The presented thesis illustrates two mass spectrometry based metabolomics studies to evaluate the nutritional impact of prebiotics (chapter II) and probiotics (chapter III) on the human gut microbiome and its related effects on host and gut metabolism. Non-targeted metabolite analyses were performed in negative and positive ionization mode and results were used to guide a subsequent series of targeted metabolite analyses. Additional 16S rRNA sequencing and shotgun proteomics complemented the analysis in order to investigate the complex interplay between organisms, metabolites and functional processes in the adult and infant gut microbiome.
Dietary starch affected the adult fecal metabolome differently, while the time point of starch consumption is negligible. Characteristic differences in the fecal metabolome between differently digestible carbohydrates (resistant starch type 2) were observed in a controlled, randomized, within-subjects’
crossover dietary intervention trial of participants suffering from insulin resistance. Diet with high resistant starch affected the fecal metabolome immensely. Thereto, the ultra-high resolution FT-ICR-MS based metabolomics approach was highly suitable for investigating the function of the gut microbiome and host metabolism and helped to evaluate the effect of different diets. Additionally, it made a comprehensive overview of metabolites from different chemical classes and pathways available. Specific fatty acids, bile acids, oxylipins and several compounds of the lipid metabolism were strongly affected by digestion of different prebiotics, compared to the non-dietary starch diet. These also include finding novel links between a RS diet and lipid metabolism by the host and microbiome.
Especially lipids, namely phosphatidic acids showed differently altered patterns by the consumption of carbohydrates varying in the amount of dietary starch. However, cyclic phosphatidic acids showed decreased intensity levels through dietary starch intake.
The high amount of resistant starch altered the fecal microbial composition arising with a high abundance of Firmicutes, especially Faecalibacterium prausnitzii and Eubacterium rectale, whereas the abundance of Bacteroides was low. The applied shotgun proteomics approach allowed classifying thousands of host and microbial proteins impacted by resistant starch and confirmed several results already observed through the metabolomics analyses, such as the impact of RS on the lipid metabolism.
Understanding the complexity arising in the human gut, including its microbial composition, host and/or microbial metabolism, protein expression, as well as the metabolic status is of top priority to understand the impact of diet, especially of varying amounts of non-digestible carbohydrates. Therefore, the applied multi-omics integration enabled to gain a deeper insight into main effects of the resistant starch diet on the gut microbiome and functions that they carry out. These results emphasize the importance of further multi-omics study designs, to clarify the effects of nutrition on the microbiome and health.
Long-term effects of probiotic baby food on gut physiology are widely unknown, wherefore research is still needed. Additionally, long-term effects of probiotics on the gut microbiota need to be investigated due to the decrease of willingness for weaning. The focus should be laid on the effect of probiotics on the healthy, developing gut microbiota of newborns. Thus, the second part of the thesis discussed the impact of diet on the infant gut microbiome. This study wanted to investigate the impact of probiotics on the gut microbiota and fecal metabolome within the first year of life. By means of metabolomics, detailed information of the fecal metabolome and differences within feeding types was achieved.
Metabolite profiles were clearly distinct between breast- and formula-fed infants, and they converged overtime. The metabolite profiles were altered differently through breastfeeding and formula, independent of type of formula. However, in the probiotics group several metabolites occurred that showed similar patterns as the breastfed infants. Several distinct metabolic effects were seen that made the use of probiotic formula feeding more similar to breast milk. Probiotic supplementation may help to approximate breast milk. However, these similar effects cannot yet be explained and their occurrence remains still unknown and these results were suggestively initial signs for the effects of probiotics.
Accordingly, the overall metabolite pattern was still more similar between the two formulas than between probiotics supplementation and breast milk. Therefore, further research is needed in order to optimize probiotics in terms of selection of strains, route of administration and dose.
Pathways analysis revealed the lipid metabolism, especially the bile acid biosynthesis to be differently impacted by breast- and formula-feeding. While breastfeeding seemed to influence sulfated bile acids, formula revealed high levels of glycine conjugated bile acids. Further metabolites altered by breastfeeding were identified as carboxylic acids and numerous fatty acids, varying in chain length and saturation degree. On the contrary, fecal sample of the formula-fed infants were dominated by several metabolites of the tocopherol biosynthesis, carboxylic acids and oxylipins.
Generally, formula-feeding lead to significantly higher proportions of propionic, butyric, valeric and isovaleric acid, whereas the fecal samples of breastfed infants were dominated by pyruvic and lactic acid. Correlations studies revealed lactic acid and pyruvic acid to be highly associated with species, such as Bifidobacterium sp., Streptococcus sp. and/or Lactobacillus sp. Additional correlation studies of the whole metabolomics and microbiome set revealed several relationships between metabolites and OTUs over time involved in the feeding-specific groups, which were shaping the fecal ecosystem and therefore contribute to the separation of the feeding groups. Of particular interest was the relation between F+ specific metabolites and two species, namely Bifidobacterium bifidum and Lactococcus sp., in contrast to the correlation of F- specific metabolites and Bacteroides spp. Breastfed specific metabolites seem not to be set into relation with characteristic species. Nevertheless, metabolites were the driving force behind the class discrimination of breastfed and formula-fed infants over time. In conclusion, the study demonstrated that the infant microbiome and metabolome can be altered by bifidobacteria-supplemented formula in early life. These results support the assumption that probiotics partially alter the infant microbiome and metabolome towards a similar impact owing to breastfeeding.
Accordingly, consideration should also be which metabolic effects are of interest through probiotic supplementation to approximate breast milk, e.g. the bile acid metabolism to improve fat digestion, customize SCFA profile or to improve protein digestion to enable the settlement of phenyl lactate- or p-cresol sulfate-producer. Depending on this, specific strains should be selected for supplementation.
It was demonstrated that fecal samples were highly suitable to evaluate the impact of diet on the human gut microbiome. Many mass signals were observed to be altered with diet as well, whose identity remains still unknown. Dealing with the unknown and identifying metabolites was difficult and demanding and is still a major drawback in metabolomics, which poses many challenges of experimental and analytical nature. This harbors several difficulties in data evaluation and interpretation. Nevertheless, fecal non-targeted metabolomics came out as a powerful discipline to discover the impact of diet on the human microbiome. A series of targeted analyses revealed in both studies, that especially the metabolite classes of fatty acids, carboxylic acids and steroids, such as bile acids were affected by diet, as well as metabolites of the lipid metabolism. In both studies a comprehensive overview of metabolites from different chemical classes and pathways were detected to be highly impacted by pre- and probiotics.