A mosaic of cross-phylum chemical interactions occurs between all metazoans and their microbiomes. A number of molecular families known to be produced by the microbiome have a profound impact on the balance between health and disease 1-9. Considering the diversity of the human microbiome, numbering over 40,000 operational taxonomic units 10 , the impact of the microbiome on the chemistry of an entire animal remains underexplored. In this study, mass spectrometry informatics and data visualization approaches 11-13 were used to provide an assessment of the impacts of the microbiome on the chemistry of an entire mammal by comparing metabolomics data from germ-free (GF) and specific pathogen Reprints and/or permissions can be provided by R. Quinn or P.C. Dorrestein.

The objective of this study was to characterize developmental differences in low birth weight (LBW) and normal birth weight (NBW) piglets with or without pre-weaning nutrient restriction using serum metabolomic profile analysis. At farrowing, 112 piglets were identified as LBW (1.22 ± 0.28 kg) or NBW (1.70 ± 0.27 kg) and were randomly assigned to receive normal nutrition (NN) or restricted nutrition (RN) (6 h/day no suckling) from d 2 to 28 post-farrow (n=8pigs/group). On d 28, piglets were weaned onto a common diet. Fasted blood samples were obtained on d 28 and 56 (n=8pigs/group) and were analyzed using quantitative metabolomics via a combination of direct injection mass spectrometry with a reverse-phase LC-MS/MS custom assay. Data were normalized using logarithmic transformation and auto-scaling. Partial least squares discriminant analysis (PLS-DA) was carried out to further explore the differential metabolites among the groups (metaboanalyst.ca) with an integrated enrichment and pathway topography analysis. On d 28, LBW piglets had lower levels of essential amino acids as well as reduced metabolites associated with fatty acid oxidation, glycolysis, and the TCA cycle compared to the NBW group. The overall reduction of metabolites associated with energy production and regulation suggests that LBW vs. NBW are in an energy-survival state. On d 56, LBW pigs had increased utilization of fatty acids and resultant ketone production, evident by increased carnitines, acetoacetate, β-hydroxybutyrate, and glycerol compared to NBW pigs. Additionally, compared to the NBW pigs LBW pigs had a consistent decrease in serum glucose and lactate as well as reduced TCA cycle metabolites: pyruvate, succinate, citrate, and α-ketoglutaric acid similar to d 28. Low reliance on glycolysis and the TCA cycle and higher glycerol production in the LBW pigs may indicate impairments in glucose tolerance at 56 d. In summary, LBW piglets appear to have more metabolic alterations in early life, which is not resolved with adequate nutrition or refeeding and may elucidate physiological and metabolic mechanisms of poor growth and life performance compared to NBW pigs later in life.

Undernutrition induced growth restriction in early life increases the risk of chronic disease in adulthood. While metabolic impairments have been observed, few studies have characterized the gut microbiome and gut-liver metabolome profiles of growth restricted animals during early-to-mid-life development. To induce growth restriction, mouse offspring were either born to gestational undernutrition (GUN) or suckled from postnatal undernutrition (PUN) dams fed a protein-restricted diet (8% protein) or control diet (CON; 20% protein) until weaning at postnatal age of 21 days (PN21). At PN21, all mice were fed the CON diet until adulthood (PN80). Livers were collected at PN21 and PN80, and fecal samples were collected weekly for gut microbiome and metabolome analyses. PUN mice exhibited the most alterations in gut microbiome and gut and liver metabolome compared to CON. These mice had altered fecal microbial Beta-Diversity (p=0.001) and exhibited higher proportions of Bifidobacteriales (Linear Mixed Model (LMM) p=7.158x10-6), Clostridiales (p=1.459x10-5), Erysipelotrichales (p=0.0003) and lower Bacteroidales (p=4.121x10-5). PUN liver and fecal metabolome had a reduced total bile acid pool (p<0.01), as well as lower Riboflavin (p=0.003), amino acids (i.e., methionine (p=0.0018), phenylalanine (p=0.0015), and tyrosine (p=0.0041)), and higher excreted total peptides (LMM p=0.0064), dipeptides (p=0.001), and tripeptides (p=0.0123) compared to CON. PUN also had abnormally reduced Epiandrosterone (p=0.02) at PN80. PUN liver and fecal metabolome varied in specific acylcarnitines: higher liver oleoyl-L-carnitine (p=0.0038) and palmitoylcarnitine (p=0.0096) vs. CON at PN21, which recovered at PN80. However, PUN also had higher fecal R-ButyrylCarnitine (p=0.0125) through adulthood, despite refeeding. Overall, protein restriction during lactation permanently alters the gut microbiome into adulthood. While the liver bile acids, amino acids, and acylcarnitines recovered, the fecal peptides and microbiome remained permanently altered into adulthood, indicating that inadequate protein intake in a specific time frame in early life can have an irreversible impact on the microbiome and fecal metabolome.

INTRODUCTION Early life growth restriction increases susceptibility to non‐communicable adult‐onset diseases. However, there is minimal mechanistic rationale for elevated disease risk. The gut microbiome assists in regulating host physiological processes, with emerging evidence suggesting microbiome dysregulation leads to increased risk of chronic disease. Thus, the purpose of this investigation was to determine differences in the gut microbiome of growth restricted mice as compared to non‐restricted mice. METHODS A cross‐fostering, protein‐restricted (8% protein) nutritive model in FVB mice was used to induce growth restriction during gestation (GUN; N=10 male,15 female) or lactation from days 1–21 of life (PUN; N=11 male, 12 female) along with a control group (CON, 20% protein, N=12 male, 11 female). At 21 postnatal days of age (PN21) all mice were weaned to a non‐restricted diet (20% protein), isolating undernutrition to early life windows. Fecal samples were collected weekly across the lifespan (until PN80) to determine longitudinal programming effects of growth restriction on the gut microbiome. Cecum samples were collected at PN21 and PN80. Microbiome analysis of fecal and cecum samples were conducted by first extracting the DNA using Qiagen Powersoil DNA extraction kits and conducting PCR amplification of the bacterial 16S rRNA gene amplicon profiles through the Qiita bioinformatics platform. Phenotypic growth markers of all three diet groups (CON, GUN, PUN) were analyzed using repeated measures ANOVA. Metabolomics data on the same samples is currently being analyzed and will be integrated with the microbiome dataset. RESULTS The fecal microbiome was strongly separated by treatment group using the Bray‐Curtis distance measure (PERMANOVA p = 0.001). A random forest machine learning classification validated this separation showing an overall out‐of‐bag error rate of 18.6% across the three groups in the whole dataset. The PUN group was distinctly separated from GUN and CON, with a group error rate of only 15.0%. Differences in the microbiome were not evident through analysis at the Phylum level (Firmicutes/Bacteroidetes ratio) but was instead driven by differential abundances of specific rare taxa. Bifidobacterium was significantly elevated in the PUN group (Kruskal‐Wallis test p=0.00017), beginning as early as PN21 and remaining elevated in the PUN group through the lifespan (PN80). CONCLUSIONS A protein‐restricted nutritive model induces strong imprinting on the structure of the murine microbiome. Changes in the PUN group are primarily driven by differences in low abundance taxa. Understanding the mechanisms that lead to the microbiome dysbiosis shown here could lead to the development of therapeutic interventions to limit the development of a stunted or immature gut microbiota. Support or Funding Information MSU Department of Biochemistry and Molecular Biology; MSU Department of Kinesiology

No abstract