GENE EXPRESSION IN THE EQUINE ENDOMETRIUM DURING MATERNAL RECOGNITION OF PREGNANCY

Abstract

Mechanisms of equine embryo-uterine signaling and maternal recognition of pregnancy are unknown. In non-pregnant mares, endometrial production of prostaglandin F2-alpha (PGF) from d14 – 18 post ovulation drives luteolysis. Endometrial recognition of a viable embryo is therefore required to avoid the loss of the corpus luteum (CL). In sows, PGF is sequestered in the uterine lumen by embryo-produced estrogen (E). Equine embryos produce E, yet luminal PGF is not found in any higher quantities. Human pregnancy maintenance is dependent on human chorionic gonadotropin. The equine conceptus also produces a chorionic gonadotropin, but not until d38, long after maternal recognition. Mice induce CL function by coitus. Luteolysis occurs following either PGF production by the placenta at parturition or another mating attempt. The ruminant conceptus secretes interferon-tau (IFN), which results in reduction of PGF production. Equine embryos do not produce INF but do, by their presence, prevent PGF production. Investigators have demonstrated the need for a viable and mobile equine conceptus, which results in the lack of PGF production. Regardless of species, signals from the conceptus impact endometrial gene expression that either directly or indirectly protects the pregnancy. A common uterine response at the time of maternal recognition in cattle, mice and primates is production of cytokines, chemokines and their receptors that classically have been described in anti-inflammatory responses. We therefore hypothesized that differential equine endometrial gene expression at the time of maternal recognition would be associated with these anti-inflammatory pathways. The objectives of this experiment were to define changes in endometrial gene expression between d18 pregnant and non-pregnant mares and to attempt to define a peripheral maternal recognition signal that might initiate that expression. Each mare (n=3) served as her own control in a cross-over design. Mares were either inseminated or not exposed to semen on alternating cycles to generate pregnant or non-pregnant condition. Ultrasonography was used to identify embryos at d12 and each subsequent day until embryo collection and endometrial biopsy on d18 post ovulation. Blood samples were collected daily from d12- 18. Total RNA from endometrial biopsies was used to screen an Affymetrix Equine chip containing over 3300 unique equine transcripts, including those associated with inflammation. Metabolite profiles of plasma samples (d16) were obtained by liquid chromatography/mass spectrometry (LC/MS). Thirty-six genes were identified to have exhibited a minimum of a 3-fold change following a modified paired t-test analysis (P< 0.0004). Eighteen genes were up regulated in endometrium from pregnant mares. nalysis of the LC/MS data using both principal component and partial least square discriminant analysis yielded a distinction between plasma proteins from pregnant and non-pregnant mares. These preliminary data suggest that a similar expression pattern of small proteins can be predictive of pregnant status and that the equine conceptus is responsible for endometrial gene expression that is differentially regulated during maternal recognition. Confirmation of microarray data by RT-PCR will substantiate antiinflammatory responses. Differentially expressed blood peptide fragments also will be subjected to MS-MS sequencing to identify peripheral markers for maternal recognition of pregnancy in the mare. (poster)