APP/PS1 Gene-Environmental Cadmium Interaction Aggravates the Progression of Alzheimer’s Disease in Mice via the Blood-Brain Barrier,Amyloid-β,and Inflammation

Abstract

Background:

There is limited information about gene-environment interaction on the occurrence and the progression of Alzheimer’s disease.

Objective:

To explore the effect of environmental low-dose cadmium (Cd) exposure on the progress of Alzheimer’s disease and the underlining mechanism.

Methods:

We administered 1 mg/L, 10 mg/L cadmium chloride (treated groups), and water (control group) to C57BL/6J and APP/PS1 mice through drinking water, from one week before mating, until the offspring were sacrificed at 6 months of age. The behaviors, Cd level, blood-brain barrier (BBB) leakage, Aβ_1-42 deposition, and inflammation expression were evaluated in these mice.

Results:

Mice of both genotypes had similar blood Cd levels after exposure to the same dose of Cd. The toxic effects of Cd on the two genotypes differed little in terms of neuronal histomorphology and BBB permeability. Cd caused a series of pathological morphological changes in the mouse brains and more fluorescent dye leakage at higher doses. Furthermore, the APP/PS1 mice had more severe damage than the C57BL/6J mice, based on the following five criteria. They were increasing anxiety-like behavior and chaos movement, spatial reference memory damage, Aβ plaque deposition in mouse brains, increasing microglia expression in the brain, and IL-6 higher expression in the cortex and in the serum.

Conclusion:

Low-dose Cd exposure for 6 months increases Aβ plaque deposition and BBB permeability, exacerbates inflammatory responses, and activates microglia, in APP/PS1 mice. APP/PS1 gene-environmental Cd interaction aggravates the progression of Alzheimer’s disease in mice.

Keywords

Alzheimer’s disease APP/PS1 gene blood-brain barrier cadmium gene-environment interaction inflammatory response

INTRODUCTION

Alzheimer’s disease (AD) is a common age-related neurodegenerative disease. It affects millions of individuals globally. 5-10% of individuals worldwide aged 65 and above suffer from AD [1]. Although the exact etiology of AD remains unclear, a potential adjunct for determining AD risk over a longer timeframe lies in genetic factors. Early-onset AD (onset before 65) has been linked to strong patterns of familial inheritance, and causative mutations have been described on the amyloid precursor protein (APP) gene, presenilin-2, and presenilin-1 [2, 3]. For late-onset AD, several genetic loci have been identified by genome-wide association studies as potential risk factors such as BIN1, CLU, and CR1 [4, 5]. However, these prevalent risk genes impart only a minute increase in risk [6], and the etiology of AD has yet to be explained by genetic factors alone.

It has been hypothesized that exposure to various environmental factors raises the risk onset and progression of AD. Prolonged exposure to various well-known environmental factors including heavy metals [7], air pollutants (particulate matter), pesticides, and industrial chemicals accelerates the progression of AD [8]. Cadmium (Cd) is a ubiquitous environmental pollutant, identified as a human carcinogen (group I) by the International Agency for Research on Cancer (IARC) [9]. It is the main route of Cd intake through eating food and drinking water [10]. Epidemiological investigation has shown a negative correlation between cognitive score and blood Cd levels in individuals over 60 years old [11]. An interquartile range (IQR; IQR = 0.51 ng/mL) increase in urinary Cd was associated with 58% higher rate of AD mortality (mean follow-up: 7.5 years) in NHANES 1999-2006 participants [12]. Several animal studies have reported that Cd is highly permeable and can cross the blood-brain barrier (BBB) and collect in developing rat brains, ultimately increasing the BBB permeability [13 –15]. However, these studies have often treated adult animals with high concentrations of Cd; the effect of short-term high-dose exposure may be completely different from that of long-term low-dose exposure.

Mounting evidence suggests that neurodegenerative diseases are the result of gene-environment interaction [16]. The gene-environment interaction between genetic risk and environmental exposure may perturb learning and memory, and accelerate cognitive decline [17]. However, studies on the gene and environmental interactions of AD are limited. The Developmental Origin of Health and Disease (DOHaD) theory proposes that the risk of developing non-communicable diseases later in life may be related to exposures during the developmental period [18]. Developmental life is a vulnerable period in the lifespan during which adverse environmental factors have the potential to disturb the processes of cell proliferation and differentiation [19]. Yet the embryonic development stage, as the starting point of individual growth, has been overlooked.

AD is characterized by a gradual decline in cognitive functions, memory impairment, and behavioral changes [20]. It is believed that the pathophysiological process of AD begins many years before its diagnosis [21] and that it conveys different mechanistic pathways. Abnormal amyloid-β peptide (Aβ) deposition and neurofibrillary tangles are regarded as the critical pathological hallmarks of AD [22]. The misfolded and aggregated proteins are recognized by astroglia and microglia cells, and trigger an innate immune response, which contributes to disease progression and severity [23]. Soluble Aβ may cause neuronal membrane damage, producing reactive oxygen and nitrogen species [24]. Amyloid-β protein precursor (AβPP) is a type I transmembrane protein expressed in many cell types, including neurons [25]. Evidence suggests that AβPP processing is a critical event in the onset and progression of AD [26]. The AβPP protein is not neurotoxic, and it needs to be hydrolyzed into Aβ_1-40 and Aβ_1-42 fragments by BACE-1 [27]. Therefore, the expression levels of the APP and BACE-1 genes influence Aβ production in the body. The LRP-1 protein is an important ligand protein that is present in the BBB and is responsible for transporting Aβ between the brain and the blood [28, 29]. It also influences the Aβ transport process.

Numerous studies have suggested that BBB disruption is an important factor in AD pathogenesis. Aging and blood vessel impairment can disrupt the structure and function of the BBB by causing chronic cerebral hypoperfusion, leading to an imbalance in brain homeostasis [30 –32]. This has been confirmed by MRI scans of AD patients, and the detection of related proteins in brain tissue [31, 32]. At the same time, it has been confirmed that in AD mouse models, BBB leakage occurs before Aβ plaque appears, and then contributes to AD pathogenesis [33]. The BBB is a closed barrier formed by tight junctions of cerebral endothelial cells that prevent toxic and harmful macromolecules from the blood from entering the brain, while also expelling neurotoxic substances and pathogens from the central nervous system [34]. Tight junctions in cerebral endothelial cells include the Occludin protein family and the Zonula occludins protein family [35, 36]. Occludin is highly expressed in brain endothelial cells, and evidence indicates that it may be a regulatory protein with a crucial role in paracellular permeability [37]. Zonula occludens (ZOs) are cytoplasmic membrane-associated accessory proteins that connect the cytoplasmic tails of Claudins and Occludin to the actin cytoskeleton to maintain the tight junction structure [38]. Occludin and ZO-1 are considered sensitive indicators of normal and disturbed functional states of the BBB [38].

Neuro-inflammation is another pathological change in AD pathogenesis. In both AD patients and transgenic mice, many microglia are activated in the brain, and numerous markers of inflammatory response have been detected [39]. Microglia is an immune cell within the central nervous system that can eliminate harmful substances from the brain. During infection, injury, and brain disease, microglia change from a quiescent to an activated state, and migrate to the damaged site to exert innate immune effects [40]. It is generally accepted that there are two phenotypes after microglia activation—a pro-inflammatory phenotype (M1) and an anti-inflammatory phenotype (M2). In the early stages of AD pathogenesis, microglia will transform into the M2 phenotype and secrete anti-inflammatory factors such as IL-10 and IL-4. In later stages of AD pathogenesis, microglia convert into an M1 phenotype and secrete the pro-inflammatory factors IL-1β, TNF-α, and IL-6, thus reducing microglia’s phagocytic capacity and damaging neurons [41].

APP/PS1 double transgenic mice (APP/PS1 mice) carry human APP and PS1 genes. They are often used as animal models for Alzheimer’s type dementia. APP/PS1 mice often develop amyloid plaque at 5-6 months old, and show typical symptoms of AD early in life [42].Thus we could observe any changes in the early stages of the disease if mice were sacrificed at 6 months of age. C57BL/6J mice have been regarded as wildtype mice, and are used in many research fields including genetics, immunology, and neurobiology [43]. Therefore, we regarded APP/PS1 genes as genetic risk factors, and Cd exposure from the embryo as an environmental factor for evaluating whether there is gene-environment interaction between environmental Cd exposure and the APP/PS1 gene during AD pathogenesis such as abnormalities in behavioral tests, Aβ accumulation, BBB permeability, and neuroinflammation, in terms of cognitive behavior and AD risk.

MATERIALS AND METHODS

Animals and treatments

Three-month-old C57BL/6J mice (Strain #000664| Common Name: B6) and three-month-old B6; C3-Tg hemizygotes mice (APPswe, PSEN1dE9, MMRRC Strain #034829-JAX|Common Name: APP/PS1) were purchased from the Guangdong Medical Laboratory Animal Center, China. APP/PS1 transgenic mice have C57BL/6J; C3 H genetic background and exhibit no seizure phenotype. The double-transgenic mice express chimeric mouse/human amyloid precursor protein (Mo/HuAPP695swe) and mutant human presenilin1 (PS1-dE9).

Females and males were mated at a ratio of 2 : 1 for both the C57BL/6J mice and the APP/PS1 mice, and they were given either filtered sterile distilled water or water with 1 mg/L and 10 mg/L CdCl₂ (analytical purity, Sigma-Aldrich, USA) one week before mating. To prevent injury to the gastrointestinal tract caused by long-term gavage, we poisoned the mice via their drinking water. The drinking water was replaced weekly. Females were examined daily for their vaginal plug and were separated into cages after pregnancy. The pregnant females of the F0 generation were exposed to Cd until the weaning of the offspring. After weaning, pups with the same genotype and similar weight were selected. The pups received 0, 1, or 10 mg/L CdCl₂ in their drinking water, the same as the F0 generation. For each genotype and Cd exposure group, there were 12 mice (5∼7 females and 5∼7 males) in total. Mice of the same-sex were housed in one cage, with 3∼4 mice in each cage. Different genotypes and Cd exposure mice were housed separately throughout. As APP/PS1 mice often develop amyloid plaque at 5-6 months old, mice were sacrificed at 6 months of age to observe any changes in the early stages of the disease. Cd exposure via drinking water continued throughout pregnancy, lactation, weaning, and adulthood, until the end of the experiment.

All animals were housed in climate-controlled rooms at 20-22°C and 50% -70% humidity with controlled lighting (12 h light and 12 h dark). Food and water were given ad libitum. All experiments were approved by the Sun Yat-sen University Animal Care and Use Committee, and complied with the Institute Animal Ethics Guidelines.

Open field test

To analyze locomotion and anxiety, we conducted an open field test, as in previous research [44]. One hour before the test, we placed the mice in the testing room for acclimatization. The mice were placed in a 40 cm×40 cm×40 cm box, with a central area in the middle of the box marked with a permanent marker (20 cm×20 cm). The mice were removed from their cages and placed in the central zone of the open field box. They were allowed to freely explore the area for 5 min, and the data was collected by TopScan (Clever Sys Inc.; Reston, Virginia, USA). We recorded the total distance covered, the time and the distance spent in the center area. After each mouse finished the experiment, we used 75% ethanol to sterilize the box.

Y-maze test

Y-maze can be used to assess spatial working and reference memory in mice. The procedures and dimension of the circular pool used in the experiment were based on the method reported by Kraeuter et al. [45].

Spontaneous alternation, a measure of spatial working memory, can be assessed by allowing mice to explore all three arms of a maze. It is driven by the rodents’ innate curiosity to explore previously unvisited areas. The Y-maze has three 30 cm arms with congruent angles between each of them. We placed each mouse at one arm of the Y-maze and allowed it free movement throughout the maze for 8 min. Spontaneous alternation was defined as entry into all three arms on consecutive choices. We calculated the percentage of spontaneous alternations as: $\begin{matrix} % Alternation = (Number of alternations / \\ [total number of arm entries - 2]) \times 100 \end{matrix}$

A mouse with intact working memory will remember the arms previously visited, and be more likely to enter a less recently visited arm.

Spatial reference memory can also be tested by placing the test mice into the Y-maze with one arm closed off during training. This arm is designated as the novel arm. During the training period, one arm was closed off and the mice were allowed free movement between the other two arms for 10 min. After 10 min, the mice were returned to their home cages and the novel arm was opened. After an inter-trial interval (2 h in our study), we placed the mice back into the maze and allowed them to move freely throughout the three arms for 5 min. The entries and times for each arm were recorded. In theory, the mice should have remembered which arm they had not explored and visited it more often.

Cerebrospinal fluid, blood, and brain sample collection

We collected cerebrospinal fluid (CSF) as described in Lim et al.’s study [46]. The mice fasted for 12 h and anesthetized with 1% sodium pentobarbital. Immediately following anesthetizing the mice, CSF from the cisterna magna was collected with glass capillaries. To avoid blood contamination, we centrifuged the CSF samples at 13,000 g for 30 s, and stored the supernatant at -80°C until use.

After the CSF was collected, we pumped 100μl of heparin anticoagulated blood and 100 of μl non-anticoagulated blood from the hearts. The non-anticoagulated blood was centrifuged at 4,000 rpm, 4°C for 15 min. The serum and blood clots were separated and stored at -80°C. Then we perfused the left ventricle with 20 ml of ice-cold PBS to remove any remaining blood.

We randomly selected three mice from each group to fix the brains in 4% paraformaldehyde (PFA). Half of each brain was fixed in 4% PFA overnight at 4°C. Then the tissues were dehydrated in 15% sucrose in PBS solution at 4°C until the tissues sank. We embedded them in OCT (Sakura, Japan), and stored them at -80°C until use. The other hemisphere of each brain was fixed in 10% formaldehyde for 24 h and dehydrated by an automatic dehydrator and embedded in paraffin at room temperature.

Another three mice from each group were randomly selected for FITC perfusion to detect any leakage in the BBB. We slowly perfused FITC-Dextran solution (Mr = 70 kDa, Sigma, USA) into the mice (0.625 g/kg body weight) under low light conditions. The mice were kept at 4°C for 45 min after perfusion, and then were perfused with some of the ice-cold PBS to remove any residual fluorescent dye. After perfusion, we collected the mouse brain tissue, embedded it in OCT (Sakura, Japan), and stored it at -80°C until use.

The brains of the other six mice in each group were divided into cortex and hippocampus, frozen in liquid nitrogen, and stored at -80°C. These brain tissues were used to detect ICP-MS and RNA.

Determining Aβ_1-42 levels in CSF

5-10μl CSF was collected from the fourth ventricle of every mouse, thawed at 4°C, and diluted for use. The Aβ_1-42 levels in the CSF were quantified with an ELISA kit provided by EIAab Science (E1151 m). All procedures were performed as described by the manufacturer’s protocol.

Measuring Cd levels by ICP-MS

Cd measurement followed the methods in the literature [8]. For blood Cd analysis, we added 100μl of whole blood into a tube, and then added a mixture solution of 0.1% TritonX-100/0.1% HNO₃ to 2.0 ml of the solution. After mixing, we centrifuged the samples for 15 min (3,000 rpm) and collected the supernatant for analysis. For brain Cd detection, we collected an appropriate amount of frozen mouse brain tissue and recorded the weight. Then, we dissolved it with 1 ml of acid consisting of a mix of HClO₄ and HNO₃ at a ratio of 1 : 3, at a temperature of 70°C, until the liquid was clear and transparent. We diluted the digestive solution to 10 ml and used the supernatant for detection after centrifugation. Finally, we determined the Cd concentrations in different samples by inductively coupled plasma mass spectrometry (ICP-MS; iCAP RQ, Thermo, USA). We purchased the standard Cd solution from ANPEL (Shanghai, China). It was diluted with 2% HNO₃ to obtain appropriate concentrations (0.01, 0.1, 1.0, 5.0, 10.0, 20.0μg/L). ICP-MS operating conditions are shown in Supplementary Table 1.

Determining inflammatory cytokines (IL-1β, TNF-α, and IL-6) in the serum

We removed the serum from storage in -80°C and placed it into a 4°C refrigerator to thaw. Then we diluted some of the serum 4-5 times to detect serum TNF-α (EMC102a, NeoBioscience) levels. The procedure was the same as done for IL-6 (EMC004, NeoBioscience) and IL-1β (EMC001b, NeoBioscience) pretreatment. All procedures were performed according to the manufacturer’s instructions.

Immunofluorescence staining

We followed the procedure of Xu et al.’s study [47]. The OCT-embedded tissues were cut into 10-μm-thick sections. Next, we balanced them at room temperature for 20 min, and then washed them with 1×PBS for 5 min. After that, we blocked the sections with 5% BSA with 0.3% TritonX-100 in PBS for 1 h. After the blocking step, we incubated the sections with primary anti-Aβ_1-42 (1 : 200, ab201061, Abcam) and anti-Iba1(1 : 200, #17198, Cell Signaling Technology) at 4°C overnight. After washing them three times with PBS, the brain tissues were incubated with Alexa Fluor^® 488 conjugated Donkey anti-rabbit IgG or 568 Donkey anti-mouse IgG secondary antibodies at 1 : 500 dilutions for 2 h, at room temperature. The nuclei were stained with DAPI for 5 min after capturing the images with the TissueFAXS System (TissueGnostics, Austria).

HE staining

We followed the procedure from Zhang et al’s. study [48]. The paraffin-embedded tissues were cut into 5-μm-thick sections. After that, we deparaffinized and rehydrated the sections, and stained them with hematoxylin and eosin (H&E). We observed the staining with the TissueFAXS System (TissueGnostics, Austria). Histopathological evaluation was performed in the stained sections. The pathological score ranged from 0 to 6, and was subdivided into the following categories: degree of cell arrangement (0 = cells arranged regularly, cell layer was rich, and cell band structure was intact, 2 = cells arranged irregularly, cell band structure missing, incomplete, 1 = between 0 and 2), cell gap size (0 = cell arranged closely, 2 = cell gap grew, 1 = between 0 and 2), neuron morphology (0 = neurons with regular morphology, large mass, intact cell structure and clear cytoplasm, 2 = a large number of neurons with small cell bodies, irregular morphology, deep staining, 1 = between 0 and 2).

Measuring BBB permeability

We cut the 70,000 MW FITC-dextran perfusion tissues several times with a freezing microtome (Leica, German) to create 20-μm-thick sections, and then captured the images with an inverter fluorescence microscope. We used Image J to analyze the data obtained from the images, as has been described by Natarajan et al. [49].

RNA extraction and quantitative real-time RT-PCR

We extracted total RNA from the hippocampi and cortexes of the brains with TRIZOL (Life, USA), and synthesized cDNA with an RT reagent kit (Takara, Japan) according to the manufacturer’s instructions. The final cDNA product was used for the subsequent cDNA amplification by an ABI 3421 Real-Time PCR System (ABI, USA). We evaluated the mRNA expression with software provided by the manufacturer. β-actin served as a housekeeping gene. APP, BACE-1, LRP-1, ZO-1, and Occludin gene expression were tested. The primer sequences are shown in Table 1.

Table 1

Real-time PCR primer sequences

Name	Forward Primer (5’-3’)	Reverse Primer (5’-3’)
β-actin	GATTACTGCTCTGGCTCCTAGC	GACTCATCGTACTCCTGCTTGC
APP	TCCGAGAGGTGTGCTCTGAA	CCACATCCGCCGTAAAAGAATG
BACE-1	CAGTGGGACCACCAACCTTC	GCTGCCTTGATGGACTTGAC
LRP-1	ACTATGGATGCCCCTAAAACTTG	GCAATCTCTTT CACCGTCACA
ZO-1	GCCGCTAAGAGCACAGCAA	TCCCCACTCTGAAAATGAGGA
Occludin	TTGAAAGTCCACCTCCTTACAGA	CCGGATAAAAAGAGTACGCTGG
IL-1β	GCAACTGTTCCTGAACTCAACT	ATCTTTTGGGGTCCGTCAACT
TNF-α	CCCTCACACTCAGATCATCTTCT	GCTACGACGTGGGCTACAG
IL-6	TAGTCCTTCCTACCCCAATTTCC	TTGGTCCTTAGCCACTCCTTC

Statistical analysis

All data are presented as mean±SD and analyzed with SPSS 22.0 and GraphPad Prism 6. The interaction between gene (APP/PS1 and WT) and environment (Cd and control) on their main effect was analyzed with two-way ANOVA and doing post hoc LSD (Least Significant Difference Test) multiple comparison procedure after any significant ANOVA. The p-value of gene-environment interaction effect was indicated the absence or presence of interaction. Post-hoc tests were used to compare the differences among the different dose groups of Cd exposure when the genotype was APP/PS1 or WT. In all cases, p < 0.05 was considered statistically significant.

RESULTS

Cadmium exposure increased Cd levels in the blood and brain tissue of both C57BL/6J and APP/PS1 mice

To assess the internal Cd exposure levels, we tested the Cd concentrations in the blood and brain tissue through ICP-MS. As shown in Fig. 1A, in both C57BL/6J and APP/PS1 mice, we detected higher blood Cd concentrations in the 10 mg/L Cd exposure group than in the control and 1 mg/L Cd exposure groups (p < 0.05). Furthermore, the brain Cd levels for 10 mg/L Cd exposure also increased compared with the control groups of both C57BL/6J and APP/PS1 mice; see Fig. 1B (p < 0.05). In both C57BL/6J and APP/PS1 mice, the internal exposure-level gradually increased along with the external exposure concentration, showing a dose-dependent response. However, long-term exposure to drinking water containing low concentrations of Cd could induce Cd accumulation in the body, especially in the brain tissue. Furthermore, the internal exposure level for the C57BL/6J in the same group was similar to that of the APP/PS1 mice, and the differences had no statistical significance. Additionally, the exposure levels were comparable between these two mice.

Fig. 1

Cadmium exposure increased the Cd levels in the blood and brain tissue of both the C57BL/6J and the APP/PS1 mice. A) Blood Cd concentration. N = 6 per group. There was a significant main effect of Cd exposure on blood Cd concentration (Two-way ANOVA: Cd exposure: F = 53.1, p < 0.001; genotype: F = 1.0, p = 0.320; Cd exposure-genotype interaction: F = 0.4, p = 0.692). B) Brain Cd concentration. N = 3 per group. There was a significant main effect of Cd exposure on brain Cd concentration (Two-way ANOVA: Cd exposure: F = 20.2, p < 0.001; genotype: F = 0.2, p = 0.666; Cd exposure-genotype interaction: F = 1.2, p = 0.347). ^* p < 0.05.

Cadmium exposure changed the mice’s brain weights

The change in the organ coefficient reflects the comprehensive toxicity of the chemicals. It is also an important clue to the target organ of toxic action. Thus, we calculated the organ coefficient of the brain. $\begin{matrix} Brain organ coefficient % = \\ brain weight / mouse weight \times 100 % \end{matrix}$

Two-way ANOVA showed a significant main effect of Cd exposure (F = 6.9, p = 0.002), genotype (F = 86.4, p < 0.001), and exposure-genotype interaction (F = 5.8, p = 0.005) on the brain’s organ coefficient (Fig. 2). Post-hoc analysis indicated that the APP/PS1 mice’s brain organ coefficient was lower than that of the C57BL/6J mice at different exposure levels (p < 0.05). Moreover, Cd decreased the brain organ coefficient in 1 mg/L Cd-exposed APP/PS1 mice (p < 0.05). The effect of Cd exposure was not significant in the brain/body ratios of the C57BL/6J mice (Fig. 2).

Fig. 2

Cadmium exposure changed mouse brain weights in the C57BL/6J and APP/PS1 mice. N = 10∼12 per group. ^* p < 0.05.

Pathological changes in the brain were detected by HE staining

To gain further insight into the pathogenesis in the mouse brains, we used hematoxylin and eosin (HE) staining to obverse the effect of Cd exposure on the pathological change in mouse brains. HE staining showed that nerve cells in the control groups (0 mg/L) were evenly stained and neatly arranged, among both the C57BL/6J and the APP/PS1 mice. In the 1 mg/L Cd-exposed group, the cell band structure in the hippocampus and the cortical area was incomplete, some cell spaces were enlarged (black arrows), and some neurons were irregular and abnormal (red arrows). In the Cd-10 mg/L exposed group, the cells were sparsely arranged, the cellular structures were unclear, and the cell shapes were irregular (red arrows). Neuron cell damage in the CA1 and CA3 regions in the APP/PS1-10 mg/L group was more serious than that in the C57BL/6J-10 mg/L group. Consistently, the pathological score of the Cd-exposed group mice was more severe than that of the control group mice. In both C57BL/6J and APP/PS1 mice, the pathological scores of the Cd-10 mg/L exposed group mice were considerably higher than those of the control group mice (p < 0.05). (Fig. 3A, B).

Fig. 3

Pathological changes in the brain. A) HE staining of the brain tissues (hippocampus CA1, CA3, and DG areas, and the cortex). Black arrows: enlarged intercellular spaces. Red arrows: cells were atrophic and irregularly shaped. Scale bar = 100μm. B) Pathological scores of the brain histology in the HE-stained sections were determined blindly. N = 3 per group. There was a significant main effect of Cd exposure on pathological score (Two-way ANOVA: Cd exposure: F = 12.0, p = 0.001; genotype: F = 0, p = 1.000; Cd exposure-genotype interaction: F = 0, p = 1.000). * p < 0.05.

Cadmium’s effect on locomotor activity and anxiety

We conducted an open field test to assess Cd’s effect on locomotor activity and anxiety (Fig. 4). Two-way ANOVA showed a significant main effect of genotype (F = 6.6; p = 0.013), and Cd exposure-genotype interaction was also statistically significant (F = 5.4; p = 0.007) on total travel distances. Post-hoc analysis indicated that Cd exposure significantly increased the locomotor activity in C57BL/6J mice (p < 0.05), but not in the APP/PS1 mice. As mice that are stressed show less activity in the open field, APP/PS1 mice showed greater stress than C57BL/6J mice at the same exposure doses. The APP/PS1 mice showed significantly more chaotic movements.

Fig. 4

Cadmium’s effect on locomotor activity and anxiety. A) Schematic of zones in the open field arena and the representative traces of mice movement. B) Total distance. C) Center area distance as a percentage of total distance. N = 9∼12 per group. ^* p < 0.05.

Furthermore, mice that prefer staying close to the walls and that travel more in the periphery can be described as showing thigmotaxis (movement towards a solid object), which is pronounced in mice showing signs of anxiety-like behavior [50]. We found a significant main effect of Cd exposure (F = 4.7, p = 0.012) and genotype (F = 30.0, p < 0.001) on the percentages of center area distance in the total distance, and Cd exposure-genotype interaction effect was not statistically significant (F = 2.0, p = 0.144). Post-hoc analysis suggested that Cd exposure caused anxiety in the APP/PS1 mice (p < 0.05) (Fig. 4).

Cadmium exposure impaired spatial working and reference memory in the Y-maze test

We used spontaneous alternation to measure the mice’s spatial working memory. Two-way ANOVA revealed a significant main effect of Cd exposure (F = 32.3, p < 0.001), genotype (F = 10.2, p = 0.002), and exposure-genotype interaction (F = 3.5, p = 0.037) on the spontaneous alternation percentage. Post-hoc analysis indicated that spontaneous alternation declined in both the C57BL/6J and the APP/PS1 mice after any dose of Cd exposure (p < 0.05) (Fig. 5A). In general, regardless of the mouse genotype or the exposure dose, Cd could impair spatial working memory in mice. Surprisingly, we found that the APP/PS1 mice’s spatial working memory was slightly superior to that of the C57BL/6J mice at exposure doses of 10 mg/L (p < 0.05). Additionally, the correlation between the spontaneous alternation effect and the Cd exposure dose in the APP/PS1 mice was not linear.

Fig. 5

Cadmium exposure impaired spatial working and reference memory. A) Spontaneous alternation rate. B) Time spent in the novel arm(s). C) Novel arm distance as a percentage of total distance. N = 9∼12 per group. ^* p < 0.05.

We also used Y-maze to measure the mice’s spatial reference memory. This can be measured by placing the mice into the Y-maze with an arm (the novel arm) closed during the training period. In the testing period, a mouse with a good memory will enter the novel arm more frequently than other arms. We found a significant main effect of Cd exposure (F = 5.8, p = 0.005) and genotype (F = 25.3, p < 0.001) on the time the mice spent in the novel arm, and Cd exposure-genotype interaction effect was not statistically significant (F = 2.0, p = 0.142); the APP/PS1 mice spent less time than the C57BL/6J mice did in the novel arm, regardless of whether they were Cd-exposed (10 mg/L, p < 0.05) or non-exposed (p < 0.05). The Cd-exposed mice spent less time in the novel arm, and the differences between the 1 mg/L exposed C57BL/6J mice and the unexposed C57BL/6J mice were statistically significant (p < 0.05). Two-way ANOVA showed a significant effect of Cd exposure (F = 5.0, p = 0.009) and genotype (F = 15.0, p < 0.001) on the percentages of novel arm distance in total distance, and Cd exposure-genotype interaction effect was not statistically significant (F = 1.6, p = 0.211). Post-hoc analysis indicated that the Cd exposure APP/PS1 mice had shorter moving distances in the novel arm than the non-Cd exposure APP/PS1 mice (p < 0.05). These results indicated that Cd exposure had impaired spatial reference memory, especially in APP/PS1 mice (Fig. 5).

Cd exposure increased the brain’s Aβ load in APP/PS1 mice

As APP/PS1 mice often develop amyloid plaque at 5-6 months old, and abnormal Aβ deposition is the critical pathological hallmark of AD, we detected the Aβ_1-42 levels both in the APP/PS1 and C57BL/6J mice. We examined the amount of Aβ_1-42 in the CSF and cortexes of the C57BL/6J mice, but detected none. ELISA analysis revealed that Cd exposure had increased the Aβ_1-42 levels in the APP/PS1 mice’s cerebrospinal fluid (Fig. 6A). Moreover, the 10 mg/L groups showed an increase in the Aβ_1-42 in the CSF levels, compared with the control group (p < 0.05).

Fig. 6

Cd exposure increased the Aβ load of the APP/PS1 mouse brains. A) Aβ_1-42 protein concentrations in the CSF of the APP/PS1 mice. N = 6 per group. B) Quantitation of the Aβ_1-42 protein expression in the brains of the APP/PS1 mice. C) Aβ_1-42- expression in the brains of the APP/PS1 mice. N = 3 per group. ^* p < 0.05.

We evaluated the expression of Aβ_1-42 in the cortexes of the APP/PS1 mice by immunofluorescence (Fig. 6B, C). The results showed a marked increase in Aβ_1-42 expression levels with the increase in Cd doses (p < 0.05), and a large amount of the Aβ plaque was deposited in the mouse brains in the 10 mg/L group (Fig. 6).

To understand how Aβ-related gene expression changed, we detected relative mRNA levels in the hippocampi and cortexes of each group of mice by qRT-PCR (Fig. 7).

Fig. 7

The mRNA expression of Aβ-related molecules in the hippocampi and cortexes. A, B) APP expression. There was no significant main effect of Cd exposure and genotype and Cd exposure-genotype interaction effect on the APP expression in hippocampus and cortex (Two-way ANOVA: hippocampus, Cd exposure: F = 3.4, p = 0.066; genotype: F = 0.3, p = 0.602; Cd exposure-genotype interaction: F = 1.7, p = 0.220. Cortex, Cd exposure: F = 3.1, p = 0.083; genotype: F = 1.1, p = 0.325; Cd exposure-genotype interaction: F = 0.8, p = 0.454). C, D) BACE-1 expression. There was no significant main effect of Cd exposure and genotype on the BACE-1 expression in hippocampus and cortex (Two-way ANOVA: hippocampus, Cd exposure: F = 1.0, p = 0.403; genotype: F = 2.7, p = 0.124; Cd exposure-genotype interaction: F = 1.0, p = 0.386. Cortex, Cd exposure: F = 1.6, p = 0.249; genotype: F = 0.6, p = 0.437; Cd exposure-genotype interaction: F = 4.6, p = 0.033). E, F) LRP-1 expression. N = 3 per group. ^* p < 0.05.

AβPP edited by the APP gene acts as a precursor to Aβ and can be hydrolyzed into Aβ_1-40 and Aβ_1-42 fragments by beta-secretase 1 (BACE-1). Therefore, an increase in the APP and BACE-1 gene expression levels may lead to an increase in Aβ in the brain. Two-way ANOVA showed that neither Cd exposure nor genotype had any effect on APP or BACE-1 gene expression levels in the hippocampi or the cortex.

Low-density lipoprotein receptors (LRP-1) are important ligand proteins that are present in the BBB and responsible for transporting Aβ from the brain to the blood. The decline in LRP-1 expression can result in deposits of Aβ in the brain. We found a significant effect for Cd exposure (F = 24.6, p < 0.001), and exposure×genotype interaction (F = 20.1, p < 0.001), but no effect for genotype (F = 0.1, p = 0.715) on LRP-1 expression levels in the hippocampi. Additionally, two-way ANOVA revealed a significant main effect of Cd exposure (F = 12.1, p = 0.005) and genotype (F = 4.8, p = 0.029), but no exposure-genotype interaction (F = 9.9, p = 0.001) on LRP-1 expression levels in the cortex. Moreover, the mRNA expression of LRP-1 in the hippocampi and cortexes of the C57BL/6J mice decreased. However, with increasing exposure doses, the APP/PS1 mice’s mRNA expressions first increased, and then decreased.

Cd exposure increased BBB permeability

A growing body of evidence has indicated that BBB dysfunction is an important pathogenic mechanism in several neurodegenerative diseases. Thus, we perfused the animals with FITC-Dextran (70 kDa) through the left ventricle to further evaluate the BBB permeability changes. We used Image J to analyze the data obtained from imaging. After removing any brightly stained capillaries, we measured the fluorescence of the entire image without capillary staining. We found a significant main effect of Cd exposure (F = 33.1, p < 0.001), but no effect of genotype (F = 0.1, p = 0.751) or Cd-exposure-genotype interaction (F = 0.3, p = 0.732) on the leakage’s mean fluorescence intensity. Additionally, BBB leakage differed little between mice with different genotypes. Moreover, the BBB leakage gradually increased as the Cd exposure level increased (p < 0.05) (Fig. 8).

Fig. 8

Cd exposure increased BBB permeability. Images were obtained by fluorescence microscope. Scale bar = 100μm. The images’ quantitation results were analyzed by measuring the average fluorescence intensity. N = 3 per group. ^* p < 0.05.

To understand how BBB-related gene expression changed, we detected relative mRNA levels in the hippocampi and cortexes of each group of mice by qRT-PCR (Fig. 9).

Fig. 9

The mRNA expression of BBB-related molecules in the hippocampi and cortexes of 6-month-old mice. A, B) ZO-1 expression. C, D) occludin expression. N = 3 per group. ^* p < 0.05.

The BBB is a closed barrier formed by tight junctions of cerebral endothelial cells. Tight junctions in cerebral endothelial cells include the Claudin protein family, the Occludin protein family, and the ZO protein family. A decrease in the tight junction proteins means that the BBB has been impaired.

Two-way ANOVA showed no effect on either Cd exposure, genotype, or exposure-genotype interaction on ZO-1 gene expression levels in the hippocampi or the cortexes. (Two-way ANOVA: hippocampus, Cd exposure: F = 3.5, p = 0.064; genotype: F = 4.3, p = 0.060; Cd exposure-genotype interaction: F = 2.7, p = 0.110. Cortex, Cd exposure: F = 1.9, p = 0.194; genotype: F = 3.2, p = 0.098; Cd exposure-genotype interaction: F = 2.9, p = 0.093). However, we found a significant main effect of Cd exposure (F = 47.3, p < 0.001), genotype (F = 32.4, p < 0.001), and exposure-genotype interaction (F = 20.4, p < 0.001) on Occludin expression levels in the hippocampi. Also, the occludin expression level in the hippocampi decreased with the increase in dose (p < 0.05). Moreover, there was a significant main effect of genotype (F = 7.1, p = 0.021) in the cortex.

Cd exposure resulted in neuroinflammation, in both the C57BL/6J and the APP/PS1 mice

Neuroinflammation, including the activation of microglia, is another pathological feature of AD. Ionized calcium binding adaptor molecule 1 (Iba-1) is the histological microglial marker [51], and we evaluated Iba-1 expression by immunofluorescence. Cd exposure resulted in increased microglia expression in the brain tissues of the mice; microglia expression in the brains of the APP/PS1 mice was higher than that in the C57BL/6J mice (Fig. 10).

Fig. 10

Cd exposure resulted in neuroinflammation in both C57BL/6J and APP/PS1 mice. Immunofluorescence staining of microglia (Iba-1) in the mice’s brain tissue. Scale bar = 100μm. N = 3 per group. ^* p < 0.05.

To understand how inflammatory-related gene expression changed, we detected relative mRNA levels in the hippocampi and cortexes of each group of mice by qRT-PCR (Fig. 11).

Fig. 11

The mRNA expression of inflammatory-related molecules in the hippocampi and cortexes of 6-month-old mice. A, B) IL-1β expression. C, D) TNF-α expression. E, F) IL-6 expression. N = 3 per group. ^* p < 0.05.

The onset of AD is also associated with the neuroinflammation and activation of glial cells, which leads to the secretion of a variety of proinflammatory cytokines, such as IL-1β, TNF-α, and IL-6. We found a significant main effect of Cd exposure (F = 40.3, p < 0.001), but no effect for either genotype (F = 0, p = 0.969) or Cd exposure-genotype interaction (F = 2.1, p = 0.160) on IL-1β gene expression levels in the hippocampus. Additionally, we found no effect of Cd-exposure, genotype, or exposure-genotype interaction on IL-1β gene expression levels in the cortex (Two-way ANOVA: Cd exposure: F = 0.2, p = 0.822; genotype: F = 1.7, p = 0.215; Cd exposure-genotype interaction: F = 1.4, p = 0.287). Moreover, we found a significant main effect of Cd exposure (hippocampus F = 19.2, p < 0.001; cortex F = 3.9, p = 0.048), but no effect of genotype (hippocampus F = 0, p = 0.888; cortex F = 2.9, p = 0.113) or Cd exposure×genotype interaction (hippocampus F = 1.2, p = 0.349; cortex F = 0.6, p = 0.577) on TNF-α genes expression levels. Two-way ANOVA showed a significant main effect of Cd exposure (F = 9.5, p = 0.003), but no effect for either genotype (F = 4.1, p = 0.067) or Cd exposure-genotype interaction (F = 1.3, p = 0.313) on IL-6 gene expression levels in the hippocampus. However, we did find a significant main effect for Cd exposure (F = 4.8; p = 0.029) and genotype (F = 48.7; p < 0.001). Cd exposure-genotype interaction was also statistically significant (F = 5.6; p = 0.020) for IL-6 gene expression levels in the cortex. The IL-1β, TNF-α and IL-6 expression levels in the hippocampi of mice first increased, and then decreased. The IL-6 expression levels in the cortexes of the APP/PS1 mice were higher than those of the C57BL/6J mice at the same dose, and the difference was statistically significant in both the 0 mg/L and the 10 mg/L groups (p < 0.05).

Cd exposure resulted in systemic inflammation in both the C57BL/6J and the APP/PS1 mice

To identify potential biomarkers, we measured the proinflammatory cytokine levels (e.g., IL-1β, TNF-α, and IL-6) in peripheral circulation with ELISA kits. Two-way ANOVA revealed a significant main effect of Cd exposure (F = 7.1, p = 0.005), genotype (F = 8.9, p = 0.008), and exposure-genotype interaction (F = 9.9, p = 0.001) on IL-1β levels. Post-hoc analysis indicated that the Cd exposure had increased the serum IL-1β levels in the APP/PS1 mice; the IL-1β concentrations in the serum of the APP/PS1 mice peaked at 1 mg/L in the Cd-exposed groups (p < 0.05). However, we observed no effect of Cd exposure on serum IL-1β levels in the C57BL/6J mice. Yet we did find a significant effect of genotype (F = 11.7, p = 0.003), and exposure-genotype interaction (F = 9.0, p = 0.002), but no effect of Cd exposure (F = 2.5, p = 0.109) on IL-6 levels. Compared with that of C57BL/6J mice, the serum IL-6 levels for the APP/PS1 mice significantly increased after Cd exposure (p < 0.05). However, we found no effect of Cd exposure, genotype, or exposure-genotype interaction on serum TNF-α levels (Two-way ANOVA: Cd exposure: F = 1.1, p = 0.366; genotype: F = 0.3, p = 0.593; Cd exposure-genotype interaction: F = 1.8, p = 0.202) (Fig. 12).

Fig. 12

Cd exposure resulted in systemic inflammation in both C57BL/6J and APP/PS1 mice. A) IL-1β serum levels. N = 4 per group. B) TNF-α serum levels. N = 4 per group. C) IL-6 serum levels. N = 4 per group. ^*p < 0.05.

DISCUSSION

This study explored behavioral tests, Aβ expression, BBB permeability, and inflammatory response in two mouse genotypes after exposure to either 1 mg/L or 10 mg/L Cd in drinking water, from mothers to offspring, at 6 months old. We explored whether the environmental risk factor Cd exhibited different pathogenic potency in mice with different genotypes. To simulate human Cd exposure, our intervention time for the mice was long—from the embryo stage through adulthood—until they were sacrificed.

Cd is a relatively rare metal, naturally occurring in its inorganic form. It is a non-biodegradable element and possesses a biological half-life of 20-30 years [52]. Our study showed that blood Cd levels are positively correlated with dietary Cd intake, that is to say, they are good indicators of the body’s Cd exposure levels [53, 54]. Mice of both genotypes had similar blood Cd levels after exposure to the same dose of Cd. Blood Cd levels in the mice exposed to 1 mg/L were about 0.5μg/L. Hongyu Li’s study suggested that blood Cd levels were comparable to adults aged 60 years or older in the US [12]. Blood Cd levels of mice exposed to 10 mg/L were about 2μg/L, and blood level was close to that in the Chinese population [55]. Thus, the Cd exposure levels of the mice in this study were close to those of the non-occupationally-exposed population.

Meanwhile, our study showed that concentrations of Cd in the mouse brains gradually increased along with increasing concentrations of Cd in drinking water. This suggested that Cd accumulates in the brain, which is one of Cd’s target organs. Studies in humans have reached similar conclusions. Once it enters the body, and owing to its long life, Cd exerts toxic effects in the brain [56], dependent on its ability to cross the BBB [57]. Although the difference was not statistically significant, the brains of the APP/PS1 mice accumulated less Cd than those of the C57BL/6J mice who drank water containing 10 mg/L of Cd. Our previous study showed that lead causes cerebrovascular endothelial function impairment in APP/PS1 mice, and decreased functional vessel numbers [58]. Therefore, Cd may also damage the cerebral vessels of the APP/PS1 mice, and reduce Cd transport through blood vessels to the brain. In future studies, we will explore potential mechanisms underlying this difference.

We observed the differences between the two types of mice after they drank water containing Cd. The toxic effects of Cd on the two genotypes differed little in terms of neuronal histomorphology and BBB permeability. Cd caused a series of pathological morphological changes in the mouse brains, such as irregular neurons, sparsely arranged cells, and enlarged cell spaces. With the increasing concentration of exogenous Cd, the pathological scores for the mouse brain neurons gradually increased. Many existing studies on Cd and neurotoxicity have used short-term exposure to high Cd concentrations [59, 60], which is a far cry from the long-term exposure to low Cd concentrations to which people are often exposed.

Many studies have revealed that Cd-exposure induces neurotoxicity in mice [61]. However, models of long-term exposure to low concentrations of Cd have been few and far between. Our study showed that life-long exposure to Cd can cause neuron damage in mice, even at extremely low concentrations of Cd. We also examined BBB permeability in mice by FITC-dextran perfusion. The results showed more fluorescent dye leakage at higher Cd doses, indicating that Cd exposure had damaged the BBB in mice with both genotypes. This has also been tested in rats with high doses of acute Cd exposure; by the end of the experiment, the BBB permeability had increased [13]. To investigate the potential causes of BBB leakage, we measured the mRNA expression levels of tight junctions in cerebral endothelial cells in the hippocampi and cortexes of mice. We found a decreased expression of Occludin in the hippocampus along with increased Cd exposure levels. Occludin is a transmembrane protein which is the primary building block of tight junctions. Occludin stabilizes the tight junction assembly which assists in its function as a barrier [62]. Several studies have reported that the alternation or decrease in occludin results in increased BBB permeability [63, 64]. Therefore, the decrease in occludin may be one reason for the diminished BBB integrity caused by Cd.

Furthermore, when exposed to environmental Cd, we found that the APP/PS1 mice had more severe damage than the C57BL/6J mice, based on the following five criteria. All the damage mentioned reflects important features and symptoms in the pathogenesis and progression of AD.

Firstly, anxiety-like behavior and chaos movement increased in the APP/PS1 mice with increasing Cd exposure, while these behaviors changed little in the C57BL/6J mice. AD is the most common neurodegenerative disease accompanied by the progressive impairment of memory and cognitive function [20]. Moreover, a range of behavioral changes and impairments in cognitive function have been observed in different AD mouse models [65]. The present study incorporated an open-field test and a Y-maze test. The open field test can detect rodents’ locomotor activity and anxiety-like behavior. The total distance traveled may reflect the animals’ locomotor activity, while the entries and time spent in the center area and the distance traveled may reflect the animal’s anxiety. Stressed mice showed less activity in the open field; the less anxious animals will spend more time exploring the center area [50]. APP/PS1 mice showed greater stress and more chaotic movements than C57BL/6J mice at the same exposure dose. Also, Cd exposure caused anxiety in the APP/PS1 mice. This is similar to previous findings on anxiety-like behavior following Cd exposure in rats and mice [66 –68].

Secondly, Cd showed significant damage to spatial reference memory in APP/PS1 mice, but not in C57BL/6J mice. Although the effects of Cd on learning and memory in mice are still not fully understood, especially in regard to hippocampus-dependent learning and memory, we explored the effect of Cd on hippocampus-dependent memory with a Y-maze test. The time and distance traveled in the novel arm decreased among the APP/PS1 mice after Cd exposure. This indicates that Cd exposure impaired spatial reference memory among the APP/PS1 mice. Several studies have led to similar findings. Wang [69] reported that Cd exposure impaired short-term spatial memory and contextual fear memory in mice. Wang et al. observed that Cd-treated mice had fewer adult-born cells, adult-born neurons, and a reduced proportion of adult-born cells that differentiated into mature neurons in the sub-granular zone of the dentate gyrus. These results suggest that Cd exposure from adolescence to adulthood is sufficient to cause cognitive deficits in mice.

Thirdly, Aβ plaque in APP/PS1 mouse brains gradually increased as the Cd exposure levels increased. AD in APP/PS1 mice develops rapidly, between the ages of 6-9 months [42]. It is characterized by the deposition of Aβ plaque in mouse brains, and it is an important pathological feature of AD. Under normal conditions, intracerebral Aβ maintains a balance between production and clearance. When Aβ clearance is insufficient, Aβ deposits in the brain. The clearance mechanisms include the ubiquitin-proteasome system, autophagy lysosomes, proteases, microglial phagocytosis, and transport across the BBB and blood-CSF barrier, from the brain to the blood [70]. BBB is a factor in Aβ clearance [71]. The results mentioned above suggest that Cd exposure had compromised BBB integrity in the mice. As such, the BBB disruption may have affected Aβ clearance and transport, which in turn caused Aβ deposition. Furthermore, some Aβ can also be transported across the blood-CSF barrier. CSF is a fluid secreted by the choroid plexus. It can deliver nutrients to the CSF and remove brain metabolites and unwanted materials from the CSF to maintain central nervous system homeostasis.

It has been observed in mice that lead exposure increases Aβ in CSF, and decreases the clearance capacity of the blood-CSF barrier. This suggests that heavy metals may damage Aβ clearance capacity [72]. In our study, we determined the concentration of Aβ_1-42 in CSF by ELISA. The results showed that Cd exposure increased the concentrations of Aβ_1-42 in CSF in the APP/PS1 mice. There was also a level-effect relationship with the Cd exposure level. In addition, we detected the Aβ_1-42 expression in the brain by immunofluorescence. We observed a considerable amount of Aβ plaque in the brains of the 10 mg/L Cd-exposed group. This indicated that Cd exposure had increased Aβ plaque deposition. The AβPP protein edited by the APP gene acts as a precursor to Aβ, and is a factor in AD pathogenesis. In this experiment, we detected the mRNA expression levels of the APP, BACE-1, and LRP-1 genes in the hippocampus and cortex by qRT-PCR. Despite the absence of statistical differences, the results revealed that Cd exposure had dose-dependently increased the mRNA expression of cortex APP in the APP/PS1 mice. This suggests that Cd exposure increases Aβ production, especially in the cerebral cortex.

Fourthly, Cd exposure resulted in increased microglia expression in the brain tissues of mice. Additionally, microglia expression in the brains of the APP/PS1 mice was higher than that of the C57BL/6J mice. Microglia activation is involved in the brain’s inflammatory responses. In this study, immunofluorescence staining results revealed that Cd exposure had activated microglia in the brain, and increased the number of microglia. The proliferation and activation of microglia in the brain, concentrated around amyloid plaque, is a prominent feature of AD [73]. Stimulation of microglial activity may prevent AD before it is established at an early stage, but become detrimental later, when the disease has reached a highly inflamed, neurodegenerative stage. It has been shown that neuroinflammation and microglial activation at rostral ventrolateral medulla, and their downstream cellular mechanisms, causally underpin Cd -induced cardiovascular dysregulation [74].

Finally, IL-6 was expressed at higher levels in the APP/PS1 mice than in the C57BL/6J mice, both in the cortex and in the serum. This indicated that the inflammatory response had been more severe in the AD model mice. The occurrence of AD is also associated with neuroinflammation and the activation of glial cells, and then secretions of a variety of proinflammatory cytokines [75]. Previous studies have shown that in the serum of AD patients, some proinflammatory cytokines, such as IL-1β, TNF-α, and IL-6, are elevated compared to control groups [76, 77]. In this experiment, we examined the levels of inflammatory cytokine IL-1β, IL-6, and TNF-α in mouse serum by ELISA. The serum levels of IL-6 in the APP/PS1 mice were higher than those in the C57BL/6J mice at the same Cd exposure level. Along with the increased expression in the brain, IL-6 expression also increased in the serum, implying that IL-6 may be a biomarker of inflammatory effects in the brains of AD patients.

Meanwhile, our study also produced some unexpected results. Like the results of many toxicity studies [78, 79], the dose-toxicity relationship was not linear. Some toxicity was more severe in mice exposed to lower doses. For instance, the mice had decreased brain/body weight ratios and increased LRP-1, IL-1β, and TNF-α in some of the 1 mg/L exposed groups. As the evidence in our study is still inadequate and other relevant studies are limited, it is difficult to elucidate the mechanisms of these results. As such, we will continue to explore the mechanisms behind the relationship between dose and toxicity.

In conclusion, we have established a model of low-dose Cd exposure in drinking water from mother to offspring throughout the life course of mice. This study has also addressed the potential mechanisms of Cd exposure-induced AD progression (Fig. 13). The potential mechanisms include BBB disruption, increased Aβ production, and the decline of Aβ clearance, as well as increased inflammatory responses. Meanwhile, gene-environment interaction analysis of some indicators suggested that low-dose Cd exposure accelerates disease progression in APP/PS1 mice. However, further experimentation is needed to assess the role of gene-environment interaction in AD pathogenesis.

Fig. 13

A potential mechanism of Cd exposure-induced AD progression.

Footnotes

ACKNOWLEDGMENTS

The authors have no acknowledgments to report.

FUNDING

This project was supported by grants from the National Natural Science Foundation of China (No. 81472998, 81872661). The funders had no role in the study design, data collection and analysis, manuscript preparation, or the decision to publish.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author. The data is not publicly available due to privacy and ethical restrictions.

The supplementary material is available in the electronic version of this article: .

References

Arvanitakis

, Shah

, Bennett

(2019) Diagnosis and management of dementia: Review. JAMA 322, 1589–1599.

Brouwers

, Sleegers

, Van Broeckhoven

(2008) Molecular genetics of Alzheimer’s disease: An update. Ann Med 40, 562–583.

Goate

, Chartier-Harlin

, Mullan

, Brown

, Crawford

, Fidani

, Giuffra

, Haynes

, Irving

, James

(1991) Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer’s disease. Nature 349, 704–706.

Jansen

, Savage

, Watanabe

, Bryois

, Williams

, Steinberg

, Sealock

, Karlsson

, Hägg

, Athanasiu

, Voyle

, Proitsi

, Witoelar

, Stringer

, Aarsland

, Almdahl

, Andersen

, Bergh

, Bettella

, Bjornsson

, Brækhus

, Bråthen

, de Leeuw

, Desikan

, Djurovic

, Dumitrescu

, Fladby

, Hohman

, Jonsson

, Kiddle

, Rongve

, Saltvedt

, Sando

, Selbæk

, Shoai

, Skene

, Snaedal

, Stordal

, Ulstein

, Wang

, White

, Hardy

, Hjerling-Leffler

, Sullivan

, van der Flier

, Dobson

, Davis

, Stefansson

, Pedersen

, Ripke

, Andreassen

, Posthuma

(2019) Genome-wide meta-analysis identifies new loci and functional pathways influencing Alzheimer’s disease risk. Nat Genet 51, 404–413.

Lambert

, Ibrahim-Verbaas

, Harold

, Naj

, Sims

, Bellenguez

, DeStafano

, Bis

, Beecham

, Grenier-Boley

, et al. (2013) Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet 45, 1452–1458.

Manolio

, Collins

, Cox

, Goldstein

, Hindorff

, Hunter

, McCarthy

, Ramos

, Cardon

, Chakravarti

, Cho

, Guttmacher

, Kong

, Kruglyak

, Mardis

, Rotimi

, Slatkin

, Valle

, Whittemore

, Boehnke

, Clark

, Eichler

, Gibson

, Haines

, Mackay

TFC

, McCarroll

, Visscher

(2009) Finding the missing heritability of complex diseases. Nature 461, 747–753.

, Tsay

, Churchill

(2011) A tutorial and mini-review of the ICP-MS technique for determinations of transition metal ion and main group element concentration in the neurodegenerative and brain sciences. Mon Chem 142, 385–398.

Mir

, Sawhney

, Pottoo

, Mohi-Ud-Din

, Madishetti

, Jachak

, Ahmed

, Masoodi

(2020) Role of environmental pollutants in Alzheimer’s disease: A review. Environ Sci Pollut Res Int 27, 44724–44742.

Wang

, Chen

, Song

, Hong

, Huang

, Li

(2021) A review on cadmium exposure in the population and intervention strategies against cadmium toxicity. Bull Environ Contam Toxicol 106, 65–74.

10.

Jurczak

, Brodowska

, Szkup

, Prokopowicz

, Karakiewicz

, Łstrokój

, Kotwas

, Brodowska

, Grochans

(2018) Influence of Pb and Cd levels in whole blood of postmenopausal women on the incidence of anxiety and depressive symptoms. Ann Agric Environ Med 25, 219–223.

11.

, Wang

, Fu

, Yan

, Wu

, Yin

(2018) Associations between blood cadmium levels and cognitive function in a cross-sectional study of US adults aged 60 years or older. BMJ Open 8, e020533.

12.

Peng

, Bakulski

, Nan

, Park

(2017) Cadmium and Alzheimer’s disease mortality in U.S. adults: Updated evidence with a urinary biomarker and extended follow-up time. Environ Res 157, 44–51.

13.

Ibiwoye

, Matthews

, Travers

, Foster

(2019) Association of acute, high-dose cadmium exposure with alterations in vascular endothelial barrier antigen expression and astrocyte morphology in the developing rat central nervous system. J Comp Pathol 172, 37–47.

14.

Karri

, Schuhmacher

, Kumar

(2016) Heavy metals (Pb, Cd, As and MeHg) as risk factors for cognitive dysfunction: A general review of metal mixture mechanism in brain. Environ Toxicol Pharmacol 48, 203–213.

15.

Fatemi

, Folsom

(2009) The neurodevelopmental hypothesis of schizophrenia, revisited. Schizophr Bull 35, 528–548.

16.

Burke

, Trinh

, Nadar

, Sanyal

(2017) AxGxE: Using flies to interrogate the complex etiology of neurodegenerative disease. Curr Top Dev Biol 121, 225–251.

17.

Engstrom

, Snyder

, Maeda

, Xia

(2017) Gene-environment interaction between lead and Apolipoprotein E4 causes cognitive behavior deficits in mice. Mol Neurodegener 12, 14.

18.

McMullen

, Mostyn

(2009) Animal models for the study of the developmental origins of health and disease. Proc Nutr Soc 68, 306–320.

19.

Barouki

, Gluckman

, Grandjean

, Hanson

, Heindel

(2012) Developmental origins of non-communicable disease: Implications for research and public health. Environ Health 11, 42.

20.

Scheltens

, Blennow

, Breteler

MMB

, de Strooper

, Frisoni

, Salloway

, Van der Flier

(2016) Alzheimer’s disease. Lancet 388, 505–517.

21.

Holtzman

, Morris

, Goate

(2011) Alzheimer’s disease: The challenge of the second century. Sci Transl Med 3, 77sr1.

22.

Reitz

, Mayeux

(2014) Alzheimer disease: Epidemiology, diagnostic criteria, risk factors and biomarkers. Biochem Pharmacol 88, 640–651.

23.

Clark

, Vissel

(2015) Amyloid β: One of three danger-associated molecules that are secondary inducers of the proinflammatory cytokines that mediate Alzheimer’s disease. Br J Pharmacol 172, 3714–3727.

24.

Lauderback

, Hackett

, Huang

, Keller

, Szweda

, Markesbery

, Butterfield

(2001) The glial glutamate transporter, GLT-1, is oxidatively modified by 4-hydroxy-2-nonenal in the Alzheimer’s disease brain: The role of Abeta1-42. J Neurochem 78, 413–416.

25.

Westmark

(2013) What’s hAPPening at synapses? The role of amyloid β-protein precursor and β-amyloid in neurological disorders. Mol Psychiatry 18, 425–434.

26.

De Strooper

, Vassar

, Golde

(2010) The secretases: Enzymes with therapeutic potential in Alzheimer disease. Nat Rev Neurol 6, 99–107.

27.

Schaeffer

, Figueiro

, Gattaz

(2011) Insights into Alzheimer disease pathogenesis from studies in transgenic animal models. Clinics (Sao Paulo) 66 Suppl 1, 45–54.

28.

Donahue

, Flaherty

, Johanson

, Duncan

, Silverberg

, Miller

, Tavares

, Yang

, Wu

, Sabo

, Hovanesian

, Stopa

(2006) RAGE, LRP-1, and amyloid-beta protein in Alzheimer’s disease. Acta Neuropathol 112, 405–415.

29.

Cai

, Qiao

P-F

, Wan

C-Q

, Cai

, Zhou

N-K

, Li

(2018) Role of blood-brain barrier in Alzheimer’s disease. J Alzheimers Dis 63, 1223–1234.

30.

Kisler

, Nelson

, Montagne

, Zlokovic

(2017) Cerebral blood flow regulation and neurovascular dysfunction in Alzheimer disease. Nat Rev Neurosci 18, 419–434.

31.

Bien-Ly

, Boswell

, Jeet

, Beach

, Hoyte

, Luk

, Shihadeh

, Ulufatu

, Foreman

, Lu

, DeVoss

, van der Brug

, Watts

(2015) Lack of widespread BBB disruption in Alzheimer’s disease models: Focus on therapeutic antibodies. Neuron 88, 289–297.

32.

Sweeney

, Sagare

, Zlokovic

(2018) Blood-brain barrier breakdown in Alzheimer disease and other neurodegenerative disorders. Nat Rev Neurol 14, 133–150.

33.

Paul

, Strickland

, Melchor

(2007) Fibrin deposition accelerates neurovascular damage and neuroinflammation in mouse models of Alzheimer’s disease. J Exp Med 204, 1999–2008.

34.

Montagne

, Barnes

, Sweeney

, Halliday

, Sagare

, Zhao

, Toga

, Jacobs

, Liu

, Amezcua

, Harrington

, Chui

, Law

, Zlokovic

(2015) Blood-brain barrier breakdown in the aging human hippocampus. Neuron 85, 296–302.

35.

Montagne

, Zhao

, Zlokovic

(2017) Alzheimer’s disease: A matter of blood-brain barrier dysfunction? J Exp Med 214, 3151–3169.

36.

Zhao

, Nelson

, Betsholtz

, Zlokovic

(2015) Establishment and dysfunction of the blood-brain barrier. Cell 163, 1064–1078.

37.

Papadopoulos

, Saadoun

, Woodrow

, Davies

, Costa-Martins

, Moss

, Krishna

, Bell

(2001) Occludin expression in microvessels of neoplastic and non-neoplastic human brain. Neuropathol Appl Neurobiol 27, 384–395.

38.

Almutairi

MMA

, Gong

, Xu

, Chang

, Shi

(2016) Factors controlling permeability of the blood-brain barrier. Cell Mol Life Sci 73, 57–77.

39.

Brosseron

, Krauthausen

, Kummer

, Heneka

(2014) Body fluid cytokine levels in mild cognitive impairment and Alzheimer’s disease: A comparative overview. Mol Neurobiol 50, 534–544.

40.

Paouri

, Georgopoulos

(2019) Systemic and CNS inflammation crosstalk: Implications for Alzheimer’s disease. Curr Alzheimer Res 16, 559–574.

41.

Kaur

, Sharma

, Deshmukh

(2019) Activation of microglia and astrocytes: A roadway to neuroinflammation and Alzheimer’s disease. Inflammopharmacology 27, 663–677.

42.

Webster

, Bachstetter

, Nelson

, Schmitt

, Van Eldik

(2014) Using mice to model Alzheimer’s dementia: An overview of the clinical disease and the preclinical behavioral changes in 10 mouse models. Front Genet 5, 88.

43.

Simon

, Greenaway

, White

, Fuchs

, Gailus-Durner

, Wells

, Sorg

, Wong

, Bedu

, Cartwright

, Dacquin

, Djebali

, Estabel

, Graw

, Ingham

, Jackson

, Lengeling

, Mandillo

, Marvel

, Meziane

, Preitner

, Puk

, Roux

, Adams

, Atkins

, Ayadi

, Becker

, Blake

, Brooker

, Cater

, Champy

M-F

, Combe

, Danecek

, di Fenza

, Gates

, Gerdin

A-K

, Golini

, Hancock

, Hans

, Hölter

, Hough

, Jurdic

, Keane

, Morgan

, Müller

, Neff

, Nicholson

, Pasche

, Roberson

L-A

, Rozman

, Sanderson

, Santos

, Selloum

, Shannon

, Southwell

, Tocchini-Valentini

, Vancollie

, Westerberg

, Wurst

, Zi

, Yalcin

, Ramirez-Solis

, Steel

, Mallon

A-M

, de Angelis

, Herault

, Brown

SDM

(2013) A comparative phenotypic and genomic analysis of C57BL/6J and C57BL/6N mouse strains. Genome Biol 14, R82.

44.

Nakajima

, Takao

, Hattori

, Shoji

, Komiyama

, Grant

SGN

, Miyakawa

(2019) Comprehensive behavioral analysis of heterozygous Syngap1 knockout mice. Neuropsychopharmacol Rep 39, 223–237.

45.

Kraeuter

A-K

, Guest

, Sarnyai

(2019) The Y-maze for assessment of spatial working and reference memory in mice. Methods Mol Biol 1916, 105–111.

46.

Lim

NK-H

, Moestrup

, Zhang

, Wang

W-A

, Møller

, Huang

F-D

(2018) An improved method for collection of cerebrospinal fluid from anesthetized mice. J Vis Exp, 56774.

47.

, He

, Wang

, Gong

, Bai

, Zhang

, Wang

(2019) Quantifying blood-brain-barrier leakage using a combination of evans blue and high molecular weight FITC-Dextran. J Neurosci Methods 325, 108349.

48.

Zhang

, Lai

, Li

, Liu

, Yang

(2017) Learning and memory improvement and neuroprotection of Gardenia jasminoides (Fructus gardenia) extract on ischemic brain injury rats. J Ethnopharmacol 196, 225–235.

49.

Natarajan

, Northrop

, Yamamoto

(2017) Fluorescein isothiocyanate (FITC)-dextran extravasation as a measure of blood-brain barrier permeability. Curr Protoc Neurosci 79, 9.58.1–9.58.15.

50.

Kraeuter

A-K

, Guest

, Sarnyai

(2019) The open field test for measuring locomotor activity and anxiety-like behavior. Methods Mol Biol 1916, 99–103.

51.

Calcia

, Bonsall

, Bloomfield

, Selvaraj

, Barichello

, Howes

(2016) Stress and neuroinflammation: A systematic review of the effects of stress on microglia and the implications for mental illness. Psychopharmacology (Berl) 233, 1637–1650.

52.

Zhou

, Yin

, Gao

, Liu

, Xie

, Ouyang

, Fan

, Yu

, Zha

, Wang

, Shao

, Feng

, Fan

(2019) Toxicity assessment due to prenatal and lactational exposure to lead, cadmium and mercury mixtures. Environ Int 133, 105192.

53.

Bakulski

, Seo

, Hickman

, Brandt

, Vadari

, Hu

, Park

(2020) Heavy metals exposure and Alzheimer’s disease and related dementias. J Alzheimers Dis 76, 1215–1242.

54.

Ikeda

, Nakatsuka

, Watanabe

, Shimbo

(2018) Estimation of dietary intake of cadmium from cadmium in blood or urine in East Asia. J Trace Elem Med Biol 50, 24–27.

55.

Chen

, Zhang

H-J

, Zhai

H-L

, Chen

, Han

, Li

, Xia

F-Z

, Wang

N-J

, Lu

Y-L

(2020) Association between blood cadmium and vitamin D levels in the Yangtze Plain of China in the context of rapid urbanization. Chin Med J (Engl) 134, 53–59.

56.

Gong

D -k

, Liu

B -h

, Tan

X -h

(2015) Genistein prevents cadmium-induced neurotoxic effects through its antioxidant mechanisms. Drug Res (Stuttg) 65, 65–69.

57.

Khan

MHA

, Parvez

(2015) Hesperidin ameliorates heavy metal induced toxicity mediated by oxidative stress in brain of Wistar rats. J Trace Elem Med Biol 31, 53–60.

58.

, Liu

, Zhao

, Wang

, Chen

, Xia

, Xiao

, Cheng

, Zhao

, He

(2020) Environmental lead exposure aggravates the progression of Alzheimer’s disease in mice by targeting on blood brain barrier. Toxicol Lett 319, 138–147.

59.

Ren

T-T

, Yang

J-Y

, Wang

, Fan

S-R

, Lan

, Qin

X-Y

(2021) Gisenoside Rg1 attenuates cadmium-induced neurotoxicity in vitro and in vivo by attenuating oxidative stress and inflammation. Inflamm Res 70, 1151–1164.

60.

Fan

S-R

, Ren

T-T

, Yun

M-Y

, Lan

, Qin

X-Y

(2021) Edaravone attenuates cadmium-induced toxicity by inhibiting oxidative stress and inflammation in ICR mice. Neurotoxicology 86, 1–9.

61.

Wang

, Zhang

, Abel

, Storm

, Xia

(2018) Cadmium exposure impairs cognition and olfactory memory in male C57BL/6 mice. Toxicol Sci 161, 87–102.

62.

Feldman

, Mullin

, Ryan

(2005) Occludin: Structure, function and regulation. Adv Drug Deliv Rev 57, 883–917.

63.

Kim

K-A

, Kim

J-H

, Shin

Y-J

, Kim

E-S

, Akram

, Kim

E-H

, Majid

, Baek

S-H

, Bae

O-N

(2020) Autophagy-mediated occludin degradation contributes to blood-brain barrier disruption during ischemia in bEnd.3 brain endothelial cells and rat ischemic stroke models. Fluids Barriers CNS 17, 21.

64.

Lakhan

, Kirchgessner

, Tepper

, Leonard

(2013) Matrix metalloproteinases and blood-brain barrier disruption in acute ischemic stroke. Front Neurol 4, 32.

65.

Eriksen

, Janus

(2007) Plaques, tangles, and memory loss in mouse models of neurodegeneration. Behav Genet 37, 79–100.

66.

Lamtai

, Chaibat

, Ouakki

, Berkiks

, Rifi

E-H

, El Hessni

, Mesfioui

, Hbibi

, Ahyayauch

, Essamri

, Ouichou

(2018) Effect of chronic administration of cadmium on anxiety-like, depression-like and memory deficits in male and female rats: Possible involvement of oxidative stress mechanism. J Behav Brain Sci 8, 240–268.

67.

Wang

, Cai

, Hu

, Li

, Kong

, Chen

, Wang

(2019) Investigation of the neuroprotective effects of crocin via antioxidant activities in HT22 cells and in mice with Alzheimer’s disease. Int J Mol Med 43, 956–966.

68.

Wei

J-P

, Wen

, Dai

, Qin

L-X

, Wen

Y-Q

, Duan

, Xu

S-J

(2021) Drinking water temperature affects cognitive function and progression of Alzheimer’s disease in a mouse model. Acta Pharmacol Sin 42, 45–54.

69.

Wang

, Abel

, Storm

, Xia

(2022) Adolescent cadmium exposure impairs cognition and hippocampal neurogenesis in C57BL/6 mice. Environ Toxicol 37, 335–348.

70.

Xin

S-H

, Tan

, Cao

, Yu

J-T

, Tan

(2018) Clearance of amyloid beta and tau in Alzheimer’s disease: From mechanisms to therapy. Neurotox Res 34, 733–748.

71.

Cockerill

, Oliver

J-A

, Xu

, Fu

, Zhu

(2018) Blood-brain barrier integrity and clearance of amyloid-β from the BBB. Adv Exp Med Biol 1097, 261–278.

72.

Shen

, Xia

, Liu

, Jiang

, Shannahan

, Du

, Zheng

(2020) Altered clearance of beta-amyloid from the cerebrospinal fluid following subchronic lead exposure in rats: Roles of RAGE and LRP1 in the choroid plexus. J Trace Elem Med Biol 61, 126520.

73.

Hansen

, Hanson

, Sheng

(2018) Microglia in Alzheimer’s disease. J Cell Biol 217, 459–472.

74.

Tsai

C-Y

, Fang

, Wu

JCC

, Wu

C-J

, Dai

K-Y

, Chen

S-M

(2021) Neuroinflammation and microglial activation at rostral ventrolateral medulla underpin cadmium-induced cardiovascular dysregulation in rats. J Inflamm Res 14, 3863–3877.

75.

Samaila Chiroma

, Baharuldin

, Mat Taib

, Amom

, Jagadeesan

, Mohd Moklas

(2018) Inflammation in Alzheimer’s disease: A friend or foe? Biomed Res Ther 5, 2552–2564.

76.

Lai

KSP

, Liu

, Rau

, Lanctôt

, Köhler

, Pakosh

, Carvalho

, Herrmann

(2017) Peripheral inflammatory markers in Alzheimer’s disease: A systematic review and meta-analysis of 175 studies. J Neurol Neurosurg Psychiatry 88, 876–882.

77.

Khemka

, Ganguly

, Bagchi

, Ghosh

, Bir

, Biswas

, Chattopadhyay

, Chakrabarti

(2014) Raised serum proinflammatory cytokines in Alzheimer’s disease with depression. Aging Dis 5, 170–176.

78.

Szturz

, Wouters

, Kiyota

, Tahara

, Prabhash

, Noronha

, Castro

, Licitra

, Adelstein

, Vermorken

(2017) Weekly low-dose versus three-weekly high-dose cisplatin for concurrent chemoradiation in locoregionally advanced non-nasopharyngeal head and neck cancer: A systematic review and meta-analysis of aggregate data. Oncologist 22, 1056–1066.

79.

Galvao

, Davis

, Tilley

, Normando

, Duchen

, Cordeiro

(2014) Unexpected low-dose toxicity of the universal solvent DMSO. FASEB J 28, 1317–1330.

APP/PS1 Gene-Environmental Cadmium Interaction Aggravates the Progression of Alzheimer’s Disease in Mice via the Blood-Brain Barrier,Amyloid-β,and Inflammation

Abstract

Background:

Objective:

Methods:

Results:

Conclusion:

Keywords

INTRODUCTION

MATERIALS AND METHODS

Animals and treatments

Open field test

Y-maze test

Cerebrospinal fluid, blood, and brain sample collection

Determining Aβ1-42 levels in CSF

Measuring Cd levels by ICP-MS

Determining inflammatory cytokines (IL-1β, TNF-α, and IL-6) in the serum

Immunofluorescence staining

HE staining

Measuring BBB permeability

RNA extraction and quantitative real-time RT-PCR

Statistical analysis

RESULTS

Cadmium exposure increased Cd levels in the blood and brain tissue of both C57BL/6J and APP/PS1 mice

Cadmium exposure changed the mice’s brain weights

Pathological changes in the brain were detected by HE staining

Cadmium’s effect on locomotor activity and anxiety

Cadmium exposure impaired spatial working and reference memory in the Y-maze test

Cd exposure increased the brain’s Aβ load in APP/PS1 mice

Cd exposure increased BBB permeability

Cd exposure resulted in neuroinflammation, in both the C57BL/6J and the APP/PS1 mice

Cd exposure resulted in systemic inflammation in both the C57BL/6J and the APP/PS1 mice

DISCUSSION

Footnotes

ACKNOWLEDGMENTS

FUNDING

CONFLICT OF INTEREST

DATA AVAILABILITY

References

Determining Aβ_1-42 levels in CSF