University of California San Francisco

Author Of 1 Presentation

Microbiome Oral Presentation

PS10.03 - Functional survey of the pediatric multiple sclerosis microbiome        

Speakers
Presentation Number
PS10.03
Presentation Topic
Microbiome
Lecture Time
09:45 - 09:57

Abstract

Background

Metagenomic sequencing reveals the functional potential of the gut microbiome, and may explain how the gut microbiome influences pediatric-onset multiple sclerosis (MS) risk.

Objectives

To examine the gut microbiome functional potential by metagenomic analysis of stool samples from pediatric MS cases and controls using a case-control design.

Methods

Persons ≤21 years old enrolled in the Canadian Pediatric Demyelinating Disease Network who provided a stool sample and were not exposed to antibiotics or corticosteroids 30 days prior were included for study. All MS cases met McDonald criteria, had symptom onset <18 years of age and had either no prior disease-modifying drug (DMD) exposure or were exposed to beta-interferon or glatiramer acetate only. Twenty MS cases were matched to 20 non-affected controls by sex, age (± 3 years), stool consistency (Bristol Stool Scale, BSS) and, when possible, by race. Shotgun metagenomic reads were generated using the Illumina NextSeq platform and assembled using MEGAHIT. Metabolic pathway analysis was used to compare the gut microbiome between cases and controls, as well as cases by DMD status (DMD naïve vs DMD exposed MS cases vs controls). Gene ontology classifications were used to assess α-diversity and differential abundance analyses (based on the negative binomial distribution) reported as age-adjusted log-fold change (LFC) in relative abundance, 95% confidence intervals (CI), and false discovery rate adjusted p-values.

Results

The MS cases were aged 13.6 mean years at symptom onset. On average, MS cases and controls were 16.1 and 15.4 years old at the time of stool collection and 80% of each group were girls. MS cases and controls were similar for body mass index (median: 22.8 and 21.0, respectively), stool consistency (BSS types 1-2: n=4, types 3-5: n=16, for both MS and controls) and race (Caucasian: 11 and 9, respectively). Eight MS cases were DMD naïve. Richness of gene ontology classifications did not differ by disease status or DMD status (all p>0.4). However, differential analysis of metabolic pathways indicated that the relative abundance of tryptophan degradation (via the kynurenine pathway; LFC 13; 95%CI: 8–19; p<0.0005) and cresol degradation (LFC 19; 95%CI: 13–25; p<0.0001) pathways were enriched for MS cases vs controls. Differences by DMD status were also observed, e.g., choline biosynthesis was enriched in DMD exposed vs naïve MS cases (LFC 21; 95%CI: 12–29; p<0.0001).

Conclusions

We observed differences in the functional potential of the gut microbiome of young individuals with MS relative to controls at various metabolic pathways, including enrichment of pathways related to tryptophan and metabolism of industrial chemicals. DMD exposure affected findings, with enrichment of pathways involved in promoting CNS remyelination (e.g., choline).

Collapse

Author Of 1 Presentation

Biomarkers and Bioinformatics Poster Presentation

P0171 - The gut mycobiome in pediatric multiple sclerosis: establishing a bioinformatics pipeline (ID 876)

Speakers
Presentation Number
P0171
Presentation Topic
Biomarkers and Bioinformatics

Abstract

Background

Studies examining the role of the microbiota in multiple sclerosis (MS) often focus on the gut bacteria; few have considered a potential role of gut mycobiota. Methods for evaluating gut mycobiota are lacking and require systematic evaluation of sequencing protocols, reference databases, and bioinformatics pipelines to properly investigate possible gut mycobiome influences on MS.

Objectives

We set out to evaluate the performance of different sequencing conditions and analytical approaches for characterizing the gut mycobiome in a cohort of healthy individuals and cases with monophasic acquired demyelinating syndrome (mono ADS) or pediatric-onset MS.

Methods

We first assessed a mock-community control pool of known, staggered quantities of 19 defined fungal organisms. We then assessed 201 stool samples obtained from our cohort of 52 healthy individuals, 49 individuals with mono ADS, and 46 participants with pediatric-onset MS. The fungal internal transcribed spacer (ITS) 2 region was sequenced using the Illumina® MiSeq platform. Varying concentrations of PhiX Control v3 Library spike-in were tested to address low-complexity amplicon sequencing. Generated sequences were characterized by the UNITE database—a curated collection of eukaryotic ITS sequences—in conjunction with three distinct fungal sequence analysis pipelines: LotuS, mothur, and PIPITS.

Results

Taxa identified in our mock-community differed across sequencing conditions but were similar between technical replicates. LotuS correctly classified 7 taxa at species-level, 7 taxa at genus-level, whereas 5 remained unclassified. Mothur correctly identified 5 species-level taxa, 11 genus-level taxa, whereas 3 remained unclassified. Lastly, PIPITS correctly identified only 3 species-level taxa, 12 genus-level, while 4 remained unclassified. We successfully generated sequence data for 112 of 147 (76%) individuals (70 females; 42 males). The mean age at stool sample collection was 17.3 (SD 5.1) years. Of the tested sequencing conditions, a spike-in of 50% PhiX produced the highest-quality reads.

Conclusions

The LotuS pipeline best identified fungal taxa in our mock-community, with optimal resolution to species level. Sequencing read quality was optimal when 50% PhiX was used for sequencing ITS2 amplicon libraries of stool samples. Establishment of this validated sequencing pipeline, confirmed using a mock-community with known fungal identities, will aid characterization of gut mycobiomes for our cohort of individuals with/without pediatric-onset MS.

Collapse