A Randomized Controlled Trial of Probiotics Targeting Gut Dysbiosis in Huntington’s Disease

Abstract

Background:

Gastrointestinal symptoms are clinical features of Huntington’s disease (HD), which adversely affect people’s quality of life. We recently reported the first evidence of gut dysbiosis in HD gene expansion carriers (HDGECs). Here, we report on a randomized controlled clinical trial of a 6-week probiotic intervention in HDGECs.

Objective:

The primary objective was to determine whether probiotics improved gut microbiome composition in terms of richness, evenness, structure, and diversity of functional pathways and enzymes. Exploratory objectives were to determine whether probiotic supplementation improved cognition, mood, and gastrointestinal symptoms.

Methods:

Forty-one HDGECs, including 19 early manifest and 22 premanifest HDGECs were compared with 36 matched-healthy controls (HCs). Participants were randomly assigned probiotics or placebo and provided fecal samples at baseline and 6-week follow-up, which were sequenced using 16S-V3-V4 rRNA to characterize the gut microbiome. Participants completed a battery of cognitive tests and self-report questionnaires measuring mood and gastrointestinal symptoms.

Results:

HDGECs had altered gut microbiome diversity when compared to HCs, indicating gut dysbiosis. Probiotic intervention did not ameliorate gut dysbiosis or have any effect on cognition, mood, or gastrointestinal symptoms. Gut microbiome differences between HDGECs and HCs were unchanged across time points, suggesting consistency of gut microbiome differences within groups.

Conclusion:

Despite the lack of probiotic effects in this trial, the potential utility of the gut as a therapeutic target in HD should continue to be explored given the clinical symptomology, gut dysbiosis, and positive results from probiotics and other gut interventions in similar neurodegenerative diseases.

Keywords

Huntington’s disease randomized controlled trial neurodegenerative disease gut microbiome cognitive impairment gut–brain axis

INTRODUCTION

Gastrointestinal (GI) disturbances are clinical features of Huntington’s disease (HD), and include diarrhea, nutrient deficiencies, gastritis, and unintended weight loss, significantly affecting quality of life [1]. Frequency of GI disturbances in HD is unclear, but there is evidence to suggest that these symptoms are often underreported [2]. To date, there are no effective treatments targeting GI disturbances in HD, and management of GI symptoms and unintended weight loss remain difficult for people with HD and treating physicians [1]. Although several studies have documented clinical evidence of gut dysfunction in HD [3], information about how the gut microbiome is affected in HD has only begun to emerge. Our research group recently reported the first evidence of altered gut microbiome profiles in 42 HD gene expansion carriers (HDGECs) compared to 36 healthy controls (HCs) [4], observed in fecal samples. We found less richness (fewer species) and evenness, and differences in the compositional structure of the gut microbiome in HDGECs compared to HCs. These findings are associated with vulnerability of the gut environment to disease [5]. We also observed group differences in gut functional pathways and enzymes, and preliminary associations between gut microbiota, motor signs, and cognition in HDGECs. Similar observations were seen in the R6/1 HD mouse model, showing gut dysbiosis (microbial imbalance) compared to wild-type controls, which was associated with weight loss and motor signs [3]. These findings highlight the gut microbiome as a potential site for tracking disease activity and as a treatment target, although to the best of our knowledge, treatments for the gut microbiome have not been trialed in HD to date.

In other neurodegenerative diseases, there is consistent evidence of GI dysfunction and gut microbiome differences in people with Parkinson’s disease (PD) and Alzheimer’s disease (AD) compared to HCs, showing lower richness, evenness and altered structure of the gut microbiome (for review see Cryan et al. [6]). Imbalances in gut microbiota have been linked to amyloid development in AD, α-synuclein pathology in PD, and immune response signaling and microglial activation in multiple sclerosis (MS) [6]. Further, imbalance of the gut microbiome has been associated with neurodegenerative processes such as oxidative stress and neuroinflammation, which are implicated in HD, AD, and PD [7]. In comparison, appropriate diversity within the gut microbiome has been shown to be essential for gut health and for intact functioning of pathways connecting the gut to the brain (gut-brain axis) and to other organs [5]. These findings highlight the important role that the gut microbiome may play in neurodegeneration and as a potential target for therapeutic interventions.

In terms of therapies for the gut microbiome, mounting evidence highlight the potential clinical utility of probiotics to drive changes in the gut microbiome in a variety of neurodegenerative diseases (e.g., AD, PD, MS [8, 9]). Probiotics are live microorganisms delivered through food, supplementation, or via pharmaceutical intervention benefiting the gut microbiome by restoring homeostasis, thereby conferring health benefits to the rest of the body [10]. Probiotics suppress the growth of harmful pathogens, increase the integrity of the GI epithelial barrier, and increase excitability within the enteric nervous system [11]. These processes regulate immune response signaling and reduce inflammatory processes, preventing neurotransmitter loss, and influencing neural excitatory-inhibitory balances [12].

To date, probiotic interventions in neurodegenerative diseases have yielded promising findings. In PD, which shares notable similarities with HD (e.g., movement disorder and basal ganglia pathology), a 4-week probiotic randomized controlled clinical trial (RCCT) decreased GI symptoms [13], and a 12-week probiotic RCCT decreased motor signs [14]. However, improvement in PD symptoms post-probiotic intervention may be due to alterations in metabolism of levodopa uptake by the gut microbiome after treatment [15]. In AD, a 12-week probiotic RCCT improved performance on the Mini-Mental State Examination [16], and in MS, 8- and 12-week probiotic RCCTs decreased depression symptoms and increased the abundance of protective bacteria (Lactobacillus) within the gut microbiome [8, 17]. Adding to this evidence, our findings of gut dysbiosis in HDGECs and links between gut microbiota and HD symptoms, highlight the importance of investigating interventions for the HD gutmicrobiome.

We conducted a RCCT to investigate the effects of a 6-week probiotic intervention in HDGECs. Our primary objective was to investigate whether probiotics caused positive changes within the gut microbiome (measured through fecal samples) in HDGECs compared to HCs, such as increased microbiome richness and evenness (alpha diversity), or changes in gut structure (beta diversity), pathways and enzymes. Exploratory objectives were to investigate whether probiotics significantly improved cognition, depression, and GI symptoms.

MATERIALS AND METHODS

Participants

Our sample comprised of 41 HDGECs (19 manifest HDGECs, 22 premanifest HDGECs) and 36 age- and gender-matched HCs, reflecting the cohort in our previous report [4], with the exception of one new HC participant (Fig. 1).

Fig. 1

Participant trial flow based on Consolidated Standards of Reporting Trials (CONSORT). HCs, healthy controls; HDGECs, Huntington’s disease gene expansion carriers. *Two HDGECs stool samples could not be analyzed because amplicon reads were less than 10,000. Thus, these samples were deemed insufficient for microbial diversity processing.

HDGECs had been genetically confirmed to have Cytosine Adenine Guanine (CAG) expansions indicative of HD (CAG ≥39), with the exception of one manifest HDGEC who had CAG = 38. All HDGECs had Unified Huntington’s Disease Rating Scale (UHDRS) Total Functional Capacity (TFC) scores ≥6, indicating early to moderate disease, based on Shoulson and Fahn Rating Scale [18]. For manifest HDGECs, clinical diagnosis had been previously confirmed by neurological examination, based on unequivocal motor signs of HD, with UHDRS Diagnostic Confidence Level scores of 4 [19]. We used scaled CAG-age product (CAP _S ) scores [calculated by CAP _S = Age₀×(CAG–3.66)/432.3326 [20]] to classify the probability that our premanifest HDGECs would become symptomatic within the next 5 years, using the optimization algorithm from the PREDICT-HD study [21]. Eight HDGECs were in the low likelihood range to develop motor signs within 5 years (CAP _S <0.67), seven in the medium likelihood range (CAP _S = 0.67–0.85), and six in the high likelihood range (CAP _S >0.85; Table 1).

Table 1

Demographic and clinical data of HDGEC and HC participants

		HDGECs = 41	HCs = 36	p/ χ² value
Gender	M:F	20:21	15:21	0.494
Age (y)	M (SD)	50.20 (12.29)	50.19 (13.60)	1.000
	Range	26–75	24–69
UHDRS-TMS	M (SD)	9.45 (12.78)	–	–
	Range	0–41	–
UHDRS-TFC	M (SD)	10.95 (2.58)	–	–
	Range	6–13	–
CAG Length	M (SD)	41.26 (1.80)	–	–
	Range	38–46	–
CAP _S Score	M (SD)	0.88 (0.25)	–	–
	Range	0.32–1.40	–
Years of Education	M (SD)	14.62 (2.97)	14.97 (3.09)	0.614
	Range	8–21	9–23
Estimated IQ	M (SD)	104.00 (7.07)	106.49 (6.30)	0.113
	Range	87–115	89–117

HCs, healthy controls; HDGECs, Huntington’s disease gene expansion carriers; M, male; F, female; TMS, Total Motor Score (range 0–124); TFC, Total Functional Capacity (range 0–13); CAP _S , Scaled CAG-age product score (for premanifest HDGECs); Estimated intelligence quotient (IQ) calculated by Wechsler Test of Adult Reading; p/χ² value signifies probability value between groups, using independent samples t-tests or chi-squared test (χ² for gender). All scores represent measurements at the baseline testing session; TMS scores for three manifest HDGECs were unavailable. CAP _S scores were not calculated for two premanifest HDGECs because their specific CAG number was unknown to them (although confirmed to be ≥39 through genetic testing).

Exclusion criteria for all participants included diagnosis of coeliac disease, diabetes, irritable bowel syndrome, recent antibiotic or anti-inflammatory use within the last two months, traumatic brain injury, current recreational substance use, psychiatric or neurological illness (except symptoms attributable to HD for HDGECs), or any other diagnosed condition that may affect gut health (e.g., Crohn’s disease, ulcerative colitis). We recruited participants from Monash University’s research participant database and the Statewide Progressive Neurological Disease Service at Calvary Health Care Bethlehem in Melbourne, Australia. This trial adhered to the CONSORT guidelines [22].

Trial design and procedure

The trial was randomized, double-blinded, placebo-controlled, and conducted at Monash University in Melbourne, Australia. A randomization matrix was used to assign participants in a 1:1 ratio to receive either probiotics or placebo. Each participant was provided with 120 capsules of probiotics or placebo (of which 84 capsules were expected to be consumed) for a period of 6 weeks. At both baseline and 6-week follow-up, we administered characterization measures, cognitive tests, and participants provided fecal samples (for gut microbiome sequencing).

We screened participants for eligibility by telephone, and then scheduled a 90-min baseline assessment. We mailed stool collection kits in advance to participants, containing instructions and materials for baseline and follow-up stool sample collection. Participants provided a fecal sample within 24 h of their in-lab testing sessions. Samples were stored on ice until participants arrived at their testing session and immediately stored at –80°C for future processing. Testing sessions at baseline and follow-up were identical, and we administered sample characterization measures and a battery of neuropsychological tests. At the end of the baseline session, we provided participants with their allocated dietary supplements (probiotic or placebo). All participants were instructed to swallow two capsules with water each day at approximately the same time. Participants were instructed not to alter their physical activity, diet, or start any different supplements during this time. At 6-weeks follow-up, participants returned the remainder of their allocated supplements, which we then counted to measure compliance. To support compliance, we contacted participants weekly via telephone.

Standard protocol approvals, registrations, and patient consents

Monash University Human Research Ethics Committee approved this study (MUHREC ID: 8031), and all participants provided written informed consent in accordance with the Declaration of Helsinki [23]. All results presented in this study are de-identified. This clinical trial is registered with the Australian New Zealand Clinical Trials Registry - ACTRN12618000102279. We reimbursed participants for their time after each testing session.

Dietary supplement

Probiotic and placebo capsules were produced and supplied by Metagenics® (Queensland, Australia), and had been approved by the Therapeutic Goods Administration, Australia. Each probiotic capsule contained 22.5 billion live probiotic organisms, including 10 billion colony-forming units (CFUs) of Lactobacillus rhamnosus (LGG®), 7.5 billion CFU (organisms) of Saccharomyces cerevisiae (boulardii), and 5 billion CFU (organisms) of Bifidobacterium animalis ssp lactis (BB-12®). Dosages were based on recommendations of previous research from Metagenics [24]. Microcrystalline cellulose was used for placebo capsules, and were indistinguishable from the probiotic capsules in appearance, taste, and color.

Fecal sequencing and processing

Fecal samples underwent bacterial genomic DNA extraction by the Australian Genomic Research Facility (AGRF) in South Australia. Extraction occurred at a single time-point to eliminate possible batch processing effects. We used 16 S V3-V4 rRNA gene sequencing (341F-806R) for tracing phylogeny, and final concentrations were determined by fluorometry (see Wasser et al., 2020 for details of fecal DNA extraction processes [4]).

Clinical and cognitive characterization

To characterize our sample, we used the Gastrointestinal Health Appraisal Questionnaire (GIHAQ; Metagenics®, 1992) to measure self-reported levels of GI symptoms. Given high prevalence of depression in HD and the negative effects of depression on the gut microbiome and on cognition [25], we measured depression symptoms using the Center for Epidemiologic Studies Depression Scale-Revised (CESD-R) [26]. We administered the Wechsler Test of Adult Reading (WTAR) to estimate premorbid intelligence [27], which we used to match our HDGEC and HC groups.

To assess cognition, we used the Huntington’s Disease Cognitive Assessment Battery (HD-CAB) and the Paired Associates Learning (PAL) from the Cambridge Neuropsychological Test Automated Battery (CANTAB® [Cognitive assessment software], Cambridge Cognition, 2017, Cambridge, UK). The HD-CAB is designed for assessing cognition in HD clinical trials and includes six cognitive tests sensitive to HD [28]: the Hopkins Verbal Learning Test-Revised (HVLT-R) [29] for assessment of verbal learning and memory, Symbol Digit Modalities Test (SDMT) [30] for processing speed, Trail Making Test Part B (TMT-B) [31] for executive functioning and visuospatial attention, Paced Tapping (PTAP) [32, 33] for motor control and timing, Emotion Recognition (EMO) [34] for emotional processing, and One Touch Stockings of Cambridge (OTS) [35] for working memory and spatial planning. The outcome measure was the HD-CAB composite score, which we calculated for each group (HDGEC and HC) at each time point (baseline and 6-week follow-up; see Supplementary Material).

We administered the PAL, which assesses spatial memory and is sensitive to HD [36]. We used alternate versions of the PAL and HVLT-R (from HD-CAB) between the two time points. For all cognitive tests (except HVLT-R), at both baseline and follow-up, participants were pre-exposed to the cognitive measures and performed brief practice on each task to limit practice effects (see Supplementary Material). Pre-exposure to the HVLT-R was not possible because this would have aided memory performance on the test.

Statistical analysis

At both baseline and follow-up, we first ran bioinformatics analysis to determine the makeup of the gut microbiome in HDGECs and HCs. We analyzed microbiome diversity indices (pertaining to the richness and structure of the microbiome), pathway and enzyme differences (relating to gut microbiome function), and taxonomic classification of bacteria (representing the constellation of bacteria that exists within the gut microbiome). To address our primary outcome, we used baseline to follow-up within group comparisons to determine the effect of probiotics on gut microbiome measures. Baseline to follow-up within group comparisons were also used to investigate the effect of probiotics on exploratory outcomes (cognition, depression, and GI symptoms). Below we provide details on the statistical analyses for each outcome.

Primary outcome: Effect of probiotic supplementation on the gut microbiome

We first compared the structure and composition of the gut microbiome between HDGECs and HCs at baseline and follow-up in terms of alpha and beta diversity. We used the Phyloseq’ R package [37] for alpha and beta diversity analyses of Amplicon Sequencing Variants (ASVs). For alpha diversity, reads were rarefied to 1984 reads (the global minimum number of reads found in this sequencing cohort). Fisher index and number of ASVs observed were used to determine the evenness and richness of individual microbial communities. Kruskal Wallis test was implemented to test for significance in alpha diversity measurements between groups. For beta diversity, we calculated unweighted UniFrac distance and ordinated in Principal Coordinated Analysis (PCoA) for visualization of variation in the data. Adonis (Permutation multivariate ANOVA) from the ‘vegan’ R package was performed with 999 permutations to test for significance in beta diversity [38]. The R² score ranges from 0 to 1, indicating the amount of variance explained by a particular factor. For taxonomic analysis, we used a cut-off of a minimum 10 reads per sample and only those that were found in at least half of the samples were included. To assess differential abundance in the phyla, family and ASV level, we performed analysis of composition of microbiomes (ANCOM) implemented in QIIME2 v2019.7 [39], which protects from multiple comparisons when testing for significant differences in total abundances of bacteria [40]. To identify enzymes and functional pathways in baseline and follow-up, enzyme commission (EC) numbers and Kyoto Encyclopedia of Genes and Genomes Ortholog (KO) abundances (for pathways) were obtained from raw sequences through PICRUSt2 pipeline to infer on the potential differences in microbiome function [41]. Alpha (Observed and Fisher) and beta (Unweighted UniFrac) diversity were analyzed using Phyloseqin R.

Similar to taxonomic analysis, differential abundances testing was performed using pairwise comparisons through ANCOM for the effects of probiotics on gut microbiome outcomes. Kruskal Wallis test was used for alpha diversity indices and pairwise Adonis PERMANOVA for beta diversity indices. Pairwise comparisons for the effect of probiotic supplements were performed on alpha and beta diversity indices, bacterial taxonomies (phylum, family, and genus levels), and on measures of gut function (enzyme and pathway levels). These comparisons were measured within groups (HDGECs vs. HCs), supplement (probiotic vs. placebo), and between males and females (given that sex can affect the gut microbiome). We also ran analyses across the whole sample (baseline vs. follow-up) to assess for changes between time points (irrespective of group or supplement). Significance threshold was set at 0.05 for all tests.

Exploratory outcomes: Effect of probiotic supplementation on cognitive performance, mood, and self-reported GI symptoms

We ran repeated measures ANOVAs to examine the effect of probiotics on cognition (HD-CAB composite score and PAL), mood (CESD-R), and GI symptoms (GIHAQ). For all analyses, we specified time as the within-subjects factor (baseline and follow-up scores), with group (HDGEC or HC) and supplement (probiotic or placebo) as between subjects factors. We ran similar analyses across our entire sample without separating participants by group (HDGEC or HC) to measure whether there was any overall effect on our exploratory outcomes more broadly.

RESULTS

None of our participants reported any adverse effects of the dietary supplementation during the trial or after trial completion. In both HDGEC and HC groups, compliance rates were very high (99%), with no participants missing more than three supplement doses (six capsules) across the 6-week period of the trial. Similar to our previous baseline data [4], gut microbiome measures did not differ between manifest and premanifest HDGECs, thus we limited our examination of the gut microbiome to comparisons between our entire HDGEC group (manifest and premanifest HD combined) and HCs.

We first examined the gut microbiome in HDGECs and HCs at baseline and 6-week follow-up in terms of diversity, enzymes, and functional pathways. On measures of richness and evenness of the gut microbiome, HDGECs showed lower alpha diversity (Observed and Fisher) compared to HCs at baseline (p = 0.001) and 6-week follow-up (p = 0.001), indicating fewer species present and lower evenness in HDGECs than HCs. In terms of the microbial community structure (beta diversity), at both time points we found significant group (HDGEC vs. HC) by gender effects (unweighted UniFrac distances, p = 0.001), but no main effect of group at baseline (p = 0.521) or follow-up (p = 0.246). At the enzyme level, HDGECs significantly differed from HCs in richness and evenness (alpha diversity) of enzyme pathways at baseline (p = 0.001) and follow-up (p = 0.001), but groups did not differ in composition (beta diversity) of enzyme pathways at baseline (p = 0.864) or follow-up (p = 0.589). Groups did not differ significantly in terms of the richness or composition of functional pathways at baseline or follow-up.

3.1 Primary outcome: Effect of probiotic intervention on the gut microbiome

Six-week probiotic supplementation did not affect any of our gut microbiome measures (Table 2). When looking at each group separately (HDGEC and HC), or across the entire sample (probiotic vs. placebo), we observed two specific differences after probiotic intervention in HCs. HCs demonstrated significant improvement (p = 0.049) in alpha diversity richness of functional pathways, and further, at the family level, in female HCs, Eggerthellaceae (gram-positive bacilli) abundance was significantly lower after probiotic intervention (p = 0.001). See summary of taxonomic analysis after probiotic intervention in Supplementary Table 3.

Table 2
Summary of group alpha (richness) and beta diversity differences in the gut microbiome after a 6-week probiotic intervention

Treatment group Group comparison ASV Enzymes Functional pathways

α-div β-div α-div β-div α-div β-div

Probiotic baseline vs. HCs 0.290 0.904 0.901 0.162 0.049^* 0.287

Probiotic follow-up HDGECs 0.980 0.995 0.790 0.617 0.679 0.685

HDGECs and HCs 0.666 0.710 0.650 0.324 0.063 0.535

Placebo baseline vs. HCs 0.188 0.998 0.709 0.372 0.555 0.849

Placebo follow-up HDGECs 0.124 0.991 0.982 0.708 0.208 0.625

HDGECs and HCs 0.545 0.703 0.996 0.469 0.096 0.585

Treatment group	Group comparison	ASV	Enzymes	Functional pathways
		α-div	β-div	α-div	β-div	α-div	β-div
Probiotic baseline vs.	HCs	0.290	0.904	0.901	0.162	0.049^*	0.287
Probiotic follow-up	HDGECs	0.980	0.995	0.790	0.617	0.679	0.685
	HDGECs and HCs	0.666	0.710	0.650	0.324	0.063	0.535
Placebo baseline vs.	HCs	0.188	0.998	0.709	0.372	0.555	0.849
Placebo follow-up	HDGECs	0.124	0.991	0.982	0.708	0.208	0.625
	HDGECs and HCs	0.545	0.703	0.996	0.469	0.096	0.585

HCs, healthy controls; HDGECs, Huntington’s disease gene expansion carriers; ASV, amplicon sequencing variants; α-div, alpha diversity; β-div, beta diversity. ^*denotes significant difference between groups where p≤0.05.

Table 3

Summary of cognitive performance and sample characterization measures of HDGECS and HCs

		HDGECs = 41	HCs = 36	Groups comparison p
HD-CAB-Baseline	M (SD)	–0.57 (0.76)	0.38 (0.46)	0.001
	Range	–1.84–0.98	–0.43–1.38
HD-CAB-Follow-up	M (SD)	–0.36 (0.88)	0.41 (0.38)	0.001
	Range	–2.51–0.89	–0.31–1.30
PAL-Baseline	M (SD)	35.98 (35.48)	21.08 (16.80)	0.024
	Range	4–120	0–58
PAL-Follow-up	M (SD)	29.15 (34.41)	12.14 (11.93)	0.006
	Range	0–113	1–50
GIHAQ-Baseline	M (SD)	20.73 (25.30)	16.31 (21.21)	0.412
	Range	0–119	0–116
GIHAQ-Follow-up	M (SD)	15.90 (24.22)	13.92 (23.03)	0.714
	Range	0–105	0–104
CESD-R-Baseline	M (SD)	10.80 (10.55)	7.17 (7.70)	0.092
	Range	0–43	0–31
CESD-R-Follow-up	M (SD)	8.51 (9.90)	4.92 (5.43)	0.056
	Range	0–38	0–24

HCs, healthy controls; HDGECs, Huntington’s disease gene expansion carriers; HD-CAB, Huntington’s Disease Cognitive Assessment Battery; PAL, Paired Associates Learning; GIHAQ, Gastrointestinal Health Appraisal Questionnaire; CESD-R, Center for Epidemiologic Studies Depression Scale-Revised. p-value calculated using independent samples t-tests between groups.

3.2 Exploratory outcomes: Effect of probiotic intervention on cognitive performance, mood, and self-reported GI symptoms

We first examined whether HDGEC and HC groups differed on cognitive performance at each time point (baseline and follow-up). As expected, HDGECs’ performance on the HD-CAB (composite score) was significantly worse than HCs at baseline (t(75) = 5.08, p < 0.001), and follow-up (t(75) = 5.13, p < 0.001), with large effect sizes (Cohen’s d = 1.14 and 1.15, respectively). Similarly, for the PAL, HDGECs performed significantly worse than HCs, reflected as a higher number of errors at both baseline (t(75) = 2.40, p = 0.020, Cohen’s d = 0.54) and follow-up (t(75) = 2.97, p = 0.005, Cohen’s d = 0.66), with medium effect sizes (Fig. 2, Table 3).

Fig. 2

Comparison between HDGECs and HCs on cognitive performance using the HD-CAB (composite z score) and PAL (spatial memory) at baseline and 6-week follow-up. HCs, healthy controls; HDGECs, Huntington’s disease gene expansion carriers. Lower mean errors on the PAL indicates better performance.

Next, we examined whether probiotic supplementation improved cognitive performance on cognition (Fig. 3). For the HD-CAB composite score, we did not find any significant effects of either probiotics or placebo for the total sample (F(1,73) = 0.064, p = 0.801), nor a significant interaction effect between supplements, time (baseline and follow-up), and group (F(1,73) = 1.33, p = 0.252). Similarly, for the PAL, we did not find significant effects of probiotics or placebo on memory performance for the entire sample (F(1,73) = 0.104, p = 0.375), or any significant interaction between supplement, time, and group (F(1,73) = 0.057, p = 0.811). Across the whole sample (regardless of supplement allocation), we found no significant changes in cognitive performance on the HD-CAB composite score from baseline to follow-up (F(1,73) = 0.008, p = 0.927), suggesting no evidence of practice effects from baseline to 6-weeks follow-up. For the PAL, there was a significant decrease in the number of errors (total adjusted) from baseline to follow-up (across the whole sample and irrespective of supplement allocation; F(1,73) = 10.63, p = 0.002), possibly because of practice effects.

Fig. 3

Effects of probiotic and placebo supplementations on the HD-CAB (composite z-score) and PAL (total adjusted errors) in HDGECs, HCs, and in both groups combined (total sample). HCs, healthy controls; HDGECs, Huntington’s disease gene expansion carriers. Error bars represent standard deviations; p-values calculated using two-way ANOVA for each group (whole sample, HDGEC, and HC) from baseline to follow-up.

Clinically, HDGECs and HCs did not differ on self-reported GI symptoms on the GIHAQ at baseline (t(75) = 0.825, p = 0.412) or follow-up (t(75) = 0.367, p = 0.714). In terms of mood, HDGECs demonstrated a trend for more depressive symptoms than HCs at baseline (t(75) = 1.707, p = 0.092) and follow-up (t(75) = 1.938, p = 0.056), although this was not statistically significant (Table 3). For both groups, at both time points and irrespective of supplement allocation, reported GI and depression symptoms were in the lowest range on the GIHAQ and CESD-R, indicating no clinical significance. At baseline, when comparing those who were allocated probiotics and those allocated placebo, irrespective of participant group (HDGECs or HCs), there were no significant differences in self-reported GI symptoms (t(75) = 0.850, p = 0.423) or depression symptoms (t(75) = 0.116, p = 0.940).

For the effect of probiotics on GI symptoms, baseline to follow-up comparisons revealed no significant effects of supplement (F(1,73) = 0.539, p = 0.465) or interaction between supplement, time and group on the GIHAQ (F(1,73) = 0.588, p = 0.446). For the effect of probiotics on depression symptoms, baseline to follow-up comparisons indicated no change in depressive symptoms, with no significant main effect of supplement (F(1,73) = 0.013, p = 0.910), or significant interaction between supplement, time and group (HDGEC and HC) on the CESD-R (F(1,73) = 0.665, p = 0.417; Fig. 4). For both GIHAQ and CESD-R there were significant effects of time (across the whole sample), showing declines in reporting of GI symptoms (F(1,73) = 4.939, p = 0.029), and depression symptoms (F(1,73) = 10.63, p = 0.002) from baseline to 6-weeks follow-up. Given the nonsignificant effects of supplement on self-reported GI and depression symptoms, the effects of time may be related to trial participation placebo effects.

Fig. 4

Effects of probiotic and placebo supplementations on self-reported GI and depression symptoms on the Gastrointestinal Health Appraisal Questionnaire (GIHAQ) and the Center for Epidemiologic Studies Depression Scale-Revised (CESD-R) in HDGECs, HCs, and in both groups combined (total sample). HCs, healthy controls; HDGECs, Huntington’s disease gene expansion carriers. Error bars represent standard deviations; p-values calculated using two-way ANOVA for each group (whole sample, HDGEC, and HC) from baseline to follow-up.

DISCUSSION

We investigated the effects of 6-week probiotic supplementation on the profile of the gut microbiome in premanifest and early manifest HDGECs and HCs, and explored potential effects of probiotics on GI symptoms, cognition, and mood. We found no evidence that 6 weeks of probiotic supplementation in HDGECs affected the gut microbiome or our exploratory outcomes. Nevertheless, regardless of supplementation, gut microbiome composition differed on several measures when compared with HCs, indicating gut dysbiosis, a difference that was consistent over the 6-week trial. Our findings support the emerging notion of an altered gut microbiome in HD [3, 4] and in other neurodegenerative diseases [12], and add evidence of consistent differences in composition of the HD gut microbiome compared to that of HCs across a 6-week period.

We designed our RCCT based on the emerging evidence of the utility of probiotics in improving motor, cognitive and mood symptoms in AD, PD, and MS [6], relevant to the triad of symptoms experienced in HD, and the clinical evidence of GI symptoms and gut dysbiosis in HD [4]. The utility of probiotics is based on the premise that these substances influence humoral, neural and metabolic pathways mediated via the gut and brain, which are thought to confer beneficial effects to the rest of the body [42]. Despite this theoretical basis, our results did not improve the gut microbiome in HDGEC participants, nor affect any HD-related symptoms. It is possible that probiotics may be unable to confer beneficial effects due to compromised condition of the gut in HD. Evidence from animal and human HD studies show GI system abnormalities including muscle wasting, decreased mucosal thickness and villus length, loss of neurons within the enteric nervous system, impaired absorption of sustenance from food, and altered density of cells within the epithelium [1, 43]. Given that the enteric nervous system interacts with, and regulates functioning of the GI epithelium, the effectiveness of probiotics may be impeded or probiotics may not be ‘welcomed’ into the GI system of people with HD [44]. The idea of reduced efficacy of probiotics in populations with GI dysfunction is supported by evidence from coeliac disease and in malnourishment, in which nutrients in the body are lacking due to dysfunction in the small intestine, affecting gut permeability [45]. Determining whether probiotics are able to reach the gut microbiome and confer beneficial effects for people with HD may become clearer once the precise mechanisms of action for probiotics are furtherunderstood.

In neurodegenerative diseases, probiotic studies have varied widely in terms of study duration. In PD and MS, 4- and 6-week probiotic RCCTs have successfully improved gut microbiome composition and GI symptoms [6, 17], and based on this evidence we designed our RCCT primarily to improve the gut microbiome. More recent probiotic studies in neurodegenerative populations, however, have utilized longer trial periods, typically over 3–6 months, yielding improvement in motor and cognitive signs in PD and AD, and increases in positive bacteria within the gut microbiome [6, 16]. Together with known GI abnormalities in HD, our trial may have been too brief to elicit beneficial effects, highlighting the need for more evidence to guide the appropriate duration for probiotic trials in HD.

Other methodological differences in probiotic studies, such as sequencing techniques implemented, may also account for the disparity in results between our trial and other probiotic trials in neurodegenerative diseases [46]. In our study, we used 16S rRNA sequencing, which is sensitive in discriminating between high-level gut diversity differences [47]. Other studies, have utilized shotgun genomics sequencing, which provides a more comprehensive sample of organisms in the gut microbiome compared to 16S rRNA sequencing [46]. Shotgun sequencing may be more sensitive in detecting species-specific changes in bacteria after probiotic intervention [47], potentially limiting the degree to which changes after our RCCT could be detected. In addition, the samples sequenced for this study had on average less reads per sample compared to our previous reported data (95,143 reads in this study, vs. 172,439 average reads in previous study [4], per average sample). Fewer reads per sample indicates less sequencing depth, which can limit the power to observe differences in the gut microbiome after probiotic intervention.

Our sequencing results of the gut microbiome indicated dysbiosis in HDGECs at both baseline and follow-up, demonstrating consistency of group gut microbiome differences between HDGECs and HCs. Stability of the gut microbiome has been supported in previous research, where across the lifespan, gut microbiome composition develops in childhood, becoming stable and resilient in early adulthood [48]. Most adults develop a ‘core’ set of microbiota, and disruption to the constitution of this ‘core’ microbiota in adulthood is thought to occur in the context of underlying disease processes or medical interventions (e.g., chemotherapy), which disrupts the richness, evenness, and composition of the microbiome [48]. Longitudinal characterization of the gut microbiome in HD is needed to establish signature HD gut microbiome differences. Given that our participants were adults and age-matched, shifts in gut diversity in HDGECs (indicative of gut dysbiosis) compared to HCs, signal the effect of HD on the gut microbiome. The concept that the gut microbiome is shaped in early life is supported by our findings that both premanifest and manifest HDGECs showed consistent gut microbiome differences to HCs, suggesting that the gut microbiome may be affected early in HD. Although we did not replicate all our previously reported microbiome results, we repeated high-level findings of reduced richness, evenness, and compositional structure of the gut microbiome in HDGECs compared to HCs, suggesting that HD has a prevailing effect on the gut microbiome. Based on these findings, it would be beneficial to study the gut microbiome in HDGECs who are far from disease onset (greater than 20 years prior to onset) and in those who are in late-stage HD, in order to understand how HD affects the gut microbiome across the disease span and the lifespan.

In light of our findings, future gut microbiome studies in HD may wish to address the methodological limitations presented in our study by increasing the duration of the probiotic intervention, trialing different probiotic cocktails, or by incorporating additional gut sequencing techniques such as shotgun genomic sequencing, which may be more sensitive in detecting effects of probiotics on the gut microbiome [47]. Alternative investigations may include prebiotics, which are dietary fibers that stimulate growth of specific bacteria [49]. Prebiotics may have indirect effects on proliferating growth within the gut microbiome, given their influence on promotion of Lactobacillus and Bifidobacterium species, protective against neuroinflammation [49].

The study of the gut microbiome in HD is in its infancy, and continued characterization of the HD gut is required in order to identify signature differences in the gut microbiome. This may aid in the development of gut biomarkers, symptomatic treatments, or potential disease-modifying therapies for HD, which is one of the few neurodegenerative diseases that can be predicted early, in advance of symptoms, and with precision [50]. Thus, HD is an ideal model for continued research pertaining to the gut microbiome, with the opportunity to translate findings to other neurodegenerative diseases.

Footnotes

ACKNOWLEDGMENTS

The authors thank all participants who voluntarily donated their time to participate in this research study.

CONFLICT OF INTEREST

Julie Stout is director of a research company, Zindametrix, which supports clinical trial sponsors in the use of the HD-CAB and other cognitive testing in clinical trials, and provides consulting services regarding cognitive assessment in Huntington’s disease to clinical trial sponsors via Stout Neuropsych Pty Ltd. The remaining authors have no competing interests to report.

DATA AVAILABILITY

As part of this clinical trial, raw de-identified data from analyses may be made available upon reasonable request. The study protocol will also be made available. Data and statistical analyses plans will be available to researchers who provide a methodologically sound proposal for need to access this information beginning 9 months and ending 36 months following article publication. Proposals to access the data should be directed to Cory.Wasser@monash.edu. To gain access to the data and other information, requestors will need to sign a data access agreement.

The supplementary material is available in the electronic version of this article: .

References

van der Burg

, Bjorkqvist

, Brundin

Beyond the brain: Widespread pathology in Huntington’s disease. Lancet Neurol2009; 8 (8): 765–74.

Andrich

, Wobben

, Klotz

, Goetze

, Saft

Upper gastrointestinal findings in Huntington’s disease: Patients suffer but do not complain, J Neural Transm (Vienna)2009; 116 (12): 1607–11.

Kong

, Cao

, Judd

, Li

, Renoir

, Hannan

Microbiome profiling reveals gut dysbiosis in a transgenic mouse model of Huntington’s disease, Neurobiol Dis2020; 135, 104268.

Wasser

, Mercieca

, Kong

, Hannan

, McKeown

, Glikmann-Johnston

, et al. Gut dysbiosis in Huntington’s disease: Associations among gut microbiota, cognitive performance and clinical outcomes, Brain Commun2020; 2 (2): fcaa110.

Dinan

, Stilling

, Stanton

, Cryan

Collective unconscious: How gut microbes shape human behavior, J Psychiatr Res2015; 63, 1–9.

Cryan

, O’Riordan

, Sandhu

, Peterson

, Dinan

The gut microbiome in neurological disorders, Lancet Neurol2020; 19 (2): 179–94.

Dumitrescu

, Popescu-Olaru

, Cozma

, Tulba

, Hinescu

, Ceafalan

, et al. Oxidative stress and the microbiota-gut-brain axis, Oxid Med Cell Longev2018; 2406594;.

Cantarel

, Waubant

, Chehoud

, Kuczynski

, DeSantis

, Warrington

, et al. Gut microbiota in multiple sclerosis: Possible influence of immunomodulators, J Investig Med2015; 63 (5): 729–34.

Tankou

, Regev

, Healy

, Tjon

, Laghi

, Cox

, et al. A probiotic modulates the microbiome and immunity in multiple sclerosis, Ann Neurol2018; 83 (6): 1147–61.

10.

Rijkers

, de Vos

, Brummer

, Morelli

, Corthier

, Marteau

Health benefits and health claims of probiotics: Bridging science and marketing, Br J Nutr2011; 106 (9): 1291–6.

11.

Eslamparast

, Eghtesad

, Hekmatdoost

, Poustchi

Probiotics and nonalcoholic fatty liver disease, Middle East J Dig Dis2013; 5 (3): 129–36.

12.

Westfall

, Lomis

, Kahouli

, Dia

, Singh

, Prakash

Microbiome, probiotics and neurodegenerative diseases: Deciphering the gut brain axis, Cell Mol Life Sci2017; 74 (20): 3769–87.

13.

Barichella

, Pacchetti

, Bolliri

, Cassani

, Iorio

, Pusani

, et al. Probiotics and prebiotic fiber for constipation associated with Parkinson disease: An RCT, Neurology2016; 87 (12): 1274–80.

14.

Tamtaji

, Taghizadeh

, Daneshvar Kakhaki

, Kouchaki

, Bahmani

, Borzabadi

, et al. Clinical and metabolic response to probiotic administration in people with Parkinson’s disease: A randomized, double-blind, placebo-controlled trial, Clin Nutr2019; 38 (3): 1031–5.

15.

Gazerani

Probiotics for Parkinson’s disease, Int J Mol Sci2019; 20 (17): 4121.

16.

Akbari

, Asemi

, Daneshvar Kakhaki

, Bahmani

, Kouchaki

, Tamtaji

, et al. Effect of probiotic supplementation on cognitive function and metabolic status in Alzheimer’s disease: A randomized, double-blind and controlled trial, Front Aging Neurosci2016; 8, 256.

17.

Tamtaji

, Kouchaki

, Salami

, Aghadavod

, Akbari

, Tajabadi-Ebrahimi

, et al. The effects of probiotic supplementation on gene expression related to inflammation, insulin, and lipids in patients with multiple sclerosis: A randomized, double-blind, placebo-controlled trial, J Am Coll Nutr2017; 36 (8): 660–5.

18.

Shoulson

, Fahn

Huntington disease: Clinical care and evaluation, Neurology1979; 29 (1): 1–3.

19.

Kieburtz

, Penney

, Corno

, Ranen

, Shoulson

, Feigin

, et al. Unified Huntington’s disease rating scale: Reliability and consistency, Neurology2001; 11 (2): 136–42.

20.

Zhang

, Long

, Mills

, Warner

, Lu

, Paulsen

, et al. Indexing disease progression at study entry with individuals at-risk for Huntington disease, Am J Med Genet B Neuropsychiatr Genet2011; 156B (7): 751–63.

21.

Epping

, Mills

, Beglinger

, Fiedorowicz

, Craufurd

, Smith

,et al. Characterization of depression in prodromal Huntington disease in the neurobiological predictors of HD (PREDICT-HD) study, J Psychiatr Res2013; 47 (10): 1423–31.

22.

Schulz

, Altman

, Moher

, Consort GrouCONSORT 2010 Statement: Updated guidelines for reporting parallel group randomised trialsBMC Med2010; 8 (11): 18.

23.

World Medical AssociationWorld Medical Associa-tion Declaration of Helsinki: Ethical principles for medical research involving human subjects, JAMA2013; 310 (20): 2191–4.

24.

Sassone-Corsi

, Raffatellu

No vacancy: How beneficial microbes cooperate with immunity to provide colonization resistance to pathogens, J Immunol2015; 194 (9): 4081–7.

25.

Foster

, McVey Neufeld

Gut-brain axis: How the microbiome influences anxiety and depression, Trends Neurosci2013; 36 (5): 305–12.

26.

Eaton

, Smith

, Ybarra

, Muntaner

, Tien

Center for Epidemiologic Studies Depression Scale: Review and revision (CESD and CESD-R) 2004.

27.

Holdnack

Wechsler test of adult reading: WTAR, San Antonio, TX: The Psychological Corporation 2001.

28.

Stout

, Queller

, Baker

, Cowlishaw

, Sampaio

, Fitzer-Attas

, et al. HD-CAB: A cognitive assessment battery for clinical trials in Huntington’s disease 1,2,3, Mov Disord2014; 29 (10): 1281–8.

29.

Benedict

RHB

, Schretlen

, Groninger

, Brandt

Hopkins Verbal Learning Test –Revised: Normative data and analysis of inter-form and test-retest reliability, Clin Neuropsychol2010; 12 (1): 43–55.

30.

Smith

Symbol digit modalities test

Los Angeles, CA

Western Psychological Services

1982.

31.

Reitan

Validity of the Trail Making Test as an indicator of organic brain damage, Percept Mot Skills1958; 8 (3): 271–6.

32.

Rowe

, Paulsen

, Langbehn

, Duff

, Beglinger

, Wang

,et al. Self-paced timing detects and tracks change in prodromal Huntington disease, Neuropsychology2010; 24 (4): 435–42.

33.

Stout

, Paulsen

, Queller

, Solomon

, Whitlock

, Campbell

, et al. Neurocognitive signs in prodromal Huntington disease, Neuropsychology2011; 25 (1): 1–14.

34.

Johnson

, Stout

, Solomon

, Langbehn

, Aylward

, Cruce

,et al. Beyond disgust: Impaired recognition of negative emotions prior to diagnosis in Huntington’s disease, Brain2007; 130 (Pt 7): 1732–44.

35.

Watkins

, Rogers

, Lawrence

, Sahakian

, Rosser

, Robbins

Impaired planning but intact decision making in early Huntington’s disease: Implications for specific fronto-striatal pathology, Neuropsychologia2000; 38 (8): 1112–25.

36.

Glikmann-Johnston

, Carmichael

, Mercieca

E-C

, Stout

‘Real-life’hippocampal-dependent spatial memory impairments in Huntington’s disease, Cortex2019; 119, 46–60.

37.

McMurdie

, Holmes

Phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data, PLoS One2013; 8 (4): e61217.

38.

Dixon

VEGAN, a package of R functions for community ecology, J Veg Sci2003; 14 (6): 927–30.

39.

Caporaso

, Kuczynski

, Stombaugh

, Bittinger

, Bushman

, Costello

,et al. QIIME allows analysis of high-throughput community sequencing data, Nat Methods2010; 7 (5): 335–6.

40.

Mandal

, Van Treuren

, White

, Eggesbo

, Knight

, Peddada

Analysis of composition of microbiomes: A novel method for studying microbial composition, Microb Ecol Health Dis2015; 26 (1): 27663.

41.

Douglas

, Maffei

, Zaneveld

, Yurgel

, Brown

, Taylor

,et al. 2020; PICRUSt An improved and customizable approach for metagenome inference, BioRxiv2, 672295.

42.

Wang

, Lee

, Braun

, Enck

Effect of Probiotics on central nervous system functions in animals and humans: A systematic review, J Neurogastroenterol Motil2016; 22 (4): 589–605.

43.

van der Burg

, Winqvist

, Aziz

, Maat-Schieman

, Roos

, Bates

,et al. Gastrointestinal dysfunction contributes to weight loss in Huntington’s disease mice, Neurobiol Dis2011; 44 (1): 1–8.

44.

Walsh

, Zemper

The enteric nervous system for epithelial researchers: Basic anatomy, techniques, and interactions with the epithelium, Cell Mol Gastroenterol Hepatol2019; 8 (3): 369–78.

45.

Keller

, Layer

The pathophysiology of malabsorption, Viszeralmedizin2014; 30 (3): 150–4.

46.

Morovic

, Hibberd

, Zabel

, Barrangou

, Stahl

Genotyping by PCR and high-throughput sequencing of commercial probiotic products reveals composition biases, Front Microbiol2016; 7, 1747.

47.

Franzosa

, Hsu

, Sirota-Madi

, Shafquat

, Abu-Ali

, Morgan

,et al. Sequencing and beyond: Integrating molecular ‘omics’ for microbial community profiling, Nat Rev Microbiol2015; 13 (6): 360–72.

48.

O’Toole

, Claesson

Gut microbiota: Changes throughout the lifespan from infancy to elderly, Int Dairy J2010; 20 (4): 281–91.

49.

Wang

, Xiao

, Tian

, Zhao

, Zhang

, Zhai

,et al. Rational use of prebiotics for gut microbiota alterations: Specific bacterial phylotypes and related mechanisms, J Funct Foods2020; 66, 103838.

50.

Goldman

Genetic testing and counseling in the diagnosis and management of young-onset dementias, Psychiatr Clin North Am2015; 38 (2): 295–308.