Abstract
Alzheimer’s disease (AD) is a neurodegenerative brain disease and is the most common form of dementia. In recent years, many studies indicated the association of gut microbiota changes with metabolic diseases. However, the gut microbiota of AD has not been investigated. The present study aims to compare the gut microbiota in APP/PS1 transgenic mice of AD and C57/Bl6 wild-type (WT) mice by pyrosequencing the V3 and V4 regions of the bacterial 16S ribosomal RNA genes. The 3-, 6-, and 8-month-old APP/PS1 and WT mice were used to explore the effects of age on the gut microbiota. First, the results indicated that impaired spatial learning and memory appeared in 6-month-old APP/PS1 mice and was further aggravated in the 8-month-old group, which was consistent with immunohistochemical studies of amyloid plaque. Second, AD histological and behavioral manifestations in the APP/PS1 mice were found to be correlated with a specific gut microbiome state. Third, the microbiota diversity of APP/PS1 mice decreased with increased age. Fourth, further inspection showed that the abundance of Helicobacteraceae and Desulfovibrionaceae at the family level and Odoribacter and Helicobacter at the genus level increased significantly in APP/PS1 mice than in WT mice, while Prevotella abundance in WT mice was significantly higher than in APP/PS1 mice. More human studies are warranted to explore the potential of gut microbiota as diagnostic biomarkers or therapeutic target for AD.
INTRODUCTION
As the most common form of dementia among the elderly, Alzheimer’s disease (AD) is expected to have rising prevalence with the aging demographics of human society worldwide [1–3]. The clinical features of AD, including progressive loss of memory, cognitive function, and behavior impairment, make this chronic disease a great threat to human health and quality of life [4–8]. Despite tremendous efforts devoted to AD drug discovery in the past decades, there is presently no agent exhibiting profound disease-modifying effects.
In recent years, an exponentially increasing number of studies have shown that alterations in the gut microbiota are associated with numerous diseases, including not only gastrointestinal diseases, such as inflammatory bowel disease and irritable bowel syndrome, but also many metabolic diseases, such as obesity, diabetes, and chronic liver diseases [9–14]. Thus, gut microbiota has been proposed to be an attractive target in preventing and treating these diseases [15, 16]. However, there is no report on whole gut microbiota for AD until now. In this study, we conducted comparative analyses on the gut microbiota of APPswe/PS1dE9 (APP/PS1) mice, a well-established mouse model of familial AD, and that of wild-type (WT) C57/Bl6 mice. The findings will provide important clues for future human studies on gut microbiota regulation in preventing and combating AD.
MATERIALS AND METHODS
Materials
Anti-Aβ1-16 6E10 was purchased from Covance (USA). Biotin-Streptavidin HRP Detection system and DAB kit were obtained from ZSGB-BIO (Beijing, China). Hematoxylin was obtained from Lanji Co., LTD (Shanghai, China). All other chemicals and reagents were of analytical reagent.
Animal treatment
18 male APP/PS1 double transgenic mice aged 3 months (group A, n = 6), 6 months (group B, n = 6), and 8 months (group C, n = 6) and 18 male WT mice aged 3 months (group D, n = 6), 6 months (group E, n = 6), and 8 months (group F, n = 6), were used in the study. All mice were housed one per cage under specific pathogen-free conditions and were maintained at standard conditions, 12 h light/dark cycle conditions with room temperature at 22±2°C and 50±5% humidity, and with ad libitum access to food and water. The animal studies were approved by the Animal Use Subcommittee at the Shandong University of Technology.
Morris water maze test
The spatial learning and memory abilities of APP/PS1 mice and WT mice were assessed by Morris water maze test with a water maze 140 cm in diameter. Water mixed with milk was added until it was 1.5 cm over the platform. The maze was surrounded by a blue curtain and the lights stayed in fixed positions. The parameters including escape latency and percent time in the target quadrant were examined.
Tissue preparation and immunohistochemistry
On the day following Morris water maze testing, all mice were sacrificed. The brain was immediately removed and fixed for 22–24 h in 4% paraformaldehyde/phosphate-buffered saline (PBS) at 4°C, dehydrated over 24 h in 20% and 30% sucrose/PBS for immunohistochemisty. The right hemisphere, covered with cryo-embedding media, was cut into 10 μm slices with a cryotome cryostat (Leica CM1850). 4 slices of the same positions from each mouse were placed on the adhesion microscope slides (CITOGLAS). Immunohistochemistry was conducted with an SPlink detection kit (Biotin-Streptavidin HRP Detection system, ZSGB-BIL, Beijing, China). Please refer to the Supplementary Material for further experimental details.
Collection of feces and DNA extraction
Fresh mice feces were collected into individual sterile EP tubes. All fecal samples were frozen immediately at –80°C until DNA extraction. DNA from each fecal sample was collected by phenol trichloromethane DNA extraction methods. The extracted DNA concentration was determined by NanoDrop (Thermo Scientific) and its molecular size was assessed by agarose gel electrophoresis.
16S rRNA gene sequencing and analysis
After DNA extraction from the feces samples, we used polymerase chain reaction amplification and pyrosequenced the V3 and V4 regions of the bacterial 16S ribosomal RNA gene. Amplicon sequencing libraries were sequenced on Illumina Miseq platform for paired end reads of 300 bp. Please refer to the Supplementary Material for further details on laboratory protocols and analysis methods.
Statistical analysis
Morris water maze experimental data were analyzed with SPSS 16.0 software. Data comparisons among different groups were analyzed by one-way ANOVA. Graphical presentations were performed with GraphPad Prism 5 software (GraphPad Software, San Diego, IL, USA).
RESULTS
Spatial learning and memory
The time that an APP/PS1 or WT mouse needed to reach the hidden platform was recorded as the escape latency for its spatial learning score. As shown in Fig. 1, there was no obvious difference in escape latency between 3-month-old APP/PS1 and WT mice, indicating that 3-month-old APP/PS1 mice had no impaired spatial learning compared with WT mice. In comparison, the escape latency of 6-month-old APP/PS1 mice was significantly longer than WT mice, and the difference further enlarged in 8-month-old mice. Thus, impaired spatial learning appeared in the 6-month-old APP/PS1 mice and was further aggravated in the 8-month-old group.
To assess the spatial memory changes in APP/PS1 mice, the percent time in the target quadrant (%) in the spatial probe test was recorded. As indicated in Fig. 2, percent time in the target quadrant in 6-month-old APP/PS1 mice decreased in comparison with WT mice, and this difference was more significant in the 8-month-old group. This suggested impaired spatial memory in 6- and 8-month-old APP/PS1 mice.
Amyloid plaque burden
A standard immunohistochemical study using an amyloid β-peptide (Aβ) antibody showed an age-dependent increase in plaque burden in the hippocampus of APP/PS1 mice. Positive Aβ immunoreactivity appeared at 6 months for APP/PS1 mice and Aβ plaque burden of 8-month-old mice increased significantly (Fig. 3). In WT mice, no positive Aβ immunoreactivity was observed.
Gut microbiota analysis
Paired-end sequencing of the V3-V4 regions of 16S rDNA genes were implemented on 36 samples. A total of 993,852 usable reads (27,607 per sample, reads length = 220–500 nt) were obtained from 36 samples. 714 operational taxonomic units (OTUs) in total were displayed at the 97% similaritylevel.
For the APP/PS1 mice and control groups, no statistically significant differences were observed for the alpha diversity indices (Shannon or Simpson indices; data not shown). However, for the APP/PS1 mice, by comparing the Shannon and Simpson indices from the 3-, 6-, and 8-months old groups (Shannon: 6.12, 5.59, and 5.39, p = 0.01; Simpson: 0.97, 0.95, and 0.94, p = 0.02), it was found that the alpha diversity of the APP/PS1 mice microbiota decreased with increased age, while this trend was not observed in the WT mice groups.
Figure 4 shows the 20 most abundant bacterial families in the gut microbiota of APP/PS1 and C57BL/6 mice in the 3-, 6- to 8-months old groups. Each bar represents the mean relative abundance of each bacterial taxon within a group. With the aim to keep as many microbiota taxa as possible for meaningful comparisons, we then identified the gut microbiome taxa with mean abundance all higher or lower in the three groups of APP/PS1 mice than the age-matched WT mice groups with significant difference (p < 0.05), respectively, at the family and genus levels with the threshold of mean relative abundance >0.0001. It was found that Helicobacteraceae and Desulfovibrionaceae mean abundance in APP/PS1 mice was significantly higher than that in WT mice at the family level (2.13 versus 0.09 and 0.47 versus 0.12). At the genus level, Odoribacter and Helicobacter in the APP/PS1 mice were significantly abundant in comparison with WT mice (mean abundance 0.84 versus 0.01 and 2.13 versus 0.09). Prevotella mean abundance in WT mice was significantly higher than APP/PS1 mice (0.09 versus 7.17).
The aforementioned impaired spatial learning/memory and Aβ plaque burden appeared at 6 months and further aggravated at 8 months. Further comparative analysis found a new species (Coriobacteriaceae) significantly more abundant in 6- and 8-month-old APP/PS1 mice than that in the age-matched WT mice at the family level, and Ruminococcus being significantly abundant in WT mice than in APP/PS1 mice at the genus level.
In addition, to explore the effect of age, we compared the gut microbiota abundance in the three groups of APP/PS1 mice. It was found that there was a significant increase of Prevotellaceae abundance from 3.05% for 3-month-old, 5.03% for 6-month-old, to 10.30% for 8-month-old APP/PS1 mice.
DISCUSSION
In recent years, there is a dramatically increasing interest in investigating the alterations in gut microbiota in many human diseases. Although a possible link between gut microbiota alteration and AD has been proposed [17–20], no previous study has compared the composition of the whole fecal microbiome between APP/PS1 and WT mice. In the present study, through pyrosequencing the V3 and V4 regions of the bacterial 16S ribosomal RNA genes, we compared the gut microbiota in APP/PS1 transgenic mice of AD and WT mice. It was found that the microbiota diversity of APP/PS1 mice decreased with increased age. Further inspection showed that Helicobacteraceae and Desulfovibrionaceae abundance at the family level, and Odoribacter and Helicobacter abundance at the genus level, increased significantly in APP/PS1 mice than in WT mice. In comparison, Prevotella abundance in WT mice was significantly higher than APP/PS1 mice.
The newly emerged concept of bidirectional microbiota–gut–brain axis in recent years should mainly account for the association between the gut microbiota alteration and AD [21, 22]. Increasing evidence supports that gut microbiota may exert effects on the central nervous system via neural, neuroendocrine, and neuroimmune mechanisms [23–26]. Harach et al. found that an axenic mouse model of AD without gut microbiota exhibited a drastic reduction of cerebral Aβ amyloid pathology when compared to controls with intestinal microbiota, which suggests a microbial involvement in the development of AD pathology [27]. In addition, Friedland proposed the potential pathways of molecular mimicry involving the gut microbiota in the initiation and development of neurodegenerative diseases [28]. More mechanistic studies are needed to elucidate the molecular basis underlying the association between gut microbiota alteration and AD.
Moreover, in addition to the concept of gut microbiota as a potential therapeutic target as discussed in many studies [29, 30], it has been proposed that gut microbiota alteration in AD will help better elucidate the pharmacology of many agents. For instance, polyphenolic compounds, which possess poor in vivo bioavailability, represent a type of promising agents for AD. In view of the expected high concentration of polyphenolic compounds in the gut tract after supplementation, it was previously proposed that gut microbiota regulation by polyphenols may account for an important pathway underlying their therapeutic benefits [31].
In conclusion, this is the first study on the composition of the whole fecal microbiome of AD transgenic mice, which suggests that histological and behavioral manifestations of APP/PS1 mice correlate with the specific gut microbiome state. More human studies are strongly encouraged to investigate the whole fecal microbiome between AD patients and controls and explore the potential microbiota diagnostic biomarkers.
Footnotes
ACKNOWLEDGMENTS
This work was supported by the Shandong Provincial Science Foundation for Distinguished Young Scholars (Grant No. JQ201508), Key program of Shandong Provincial Science Foundation (Grant No. ZR2015JL010), the National Science and Technology Major Projects of New Drugs (Grant No. 2015ZX09102015) and Shandong Provincial Science Foundation (ZR2014CL008).
