In cattle, as in other ruminants, the transition from a blastocyst to a fully elongated conceptus is one of the major morphological events that take place during the peri-implantation period of pregnancy. This process is a maternally-driven event, as elongation does not occur in vivo in the absence of uterine gland secretions in sheep , and has yet to be recapitulated in vitro in cattle [8, 9]. Transfer of in vitro derived blastocysts  or trophoblastic vesicles  into the uterus of synchronized recipients results in successful conceptus elongation; ULF must therefore contain molecules that drive conceptus elongation. One approach to predicting which proteins are present in ULF is to examine the endometrial expression of selected genes that encode for proteins that are known to be secreted or extracellular in nature. If expression of these genes is modulated during early pregnancy and/or by altered concentrations of P4 in vivo (which have demonstrable effects on the elongation of the conceptus), it is plausible that they are candidate genes likely to influence conceptus elongation. Using this candidate gene approach, we found that expression of ARSA, MUC13, NXPH3 and PENK increased with advancing stages of the estrous cycle and early pregnancy. Similarly, LCAT, NCDN, NMN, PLIN2 and TINAGL1 expression increased during the later luteal phase, with the presence of the conceptus significantly increasing their expression. In contrast, APOA1, DCN and NPNT expression decreased in pregnant compared with cyclic heifers. A delay in the postovulatory increase in circulating concentrations of P4 delayed the increase in expression of LCAT and NPNT on Day 7 and APOA1, PENK and TINAGL1 on Day 13, the day on which conceptus elongation is initiated. Finally, comparison of this list of candidate genes with the protein content of ULF of pregnant heifers during conceptus elongation (Days 13 through 19) indicated the presence of APOA1, ARSA, LCAT, NCDN as well as members of the PLIN family, lending validity to this candidate gene expression approach.
In cattle, two of the main molecules that affect endometrial gene expression are circulating concentrations of P4 and conceptus-derived IFNT, both of which can affect the composition of the ULF and alter the rate of conceptus elongation. The comparison of gene expression in the endometrium and proteins present in ULF identified four proteins (APOA1, ARSA, LCAT and NCDN) from 14 candidate genes, while two members of the perilipin family were detected. At first glance, the number of proteins detected in ULF in comparison to the number of candidate genes may seem low. However, it is important to note that there is at least an order of magnitude in the difference between the number of total transcripts detectable in the endometrium by RNA-SEQ (~21,000 transcripts: ) versus the number of total proteins detected in ULF by proteomics . In addition, analysis of gene expression was performed on endometrial homogenates which contain a heterogenous cell population. It is possible that, despite their status as genes that encode for extracellular/secreted proteins, their expression may be predominantly stromal in nature, similar to the expression of HGF and FGF10 in the sheep endometrium. As such, the protein products of these genes may not be secreted into ULF, but may act in a more autocrine or paracrine manner within the endometrium.
The patterns of expression of ARSA, MUC13, NXPH3 and PENK were similar in endometria from both pregnant and cyclic heifers. MUC13 exhibited minimal expression on Day 7 and 10, while ARSA, NXPH3 and PENK expression decreased on Day 10, but increased significantly during the late luteal phase of the estrous cycle and as pregnancy progressed. Given that the expression of these genes increased during the luteal phase of the estrous cycle and early pregnancy, we propose that they play a role in establishing uterine receptivity to implantation. Although the protein products of NXPH3 or PENK were not detected in ULF, the increased mRNA abundance for these genes in endometria of P4-supplemented heifers strongly suggests that they contribute to advancing conceptus elongation after both artificial insemination  and embryo transfer .
In contrast, endometrial expression of APOA1, DCN and NPNT decreased in to the presence of the developing conceptus. Expression of both APOA1 and NPNT was modulated by P4, with a delay on Day 13 in APOA1 expression, while P4 supplementation resulted in an early increase in NPNT on Day 7. We propose that the decreased expression of these genes in endometria in the peri-implantation period of pregnancy, may be a pre-requisite for establishing uterine receptivity to implantation, similar to that observed for nuclear P4 receptors in all species studied  as well as MUC1 in sheep .
The modulation of circulating concentrations of P4 in cattle and sheep has demonstrable effects on gene expression in the endometrium and timing of conceptus elongation. In cattle, this results in altered trajectories of conceptus elongation on Day 14 following embryo transfer on Day 7. Exogenous supplementation of P4 advances conceptus elongation ; conversely, when P4 output from the CL is diminished , conceptus elongation is delayed. High circulating concentrations of P4 advanced the normal increase of NXPH3 expression on Day 7 of the estrous cycle, while in heifers with low P4 the normal expression level of NPNT was decreased. In addition, decreased expression of APOA1 and PENK on Day 13 in heifers with low P4 and delayed conceptus elongation suggests that these endometrial genes and their protein products contribute to the process of conceptus elongation. This hypothesis is supported by evidence for the presence of APOA1 protein in ULF on both Day 13 and 16.
Of the 14 candidate genes examined, changes in expression of NMN, PLIN2 and TINAGL1 were of particular interest. Neurotensin (NT) was first isolated from the bovine hypothalamus  and small intestine  and NMN protein, along with NT is derived from the same precursor peptide with the resulting active form of the peptides (i.e., NT or NMN) being dependent of the site of cleavage of the pro-peptide (reviewed by ). Interestingly, the predominant forms of the processed peptides are tissue-specific, e.g., the long form of NMN is the predominant form in the gut . NMN expression was similar in pregnant and cyclic heifers up to Day 16 when pregnancy recognition occurs and was modulated by low circulating concentrations of P4. NMN protein was not detected in ULF, but a protein that degrades NMN, aminopeptidase M , has been detected in ULF . This may explain in part why NMN protein was not detected and, given its expression pattern, may suggest that NMN is involved in the pregnancy recognition response in the endometrium rather than conceptus elongation itself.
TINAGL1 is a matri-cellular protein that can interact with both extracellular matrix proteins and cell surface receptors. In mice, TINAGL1 is expressed by the trophectoderm and interacts with laminin 1 . TINAGL1 expression increases in the endometrium of pregnant mice as well as at the sites of implantation where the abundance of protein increases during the later stages of implantation . In addition, TINAGL1 has been linked to the integrins ITGA5 and ITGB1 in the decidua . In the current study, TINAGL1 expression was minimal from Day 7 to 13, but increased on Day 16 and to a greater extent in pregnant than cyclic heifers. This is interesting as there is basal expression when the uterus is not receptive to implantation (i.e., when PGR is expressed by uterine LE and GE). TINAGL1 expression increases after uterine receptivity to implantation is established, but interestingly, its expression on Day 13, when conceptus elongation is initiated, is minimal. Moreover, TINAGL1 protein is not detectable in ULF. Thus, TINALG1 in the endometrium of cattle is not directly involved in conceptus elongation, but instead plays a role in implantation as evidenced by its increase in expression in pregnant heifers on Days 16 and 19, similar to its role in mice.
PLIN2 (a.k.a. ADRP) protein is involved in lipid droplet formation and, in particular, the intracellular storage of triaclyglycerols (TAG: reviewed by ). It is also detectable in milk lipid droplets secreted from mammary epithelial cells in a variety of species including humans, rats and cattle . Triaclyglycerols have been detected in the ULF of cattle  and in vitro TAG is utilized as an energy source by the embryo up to the blastocyst stage of development . In both breast cancer and leiomyoma human cell lines, the PLIN2 gene has a PGR response element in its promoter region. Given its modulation by P4 in vivo and its increased expression in pregnant heifers likely due to IFNT , as well as detection of PLIN2 family members in ULF of cattle, it is likely that PLIN2 contributes to the secretion and /or transport of TAG into the ULF as an energy source for the developing conceptus during elongation.