Benzo(a)pyrene exposure induced neuronal loss, plaque deposition, and cognitive decline in APP/PS1 mice

The APPswe/PSEN1^ΔE9 (also called APP^K670N,M671L/PSEN1^ΔE9 or APP^K595N,M596L/PSEN1^ΔE9) transgenic (APP/PS1) mice and their wildtype (WT) littermates were obtained from Model Animal Research Center of Nanjing University. All housing and breeding of the animals were performed in strict accordance with the guidelines of the Animal Care and Committee of Nanjing University Animal Center.

To investigate spatial learning and working memory abilities, mice were tested in the Morris water maze [16]. The Morris water maze consisted of a large circular plastic pool (diameter 120 cm), which was divided into four quadrants. A circular platform (diameter 10 cm) was located 1 cm below the surface of the opaque water, and the water temperature was kept at 22 °C. Four geometric images, namely a circle, square, triangle, and irregular shape with different colors were stick on the pool wall as visual spatial cues. At the acquisition phase, all mice were trained four trails per day for 5 consecutive days with the submerged platform in a fixed position. For each training trial, the mouse was given 60 s to find the platform. Any mouse that did not reach the platform within 60 s was led to the platform by the experimenter and allowed the mouse to stay on the platform for 10 s. After 24 h from the last training trial, the mice were given a 60-s probe trial without the platform in the pool. Swimming activity of each mouse was tracked via a camera linked to a computer monitoring system. All the parameters were calculated by ANY-maze software.

The open field test measures exploratory behaviors and anxiety-related behaviors. The open field box consisted of four-square blue arenas, 40.5 cm × 40.5 cm × 35.5 cm. The center area was scored as 20 cm × 20 cm area in the center of the open field. The mice were habituated to the behavioral room for 30 min prior to experimental sessions. During the test, the mouse was placed in the center of an arena and left to explore the environment freely for 20 min. Between trials, the arena was cleaned with 70% ethanol and ddH₂O. All the activities of each mouse were tracked via a camera linked to a computer monitoring system. Average speed, distance traveled, central zone entries, and time spend in the central area were calculated by ANY-maze software.

After behavior tests conducted one month after 2-month BaP exposure, the mice were anesthetized using 5% chloral hydrate and perfused transcardially with 0.9% saline; then, brains were harvested. To conduct immunohistochemistry and double-label immunofluorescence, mouse brains were fixed in 4% (w/v) paraformaldehyde (PFA) and then cryopreserved in 30% (w/v) sucrose/phosphate buffer saline (PBS), and stored at 4 °C. The brains were serially sectioned at 10 μm in the coronal plane using Leica VT1200S Fully Automated Vibrating Blade Microtome, and sections were stored in PBS. For biochemical assay and gene array, brain regions were dissected and immediately stored at the − 80 °C. While proteins extracted from the cerebral cortex and the hippocampus were used for Western blot analysis, the total RNA isolated from cortical tissues was used for gene expression profiling.

Immunostaining was performed as previously described [17, 18]. Briefly, PFA-fixed cell cultures or brain sections containing the cerebral cortex and the hippocampus were treated with 3% hydrogen peroxide for 10 min followed by three washes with PBS and incubated with blocking solution containing 4% goat serum, 0.4% Triton X-100, and 1% bovine serum albumin (BSA). Brain sections were then incubated overnight at 4 °C with primary antibodies. Human Aβ were detected with 6E10 (1:500, BioLegend, USA, CAT: 803002) antibody. Neurons were stained with an antibody specific for NeuN (neuron-specific nuclear-binding protein; a neuron-specific marker) (1:200, Millipore, USA, CAT: MAB377) or tyrosine hydroxylase (TH, a marker of dopaminergic [DA] neurons) (1:1000, Millipore, MA, CAT: AB152). Microglia and astrocytes were detected with anti-Iba1 (ionized calcium-binding adaptor molecule 1; a microglia marker) (1:500, Wako, Japan, CAT: 019-19741) and anti-GFAP (glial fibrillary acidic protein, as an astroglia marker) (1:800, Millipore, USA, CAT: MAB360) antibody, respectively. Brain sections were incubated with biotinylated anti-rabbit (1:500, Vector Laboratories, USA, CAT: BA-1000) or anti-mouse IgG (1:500, Vector Laboratories, USA, CAT: BA-2000) antibody for 1 h followed by incubation with avidin-biotin complex reagents (Vector Laboratories, USA) for 1 h, and color was developed with 3,3′-diaminobenzidine. Digital images of brain sections were acquired on an Olympus DP72 microscope. Digital images of cultured cells were recorded with a CCD (charge-coupled device) camera.

The optical density of NeuN and GFAP immunoreactivity was measured and analyzed by using Image J software, and a mean value was then deduced by averaging the density of a series of 12 sections that covered the entire cortex/hippocampus. Results were indicated as percentage of WT-vehicle group. The number of Iba-1⁺ cells from a series of 12 brain sections that covered the entire cortex/hippocampus was counted by two investigators individually in a double-blind manner. For visual enumeration of the immunostained cells, images from ten areas per well were randomly taken and the TH-IR neurons were counted. For each experiment, three to four wells per treatment condition were used and results from three to four independent experiments were obtained.

To detect Aβ plaques, brain sections containing the cortex and the hippocampus mounted onto glass slides were rinsed in PBS three times and incubated in the humidity chamber for 5 min with the solution of 0.5% thioflavin T (Sigma-Aldrich, USA, CAT: T3516) solution in 0.1 N HCl. Brain sections were then briefly rinsed in PBS and ddH₂O and cover-slipped with aqueous mounting media. Thioflavin T-stained plaques were viewed with Olympus BX53 semi-motorized fluorescence microscope (excitation wavelength, 488 nm).

The cortex and the hippocampus were gently homogenized in cold RIPA buffer (Radioimmunoprecipitation assay buffer; 50 mM Tris-HCl, pH 8.0, 5 mM EDTA, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, and protease inhibitor cocktail) on ice, centrifuged at 175,000×g for 1 h at 4 °C and the soluble supernatant was collected. The protein concentration was determined with Pierce BCA assay kit (ThermoFisher, USA). Equal amounts of total protein (50 μg per lane) were separated on 4~12% Bis-Tris-polyacrylamide electrophoresis gel and transferred to polyvinylidene difluoride membranes. Membranes were blocked with 5% nonfat milk and incubated with the following primary antibodies overnight at 4 °C: anti-NeuN (1:1000, Millipore, USA, CAT: MAB377), anti-Aβ (6E10; 1:1000, BioLegend, USA, CAT: 803002), anti-GFAP (1:1000, Millipore, USA, CAT: MAB360), anti-gp91 (1:1000, BD, USA, CAT: 611414), anti-iNOS (1:1000, Santa Cruz, USA, CAT: sc-7271), and anti-β-actin (1:1000, Cell Signaling Technology, USA, CAT: 8457). The membrane was then incubated with horseradish peroxidase-linked anti-rabbit (1:2000, Cell Signaling Technology, USA, CAT: 7074) or anti-mouse IgG (1:2000, Cell Signaling Technology, USA, CAT: 7076) for 1 h. The blots were detected with enhanced chemiluminescence (ECL) reagent. Images were analyzed by Image J software.

The gene expression profile related to Alzheimer’s disease was quantified using Alzheimer’s disease gene expression profile array (GeneCopoeia, USA). The gene primer pairs were used for the mRNA expression. In short, the total RNA was isolated from cortical tissues using Trizol reagent following the manufacturer’s direction. An aliquot of RNA (1 μg) was reverse-transcribed into cDNA using the First Strand cDNA synthesis kit (GeneCopoeia, USA). The final cDNA product was used for array using SYBR-Green-based real-time PCR according to the manufactures’ protocol. The amplification was run in the CFX Connect Real-Time PCR Detection system (Bio-Rad, USA). A dissociation curve was used to verify that majority of fluorescence detected was attributed to the labeling of specific PCR products. Relative gene mRNA expression ratios between groups were calculated using the ΔΔCt formulation. All the data were normalized to that of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in the same sample.

Sprague-Dawley (SD) rats were obtained from the Animal Center of Peking University Health Science Center. Mesencephalic primary neuron-glia cultures were prepared from the ventral mesencephalon of embryonic day 14 ± 0.5 SD rats as described before [18]. Briefly, neuron-glia cultures were maintained in MEM (Minimum Essential Medium) supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 10% heat-inactivated horse serum (HS), 1 g/l glucose, 2 mM L-glutamine, 1 mM sodium pyruvate, and 0.1 mM nonessential amino acids. Seven-day-old cultures were treated with vehicle or desirable reagents in the treatment medium (MEM containing 2% FBS, 2% HS, 2 mM L-glutamine and 1 mM sodium pyruvate). At the time of treatment, the neuron-glia cultures were made up of 10% microglia, 50% astrocytes, and 40% neurons of which 3–4% were TH-immunoreactive (IR) neurons.

Synthesized human Aβ_1-42 peptide (China peptides, China) was purchased to prepare aged Aβ solution as described before [19]. The Aβ peptide was equilibrated to room temperature for 30 min before dissolved in 100% 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP, Sigma) to 1 mM. After being incubated for 2 h at room temperature, the peptide was dried under vacuum to form peptide films. The peptide film was dissolved in cold sterile phosphate butter saline (PBS) to 100 μM, and after vortexed for 30 s, Aβ solution was incubated at 37 °C for 24 h to form the aged Aβ, which contained Aβ monomer, oligomers, and fibrils.

The production of superoxide was determined by measuring the superoxide dismutase (SOD)-inhibitable reduction of WST-1 (water-soluble tetrazolium 1) as previously described [20]. Neuron-glia cultures grown in 96-well plates were washed and refilled with 150 μl of phenol red-free treatment medium. Fifty microliters of HBSS (Hank's Balanced Salt Solution) containing vehicle or corresponding treatment reagents were then added to each well, followed by addition of 50 μl of WST-1 (1 mM) in HBSS with or without 600 U/ml SOD. The cultures were incubated for 30 min at 37 °C, and the absorbance at 450 nm was read with a SpectraMax Plus microplate spectrophotometer (Molecular Devices, CA).

Data were presented as the mean ± standard error (SE). Differences of parameters in Morris water maze and open filed were detected using repeated measures ANOVA. Statistical significance of other parameters between multiple groups was performed using one- or two-way analysis of variance (ANOVA). When ANOVA showed significant difference, least significant difference (LSD) multiple comparisons post-hoc test was performed. A value of P < 0.05 (two-tailed) was considered statistically significant. Statistical analyses were conducted by SPSS 13.0 software (Chicago, IL, USA).

Pathways and functional systems of the differentially modulated genes were further analyzed. Among the highlighted categories, inflammation and immunoregulatory process, Aβ secretes, and degradation and synaptic formation pathways/systems emerged as the most robustly altered in AD progression after BaP exposure. As shown in Table 1, almost all the Aβ secretion-related proteins tested in this study were significantly increased in APP/PS1 mice after BaP exposure but not in WT mice. For instance, in APP/PS1 mice, beta-secretase 1 (Bace1, as known as beta-site APP cleaving enzyme 1), presenilin proteins (Psen1/2), and aph-1 homolog b (Aph1b) were raised to 2.93-, 1.88-, and 3.20-folds respectively after BaP exposure. However, mild changes of these proteins in WT-BaP mice (Table 1) suggested that BaP facilitated abnormal Aβ secretion in APP/PS1 mice, which was in concordance with accelerated/exacerbated Aβ burden observed in APP/PS1-BaP mice (Fig. 5c). Interestingly, Aβ degradation-related proteins such as membrane metalloendopeptidase (MME), plasminogen activator (Plau), and insulin degrading enzyme (IDE) were also elevated in APP/PS1-BaP mice, but not in WT-BaP mice, which is consistent with deteriorated Aβ burden in APP/PS1 mice after BaP exposure.

The gene expression profiles related to synaptic formation were also remarkably changed in BaP-treated mice. Agrin, known to be important for synaptic differentiation, formation, and/or maintenance [21–23], was significant decreased in APP/PS1 mice after BaP exposure (Table 1). Altered expression and abnormal distribution of agrin in AD brain contribute to AD pathogenesis [21]. Slc6a4, a gene encoding SERT (serotonin transporter), was significantly downregulated in APP/PS1 mice with or without BaP exposure (Table 1). The brain of AD patients show reduced SERT in both protein and mRNA, and reduced SERT is associated with depression and anxiety in AD [24, 25]. Acetylcholinesterase (AChE) was remarkably upregulated in WT and APP/PS1 mice after BaP exposure (Table 1). Studies have shown increased AChE activity in plasma and elevated AChE levels around amyloid plaques in AD patients [26, 27]. Interestingly, neurotropic factors BDNF (brain-derived neurotrophic factor) and GNDF (glial cell line-derived neurotrophic factor) were upregulated in WT and APP/PS1 mice, which could be protective responses to BaP-induced neuronal damages. These findings together suggested that BaP exposure might induce synaptic loss/dysfunction.

It is notable that neuroinflammatory responses were the pathway with the most robust alteration after BaP exposure in both WT and APP/PS1 mice. Neuroinflammatory cytokines (e.g., TNFα, IL3, TREM2, IL1β, and IL6) were slightly increased in WT-BaP mice, but remarkably upregulated in APP/PS1-BaP mice, which indicated persistent neuroinflammation in BaP-induced APP/PS1 mice. In addition, multiple subunits of NADPH oxidase including gp91, p67, p40, and p47 were remarkably upregulated. In particular, gp91 (the catalytic subunit of NADPH oxidase) was upregulated by 3.19-folds and 5.71-folds in WT-BaP mice and APP/PS1-BaP mice, respectively (Table 1). Taken together, these alternation patterns of gene expression in WT and APP/PS1 mice after BaP exposure indicated that BaP could initiate AD-related pathological and biochemical pathways to precipitate in AD progression. Notably, sustained neuroinflammatory responses may play a crucial role in BaP-/Aβ-induced AD-like phenotypes. Neuroinflammation has been believed to be the key molecular event in the initial phase of AD; We next investigated how neuroinflammation and NADPH oxidase activation affected BaP’s aggravation of Aβ-induced AD progression.

NADPH oxidase has been shown to play a pivotal role in Aβ-induced neurotoxicity [28–31]. Our array results showed significant upregulation of multiple subunits of NADPH oxidase in the cortex of BaP-exposed mice (Table 1). Western blot results also revealed upregulation of gp91 (the catalytic subunit of NADPH oxidase) in the cortex/hippocampus of BaP-exposed mice. Therefore, we investigated whether upregulated NADPH oxidase mediated Aβ/BaP-induced neurodegeneration. We first measured extracellular superoxide production in neuron-glia cultures treated with Aβ and/or BaP. BaP and aged Aβ showed synergistic effects on extracellular superoxide production (Fig. 7c). Moreover, pre-treatment with NADPH oxidase inhibitor apocynin (0.25 mM) for 30 min before treatment with aged Aβ alone or in combination with BaP only blocked extracellular superoxide production but also neuronal loss, which were detected at 1 day or 7 days after Aβ/BaP treatment respectively (Fig. 7c, d). Neuroprotective effects of NADPH oxidase inhibitor on neurodegeneration induced by BaP and aged Aβ indicated that the upregulation and the activation of NADPH oxidase and resultant oxidative stress were important mediators of BaP/Aβ-induced neuronal loss seen in both in vitro and in vivo studies.

Additional file 1: Supplemental Table S1. Mouse Alzheimer’s Disease RT2 Profiler PCR Array.