (full-length): Rabbit pAb, Cell Signaling Cat. # 9662, 1:1000; p25 and p35: Rabbit pAb, C-19, Santa Cruz Biotechnology Cat. # sc-820, 1:1000) for 2 hrs at RT or overnight at 4°C. Blots were washed in TBST and incubated for 1 h in horseradish peroxidase (HRP)-conjugated goat anti rabbit (Jackson ImmunoResearch Laboratories, West Grove, PA) or horse anti-mouse (Vector Laboratories) secondary antibodies diluted 1:10,000 in 5% milk in TBST. Immunosignals were detected using enhanced chemiluminescence (EMD Millipore Luminata Classico, Crescendo, or Forte) and quantified using a Kodak Image Station 4000R imager (Rochester, NY). Densitometric analyses of immunoblots and images of Ponceau S-stained blots were performed using Kodak Molecular Imaging Software SE. Immunosignals were normalized to the measured intensity of whole lane Ponceau S staining. Values were expressed as percentages of the mean of the control.
Immunofluorescence microscopy of tissue sections
Coronal sections of 30 μm were cut on a freezing sliding microtome and were selected with equivalent rostral-caudal locations using anatomical landmarks and the size and shape of the hippocampus and ventricles. Free-floating sections were washed 3× (10 min each) in TBS + 0.25% Triton-× 100 (TBS + T) and blocked for 90 min in 5% goat serum with TBS + T. Then they were washed 2× (10 min each) in 1% BSA with TBS + T before being incubated in anti-Aβ42 C-terminus specific rabbit polyclonal antibody (Invitrogen, Cat. No. 44-344, 1:1000) solution at 4°C overnight on an orbital shaker.
The sections were washed again the next day 2× (each 10 min) in 1% BSA with TBS + T before being incubated in secondary antibody solution of Alexa-Fluor Donkey anti rabbit-488 (A21206 along with DAPI (300nM) and thiazine red (1:60 k). Sections were washed in TBS in a dark room before mounting with ProLong gold antifade reagent (Life Technologies #P36934) and coverslipping #1.5 (VWR). Sections were imaged with a 10× air objective of a Keyence BZ-9000 Series microscope. Images were stitched using Keyence proprietary software. High magnification images of plaques were acquired on UV LSM510 laser scanning confocal microscope using laser lines of 405 nm (blue) 488 nm (green) and 561 nm (red).
Amyloid plaque density
The total number of Aβ42-positive plaques was counted manually from one coronal section per mouse taken from the same mid-rostral-caudal location in the brain. 10-12 mice of each genotype (APP/Ld monogenic or D257A; APP/Ld bigenic) were used for plaque counting. The plaque density was then determined by dividing the total number of Aβ42-positive plaques by the total area of the section analyzed per mouse. The mean number of Aβ42-positive plaques per cm2 was then calculated for a given genotype. Additionally, mean plaque density was normalized to transgenic human APP protein level by dividing the plaque density of each mouse by that mouse’s respective NF value (see Human Aβ42 ELISA methods). SEMs and p-values were determined using the two-tailed Student’s t-test.
Coronal sections with equivalent mid-rostral-caudal locations in the brain were selected for histological staining and all related measurements. Hematoxylin staining was performed according to the manufacturer’s protocol (Vector Labs). Sections were dried on Superfrost Plus micro glass slides (VWR) for ~10 minutes to facilitate adhesion. Once stuck to the slide, the sections were then rehydrated with a few drops of water. The sections were then stained with Vector Hematoxylin QS for about 1 minute, and then rinsed with tap water until water washed over the sections was colorless. Afterward, sections were dehydrated in a series of alcohols (70%, 95%, 95%, 100%, and 100%), cleared in xylene, and coverslipped with Permount mounting medium (Fisher Scientific, SP15).
On hematoxylin stained sections, lateral cortical thickness was measured by the length of a line running perpendicular to the cortical layers connected from the tip of the CA1 region to the visible boundary of the cortex. Higher magnification images of the cortex stained with hematoxylin was acquired on Leica M165 FC or Zeiss Axioskop/CRi Nuance camera microscopes. Using images acquired on a Nikon Eclipse E800 microscope, the hippocampal area was measured by drawing a freehand selection around the hippocampus in ImageJ software (NIH).
Neuron density (number of neurons per mm
2) was determined by counting the number of neurons in middle layers of the dorsal cortex in 30 μm thick hematoxylin stained coronal brain sections. Three sections spaced 15 sections apart that had the same relative rostral-caudal positions (centered over the mid-hippocampus; Figure
5A) in each mouse were chosen and two representative fields per section were counted using a Zeiss Axioskop microscope with a 100× oil objective in bright-field. All neurons (as identified by their large round nuclei) were counted in each 100× field, while glial cells (as identified by their small dense nuclei) were not counted. A total of 5 mice per genotype were counted and the mean neuron number per genotype was determined and converted to neuron density.
The Vector Laboratories (Burlingame, CA) ABC kit was used with DAB as chromogen to visualize the reaction product of cleaved caspase-3. On Day 1 of staining, sections were first incubated in 0.4% Triton at room temperature for 30 min. The sections were briefly rinsed in PBS before and after each incubation step, unless otherwise specified. The sections were subsequently incubated in vehicle/Serum (0.1% Triton X-100, 3% Goat Serum in PBS) at 4°C on shaker for 1 h. Sections were then incubated in H2O2 (1% in PBS) at 4°C for 1 h, followed by incubation in the primary antibody (anti-cleaved caspase 3, Rabbit mAb, 5A1E, 1:500), diluted in 1% BSA in TBS with 0.25% Triton X-100. On Day 2, sections were incubated in secondary antibody (Vector, Goat anti-Rabbit-HRP, 1:500) at 4°C for 2 hrs, followed by incubation in Reagent A + B (1% of each reagent in Vehicle solution) at 4°C for 2 hrs. The reagent solution was made at least 30 minutes prior to use. The sections were submerged in urea (120 mg/ml diluted in H2O)—no shaking—then briefly rinsed 2× in PBS followed by 3× in 0.05 M TRIS. Two development solutions (0.07% H2O2 in H2O; 1 mg/ml Diaminobenzidine in 0.05 M TRIS) were made and then mixed just before the development step. Sections were incubated in the development solution from 30 sec to 5 min until sufficient degree of color development occurred by visually inspecting the sections under a microscope. Finally, the sections were briefly rinsed in 0.05 M TRIS, dried overnight, dehydrated in a series of alcohols, cleared in xylene and coverslipped with Permount mounting medium.
Statistical analysis
Data are presented as means and standard errors of the mean (SEMs, represented by error bars in histograms). N-values are stated in figure legends. GraphPad Prism (GraphPad Software, Inc., San Diego, CA) was used for all statistical analysis. The statistical significance between means of experimental and control groups was determined using the two-tailed Student’s t-test. For comparisons involving more than 2 groups, one-way ANOVA was used followed by post hoc Newman-Keuls multiple comparison test (*p < 0.05, **p < 0.01, ***p < 0.001).
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
LK bred the mice, performed the experiments, and wrote the manuscript. RV conceived of the study, participated in its design and coordination, and edited the manuscript. GCK and TAP provided PolgAD257A/D257A mitochondrial DNA mutator mice, and FVL provided APPV717I (APP/Ld) mice. All authors participated in the interpretation of results, read and approved the final manuscript.