Multi-omics approach uncovers gut microbiome-host interactions in myalgic encephalomyelitis/chronic fatigue syndrome patients

By Neha MathurReviewed by Aimee MolineuxFeb 10 2023

In a recent study published in Cell Host & Microbe, researchers performed a multi-omics investigation of gut microbiome-host interactions and plasma metabolomics among myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) patients.

For high-resolution characterization of these two parameters, they also integrated clinical and lifestyle data of all the patients for association analysis. They profiled short-term ME/CFS patients diagnosed within the last four years and long-term patients suffering from ME/CFS for over 10 years vs. healthy controls. This cohort design helped the researchers identify their distinguishing microbial and metabolic features (biomarkers) contributing to ME/CFS progression.

Background

ME/CFS is a complex, debilitating syndrome for which research studies have gathered limited understanding, be it the physiological changes related to the syndrome or the governing biological mechanisms. Due to a shortage of data, scientists still need to succeed in developing ME/CFS-specific therapies and robust biomarker-based diagnostic tools.

About the study

In the present study, researchers pursued baseline molecular mechanisms by which the ME/CFS microbiome changes might reflect in metabolic markers circulating in the blood, which, in turn, potentiate more alterations in host physiology. They enrolled 149 ME/CFS patients, of which 74, 75, and 79 were in the short-, long-term, and healthy control groups, respectively. They followed up for one year with 107 ME/CFS patients and 59 healthy controls.

These participants fulfilled three inclusion criteria:

1. International Chronic Fatigue Syndrome Study Group research criteria,
2. the IOM clinical diagnostic criteria, and
3. the Canadian Consensus Criteria.

They also collected demographic and health-related information from all the participants. They provided information about their age, gender, diet, education, medical history, etc. Likewise, they furnished information regarding their general health, sleep quality, and gastrointestinal (GI) health via filling out questionnaires. While the participants self-collected their stool samples, the team collected their blood samples. They performed shotgun sequencing of the fecal microbiome of ME/CFS patients and healthy controls.

The team first applied Over Representation Analysis (ORA) in two patient cohorts vs. controls. Next, they conducted a Wilcoxon rank-sum test on all metabolites and counted the significantly differential metabolites in every sub-pathway. Likewise, they used the Fischer test to identify the distributions of the over-represented genes in the pathway. Stratified by disease duration, they also computed the association between a group of metabolites in every pathway and related sub-pathways.

A Bayesian classification model helped identify disease markers, both metagenomic and phenotypic, at the disease onset for controls vs. patients. It accounted for microbial species abundance, normalized Kyoto Encyclopedia of Genes and Genomes (KEGG) gene abundance, and metabolite profile. The all-inclusive multi-omics model with ten features collected from each classification model.

Results

The short-term ME/CFS patients had more significant microbial dysbiosis, including reduced microbial diversity, altered Bacteroides: Firmicutes microbiota ratio, and augmented heterogeneity of low-abundance microbes and GI abnormalities. In addition, these patients had decreased populations of immunomodulatory microbes, e.g., butyrate and tryptophan synthesizing F. prausnitzii. Together with the reduced prevalence of the butanoate synthesis pathway, it resulted in a loss of butyrate in the intestinal environment, which possibly led to long-term metabolic dysbiosis. The abnormalities in the short-term cohort potentially affected host immune and metabolic processes.

Some of the interpretations for early-stage dysbiosis in short-term ME/CFS patients are as follows. First, short-term patients experienced more GI disturbances, which reflected their environmental changes. Second, it highlighted that these patients might adopt a range of interventions impacting their gut microbiome. The gut microbiome is dynamic and influenced by age and diet, both of which could dramatically shift gut microbiome composition, alter their metabolic potential, and the production of immunomodulatory metabolites.

Conversely, long-term ME/CSF patients showed a stable, individualized gut microbiome. Despite their 'control-like' gut microbiome, they had irreversible clinical symptoms and progressive metabolic dysbiosis. So, the researchers hypothesized that these patients might have more widespread GI phenotypes, that later mirrored in plasma metabolite levels. It is also possible that their disease progression did not normalize. It was due to survival bias as the study cohort design was cross-sectional; thus, the long-term cohort might have comprised individuals who had not recovered and showed ME/CSF symptoms consistently.

Nevertheless, the return of the gut microbiome to a configuration similar to healthy controls suggested a return to homeostasis. Another possibility is that the markers identified in this study were not very specific to ME/CFS but predicted future disease risk. It also seems likely because there is no standardized diagnostic tool for ME/CFS as it shares phenotypes with other diseases, e.g., fibromyalgia.

Conclusions

The study used high-resolution multi-omics data to develop a highly accurate ME/CFS classifier and a mechanistic hypothesis of host-microbiome interactions in this disease. The study identified microbial and metabolic biomarkers that could aid further investigation and therapeutic strategies in the context of ME/CFS and other diseases.

Longitudinally sampling short-term patients as they proceed to long-term disease states could also help disentangle the path of microbial dysbiosis and its subsequent effects on the blood metabolome. Overall, multi-omics workflows could help develop a better understanding of host-microbiome interactions.