Targeted metabolomics reveals reduced levels of polyunsaturated choline plasmalogens and a smaller dimethylarginine/arginine ratio in the follicular fluid of patients with a diminished ovarian reserve

STUDY QUESTION: Does the metabolomic profile of the follicular fluid (FF) of patients with a diminished ovarian reserve (DOR) differ from that of patients with a normal ovarian reserve (NOR)? SUMMARY ANSWER: The metabolomic signature of the FF reveals a significant decrease in polyunsaturated choline plasmalogens and methyl arginine transferase activity in DOR patients compared to NOR patients. WHAT IS KNOWN ALREADY: The composition of the FF reflects the exchanges between the oocyte and its microenvironment during its acquisition of gametic competence. Studies of the FF have allowed identification of biomarkers and metabolic pathways involved in various pathologies affecting oocyte quality, but no large metabolomic analysis in the context of ovarian ageing and DOR has been undertaken so far. STUDY DESIGN, SIZE, DURATION: This was an observational study of the FF retrieved from 57 women undergoing in vitro fertilization at the University Hospital of Angers, France, from November 2015 to September 2016. The women were classified in two groups: one including 28 DOR patients, and the other including 29 NOR patients, serving as controls. PARTICIPANTS/MATERIALS, SETTING, METHODS: Patients were enrolled in the morning of oocyte retrieval after ovarian stimulation. Once the oocytes were isolated for fertilization and culture, the FF was pooled and centrifuged for analysis. A targeted quantitative metabolomic analysis was performed using high-performance liquid chromatography coupled with tandem mass spectrometry, and the Biocrates Absolute IDQ p180 kit. The FF levels of 188 metabolites and several sums and ratios of metabolic significance were assessed by multivariate and univariate analyses. MAIN RESULTS AND THE ROLE OF CHANCE: A total of 136 metabolites were accurately quantified and used for calculating 23 sums and ratios. Samples were randomly divided into training and validation sets. The training set, allowed the construction of multivariate statistical models with a projection-supervised method, i.e. orthogonal partial least squares discriminant analysis (OPLS-DA), applied to the full set of metabolites, or the penalized least absolute shrinkage and selection operator with logistic regression (LASSO-LR), applied to the ratios and sums of the metabolites. Both multivariate models showed good predictive performances when applied to the validation set. The final penalized model retained the three most significant variables, i.e. the total dimethylarginine-to-arginine ratio ( Total DMA/Arginine), the sum of the polyunsaturated choline plasmalogens (PUFA ae), and the patient's age. The negative coefficients of Total DMA/Arginine and PUFA ae indicated that these FF variables had lower values in DOR patients than in NOR patients. LARGE SCALE DATA: N/A. LIMITATIONS REASONS FOR CAUTION: This study presents two limitations. First, with this targeted metabolomics analysis, we have explored only a limited portion of the FF metabolome. Second, although the signature found was highly significant, the mechanism underlying the dysfunction remains undetermined. WIDER IMPLICATIONS OF THE FINDINGS: The understanding of the mechanisms implied in ovarian ageing is essential for providing an adequate response to affected women desiring pregnancy. Our study proposes an incoming signature that may open new paths towards this goal.