Characterization of Diet-Dependent Metabolic Serotypes: Primary Validation of Male and Female Serotypes in Independent Cohorts of Rats

Open Access
Publisher Website
Google Scholar
Abstract
Our research seeks to identify serum profiles, or serotypes, that reflect substantial changes in food intake in both male and female rats. This report validates previously defined subsets of redox-active low-molecular-weight metabolites using independent cohorts of ad libitum consumption (AL) and energy or dietary restricted (DR) 6-mo-old male and female rats. In the male study, both hierarchical cluster analysis (HCA) and principal component analysis (PCA) distinguished the dietary groups of origin in the second male cohort with >85% accuracy using 56 analytically and biologically valid metabolites. Further analysis revealed that 29 metabolites (nine previously unidentified metabolites + 20 chosen from the 56 metabolites) enabled HCA to distinguish dietary groups at 100% efficacy. In the female study, the 63 previously identified serum metabolites were sufficiently robust to enable classification of the dietary intake of two female cohorts (cohorts 2 and 3) that were independent of the cohort on which these markers were initially identified (cohort 1). Classification accuracy was 94 and 100% using HCA and PCA, respectively, in the female cohort 2. HCA and PCA revealed that the 63-metabolite profile distinguished AL and DR samples at 91 and 100% accuracy in the female cohort 3, establishing the 63-metabolite dataset as our baseline profile. These studies used independent cohorts to validate and potentially improve upon previously defined metabolic serotype in male and female rats and set the stage for pattern recognition–based approaches to establish metabolome-based categorical separations.