Male albino Wistar rats (P30-40; 200–250 g) were purchased from Charles River Laboratories, United Kingdom. Animals were housed in groups of 4/5 at the UCL School of Pharmacy and had ad libitum access to food (Teklad Global 18% Protein Rodent Diet, Envigo) and water. Conditions of humidity (40–60%), temperature (18–22°C) and a 12-h light–dark cycle were kept constant in line with the Home Office regulations. All experiments were approved by the Bloomsbury Animal Welfare and Ethical Review Body (AWERB) and United Kingdom Home Office (PPL PP3144142). Animals were randomly allocated to experimental groups. The early-stage PD rat model was induced as previously described (Sancandi et al., 2018). Briefly, the noradrenergic neurotoxin N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine hydrochloride (DSP-4, 25 mg/kg in saline solution, Sigma Aldrich, Gillingham, United Kingdom) or saline for sham animals were injected intraperitoneally, 4 days prior to stereotaxic striatal bilateral injection of the dopaminergic neurotoxin 6-hydroxydopamine (6-OHDA, 5 mg/mL in saline solution containing 0.9% ascorbic acid, Sigma Aldrich, Gillingham, United Kingdom) or saline for sham animals. The following coordinates were used, from Bregma: AP + 1.0 mm, ML + 3.0 mm, DV-6.5 mm. Each animal received 15 mg of 6-OHDA per striatum (or vehicle) at a flow rate of 1 mL/min⁻¹. Both Symprove™ (original flavor) and placebo (matched flavor) were kindly supplied by Symprove Ltd. (Farnham, United Kingdom). 0.5 mL of placebo or Symprove™ (7×10⁷ colony forming units (CFU)/day) was orally delivered daily from Day 1 for 24 days (Figure 1). Placebo (sham and model + placebo) and Symprove™ (sham and Model + Symprove™) (n = 4/5 animals per group) experiments were carried out separately to minimize any risk of contamination. The weight of the animals was monitored daily throughout the duration of the experiments. At Day 24, stool samples were collected and frozen. Animals were anaesthetized with inhaled isoflurane (4% v/v in O₂) followed by an intraperitoneal injection of euthatal (Merial, Harlow, United Kingdom) at a dose of 60 mg/kg. Blood samples were drawn from the heart of anaesthetized animals, and animals were then perfused with ice-cold artificial cerebrospinal fluid (aCSF) containing in mM: 124 NaCl, 25.5 NaHCO₃, 3.3 KCl, 1.2 KH₂PO₄, 1 MgSO₄, 2.5 CaCl₂, 15 mM D-Glucose equilibrated with 95% O₂/5% CO₂. The ileum (5 cm above the caecum) of each animal from each experimental group was extracted as previously described (Citraro et al., 2021) and frozen for further investigation. Brains were also removed and fixed overnight (4% paraformaldehyde, 0.2% saturated picric acid solution, 0.025% glutaraldehyde solution in 0.1 M phosphate buffer) prior to histological procedures.

Small intestine samples were extracted and homogenized in ice-cold lysis buffer [20 mM Tris–HCl (pH 7.5), 10 mM NaF, 150 mM NaCl, 1% Nonidet P40, 1 mM phenylmethylsulfonyl fluoride, 1 mM Na₃VO₄, leupeptin and trypsin inhibitor 10 mg/mL; 0.25/50 mg tissue]. After 1 h, tissue lysates were obtained by centrifugation at 20,000 g for 15 min at 4°C. Protein concentrations were estimated by the Bio-Rad protein assay (Bio-Rad Laboratories, Milan, Italy) using bovine serum albumin as standard. Small intestine lysate proteins (70 μg) were dissolved in Laemmli sample buffer, boiled for 5 min, and separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis, and then transferred to a nitrocellulose membrane (240 mA for 40 min at room temperature). The filter was then blocked with 1× phosphate-buffered saline (PBS) and 3% non-fat dried milk for 40 min at room temperature and probed with anti-inducible nitric oxide synthase (iNOS) antibody (dilution 1:1,000; cat. no. 610204, BD Bioscience, Becton Dickinson, Buccinasco, Italy), or anti-cyclooxygenase (COX)-2 (dilution 1:1,000; cat. no. 610431, BD Bioscience,), or anti-NF-κB p65 (dilution 1:500; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, United States), or anti-Iκ-Bα (dilution 1:500; Santa Cruz Biotechnology, Inc.), or anti-occludin (dilution 1:500, cat. no. sc-492, Santa Cruz Biotechnology, at 4°C overnight). The secondary antibody was incubated for 1 h at room temperature. Subsequently, the blot was developed using enhanced chemiluminescence detection reagents (Amersham Pharmacia Biotech, Piscataway, NJ, United States) according to the manufacturer’s instructions. The detection of filter was performed by ChemiDoc Imaging System (Bio-Rad Laboratories). Blots were also incubated in the presence of the antibody against the β actin protein (Sigma-Aldrich, cat.no. G9545, Sigma-Aldrich) to ascertain that they were loaded with equal amounts of protein lysates (De Caro et al., 2020).

Plasma was obtained from blood collected prior to perfusion (as described above) by centrifugation at 1,000 × g for 15 min under 4°C and stored in aliquots at −80°C. The levels of pro-inflammatory cytokines, namely IL-1β, IL-6, TNF-α, and anti-inflammatory cytokine IL-10 as well as bacterial-derived pro-inflammatory lipopolysachharide (LPS), in plasma were measured by ELISA kits (Thermo Fisher Scientific, Monza, Italy) according to the manufacturers’ instructions as previously described (Russo et al., 2014; De Caro et al., 2020; Leo et al., 2020).

Freshly evacuated fecal pellets were kept directly in a sterile microtube and stored at −80°C until assayed. Bacterial genomic DNA was extracted from frozen fecal samples using the QIAamp DNA Stool Mini Kit according to manufacturer’s instructions (Qiagen). DNA concentration was measured fluorometrically using a Qubit dsDNA BR assay kit (Invitrogen). Samples were stored at −20°C until processed for amplification. Sequencing samples were prepared according to the protocol 16S Metagenomic Sequencing Library Preparation for Illumina Miseq System with some modifications. The V3–V4 regions of the 16S rDNA gene were amplified and PCR was performed using the following cycle conditions: an initial denaturation step at 95°C for 3 min, followed by 25 cycles of 95°C for 30 s, 55°C for 30 s, 72°C for 30 s and ended with an extension step at 72°C for 5 min. After a purification step with Agencourt AMPure XP (Beckman Coulter Inc), the amplicons were indexed with 10 subsequent cycles of PCR using the Nextera XT Index Kit (Illumina). Each PCR reaction contained 5 μL of amplicons from first PCR, 5 μL index 1 primer (N7xx), 5 μL index 2 primer (S5xx), 25 μL 2× KAPA HiFi HotStart Ready Mix and 10 μL PCR grade water. Library sizes were assessed using an Agilent High Sensitivity 2200 Tape Station System (Agilent Technologies) and quantified with Qubit. Normalized libraries were pooled, denatured with NaOH, then diluted to 10 pM and combined with 25% (v/v) denatured 10 pM PhiX, according to Illumina guidelines. Sequencing run was performed on an Illumina Miseq system using v3 reagents for 2 × 281 cycles (Citraro et al., 2021).

Paired-end demultiplexed Illumina sequencing reads were imported into the Quantitative Insights Into Microbial Ecology 2 (QIIME 2; 2021.2 distribution)¹ software suite for downstream analysis (Caporaso et al., 2010). Sequences were then quality filtered, dereplicated, chimeras identified, and paired-end reads merged in QIIME2 using DADA2 (Callahan et al., 2016); quality filtering was performed using default settings, trimming was set at position 60 (forward and reverse), and truncation lengths were set at 240 for forward and reverse, respectively. A phylogenetic tree was generated using the align-to-tree-mafft-fasttree pipeline in the q2-phylogeny plugin. Bray–Curtis dissimilarity between samples was calculated using core-metrics-phylogenetic method from the q2-diversity plugin. Classification of Amplicon Sequence Variants (ASVs) was performed using a naïve Bayes algorithm trained using sequences representing the bacterial V3-V4 rRNA region available from the SILVA database² (Silva_132-99) (Quast et al., 2013), and the corresponding taxonomic classifications were obtained using the q2-feature-classifier plugin in QIIME2. The classifier was then employed to assign taxonomic information to representative sequences of each ASV. Statistical analyses were performed using the Microbiome analyst software.³

Fecal samples (n = 5 per group) were stored frozen at the end of the experiment. A fixed amount was weighed (0.5 g) and suspended in ultra-pure water (2.5 mL) in a 15 mL tube, vortexed briefly, and then a total of 7.5 mL 1.33% HCl/ethanol solution was added. The samples were homogenized for 2 min and centrifuged at 17,968 g for 10 min at room temperature. Supernatant was promptly transferred to a 2 mL sample vial. This step was repeated three times. A quantity of the pooled extract containing acetate, propionate and butyrate were transferred into a 2 mL glass vial and loaded onto an Agilent 7890 gas chromatograph (GC) system with automatic loader/injector (Agilent Technologies, Santa Clara, CA, United States). Pure fatty acid standards (cod. 71251 Acetic Acid, cod. 94425 Propionic Acid, cod. 19215 Butyric Acid Merck Life Science, Milano, Italy) were also prepared to prepare a calibration curve and each sample was analyzed three times on the same day (De Caro et al., 2020).

Tyrosine hydroxylase (TH) and neuroinflammation immunostainings were carried out on coronal brain sections containing either SNpc or striatum. 50 μm coronal sections were cut using a vibratome (Agar Scientific, United Kingdom). One in every 4 sections containing the SNpc and one in every 12 sections containing the striatum were collected. 4–7 slices were used per animal and per staining. All histological procedures have been previously described (Economides et al., 2018; Sancandi et al., 2018). Sections were incubated in 1% H₂O₂ for 30 min and then in 1% sodium borohydride (NaBH₄) for 30 min to reduce background staining. Sections were then incubated in 10% normal goat serum for another 60 min to prevent non-specific antibody binding. Sections were incubated overnight at 4°C in primary antibodies (Table 1) and 0.1% triton X-100 (Sigma Aldrich). The next day, sections were incubated overnight in secondary antibodies [biotinylated goat anti-mouse or anti-rabbit antibody (1:500 made up in PBS, Vector Laboratories)]. Sections were then incubated in ABC (Vector Laboratories) overnight, washed in TRIS buffer, incubated in 3, 3′ diaminobenzidine (DAB, Sigma Aldrich) for 20 min. Staining was then revealed using H₂O₂. All sections were processed using the same immunoreagents and the DAB reaction was stopped by adding TRIS buffer at the same time allowing comparison between experimental groups. Sections were then mounted onto Superfrost slides, dehydrated, cleared with Histoclear and coverslipped using DPX (Sigma Aldrich). Assessors were blinded to the treatment allocation. Specificity of primary antibodies used in this study has been described previously. A SDS-denaturated rat tyrosine hydroxylase purified from pheochromocytoma was used to produce the rabbit polyclonal anti-TH antibody. It reacts with all mammalian species. This antibody specifically recognizes the 60 kDa rat tyrosine hydroxylase (Zhou et al., 1995). Anti-Glial Fibrillary Acidic Protein Antibody, clone GA5 is an antibody against Glial Fibrillary Acidic Protein (GFAP) used to stain astrocytes and Bergman glia cells (Debus et al., 1983). According to the manufacturer’s specifications, in western blots, this antibody labeled a single band at 51 kDa (molecular mass of glial acidic protein). The antibody reacts with GFAP from human, pig, chicken and rat. C-terminus of Iba1 (ionized calcium-binding adapter molecule1) was used for the production of the polyclonal anti-Iba1 antibody that reacts with Iba1 from human, rat and mouse. This antibody stains both resting and activated microglia (Imai et al., 1996; Mori et al., 2000). No staining was observed in sections incubated either with a non-immune serum and secondary antibodies or with primary antibodies only.

The optical fractionator probe was used to determine the number of TH-immunopositive neurones in the SNpc (StereoInvestigator, MicroBrightField) using a Nikon microscope coupled to a computer-controlled x-y-z motorized stage and an MBF video camera system. Unbiased stereology was carried out on 5 slices per animal (n = 4 per experimental group) with the following parameters: counting frame 70 μm × 50 μm; grid size 175 μm× 175 μm; section thickness 50 μm, dissector height 16 μm. Only neurones with visible nuclei and dendrites were counted.

Levels of GFAP immunohistochemical staining were measured by quantitative thresholding image analysis as previously described (Rahim et al., 2012). Four non-overlapping images of the striatum in each section were captured using a DMR microscope and Leica Application Suite V4 (Leica Microsystems) at X20 magnification with constant light intensity, microscope calibration and video camera settings. Image-Pro Premier (Media Cybernetics, Cambridge, United Kingdom) was used to analyze the images and measure immunoreactivity using a constant threshold that was applied to all images for each respective antigen. Data are presented as the mean percentage area of immunoreactivity ± SEM. Morphological characteristics of Iba1-immunopositive cells (Sholl and branched analyses; n = 20 per experimental group) were analyzed using a Neurolucida software (MBF Bioscience). Data are represented as mean ± SEM.

Data were analyzed using GraphPad Prism v.9. The Shapiro–Wilk test and the Kolmogorov–Smirnov test were carried out prior to statistical analysis to determine whether the data followed a normal distribution. The parametric one-way ANOVA test was then used as all data followed a normal distribution.

To assess the tolerability of the probiotics, the weight of the animals was monitored and recorded daily. Animals treated with Symprove™ showed a weight gain comparable to the other groups (Figure 2). Interestingly, rats injected with toxins (PD models) and treated with Symprove™ did not display the characteristic drop in weight following surgery at Day 5.

Expression of occludin in the ileum was measured in the four experimental groups to evaluate intestinal barrier integrity (Figure 3 and Supplementary Figure 1A). Occludin levels measured in the PD model treated with placebo were significantly lower than those in the sham animals treated with placebo and Symprove™ (p < 0.05). Symprove™ treatment prevented this decrease (p < 0.05).

The inflammatory state in the ileum of all experimental groups was then assessed by evaluating different pro-inflammatory protein expression, including iNOS, NF-κB, COX-2 and the anti-inflammatory protein Iκ-Bα (Figure 4 and Supplementary Figures 1B,C). iNOS expression was increased in the early-stage PD model compared with that in the sham animals treated with placebo and Symprove^™ (Figures 4A,B, p < 0.05). Expression of iNOS in the PD model animals treated with Symprove^™ was similar to that in the model animals treated with placebo (p > 0.05). Expression of NF-κB in the PD model animals treated with placebo and Symprove™ was similar to that in the sham animals treated with placebo (Figures 4A,C). Surprisingly, sham animals treated with Symprove™ displayed a reduced NF-κB expression compared with the other three experimental groups (p < 0.01).

COX-2 expression was significantly decreased in the ileum of PD model animals treated with placebo compared with that in the Sham animals treated with placebo and models treated with Symprove™ (Figures 4D,E, p_{sham + placebo} < 0.05; p_{model + symprove} < 0.001). Notably, low levels of COX-2 were found after Symprove™ treatment in the sham animals compared with those in the sham animals treated with placebo (p < 0.001). IκBα expression was reduced in the model animals treated with placebo compared with that in the sham group treated with placebo and Symprove™ (Figures 4D,F, p < 0.01). Symprove™ in the model group prevented this decrease (p < 0.05).

To investigate the effects of Symprove™ supplementation on inflammatory markers in plasma, the level of lipopolysaccharide (LPS), the pro-inflammatory cytokines IL-6, TNF-α, IL-1β and anti-inflammatory cytokine IL-10 were measured in all experimental groups (Figure 5). An increased level of LPS was observed in the PD model treated with placebo compared with the sham animals (p_{sham + placebo} < 0.05; p_{sham + symprove} < 0.01) and Symprove™ supplementation significantly decreased its circulating levels (Figure 5A, p < 0.05). Plasma levels of IL-6, TNF-α and IL-1β in the PD models treated with placebo were increased compared with those in the Sham animals treated with either placebo and Symprove™ (Figures 5B–D, p < 0.05). Treatment with Symprove™ in the PD models prevented this increase (p_{TNF, IL6} < 0.05; p_IL1 < 0.001). In addition, animals in the model group treated with placebo exhibited a 25% decrease of IL-10 in plasma compared with the sham groups (Figure 5E). Symprove™ treatment in the model prevented this decrease albeit not significantly.

Fecal microbiota composition following treatment with either placebo or Symprove™ was analyzed in all experimental groups (Figure 6). The α-diversity results showed that there was a significant difference in the indexes of Shannon among the four groups (Figure 6A). PD model induction was able to significantly increase α-diversity also modifying ß-diversity (Figure 6B); moreover, Symprove™ further increased α-diversity, also significantly modified ß-diversity both in the model and sham control group (Figure 6B). Symprove™ induced a different shift within groups that likely depended on the initial GM composition. GM analysis following treatment with either placebo or Symprove™ reported that, at the phylum level, the predominant bacterial communities were Firmicutes and Bacteroidetes (Figure 7); only in the sham group, Symprove™ was able to significantly increase Bacteroidetes. However, other phylum such as Proteobacteria, Cyanobacteria and Actinobacteria were not significantly modified. On the other hand, several significant differences were evidenced at genus level among groups. PD model induction significantly modified microbiota composition (Figure 8A and Supplementary Table 1); specifically, Eubacterium_xylanophilum were decreased while 15 genera were significantly increased. Notably, among these latter genera, 5 (Acetatifactor, Alloprevotella, Lachnospiraceae_NC2004, Ruminococcus_torques and UCG_009) were reduced by Symprove™ treatment in PD animals (Figure 8B and Supplementary Table 2); furthermore, treatment was overall able to modify 22 genera (Figure 8B and Supplementary Table 2).

Short chain fatty acid concentrations were measured in stool samples from all experimental groups (Figure 9). Acetate levels were similar in all four experimental groups (Figure 9A). Propionate levels were reduced by 60% in the PD model treated with placebo compared with those in the sham animals treated with placebo (Figure 9B – p < 0.0001). PD model treated with Symprove™ displayed higher levels than those treated with placebo although this 40% increase was not significant (p > 0.05). Levels of butyrate in the PD model treated with placebo were decreased by around 60% compared to those in the sham animals treated with placebo and Symprove™ (Figure 9C – p_{sham + placebo} < 0.001; p_{sham + symprove} < 0.0001). Treatment with Symprove™ in the model group prevented the decrease in butyrate levels (p_{model placebo vs. symprove} < 0.0001).

To determine the effect of Symprove™ on neuroinflammation, levels of GFAP and Iba1 (markers of astrocytic and microglia activation respectively) were assessed in the striatum of the four experimental groups (Figure 10). Astrocytic activation, shown by a stronger GFAP staining, in the PD models treated with placebo was prevented by treatment with Symprove™ (Figures 10Aa,Ab; p < 0.01). The change in cell morphology as a result of microglial activation observed in the striatum of the PD model treated with placebo was also prevented by Symprove™ (Figures 10Ba–Bc).

Immunohistochemical staining of TH cells in the SNpc (Figure 11A) and unbiased stereology (Figure 11B) revealed that the TH level in the early-stage model treated with placebo was significantly decreased by 45% in the SNpc compared with that observed in sham animals treated with placebo (p < 0.001). Treatment with Symprove™ prevented the loss of TH-immunopositive cells in the early-stage model by 17% compared to the model treated with placebo.

The state of the gut integrity and GM was investigated in the early-stage PD rat model for the first time in this study. As observed in PD patients (Forsyth et al., 2011; Wang et al., 2021), gut permeability was disrupted in the model and a model-dependent modification of the GM composition was observed in terms of several genera although it is important to note that gut dysbiosis in rodent models of 6-OHDA-induced PD remains poorly characterized and often controversial with opposite results being found (e.g., increased/decreased abundance of Lactobacillus; Choi et al., 2019; Tsao et al., 2021). As it is often difficult to establish the exact mechanisms underlying the impact of microbiota composition alone on MGBA and the CNS, a recent study assessed common changes in the GM composition observed in studies on PD patients based on the current concept that an alteration of the GM composition and the subsequent imbalance in the MGBA results in increased gut permeability, immune function dysregulation, impaired lipid metabolism and SCFA production contributing to PD pathogenesis (Shen et al., 2021). Although several limitations may be found, this study indicated that PD patients in comparison to healthy controls have a significantly lower abundance levels of Prevotellaceae, Faecalibacterium, and Lachnospiraceae and an increased abundance of the families of Bifidobacteriaceae, Ruminococcaceae, Verrucomicrobiaceae, and Christensenellaceae (Shen et al., 2021). Although comparisons between rats’ and PD patients’ GM compositions cannot be easily translated, our results showed that Ruminococcaceae and Bifidobacteriaceae were also increased in the model as in PD patients.

Intestinal inflammation, circulating inflammatory markers and SCFA levels were also investigated in the early-stage model in order to determine whether an alteration was present in the model. GM modification was found not to be accompanied, further than by the above reported increased permeability, by a markedly increased inflammatory response in the ileum, at least at this disease stage. In contrast, we found that pro-inflammatory cytokines (i.e., IL-1β, IL-6 and TNF-α) were significantly increased in plasma and the anti-inflammatory cytokine IL-10 was reduced along with an increased concentration of circulating LPS and reduced concentrations of the gut microbial SCFA metabolites butyrate and propionate, considered (at least experimentally) as neuroprotective in other PD models (Ostendorf et al., 2020; Hou Y. et al., 2021; Hou Y. F. et al., 2021). Altogether, these results indicate that gut integrity is damaged in the model and GM is modified to a condition in which SCFA production is disrupted while LPS may either or both be increased in production and/or more easily penetrate into the blood. Similar results have been found in both patients and animal models (Rutsch et al., 2020; Aho et al., 2021; Wang et al., 2021). Association between gut dysbiosis and increased plasma levels of LPS has also been shown in different diseases such as Type 2 diabetes (T2DM), chronic kidney disease (CKD; Salguero et al., 2019), Alzheimer’s Disease (AD; Zhao et al., 2017), inflammatory bowel diseases (IBD; Candelli et al., 2021) and autism (Parracho et al., 2005; Emanuele et al., 2010). However, to our knowledge, elevated blood LPS derived from intestinal dysbiosis in PD subjects has not, so far, been demonstrated. Interestingly, plasma levels of the alternative inflammatory biomarker LPS binding protein (LBP) were found to be consistently lower in PD patients than in healthy patients (Forsyth et al., 2011; Pal et al., 2015; Chen et al., 2021). This inverse correlation between LPS and LBP levels and neuroinflammation in PD patients may have some diagnostic relevance for PD and is worthy of further study (see also Zweigner et al., 2001).

Treatment with the probiotic suspension prevented gut barrier impairment and modified GM composition in the early-stage PD model with 5 genera that were specifically altered by the model being normalized after treatment (e.g., Acetatifactor, Alloprevotella, Lachnospiraceae_NC2004, Ruminococcus_torques, and UCG_009). Besides this effect, probiotic treatment had a positive impact on circulating cytokines, preventing their alteration by model induction and increasing SCFA levels; this is in agreement with previous observations indicating that probiotic, prebiotic, and synbiotic supplementation may suppress the levels of pro-inflammatory cytokines and decrease the level of NF-ĸB (Romo-Araiza et al., 2018; Tsao et al., 2021). Similarly, SCFA levels were also found to be increased by microbiota manipulation being accompanied by anti-inflammatory and antioxidant effects (Tsao et al., 2021). In line with these effects, the treatment also prevented the LPS increase in plasma. Overall, treatment with the probiotic suspension prevented gut involvement in the model by maintaining a healthier GM, being able to avoid local and systemic inflammation, protecting gut integrity and improving SCFA production.

As neuroinflammation is a major player in PD etiology, pathophysiology, and progression of the disease (Cicchetti et al., 2002; Doty, 2012; O'Neill and Harkin, 2018; Arena et al., 2022; Teleanu et al., 2022), the effect of Symprove™ on striatal neuroinflammation in the early-stage PD model was investigated in this study. Supplementation with the probiotics prevented activation of both astrocytes and microglia in the striatum that was observed in the early-stage PD model treated with placebo. This is in agreement with previous studies in which probiotics have been shown to have anti-inflammatory properties (Klaenhammer et al., 2012; Srivastav et al., 2019; Tamtaji et al., 2019; Castelli et al., 2021). Modulation of neuroinflammation was associated with a prevention of the loss of TH-positive neurones in the SNpc of ~17% which was in agreement with recent studies that showed a neuroprotective effect of probiotics in animal models of more advanced stages of PD (Srivastav et al., 2019;Hsieh et al., 2020; Castelli et al., 2021). Castelli and colleagues (2020) demonstrated that treatment with probiotics prevented 6-OHDA-induced neurodegeneration both in vitro and in vivo, mainly due to anti-inflammatory effects, restoration of pro-survival and neuroprotective pathways and partial prevention of dopaminergic loss. Although the precise mechanisms by which probiotics exert neuroprotective effects are yet to be uncovered, it has been suggested that they may be due to their abilities to both enhance peripheral tyrosine decarboxylase production (van Kessel et al., 2019) and modulate neuroinflammation (Hsieh et al., 2020; Castelli et al., 2021). The early-stage PD model used in this study was shown to display neuroinflammation soon after toxin injection and a reduction in dopaminergic cells in the SNpc (Sancandi et al., 2018) and therefore represents a useful tool to study the etiology of the disease including the cross-talk between the gut and the brain, and to determine new therapeutic targets. Further studies will now be needed to determine the effects of probiotics on PD symptoms (non-motor and motor) and α-synuclein aggregation (in genetic PD models) and to identify their role in modulating the gut-brain axis. Although none of the current animal PD models recapitulates all the symptoms experienced by PD patients exactly, they allow the investigation of the gut-brain axis, how these two organs can influence each other, and which factors are involved in the cascade of events that lead to neurodegeneration.