Fecal microbiota transplantation protects rotenone-induced Parkinson’s disease mice via suppressing inflammation mediated by the lipopolysaccharide-TLR4 signaling pathway through the microbiota-gut-brain axis

Male C57BL/6J mice aged 8 weeks (20–22 g) were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China). The mice were then acclimatized (12-h light/dark cycle) under standard conditions (temperature 22 ± 2 °C, humidity 50–60%) with ad libitum access to food and water for 7 days. All the procedures were performed following the guidelines developed by the Beijing Municipal Ethics Committee for the care and use of laboratory animals. The animal experiments were approved by the Animal Care & Welfare Committee, Institute of Materia Medica, CAMS & PUMC (No. 00005402).

Rotenone (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in 4% carboxymethylcellulose (CMC, Sigma-Aldrich) with 1.25% chloroform (Beijing Chemical Works, Beijing, China). The fresh rotenone solution was orally administrated (30 mg/kg body weight) by gavage once a day for 4 weeks [32, 38].

For FMT administration, fresh stools were collected from the control group mice, and 600 mg of feces were immediately placed into 6 mL of sterile saline and were steeped for 1 min. Then, the dissolved feces were centrifuged at 1000 g (4 °C) for 3 min. The suspension was collected and 100 μL of bacterial suspension was then delivered to each recipient mouse via oral gavage within 10 min.

To assess the motor function of each mouse, 4 behavioral tests were conducted with an interval of 1 h.

After fasting for 2 h, each mouse was removed from their home cages and placed in a clean and transparent plastic cage for 2 h. We collected the fecal pellets and counted the numbers. Subsequently, the wet weight was obtained by weighing the fresh feces and the dry weight was weighed after 24-h drying at 85 °C. The water content percentage was calculated based on the difference in the wet and dry weights of the fecal pellets.

To evaluate the gut motor function, 20-min fecal collection was performed. Each mouse was assigned into a clean and transparent plastic cage. The numbers of stool pellets were recorded every 5 min for 20 min.

30 min before the sacrifice, mice were orally treated with 0.3 mL 2.5% Evans blue (Sigma-Aldrich) dissolved in 1.5% CMC-Na (Sigma-Aldrich) to detect the intestinal transit distance. Then, the distance from the pylorus to the farthest end was recorded, which was defined as intestinal transit distance [40]. In addition, the colon length between the end of the cecum and the anus was measured.

Five mice randomly chosen from each group were well sedated, and then they were transcardially well-perfused with 0.9% saline infusion followed by 4% paraformaldehyde in 0.1 M phosphate buffer. The tissues of the colons and the brains were dissected and immediately placed in 4% paraformaldehyde for 24 h, and next in 4% paraformaldehyde containing 30% sucrose. The brain and the colon tissues were embedded in paraffin and sliced into 5-μm-thick slides. The immunofluorescence staining was conducted as described in our previous study [41]. Briefly, the slides went through antigen retrieval using sodium citrate solution (pH 6.0). After being blocked with 3% bovine serum albumin (Servicebio, Wuhan, China), the slides were incubated with primary antibodies at 4 °C overnight. Primary antibodies included: anti-tyrosine hydroxylase (TH, 1:1000, Servicebio), anti-α-synuclein (1:250, Invitrogen, Waltham, MA, USA), anti-ionized calcium-binding adapter molecule 1 (Iba-1, 1:1000, Servicebio), anti-ZO-1 (1:100, Thermo Fisher Scientific, Waltham, MA, USA), anti-TLR4 (1:50, Santa Cruz Biotechnology, California, USA), anti-glial fibrillary acidic protein (GFAP, 1:1000, Servicebio) antibodies. Then, the appropriate secondary antibodies, including FITC-labeled goat anti-mouse (1:400, Servicebio), goat anti-rabbit IgG-CY3 (1:300, Servicebio), and goat anti-mouse IgG-CY3 (1:300, Servicebio) were used to detect the corresponding primary antibodies. DAPI solution was used to detect nuclei. The representative images were captured by a fluorescence microscope (Nikon Eclipse C1, Tokyo, Japan). The numbers of positive cells were analyzed by Image Pro Plus 6.0 software. Each section was calculated based on 5 randomly chosen fields.

Five mice from each group were randomly chosen. The sliced colon slides of each mouse embedded in paraffin were stained with hematoxylin and eosin (H&E) following the descriptions in our previous research [41]. The histological score was blindly assessed by 2 researchers. Tissue damage was scored as the following criteria [42]: 0, no damage; 1, lymphoepithelial lesions; 2, focal ulceration or surface mucosal erosion; 3, broad mucosal damage involving deeper structures of the intestinal wall. Inflammatory cell infiltration score was evaluated according to the criteria below [42]: 0, few inflammatory cells in the lamina propria; 1, increased infiltration of inflammatory cells into the lamina propria; 2, the group of inflammatory cells infiltrating into the submucosa; 3, transmural infiltration of inflammatory cells. Then, the histological score was determined by combining the scores of tissue damage and inflammatory cell infiltration. Each section was calculated based on 5 randomly chosen fields.

The mice were well anesthetized and perfused with 0.9% saline, and both brain and colon tissues were dissected. Each sample was cut into cubes of 1 mm³ in size and placed into fixative at 4 °C for 4 hours. Then, the tissues were fixed in 1% osmium tetroxide at room temperature for 2 h, followed by dehydration using gradient alcohol. Subsequently, the tissues were embedded in resin and cured by baking at 60 °C for 48 h. Later, ultra-thin sections (60 nm) were cut by ultramicrotome. Finally, a transmission electron microscope (HITACHI, HT7700, Japan) was utilized to observe the ultrastructure of tight junctions both in the SN and the colon.

At sacrifice, the samples of spleen, liver, and mesenteric lymph nodes (MLNs) were aseptically collected from each mouse. The samples were smashed and suspended in 0.9% sterile saline at a ratio of 1:9, which meant that 0.1-g tissue samples were lysed with 0.9 mL 0.9% sterile saline. Then, 100 μL of the suspension was plated on a Lysogeny Broth agar plate. The numbers of colony-forming unit (CFU) were counted and analyzed after aerobic incubation at 37 °C for 24 h.

The intestinal barrier integrity was evaluated by the in vivo permeability assay using fluorescein isothiocyanate-dextran (FITC-dextran, MW: 4 kDa, FD4, Sigma-Aldrich). Briefly, mice were maintained without food for 4 h and then they were orally treated with FD4 at 0.6 mg/g body weight. After another 4 h, mice were euthanized and exsanguinated by retro-orbital puncture. Immediately, serum fluorescence intensity was detected by a fluorescence spectrophotometer (485/525 nm) and FD4 concentration in the serum was further calculated based on a standard curve generated by serial dilutions of FD4.

For total protein extraction, RIPA lysis buffer (C1055, APPLYGEN, China) with protease inhibitor cocktail (P1265, APPLYGEN, China) and phosphatase inhibitors mixture (P1260, APPLYGEN, China) was utilized to lyse the tissues. Then, the lysate was centrifuged at 12,000 g, 4 °C for 20 min to obtain the total protein. To separate cytoplasmic and nuclear proteins, the Nuclear/cytoplasmic Isolation Kit (P1201, APPLYGEN, China) was used according to the manufacturer's instruction. As described in our previous study [43], the protein concentrations were detected by the BCA Protein Assay Kit (P1511, APPLYGEN, China) and western blot was performed using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) method. The transferred membranes were incubated at 4 °C overnight with the following primary antibodies: mouse anti-TH antibody (1:1000, MAB318, Millipore), rabbit anti-tumor necrosis factor-α (TNF-α) antibody (1:500, AF7014; Affinity Biosciences, Cincinnati, OH, USA), rabbit anti-interleukin-1β (IL-1β) antibody (1:1000, ab200478; Abcam, Cambridge, MA, USA), rabbit anti-interleukin-6 (IL-6) antibody (1:1000, ab229381; Abcam), rabbit anti-inducible NO synthase (iNOS) antibody (1:1000, A0312; Abclonal, China), goat anti-cyclooxygenase 2 (COX2) antibody (1:1000, ab23672; Abcam), mouse anti-α-synuclein (1:1000, 32-8100; Invitrogen), mouse anti-TLR4 antibody (1:400, ab22048; Abcam), rabbit anti-MyD88 (1:400, ab2064; Abcam, Cambridge, MA, USA), rabbit anti-IκB-α (1:400, sc-371; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), rabbit anti-ZO-1 antibody (1:1000, ab96587, Abcam), mouse anti phosphorylated-IκB-α (1:400, sc-8404; Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA), rabbit anti-NF-κB p65 antibody (1:1000, A2547, Abclonal, China), mouse anti-claudin-1 antibody (1:1000, 2H10D10, Invitrogen), rabbit anti-occludin antibody (1:1000, 27260-1-AP, ProteinTech), mouse anti-claudin-5 antibody (1:1000, A10207, Abclonal, China). The membranes were then incubated with the corresponding secondary antibodies, including HRP anti-mouse antibody (1:2000, AS003; Abclonal), HRP anti-rabbit antibody (1:2000, AS014; Abclonal), HRP anti-goat antibody (1:2500, AS031; Abclonal) for 2 h at room temperature. The blots were visualized using LAS4000 chemiluminescence system (Fujifilm, Tokyo, Japan), and the densities were analyzed by Gel-Pro Analyzer 4.0 software.

The enzyme-linked immunosorbent assay (ELISA) kits used to detect the levels of mouse TNF-α, IL-1β, IL-6, LPS endotoxin, and lipopolysaccharide binding protein (LBP) were purchased from Jianglai Industrial Limited By Share Ltd, Shanghai, China. The experimental procedures were carried out according to the manufacturer's instructions, which is similar to the descriptions in our previous study [41]. The concentrations of target proteins were determined by standard protein curves.

Mice were randomly chosen from each group at week 4 and week 6 for microbiota sequencing analysis. After placing each mouse in a separate empty autoclaved cage, 6-8 fresh fecal pellets from each mouse were collected and immediately put into a sterile EP tube. All the fecal samples were instantly frozen and stored at – 80 °C for further analysis. Microbial genomic DNA was extracted from fecal samples by a QIAamp Fast DNA Stool Mini Kit (Qiagen, Germany) following the improved protocol based on the manufacturer's instructions. The V3-V4 regions of the microbial 16S RNA were amplified with the paired primers (forward primer: 5′-CCTACGGGRSGCAGCAG-3′; reverse primer: 5′-GGACTACVVGGGTATCTAATC-3′). The following condition was used: 95 °C for 3 min, followed by 30 cycles at 98 °C for 20 s, then 58 °C for 15 s, and 72 °C for 20 s and a final extension at 72 °C for 5 min. All the quantified amplicons were pooled together at equalized concentrations for Illumina MiSeq sequencing (Illumina, Inc., CA, USA). Experiments including DNA extraction, quality assessment, library construction, and high-throughput sequencing were performed by Realbio Genomics Institute (Shanghai, China).

Sequencing data were processed and analyzed by Realbio Genomics Institute (Shanghai, China). Briefly, PANDAseq (V2.9) was used to filter the clean reads by removing the reads with the average Phred score lower than 20, removing the reads containing more than 3 ambiguous bases, and retaining the reads of 250–500 nt [44]. Then, operational taxonomic units (OTUs) were obtained using Usearch (V7.0.1090) [45] by clustering the sequences with similarities over 97%. Then, OTUs were annotated by comparing representative sequences to the Ribosomal Database Project (RDP, http://rdp.cme.msu.edu↗) [46, 47]. Alpha-diversity indices (Shannon index and Simpson index) were calculated by QIIME (V1.9.1) [48] and the differences among the 3 groups were analyzed by the Kruskal-Wallis test on R (V3.5.1). Beta-diversity analysis was performed using weighted UniFrac distances. Thereafter, the difference comparisons among different groups were conducted by Adonis and were displayed by the principal coordinate analysis (PCoA) method. Besides, the analysis of similarity (ANOSIM) method was also performed to compare the group differences. Linear discriminant analysis (LDA) was used for comparing the differences of microbial abundances at different taxon levels by LDA EffectSize Tools (V1.0) [49]. Then, the microbiota community structures were displayed by Graphical Phylogenetic Analysis software. The correlations between the bacterial relative abundances and other experimental results were performed using Spearman correlation analysis and were displayed by heatmap on R (V3.5.1). Eventually, the functional prediction analysis was performed by the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) method to determine Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways [50]. The analysis was conducted using R (V3.5.1).

The statistical analysis was conducted by using SPSS software (version 20.0). All the data were displayed as mean ± standard deviation (SD). Statistical analysis for multiple comparisons was performed by one-way analysis of variance (ANOVA) followed by the least significant difference (LSD) post-hoc test. For pole test performance, grip strength test performance, histological score, ZO-1 integrity score, and cell numbers, the non-parametric Kruskal-Wallis test was used, followed by the Mann-Whitney U post-hoc test. For 2 group comparisons, the independent t test was utilized. To obtain the correlations between different experiments, Spearman correlation analysis was performed by R (V3.5.1). When the P value was < 0.05, the results were considered statistically significant.

Weight loss, motor disorders, and GI dysfunctions commonly occur in PD animal models. To evaluate the protective effects of FMT administration on PD, we established a chronic PD mouse model by orally treating the mice with rotenone for 4 weeks. In the following 2 weeks, the rotenone-challenged mice were administrated with FMT or vehicle (Fig. 1A). During the establishment of the PD model, the rotenone group mice began to show significant weight loss from day 20 (data not shown). As illustrated in Fig. 1B, FMT treatment markedly alleviated the weight loss of rotenone-challenged mice in weeks 5 to 6. At week 6, 4 kinds of behavioral tests were performed to assess the motor functions of mice from different groups, including the Rota-Rod test for motor coordination, the adhesive removal test for sensorimotor integration, the grip strength test for muscle force of limbs, and the pole test for motor balance (Fig. 1A). The rotenone-intoxicated mice were present with significant motor disorders compared to the control group mice, including decreased time on the rod (P < 0.001, Fig. 1C), delayed removal from the stickers (P < 0.001, Fig. 1D), poorer performance of the limb muscle force (P < 0.01, Fig. 1E), and decreased scores of the pole test (P < 0.01, Fig. 1F). On the contrary, the performances of the FMT group mice improved remarkably in the Rota-Rod test (P < 0.001, Fig. 1C), the adhesive removal test (P < 0.001, Fig. 1D), the grip strength test (P < 0.01, Fig. 1E), and the pole test (P < 0.001, Fig. 1F) compared to the rotenone group.

In addition, GI dysfunctions were also measured in our study. Evans blue was applied to indicate the intestinal transit functions of mice and the colon length was measured. The rotenone group mice showed a significantly decreased intestinal transit distance (P < 0.001, Fig. 1G) and colon length (P < 0.001, Fig. 1H) compared to the control group mice. On the contrary, FMT treatment remarkably attenuated these two dysfunctions (both P < 0.001, Fig. 1G-H) induced by rotenone. Besides, fecal pellets were collected to measure the water content percentage and the colon motility. The rotenone-challenged mice showed markedly reduced fecal water content percentages (P < 0.001, Fig. 1I) and pellet numbers (P < 0.001, Fig. 1J), which were significantly elevated by FMT administration (water content percentage, P < 0.001; pellet number, P < 0.01) (Fig. 1I-J). Moreover, remarkable differences in fecal output frequencies (indicated by a 20-min fecal collection experiment) between the rotenone group mice and the control group mice were observed in the study (Fig. 1K) while FMT treatment significantly increased the fecal output frequencies of rotenone-challenged mice (Fig. 1K). Taken together, these data suggest that rotenone intoxication induces weight loss, motor disorders, and GI dysfunctions in the PD mouse model whereas FMT administration significantly attenuates these PD manifestations.

It has been reported that the activation of microglial cells and astrocytes may contribute to neuroinflammation and dopaminergic neuronal death in the SN during PD development [51, 52]. In our present study, we performed immunofluorescence staining of the SN region to detect the activation of glial cells, using GFAP as the marker of astrocytes, and Iba-1 as the marker of microglia. Both GFAP⁺ and Iba-1⁺ cells showed significant elevation in the SN region of the rotenone group mice (both P < 0.001, Fig. 2A, C, D, and F). However, FMT administration markedly attenuated the elevation of GFAP⁺ cells (rotenone group, 28 cells/field; FMT group, 15 cells/field; P < 0.001) and Iba-1⁺ cells (rotenone group, 37 cells/field; FMT group, 21 cells/field; P < 0.001) in the SN region of rotenone-intoxicated mice (Fig. 2A, C, D, and F).

In addition, immune infiltration and epithelium impairment in the colon were evaluated by H&E staining. The H&E histological scores were remarkably higher in the rotenone group than the control group (P < 0.01) while FMT administration reduced the scores significantly (P < 0.05) (Fig. 2M, N). Collectively, these results illustrate that FMT treatment significantly attenuates PD-related histological characteristics and inflammation both in the SN and the colon of rotenone-challenged mice.

Besides, the functions and the structures of the intestinal barriers were also tested. First of all, the in vivo intestinal permeability assay was performed, in which the higher intensity of FD4 detected in the serum represents the elevated gut permeability. The increased concentrations of FD4 were shown in the serum of the rotenone group in comparison with the control group (P < 0.001) while FMT administration remarkably attenuated the elevation of serum FD4 concentrations (P < 0.01) (Fig. 3G). Moreover, we conducted a bacterial translocation experiment to reflect the integrity of the gut barriers. Translocated bacteria were detected in the germ-free organs, including the liver (P < 0.01), spleen (P < 0.01), and MLNs (P < 0.01) of the rotenone group mice while FMT treatment remarkably suppressed the translocation (Liver, P < 0.01; Spleen, P < 0.01; MLNs, P < 0.05) (Fig. 3P). Furthermore, we conducted TEM analysis to detect the structures of intestinal barriers. The results illustrated the impaired epithelium with sparse microvilli, as well as decreased and discontinuous electron-dense materials with wider paracellular gaps in the colon slides of rotenone-challenged mice (Fig. 3F). However, the epithelium was present with regularly organized microvilli and intact tight junctions after FMT treatment (Fig. 3F). Additionally, the expression of tight junction proteins in the colon was also measured by different methods. Immunofluorescence staining results showed disrupted and decreased expression of ZO-1 in the rotenone group (P < 0.001) whereas FMT administration significantly preserved the ZO-1 integrity (P < 0.001) (Fig. 3H, I). In addition, western blot results demonstrated that the expression of ZO-1 (P < 0.05), occludin (P < 0.05), and claudin-1 (P < 0.01) was remarkably reduced in the colon samples of the rotenone group mice compared to the control group mice but elevated in the FMT group mice (ZO-1, P < 0.01; occludin, P < 0.05; claudin-1, P < 0.05) (Fig. 3J-M). Consistently, mRNA expression of these important genes in the colon also decreased significantly in the rotenone group (all P < 0.001) while FMT treatment restored the declined expression (ZO1, P < 0.01; Ocln, P < 0.01; Cldn1, P < 0.001) (Fig. S1F-H).

To further test whether the restoration of the intestinal barriers can lead to the decline of leaked pro-inflammatory molecules, we measured the levels of LPS, LBP, and inflammatory cytokines in the circulation. The ELISA analysis revealed that both serum LPS and LBP in the rotenone group were elevated to more than twice those of the control group (both P < 0.001, Fig. 3N-O). By contrast, the levels of LPS (rotenone group, 1.09 EU/mL; FMT group, 0.95 EU/mL) (Fig. 3N) and LBP (rotenone group, 0.42 EU/mL; FMT group, 0.35 EU/mL) (Fig. 3O) in the serum were remarkably reduced by FMT administration (both P < 0.001). In addition, significantly higher levels of pro-inflammatory cytokines (TNF-α, IL-1β, and IL-6) were detected in the systemic circulation of the rotenone group compared to the control group (all P < 0.001) by ELISA analysis whereas FMT administration restored the elevation (all P < 0.001) (Fig. 3Q-S). All these data suggest that FMT treatment preserves the BBB damage and the intestinal barrier impairment in the rotenone-induced mouse model.

Likewise, we measured the TLR4 pathway activation in the colon samples, which was consistent with the results in the midbrain tissues. As revealed by western blot, the rotenone group showed a significantly higher expression of proteins, including TLR4 (P < 0.01), MyD88 (P < 0.001), p-IκB-α (P < 0.001), and NF-κB (P < 0.05), as well as the lower expression level of IκB-α protein than the control group (P < 0.01) (Fig. 4L-Q). On the contrary, FMT administration illustrated the inhibition of TLR4 pathway in the colon by reducing the expression of TLR4 (P < 0.05), MyD88 (P < 0.001), p-IκB-α (P < 0.001), NF-κB (P < 0.05) and increasing the expression level of IκB-α (P < 0.01) (Fig. 4L-Q)). Consistently, the relative mRNA expression of Tlr4 in the colon was also remarkably elevated in the rotenone group (P < 0.001) but was suppressed by FMT administration (P < 0.001) (Fig. 4R). In addition, the immunofluorescence staining results revealed that the numbers of TLR4⁺ cells notably increased in the rotenone group (P < 0.001) whereas FMT treatment significantly reduced the numbers (P < 0.001) (Fig. 4U, V). Clearly, all the evidence supports that FMT administration inhibits the TLR4/MyD88/NF-κB signaling pathway activated by LPS both in the SN and the colon of rotenone-challenged mice.

As demonstrated in Fig. 6E and F and Table 2, the overall significant differences among the 3 groups were detected in the genera Akkermansia (P = 0.0001), Alistipes (P = 0.0095), Barnesiella (P = 0.0427), Butyricicoccus (P = 0.0108), Desulfovibrio (P = 0.0066), Helicobacter (P = 0.0008), Lactobacillus (P = 0.0054), and Saccharibacteria genera incertae sedis (P = 0.0191). We further selected 6 genera and utilized boxplots to display their detailed information (Fig. 6G-L). Compared to the control group, the genera Akkermansia (P = 0.0002; Fig. 6G) and Desulfovibrio (P = 0.0018; Fig. 6J) showed significantly elevated relative abundances whereas the genera Helicobacter (P = 0.0180; Fig. 6K) and Roseburia (P = 0.0124; Fig. 6L) had the notably decreased abundances in the rotenone group. However, FMT treatment remarkably reduced the abundances of Akkermansia (P = 0.0009; Fig. 6G) and Desulfovibrio (P = 0.0500; Fig. 6J), while the increased relative abundances of Barnesiella (P = 0.0207; Fig. 6H), Butyricicoccus (P = 0.0053; Fig. 6I), Helicobacter (P = 0.0002; Fig. 6K) and Roseburia (P = 0.0029; Fig. 6L) were witnessed in the FMT group. In addition, the comparisons of bacteria at different taxon levels between the control group and the FMT group revealed that most of them had no significant difference, suggesting that these 2 groups shared similar gut microbiome profiles (Table 2 and Fig. 6E-L and Fig. S3A-F). Collectively, differential abundances of gut microbes at various taxon levels among all the 3 groups are summarized in Fig. 6M.

To better interpret the functions of these significantly altered bacteria, we performed a functional analysis of KEGG pathway enrichment. A total of 21 KEGG pathways at level 2 were identified (Fig. S3G). Among them, the immune system was enriched in the control group. Meanwhile, pathways such as cellular processes and signaling, xenobiotics biodegradation and metabolism, and lipid metabolism were enriched in the rotenone group. Besides, replication and repair, nucleotide metabolism, translation, and other pathways were enriched in the FMT group (Fig. S3G). Additionally, level 3 analysis revealed that 2 pathways (DNA repair and recombination proteins and energy metabolism) were enriched in the control group whereas 17 pathways, including bacterial secretion system, secretion system, lipopolysaccharide biosynthesis, transcription machinery, replication recombination, and repair proteins, were enriched in the rotenone group (Fig. S3H). In addition, there were 16 level 3 pathways enriched in the FMT group, including pyrimidine metabolism, ribosome, methane metabolism, and chromosome (Fig. S3H). The correlations between the abundances of bacterial genera and other experimental results were analyzed to further demonstrate the effects of microbiota on the rotenone-challenged mice. As illustrated in Fig. 6N, the abundances of some genera like Akkermansia, Barnesiella, Desulfovibrio, and Helicobacter showed significant correlations with many other experimental results, including the behavioral tests, the GI function tests, the TH-positive cell numbers in the SN, and some inflammation-related data. Together, these data provide evidence that FMT administration attenuates microbiota dysbiosis to protect the rotenone-intoxicated mice, suggesting the critical role of microbiota dysbiosis in the development of PD.

Additional file 1: Fig. S1 FMT treatment alleviates the expression of tight junction proteins reduced in the rotenone-challenged mouse model. (A) Representative captures of immunofluorescence of ZO-1 in the SN. (B) The intensity analysis of ZO-1 immunofluorescence staining in the SN. (C-E) mRNA expression of tight junction proteins ZO1, Ocln and Cldn5 in the midbrain containing the SN. (F-H) mRNA expression of tight junction proteins ZO1, Ocln and Cldn1 in the colon. For (B), n = 5 for each group. For (C-H), n = 3 for each group. Data are presented as mean ± SD. ## p < 0.01, ### p < 0.001 versus the control group; ** p < 0.01, *** p < 0.001 versus the rotenone group.Additional file 2: Fig. S2 FMT administration suppresses the generation of pro-inflammatory molecules both in the SN and the colon of rotenone-challenged mice. (A-E) mRNA expression of pro-inflammatory cytokines (Tnf, Il1b, Il6, Nos2, COX2) in the midbrain containing the SN and the colon. For (A-E), n = 3 for each group. Data are presented as mean ± SD. ## p < 0.01, ### p < 0.001 versus the control group; * p < 0.05, ** p < 0.01, *** p < 0.001 versus the rotenone group.Additional file 3: Fig. S3 FMT treatment attenuates microbiota dysbiosis of rotenone-intoxicated mouse model. (A) Relative abundances of gut microbiota at the phylum level in the 3 groups. (B-C) Relative abundances of significantly altered bacterial phyla: Verrucomicrobia and Proteobacteria. (D) Relative abundances of gut microbiota at the family level in the 3 groups. (E-F) Relative abundances of significantly altered bacterial families: Verrucomicrobiaceae and Lactobacillaceae. (G) Bar graph of LDA scores of enriched KEGG pathways at level 2. LDA scores (log10) > 2 and P < 0.05 are shown. (H) Bar graph of LDA scores of enriched KEGG pathways at level 3. LDA scores (log10) > 2.7 and P < 0.05 are shown. In this figure, n = 8 for each group. Each boxplot represents the median, interquartile range, minimum and maximum values. # p < 0.05, ### p < 0.001 versus the control group; ** p < 0.01, *** p < 0.001 versus the rotenone group.