Our results demonstrate the expression of PR, PGRMC1 and, for the first time, of PGRMC2, in several compartments of the bovine oviduct between Day 1.5 and Day 5 post-ovulation. To our knowledge, this is the first exploration of PGRMC1 and PGRMC2 gene expression in this organ. At Day 3.5 post-ovulation, the PR gene was more highly expressed in epithelial cells from the isthmus than from the ampulla, whereas PGRMC2 displayed an opposite pattern of expression. However, neither the side relative to the CL nor the pregnancy status changed the gene expression of P4 receptors.
Western blot analysis of PR detected both PRA and PRB subtypes in oviduct samples, as previously reported in bovine epithelial cells . Although both PRA and PRB bind P4, several studies have shown that, depending on the cell and target genes, PRA and PRB have remarkably different transcriptional activities . However, little information is available so far with respect to the respective roles of PRA and PRB in mediating P4 action in the mammalian oviduct [32, 33]. Nevertheless, in the present study, nuclear P4 receptors were expressed between Day 1.5 and Day 5 post-ovulation in both the ampulla and the isthmus. These results are in keeping with earlier reports on PR expression in bovine whole-oviduct sections [17, 34] and in isolated oviduct epithelial cells . A nuclear immunostaining was observed in the luminal epithelial cell layer, as well as in the muscle layer and stromal tissue, in agreement with earlier reports using the same anti-PR antibody  or some others [17, 34, 35] in bovine oviduct samples collected during the first five days of the oestrous cycle. Accordingly, one of the latter studies reported that the muscle layer was positively stained only during the early luteal phase (days 1–5) compared with other phases of the oestrous cycle in the cow .
We observed that PR, PGRMC1 and PGRMC2 were expressed in both the ampulla and isthmus at all stages examined. To our knowledge, this is the first report of PGRMC2 expression in the mammalian oviduct. Western blot analysis for PGRMC2 revealed one band at 26 kDa, as previously described in bovine cumulus and luteal cells . In contrast, two signals at 23 kDa and 28 kDa were observed on PGRMC1 western blots. While the PGRMC1 protein is usually detected as a single signal between 24 and 28 kDa [21, 23, 25, 26], multiple bands were previously reported at the same apparent molecular weights in murine placental tissues , which were assumed to represent PGRMC1 oligomers . Various protocols of sample fixation/preparation were used in the latter and present studies and could explain these discrepancies. These conflicting findings could be also attributed to the different tissues/cells and subcellular compartments analysed. Indeed, an antibody from the same source was reported to detect a single band in membrane preparations from human and rat granulosa/luteal cells [21, 37] and multiple bands in whole murine uterine tissues .
PGRMC1 has been reported in various subcellular locations, including plasma membrane [21, 38], cytosol  and nucleus [26, 36, 37]. In this study, PGRMC1 was mainly localised in the cytosol of luminal epithelial cells, and to a lesser extent in muscular and stromal cells. This differs from findings in a previous study  that localised PGRMC1 in the bovine oviduct: the protein was observed in both the nucleus and cytoplasm of the luminal epithelium, endothelium and muscle layer at equivalent staining levels. This divergence may be due to differences in stages of estrous cycle examined, which were not precisely timed in the latter study , as well as to differences in antibodies and protocols used. The fact that we had a clear signal in the nucleus for PR and PGRMC2 demonstrates, however, the efficacy of the method used. Furthermore, our results are in keeping with those of Aparicio et al., , who localised PGRMC1 only in the cytoplasm of bovine cumulus cells. With the antibody used in the present study, the immunostaining for PGRMC1 was intense in the nucleus of human and rat granulosa/luteal cells [21, 37, 39]. However, the subcellular localization of PGRMC1 was reported to change dramatically after eCG treatment in the immature rat ovary  and depending on the pregnancy stage in the mouse uterus . Thus, the expression of PGRMC1 in various tissues and subcellular compartments of the bovine oviduct might be regulated by the hormonal environment and requires further investigations. The mRNA and/or protein for PGRMC2 have been detected previously in the murine  and macaque  endometrium, human choriodecidua , newborn rat ovary , and bovine cumulus cells . Various subcellular localisations for the PGRMC2 protein were previously observed depending on the species and cell tissue: throughout the cytoplasm in bovine cumulus cells , in the cytoplasm of the luminal and glandular epithelia and in the nuclei of the stromal cells in the macaque endometrium .
Oviducts in the present study were collected from cows treated for estrous synchronisation. Although not well documented, compared with natural cycles, synchronisation protocols may have some impact on the oviduct physiology. Furthermore, oviduct epithelial cells from pregnant animals in the present study were collected from cows inseminated twice at Day 0 and Day 0.5 post-ovulation. Although the lumen of the oviducts was rinsed prior to their use for RNA extraction, sperm cells attached to the luminal epithelium, especially in the isthmic sperm reservoir , may have been picked up in these extraction processes. Since PR and also PGRMC1 and PGRMC2 have been identified in porcine  and human  sperm, one may question whether our results include expression data from bovine sperm cells. Several sets of data seem to exclude this hypothesis. First, the number of spermatozoa counted per entire bovine oviduct between 17 and 20 h after insemination is around a few tens . Second, nuclei from spermatozoa that transit in the female genital tract are transcriptionally inactive  and therefore, mRNAs from spermatozoa have no chance of being co-amplified during the RT-qPCR performed on oviduct samples. Thus, it appears reasonable to exclude possible contamination with bull spermatozoa in our expression data.
While the gene expression of PR in target tissues is classically up-regulated by estradiol and down-regulated by P4 , the factors that regulate the expression of PGRMC1 and PGRMC2 are largely unknown. Nonetheless, P4 has been identified as a good candidate for the regulation of PGRMC1 gene expression in mouse brain and placenta [36, 38] as well as in the endometrium of human [19, 46, 47] and rhesus monkey . Furthermore, the gene expression of both PGRMC1 and PGRMC2 was recently demonstrated in the canine oviduct, in which their mRNA levels varied significantly with the stage of the estrous cycle . The bovine oviduct ipsilateral to the CL has been shown to contain higher levels of P4 compared with the contralateral oviduct during the luteal phase of the estrous cycle . However, in this study, quantitative RT-PCR applied on oviduct epithelial cells did not reveal any significant difference in the gene expression of PR PGRMC1 or PGRMC2 between ipsilateral and contralateral oviducts. Accordingly, Rottmayer et al.  reported no difference in PR mRNA expression between epithelial cells collected from bovine ipsilateral and contralateral oviducts. By a transcriptomic approach, 35 genes were found to be differentially expressed between epithelial cells from ipsilateral versus contralateral bovine oviducts collected at Day 3.5 post-oestrus . However, in accordance with our findings, genes for P4 receptors were not among the differentially expressed genes. Furthermore, the localisation pattern of the three P4 receptors observed in this study did not reveal any difference between the two oviducts. In accordance, PGRMC1 was immunolocalised by others in a similar pattern in both oviducts from cyclic cows . In this study, the effect of the pregnancy status on the gene expression of nuclear and membrane P4 receptors was explored for the first time in the bovine oviduct. Although plasma P4 concentrations do not differ between cyclic and pregnant cows at Day 3.5 post-ovulation , local differences in tubal vascularisation and ciliary beating in the bovine oviduct epithelium were previously shown to depend on the presence or absence of an embryo [8, 9]. However, here, the pregnancy status did not influence the expression of any of the P4 receptors at Day 3.5 post-ovulation. Similarly, in the bovine endometrium, the level of PR mRNAs was not influenced by the pregnancy status during the early post-ovulation period (Days 5–7) . Our results do not exclude differential expression of these receptors between cyclic and pregnant cows in other oviductal tissue layers or at later stages post-ovulation. Indeed, in the bovine endometrium, PR mRNA concentration was lower in pregnant cows than in cyclic cows at Day 13 post-ovulation, but not earlier .
In the present study, the PR gene was more highly expressed in epithelial cells from the isthmus than from the ampulla, whereas PGRMC2 displayed an opposite pattern of expression. Higher amounts of PR mRNAs in the isthmus compared with the ampulla were previously reported in the bovine [18, 34] and rat  oviduct luminal epithelia. However, in the pig oviduct, PR mRNAs were measured at comparable levels in both oviduct regions . As the ampulla is mainly implicated in oocyte transport and fertilisation , a positive effect of PGRMC2 in these tubal events might be assumed. Furthermore, a higher expression of PR in the isthmus might have beneficial effects in the specific functions of this part of the oviduct, i.e. in the storage of spermatozoa before fertilisation and/or in the timed transport of the embryo toward the utero-tubal junction . The factors that could regulate the differential gene expression of PR and PGRMC2 in the two parts of the bovine oviduct are still unknown. Previous measurement of P4 levels in whole-oviduct sections collected during the first 5 days post-ovulation revealed no difference in P4 concentrations between the ampulla and isthmus . Thus, it seems likely that factors other than local P4 regulate the region-specific gene expression of PR and PGRMC2 in the oviduct epithelium.
An important question remains the potential roles of nuclear and membrane P4 receptors in the regulation of tubal functions. A role for PGRMC1 in the mediation of the anti-apoptotic effects of P4 has been reported in ovarian granulosa cells [19–21, 56]. Furthermore, PGRMC1 has been recently implicated in the mediation of myometrium relaxation during pregnancy in human . Of interest, a proteomic approach showed that PGRMC2 was up-regulated in the human choriodecidua during term and preterm labor . A central role of the oviduct during the post-ovulation period is to allow the transport of the cumulus-oocyte complex to the site of fertilisation, then of the early embryo toward the uterus, achieved by synchronised ciliary beating, smooth muscle contraction and flow of tubal secretions . It might be assumed that both PGRMC1 and PGRMC2 are involved in tubal contraction and relaxation. However, as both receptors were expressed in the luminal epithelium, the mediation of P4 action on tubal ciliary beating or fluid secretion cannot be excluded. Further experiments would be required to determine the potential functions of P4 nuclear and membrane receptors in the bovine oviduct.