Background
Female domestic cats (
Felis catus) are traditionally classified as seasonally polyestrous with ovulation provoked by coitus [
1,
2]. However, in as many as 50% of domestic cats, ovulation occurs without cervical or vaginal stimulation [
3,
4]. If ovulation is not followed by pregnancy, the queen enters pseudopregnancy during which the corpora lutea (CL) remain active. Hence, the reproductive cycle in the domestic cat differs significantly from that in ungulates [
5,
6] and dogs [
7,
8], in which spontaneous ovulation is followed by the formation of CL in each estrous cycle. In the cat [
9], rabbit [
10] and rat [
11], the luteal phase of non-pregnant animals lasts only one-half of a normal gestation. This means elevated progesterone (P
4) levels measured in circulating blood for 35–40 days in pseudopregnant and for approximately 60–65 days in pregnant queens. In contrast, in the dog [
12], mink [
13] and ferret [
14], the length of the luteal phase is similar in the presence or absence of pregnancy. Progesterone levels in pregnant and non-pregnant dogs are similar within the first 60 days after ovulation [
15]. Moreover, there are no indications of an active luteolytic principle in non-pregnant dogs, and, the canine CL appears to be devoid of PGF
2α-synthase (PGFS) activity [
16,
17]. Consequently, luteal regression in non-pregnant bitches seems to be a passive degenerative process in the absence of endogenous PGF
2α[
16,
17]. The domestic cat thus seems to have a reproductive advantage over some other carnivores because its shorter luteal phase allows for an earlier return to estrual cyclicity. In the cat, ovarian activity returns within 7 to 10 days following pseudopregnancy [
10]. The histological and endocrine patterns of early CL formation are the same in pregnant and pseudopregnant cats until days 10–12 after coitus [
18]. However, after days 12–13, which coincides with the time of implantation [
19,
20], plasma P
4 concentration slowly declines in pseudopregnant queens, whereas during pregnancy it remains on a plateau until day 30 and then gradually declines towards the end of gestation [
21].
Until the 1970s, the CL was considered to be the main source of P
4 during feline gestation. However, the finding that ovariectomy in queens after day 45 either did not interrupt pregnancy [
22] or only did in some animals [
23] supported an assumption that the placenta is a source of steroidogenic enzymes [
18]. Subsequently, feline placental homogenates were found to be capable of synthesizing P
4[
19]; however, the expression of steroidogenic factors or P
4 content has not been examined in the placenta to date.
Cholesterol is the precursor of all steroid hormones. Steroidogenesis is regulated by steroidogenic acute regulatory (StAR) protein, which controls cholesterol transport from the outer to the inner mitochondrial membranes of the steroidogenic cells [
24]. Conversion of pregnenolone to P
4 is a step catalyzed by 3β- hydroxysteroid dehydrogenase/isomerase (3βHSD). In this study, expression of these two factors was investigated in the feline CL and placenta throughout pregnancy and pseudopregnancy.
Discussion
The present results confirm that the placenta is an additional source of P4 in pregnant queens, possibly acting as an important endocrine organ during pregnancy. Both StAR and 3βHSD were immunolocalized in the placenta and CL in each luteal phase. In the present study, the luteal 3βHSD mRNA expression patterns in pregnant and pseudopregnant animals were similar, both peaking during the mid-luteal phase. However, luteal mRNA was 2.5-fold higher in pregnant than in pseudopregnant queens. Luteal StAR mRNA remained at a relatively constant low level in pseudopregnant animals, but followed the pattern of 3βHSD mRNA expression in pregnant animals. However, no quantitative assessment was performed for the expression of StAR and 3βHSD at the protein level.
A rapid development of the P
4-producing CL is observed in both pregnant and pseudopregnant cats. Plasma P
4 levels are about the same in pregnant and pseudopregnant queens in the first 10–12 days after coitus [
18] but their profiles diverge as early as on day 12 or day 13 after coitus, which is the period of implantation in the cat [
18,
20]. The luteal
StAR- and
3βHSD- mRNA were highest during the mid-phase of pregnancy. Additionally, increased
3βHSD-levels were also observed at the mid-phase of pseudopregnancy. The high expression of both factors during both mid-luteal phases may be responsible for the strongly elevated P
4 levels in the circulating blood reported previously [
2,
18,
21,
37] and observed also in the present study. As reported previously, P
4 values rose rapidly in pseudopregnant queens between Day 4 and Day 9, peaking around Day 14 post-ovulation with a plateau between Day 9 and Day 23. The range of serum P
4 concentrations at that time was 30 ng/mL to more than 87 ng/mL [
2]. The present results concur with data reported earlier, in which strong individual variations in plasma P
4 concentrations were observed. In one queen experiencing a nongravid luteal phase, P
4 peaked at 88 ng/mL,
vs. a range of 20.1-28.7 ng/mL observed in other animals. The mean value of plasma P
4 observed in queens at the mid-luteal stage was 39.9 ng/mL and is consistent with the those reported in the literature [
2]. However, in another study, P
4 values in pseudopregnant cats were shown to peak at levels of 17 ng/mL on Day 18 postcoitum, followed by a gradual decline to values < 1 ng/mL between Days 30–46 after coitus [
25]. In pseudopregnant queens at the late luteal stage, we usually observed distinctly diminished P
4 values compared to the mid luteal phase. Wildt and coworkers [
2] reported that the serum P
4 values were lower than 1 ng/mL by Day 42 post ovulation, however, CL remnants were visually evident throughout the following interestrus period. In the present study, the P
4 values in plasma reached approximately 4.2 ± 1.2 ng/mL. However, blood samples were collected from pseudopregnant queens only from Day 25 to Day 35 (Day 1 signified the first day of pseudopregnancy and was consistent with silencing of estrus behavior), so this difference might explain the discrepancy in observed P
4 values.
During mid-pregnancy, the serum P
4 values varied significantly between individual cats (range 34 – 74.4 ng/mL), with a mean P
4 concentration of 46.5 ng/mL. Although P
4 concentrations in the blood of mid-pregnant queens were numerically higher than in non-pregnant animals in the mid-luteal phase, no statistical differences were seen in the present study between these two groups. It cannot be ruled out that these results might differ significantly if the number of females in each group were increased; however, it should also be mentioned that the present data are very similar to those reported by others (for review see: 2, 37). Distinct individual variations in the plasma P
4 levels collected from pregnant cats were previously reported [
37],
i.e., peak serum P
4 from 13.5 to 57 ng/mL extending over Days 11 to 60 in individual females. After peripheral P
4 peaked, a gradual decline began, usually at Day 44 in most of the queens that were examined. The P
4 levels were maintained at a relatively constant level until a few days before parturition, when they sharply decreased [
37]. In the present study, serum P
4 decreasing to values lower than 10 ng/mL (range 3.8 – 9.2 ng/mL) was observed in queens in the 7
th and 8
th week of pregnancy; however, in three individuals at the 6
th and 7
th week of pregnancy, the P
4 values were still near 30 ng/mL. Summarizing these data, it might be concluded that, besides visible differences existing in plasma P
4 levels between individuals, the distinct decrease in peripheral P
4 precedes onset of parturition.
The successful extraction of P
4 from the placenta further supports its role as an additional source of P
4 during pregnancy in the cat. Placental P
4 concentrations seem to be dependent on gestational age. Thus, the higher P
4 values reported for pregnant queens may result from P
4 supplementation by placental tissue, as hypothesized previously [
18]. Still remaining, however, is the question whether the placenta-originating P
4 would be alone able to maintain pregnancy in domestic cats ovariectomized after 45 days of pregnancy. In contrast to cats, in mice it was proven that ovariectomy at any time of pregnancy causes abortion [
38]. It was thought that P
4 synthesis is not possible
de novo in the rodent placenta. Nevertheless, trophoblast giant cells from mouse placenta were shown to produce P
4 from cholesterol [
39]. Moreover, recent studies using molecular tools have shown that P
4 synthesis is possible in rodent maternal decidual cells and occurs upon decidualization, but is terminated at mid-gestation [
40]. It is further assumed that this local synthesis of P
4 may act as an immunosuppressive factor at the implantation sites [
41]. In contrast in cows placenta-derived progesterone and oestrone are supposed to be auto- or paracrine factors involved in placental growth and differentiation [
42]. The synthesis of P
4 was recently demonstrated in uninucleate trophoblast cells and the trophoblast giant cells of the bovine placenta; however, these two kinds of cells were reported to have different steroidogenic capacities [
43].
The morphological comparative study in the epitheliochorial and endotheliochorial placenta types published by Leiser and coworkers in 1998 [
44] presents a photograph of the typical labyrinthine-like system in the cat placenta with very easily visible and differentiated decidual cells [
44]. In the present study, anti-vimentin staining was applied in order to distinguish between cells of mesenchymal origin and other sources, as described and confirmed for the canine and feline placenta by Bezler [
34]. Since the StAR- and 3βHSD-positive placental cells were identified as maternal decidual cells, and were found mostly in placentas from the second half of gestation, the present findings on the spatio-temporal expression of the key steroidogenic genes differ from data obtained for other species. Even though some 3βHSD-positive cells were observed in the trophoblast, the strongest signals were present in decidual cells suggesting that these cells may be the main source of steroid synthesis within the feline placenta. Although the reproductive biology characteristics of the dog and cat are often compared, especially on this point distinct differences are observed, because in dogs the CL is the only source of P
4 during both pseudopregnancy and pregnancy [
7].
Acknowledgements
This study was supported by Grants-in-Aid for Scientific Research from the Polish Ministry of Scientific Research and High Education (IP2010 037570 and IP2011 048971) and was partially financed from the Institute of Veterinary Anatomy, University of Zurich, Zurich, Switzerland. Authors are indebted to Dr Ian J. Mason, Centre for Reproductive Biology, University of Edinburgh, for donation of rabbit immunoprotein against recombinant human type II 3βHSD (R1484); Dr Douglas M. Stocco, Department of Cell Biology and BiochemistryTexas Tech University Health Sciences Center, Lubbock, Texas, for donation of rabbit immunoprotein against StAR; Dr Stanislaw Okrasa, University of Warmia and Mazury, Olsztyn for donation of P4-antibodies, and to Elisabeth Högger and Urs Büchler, Institute of Veterinary Anatomy, Vetsuisse Faculty, University of Zurich, Zurich, Switzerland for their excellent technical help, and Dr. Barry Bavister for help with editing this manuscript.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MJS conceived of the study, and participated in its design, carried out molecular genetic studies and drafted the manuscript. EJ and AZS carried out the immunoassays and participated in molecular studies. AB and DJS helped to draft the manuscript. MPK participated in the design of the study and interpretation of the results, carried out the molecular and genetic studies, provided the knowledge- and methods-transfer and helped to draft the manuscript. The study was partially financed from the Institute of Veterinary Anatomy, University of Zurich, Zurich, Switzerland. All authors read and approved the final 552 manuscript.