A prebiotic dietary pilot intervention restores faecal metabolites and may be neuroprotective in Parkinson's Disease

Abstract

Current treatment of Parkinson's Disease (PD) remains symptomatic, and disease-modifying approaches are urgently required. A promising approach is to modify intestinal microbiota and key metabolites of bacterial fermentation: short-chain fatty acids (SCFA), which are decreased in PD. A prospective, controlled pilot study (DRKS00034528) was conducted on 11 couples (PD patient plus healthy spouse as control (CO)). Participants followed a 4-week diet rich in dietary fibre, including intake of the prebiotic Lactulose. Gut metagenomes, faecal and urinary metabolites, and clinical characteristics were assessed. The dietary intervention significantly augmented faecal SCFA and increased Bifidobacteria spp., reducing PD-related gastrointestinal symptoms. The pre-existing bacterial dysbiosis in PD (depletion of Blautia, Dorea, Erysipelatoclostridium) persisted. Bacterial metabolite composition in faeces and urine positively changed with the intervention: Brain-relevant gut metabolic functions involved in neuroprotective and antioxidant pathways, including S-adenosyl methionine, glutathione, and inositol, improved in PD. These promising results warrant further investigation in larger cohorts.

Fig. 1. SCFA and clinical scales.

A Study design; B clinical measures, disease severity measured with MDS-Unified Parkinson's Disease Rating Scale Part III (UPDRSIII), gastrointestinal symptom scale (GSRS) with improvement in PD after prebiotics, and stool frequency score before and after prebiotics. Asterisk indicates the level of significance (single asterisk = p < 0.05, double asterisk = p < 0.01); C targeted SCFA measurements in faecal samples show increasing SCFA concentrations in both groups after prebiotics (Wilcoxon signed-rank test). “Wilcox all” represents testing of the entire sample set, and r gives the respective effect size for post-hoc tests; an asterisk indicates a significant difference in post-hoc tests; and D partial correlation of SCFA concentrations with clinical measure shows a sign. inverse correlation of acetate, propionate, and butyrate with disease severity measured with UPDRSIII, while stool frequency was inversely correlated to propionate, butyrate, and valerate. GSRS shows a sign. negative correlation with butyrate and valerate, red = positive and blue = negative correlation, numbers indicate the correlation coefficient r.

Fig. 2. Household-effect and taxonomic differences between PD patients and healthy CO.

A The distance-based redundancy analysis (dbRDA) shows a strong clustering of metagenomes by household, i.e. from PD with their respective healthy spouse/CO, further termed “household-effect”; this effect explained 46% of data variability on MGS species level; B dbRDA conditioned for households reveals a sign. difference in taxa dissimilarity between PD and CO after prebiotics, but not before prebiotics; C top 15 correlated taxa (MGS species) between the constrained dbRDA (conditioned for households) and study group (PD/CO) or diet intervention (before/after prebiotics), indicating a strong association of different Bifidobacteria spp. with the intervention. D, E Several genera were markedly reduced in the PD group after prebiotics, paralleled by several trends, e.g. enrichment in Eubacterium F and Methanobrevibacter_A in PD; some of these Genera already tended to differ before prebiotics (p < 0.05, but q > 0.1 before prebiotics; p < 0.05, q < 0.1 after prebiotics; data in D is presented with p-values generated by univariate tests between PD and CO on a log scaled axis, Wilcoxon signed-rank test; data in E is presented as relative abundance of the different genera). Asterisk indicates p < 0.05, q < 0.1, and r gives the respective effect size for Wilcoxon signed-rank tests.

Fig. 3. Gut bacterial communities affected by the prebiotic diet intervention.

A several taxa were significantly different between before and after prebiotics tests in both groups; n = 6 Bifidobacteria spp. were enriched after prebiotics in PD patients, of which n = 4 were also enriched in healthy CO individuals; n = 2 taxa are only changed in PD (S. Thermophilus, UMGS1975, data is presented as p-values generated with univariate tests/Wilcoxon signed-rank test between before and after prebiotics on log scaled axes); B Enterosignatures (ES) ES-Bifidobacterium (ES-Bifi) in both groups while ES-Firmicutes (ES-Firmi) decreased (sign. only in PD group, Wilcoxon rank sum test; C top 15 Kyoto Encyclopaedia of Genes and Genomes (KEGG) modules in relation to the study group (PD/CO and diet intervention (before/after prebiotics), constrained dbRDA, conditioned for household),shows a strong correlation of fatty acid (FA) metabolism with the intervention (permutation test, R² = 0.3, p < 0.001) as well as several functions correlated to either PD or CO (permutation test, R² = 0.16, p < 0.001, SPL saccharide, polyol and lipid transport system); and D Differences in KEGG Modules before vs after prebiotics (relative abundance, Wilcoxon signed-rank test), increased microbial genes related to fatty acid metabolism after prebiotics were similarly observed in PD and CO subjects (upper panel), on the other hand several genes were only changed in PD (lower panel). Single asterisk = p < 0.05, double asterisk = p < 0.01 (AA amino acid), r gives the respective effect size for post-hoc tests.

Fig. 4. Functional metabolic potential and metabolite levels.

A Brain-relevant functional metabolic potential (gut-brain modules (GBM)) improves after prebiotics in PD in a potentially neuroprotective manner, e.g. p-cresol synthesis decreased and quinolinic acid degradation increased, both leading to neurotoxic compounds. The x- and y-axis correspond to p-values of the Wilcoxon signed-rank test for changes due to the prebiotic intervention of healthy CO (x-axis) and PD (y-axis) patients. B The relative abundance of bacterial functional GBM modules improved only in PD patients after the prebiotic intervention compared to baseline before prebiotics (upper panel, p < 0.05, q < 0.1, Wilcoxon signed-rank test). This was also reflected in corresponding urinary metabolites (bottom panel, p > 0.05, Wilcoxon signed-rank test, relative concentrations, a.u., arbitrary units). C Urine metabolite composition differed between PD and CO before prebiotics, but normalized after prebiotics (permutation test, before prebiotics: R² = 0.08, p = 0.045; after prebiotics: R² = 0.05, p = 0.36). Single asterisk p < 0.05, double asterisk = p< 0.01; D faecal metabolite profiles before prebiotics were different in SCFA between PD and CO. After prebiotics these differences are no longer observed, instead amino acid concentrations were different between PD and CO (all q > 0.1, Wilcoxon rank sum test); But butyrate, Isobut isobutyrate, Prop propionate. The numerical suffix (e.g. _2) associated with each metabolite represents a distinct nuclear magnetic resonance (NMR) peak for that particular metabolite.