Urinary metabolomics reveals the biological characteristics of early pregnancy in pigs

Abstract

Background: Embryo implantation in sows is an important event during pregnancy. During this process, blastocysts undergo dramatic morphologic changes, and the endometrium becomes receptive. Studies have shown that developmental changes associated with the crosstalk between peri-implantation embryos and embryo-uterine are driven by various biomolecules secreted by the endometrium and embryos. In sows, changes in the uterus are also reflected in circulating body fluids and urine. Metabolomics reveals the metabolic state of cells, tissues, and organisms. In this study, we collected urine samples from large white sows during the peri-implantation period. The levels of urinary metabolites at different periods were analyzed using ultra-performance liquid chromatography/tandem mass spectrometry (UPLC-MS/MS) analysis techniques.

Results: A total of 32 samples were collected from 8 sows during the estrus period and at each phase of early pregnancy (9, 12, and 15 days of gestation). A total of 530 metabolites were identified with high confidence in all samples. Compared with samples collected during the estrus phase, 269 differential metabolites were found in samples obtained during early pregnancy.

Conclusions: The identified metabolites included lipids and lipid-like molecules, organic acids and their derivatives, organic oxygen compounds, organoheterocyclic compounds, benzenoids, among others. Metabolites, such as choline and pregnanediol-3-glucuronide, play important roles in pregnancy in sows and other animals. These results reveal the metabolic changes in urine of sows during early pregnancy phase. The differential urinary metabolites can be used for assessing peri-implantation status in sows. Understanding these metabolic changes may promote the management of pregnant sows through various interventions such as provision of proper nutrition.

Keywords: Early pregnancy; Metabolome; Pig; Pregnanediol-3-glucuronide; Urine.

Fig. 1

Typical total ion chromatograms (TICs), A positive ions and B negative ions, obtained from the UHPLC-Q Exactive ESI of Quality Control samples

Fig. 2

Pie chart showing the percentages of diverse urine metabolites based on counts of HMDB chemical taxonomy (“Super class”). The selected HMDB level (“Super class”) and the percentage of metabolites are displayed in a descending order. Different colors on the pie chart represent different HMDB categories while the area represents relative proportions of metabolites in that category

Fig. 3

Principle component analysis score plots of metabolites identified in urine during estrus and pregnancy periods. A represents positive ions, B represents negative ions. Each point on the PCA score chart represents a sample

Fig. 4

VIP score analysis based on weighted coefficients of the OPLS-DA model used to rank the contribution of top 30 metabolites to the discrimination between estrus and early pregnancy groups. A Heat map showing VIP of metabolites between DP9 and DP0, B A heat map showing VIP of metabolites between DP12 and DP0, C A heat map showing VIP of metabolites between DP15 and DP0, D The overlap of metabolites between early pregnancy (DP9, DP12, DP15) and estrus (DP0)

Fig. 5

Histogram showing differential metabolites annotated by comparisons to the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Differential metabolites were classified by KEGG pathway enrichment and significance analysis. Fold enrichment is presented as the ratio of the number of metabolites assigned to the modified pathway by enrichment analysis to the theoretical number of metabolites assigned to the modified pathway by random distribution. The degree of enrichment is indicated by different colors in the histogram, according to the p value. Histogram (A) shows DP9 versus DP0, histogram (B) shows DP12 versus DP0, whereas histogram (C) shows DP15 versus DP0

Fig. 6

Relative expression levels of urinary metabolites at 4 different stages of porcine gestation. All metabolites were clustered into 8 soft partitioned clusters based on expression patterns