The microbial composition in the human gut has been associated with a range of chronic diseases. Genetic variation within bacteria species impacts their biological properties and how they interact with their environment, in this case their human host. “In the department of genetics, we are interested in how genetic variants contribute to human health. But don’t forget that more than 90% of the genes in our body is actually not from human cells, but from microorganisms in our gut. We should not overlook the impact of the bacteria genome on human health,” says Jingyuan Fu, professor of Systems Medicine at the UMCG.
Microbial profiling can help us find associations between our gut microbiome and our health. While conventional microbial research focuses on abundance of certain species, this research explores the impact of the smallest genetic differences between microbes: single nucleotide variants (SNVs). These variants are the result of mutations or evolution and can significantly impact how bacteria work and which proteins they produce. “If you only do abundance-based studies, you will miss a lot of links,” explains Daoming Wang, Veni fellow at the UMCG. “We try to understand which human health aspects can be linked to these small changes in the bacteria’s genetic instruction book.”
This study is, to the knowledge of the authors, the first large-scale systematic study into the associations between gut microbial SNVs and the human host’s phenotypes (phenome). Fu: “We barrowed the concept from human genetics-based phenome-wide association studies (PheWAS) to microbiome-based PheWAS.”
In contrast to human PheWAS, microbiome-based PheWAS deals with hundreds of different species, which all have a very different complexity and diversity of the genome and population structure. “Finally, we managed to develop a workflow to deal with this problem and make it feasible,” says Wang.
In this study 12.8 million gut microbe SNVs and their associations with 244 human phenotypes and environmental exposures were explored. The researchers identified 940 independent associations involving 101 microbial species and 69 host factors. Importantly, many of these associations were not detectable at the species-abundance level, suggesting that studying the microbiome at base-pair resolution can reveal biological information that conventional microbiome profiling misses.
The associations can be explained in multiple ways. Most SNVs occur in the coding regions of the bacterial genome. The proteins produced by gut microbes can act as catalysts, where a SNV may influence the efficiency of the enzyme. Additionally, there may be direct interactions between human proteins and microbial proteins, where a mutation at the interface level can influence that interaction.
Do these associations imply causality? “What we believe is that both can exist. Maybe microbes can adapt in response to our physiological changes, caused by factors such as diet, disease or age. On the other hand, we believe that some of the genetic variants, because they affect the bacteria functionality, can contribute to human health,” explains Wang. Fu further explains why finding functional consequences is hard: “The challenge is much, much bigger in the microbiome than in human PheWAS. There are so many different species and so many different proteins.”
One example of this is a SNV in Faecalibacterium prausnitzii, associated with age. This particular SNV was found in a protein belonging to a glutathione transport system. Glutathione is a component that can prevent cell damage caused by oxidative stress. The found association suggests that age-associated mutations may affect glutathione uptake efficiency. This can help the microbe better resist ageing-related oxidative stress in the intestinal environment.
Another example comes from preliminary results regarding a SNV associated with a lower BMI. “The SNV is found in a gene that encodes a protein that mimics a human hormone. The function of that hormone is to regulate human appetite and influence eating behaviour,” says Wang. Fu adds: “There is a lot of potential there, but we cannot conclude anything yet.”
The researchers have made all study summary statistics available in an online catalogue. “Most of the participants from our study are from the Netherlands. We hope that in the future other researchers working in the microbiome field can also conduct a similar study in other geographically and genetically different populations and try to compare their results with our results,” explains Wang. He and Fu are interested to learn more about differences and similarities in the microbiome of populations on different geographical locations. Additionally, the researchers hope that others will validate the associations they found.
“In the future we would also like to use the longitudinal data of the Lifelines cohort to look at how microbiomes change over time, how health changes over time and dive deeper into the causality as well,” says Fu. Once the insight in causality is more precise, microbiome-based therapy approaches such as probiotics can be developed to improve human gut health.
The genomic data came from 11,831 stool samples of six independent cohorts, including Lifelines. This was combined with health data form the same cohorts. In this study, the phenome studied consisted of 200 different factors related to health, from age, sex and BMI to the expression of certain proteins in the blood and metabolic parameters. In addition, environmental exposures of the participants were considered.
Read the full study in Nature.
