Introduction

The foetal pituitary-adrenal axis is the route by which parturition is initiated in cattle [4]. In late gestation, ACTH from the foetal pituitary stimulates the foetal adrenals to produce increased amounts of cortisol. This increase induces synthesis of placental 17α-hydroxylase and aromatase and increases production of oestrogen at the expense of progesterone [2]. Other steroids, like the synthetic cortisol analogue dexamethasone, can induce placental 17α-hydroxylase and aromatase in pregnant cattle in a similar way [13, 18]. The subsequent decrease in progesterone production together with increased levels of oestrogens and induction of endometrial cyclooxygenase-2 (COX-2) synthesis, prepare the uterus for parturition.

Prostaglandin F2α has this far been shown to be the major luteolytic hormone produced by the bovine endometrium (for a review see [15]). The release of PGF2α at luteolysis in the oestrous cycle has been studied [3] and found to be released into the uterine veins in an "on-off " fashion [10]. Each pulse lasts about 4 h and is followed by a period of several h with only basal PGF2α release. This pulsatile release continues during luteolysis as the progesterone falls to basal levels. After luteolysis, the PGF2α pulsatility ceases. In cattle reaching the end of pregnancy progesterone is produced mainly by the corpus luteum and parturition does not occur until this progesterone production has ceased [12, 8]. The aim of this experiment was to study the profile of the PGF2α metabolite, 15-keto-13,14-dihydro-PGF2α, and to relate this to cortisol and progesterone levels as well as clinical findings during the period after induction of parturition by dexamethasone in cattle.

Materials and methods

Animals

In this study 4 late pregnant heifers (3 of the Swedish red and white breed (A, B and C) and one of the Swedish black and white breed (D)) were used. The animals were divided into 2 groups (A and B in the first group and C and D in the second) according to expected date of calving. All heifers were examined clinically and found healthy. Rectal palpation was used for pregnancy diagnosis. At the clinic, the heifers were fed according to Swedish standards [19].

The local ethical committee approved the study.

Induction of parturition

Twenty mg dexamethasone (Vorenvet® vet 1 mg/ml, BI-vet, Malmö, Sweden) was injected twice intramuscularly at a 24 h interval. The injections were made 3 to 4 weeks before expected calving (days 254–265).

The experimental period was divided into 4 phases, I to IV. Phase I started at the first dexamethasone injection and ended with either rupture or the first view of the foetal membranes. The subsequent phase II ended with the first sight of any part of the foetus and was followed by phase III. This phase ended at the final expulsion of the calf. Phase IV ended 12 h after parturition.

Blood sampling

Blood was collected via a polyurethane cannula (Cook central venous catheter, Cook, Brisbane, Australia) inserted 24 h before the first injection of dexamethasone. After cutaneous infiltration of local anaesthetics (Lidocain, Astra, Södertälje, Sweden) and a stab incision in the superficial skin, the catheter was inserted in the V. jugularis externa. Samples were collected once per hour from 2 h before the first injection of dexamethasone and until the start of parturition (phase I). During phase II, blood samples were collected at 10 min intervals. As soon as any part of the calf was visible from the outside, the sampling interval was changed to 5 minutes (phase III) and this sampling interval continued until at least 15 min after the calf was born. After parturition (phase IV), samples were collected once per hour for 12 h. The blood was transferred both to glass tubes containing NaHeparin (Venoject, Terumo, Leuven, Belgium) and to glass tubes containing NaEDTA with addition of 2000 KIE of Aprotinin (Trasylol® 10000 KIE/ml, Bayer, Göteborg, Sweden). The tubes were agitated and centrifuged for 10 min at 1000 × g (3000 rpm). Plasma was stored at -20°C until analysis.

Samples for analysis of progesterone were selected as follows: one sample every 8th hour until the day of luteolysis, then 1 sample every 4th h. From 12 h before parturition 1 sample every hour, and after parturition had started (phases II and III) 1 sample every 30th min. During phase IV, 1 sample per hour was selected.

Samples for cortisol analysis were selected as follows: a set of 5 consecutive samples, 1 per hour, were analysed. Twelve h after the first sample another set of five samples were analysed. This continued until 12 h before parturition. Then samples were selected once per h until the start of parturition. During phase II, samples were selected every 20th min and during phase III, one sample every 10th minute was selected. During phase IV, 1 sample per hour was selected.

Analytical methods

15-Ketodihydro-PGF2α was analysed using a radioimmunoassay [6]. Heparin plasma was used for the analysis and all samples were analysed in duplicates. The sensitivity of the method was 30 pmol/L. The intra-assay coefficients of variation ranged between 6.6% and 11.7% for the different ranges of the standard curve and the inter-assay coefficient of variation was 14%.

Heparin plasma was used for analysis of progesterone. This was done by the use of a solidphase radioimmunoassay technique (Coat-A-Count Progesterone, Diagnostic Products Corporation, Los Angeles, CA, U.S.A.). The sensitivity of the assay was 0.1 nmol/L. The intra-assay coefficients of variation for 3 control samples (2.6 nmol/L, 21.9 nmol/L and 53.1 nmol/L) assayed in duplicates in 20 assays were 11.9%, 5.8% and 7.0%, respectively. The inter-assay coefficients of variation were 12.6%, 12.1% and 13.3%, respectively.

For the cortisol analysis, EDTA plasma was used with an addition of Trasylol. Cortisol concentrations were determined directly by a rapid EIA in 20 μl plasma diluted 1:40 without prior extraction [14]. The cross-reactivities for the method are as follows: cortisone 45%, corticosterone 15%, desoxycorticosterone 8%, progesterone 8% and testosterone 3%. Parallelism between standards and unknowns in plasma were demonstrated for the range between 8 and 44 nmol/L plasma. The intra- and interassay coefficients of variation were 8.9% and 12.6%, respectively.

Statistical methods

For determination of the cortisol baseline a method was used that calculated the mean value of the base line after removal of all high values. Cortisol levels were judged as elevated when they exceeded 2 standard deviations above this mean value. Mean values and standard deviations were calculated by use of Minitab for Windows 95, release 12 (Minitab inc. PA, U.S.A.). Initial levels of progesterone and PGF2α metabolite are calculated as the mean and standard deviation of the first 5 and 10 samples, respectively. PGF2α metabolite levels during luteolysis are calculated as mean and range of the values during the period when progesterone levels decline most rapidly. Start and end of luteolysis are defined as the last progesterone value before onset and the first progesterone value after the end of luteolysis.

Results

Clinical observations

Clinical results in individual animals are shown in table 1. Parturition took place 7.7 (6.6–8.9) days (mean (range)) after the first dexamethasone injection in the 4 heifers. The parturitions were uneventful in 3 (A, B and C) of the heifers. In heifer D, gentle traction of the calf was applied during the last part of the second stage of labour. All 4 heifers delivered healthy calves of normal size. In 2 of the heifers (A and D), the foetal membranes were retained after parturition (RFM).

Table 1 Clinical data after induction of parturition by intramuscular injection of dexamethasone to four late pregnant heifers.
Full size table

Prostaglandin metabolite

The levels of PGF2α metabolite before first dexamethasone injection ranged from 150 to 300 pmol/L in all animals. After injection, the PGF2α metabolite levels showed two different kinds of patterns. In heifers B and D (see Fig.), the PGF2α metabolite levels continuously increased from the time of dexamethasone injection until parturition. In heifers A and C, however, the levels of PGF2α metabolite started to increase initially as for heifers B and D, but after a few days the levels declined to levels similar to the pre-experimental levels. In heifers A and C, the PGF2α metabolite levels then started to increase a second time and this increase continued until parturition. The nadir of this decrease appeared at three and four days before parturition in heifers A and C, respectively. Luteolysis occurred in all animals during the final increase of PGF2α. During this period the PGF2α metabolite levels increased rapidly but showed no signs of a pulsatile release. Mean values of PGF2αmetabolite during luteolysis are shown in table 2. The high PGF2αrelease during parturition was prolonged in heifers A and D relatively to B and C, and this prolongation corresponded to an increased length of phases II and III. The peak value of PGF2α metabolite at parturition was lower in heifers A and D than in heifers B and C. Immediately after foetal expulsion, the levels of PGF2α metabolite declined rapidly in all heifers. In A and D (RFM heifers), however, the quick decline soon was interrupted by a new period of increasing PGF2α metabolite levels. The post-partal levels in these animals were as high as during parturition. The post-partal increase was absent in heifers B and C (non-RFM heifers).

Table 2 Changes in PGF2α metabolite and cortisol levels after induction of parturition by intramuscular injection of dexamethasone to 4 late pregnant heifers.
Full size table

Progesterone

Progesterone levels at the time of dexamethasone injection were 12–18 nmol/L in all heifers. Luteolysis occurred during a period of time starting at 1.3 ± 0.3 and ending at 0.6 ± 0.1 days (mean ± SD) before parturition. After luteolysis, the progesterone levels remained elevated (1–2 nmol/L) until parturition. The progesterone profile around parturition is shown in the figure (inserted panels). After parturition, progesterone levels remained slightly elevated throughout the experiment in A and D (RFM). In B and C (non-RFM), the levels declined to levels below the sensitivity of the assay after the expulsion of the placenta.

Cortisol

Cortisol showed a basal level of 5.6–7.5 nmol/L in all heifers (Fig. 1 and Table 2) during the initial part of the experiment. In heifers A, B and D the variation of the cortisol levels was low during the period preceding luteolysis. In heifer C, the cortisol levels during this period were undulating with an interval between the peaks of about three days. The levels increased markedly immediately before and peaked during parturition in heifers A and D (RFM). Also in heifer C, there was a slight increase in cortisol levels but this was more pronounced immediately after parturition than during parturition. In heifer B, no such elevations could be seen. The cortisol increase started at 6.5 h and 4.3 h before parturition in heifers A and D, respectively. After parturition the levels declined rapidly in all animals.

Figure 1
figure 1

15-Ketodihydro-PGF2α (solid line), cortisol (dashed line) and progesterone (solid circles) profiles in four heifers after induction of parturition by intramuscular injections of dexamethasone in late pregnancy. Parturition takes place at day 0. Progesterone levels during parturition are shown in the small figure. Arrows indicate time for injection of dexamethasone. (Note – Logarithmic scale for the prostaglandin F2α metabolite)

Full size image

Discussion

Studies of parturition should ideally be performed on late pregnant females without pharmacological intervention. However, the exact time of parturition is difficult to predict in cattle and a model where parturition is induced in a physiological manner can offer an alternative. Induction of parturition with dexamethasone gives a defined start of the initiation of parturition and thereby facilitates the intensive blood sampling that is necessary for the investigation of the rapid hormonal changes around parturition.

The main finding of this study was that there were no signs of pulsatile PGF2α release leading to prepartal luteolysis. This is in agreement with studies by [1] in cows, and by [5] in goats, but unlike the situation in the bovine oestrous cycle [10]. In contrast to the pulsatile pattern observed in the oestrous cycle, the PGF2α metabolite levels increased in a continuous way, showing a completely different profile. However, even though luteolysis is essential both in the oestrous cycle and before parturition the prerequisites are different at the two occasions. The prerequisite for luteolysis in the oestrous cycle includes 2 options: luteolysis in the case of non-pregnancy and non-luteolysis in case of pregnancy. Prepartal luteolysis, however, only includes one option, luteolysis without exceptions. The 2 kinds of release patterns possibly reflect this difference.

The absolute levels of PGF2α metabolite at the time of prepartal luteolysis (1.6-0.4 days antepartum) are comparable to those observed during the luteolytic pulses in the oestrous cycle [3], but the levels observed after progesterone decline differ between prepartal and preovulatory luteolysis. After prepartal luteolysis, in this experiment, the PGF2α metabolite levels continue to increase (5–10 times) until the end of calving while after luteolysis in the oestrous cycle the pulsatility ceases and PGF2α metabolite levels decrease to basal levels. However, in a study by [11], parturition in heifers was induced with PGF2α. In that study, the PGF2α metabolite levels at the time of foetal expulsion (which was uneventful and occurred approx. 2 days post injection) were found to be around 10 times lower than what was observed in our experiment. The discrepancy between the results suggests that although the peripheral PGF2α metabolite levels are several times higher during parturition after dexamethasone injections than during parturition after PGF2α injections, this difference did not affect the clinical outcome of the birth process.

PGF2α metabolite profile immediately after calving differed between RFM and non RFM heifers. In both groups, the levels of the PGF2α metabolite were high at the time of calving and there was an immediate decrease after the foetal expulsion. But, unlike the non-RFM heifers, the postpartal decline was soon interrupted by a new period of increasing PGF2α metabolite levels in the RFM heifers. [20] showed that, in sheep, COX-2 expression in cotelydonary tissue increased and was the enzyme predominantly responsible for prostaglandin synthesis in late gestation. Thus, a separation of the foetal and maternal placentas as seen when the foetal membranes are shed immediately after calving resulted in an abrupt removal of the source of PGF2α and, consequently, to a quick decline in PGF2α metabolite levels. In RFM heifers, on the other hand, the non-shed placenta might have stimulated continuous PGF2α synthesis also after calving. In this study, the experiment ended only 12 h after calving but other studies have shown that post-partal PGF2α metabolite levels in RFM cows are as high as during parturition, or even higher [11]. In a study by [9] it was shown that during the first 2 weeks post partum, cows with retained foetal membranes have levels of PGF2α metabolite that clearly exceed the levels seen in cows, where the placenta was shed immediately after parturition.

An interesting feature of the RFM heifers in this study was the distinct peak of cortisol at calving. The cortisol response might reflect stress due to a prolonged or difficult parturition as suggested by [7] but might also be an effect of the retained foetal membranes per se. There are, however, studies that show a positive correlation between PGF2α metabolite levels and cortisol release. This has been shown, after massive intravenous injection of a synthetic ACTH-analogue (tetracosactide) to pigs [16] and after intravenous endotoxin injections to cattle [17]. Cortisol and PGF2α metabolite levels also increase simultaneously after starvation. The link between these 2 parameters remains unknown. But since only the 2 RFM heifers had cortisol peaks at parturition, although the levels of PGF2α metabolite were as high as in the non-RFM heifers, the mechanism for this correlation must differ from the one that can be explained by the high levels of PGF2α. In conclusion, the release of PGF2α after induction of parturition by injection of dexamethasone in the bovine does not show a pulsatile release as it does during luteolysis in the oestrous cycle. Instead, the pre-partal profile of PGF2α metabolite in the cow is characterised by an ever-increasing release initiated by the dexamethasone injection and terminated by the parturition. The PGF2α metabolite levels then decrease immediately after the parturition. In heifers with retention of the foetal membranes, however, this decrease is soon interrupted by a new increase with PGF2αmetabolite levels as high as during the parturition. Furthermore in this study, heifers with retained foetal membranes had higher levels of cortisol at parturition than heifers where the placenta was shed immediately post partum.