Abstract :
[en] The equine placenta produces conjugated estrogens: research mainly focused on sulfonated estrogens although glucuronide estrogens were observed in early studies. These preliminary investigations were performed using immunoassays that lack of specificity for small molecules. The higher analytic accuracy of liquid chromatography tandem mass spectrometry (LC-MS/MS) enables to refine the knowledge about estrogens production during equine pregnancy. This study aims to describe the evolution of estrone and estradiol glucuronide (E1G and E2G) during equine pregnancy in two different breeds and to determine if their secretion is linked to native and sulfonated estrone or estradiol (E1, E1S or E2, E2S). Between 2020 and 2024, serum samples were collected monthly from 18 Warmbloods (WB) and 24 Spanish Pure-Breed (SPB) pregnant mares. From 4 months of pregnancy onward, the Combined Thickness of the Uterus and Placenta (CTUP) was measured by ultrasonography. Mares with enlarged, heterogeneous CTUP or clinical signs of placentitis pre- or post-partum (premature lactation, vulvar discharge, abortion, abnormal macroscopic placenta, foal weakness...) were excluded from this study. A dedicated LC-MS/MS method was used to assay serum E1, E2, E1S, E2S, E1G and E2G. For the 93 pregnancies included, monthly hormones’ evolutions were assessed using the Kruskal-Wallis test, and potential concentration differences between breeds were studied with Mann-Whitney test. Non-parametric Spearman’s test was used to assess correlations between E1G, E2G, E1, E2, E1S and E2S. Data are reported as median (percentiles 25 and 75) with statistical significance set at p < 0.05. The peak concentration of E1G (12,068 (6,530-15,751) pg/mL) was observed at 2 months, whereas the peak value of E2G was lower (1,692 (1,036-2,000) pg/mL) and observed at 4 months. There was a breed effect for both hormones, but profiles differed: E1G concentrations were higher in WB from 6 to 10 months of pregnancy, whereas higher E2G levels were observed in SPB at 3, 7 and 8 months. Concentration of E1G was poorly correlated with other studied estrogens, with its strongest correlation observed with E1S (rs = 0.47, p < 0.0001). Positive correlations were found between E2G and E1 (rs = 0.66, p < 0.0001), E1S (rs = 0.71, p < 0.0001), and E2S (rs = 0.70, p < 0.0001). To the best of our knowledge, LC-MS/MS was never used to describe E1G and E2G kinetics during pregnancy, and differences in their concentrations between breeds of different sizes was not reported. Although glucuronosyltransferase genes are constantly expressed by the endometrium during pregnancy, maximal E2G concentration was observed at 4 months, before the E1, E2 and E1S peaks at 5 months. Surprisingly, the E1G peak occurred earlier in pregnancy, one month after the onset of endometrial cups’ eCG production, which has an FSH-like effect normally promoting follicular E2 secretion. This unique E1G kinetic during pregnancy was confirmed by the absence of correlation with E1, E1S, E2G, and E2S, which all correlate, suggesting different pathways of production and effects of E1G. Further studies should confirm the origin of early E1G production in pregnant mares and physiological effects of its secretion after the implantation.
