Human donors

Fecal material from human donors was obtained from patients followed at the Movement Disorders Unit of the Centro Hospitalar e Universitário de Coimbra. Outpatients with PD were recruited from outpatient clinics based on the following inclusion criteria: (a) meet the diagnosis of idiopathic PD according to the UK Brain Bank diagnostic criteria [38], (b) disease onset > 50 years of age (c) to have symptoms suggestive of constipation. As exclusion criteria: (a) hepatic, renal or cardiac failure, (b) severe hypertension, (c) other neurological disease, (d) head trauma, (e) any anti-inflammatory, antineoplastic or immunosuppressive treatment in the 3 months prior to sampling, drugs treatments, (f) any antibiotic treatment on the 3 months before sampling. Inclusion criteria for control subjects were limited to age > 50 years, and exclusion criteria were the same plus (a) consanguinity to any of the recruited individuals diagnosed with PD (b) family history of PD.

Blood samples were collected in anticoagulant tubes during a clinical visit by a trained practitioner as described below.

Ileal biopsies were performed by gastroenterology experts. A small piece of the terminal ileum (1–3 mm in diameter) was removed with a biopsy needle from patients who were previously sedated. The procedure was performed via colonoscopy. The tissue was postfixed for 24 h in fixative solution at 4 °C. For microbiome analysis, tissue samples were snap frozen at -70ºC. For immunohistochemistry tissue samples were cryoprotected using increasing concentrations of sucrose in PBS (10, 20 and 30%) and embedded in optimal cutting temperature (O.C.T.) (Tissue Tekª, ThermoFisher) as described previously [4]. Sections were cut at 20 µm of thickness on a cryostat (Cryostar NX50, ThermoScientific) at − 20 °C and mounted on SuperFrost© microscope slides (Thermofisher).

All procedures were approved by the Ethics Committees of the Faculty of Medicine and the University Hospital of University of Coimbra, and all patients and controls signed the informed consent report.

Animal model and experimental design

A total of 115, 32-week-old (adult) C57BL/6 male mice, divided during time into eight different cohorts, were used in this study in (46 untreated; 23 were colonized with healthy donor fecal material (HC mice) and 46 were colonized with PD patient fecal material (PD mice). Mice were purchased from Charles River (Barcelona, Spain) and housed in our animal colony (Animal Research Center, University of Coimbra), under controlled light (12 h day/night cycle), temperature and humidity (45–65%), with free access to standard hard pellet chow and water. Signs of distress were carefully monitored and although none occurred, a rapid decrease in body weight > 15–20% was defined as a potential humane endpoint for the study. The EU and Portuguese legislation (Directive 2010/63/EU; DL113/2013, 7 August) for the care and the use of animals was followed. All procedures were in accordance with the ethical standards of the Animal Welfare Committee of the Center for Neuroscience and Cell Biology and the Faculty of Medicine, University of Coimbra and the researchers received adequate training (FELASA certified course) and certification from the Portuguese authorities (Direção Geral de Veterinária) before performing the experiments.

To determine the impact of gut dysbiosis on the development and progression of PD pathology, we transplanted fecal material from 5 PD patients and 4 age-matched healthy donors into wild-type male C57BL/6 mice (each mouse received microbiota from a single donor). Seminal experiments were conducted utilizing samples from these 9 donors to substantiate the efficacy of fecal microbiota transplantation (FMT) from PD patients' gut microbiomes as a significant instigator of pathology. Individual WT mice colonized with microbiomes from either 4 healthy control (HC) individuals or 5 Parkinson's disease (PD) patients were used to assess various parameters: motor behavior, midbrain TH + neurons, ileum-associated mucosa SFB percentage, ileum CD11b + cells, ileum ZO-1 levels, ileum aSyn aggregates (Figure S3), ileal levels of TNF, IL-6, and IL-17, as well as midbrain levels of IL-1β and IL-17. The remaining experiments utilized mouse tissues sourced from healthy control donor HC1 and Parkinson's disease donor PD1. Mice were generated by colonizing 32-week-old wild-type animals with PD or HC fecal material, administered in gelatin once a day during the first 1 week and twice a week for 5 weeks (Fig. 1A). Additionally, we performed a longitudinal study to determine temporal propagation of pathology using 4 Unt and 5 PD mice after 1-week colonization (33 wks.); 4 Unt and 5 PD mice after 2-weeks colonization (34 wks.); 5 Unt and 5 PD mice after 3-weeks colonization (35 wks.) and 3 Unt and 3 PD mice after 4-weeks colonization (36 wks.) (Fig. 9A).

Fig. 1

Dopaminergic neurodegeneration and motor alterations in WT mice transplanted with fecal microbiota. A Schematic representation of experimental design (B) Balance and motor coordination performance were assessed with the beam walking test (n = 6–16 mice per group). C Hindlimb clasping reflex was monitored, as a quick phenotypic neurological scoring system to assess disease progression (n = 6–17 mice per group). D The inverted grip test was used to assess of limb muscle strength (n = 4–12 mice per group). E Odor discrimination and habituation was tested in time spent (s) non-social odors (n = 6–8 mice per group). F Representative photomicrographs of brain coronal sections immunostained for TH.+ in the striatum (STR) and substantia nigra (SN). G Total number of nigral TH-positive neurons in the SN assessed by stereological analysis (n = 8–11 mice per group). H OD analysis of the TH-positive fibres in the STR group normalized to the untreated group (n = 4–6 mice per group). I Striatal dopamine levels in pg/mL (n = 7 mice per group). *p < 0.05, **p < 0.01, ***p < 0.001, using one-way ANOVA with Dunnet´s test (G and H) or Kruskal–Wallis with Dunn´s test (B-D and I). Data are expressed as mean ± SEM. Scale bars are 100 µm (magnified inner square) and 1 mm. See also Tables S1-S2 and Figures S1-S4

Body weight was monitored twice a week throughout the study. Animals were also weighed immediately before euthanasia. Results were expressed in terms of body weight (g). Immediately after euthanasia, whole blood was collected from selected animals to determine occasional blood glucose levels by the glucose oxidase reaction, using a glucometer (Glucometer-Elite, Bayer SA, Portugal) and compatible strips. Results were expressed as mg glucose/dL blood. At the end of the experiments (38 weeks), fecal pellets were collected from animals individually housed in a clean cage.

Microbiome profiling

Human fecal material and terminal ileum mucosa biopsies, mouse fecal pellets and terminal ileum mucosa-associated material were collected for microbiome profiling. The mouse's ileum was extracted, flushed with sterile PBS, sectioned into pieces, and the luminal area was gently scraped and frozen. Human ileal biopsies were frozen. Microbial genomic DNA from frozen fecal and ileal samples was extracted using the NZY Soil gDNA Isolation Kit (NZYTech, Portugal), which includes a mechanical lysis step (with beads). The amount and quality of the extracted genomic DNA was assessed in a Nanodrop 2000 (Thermo Scientific). DNA integrity was assessed by PCR using universal primers for the 16S rRNA gene [27 F (5’-GAGTTTGATCCTGGCTCAG-3’) and 1525 R (5’-AGAAAGGAGGTGATCCAGCC-3’)] as previously described [39]. Total microbial DNA was sequenced at Genoinseq sequencing facilities (Cantanhede, Portugal) or at Novogene (UK) using the Illumina MiSeq® platform (Illumina, USA). Universal forward primer 515F-Y (5’- GTGYCAGCMGCCGCGGTAA-3’) and reverse primer 926R (5’-CCGYCAATTYMTTTRAGTTT-3’) [40] were used to target the V4-V5 hypervariable region using a standard protocol. Raw data processing, clustering and taxonomic annotation were performed using the mothur package version 1.44.1 (www.mothur.org) [41] and Silva reference files, release 138 [42]. To perform a comprehensive meta-analysis of the microbiome data, including community profiling and differential abundance, R (v 4.0.4) and the tidyverse package (v 1.3.0) were used to manipulate and visualize the data [42], while compositional analysis of the microbiota was performed using the phyloseq [42] and microbiome packages. Alpha diversity was measured on unfiltered data using the Shannon index, with statistical significance being assessed using the Mann–Whitney test and the Nemenyi test for pairwise comparisons, with Benjamini–Hochberg correction. For the remaining analysis, genera not detected in more than 1/3 of each sample group of samples were filtered. For beta diversity assessment, a non-metric multidimensional scaling (NMDS) analysis based on the Bray–Curtis dissimilarity index was used to visualize community-level similarity. Uneven sequencing depth in ileum samples was corrected by standardization to the median sequencing depth, and permuted analysis of variance (PERMANOVA) was used to assess statistical significance. Pie charts were generated from data transformed to relative abundances, after filtering for low prevalence (retaining only the genus present in 1/3 of the samples). Filtered data were also used to perform differential abundance analysis, with fold-change calculation and statistical assessment performed using the DESeq2 Wald test algorithm [42].

Behavior

At the end of the treatments, motor activity, anxiety-like behavior, spatial memory and olfactory function were assessed in our animals, as previously described [4]. All tests were performed under red light, in a sound-attenuated observation room where the mice had been habituated for at least 1 h. The apparatus was cleaned with 10% ethanol between animals.

Balance and subtle alterations in motor coordination, limb strength of the mice and severity of motor dysfunction were assessed using the beam walking, the inverted grid and the hindlimb clasping tests, respectively. Briefly, in the beam walking test, mice were allowed to traverse a narrow beam (8 diameter beam) to reach an enclosed safety platform in two consecutive trials, with a maximum time per trial of 90 s. Any animal that did not cross within the allotted time was assigned a maximum score of 90 s for analysis. For the grip test, each mouse was placed in the center of a metal grid and turned upside down so that the mouse was hanging and clinging to the grid (40 cm above the floor and with soft padding to cushion falling). The mice were allowed to move freely on the underside of the grid during a maximum “cling time” of 90 s. As an indicator of motor deficits, disease progression and neurodegeneration, the posture of the mice was assessed in the hindlimb clasping test. The tail of the mouse was grasped near its base and the animal was lifted away from all surrounding objects. The hindlimb position was then observed for 10 s and scored. A score between 0 and 3 was assigned depending on the clasping reflex observed for each mouse: a score of 0 was assigned if the hindlimbs were consistently extended away from the abdomen; a score of 1 was assigned if one hindlimb was partially retracted toward the abdomen for more than 50% of the time; a score of 2 was assigned if both hindlimbs were partially retracted toward the abdomen for more than 50% of the time; a score of 3 was assigned if both hindlimbs were completely retracted for more than 50% of the time.

Evaluation of locomotion, anxiety and stereotypical behaviors in mice was carried out with the open field test, through the analysis of the parameters of percentage of time resting and time spent in the center of the arena, number of feces expelled in the experimental time, mean velocity of mice and total distance travelled. This test used an Acti-Track System (PanLab, Barcelona, Spain) to track the activity of mice, for 30 min, in a 50 cm wide × 50 cm deep × 50 cm high arena.

Short-term spatial memory was assessed using the T-maze test (30 cm long × 10 cm wide × 20 cm high). The T-maze spontaneous alternation test was chosen because of its simplicity and minimal stress on the animals. At the beginning of each run, mice were placed at the start arm (bottom of the "T") and given 2 min to choose the right or left target arm. Once committed to a particular target arm (all four paws entered the arm), the "T" junction between the start arm and the opposite target arm was blocked to prevent the mice from entering the opposite target arm. The mice were allowed to explore the target arm for 30 s. With the "T" junction block removed, the mice were returned to the start arm and allowed to choose a target arm. An alternation was defined as the mouse entering an arm opposite to the one it entered on the previous run. Mice completed 5–6 trials. The time taken to choose one arm was also scored.

To investigate olfactory ability and the ability to discriminate between stimuli (odors), we used the Odor Discrimination and Habituation Test. In this non-invasive, spontaneous behavioral task, mice were presented to one of five odors (water and almond and banana extracts (non-social odors) and two social odors (obtained by swabbing the outside of two different cages of same-sex mice from this study) using a saturated cotton-tipped wood applicator (located on the bottom of the cage, 1 cm above the cage floor). Each odor was presented three times in a row. Mice with their nose within 2 cm of the applicator tip were considered to be exploring the odors. The evaluation consisted of a cross-habituation phase: time spent sniffing the applicator with a novel odorant stimulus (mice spontaneously recognize new odors) versus a habituation phase: time spent on the repeated stimulus.

Immunohistochemistry, immunofluorescence and microscopy analyses

Histological samples and post-mortem microscopic analyses were performed as described in previous studies [4, 43]. Briefly, tissue samples were collected after transcardial perfusion with saline (0.9% NaCl) followed by 50 mL of fixative solution (4% paraformaldehyde (PFA) and 0.1% glutaraldehyde in PBS). Brains and intestines were postfixed in fixative solution for 24 h at 4 °C. The intestines were previously rinsed with PBS and cut into 1 cm pieces. Tissue samples were cryoprotected using increasing concentrations of sucrose in PBS (10, 20 and 30%) as previously described [4, 43]. Coronal sections were cut at 20 µm of thickness on a cryostat (Cryostar NX50, ThermoScientific) at − 20 °C and mounted on SuperFrost© slides (Thermofisher).

In the gut samples, we determined CD11+ cell enrichment, ZO-1 integrity scoring, mitochondrial morphology of enteric neurons, aSyn oligomers and the percentage of Th17 cells (CD4+/IL-17+) were determined by immunofluorescence. CD4+ infiltration, aSyn and phosphorylated-aSyn (p-aSyn) expression by immunohistochemistry. The immunofluorescence method was performed as previously described [4]. Briefly, samples were incubated with rabbit anti-ZO-1 (Abcam, 1:300), mouse anti-CD11b (BioRad, 1:200), rabbit anti-CD4 (Cell Signaling, 1:200), rabbit anti-TOM20 (Santa Cruz, 1:400), mouse anti-β3-tubulin (Cell Signaling, 1:200), rabbit anti-aSyn aggregate (Abcam, 1:300) or mouse FITC-conjugated anti-IL-17 (Santa Cruz Biotechnology, 1:50) in PBS containing 1% donkey or goat serum and 0.25% Triton-X-100 for 24 h at 4° C. For mouse primary antibodies on mouse tissue, M.O.M. mouse Ig blocking reagent was applied 1 h prior to the blocking step. Secondary antibodies were donkey or goat anti-rabbit, or anti-mouse conjugated to Alexa Fluor 488 or Alexa Fluor 594 (Life Technologies, 1:250). Sections were stained with Hoechst 33342 (Sigma, 1:1000) before mounting with Mowiol© (Sigma). Immunofluorescence images were captured using an LSM710 (Zeiss) confocal microscope with different magnification objectives (10 × , 20 × and 40 ×) at a resolution of 1024 × 1024.

Immunofluorescence

For intestinal barrier integrity, the analysis was scored using a grading scale as previously described [4, 44, 45]. Between 7–10 images with 3–5 villi per image and animal were randomly acquired and blindly scored. b) For CD11b+ scoring, ten images were randomly acquired per animal, a total of 304 villi were analyzed (21–24 villi per animal), the number of CD11b+ cells was counted and divided by the total counting area (mm2). c) Mitochondrial network analysis was performed by TOM20 expression detected by immunofluorescence at 63 × magnification using an LSM710 microscope. Network parameters were obtained using an ImageJ macro MiNA [46] applied on at least five βIIITubulin-positive cells (outlined using the square tool) per image in tissue sections, and applied on total image in tissue sections. d) aSyn quantification was conducted using ImageJ, wherein the percentage of submucosa/muscularis propria area occupied by aSyn aggregates was measured. e) To calculate the percentage of Th17 cells (CD4+/IL-17+), ten images were taken randomly per animal. The number of CD4+ cells expressing IL-17 was expressed as a percentage, considering the total of CD4+ cells as 100%. Human biopsies were analyzed using the same criteria, but images were taken at 40 × magnification. Between 7–10 images with 2–4 villi per image and sample were randomly acquired and blindly scored.

Immunohistochemistry

Immunohistochemistry was performed as previously described [4, 43]. Briefly, sections were thawed, hydrated, treated for antigen retrieval, quenched and blocked. For mouse primary antibodies on mouse tissue, M.O.M. mouse Ig blocking reagent was applied 1 h prior to the blocking step. We determined: a) the number of CD4+ cells divided per the total counting area (mm2), b) the optical density (OD) of aSyn expression and c) the number of p-aSyn positive cells with rabbit anti-CD4 (Cell Signaling, 1:200), rabbit anti-aSyn (Abcam, 1:500) or mouse anti-p-aSyn (WAKO, 1:500) in PBS containing 1% goat or horse serum and 0.25% Triton-X-100 for 24 h at 4 °C. Secondary antibodies biotinylated goat anti-rabbit or anti-mouse IgG (Vector, 1:200) were diluted in PBS containing 0.25% Triton-X-100, followed by incubation with the avidin/biotin complex-HRP (VECTASTAIN Elite ABC Kit Standard, Vector Laboratories, CA, USA) for 30 min. Finally, the sections were counterstained with 1% cresyl violet. The tissue was dehydrated and mounted on DPX mounting medium (Sigma). CD4+ cell counts were estimated using the optical fractionator method in combination with the dissector principle and unbiased counting rules [47]. Between 4–5 coronal sections of the ileum were analyzed using Stereo Investigator software (MBF Bioscience) attached to an Axio Imager Z2 microscope (Zeiss). CD4+ cells were counted at 40 × magnification (1.4 numerical aperture, oil immersion). The grid size was 250 × 250 µm and the counting frames were 150 × 150 µm. The coefficient of error was calculated according to Gundersen and coworkers [47]. An error of CE < 0.1 (m = 1 class) was accepted for the analysis.

The level of aSyn expression was assessed by OD. Between 7–9 sections per animal were captured at 20 × magnification using the AxioScan slide scanner (Zeiss) and color deconvoluted using the “Color Deconvulation” plugin (https://imagej.net/Colour_Deconvolution) to measure the OD of DAB staining. OD was measured using ImageJ software (version 1.40 National Institute of Health). Images were converted to 8-bit grayscale and the mean intensity was quantified. Values were converted from pixels to OD using the Kodak No. 3 Calibrated Step Tablet template as a pattern curve. To assess the number of p-aSyn+ cells, 4 random images of coronal sections of the ileum were taken and quantified the number of p-aSyn+ cells in the entire section was quantified by counting the presence of intracellular staining. Four sections per animal were used. In human samples, the number of p-aSyn+ cells were quantified using a 150 × 150 µm counting frame in 10 different areas per Sect. (5 sections per sample).

In brain samples, immunofluorescence and immunohistochemistry protocols were performed as described above, except for a) IgG staining, which is a direct immunohistochemistry using biotinylated anti-mouse IgG (Vector, 1:1000). By immunofluorescence, we determined: b) the expression of Trem2 in Iba1+ cells in the SN and c) the estimated number of TH + and choline acetyltransferase positive (ChAT +) cells in the dorsal motor nucleus of the vagus nerve (DMV); and by immunohistochemistry, we assessed d) the expression of CD4 + in the SN; e) aSyn in STR, SN, DMV and CX and f) the expression of TH in STR by measuring the immunoreactivity by OD. We also performed g) the stereological quantification of TH + cells in the SN to determine the degree of dopaminergic neurodegeneration and the number of perivascular IgG-immunopositive staining per total area (mm2) in the cortex, striatum (STR) and SN to assess BBB integrity. Primary antibodies were rabbit anti-CD4 (Cell Signaling, 1:200), rabbit anti-Iba1, (Wako, 1:500), sheep anti-Trem2 1:200, R&D Systems), rabbit anti-TH (Millipore, 1:300), mouse anti-ChAT (ThermoFisher Scientific, 1:200), rabbit anti-aSyn (Abcam, 1:500 or mouse anti-p-aSyn (WAKO, 1:500). Secondary antibodies were donkey or goat anti-rabbit Alexa Fluor 488 (Abcam, 1:500), donkey anti-sheep Alexa Fluor 647 (Abcam, 1:500), goat anti-mouse Alexa Fluor 488 or anti-mouse Alexa Fluor 594 (Molecular Probes, 1:500), for immunofluorescence or biotinylated goat anti-rabbit or anti-mouse IgG (Vector, 1:200) for immunohistochemistry.

To assess the expression of Trem2 in Iba1+ cells in the SN, images of identical regions were acquired using a confocal microscope LSM710 (Zeiss) with a Plan-Apochromat 40 × /1.4 Oil DIC M27 objective at a resolution of 1024 × 1024. A total of six images per animal were randomly acquired across three different sections of the SN. Z-stacks were converted to maximum projection images using Fiji image software. Images were thresholded using the Triangle algorithm. To quantify the % area of Trem2 contained in Iba1+ cells, images were split into red and green channels and were converted to 8-bit images. To create a binary mask, a threshold was applied to both images to remove the background. The mask of the green channel (Iba1) was overlapped with the red mask (Trem2) and the ratio (%) of the red area (Trem2) within the green area (Iba1) was calculated. This was done using Fiji image software and the acquisition and analysis were performed blindly.

To determine TH immunoreactivity in the striatum, slides were scanned at 20 × magnification using the AxioScan slide scanner (Zeiss). A total of eight coronal sections systematically distributed along the anteroposterior axis of the striatum, with an evaluation interval of ten per animal, were quantified. OD was measured as described above. In this case, however, background staining was corrected by subtracting values obtained from adjacent cortical areas.

The number of TH + cells in the SN was estimated by stereological analysis as described above. In this case, a total of eight sections systematically distributed along the anteroposterior axis of the SN, with an evaluation interval of seven per animal, were included in the counting procedure. TH-positive cells were counted using a 40 × magnification objective (1.4 numerical aperture, oil immersion). The grid size and the counting frames were the same as described above. The coefficient of error was calculated according to Gundersen and coworkers [47]. An error of CE < 0.1 (m = 1 class) was accepted for the analysis.

To determine the expression of aSyn and the number of p-aSyn positive cells in the SN and DMV, we performed the same analysis as described above for the ileum samples. For the DMV and SN, between five and eight coronal sections systematically distributed along the anteroposterior axis, were quantified with an evaluation interval of five and seven per animal, respectively.

IgG immunostaining was performed as direct immunohistochemistry as it has been described in previous studies [4, 43]. Briefly, slides were scanned at 20 × magnification using the AxioScan slide scanner (Zeiss). The number of brain microvascular vessels with blood–brain barrier disruption was quantified in the cortex, STR and SN. Eight coronal sections systematically distributed along the anteroposterior axis were stained and quantified with an evaluation interval of ten per animal for cortex and striatum, and eight for SN. To assess BBB integrity, we quantified the number of microvascular leakages (IgG immunopositive staining in the perivascular area) per total area (mm2).

Plasma and peripheral blood mononuclear cells (PBMCs)

Twenty milliliters of venous blood was collected by venipuncture in K2- EDTA-containing tubes from both PD and disease-free HC individuals. Erythrocytes, plasma and PBMCs were isolated by Ficoll-Histopaque density gradient centrifugation, according to the manufacturer’s instructions. Briefly, blood samples were diluted with equal volume of Hanks’s balanced salt solution (HBSS), layered on Ficoll-Histopaque and centrifuged at 300 × g for 30 min at room temperature. Mouse blood was collected by cardiac puncture using a 23G needle syringe in animals previously deeply anesthetized with sodium pentobarbital (150 mg/kg) and placed in tubes coated with EDTA (0.5 M). Blood samples were transferred to 15 mL tubes containing Histopaque© 1083 solution (Sigma) and diluted (1:1) in phosphate-buffered saline (PBS). Tubes were centrifuged at 400 × g for 30 min at RT. PBMC halo and plasma were carefully collected with a Pasteur pipette and transferred to new tubes. PBMC halo in tubes containing 5 mL PBS were washed twice with PBS and centrifuged at 250 × g for 10 min at 4 ºC. Pellets of centrifuged PBMC halos were resuspended in lysis buffer, followed by three cycles of freezing and thawing in liquid nitrogen. Lysed samples were then centrifuged at 17,968 × g for 10 min, at 4ºC. The protein content of the resulting supernatants was determined using the Pierce™ BCA Protein Assay Kit.

Flow cytometry

PBMC pellet was incubated with anti-mouse CD45 PerCP (clone 30F11), anti-mouse CD3 FITC (clone REA641), anti-mouse CD4 APC (clone REA604) and anti-mouse CD8 PE (clone REA601) (1/50) (Miltenyi biotec) for 10 min at 4 ºC. The cell suspension was washed with PBS, centrifuged at 250 × g for 10 min at 4 ºC and the pellet was fixed with 2% PFA solution for 10 min at 4 ºC. Finally, the cells were centrifuged at 250 × g for 10 min at 4 ºC and the pellet was resuspended in PBS and analyzed by flow cytometry. The analysis was performed on the BD FACSCalibur cytometer (BD Bioscience), which was pre-set with voltage adjustments, compensated using single-stained cells, and the true background level was defined using the isotype control antibodies Rat Anti-IgG2a PerCP, REA Control-FITC, REA Control-PE and REA Control-APC (Miltenyi Biotec). The gating strategy was performed using FlowJo© software (BD Bioscience). More than 10,000 events were acquired in the region of interest (ROI), identified as the lymphocyte area in the forward versus side scatter dot plot. The percentage of CD4 and CD8 was obtained by gating the CD45+CD3+ events contained in the ROI.

Western blotting, spectrophotometry and ELISA determinations in brain and intestinal homogenates

At the end of the behavioral tests, mice were deeply anesthetized with isoflurane before being euthanized by cervical dislocation for tissue isolation. Specifically, the midbrain, striatum and ileum were isolated and immediately snap frozen and stored at − 80 °C until further analysis. For Western blotting, spectrophotometry and ELISA analysis, ileum, duodenum, jejunum and cecum were homogenized as follows: 1 cm pieces were cut, washed in ice-cold PBS and homogenized in hypotonic lysis buffer (0.1% Triton X-100, 25 mM HEPES, 2 mM MgCl2, 1 mM EDTA and 1 mM EGTA, pH 7.5) supplemented with 2 mM DTT, 0.1 mM PMSF and a 1:1000 dilution of a protease inhibitor cocktail. Samples were then frozen three times in liquid nitrogen, sonicated on ice (3 pulses) and centrifuged at 17,968 × g for 10 min at 4 ºC. Supernatants were collected and protein content was determined using the Pierce™ BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA) according to the manufacturer’s instructions for the plate reader. For Western blotting, mesencephalic tissue was homogenized in hypotonic lysis buffer (10% Triton X-100, 25 mM HEPES, 2 mM MgCl2, 1 mM EDTA and 1 mM EGTA, pH 7.5) supplemented with 2 mM DTT, 0.1 mM PMSF, 2 mM sodium orthovanadate, 50 mM sodium fluoride and a 1:1000 dilution of a protease inhibitor cocktail from Sigma (St. Louis, MO, USA). Samples were then frozen three times in liquid nitrogen, centrifuged at 20,000 × g for 10 min at 4ºC and the supernatants collected. For caspase-1 determination and ELISA kits, mesencephalic tissue was homogenized in 0.1% Triton X-100 containing hypotonic lysis buffer (10 mM HEPES; 3 mM MgCl2; 1 mM EGTA; 10 mM NaCl, pH 7.5), supplemented with 2 mM DTT, 0.1 mM PMSF and a 1:1000 dilution of a protease inhibitor cocktail from Sigma (St. Louis, MO, USA). Samples were then incubated on ice for 40 min and centrifuged at 2,300 × g for 10 min at 4 ºC. Supernatants containing the cytosolic fraction were collected. For the determination of dopamine levels, striatal tissues were sonicated in ice-cold 0.2 M perchloric acid, centrifuged at 13000 rpm, 7 min, 4 °C. Supernatants were collected while pellets were resuspended in 1 M NaOH. Protein content was determined using the Pierce™ BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA) according to the manufacturer’s instructions for the plate reader.

For the analysis of TLR4 and pro-IL-1β tissue lysates were loaded on SDS-PAGE gels under reducing conditions (resuspended and boiled for 5 min at 95ºC in 4 × Tris–Cl/SDS, pH 6.8, 30% glycerol, 10% SDS, 0.6 M DTT, 0.012% bromophenol blue). For the analysis of aSyn oligomers tissue lysates were loaded onto PAGE gels under non-reducing and non-denaturing conditions (suspended in 40% glycerol, 2% SDS, 0.2 M Tris–HCl pH 6.8, 0.005% Coomassie Blue). After electrophoresis, the samples were transferred to PVDF membranes (Millipore, Billerica, MA, USA) and after transfer the membranes were blocked for 1 h in Tris-buffered solution (TBS) containing 0.1% Tween-20 and 3% BSA. The membranes were then incubated with the appropriate primary antibodies at 4 °C with gentle agitation: 1:100 anti-TLR4 from Santa Cruz Biotechnology (Santa Cruz, CA, USA); 1:500 anti-pro-IL1β from Santa Cruz Biotechnology (Santa Cruz, CA, USA); 1:1000 polyclonal anti-aSyn, from Cell Signaling (Danvers, MA, USA). Membranes were reprobed with 1:1000 β-III tubulin from Cell Signaling (Danvers, MA, USA) to confirm equal protein loading. After incubation with the primary antibody, the membranes were washed three times with TBS containing 3% BSA and 0.1% Tween (5 min each time) and then incubated with the appropriate horseradish peroxidase-conjugated secondary antibody for 2 h at RT with gentle agitation. Membranes were washed three times and bound antibodies were detected by developing with an alkaline phosphatase enhanced chemical fluorescence reagent (ECF from GE Healthcare, Piscataway, NJ, USA). Fluorescence signals were detected using a Biorad Chemidoc Imager. Analysis of Western blot band densities was performed using Quantity One software (Bio-Rad).

To assess caspase-1 activation, 40 μg of tissue lysates were incubated with 100 μM of the colorimetric substrate for caspase-1 (Sigma Chemical Co., St. Louis, MO, USA) in reaction buffer (25 mM HEPES pH 7.5, 0.1% (w/v) 3-[(3-cholamidopropyl) dimethylammonio]-propanesulfonic acid (CHAPS), 10% (w/v) sucrose, 2 mM DTT) for 2 h at 37 °C, protected from light. Enzymatic cleavage of the substrate was detected at 405 nm using a Spectramax Plus 384 spectrophotometer (Molecular Devices, Sunnyvale, CA, USA).

Ileum, midbrain lysates and blood (25 μg) were used to determine inflammatory markers using ELISA kits, specifically TNF, IL-17, IL-10, IL-8, IL-6, IFNγ, IL-1β, and NFκB p65, according to the manufacturer’s instructions. Jejunum, duodenum and cecum lysates were used to determine IL-17 levels using an ELISA kit. Absorbance was detected at 450 nm using a SpectraMax Plus 384 multiplate reader. Results are expressed as μg/mL protein for NFκB p65 and as pg/mL for the other markers.

STR homogenates (50 μL) were used to calculate dopamine levels using an ELISA kit, according to the manufacturer’s instructions. Absorbance was detected at 450 nm in a SpectraMax Plus 384 multiplate reader. Results are expressed as pg/mL.

Mitochondria isolation and oxygen consumption rate (OCR)

Fresh mitochondria were isolated from mesencephalic and cortical areas using a discontinuous Percoll density gradient centrifugation as previously described by Ferreira and co-workers [48]. Small pieces from each area were isolated and washed in ice-cold isolation buffer (225 mM mannitol, 75 mM sucrose, 1 mM EGTA, 5 mM HEPES, pH 7.2/KOH). Mesencephalic and cortical mitochondria were then homogenized with 25 up and down strokes in a Dounce All-Glass Tissue Grinder (Kontes Glass Co., Vineland, NJ, USA) using pestle A (clearance: 0.07–0.12 mm) followed by 25 up and down strokes with pestle B (clearance: 0.02–0.056 mm). The tissues were then briefly centrifuged at 1100 × g for 2 min at 4 °C. Supernatants were collected and mixed with fresh ice-cold 80% Percoll solution prepared in 1 M sucrose, 50 mM HEPES, 10 mM EGTA, pH 7.0, then carefully layered on the top of fresh ice-cold 10% Percoll solution and further centrifuged at 18,500 × g for 10 min at 4 °C. The mitochondrial-containing pellet was then gently resuspended in 1 mL washing buffer (250 mM sucrose, 5 mM HEPES–KOH, 0.1 mM EGTA, pH 7.2). The mitochondrial-containing fractions were centrifuged again at 10,000 × g for 5 min at 4 °C and the final mitochondrial pellet was resuspended in ice-cold washing buffer. Protein content was determined using a Bio-Rad protein assay. 5 µg of fresh mesencephalic or cortical mitochondria were used for OCR measurements using a Seahorse XF24 Extracellular Flux Analyzer (Seahorse Bioscience, Billerica, MA, USA). Mitochondria were centrifuged at 2,200 × g for 20 min at 4 °C in a 24-well XF culture plate precoated with 1:15,000 polyethyleneimine (PEI) solution in mitochondrial assay solution (MAS: 70 mM sucrose, 220 mM mannitol, 10 mM KH2PO4, 5 mM MgCl2, 2 mM HEPES, 1 mM EGTA, pH 7.2) to adhere to the bottom of the plate [48, 49]. Isolated mitochondria were then incubated in MAS containing succinate (10 mM; Complex II substrate) plus rotenone (2 mM; Complex I inhibitor) for 8 min at 37ºC in a CO2-free incubator and then the plate was transferred to the Seahorse XF24 flux analyzer. First, 4 mM ADP was added to energize the mitochondria and then 2.5 μg/mL of oligomycin (inhibitor of ATP synthase) was added to prevent respiration derived from ATP synthesis. Next 4 μM of the uncoupler FCCP was added, which caused an increase in OCR reflecting the maximum respiratory chain activity as well as the maximum substrate oxidation rate. Finally, 4 μM of antimycin A (Complex III inhibitor) was added to block the respiratory chain and the remaining OCR.

The following determinations were calculated as previously described [48]: basal respiration: last rate measurement before first injection; maximal respiration: last rate measurement after FCCP injection; ATP synthesis: last rate measurement before oligomycin injection minus minimum rate measurement after oligomycin injection.

Statistical analysis

Microbiome population statistics are described in detail in the Microbiome Profiling section above. Statistical analysis of the datasets was performed using GraphPad Prism 8 software (GraphPad Software, San Diego, CA, USA). All data are presented as mean ± SEM. Normality analysis (Shapiro–Wilk’s test) was used to determine the subsequent parametric or non-parametric tests. Pairwise comparisons were performed using unpaired Student’s t-test or Mann–Whitney test. Multi-group comparisons were performed using one-way ANOVA followed by Dunnet’s post-hoc test or Kruskal–Wallis test followed by Dunn’s post-hoc test. All statistical tests were two-tailed, and the significance values were annotated as follows: *p < 0.05, **p < 0.01, ***p < 0.001. P and N values are given in each figure legend.