Expression of cyclooxygenase 2 (COX2, now known as PTGS2), prostaglandin E2 synthase (PTGES, PGES), and prostaglandin F2α synthase (PGFS), of the respective receptors PTGFR (FP), PTGER2 (EP2), and PTGER4 (EP4) and of the progesterone receptor (PGR, PR) was assessed by real-time PCR, immunohistochemistry (IHC), or in situ hybridization (ISH) in utero/placental tissue samples collected from three to five bitches on days 8–12 (pre-implantation), 18–25 (post-implantation), and 35–40 (mid-gestation) of pregnancy and during the prepartal luteolysis. Additionally, ten mid-pregnant bitches were treated with the antiprogestin aglepristone (10 mg/kg bw (2×/24 h)); ovariohysterectomy was 24 and 72 h after the second treatment. Plasma progesterone and 15-ketodihydro-PGF2α (PGFM) concentrations were determined by RIA. Expression of the PGR was highest before implantation and primarily located to the endometrium; expression in the placenta was restricted to the decidual cells. PTGS2 was constantly low expressed until mid-gestation; a strong upregulation occurred at prepartal luteolysis concomitant with an increase in PGFM. PGFS was upregulated after implantation and significantly elevated through early and mid-gestation. PTGES showed a gradual increase and a strong prepartal upregulation. PTGFR, PTGER2, and PTGER4 were downregulated after implantation; a gradual upregulation of PTGFR and PTGER2 occurred towards parturition. ISH and IHC co-localized PGFS, PTGFR, PTGES, and PTGS2 in the trophoblast and endometrium. The changes following application of aglepristone were in the same direction as those observed from mid-gestation to prepartal luteolysis. These data suggest that the prepartal increase of PGF2α results from a strong upregulation of PTGS2 in the fetal trophoblast with the withdrawal of progesterone having a signalling function and the decidual cells playing a key role in the underlying cell-to-cell crosstalk.
In the dog, corpora lutea (CL) reach full functional capacity between days 20 and 30 after ovulation; thereafter, the more gradual decrease indicates luteal regression. The resulting course of progesterone (P4) concentrations in peripheral plasma is almost identical in pregnant and nonpregnant dogs until about day 60 of luteal lifespan when the more gradual decline observed so far turns into a steep one as a precondition for parturition, which generally occurs 63 days after mating. In the nonpregnant dog, however, the gradual decline continues to levels <1 ng/ml about 10–20 days later (Concannon et al. 1989, Hoffmann et al. 1994). These deviating patterns in P4 secretion may be seen as an indication for different mechanisms regulating a) gradual luteal regression in nonpregnant dogs and b) prepartal luteolysis.
Both luteotropic and luteolytic mechanisms in pregnant and nonpregnant dogs have been addressed in a number of papers. Gonadotropic support is required for luteal maintenance in pregnant and nonpregnant bitches during somewhat more than the second half of luteal lifespan with prolactin rather than LH being the major luteotropic factor (Okkens et al. 1990). Yet, luteal regression occurs in spite of an increased availability of prolactin (Gräf 1978) and LH (Hoffmann & Schneider 1993).
Recently, the role of prostaglandins in luteal regression in the nonpregnant bitch has been addressed (Hoffmann et al. 2004, Kowalewski et al. 2006a, 2008a). As hysterectomy does not interfere with normal ovarian function (Hoffmann et al. 1992), a role of CL-derived prostaglandin F2α (PGF2α) acting as a luteolytic agent via para-/autocrine mechanisms was postulated, also in analogy to the situation observed in other species, i.e. cattle (Diaz et al. 2002). However, the results obtained rather support a luteotropic effect of PGE2 during the first third of luteal lifespan (Kowalewski et al. 2008b) than a luteolytic role of PGF2α.
The observed luteolytic response of the canine CL to systemic application of PGF2α agonists (Concannon & Hansel 1977, Romagnoli et al. 1991) does not contradict this hypothesis as the CL rather constantly express the PGF2α receptor (PTGFR) following its formation around day 15 after ovulation (Kowalewski et al. 2008a).
In pregnant bitches for the immediate prepartal release of PGF2α, a clear role in respect to inducing myometrial contractions could be deduced (Nohr et al. 1993, Hoffmann et al. 1999). Indications on a luteolytic role are equivocal and only relate to the observations by Hoffmann et al. (1996), who reported about a somewhat prolonged pregnancy in two dogs following application of the cyclooxygenase inhibitor indomethacin.
The situation is further hampered by the fact that there are no data on the origin of the prepartally released PGF2α and the underlying mechanisms triggering this release.
Cyclooxygenase 2 (PTGS2) is the essential enzyme allowing for the formation of PGH2, the common precursor of PGF2α and PGE2. The reduction of PGH2 by 9,11-endoperoxidase reductase activity of prostaglandin F2α synthase (PGFS) is the major route for PGF2α synthesis leading to the synthesis of PGF2α directly from PGH2. However, alternatively PGF2α could also be formed via 9-ketoprostaglandin reductase or 11-ketoprostaglandin reductase activity using the PGH2-derived PGE2 and PGD2 as substrates respectively (Asselin & Fortier 2000, Madore et al. 2003).
As the canine PGFS cDNA sequence was originally cloned and described based on the mRNA isolated from utero/placental cross sections of late pregnant dogs (Kowalewski et al. 2008a), there is a distinct likelihood that the utero/placental unit could be the origin of the prepartal PGF2α release. This would resemble the situation in cattle, where strongly upregulated PTGS2 levels are observed during parturition in the uninucleated trophoblast cells indicating that the fetal part of the placenta (cotyledon) may be involved in the prepartal output of PGF2α and PGE2 in this species (Schuler et al. 2006).
Both prostaglandins show strong functional interrelationships; thus, PGE2 acts on softening the cervix prior to and during the uterine contractions, triggered by the action of PGF2α (Stys et al. 1981, Fuchs et al. 1984). Additionally, it has been suggested that PGE2 plays an essential role in controlling the differentiation of endometrial stromal cells into decidual cells (decidualization) and hence in implantation, as has been shown in mice (Pakrasi & Jain 2008).
Taking together the information available so far, it was hypothesized that in the dog placenta-derived prostaglandins might play a role in the prepartal preparation of the genital tract possibly also contributing to prepartal luteolysis. We therefore tested for the expression of PTGS2, PGFS, prostaglandin E2 synthase (PTGES), and the respective PGF2α and PGE2 receptors (PTGFR, PTGER2, and PTGER4). As the decidual cells of maternal origin are the only cells in canine placenta expressing the progesterone receptor (PGR; Vermeirsh et al. 2000), an immediate functional interrelationship between CL-derived P4 and placental function is suggested. Therefore, also expression of the PGR was assessed as well as the concentrations of P4 and 15-ketodihydro-PGF2α (PGFM), the major metabolite of PGF2α, in peripheral plasma. In order to gain further information on the underlying mechanism, the same parameters were assessed after application of the PGR blocker aglepristone to terminate pregnancy.
Normal pregnancy and parturition
Progesterone and PGFM concentrations
As was already shown for P4 (Kowalewski et al. 2009), mean progesterone concentrations were: 35.71±7.9 ng/ml in the pre-implantation period, 29.73±13.23ng/ml in the post-implantation period, 13.32±8.66ng/ml at mid-gestation, and 2.07±0.99 ng/ml during the prepartal progesterone decline; the effect of time was highly significant (P<0.0019; Fig. 1A). The course of the PGFM concentrations also showed a highly significant effect of time (P<0.0001) with highest values occurring during prepartal luteolysis revealing an inverse relationship in the course of P4 and PGFM concentrations (Fig. 1A). The mean PGFM concentrations were: 0.85±0.26 nmol/l in the pre-implantation period, 4.82±2.3 nmol/l in the post-implantation period, 5.44±2.2 nmol/l at mid-gestation, and 18.2±3.67 nmol/l during prepartal luteolysis (Fig. 1A).
Expression of PGR mRNA and immunohistochemical localization in the utero/placental unit
Expression of mRNA revealed a highly significant effect of time (P<0.0001). It was highest in the pre-implantation period (Fig. 2A) and decreased significantly (P<0.001) thereafter with no further changes observed towards the end of gestation (Fig. 2A). In the pre-implantation group, immunohistochemistry (IHC) located PGR to the endometrial stroma and to nuclei of both glandular and superficial epithelial cells, as well as in the smooth muscle cells of the myometrium (Fig. 3A). Following formation of the placenta, decidual cells were the only cells of the placental labyrinth strongly expressing the PGR (Fig. 3C). No signals could be detected in the endothelial cells of maternal and fetal vessels or in the trophoblast (Fig. 3C). No or only weak signals were seen in the maternal stroma and epithelial cells of deep endometrial glands and in the superficial endometrial glands referred to as glandular chambers of the utero/placental unit (not shown). The myometrial PGR signals remained detectable throughout pregnancy.
Expression of PTGS2 mRNA and immunohistochemical localization in utero/placental unit
Real-time RT-PCR revealed a significant (P<0.0005) effect of time; expression was low until prepartal luteolysis when a highly significant (P<0.001) increase in PTGS2 mRNA expression occurred (Fig. 4A). As observed by IHC, the endometrium of the pre-implantation group stained negative, while strong signals were observed in the myometrium at all stages of pregnancy (Fig. 3B).
Following formation of the placenta, a distinct staining for PTGS2 was localized to the invading trophoblast surrounding large maternal vessels at the base of the placental labyrinth; at the time of prepartal luteolysis, it had spread over the entire trophoblast (Fig. 3D). No staining was seen in the endothelial cells of maternal or fetal vessels or stromal cells. Only occasionally, some decidual cells showed a weak staining (Fig. 3D). With the beginning of implantation, some positive signals were also detected in the endometrial glands, especially in the deep endometrial glandular epithelium (not shown).
Expression of PGFS and PTGFR mRNA and localization in the utero/placental unit
Expression of PGFS and PTGFR mRNA showed a significant effect of time (P<0.0008 and P<0.02 respectively). Expression of PGFS mRNA was lowest during the pre-implantation period, followed by an increase (P<0.05) in the post-implantation and mid-gestation periods (Fig. 4C). It decreased thereafter until prior to parturition by ∼3.7-fold.
Expression of PTGFR mRNA revealed a biphasic expression pattern (Fig. 4E) and was similarly high in the pre-implantation period and during prepartal luteolysis, but distinctly lower in the post-implantation (P<0.05) and mid-gestation periods (Fig. 4E).
Cellular localization of PGFS and PTGFR mRNA by in situ hybridization (ISH) reflected localization of PTGS2. Virtually, all trophoblast cells stained positive during prepartal luteolysis, especially the invading trophoblast at the base of the placental labyrinth and in the area of the glandular chambers (Fig. 5A and B) where glandular epithelial cells were also positive.
Expression of PTGES, PTGER2, and PTGER4 mRNA and localization in the utero/placental unit
Expression patterns of PTGES and of the PGE2 receptors, PTGER2 and PTGER4, showed a significant effect of time (P<0.0001, P<0.0001, and P<0.05 respectively). PTGES mRNA showed the lowest expression during the pre-implantation period, a gradual increase until mid-gestation and was significantly elevated (P<0.001) during prepartal luteolysis (Fig. 6A).
The biphasic expression pattern of PTGER2 (Fig. 6C) with a significant upregulation during the pre-implantation period and prepartal luteolysis (P<0.001 and P<0.05 respectively) resembled that of PTGFR (Fig. 4E). Expression of PTGER4 mRNA was highest during the pre-implantation period and significantly lower (P<0.05) at prepartal luteolysis (Fig. 6E).
ISH located PTGES mRNA in the same cells as PGFS mRNA with strongest signals observed in the trophoblast cells (Fig. 5C). The same location was observed for the mRNA of PTGER2 and PTGER4 (not shown).
As parturition was induced around days 40–45, the expression of all factors assessed was compared with the mid-gestation group, which was used as a nontreated control in the statistical evaluation applying the Dunnett's multiple comparison test. For all factors, the observed changes in mRNA expression following application of the antiprogestin were in the same direction as those observed from mid-gestation to prepartal luteolysis. Thus, there was a significant upregulation of PTGS2 (P<0.001; Fig. 4B), PTGES (P<0.01), and PTGER2 (P<0.01; Fig. 6B and D). The increase of PTGFR (P>0.05) was not significant; all effects were more distinct at 24 h than at 72 h after treatment (Figs 4B and F and 6B and D).
Expression of PGFS and PTGER4 was decreased (P<0.01 and P<0.05 respectively) with the effect being more pronounced at 72 h than at 24 h after treatment (Figs 4D and 6F). No changes in the PGR mRNA expression were observed (Fig. 2B).
IHC (Fig. 3E and F) and ISH localized expression of all factors assessed in the same cell types as in nonantiprogestin-treated dogs.
Similarly, changes in PGFM and P4 plasma levels showed a significant effect of time (P<0.0047 and P<0.0163 respectively). The initial concentration of PGFM increased from 2.75±0.68 to 5.04±1.36 nmol/l at the second treatment with the antiprogestin. It was high 48 h later (9.33±1.1 nmol/l) and decreased by 72 h (4.97±1.5 nmol/l). P4 decreased significantly (P<0.01) from 15.11±6.7 ng/ml at the first treatment to 5.1±2.7 ng/ml and 1.2±0.6 ng/ml 24 and 72 h respectively after the second alizine treatment (Fig. 1B).
The data obtained clearly indicate that all major components of the prostaglandin system are expressed in the utero/placental unit of the pregnant dog. However, the quantitative data obtained by real-time RT-PCR do not allow a distinction between the uterus and/or the placenta (trophoblast) as the site of origin except for the pre-implantation period when the placenta is not yet formed. During this period, PTGS2, PGFS, and PTGES are expressed on a low level and IHC located PTGS2 solely to the myometrium. The relatively high expression of the PGF2α receptor PTGFR and of the PGE2 receptors, PTGER2 and PTGER4, may be seen as signs of a high responsiveness of the uterus to PGF2α and PGE2 during the pre-implantation period.
Also expression of the PGR was high during the pre-implantation period and IHC located nuclear staining predominantly to uterine epithelial cells. Thus, the responsiveness of the epithelium to P4 is maintained at least until implantation, securing the production of uterine milk (embryotrophe) and hence embryonic survival.
Following formation of the placenta, IHC localized expression of PTGS2 to the epithelial trophoblast cells; staining was initially restricted to the invading trophoblast surrounding the large maternal vessels at the base of the placental labyrinth but had spread over the entire trophoblast in the period of prepartal luteolysis, an observation accompanied by a dramatic increase in the expression of the respective mRNA and coinciding with a substantial prepartal increase of PGFM. This points towards a functional interrelationship and the role of placental PTGS2 as a rate-limiting factor in the provision of prepartal prostaglandins. Such a role of PTGS2 as a rate-limiting factor has also been observed in the horse (Boerboom et al. 2004), where blocking of endometrial PTGS2 expression by the conceptus at day 15 of early pregnancy prevented PGF2α-induced luteal regression allowing for continuation of pregnancy; along this line is the observation that neither PGFS nor PTGES – the essential downstream enzymes – were significantly upregulated both during the time of luteolysis in cyclic animals and during early pregnancy. Hence, an important PTGS2-related mechanism leading to the suppression of uterine luteolytic PGF2α output during pregnancy in a horse has been postulated (Boerboom et al. 2004).
Based on our data and concerning the canine placenta, this increase must mainly originate in the trophoblast cells as ISH co-localized expression of PGFS to these cells. The observation that PGFS mRNA expression is higher during the post-implantation and mid-gestation periods may indicate that the expression of PGFS is regulated at the post-transcriptional level resulting in the low PGFM levels detected in peripheral plasma. On the other hand, decreased mRNA levels observed at prepartal luteolysis could be indicative for enhanced substrate turnover due to the increased PTGS2 availability during this time. However, since no corresponding data on the PGFS expression at the protein level are yet available, no final conclusion can be drawn.
Though the reduction of PGH2 by 9,11-endoperoxidase reductase activity is the major route for PGF2α synthesis, it cannot be ruled out for sure that the formation of PGF2α is through the alternative pathway via 9-ketoprostaglandin reductase or 11-ketoprostaglandin reductase activity using PGH2-derived PGE2 and PGD2 as substrates respectively (Asselin & Fortier 2000, Madore et al. 2003). Such an alternative PGF2α-biosynthetic pathway has been identified in cattle (Madore et al. 2003) in which at least three closely related enzymes possessing the PGFS activity have been characterized within the AKR1C subclass of aldo-keto reductases family (reviewed by Madore et al. (2003)); none of them is, however, upregulated during the time of luteolysis. As reported by the same group, also another previously described enzyme 20α-hydroxysteroid dehydrogenase (AKR1B5) is capable of converting PGH2 into PGF2α and reveals an enhanced endometrial expression in cyclic cows around the time of luteolysis (Madore et al. 2003). Similarly, in the rabbit the aldo-keto reductase activity is associated with both the PGE 9-reductase and 20α-hydroxysteroid dehydrogenase activity (Wintergalen et al. 1995).
Following these observations, further studies concerning the alternative PGF2α-biosynthetic pathways in canine placental compartment should be considered.
Similar to PGF2α, PGE2 also seems to originate from the trophoblast cells as ISH located mRNA expression of PTGES to these cells, with the expression being significantly elevated during prepartal luteolysis. As there are no data on the expression on the protein level and PGE2 plasma concentration, this observation must be interpreted very carefully and only seen as a hint that the increased availability of PTGS2 during prepartal luteolysis is associated with a basic increased capacity to release PGE2 by the placental compartment during prepartal luteolysis. Within the utero/placental unit, the trophoblast seems to be the major target for PGF2α and PGE2 as ISH localized their receptors primarily to these cells.
Interestingly, both PTGFR and PTGER2 mRNA were upregulated during prepartal luteolysis, while PTGER4 mRNA was downregulated. This allows the conclusion that prepartal trophoblast function is affected by PGF2α and PGE2, triggering different signalling cascades possibly in relation to the mechanisms allowing for final placental maturation and release. The expression of PGE2 receptors, EP1 and EP3, was either not detectable or only very weak in the canine corpus luteum (Kowalewski et al. 2008b) pointing towards its organ/species-specific expression as has also been shown for the mouse CL (Segi et al. 2003). Hence, our conclusions on the role of PGE2 in regulation of canine placenta function would require further information on the expression of these receptors in the canine uterine/placental compartment.
Taken together, our data suggest a paracrine/autocrine role of the PTGS2-derived arachidonic acid metabolites in the processes of canine decidualization and placentation. Such an autocrine/paracrine effect of PGF2α has been previously proposed in bovine placentomes from early to late gestation (Arosh et al. 2004). Similarly, PGE2 has been shown to be involved in the autocrine/paracrine regulation of decidualization in mice (Pakrasi & Jain 2008). During the later stages of pregnancy and prepartal luteolysis, they might contribute to the final maturation and release of the placenta.
In addition, the increased availability of PGE2 in the prepartal period could be utilized as a substrate for PGF2α synthesis (Asselin & Fortier 2000, Madore et al. 2003); further studies are, however, necessary to support this conclusion.
As we have recently reported (Kowalewski et al. 2009) and except for the upregulation of PTGER2 mRNA, the expression patterns of luteal STAR, HSD3B, PTGS2, PGFS, PTGES, PTGER4, and of the PGR observed during antiprogestin (aglepristone)-induced luteolysis resemble those of physiological luteal regression. Hence, the hypothesis was forwarded (Kowalewski et al. 2009) that luteal P4 acts as an autocrine/paracrine factor within a positive loop feedback mechanism involving STAR and HSD3B, and that luteal regression is not an actively controlled but rather permissive process related to aging of the CL as observed by electron microscopy (Hoffmann et al. 2004).
Following treatment with aglepristone on day 58 of pregnancy, Baan et al. (2008) observed an increase of PGFM parallel to a decrease of P4 levels. Our studies with aglepristone treatment between days 40 and 45 of pregnancy confirm this observation. Interestingly, also in the utero/placental unit, all changes observed concerning mRNA expression of the prostaglandin system after treatment with the antiprogestin aglepristone were in the same direction as observed during prepartal luteolysis. As the PGR is only expressed by the maternal stroma-derived decidual cells as observed by Vermeirsch et al. (2000) and confirmed in this study, these cells must have an important signalling function in respect to induction of luteolysis and prepartal preparation of the utero/placental unit for parturition.
The induction of prepartal luteolysis might then be brought upon by the fact that CL-derived P4 levels reach a critical lower threshold during the course of luteal regression affecting the crosstalk between the maternal decidual cells and the fetal trophoblast, leading to prepartal activation of the prostaglandin system as evidenced by upregulation of PTGS2 during prepartal luteolysis.
Our results also provide an explanation of the observation that aglepristone-induced parturition in the dog may require or may not require ecbolic support (Hoffmann et al. 1999, Baan et al. 2005). As provision of adequate amounts of PTGS2 seems to be restricted to the immediate phase prior to parturition, the time point of treatment with the antiprogestin seems to be the critical issue.
In summary, these observations point to an essential role of P4 not only in respect to maintenance of pregnancy by securing uterine quiescence, but also by acting as an important autocrine/paracrine and also endocrine factor in respect to luteal regression and luteolysis.
Materials and Methods
Animals and tissue samples
Animal experiments were performed in accordance with animal welfare legislation (permit no. II 25.3-19c20-15c GI 18/14 and VIG3-19c20/15c GI 18,14 (Gießen) and permit no. Ankara 2006/06 (Faculty of Veterinary Medicine, University of Ankara)), and all utero/placental tissue samples were obtained together with the collection of CL as described earlier (Kowalewski et al. 2009).
Eighteen clinically healthy pregnant bitches of different breeds (aged 2–8 years) were divided into four groups, and utero/placental unit samples were collected via ovariohysterectomy (OHE) on one of the following days of pregnancy:
Group 1; pre-implantation, days 8–12, n=5.
Group 2; post-implantation, days 18–25, n=5.
Group 3; mid-gestation, days 35–40, n=5.
Group 4; prepartal, progesterone decline, n=3.
In group 4, P4 concentrations were monitored every 6 h beginning on day 58 of pregnancy, and OHE was performed when P4 continued to decrease below 3 ng/ml in two consecutive measurements.
Ten bitches were treated with the antiprogestin aglepristone between 40 and 45 days of pregnancy using the dose recommended for induction of abortion (10 mg/kg bw, 2 times 24 h apart). OHE was performed 24 h (n=5) and 72 h (n=5) after the second injection.
Tissue preservation and determination of P4 and PGFM
Immediately after their removal, tissue samples were washed with PBS, trimmed off the surrounding connective tissue, immersed for 24 h in RNAlater (Ambion Biotechnologie GmbH, Wiesbaden, Germany), and then stored at −80 °C until further use. For IHC and ISH, tissue samples were fixed for 24 h in 10% neutral phosphate-buffered formalin, washed with PBS, dehydrated in a graded ethanol series, and embedded in paraffin-equivalent Histo-Comp (Vogel, Giessen, Germany).
Established RIA procedures were used for the assay of peripheral blood plasma progesterone (Hoffmann et al. 1973) and the prostaglandin F2α metabolite PGFM (Kindahl et al. 1976, Granström & Kindahl 1982).
RNA extraction, semi-quantitative real-time (TaqMan) PCR, and data evaluation
Total RNA was isolated from utero/placental unit samples using TRIzol Reagent (Gibco-BRL, Life Technologies) and treated with DNase I recombinant, RNase-free, to eliminate genomic DNA contaminations (Roche Molecular Biochemicals) according to the manufacturer's instructions.
The primers were ordered from MWG Biotech AG; the 6-carboxyfluorescein (6-FAM) and 6-carboxytetra methylrhodamine (TAMRA)-labeled probes were from Eurogentec, B-4102 Seraing, Belgium. Semi-quantitative real-time PCR experiments and the comparative Ct method (ΔΔCt method) for the relative quantification of genes were performed as previously described (Kowalewski et al. 2006a). The results are expressed as the fold change in gene expression over the calibrator.
Sequences for primers and TaqMan probes were as follows:
PTGS2 (forward): 5′-GGA GCA TAA CAG AGT GTG TGA TGT G-3′;
PTGS2 (reverse): 5′-AAG TAT TAG CCT GCT CGT CTG GAA T-3′;
PTGS2 (TaqMan probe): 5′-CGC TCA TCA TCC CAT TCT GGG TGC T-3′;
PTGES (forward): 5′-CTG TCA TCA CCG GCC AAG T-3′;
PTGES (reverse): 5′-CCT GGT CAC TCC GGC AAT A-3′;
PTGES (TaqMan probe): 5′-ACG CCC TGA GAC ACG GAG GCC T-3′;
PTGER2 (forward): 5′-CAC CCT GCT GCT GCT TCT C-3′;
PTGER2 (reverse): 5′-CGG TGC ATG CGG ATG AG-3′;
PTGER2 (TaqMan probe): 5′-TGC TCG CCT GCA ACT TTC AGC GTC-3′;
PTGER4 (forward): 5′-AAA TCA GCA AAA ACC CAG ACT TG-3′;
PTGER4 (reverse): 5′-GCA CGG TCT TCC GCA GAA-3′;
PTGER4 (TaqMan probe): 5′-ATC CGA ATT GCT GCT GTG AAC CCT ATC C-3′;
PGFS (forward): 5′-AGG GCT TGC CAA GTC TAT TGG-3′;
PGFS (reverse): 5′-GCC TTG GCT TGC TCA GGA T-3′;
PGFS (TaqMan probe): 5′-TCC AAC TTT AAC CGC AGG CAG CTG G-3′;
PTGFR (forward): 5′-ACC AGT CGA ACA TCC TTT GCA-3′;
PTGFR (reverse): 5′-GGC CAT CAC ACT GCC TAG AAA-3′;
PTGFR (TaqMan probe): 5′-CAT GGT GTT CTC CGG TCT GTG CCC-3′;
PGR (forward): 5′-CGA GTC ATT ACC TCA GAA GAT TTG TTT-3′;
PGR (reverse): 5′-CTT CCA TTG CCC TTT TAA AGA AGA-3′;
PGR (TaqMan probe): 5′-AAG CAT CAG GCT GTC ATT ATG GTG TCC TAA CTT-3′;
GAPDH (forward): 5′-GCT GCC AAA TAT GAC GAC ATC A-3′;
GAPDH (reverse): 5′-GTA GCC CAG GAT GCC TTT GAG-3′;
GAPDH (TaqMan probe): 5′-TCC CTC CGA TGC CTG CTT CAC TAC CTT-3′.
Owing to the restricted availability of canine-specific and/or cross-reacting antibodies, the immunohistochemical detection of target genes was limited to PTGS2 and the PGR.
Utero/placental cross sections were cut (4 μm thick) and mounted on SuperFrost Plus microscope slides (Menzel-Gläser; Braunschweig, Germany). The standard immunoperoxidase detection method applied was as previously described (Kowalewski et al. 2006a). The antibodies used were: monoclonal mouse anti-rat PTGS2 IgG, clone 33, BD Pharmingen (Heidelberg, Germany) and mouse mAb generated against amino acids 922–933 (C-terminal) of the human PGR, clone 10A9, Immunotech, Hamburg, Germany. An isotype-specific irrelevant mAb IgG1 (Dianova, Hamburg, Germany) was used as a negative control. The biotinylated horse anti-mouse IgG BA2000 was from Vector Laboratories Inc., Burlingame, CA, USA. The peroxidase activity was detected with NovaRed substrate kit according to the manufacturer's instructions (Vector Laboratories). Finally, sections were counterstained with hematoxylin and mounted with Histokit (Assistant, Osterode, Germany).
In situ hybridization
The following primers derived from canine-specific gene sequences were used in RT-PCR to generate templates for subsequent cRNA probe synthesis:
PGFS (forward): 5′-GAT CTC TGT GCC ACA TGG GAG-3′;
PGFS (reverse): 5′-TGG GTC CTT CAG GAG AAC TGG-3′;
PTGFR (forward): 5′-TGT GCC CAC TTT TTC TAG GC-3′;
PTGFR (reverse): 5′-TCT TCC CAG TCT TTG ATG TG-3′;
PTGES (forward): 5′-ACC ATC TAC CCC TTC CTG T-3′;
PTGES (reverse): 5′- CTG CTT CCC AGA CGA TCT -3′;
PTGER2 (forward): 5′-TTC TCC TGG CTA TTA TGA CC-3′;
PTGER2 (reverse): 5′-ATC TAC TGG CGT TTG ACT G-3′;
PTGER4 (forward): 5′-GGT ACG GGT GTT CAT CAA C-3′;
PTGER4 (reverse): 5′-AGA GGA GGG TCT GAG ATG TG-3′.
The length of the amplicons was: 235, 291, 214, 273, and 323 bp for PGFS, PTGFR, PTGES, PTGER2, and PTGER4 respectively.
The further procedure was as previously described (Kowalewski et al. 2006b). Briefly, the PCR products were cloned into the pGEM-T plasmid (Promega). Digestion of the harvested pGEM-T plasmid clones containing the respective inserts was by using the restriction enzymes NcoI (antisense cRNA) and Not (sense cRNA; New England Biolabs, Frankfurt, Germany). The synthesis of digoxigenin (DIG)-labeled cRNA was achieved with DIG RNA labeling kit from Roche Molecular Biochemicals. Semi-quantitation of the DIG-labeled cRNA was by dot blot analysis of serial dilutions of cRNA probes on a positively charged Nylon Membrane (Roche Molecular Biochemicals). Paraffin-embedded utero/placental cross sections were mounted on the SuperFrost Plus slides (Menzel Glaeser, Braunschweig, Germany), dewaxed, digested with 70 μg/ml proteinase K (Sigma–Aldrich Chemie GmbH), and postfixed with 4% paraformaldehyde and then processed according to the procedure by Lewis & Wells (1992) and Klonisch et al. (1999). Detection of the DIG-labeled cRNA probes was by alkaline phosphatase-conjugated sheep anti-DIG Fab fragments (Roche Molecular Biochemicals) diluted 1:5000 in 1% ovine serum followed by incubation with the substrate 5-bromo-4-chloro-3-indolyl phosphate in the presence of nitroblue tetrazolium (Roche Molecular Biochemicals).
To test for an effect of time in the respective group, a parametric one-way ANOVA was applied. Multiple comparison post-tests were performed in case of P<0.05. Those were Tukey–Kramer multiple comparison test in experiments measuring the expression of target genes during the course of pregnancy and Dunnett's multiple comparison test in experiments showing the expression of genes after aglepristone-induced luteolysis. In the latter, the results present the fold change in gene expression compared to its expression at mid-gestation. Owing to the uneven distribution of the real-time data obtained for the PTGS2, PGFS, and PTGER4, data were analyzed using the Kruskal–Wallis test (a nonparametric ANOVA) followed by Dunn's multiple comparison test. Numerical data were presented as the mean±s.d. For all tests, the statistical software program, GraphPad 3.06 (GraphPad Software, Inc., San Diego, CA, USA) was used.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
This study was funded with support of the German Research Foundation. The authors also greatly appreciate the support of Virbac, Germany, in providing the aglepristone and financial support of the aglepristone associated studies.
MadoreEHarveyNParentJChapdelainePAroshJAFortierMA2003An aldose reductase with 20α-hydroxysteroid dehydrogenase activity is most likely the enzyme responsible for the production of prostaglandin F2α in the bovine endometrium. Journal of Biological Chemistry27811205–11212.
SchulerGTeichmannUKowalewskiMPHoffmannBMadoreEFortierMAKlischK2006Expression of cyclooxygenase-II (COX II) and 20a-hydroxysteroid dehydrogenase (20a-HSD)/prostaglandin F-synthase (PGFS) in bovine placentomes: implications for the initiation of parturition in cattle. Placenta271022–1029.
SegiEHaraguchiKSugimotoYTsujiMTsunekawaHTambaSTsuboiKTanakaSIchikawaA2003Expression of messenger RNA for prostaglandin E receptor subtypes EP4/EP2 and cyclooxygenase isozymes in mouse periovulatory follicles and oviducts during superovulation. Biology of Reproduction68804–811.
H B Beceriklisoy is now at Clinic for Obstetrics and Reproductive Diseases, Faculty of Veterinary Medicine, University of Adnan Menderes, 09100 Aydin, Turkey
C Pfarrer is now at Department of Anatomy, University of Veterinary Medicine, 30173 Hannover, Germany