Altered Gut Microbiota Related to Inflammatory Responses in Patients With Huntington’s Disease

Emerging evidence indicates that gut dysbiosis may play a regulatory role in the onset and progression of Huntington’s disease (HD). However, any alterations in the fecal microbiome of HD patients and its relation to the host cytokine response remain unknown. The present study investigated alterations and host cytokine responses in patients with HD. We enrolled 33 HD patients and 33 sex- and age- matched healthy controls. Fecal microbiota communities were determined through 16S ribosomal DNA gene sequencing, from which we analyzed fecal microbial richness, evenness, structure, and differential abundance of individual taxa between HD patients and healthy controls. HD patients were evaluated for their clinical characteristics, and the relationships of fecal microbiota with these clinical characteristics were analyzed. Plasma concentrations of interferon gamma (IFN-γ), interleukin 1 beta (IL-1β), IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13, and tumor necrosis factor alpha were measured by Meso Scale Discovery (MSD) assays, and relationships between microbiota and cytokine levels were analyzed in the HD group. HD patients showed increased α-diversity (richness), β-diversity (structure), and altered relative abundances of several taxa compared to those in healthy controls. HD-associated clinical characteristics correlated with the abundances of components of fecal microbiota at the genus level. Genus Intestinimonas was correlated with total functional capacity scores and IL-4 levels. Our present study also revealed that genus Bilophila were negatively correlated with proinflammatory IL-6 levels. Taken together, our present study represents the first to demonstrate alterations in fecal microbiota and inflammatory cytokine responses in HD patients. Further elucidation of interactions between microbial and host immune responses may help to better understand the pathogenesis of HD.


INTRODUCTION
Genetic components are the dominant factors in the pathogeneses of monogenic neurodegenerative diseases. At present, increased attention has been focused on the role of the gut-brain axis and its related humoral response in the development of neurodegenerative diseases. However, the role of the gutbrain axis in monogenic neurodegenerative disease remains unclear.
Huntington's disease (HD) is a monogenic, fully penetrant, progressive neurodegenerative disorder characterized by motor, cognitive, and psychiatric disturbances. HD is caused by the expansion of CAG trinucleotide repeats in exon 1 of the huntingtin (HTT) gene on chromosome 4, and HTT is widely expressed in the brain and in peripheral tissues such as skeletal muscles and the gut (1)(2)(3)(4). The mutant huntingtin (mHTT) protein, which is expressed in the gastrointestinal (GI) tract, causes GI dysfunction, including impaired gut motility, diarrhea, and malabsorption of food (5). Malabsorption correlates with the amount of weight loss that is a hallmark of HD, both in patients with HD (6)(7)(8)(9) and in several transgenic mouse models of HD (10). A recent study has suggested that gut dysbiosis may play a regulatory role in the onset age and progression of HD symptoms (11). In addition, mHTT is expressed in peripheral myeloid cells, including monocytes and macrophages (12,13). Furthermore, monocytes from the blood of HD patients produce increased levels of cytokines ex vivo when stimulated (13,14).
Gut microbiota may play a crucial role in the bidirectional gut-brain axis that affects brain activity under both physiological and pathological conditions (15). A growing number of studies suggests that gut microbiota exhibit physiological functions associated with neurodevelopment, brain function, and behavior (16)(17)(18)(19). Maladaptive changes in the composition of gut microbiota, referred to as gut dysbiosis, have been linked to a number of gastrointestinal and metabolic diseases, including inflammatory bowel disease (IBD), obesity, and diabetes (20)(21)(22). In addition, gut dysbiosis has been implicated in various neurological and psychiatric diseases, such as autism spectrum disorder (ASD), major depression, amyotrophic lateral sclerosis (ALS), Parkinson's disease (PD), and Alzheimer's disease (AD) (23)(24)(25)(26)(27)(28). Furthermore, R6/1 transgenic HD mice exhibit weight loss and gut-microbiota dysbiosis, the latter of which has been confirmed via 16S ribosomal RNA (rRNA) gene sequencing (11). Notably, alterations in circulating metabolites related to gut microbiota in HD patients and transgenic animals have suggested that changes in gut microbiota may occur before the onset of HD (29,30).
However, there are currently no studies that have reported the composition of gut microbiota and related host cytokine responses in HD patients. Therefore, in the present study, we analyzed and compared the microbiota communities and peripheral cytokine levels of HD patients with those of healthy controls. Additionally, we analyzed the relationships between fecal microbiota and clinical characteristics in HD patients.

Study Subjects
All HD patients and controls in this study were recruited through a longitudinal study for aging and neurodegeneration project, which was conducted by the neurogenetic group at the China National Clinical Research Center for Neurological Diseases from September 2018 following international standard Enroll HD protocol (31), recommended by China HD Network (CHDN). All participating subjects signed informed consents prior to enrollment. This study was approved by the Research Ethics Committee at Beijing Tiantan Hospital, Capital Medical University, Beijing, China.
By October 2019, 33 HD patients (24 manifest, 9 premanifest) and 33 sex-and age-matched healthy controls were recruited from 14 provinces across China. A study flow chart is presented in Figure 1. Each HD patient eligible for the present study received a diagnosis of HD according to a confirmed family history, positive genetic test, and motor disturbance as defined by the Unified HD Rating Scale (UHDRS) total motor score (TMS) diagnostic confidence score. The exclusion criteria were defined as severe chronic diseases such as diabetes, heart failure, liver cirrhosis, malignancy, hematological/autoimmune diseases, irritable bowel syndrome, and other movement disorders such as Wilson disease and chorea-acanthocytosis. The healthy controls were neurologically normal individuals and close relatives of the HD patients, such as neurologically normal spouse or parents, to ensure they shared similar environmental and dietary factors. If the patient's offspring or siblings were included, genetic testing was conducted to confirm the offspring or siblings were not mutant HTT gene carriers. Individuals taking antibiotics or probiotic supplements within one month prior to sample collection were also excluded.

Clinical Data Collection
Clinical data of subjects were collected via face-to-face interviews with HD researchers during the enrollment process. The HD researchers had already undertaken clinical-scales training under the instruction of the CHDN, which provided the Enroll HD clinical assessment package. The weight and height of each participant were measured, and body mass index (BMI) was calculated. HD clinical characteristics included disease duration and CAG repeats, as well as the motor section of the Unified Huntington's Disease Rating Scale (UHDRS-M), the Functional Assessment Scale (FAS), Total Functional Capacity (TFC), Category Fluency Test (CFT), Symbol Digit Modalities Test (SDMT), Stroop Interference Test (SIT), Mini-Mental State Examination (MMSE), and Beck Depression Inventory II (BDI-II). To elaborate the quantitative clinical measures further, UHDRS-M ranges from 0 to 124 points with higher scores indicating more severe motor symptoms (32). FAS is a more detailed measure of functional capacity evaluation test and consists of 25 yes/no questions about specific functional abilities. The FAS examines tasks related to occupation (e.g., accustomed work, volunteer work), finances (e.g., cash transactions, financial management), activities of daily living (e.g., driving, hygiene), domestic chores (e.g., home maintenance, laundry), level of care (e.g., home or supervised care), and physical abilities (e.g., walking, getting out of bed, falls). FAS scores range from 0 to 25 with higher scores indicating greater functionality (33). The TFC provides a measure of broad functional capacity and consists of five global items that assess occupation, finances, domestic chores, activities of daily living, and care level. The scores on each item range from 0 to either 2 or 3. TFC total scores range from 0 to 13 with higher scores indicating greater functioning (33). The CFT requires the subject to name as many examples of the category ''animal'' as possible within 1 min (34). The SDMT requires participants to make as many symbolnumber associations as possible within 90 s, and provides a measure of speed in information processing (35). SIT is a component of the Stroop Test that provides a measure of Executive Function (EF) including cognitive flexibility and resistance to interference. Scores reflect correct number in 45 s and higher scores indicate better performance (36). The MMSE provides a measure of general cognitive functioning (37). The BDI-II contains 21 items used to assess the intensity of the depression in clinically depressed or nondepressed patients. Each component is scored on a 4-point scale from 0 to 3 with higher scores indicating a more severe depressed mood (38).

Sample Collection and DNA Extraction
Fecal samples were collected in tubes (SARSTEDT, Germany) with fecal preservation solution at home by the study participants according to our instructions and transported to our laboratory. Then, all samples were frozen immediately and stored at −80°C until further analysis. The following experiments were carried out commercially by Realbio Genomics Institute (Shanghai, China). DNA was extracted from each fecal sample via an improved protocol according to the manual of the QIAamp Fast DNA Stool Mini Kit (Qiagen, Germany). The concentration of genomic DNA in each fecal sample was quantified using a NanoDrop 2000 spectrophotometer (Thermo Scientific, MA, U.S.). The integrities and sizes of fecal DNA samples were assessed using 1% agarose gel electrophoresis.
Blood samples of 6 ml of venous blood were drawn from each participant at enrollment and were centrifuged (1,000 g for 15 min) within 1 h of collection. The plasma was aliquoted into cryotubes following centrifugation and stored at −80°C until further use for cytokine analysis.

16S rDNA Gene Amplicons and Sequencing
The V3-V4 regions of bacterial 16S rDNA genes were amplified by PCR (95°C for 3 min, followed by 30 cycles at 98°C for 20 s, 58°C for 15 s, and 72°C for 20 s, as well as a final extension at 72°C for 5 min) using universal primers (341F and 806R) linked with indices and sequencing adaptors. PCR amplification was performed in a 30-mL mixture containing 15 mL of 2 × KAPA Library Amplification ReadyMix, 1 mL of each primer (10 mM), 50 ng of template DNA, and ddH 2 O. Amplicons were extracted from 2% agarose gels and purified using an AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, U.S.) according to the manufacturer's instructions. Purified amplicons were quantified using Qubit 2.0 (Invitrogen, U.S.). All quantified amplicons were sequenced using an Illumina NovaSeq PE250.

Processing of Sequencing Data
Assembled tags, trimmed barcodes, and primers were further checked in terms of their rest lengths and average base qualities. 16S tags were restricted between 220-500 bp so that the average Phred score of bases was no worse than 20 (Q20) and no more than 3 ambiguous N. The copy number of tags was enumerated and redundancy of repeated tags was removed. Only tags with a frequency greater than 1, which tend to be more reliable, were clustered into operational taxonomic units (OTUs), each of which had a representative tag. OTUs were clustered based on 97% similarities using UPARSE and chimeric sequences were identified and removed using USEARCH (version 7.0.1090). Each representative tag was assigned to a taxa by the RDP Classifier (http://rdp.cme.msu.edu/) against the RDP database (http://rdp.cme.msu.edu/) using a confidence threshold of 0.8. The a-diversity and b-diversity indices were calculated based on the rarefied OTU counts via Qiime v1.9.1. Specifically, adiversity represents an analysis of diversity in a single sample reflected by parameters including good coverage, Chao 1, PD whole tree, Shannon index, and Simpson index, using Qiime (39). The Wilcox test in R software was used to compare each adiversity index. b-diversity is used as a measure of the microbiota structure between groups. The results of both the weighted and unweighted Unifrac distance matrices were plotted in the principal coordinate analysis (PCoA), and Adonis was performed using R software. Microbial features used to distinguish fecal microbiotas specific to HD were identified using the linear discriminant analysis (LDA) effect size (LEfSe) method (http://huttenhower.sph.harvard.edu/lefse/) with an alpha cutoff of 0.05 and an effect-size cutoff of 2.0.

Statistical Analysis
The SPSS (ver. 21.0, SPSS Inc., Chicago, IL, USA) and R software (ver. 3.5.1, the R Project for Statistical Computing) were used for statistical analysis. Differences between groups were determined via Student's t-test and Pearson's Chi-squared tests for quantitative and categorical variables, respectively. Variables that violated the assumptions of normality were compared via nonparametric Mann-Whitney U tests. Correlations among components of fecal microbiota with clinical parameters and cytokines were determined via Spearman correlation analysis. Statistical significance was set at p < 0.05.

Demographic and Clinical Characteristics of HD Patients and Healthy Controls
The demographic characteristics of HD patients and healthy controls in this study are presented in Table 1. There were no significant differences between the HD and controls in terms of age or gender ( Table 1), but the age at assessment in the premanifest HD is much younger compared to the manifested HD (Supplementary Table 1). The median duration of HD patients at enrollment was 4 years, and the median number of CAG repetitions was 42. HD patients has a significantly lower BMI compared to that of healthy controls ( Table 1).

Alpha and Beta Diversity Between HD Patients and Healthy Controls
The dilution curves of a-diversity indices were plotted to demonstrate that the sample size in this study is adequate for valid analysis (Supplementary Figure 1). The Chao 1, observed species, and PD whole tree of the HD group were significantly higher than those of the healthy control group, whereas there were no significant differences between these two groups in terms of the Shannon and Simpson index (Figures 2A-E). These results indicate that the richness of the gut microbiota in the HD group was significantly higher than that of the healthy control group. Significant differences were also found in bdiversity based on the unweighted (qualitative, Adonis R 2 = 0.03, p = 0.02) but not the weighted (quantitative, Adonis R 2 = 0.022, p = 0.197). UniFrac between HD and healthy control groups ( Figures 2F, G), indicating that the fecal microbial structure, but not the abundance, in the HD group was significantly different from that of the healthy control group.

Compositions of Microbial Taxa at Multiple Phylogenetic Ranks Between HD Patients and Healthy Controls
The top-20 taxa at multiple phylogenetic ranks were analyzed.  Figure 3). The microbial species profiling histogram of the HD cohort were also demonstrated according to the disease duration ( Supplementary Figures 2-6).

Altered Microbiota Between HD Patients and Healthy Controls
To determine the significantly increased bacteria in the HD group or healthy control group, supervised comparisons via LEfSE (LDA > 2.0) were performed. This LEfSe analysis revealed many significant differences in the fecal microbiota between the HD group and healthy control group. Specifically, the following relative abundances in the HD group were significantly higher than those in the healthy control group: Actinobacteria at the phylum level; Deltaproteobacteria and Actinobacteria at the class level; Desulfovibrionales at the order level; Oxalobacteraceae, Lactobacillaceae, and Desulfovibrionaceae at the family level; and Intestinimonas, Bilophila, Lactobacillus, Oscillibacter, Gemmiger, and Dialister at the genus level. In contrast, Clostridium XVIII at genus level was significantly higher in the healthy control group compared to that in the HD group ( Figure 4).

Relationships Between Fecal Microbiota and HD Clinical Characteristics
To determine the relationship between clinical characteristics and gut microbiota in HD patients, Spearman correlation analysis was performed to evaluate correlations among clinical characteristics and gut-microbiota genera obtained by LEfSe analysis. Our analysis revealed two correlations between fecal microbiota at the genus level with specific clinical scores, namely Intestinimonas with TFC scores and Lactobacillus with MMSE scores ( Figure 5).

DISCUSSION
The present study provides the evidence for gut dysbiosis in human patients with HD, providing clinical relevance to a previous study that reported gut dysbiosis in a transgenic mouse model of HD (11). We found that Intestinimonas and Bilophila correlated with concentrations of IL-4 and IL-6, respectively, in HD patients, suggesting the occurrence of a systemic chronic inflammatory status associated with altered gut microbiota. In our present study, we used 16S rDNA gene sequencing on DNA isolated from fecal samples to systematically analyze characteristics of gut microbiota between HD patients and healthy controls. We observed that a-diversity (richness) in HD patients was significantly higher than that of healthy controls, as was b-diversity (structure), consistent with previous results from a mouse model of HD (11). However, in a recent study of the gut microbiota of HD patients (40), the adiversity of HD patients was lower than that of healthy controls. The disparity is likely due to different ethnic origins, geography, host, genetics, age and other factors (41). Previous study showed an increase in gut microbiota richness was detected in autistic children (42), and it shared great similarities in the altered microbiota species with HD. Given HD was also considered as a neurodevelopmental disorder similar to autism (43), the greater bacterial diversity is potentially beneficial to the development of central nervous system (CNS) function, which remains subject to debate.
Furthermore, we analyzed the top-20 taxa at multiple phylogenetic ranks. At the genus level, the average ratios of Bacteroides and Prevotella between groups HD and healthy control were 0.78 and 1.18. The significant bacterial differences identified by LEfSe (LDA > 2.0) showed that Clostridium XVIII at the genus level was significantly higher in the healthy control group compared to that in the HD group, whereas the relative abundances of the following were higher in the HD group compared to those in the healthy control group: Actinobacteria at the phylum level; Deltaproteobacteria and Actinobacteria at the class level; Desulfovibrionales at the order level; Oxalobacteraceae, Lactobacillaceae, and Desulfovibrionaceae at the family level; and Intestinimonas, Bilophila, Lactobacillus, Oscillibacter, Gemmiger, and Dialister at the genus level. These results indicate that particular gut microbiota components are associated with HD. The increased abundance of Actinobacteria at the phylum level in the HD group compared to that in the healthy control group in our present study is similar to previous findings reported in patients with IBD (44), type-2 diabetes mellitus (T2DM) (45), AD (28), ASD (46), and major depressive disorder (47,48). The increased abundance of Actinobacteria at the class level in the HD patients compared to that in the healthy controls in this study is consistent with previous studies in AD and ASD patients (28,46). Family level analysis in our present study revealed significant increases in the abundance of Desulfovibrionaceae in the fecal samples of HD patients compared to those from healthy controls, whereas a previous study found that the abundance of Desulfovibrionaceae was increased in the feces of IBD patients (49), which is likely a breaker of the intestinal mucosal barrier that induces IBD (50).
The abundance of the Intestinimonas genus was higher in HD patients than in healthy controls and correlated with TFC scores. A previous study has shown that Intestinimonas, which is correlated with propionic/butyric acid, plays a key role in anti-inflammation (51,52). In the present study, we observed a positive correlation between Intestinimonas and plasma concentrations of IL-4, an antiinflammatory cytokine mainly produced by T-helper-2 cells (53). Since butyrate has the ability to regulate T cell differentiation (54), it suggests that Intestinimonas may confer efficacious antiinflammatory effects on systemic inflammatory responses in HD patients. However, future studies are needed to reveal the interactions between changes in symbiotic gut microbiota and the immune reactions in HD pathogenesis.
Bilophila contains only species of B. wadsworthia, and a higher abundance of B. wadsworthia induces systemic inflammatory responses in specific pathogen-free (SPF) mice (55). In addition, B. wadsworthia plays an important role in the development of IBD-like colitis in IL-10 -/mice (56). Additionally, B. wadsworthia has been found to be positively associated with proinflammatory cytokines, such as IL-6 (57) and IL-1b (58). A previous study found that Bilophila was significantly negatively correlated (p < 0.05) with IL-6 mRNA levels in a T2DM rat model treated with stachyose (59). Interestingly, our present study also revealed that Bilophila were negatively correlated with proinflammatory IL-6 levels, suggesting that Bilophila may play an anti-inflammatory role in systemic inflammatory responses in HD patients. That said, further research is needed to determine the precise relationships between gut microbiota compositions and immune reactions in HD patients.
Lactobacillus has the ability to ferment a series of carbon sources, primarily to lactic acid, and is widely recognized as a source of probiotics (60). Several recent studies have found that Lactobacillus is more abundant in patients with T2DM (61), ASD (62,63), IBD (64), and rheumatoid arthritis (65,66), which is consistent with our present findings in HD patients. Interestingly, we found that the abundance of Lactobacillus was negatively correlated with MMSE scores, although we did not find any correlation between Lactobacillus and levels of inflammatory cytokines. These findings suggest that if the abundance of Lactobacillus adversely impacts MMSE scores, and consequently cognitive function, it likely does so via a non-cytokine-induced mechanism.
In the present study, we found that the abundance of Clostridium XVIII was significantly decreased in the HD group compared to that in healthy controls, indicating higher fecal water content in the HD group, similar to findings in HD mice (11). Strati et al. observed that Clostridium XVIII was significantly more abundant in constipated autistic subjects compared to that in non-constipated autistic subjects, and this increased abundance was associated with clinical manifestations of GI problems related to alterations in gut microbiota (62).
There are several limitations of this study. Although we managed to recruit HD and control pairs from the same family and at the same geographic location across China, the sample size is minimal and the nature of the present study is exploratory. Although this study was controlled for factors affecting the gut  microbiota, such as age, gender, diet, and region, physical activity, smoking, alcohol, drugs and other factors should be fully considered in the further study. Further studies with larger sample sizes comprised of independent cohorts originating from a different population are required to consolidate our findings of different microbial species between HD patients and healthy controls, as well as the microbial species associated with clinical symptoms. In addition, more advanced technology such as shotgun metagenome analysis, and more advanced association analysis platform such as HALLA (67) can be used in the future to provide more detailed information regarding any implicated microbiota in future studies to more comprehensively understand fecal microbiota composition in HD patients.
In summary, the present study first elucidates that gut microbiota are altered in HD patients and are correlated with specific clinical characteristics. Furthermore, we demonstrated that fecal microbiota was related to specific cytokine levels. Revealing the precise interactions between microbial and host immune responses may help to better understand the pathogenesis of HD.

DATA AVAILABILITY STATEMENT
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and A B FIGURE 6 | Cytokine profiles in HD patients and healthy controls, as well as correlations between cytokines and gut-microbiota genera obtained by LEfSe analysis. (A) Plasma concentrations of interferon-gamma (IFN-g), interleukin-1b (IL-1b), IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13, and tumor necrosis factor alpha (TNF-a). (B) Correlation heatmaps showing the relationships among cytokines and gut-microbiota genera. Each row in the heatmap represents a genus. Each column represents a cytokine. The color temperature encodes the Spearman correlation coefficient (r). The sample size was n = 33. Statistical analyses were performed by Mann-Whitney tests (A) and Spearman's tests (B). The error bars indicate the standard error of mean (SEM). HD denotes Huntington's disease. + p < 0.05, *p < 0.01. accession number can be found below: National Center for Biotechnology Information (NCBI) BioProject database with project number PRJNA667318.

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Ethics Committee, Beijing Tiantan Hospital, Capital Medical University, Beijing, China. The patients/participants provided their written informed consent to participate in this study.

ACKNOWLEDGMENTS
We thank the healthy individuals, the patients, and their close relatives who donated their fecal and blood samples.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2020. 603594/full#supplementary-material Supplementary Figure 1 | Dilution curves of a-diversity index. As the amount of sequencing increased, more species were found, and no new OTU was found by increasing the number of sampling strips until the species was saturated.