Chronic rotenone administration results in loss of tyrosine hydroxylase motor neurons independent of the gut microbiota

Rotenone administration has been shown to cause TH-neuron loss in a previous study,¹³ hence we quantified TH neuron using IHC in both GF (Figure 1b,d) and CR (Figure 1a,c) mice following rotenone treatment. We found that rotenone administration resulted in significantly greater TH-neuron loss within the substantia nigra region in both GF (70 ± 11.2 vs 22 ± 3.5, n = 4–5, P < .01, Figure 1b) and CR (64.9 ± 8.0 vs 34.6 ± 5.9, n = 4, P < .05, Figure 1a) mice when compared to vehicle-administered control mice (Figure 1a–d).

Figure 1.

Chronic rotenone administration causes motor dysfunction in CR mice but not in GF mice. Change in TH-neuron number in substantia nigra following treatment with rotenone (10mg/kg) or vehicle with representative images in CR (A, C) and GF mice (B, D). Changes in grip strength and time spent in rotarod in CR mice (E, G) and GF mice (F, H) respectively following treatment with rotenone (10mg/kg) or vehicle. The data are presented as mean ± SEM [unpaired Student's t-test (two-tailed); *P <0.05]  

The development of motor dysfunction following chronic rotenone administration is dependent on the presence of gut microbiota

To investigate the role of gut microbiota in PD pathogenesis, we examined the effect of chronic rotenone administration on motor function in GF and CR mice. There was a significant decrease in grip strength (0.49 ± 0.015 vs 0.42 ± 0.017, n = 9–10, P < .05) and an overall decrease in motor (19.75 ± 5.4 vs 8.0 ± 3.4, n = 4–5, P = .09) coordination as measured by rotarod test following administration of rotenone (10 mg/kg) for 6 weeks in CR mice (Figure 1e,g) but not in GF mice (Figure 1f,h) when compared to vehicle-treated mice. This suggests that motor dysfunction following rotenone administration is dependent on the presence of gut microbiota.

Chronic rotenone administration is associated with disruption in epithelial barrier only in the presence of gut microbiota

Chronic (6 weeks) but not acute (4 hours) rotenone administration in CR mice resulted in a significant increase in ex-vivo (proximal colon mucosa-submucosa preparation) 4 kDa FITC-flux (n = 3–4, Figure 2a–b). Chronic rotenone administration (6 weeks) also resulted in a significant increase in in-vivo 4 kDa FITC (n = 3–4, P < .05, Figure 2c) but not creatinine flux (data not shown) compared to vehicle-treated CR mice suggesting disruption of the leak but not the pore pathway. The lack of change with acute rotenone administration in CR mice suggests that the effect of rotenone on intestinal permeability is likely not due to a direct effect of rotenone on the epithelium. Next, to determine if the effect of rotenone on intestinal permeability is dependent on the presence of gut microbiota, we repeated the same experiments in GF mice. We found that chronic rotenone administration in GF mice had no effect on 4 kDa FITC-flux in ex-vivo proximal colon mucosa-submucosa preparation. However, in contrast to the findings in CR mice, acute rotenone administration in GF mice significantly decreased 4 kDa FITC-flux in ex-vivo proximal colon mucosa-submucosa preparation compared to vehicle-treated control mice (n = 4–5, Figure 2d–e) suggesting reinforcement of epithelial barrier in acutely rotenone treated mice. Together these data suggest that increased intestinal permeability following chronic rotenone administration in CR mice is dependent on the presence of gut microbiota.

Figure 2.

Chronic rotenone administration disrupts epithelial permeability and alters gut microbiota composition in CR mice. The cumulative fluorescein isothiocyanate (FITC) flux across the ex-vivo colonic mucosal explants obtained from (A) acutely rotenone and vehicle treated CR mice and (B) chronically rotenone and vehicle treated CR mice. (C) In-vivo FITC flux in CR mice subjected to chronic rotenone and vehicle treatment. Ex-vivo FITC flux across colonic mucosal explants in (D) acutely rotenone and vehicle treated GF mice and (E) chronically rotenone and vehicle treated GF mice. The data is presented as mean ± SEM. Difference between the slopes are calculated by F-test; in chronically treated CR: DFn = 1, DFd = 20; **P<0.01 and in acutely treated GF: DFn = 1, DFd = 28; **P<0.01. Plasma absorption of orally administered FITC in chronically vehicle control versus rotenone treated CR mice [Student’s t-test (two-tailed); *P <0.05]

Chronic rotenone administration in CR mice results in changes in gut microbiota composition similar to those observed in PD patients

To investigate the effect of chronic rotenone administration on gut microbiota composition, we characterized changes in gut microbiota composition in fecal samples from CR mice before and after treatment with rotenone or vehicle solution (control group) using 16S rRNA amplicon-based sequencing. There were no significant differences in beta diversity (based on Bray Curtis metric) or alpha diversity (based on Faiths Phylogenetic Diversity) following rotenone treatment when compared to vehicle treatment (Figure 3a,b). There were changes in the relative abundances of multiple bacterial classes after rotenone treatment such as Clostridia which was not seen following vehicle treatment (Figure 3c). We also found rotenone treatment resulted in an increase in the relative abundance of bacterial genera Lactobacillus, Bifidobacterium, Akkermansia, and Bacteroides spp and a decrease in Lachnospiraceae, Ruminococcaceae_UCG-014, Turicibacter, Faecalibaculum, and Clostridium spp when comparing changes in bacterial genera from baseline to week 6 in rotenone and vehicle-treated mice (Figure 3d). Interestingly, the microbial changes that we observed in our rotenone-treated mice such as increased Lactobacillaceae, Bacteroides, and decreased Lachnospiraceae have also been described in human PD studies and have been associated with worsening clinical symptoms, such as higher frequencies of gait disturbances, and postural instability.¹⁴ We also imputed changes in microbial metabolic pathways from 16S rRNA using the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States Predicted pathway function (PICRUSt2) and found the LipidII: glycine glycyltransferase pathway involved in bacterial peptidoglycan synthesis was significantly upregulated following rotenone treatment when compared to changes following vehicle treatment in control mice (Figure 3e).

Figure 3.

Changes in gut microbiota composition following rotenone or vehicle treatment in CR mice. (a) Change in α-diversity from baseline (BL) to week 6 (W6) following rotenone (blue) or vehicle (orange) treatment in conventionally raised (CR) mice was assessed using Faith’s phylogenetic diversity index. Kruskal–Wallis (pairwise) test, vehicle control: BL vs W6, q = 0.6; rotenone: BL vs W6, q = 0.6; BL: rotenone vs vehicle control, q = 0.7; W6: rotenone vs vehicle control, q = 0.7. A comparison of the change in α-diversity from BL to W6 following rotenone [RotenoneΔ(W6, BL)] or vehicle treatment [ControlΔ(W6, BL] shows no statistically significant difference (Wilcoxon Sign-Rank Test) between the treatment groups. (b) Bray Curtis-based PCoA plot (2-dimensional representation of 3-dimensional plot) of gut microbial communities in CR mice treated with vehicle solution (orange; control group) or rotenone solution (blue). BL samples are represented by open circles while W6 samples are represented by filled circles. There was no statistically significant difference in beta diversity at BL (PERMANOVA test; q = 0.26) or W6 (PERMANOVA test; q = 0.23) between rotenone and vehicle-treated mice. A comparison of the change in beta diversity from BL to W6 following rotenone [RotenoneΔ(W6, BL)] or vehicle treatment [ControlΔ(W6, BL] shows no statistically significant difference (Wilcoxon Sign-Rank Test) between the treatment groups. (c) Bar plot showing changes in relative abundance of different bacterial classes from BL to W6 following treatment of CR mice with rotenone (blue) or vehicle (orange). (d) Volcano plot showing compositional difference in gut microbiota between vehicle (control) and rotenone treated CR mice. Log2-transformed fold change in microbiota composition (RotenoneΔ(W6, BL)/ControlΔ(W6, BL)) is plotted on the x-axis and log-transformed, FDR adjusted p-values are plotted on the y-axis. Bacterial genera that are significantly increased [FDR adjusted P < .05] following rotenone administration are represented in red while those that are significantly decreased are represented in green. (e) Comparison of the change in predicted microbial metabolic pathways (PICRUSt) FDR<0.25) from BL to W6 following vehicle and rotenone treatment in CR mice [RotenoneΔ(W6, BL)/ControlΔ(W6, BL)]. Pathways (FDR<0.25) that are significantly increased are shown in red while those that are significantly decreased following rotenone administration are shown in green

Mice

Swiss Webster germ-free (GF) male and female mice aged 10–12 weeks were housed in a flexible film isolator in the Mayo Clinic gnotobiotic facility.^30,31 CR age/gender-matched Swiss Webster mice purchased from Taconic Farms. Mice were gavaged every day for 6 weeks with rotenone or vehicle as described in the protocol below. At the end of the sixth week, motor function was assessed and an in-vivo permeability assay was performed. Mice were then euthanized followed by a collection of brain and colon samples for IHC and ex-vivo permeability studies. Pellet samples were collected and stored at −80°C for 16S microbiota analysis. All mice studies were approved by the Mayo Clinic Institutional Animal Care and Use Committee and conducted in compliance with regulatory guidelines.

Gavage protocol

Rotenone solution was prepared by dissolving 10 mg/kg of rotenone (Tocris. Bio-Techne, Minneapolis, MN, USA) in a vehicle solution containing 4% carboxymethylcellulose and 1.25% chloroform. Age and weight-matched CR mice were randomly assigned to be gavaged every day either with freshly prepared rotenone solution (rotenone group) or vehicle solution (control group) for 6 weeks. GF mice were treated similarly with daily gavage, but sufficient volume of a freshly prepared batch of rotenone and vehicle only solution were made every 3–4 days under sterile conditions, and transferred into the isolators to reduce chances of contamination.

Quantification of tyrosine hydroxylase (TH) neurons

Loss of tyrosine hydroxylase (TH) motor neurons in substantia nigra (SN) was evaluated using immunohistochemistry (IHC). Briefly, brain samples from rotenone and vehicle solution administered mice were fixed in 4% paraformaldehyde for 48–72 h followed by 24–48 h of cryoprotection with 30% sucrose solution. Midbrain sections (~3 mm) were collected using Zivic mouse brain slicer (Zivic instruments, Pittsburgh, USA) and embedded in Tissue-Tek® OCT. freezing medium for cryostat sectioning. About 40 µm brain slices were washed 3X with 1× PBS solution at 10 min interval. Sections were blocked for 4 h at room temperature in 10% normal donkey serum (NDS) containing 0.6% Triton X-100 in 1× PBS. After 4 hours, rabbit anti-tyrosine hydroxylase (TH; Abcam ab75875 1:600) was added in a blocking serum containing 2% NDS + 0.1% Triton X-100 PBS. Sections were incubated overnight at 4°C and the primary antibody was removed and washed (3X; 15 min/wash). After washing, sections were incubated in secondary antibody solution containing donkey anti-rabbit FITC (Abcam ab6798; 1:400) in blocking serum for 4 h at room temperature shielded from light. The tissues were finally rinsed 3X with 1× PBS solution and mounted on slides with DAPI mounting medium followed by coverslips. Brain slices were analyzed using a fluorescent microscope (Leica microsystems, Leica Application suite X v 3.6.0).

Measurement of grip strength and motor coordination

Grip strength test was performed to evaluate forelimb motor strength using a grip strength meter (Chatillon, Columbus Instruments, Columbus, USA). The grip strength measurement mode was set to peak tension mode and units set to pounds-force (lbf). The mice were then gently removed from the cage and allowed to grip the metal bar tightly with both paws. Mice were slowly pulled away from the metal bar (or grid) in a straight horizontal line until their grip was broken and the peak force was recorded. This step was repeated 3 times without resting. After the third round of force measurement, mice were put back in their cages and allowed to rest for at least a minute. The entire test was repeated 5 times for a total of 15 pulls/mice. The grip strength was then averaged for the final analysis.

Rotarod test was used to examine motor coordination and sensory-motor function in mice. Mice were briefly trained and screened by placing mice on the rod at a fixed speed of 10 and 20 RPM. All test mice were screened for their ability to remain on the bar for at least 5 sec at 10 and 20 RPM. Mice were allowed at least three trials, and if a mouse was unable to stay in the bar for more than 5 sec during trials, the mouse was excluded from the study. After initial screening, the rotarod was used in an accelerating ramp mode with a minimum acceleration of 4 RPM and a maximum acceleration of 40 RPM with a ramp duration of 4 min. The time duration that the mice were able to remain on the bar was recorded. The process was repeated 3 times with a 2 min interval between each trial.

Ex-vivo intestinal permeability measurement

Proximal colon sections were obtained from rotenone and vehicle administered GF and CR mice. The colonic luminal contents were then quickly but gently flushed with chilled Krebs solution (composition in mM/l: 11.5 d-glucose/d-mannitol, 120.3 NaCl, 15.5 NaHCO3, 5.9 KCl, 1.2 NaH2PO4, 2.5 CaCl2 · 2H2O, and 1.2 MgCl2; pH 7.3–7.4) using a 30 ml syringe hooked to a 200 µl pipette tip. Tissues were then opened along the mesenteric border, flipped, and pinned flat on the Sylgard with mucosa side facing down. The underlying muscle layer was peeled away under a stereomicroscope with fine forceps and mucosa-submucosa preparations were mounted on Ussing chamber cassette. A 4 ml aliquot of 4 mg/ml of FITC (Fluorescein isothiocyanate 4 KDa; Sigma–Aldrich Catalog 60842–46-8), a well-characterized paracellular diffusion marker, was added to the mucosal side, and samples were obtained from the serosal side at regular intervals over a period of 3 h. The cumulative amount of FITC fluorescence accumulated in the serosal samples was assayed and plotted against time. The slope of the linear portion of the cumulative amount versus time profile provides an estimate of FITC flux across the mucosal explants.

In-vivo permeability measurement

In vivo permeability was measured using the method described previously.³² Briefly, on the day of measurement mice were fasted for 2 h then given a 200 µl oral gavage containing 100 mg/mL creatinine, 60 mg/mL FITC-4 kDa Dextran, and 40 mg/mL of rhodamine B isothiocyanate-70 kDa dextran (Sigma) to allow assessment of pore (radius 2.6 Å), leak (radius 13 Å) and unrestricted (radius 64 Å) barrier pathways, respectively.³² Mice were fasted for additional 2 h after which food was replaced. Five hours post-gavage, the intracardiac puncture was performed to harvest blood and collect serum samples. Serum samples were sent to the Mayo Clinic Immunochemical core for creatinine analysis while fluorescence for FITC and RITC was measured using a Synergy Mx (BioTek) at excitation wavelengths of 485 and 553 nm and emission wavelengths of 528 and 627 nm respectively. These were converted into absolute concentrations using standard curves.

16S rRNA amplicon-based sequencing and data analysis

Bacterial DNA was extracted from a stool after bead beating using the MoBio fecal DNA extraction kit (Mo Bio Laboratories, Carlsbad, CA, USA), followed by 16S ribosomal ribonucleic acid (rRNA) amplification using Nextera library compatible primers flanking the V4 hypervariable region ([forward overhang] +515 F: [TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG]GTGCCAGCMGCCGCGGTAA; and [reverse overhang] +806 R: [GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG]GGACTACHVGGGTWTCTAAT) and prepared for sequencing using a dual-indexing protocol. All samples were sequenced together in 2 × 300 paired-end mode on an Illumina MiSeq instrument by the University of Minnesota Genomics Center. All sequences and metadata are available at the short-read archive (Accession number: PRJNA673724).

Microbiome diversity analyses was done using QIIME 2 v2020.2,³³ while other data analyses, statistical tests, and visualizations were performed in R. Differences in alpha diversity were assessed based on Faith’s Phylogenetic Diversity metric. The Kruskal–Wallis (pairwise) test was used to test differences in alpha diversity between groups and within groups at baseline and week 6. Differences in beta diversity were computed based on the Bray Curtis metric and statistical significance between groups at each time point was assessed using PERMANOVA. Comparison of change in alpha or beta diversity from baseline to week 6 following treatment of CR mice with vehicle or rotenone was done in R using Wilcoxon Sign–Rank Test. The operational taxonomic unit (OTU) table for all samples generated in QIIME 2 was used to assess differences in the relative abundance of microbial genera using the DEseq2 package in R. Predictive functional profiling from 16S rRNA was performed using the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2)³⁴ and significant differences (FDR <0.25) among the groups are shown.