Metabolomics – the large-scale study of small molecules involved in metabolic processes – is an area of vast and ever growing complexity. But it is a research approach, which along with other emerging -omics type technologies, could help us to uncover and define more about the way the human body works in health and illness.
Speaking to NutraIngredients ahead of his talk at the forthcoming Probiota 2017 conference in Berlin next month, Dr Christian Hellmuth from the division of nutritional and metabolic medicine, at Dr von Hauner Childrens’ Hospital in Munich, told us that the prospect of integrating metabolomic technologies with research investigating how the gut microbiota influences human health holds a lot of promise.
The expert in metabolomic testing and analysis will present an introduction to the topic at Probiota – where he will outline how studying metabolites on such a large-scale works, and the challenges and opportunities such methods present to research on gut health and nutrition.
Better biomarkers
Hellmuth noted that while using certain metabolites in the blood as a biomarker for health or dietary intake is not a new idea, the measurement and analysis of large arrays of metabolites is relatively new.
“You can think of fatty acids or amino acids which are quite famous metabolites that have been analysed for many years, but the new take on metabolomics (…) was to cover a very wide range of metabolites in biological systems.”
Indeed, he suggested that larger-scale metabolomic information could be a reliable way to measure the intakes of different food groups: “There has already been a lot of work with levels of fatty acids in the blood that are markers for the dietary intake in the long term and the short term,” said Hellmuth. “If we can uncover these biomarkers for dietary intake then we can get an idea of what a person has been eating in the last week, month, or even year.”
In this way, the expert suggests metabolomics can be used to increase the accuracy of studies by avoiding often criticised food frequency questionnaires.
“Food frequency questionaries’ and other dietary surveys used in a study are often influenced by what a person is thinking or the mood and memory of the person that gave the information. With reliable metabolomic biomarkers of dietary intake we can just look at the blood metabolites that show intake levels.”
Uncovering complex relationships
Hellmuth and his group focus on how different exposures at various stages of life – for example the first 1000 days of life – influence later risk of obesity, insulin resistance and diabetes.
“We try to look how these exposures affect the metabolism of pregnant women, the foetus, a new-born baby, and children,” he said. “We look at how these changes in metabolism are affecting the risk of obesity so that we can get an idea of the pathway.”
“It’s always interesting to look at the biochemical pathways from the genome and environmental influences down to the metabolome,” Hellmuth told us.
“To me it’s always interesting to see where to put the microbiome because it is obviously influenced by metabolism and environmental factors like our diet … but is it then affecting our genomes?” he said. “That is really interesting.”
By using different -omics technologies to look at biomarkers, each of which tell a different part of the story, it is possible to look at these complex issues and identify relationships and pathways to both good and bad health, the expert suggested.
But while many people have shown a strong interest in these ‘big’ projects, the high cost and complexity of research, data management, and statistical analysis have so far been a barrier, he said.
Standardisation needed?
Hellmuth added that one of the main problems with current metabolomic techniques currently is that while no method can measure all metabolites, each technique used by a different research group can identify different molecules.
“In the end, each group is then referring to this as a metabolome – and if you do not understand the field and are just reading a paper, you may wonder why each group finds and uses different biomarkers,” he said. “It’s basically because there is no standardised equipment’s and different methods which all cover different metabolites and tissues.”
Further complicating the matter is the fact that teams often then use different statistical and data analysis methods to look at findings – often meaning that biomarker findings made by one group can vary to those made by another.
“I think that if you get a standard design, then metabolomics has a big advantage because it is at the end of everything,” Hellmuth said – suggesting that standardisation across the metabolomics field, or indeed in a subgroup of metabolomics such as those interested in gut health or the influence of the microbiota, could be possible ‘in theory’.
However, he warned that because most labs have very expensive – but different – equipment, standardisation across the board is ‘nearly impossible.’
“The idea for me is first to get the audience who are interested and want to do these studies, and fund these studies. To speak and ask what they want to do – not just in metabolomics but in all the possible measures,” he said. “Then you can start to build an understanding of which techniques can be used.”
Furthermore, if different parties interested in the same research area can agree to treat metabolome data in the same way throughout its statistical analysis then there may be more opportunity for integration of large data sets at a later time, he said.
“You maybe cannot influence what metabolome data you get because of the equipment used, but you can certainly agree as a community to use the same data analysis of all your samples,” he noted – adding that a recommendation paper could be put together on such points.
“I think it is helpful to have a recommendation paper where you give examples of what can be done with data sets, because then after this every group can refer to them and say that they handled their data according to those recommendations. That way every reader then knows that different studies use the same methods and can perhaps have data combined.”