Feasibility and accuracy of using different methods to detect pregnancy by conceptus-stimulated genes in dairy cattle

Development of new methods for early diagnosis of pregnancy can be important to increase the reproductive efficiency and profitability of dairy herds. The bovine conceptus secretes IFN-τ that stimulates the transcription of several genes in circulating immune cells and extrauterine tissues. The aims of this study were to evaluate the mRNA abundance for pregnancy predictability of a classic gene stimulated by IFN-τ (ISG15) and a novel potential pregnancy marker (LGALS3BP) in peripheral blood mononuclear cells (PBMC), total blood leukocytes (TBL) or milk leukocytes (TML), and cervical cells (CC) on d 20 after timed artificial insemination (TAI) in dairy cattle. Eighteen Holstein females (12 cows and 6 heifers) were submitted to an estrous synchronization protocol for TAI (d 0). On d 20 post-TAI, blood samples were collected from coccygeal vessels for isolation of PBMC and in Tempus Blood RNA tubes (Applied Biosystems) for TBL. Samples of CC were collected using a cytological brush, and the TML were isolated from milk samples collected before routine milking. Pregnancy diagnosis was performed on d 30 post-TAI using transrectal ultrasonography, and females were classified as pregnant (n = 8) or nonpregnant (n = 10). Total RNA was extracted and mRNA abundance of target genes (ISG15 and LGALS3BP) was quantified by reverse-transcription quantitative PCR and normalized in relation to reference genes. Data were analyzed by ANOVA using the MIXED procedure of SAS (SAS Institute Inc.). The mRNA abundance of ISG15 was greater in pregnant than in nonpregnant animals for PBMC, TBL, and CC. No difference was detected for TML based on pregnancy status. For LGALS3BP mRNA abundance, no difference was detected between pregnant and nonpregnant animals for PBMC, TBL, and TML, but a tendency for greater abundance in pregnant animals was observed for CC. The fold change for ISG15 in each pregnant cow related to the mean of nonpregnant animals was 2.73 ± 0.31, 3.40 ± 2.17, 1.64 ± 0.29, and 0.005 ± 0.002 for PBMC, CC, TBL, and TML, respectively. The fold change for LGALS3BP in each pregnant cow related to the mean of nonpregnant animals was 0.97 ± 0.38, 1.77 ± 0.39, 0.20 ± 0.08, and 0.70 ± 0.11 for PBMC, CC, TBL, and TML, respectively. The receiver operating characteristic curve analysis indicated that ISG15 abundance predicted pregnancy in PBMC (area under curve, AUC = 0.92) and CC (AUC = 0.77) but not in TBL (AUC = 0.72) or TML (AUC = 0.52). In conclusion, mRNA abundance for ISG15 in PBMC was the best predictor for pregnancy at d 20 post-TAI, whereas TBL and TML were not good predictors of pregnancy on d 20 post-TAI. The mRNA abundance of LGALS3BP was not associated with pregnancy status in any type of cell evaluated.