ABSTRACTAlthough metabolomics data acquisition and analysis technologies have become increasingly sophisticated over the past 5–10 years, deciphering a metabolite’s function from a description of its structure and its abundance in a given experimental setting is still a major scientific and intellectual challenge. To point out ways to address this “data to knowledge” challenge, we developed a functional metabolomics strategy that combines state-of-the-art data analysis tools and applied it to a human scalp metabolomics data set: skin swabs from healthy volunteers with normal or oily scalp (Sebumeter score 60–120, n = 33; Sebumeter score > 120, n = 41) were analyzed by liquid chromatography-tandem mass spectrometry (LC-MS/MS), yielding four metabolomics data sets for reversed phase chromatography (C18) or hydrophilic interaction chromatography (HILIC) separation in electrospray ionization (ESI) + or − ionization mode. Following our data analysis strategy, we were able to obtain increasingly comprehensive structural and functional annotations, by applying the Global Natural Product Social Networking (M. Wang, J. J. Carver, V. V. Phelan, L. M. Sanchez, et al., Nat Biotechnol 34:828–837, 2016, https://doi.org/10.1038/nbt.3597), SIRIUS (K. Dührkop, M. Fleischauer, M. Ludwig, A. A. Aksenov, et al., Nat Methods 16:299–302, 2019, https://doi.org/10.1038/s41592-019-0344-8), and MicrobeMASST (S. ZuffaS, R. Schmid, A. Bauermeister, P. W, P. Gomes, et al., bioRxiv:rs.3.rs-3189768, 2023, https://doi.org/10.21203/rs.3.rs-3189768/v1) tools. We finally combined the metabolomics data with a corresponding metagenomic sequencing data set using MMvec (J. T. Morton, A. A. Aksenov, L. F. Nothias, J. R. Foulds, et. al., Nat Methods 16:1306–1314, 2019, https://doi.org/10.1038/s41592-019-0616-3), gaining insights into the metabolic niche of one of the most prominent microbes on the human skin, Staphylococcus epidermidis.IMPORTANCESystems biology research on host-associated microbiota focuses on two fundamental questions: which microbes are present and how do they interact with each other, their host, and the broader host environment? Metagenomics provides us with a direct answer to the first part of the question: it unveils the microbial inhabitants, e.g., on our skin, and can provide insight into their functional potential. Yet, it falls short in revealing their active role. Metabolomics shows us the chemical composition of the environment in which microbes thrive and the transformation products they produce. In particular, untargeted metabolomics has the potential to observe a diverse set of metabolites and is thus an ideal complement to metagenomics. However, this potential often remains underexplored due to the low annotation rates in MS-based metabolomics and the necessity for multiple experimental chromatographic and mass spectrometric conditions. Beyond detection, prospecting metabolites’ functional role in the host/microbiome metabolome requires identifying the biological processes and entities involved in their production and biotransformations. In the present study of the human scalp, we developed a strategy to achieve comprehensive structural and functional annotation of the metabolites in the human scalp environment, thus diving one step deeper into the interpretation of “omics” data. Leveraging a collection of openly accessible software tools and integrating microbiome data as a source of functional metabolite annotations, we finally identified the specific metabolic niche of Staphylococcus epidermidis, one of the key players of the human skin microbiome.

The change in the skin microbiome as individuals age is only partially known. To provide a better understanding of the impact of aging, whole-genome sequencing analysis was performed on facial skin swabs of 100 healthy female Caucasian volunteers grouped by age and wrinkle grade. Volunteers’ metadata were collected through questionnaires and non-invasive biophysical measurements. A simple model and a biological statistical model were used to show the difference in skin microbiota composition between the two age groups. Taxonomic and non-metric multidimensional scaling analysis showed that the skin microbiome was more diverse in the older group (≥55 yo). There was also a significant decrease in Actinobacteria, namely in Cutibacterium acnes, and an increase in Corynebacterium kroppenstedtii. Some Streptococcus and Staphylococcus species belonging to the Firmicutes phylum and species belonging to the Proteobacteria phylum increased. In the 18–35 yo younger group, the microbiome was characterized by a significantly higher proportion of Cutibacterium acnes and Lactobacillus, most strikingly, Lactobacillus crispatus. The functional analysis using GO terms revealed that the young group has a higher significant expression of genes involved in biological and metabolic processes and in innate skin microbiome protection. The better comprehension of age-related impacts observed will later support the investigation of skin microbiome implications in antiaging protection.

Cutibacterium acnes is an important member of the human skin microbiome and plays a critical role in skin health and disease. C. acnes encompasses different phylotypes that have been found to be associated with different skin phenotypes, suggesting a genetic basis for their impact on skin health. Here, we present a comprehensive comparative analysis of 255 C. acnes genomes to provide insights into the species genetic diversity and identify unique features that define various phylotypes. Results revealed a relatively small and open pan genome (6,240 genes) with a large core genome (1,194 genes), and three distinct phylogenetic clades, with multiple robust sub-clades. Furthermore, we identified several unique gene families driving differences between distinct C. acnes clades. Carbohydrate transporters, stress response mechanisms and potential virulence factors, potentially involved in competitive growth and host colonization, were detected in type I strains, which are presumably responsible for acne. Diverse type I-E CRISPR-Cas systems and prophage sequences were detected in select clades, providing insights into strain divergence and adaptive differentiation. Collectively, these results enable to elucidate the fundamental differences among C. acnes phylotypes, characterize genetic elements that potentially contribute to type I-associated dominance and disease, and other key factors that drive the differentiation among clades and sub-clades. These results enable the use of comparative genomics analyses as a robust method to differentiate among the C. acnes genotypes present in the skin microbiome, opening new avenues for the development of biotherapeutics to manipulate the skin microbiota.