Restoration of LH output and 17β-oestradiol responsiveness in acutely ovariectomised holstein dairy cows pre-treated with a GnRH agonist (deslorelin) for 10 days

Animal Reproduction Science 70 (2002) 49–63

Restoration of LH output and 17␤-oestradiol responsiveness in acutely ovariectomised holstein dairy cows pre-treated with a GnRH agonist (deslorelin) for 10 days A.M. Padula, J.M. Borman, P.J. Wright, K.L. Macmillan∗ University of Melbourne, Veterinary Clinical Centre, 250 Princes Hwy, Werribee, Vic. 3030, Australia Received 8 July 2000; received in revised form 16 November 2001; accepted 19 November 2001

Abstract The objectives of the study were firstly to identify the role of the ovary in maintaining plasma luteinising hormone (LH) concentrations in cows treated with an implant of a potent GnRH agonist (deslorelin), and secondly to characterise the changes in LH following ovariectomy (OVX) in the same animals. Oestrus was synchronised in mature Holstein dairy cows and deslorelin implants were inserted 17 days later into two-third of the cows. A further 10 days later (day 0) all cows had bilateral OVX performed. A control group (CON; n = 4) received no treatment and had blood samples collected at 15-min intervals for 8 h on the day prior to OVX (day −1) and similarly on days 4 and 10. One group (DES IN; n = 4) had implants in place for the duration of the study while another group had implants removed (DES OUT; n = 4) at the time of OVX. DES IN cows were sampled hourly at each sampling session (days −1, +4 and +10), whereas DES OUT cows were sampled similarly to CON except on day −1 when hourly samples were collected. Predictable post-operative increases in mean LH (0.61 ng/ml versus 1.79 ng/ml; P < 0.01) and LH pulse amplitude (0.66 ng/ml versus 1.56 ng/ml; day −1 versus day +10; P < 0.01) occurred after CON cows were ovariectomised. Smoothed LH means showed a delayed effect of time compared to arithmetic means. Pulse frequency was unchanged following OVX in CON cows. A comparison of all cows that had been treated with deslorelin from day −1 showed a significant elevation of smoothed mean LH compared to untreated cows (0.80 ng/ml versus 0.34 ng/ml; DES IN and DES OUT versus CON; P < 0.05). DES IN cows had a 54% reduction in mean LH from day −1 to +4 following OVX (1.05 ng/ml versus 0.48 ng/ml; P < 0.01) indicating the probable involvement of the ovary in the maintenance of elevated basal LH. No further reduction was detected by day +10. The LH response to an intramuscular (IM) injection of 500 ␮g 17␤-oestradiol (E2) on day +11 varied significantly between treatment groups (P < 0.01). CON cows showed ∗ Corresponding author. Tel.: +61-3-9731-2234. E-mail address: [email protected] (K.L. Macmillan).

0378-4320/02/$ – see front matter © 2002 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 4 3 2 0 ( 0 1 ) 0 0 1 9 2 - 0

50

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

a typical LH surge, reaching maximum concentrations (10.3 ng/ml) at 17.3 h post-injection. Even though low amplitude LH pulsatility had been restored in DES OUT cows by day +4, there was an inconsistent response to E2 on day +12; one cow had an apparently normal surge yet, others showed only attenuated responses. Pulse amplitude in DES OUT cows was lower at days +4 and +10 compared to CON (P < 0.05). DES IN cows did not produce any surge after E2. Mean LH prior to OVX (day −1) remained unchanged following the 500 ␮g oestradiol injection (0.38 ng/ml versus 0.45 ng/ml pre-E2 versus post-E2 compared to 1.05 ng/ml pre-OVX). The results of this experiment implicated ovarian involvement in maintaining elevated basal LH output in cows that were chronically treated with a GnRH agonist. Individual cows varied in their LH surge response to exogenous E2 given 12 days after implant removal, even though LH pulse amplitude and frequency had been restored. © 2002 Elsevier Science B.V. All rights reserved. Keywords: GnRH agonist; Deslorelin; Dairy cows; Ovariectomy; LH

1. Introduction The chronic treatment of female cattle with a GnRH agonist is characterised by an initial stimulatory phase (Gong et al., 1995), which is soon followed by suppression of pulsatile gonadotrophin output and ultimately inhibition of follicle development (Gong et al., 1996). Pulsatile secretion of luteinising hormone (LH) was not detected in heifers chronically administered buserelin with an osmotic mini-pump (Gong et al., 1996), in bulls injected four times daily with nafarelin (Melson et al., 1986), in bulls treated with deslorelin implants (D’Occhio and Aspden, 1996) or in bulls given leuprolide with an osmotic mini-pump (Ronayne et al., 1993). A feature of treatment of female cattle with a GnRH agonist has been the continued maintenance of basal LH output, despite the loss of sensitivity to exogenous GnRH (Bergfeld et al., 1996a; Gong et al., 1996). Chronic treatment of bulls reduced the number of GnRH receptors in the anterior pituitary gland to only 16% of control animals and substantially lowered the total pituitary LH and FSH content to about 5% of controls (Melson et al., 1986). Elevated basal LH was observed during the first 48 h of a chronic deslorelin treatment in pre-pubertal heifers, but was similar to untreated controls over the subsequent 28 days (Bergfeld et al., 1996a). A similar pattern was reported in two separate studies using heifers where mean LH was elevated above control animals for the first 3 days of a 21-day treatment (Gong et al., 1995), and the first 8 days of a 48-day treatment (Gong et al., 1996). Some studies in the bull have reported elevated mean LH concentrations during chronic deslorelin treatment (Aspden et al., 1997c; D’Occhio and Aspden, 1996) associated with an increase in plasma testosterone. Other reports in the bull have found no change in mean LH concentration (Aspden et al., 1997a; Bergfeld et al., 1995; Ronayne et al., 1993; Melson et al., 1986; Rechenberg et al., 1986) yet increased testosterone concentrations. One study demonstrated a necessity for endogenous LH to maintain elevated plasma testosterone levels (Aspden et al., 1997b). In a number of other species studied (mice, rats, dogs, rams, pigs, baboons, rhesus monkey and men) the pituitary axis has been suppressed following GnRH agonist treatment (Aspden et al., 1997a). Gonad intact Marmoset monkeys treated with buserelin implants for 3 months maintained continued basal plasma LH concentrations yet were unresponsive to

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

51

exogenous GnRH (Lunn et al., 1992). In the same study, long term ovariectomised female Marmoset monkeys treated with buserelin implants showed markedly reduced plasma LH levels compared to untreated ovariectomised contemporaries. The authors suggested that the maintenance of basal LH secretion might be the result of a protective influence of gonadal origin. In contrast, castration of bulls after treatment with deslorelin did not result in typical post-castration rises in plasma LH, even though pituitary desensitisation to GnRH was demonstrated (D’Occhio and Aspden, 1996). Deslorelin is a highly potent GnRH agonist in rodents, exhibiting approximately 50–200 times the potency of native GnRH, depending on the evaluation model (Karten and Rivier, 1986). Reversible anoestrus has been induced in Bos indicus heifers with deslorelin delivered using subcutaneous controlled release implants (D’Occhio et al., 1996). Oestradiol has been shown to induce the transcription of mRNA coding for the GnRH receptor in ewes infused for 5 days with GnRH (Turzillo et al., 1998b). Studies in the hypothalamo–pituitary disconnected ewe suggest that GnRH is of greater importance for inducing transcription of GnRH receptors than oestradiol alone, which appear to be dependent on the presence of GnRH. The first objective of the study is to investigate whether the ovary was required to maintain basal LH pituitary output by ovariectomising cows that had been pre-treated with an implant of a GnRH agonist (deslorelin) for 10 days. It was suspected that an ovarian product, possibly oestradiol, was acting directly at the pituitary to increase LH output and that OVX would result in an immediate fall in basal LH. The blood-sampling schedule was designed to detect a change in mean LH in the 10-day period immediately following ovary removal. The second objective is to describe the pattern of restoration of pulsatile LH output and pituitary responsiveness to oestradiol in acutely ovariectomised cattle that had been pre-treated with deslorelin and either had the GnRH agonist implant removed (DES OUT) at OVX or were treated continuously with the agonist (DES IN). The deslorelin treated groups were compared to control cows that had not received deslorelin at any stage yet were OVX and sampled identically. 2. Materials and methods 2.1. Cow management Oestrus was synchronised in 12 non-lactating Holstein-Friesian dairy cows using an intravaginal progesterone releasing device (CIDR-B® ; Genetics Australia; Bacchus Marsh, Australia) and an intramuscular (IM) injection of 2.0 mg oestradiol benzoate (ODB) (CIDIROL® , Interag, New Zealand) at device insertion. Each device was removed 8 days later when each cow was injected IM with 500 ␮g cloprostenol (CIDR-PG® ; Genetics Australia). At 24 h post device removal each cow was given an IM injection of 1.0 mg ODB at 24 h after device removal. Oestrus was detected by observation of a standing oestrus within the ensuing 24 h. Twelve cows detected in oestrus (day 0) during this time were used in the study. They were housed indoors and fed grass hay at the Victorian Institute of Animal Science (VIAS) in Werribee for the duration of the experiment.

52

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

Prior ethics approval was obtained through VIAS and the University of Melbourne Animal Experimentation and Ethics Committee for the study, involving the experimental management of the animals. 2.2. Treatment groups On the 17th day post-oestrus cows were randomly allocated to receive either a single 6.0 mg [D-Trp6 , Pro9 −des-Gly10 -NH2 ]GnRH-ethylamide (deslorelin) implant (Peptech Animal Health, North Ryde, Australia) placed subcutaneously into the ear (n = 8), or no treatment (CON; n = 4). The implants were designed to release deslorelin at approximately 50 ␮g per 24 h for at least 3 months. Bilateral ovariectomy was performed 10 days after implant insertion (day 27 post-oestrus) when half of the cows with deslorelin implants (DES OUT; n = 4) also had their implants removed. The remaining group had implants left in place (DES IN; n = 4) for the duration of the experiment. An IM injection of prostaglandin F2␣ (CIDR-PG® ) was given on the 23rd day post-oestrus to every cow that had received a deslorelin implant (i.e. DES IN and DES OUT) to remove any induced corpora lutea arising from the LH surge associated with implant insertion (Fig. 1). 2.3. Ovariectomy Cattle were prepared for ovariectomy by tranquillisation with 20 mg of acepromazine maleate IM (Promex 10® ; Apex Laboratories, Australia) 10 min prior to surgery. The left flank was clipped and prepared aseptically for flank laparotomy. The area to be incised was anaesthetised with 50–100 ml of lignocaine (Lignomav® ; Mavlab Pty. Ltd., Australia),

Fig. 1. Overview of the experimental design.

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

53

muscle layers separated and the peritoneal cavity opened. The ovaries were located and removed using a chain loop ecrasuer (Youngquist, 1997). The flank incision was closed and each cow given a single IM injection of 25 ml of procaine penicillin (Propen® ; Ilium Laboratories, Australia). 2.4. Intensive bleeding protocol Indwelling jugular catheters were inserted 24 h prior (i.e. day −2) to the first day of bleeding. Intensive bleeds were performed on days −1, +4 and +10 relative to ovariectomy (Fig. 1). Sampling sessions began at 9:00 a.m. on the scheduled day. Blood samples (10 ml) were collected every 15 min for 8 h into heparinised tubes. Catheters were flushed after every sample with heparinised saline (50 IU/ml). Plasma was separated by centrifugation and frozen at −20 ◦ C until subsequently assayed for LH. Cows with deslorelin implants in situ were sampled at hourly intervals (DES IN days −1, +4 and +10; DES OUT day −1). 2.5. Oestradiol administration Each cow received an IM injection of 500 ␮g of 17␤-oestradiol in peanut oil on the day 11 after ovariectomy (Fig. 2). Blood samples were collected at hourly intervals for 8 h beginning at 14 h after injection (day +12) to detect an LH surge. 2.6. Luteinising hormone assay Every sample from days −3, −1, +2, +4 and +6 from an individual cow was evaluated in a single assay to avoid adding any effect of inter-assay variation. Samples collected at hourly intervals from all cows on day 0 were run in one assay. LH was assayed using a modified double antibody precipitation radioimmunoassay (Wright et al., 1980). Phosphate buffered saline (PBS) was used throughout the assay. Ovine LH standards (80, 40, 20, 10, 5, 2.5, 1.25, 0.6, 0.4, 0.15 ng/ml) were prepared in plasma from a hypophysectomised–ovariectomised (HYPOX) ewe. Internal quality control samples were made by spiking HYPOX plasma with the stock ovine standard to give approximate final concentrations of 10.0, 5.0 and 0.5 ng/ml. The primary antibody was equine anti-bovine LH (Snook, 1968) diluted 1:300,000. Bovine 125 I -LH was prepared using the chloramine T method, purified by high-pressure liquid chromatography and diluted to produce 12,000 cpm in a 100 ␮l aliquot. A 100 ␮l sample volume was used for standards, quality controls and unknown samples. These were incubated along with 100 ␮l of the first antibody, 100 ␮l of bovine 125 I -LH and 100 ␮l of equine gamma globulin at 37 ◦ C for 24 h. Separation was performed by addition of the second antibody (goat anti-horse) using 100 ␮l of a 1:160 dilution. This mixture was incubated at 4 ◦ C overnight. A 1.0 ml volume of a 10% polyethylene glycol solution was then added and the mixture incubated at 4 ◦ C for 30 min before being centrifuged at 3000 × g for 15 min so that the supernatant could be aspirated and the remaining pellet counted for 1 min in a gamma counter. The minimum detectable concentration was 0.15 ng/ml and within and between assay variation averaged less than 15% across all assay batches.

54

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

Fig. 2. LH response to an IM injection of 500 ␮g 17␤-oestradiol given on day +11 following ovariectomy. Data presented as group mean LH concentration (top panel) and individual cows from DES OUT (lower panel) at hourly intervals. Error bars represent S.E. of the mean.

2.7. Statistical analysis The software package PULSTER (Merriam and Wachter, 1982) was used to calculate the number of LH pulses and the amplitude of each pulse (Turek and Van Cauter, 1994) (Figs. 3 and 4). A linear CV function was used to define the assay variation:

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

55

Fig. 3. Individual LH pulse profiles for a typical untreated CON cow relative to day of ovariectomy.

Fig. 4. LH pulse profiles for a typical cow that had its deslorelin implant removed at the time of ovariectomy (DES OUT).

56

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

CV (%) = 9.53 + 0.168 × X. The mean pulse amplitude per cow per 8 h period was used whenever multiple pulses were detected. Arithmetic mean and PULSTER smoothed mean LH concentrations for each group across time was examined using a repeated measures analysis of variance (ANOVA) approach in a mixed model with the computer software package SAS Analyst® 8.0 (SAS Institute, 1999). The PROC MIXED command with REML estimation was used to create a fixed and random effects model suitable for repeated measurements within cows (Littell et al., 1998). Cow was classified as a random effect. The model was of the form: mean LH = day, treatment and the interaction of day across treatment. The Tukey–Kramer method for multiple comparisons was used to compare groups within and across days. The SLICE option of the LSMEANS subcommand was used to partition the interaction term when the interaction was significant (see SAS online docs at www.sas.com). LH pulse frequency was compared by non-parametric methods. Pulse amplitude was compared using non-parametric one-way ANOVA (Kruskal–Wallis test, Minitab V12.23). Peak LH response to oestradiol was compared by one-way ANOVA. The area under the curve (AUC) for the oestradiol administration was calculated using the trapezoidal rule and log transformed before one-way ANOVA.

3. Results 3.1. Arithmetic mean LH concentration The differences associated with day of blood sampling were not statistically significant (P = 0.08) (Table 1). The fixed effects of treatment (P < 0.05) and the treatment–day (P < 0.001) interaction necessitated using the SAS Analyst® SLICE command to partition the interaction term. Partitioning by treatment group showed that the interaction across days in mean LH concentration was confined to the DES IN (P < 0.01) and CON groups (P < 0.01) while the DES OUT (P > 0.1) was not involved. Partitioning the interaction by day revealed no effect at day −1 (P > 0.1), but the interaction became significantly greater

Table 1 Comparative arithmetic mean LH concentrations (ng/ml) by day of experiment relative to ovariectomy (OVX = day 0). Data presented as means (±S.E.M.) of all samples collected over the 8 h sampling perioda Group

Mean LH concentration (ng/ml) Day −1

Day +4

Day +10

DES IN DES OUT Control

1.05 ± 0.29 a 0.56 ± 0.12 0.61 ± 0.16

0.48 ± 0.13 b 0.42 ± 0.10 1.31 ± 0.50 b

0.38 ± 0.10 bxb 0.76 ± 0.21 x 1.79 ± 0.41 by

Day mean

0.73 ± 0.1

0.74 ± 0.17

0.98 ± 0.12

a

Within each cow there was no inter-assay variation only within assay variation. Means within rows without common letters (a, b) differ (P < 0.05). Means within columns without common letters (x, y) differ (P < 0.05). b

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

57

Fig. 5. Mean plasma LH concentration for cows that had implants in place for the duration of the study (DES IN; n = 4). Error bars represent S.E. of the mean.

over time (day +4: P < 0.01; day +10: P < 0.001) indicating that divergent changes in mean LH concentrations had occurred after ovariectomy (Fig. 5). There were no differences in mean LH between groups (P > 0.05) on the day prior to ovariectomy (day −1). Since that at day −1 both DES IN and DES OUT cows had received identical treatment, the means were pooled and compared with non-deslorelin treated CON cows. This revealed similar mean LH concentrations (0.849 ng/ml versus 0.606 ng/ml deslorelin versus control; P = 0.21, t-test). Ovariectomy of CON cows resulted in an increase in mean LH to almost three times the pre-ovariectomy levels by day +10 (0.606 ng/ml versus 1.785 ng/ml; day −1 versus day +10; P < 0.001) to over twice that of either DES IN or DES OUT group. Mean LH concentration for DES IN cows had declined by 54% at day +4 after ovariectomy (P < 0.01), but did not decline further by day +10 (P > 0.05). There was no effect of ovariectomy on mean LH in DES OUT cows (P > 0.05), although there did appear to be a consistent trend with a small decline in LH at day +4 (similar to DES IN) followed by a small increase by day +10. 3.2. Smoothed mean LH concentration Smoothed means were generated using the PULSTER software with the LOWESS algorithm (Merriam and Wachter, 1982) and then compared to arithmetic means where samples had been collected at hourly intervals. It was assumed for the analysis that the arithmetic mean would be a close approximation of the smoothed mean and thus represent a valid comparison for cows with deslorelin implants in place when serial sampling was performed. The use of smoothed means was used in an attempt to remove the effect of individual cow variation in pulse frequency and amplitude biasing the arithmetic means. The smoothed

58

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

Table 2 Summary of PULSTER derived smoothed LH meansa Group

DES IN DES OUT Control Day mean

Smoothed mean LH concentration (ng/ml) Day −1

Day +4

Day +10

(1.05 ± 0.29 xab ) (0.56 ± 0.12 x) 0.34 ± 0.02 a

(0.48 ± 0.13 b) 0.31 ± 0.09 0.65 ± 0.16 a

(0.38 ± 0.10 bx) 0.51 ± 0.13 yx 1.11 ± 0.14 by

0.65 ± 0.11

0.48 ± 0.04 a

0.67 ± 0.11

a

Means (±S.E.M.) in brackets represent arithmetic means for cows where samples were collected at hourly intervals and known to be non-pulsatile, but were used for comparison. b Means within rows without common letters (a, b) differ (P < 0.05). Means within columns without common letters (x, y) differ (P < 0.05).

data should better represent the basal pituitary LH output between endogenous pulses of GnRH (Table 2). Smoothed means generally followed a similar pattern to the arithmetic means but they were lower. Day of sampling effect was significant (P = 0.049) because of the lower overall mean value on day +4 (0.479 ng/ml). There was a treatment–day interaction (P < 0.0001). The SLICE command showed both CON and DES IN groups were involved in the interaction (P < 0.01) while day +4 was not involved in the treatment–day interaction (P > 0.1). Since both DES IN and DES OUT had received identical treatment up to day −1, data from both groups were pooled. A Student’s t-test of these pooled means showed that the presence of deslorelin implants significantly increased the smoothed mean LH concentrations (0.801 ng/ml versus 0.342 ng/ml; deslorelin versus CON; P < 0.05). There was a tendency towards no difference in smoothed mean LH for CON versus DES OUT by day +10 after ovariectomy (0.51 versus 1.1; DES OUT versus CON; P = 0.11Tukey; P = 0.0067; t-test). In summary, examination of smoothed LH means revealed a slower rise in basal LH following ovariectomy of CON cows, and almost similar means for CON and DES OUT by day +10. Overall a similar pattern to arithmetic means was observed. Use of smoothed means significantly reduced the variation in the CON group (Fig. 6). 3.3. LH pulse amplitude CON cows showed typical initial increases in LH pulse amplitude following ovariectomy which had appeared to stabilise by day +10 (1.49 ng/ml versus 1.56 ng/ml; day +4 versus day +10; P > 0.1) (Table 3). The recovery of pulsatile LH following ovariectomy of DES OUT cows had resulted in a 107% increase in mean LH pulse amplitudes by day +10 (0.28 versus 0.58 per 8 h; day +4 versus day +10; P < 0.05), but was still lower than that of CON cows at both times. One DES OUT cow showed a large increase in pulse amplitude between day +4 and +10 (0.31 ng/ml versus 0.91 ng/ml). This cow also had an almost normal surge response to oestradiol on day +12 (LHmax = 9.30 ng/ml; AUC = 35.7).

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

59

Fig. 6. Graphical comparison of the use of smoothed or arithmetic mean LH concentration for control cows (OVX = day 0). Error bars represent S.E. of the mean.

3.4. LH pulse frequency There was no change in LH pulse frequency in CON cows following OVX. DES OUT cows had pulsatile LH release at day +4, although amplitudes were low (0.28 ng/ml). There was a small increase in their pulse frequency (5.8 pulses per 8 h versus 8.8 pulses per 8 h; P < 0.05) between days +4 and +10 (Table 4; Figs. 3 and 4).

Table 3 Summary of LH responses to a single IM injection of 500 ␮g 17␤-oestradiol given 14 h previously, samples were collected at hourly intervals over the period 14–22 h post-injection (n = 4 per treatment group)a Group DES IN DES OUT Control a

Mean Minimum, maximum Mean Minimum, maximum Mean Minimum, maximum

LHmax (ng/ml)

Mean LH (ng/ml)

Hours to peak

AUC

0.61 ± 0.14 c 0.29, 0.88 4.32 ± 1.68 b 2.02, 9.30 10.34 ± 0.37 a 9.40, 11.22

0.45 ± 0.10 c 0.22, 0.63 2.49 ± 0.64 b 1.69, 4.25 5.00 ± 0.42 b 3.96, 5.88

15.8 ± 1.2 b 14,19 21.0 ± 0.6 a 20, 22 17.3 ± 0.7 b 16, 19

3.6 ± 0.8 c 1.7, 5.1 20.2 ± 5.4 b 12.0, 35.7 42.3 ± 3.1 a 34.6, 49.0

Ovariectomy was performed 12 days earlier. Data presented as peak LH concentration, AUC and time to peak (mean ± S.E.). Means within columns without common letters (a, b, c) differ significantly (P < 0.05).

60

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

Table 4 Summary of LH pulse data for individual cows that had deslorelin implants removed at ovariectomy (DES OUT) and untreated control cows (CON)a Group

Control DES OUT

LH pulses per 8 h

LH pulse amplitude (ng/ml)

Day −1

Day +4

Day +10

Day −1

Day +4

7.5 ± 0.3

9.0 ± 2.2 5.8 ± 1.5 bb

8.8 ± 1.3 8.8 ± 0.5 a

0.66 ± 0.2 b

1.49 ± 1.3 0.28 ± 0.04 bx

∗c

∗c

Day +10 ayb

1.56 ± 0.7 ay 0.58 ± 0.2 ax

a

Pulse amplitudes are presented as the mean pulse height (ng/ml) for all pulses detected. S.D. follow group means. b Means within rows (a, b, c) and columns (x, y) without common letters differ significantly (P < 0.05). c Samples collected at hourly intervals only.

3.5. LH response to 500 µg IM injection of 17β-oestradiol There was a noticeable increase in plasma LH levels in CON and DES OUT groups (Fig. 2) that began 15 h after injection of oestradiol. A surge release of LH was observed in all CON cows. The interval to maximum LH was slighter longer in DES OUT cows than CON cows (17.3 h versus 21.0 h; CON versus DES OUT; P < 0.05) (Table 3; Fig. 2). The maximum concentration of LH and AUC differed between groups (P < 0.001). Individual cows within DES OUT showed a variable response as indicated by the wide range of LH responses observed (Table 3; Fig. 2). Comparison of mean LH over the sampling period for DES IN cows pre- and postoestradiol injection indicated an inability of oestradiol to restore LH to pre-OVX levels (0.38 ± 0.19 ng/ml on day +10 versus 0.45 ± 0.20 ng/ml on day +12 compared to 1.05 ± 0.57 ng/ml pre-OVX). 4. Discussion The results obtained here demonstrate that the ovary is required to maintain elevated basal LH in cows chronically treated with a GnRH agonist. The evidence comes from the observation that within 4 days of ovariectomy, the mean LH levels in DES IN cows had declined to 46% of pre-OVX levels. The effect of time between OVX and LH profiling (5 days) could be argued to have contributed to the drop in LH. The evidence against this comes from other studies in heifers where LH had reached its lowest level between 2 and 8 days after treatment initiation when it was no different to untreated females (Bergfeld et al., 1996a; Gong et al., 1996, 1995). The formation of a fourth treatment group consisting of deslorelin without ovariectomy was considered during the design of the study, but due to the above-published evidence and space limitations, it was not included. There was also no further decline of mean LH concentrations between days +4 and +10 for DES IN cows (0.48 ng/ml versus 0.38 ng/ml; day +4 versus +10; P > 0.05). A large reduction in mean LH was observed after buserelin implants were inserted into long-term ovariectomised Marmoset monkeys (Lunn et al., 1992). Weekly sampling was used and the decline following insertion of implants appeared to take up to 14 days to reach its lowest point in this study, but the greatest decline was noted at the first sample collected only 1 week after implants

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

61

were inserted. The authors suggested that a gonadal product was acting to maintain elevated LH, since sham-treated ovariectomised control animals showed no reduction in LH. Deslorelin treatment for 10 days prior to ovariectomy altered the recovery pattern of LH output and oestradiol responsiveness compared to untreated cows. Control cows showed typical post-ovariectomy increases in mean LH pulse amplitude, mean LH and smoothed mean LH concentrations similar to those reported in other studies (Hobson and Hansel, 1972). The LH response to oestradiol in CON cows was surge like and began at a time similar to previously reported (Hobson and Hansel, 1972). A 500 ␮g 17␤-oestradiol injection did not restore mean basal LH to pre-OVX levels in cows treated continuously with deslorelin (DES OUT). Although this is a tenuous comparison considering between assay variation, the difference was larger than could be expected entirely from assay variation (0.38 ± 0.19 ng/ml day +10 versus 0.45 ± 0.20 ng/ml day +12 compared to 1.05 ± 0.57 ng/ml pre-OVX). Oestradiol has been shown to increase the number of GnRH receptors expressed in the pituitary in OVX ewes but only in the absence of progesterone (Turzillo et al., 1998a). In these ewes infused continuously with GnRH for 5 days, oestradiol was able to increase the expression of GnRH receptor mRNA and receptor number. The recovery of LH pulse amplitude in one DES OUT cow and the observation that this cow also had a near normal LH surge after oestradiol may indicate a link between pulse amplitude and pituitary content. In the hypothalamo–pituitary-disconnected ewe, LH pulse amplitude was a direct representation of total releasable stores of LH (Clarke and Cummins, 1985). Interestingly, one DES OUT cow that had the highest pulse amplitude also had the highest basal LH and produced the largest LH surge response. A larger group size may have revealed more about the variation between cows as they were recovering from the down-regulation produced by the deslorelin implant. One hypothesis to explain the variation in response to oestradiol in the DES OUT group at day +12 may be that pituitary stores of LH have not been sufficiently restored and that significant variation between animals exists in rate of restoration. GnRH responsiveness has been examined in pre-pubertal heifers that were previously treated for 28 days and then challenged at six intervals within the first 20 days after deslorelin implant removal (Bergfeld et al., 1996b). The results indicated that the differences between treatment and control groups did not begin to occur until at least 12 days after implant removal. The variation across days was large but there was a significant and gradual mean increase from day +12 onwards. Bulls injected four times daily with nafarelin had 20-fold reductions in gonadotrophin content when killed after 15 days (Melson et al., 1986). The recovery process must partly involve resynthesis of pituitary gonadotrophins. Individual cows appear to vary in the pattern in which they recover pituitary function following GnRH receptor down-regulation as expressed in the variability between cows in their response to oestradiol (Table 3; Fig. 2) and intervals to resumption of ovulation when postpartum cows have been treated with deslorelin implants (Padula et al., 2000, 1999).

5. Conclusions These results indicate that ovarian factors were involved in the maintenance of the marginally elevated basal LH levels seen in cows chronically treated with a

62

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

GnRH agonist because large reductions were observed immediately following ovary removal. The pattern of LH pulse amplitude, frequency and response to oestradiol administration restoration following either pre-treatment for 10 days with deslorelin or continuous treatment are described. The response to oestradiol administration varied widely amongst cows at 12 days following implant removal.

Acknowledgements The authors sincerely thank Dr. Tim Trigg from Peptech Animal Health Australia for the generous supply of deslorelin implants used in this study. Thank you to Terry Squires of VIAS for technical assistance with this study. Thank you to Prof. Iain Clarke of Prince Henry’s Institute of Medical Research for support with the LH assay and general comments on the study. References Aspden, W.J., Rao, A., Rose, K., Scott, P.T., Clarke, I.J., Trigg, T.E., Walsh, J., D’Occhio, M.J., 1997a. Differential responses in anterior pituitary luteinizing hormone (LH) content and LH beta- and alpha-subunit mRNA, and plasma concentrations of LH and testosterone, and plasma concentrations of LH and testosterone, in bulls treated with the LH-releasing hormone agonist deslorelin. Dom. Anim. Endocrinol. 14, 429–437. Aspden, W.J., van Reenen, N., Whyte, T.R., Maclellan, L.J., Scott, P.T., Trigg, T.E., Walsh, J., D’Occhio, M.J., 1997b. Increased testosterone secretion in bulls treated with a luteinizing hormone releasing hormone (LHRH) agonist requires endogenous LH but not LHRH. Dom. Anim. Endocrinol. 14, 421–428. Aspden, W.J., Vanreenen, N., Whyte, T.R., Maclellan, L.J., Scott, P.T., Trigg, T.E., Walsh, J., D’Occhio, M.J., 1997c. Increased testosterone secretion in bulls treated with a luteinizing hormone releasing hormone (LHRH) agonist requires endogenous LH but not LHRH. Dom. Anim. Endocrinol. 14, 421–428. Bergfeld, E.G.M., D’Occhio, M.J., Kinder, J.E., 1995. Continued desensitisation of the pituitary gland in bulls after treatment with the luteinising hormone agonist deslorelin. Biol. Reprod. 54, 769–775. Bergfeld, E.G.M., D’Occhio, M.J., Kinder, J.E., 1996a. Pituitary function, ovarian follicular growth, and plasma concentrations of 17 beta-estradiol and progesterone in prepubertal heifers during and after treatment with the luteinizing hormone-releasing hormone agonist deslorelin. Biol. Reprod. 54, 776–782. Bergfeld, E.G.M., D’Occhio, M.J., Kinder, J.E., 1996b. Pituitary function, ovarian follicular growth, and plasma concentrations of 17-beta-estradiol and progesterone in prepubertal heifers during and after treatment with the luteinizing hormone-releasing hormone agonist deslorelin. Biol. Reprod. 54, 776–782. Clarke, I.J., Cummins, J.T., 1985. GnRH pulse frequency determines LH pulse amplitude by altering the amount of releasable LH in the pituitary glands of ewes. J. Reprod. Fertil. 73, 425–431. D’Occhio, M.J., Aspden, W.J., 1996. Characteristics of luteinizing hormone (LH) and testosterone secretion, pituitary responses to LH-releasing hormone (LHRH), and reproductive function in young bulls receiving the LHRH agonist deslorelin: effect of castration on LH responses to LHRH. Biol. Reprod. 54, 45–52. D’Occhio, M.J., Aspden, W.J., Whyte, T.R., 1996. Controlled, reversible suppression of estrous cycles in beef heifers and cows using agonists of gonadotropin-releasing hormone. J. Anim. Sci. 74, 218–225. Gong, J.G., Bramley, T.A., Gutierrez, C.G., Peters, A.R., Webb, R., 1995. Effects of chronic treatment with a gonadotrophin-releasing hormone agonist on peripheral concentrations of FSH and LH, and ovarian function in heifers, and ovarian function in heifers. J. Reprod. Fertil. 105, 263–270. Gong, J.G., Campbell, B.K., Bramley, T.A., Gutierrez, C.G., Peters, A.R., Webb, R., 1996. Suppression in the secretion of follicle-stimulating hormone and luteinizing hormone, and ovarian follicle development in heifers continuously infused with a gonadotropin-releasing hormone agonist, and ovarian follicle development in heifers continuously infused with a gonadotropin-releasing hormone agonist. Biol. Reprod. 55, 68–74.

A.M. Padula et al. / Animal Reproduction Science 70 (2002) 49–63

63

Hobson, W.C., Hansel, W., 1972. Plasma LH levels after ovariectomy, corpus luteum removal and oestradiol administration. Endocrinology 91, 185–190. Karten, M.J., Rivier, J.E., 1986. Gonadotropin-releasing hormone analog design. Structure-function studies toward the development of agonists and antagonists: rationale and perspective. Endocrine Rev. 7, 44–66. Littell, R.C., Henry, P.R., Ammerman, C.B., 1998. Statistical analysis of repeated measures data using SAS procedures. J. Anim. Sci. 76, 1216–1231. Lunn, S.F., Cowen, G.M., Morris, K.D., Fraser, H.M., 1992. Influence of the gonad on the degree of suppression induced by an LHRH agonist implant in the Marmoset monkey. J. Endocrinol. 132, 217–224. Melson, B.E., Brown, J.L., Schoenemann, H.M., Tarnavsky, G.K., Reeves, J.J., 1986. Elevation of serum testosterone during chronic LHRH agonist treatment in the bull. J. Anim. Sci. 62, 199–207. Merriam, G.R., Wachter, K.W., 1982. Algorithims for the study of episodic homrone secretion. Am. J. Physiol. 243, E310–E318. Padula, A.M., Trigg, T.E., Macmillan, K.L., 1999. Deslorelin implants prevent early ovulations in early pospartum Holstein dairy cattle. In: Proceedings of the Australian Society for Reproductive Biology, Melbourne. Vol. 30, 123 (Abstract). Padula, A.M., McGowan, M.R., Verrall, R., Trigg, T.E., Macmillan, K.L., 2000. Absorbable deslorelin implants delay the resumption of ovulatory cycles in postpartum dairy cattle. In: Proceedings of the 4th International Congress on Animal Reproduction, Stockholm, Sweden (Abstracts). Rechenberg, W., Sandow, J., Klatt, P., 1986. Effect of long term infusion of an LH releasing hormone agonist on testicular function in bulls. J. Endocrinol. 109, R9–R11. Ronayne, E., Enright, W.J., Roche, J.F., 1993. Effects of continuous administration of gonadotropin-releasing hormone (GnRH) or a potent GnRH analogue on blood luteinizing hormone and testosterone concentrations in prepubertal bulls. Dom. Anim. Endocrinol. 10, 179–189. Snook, R.B., 1968. Immunological and biological properties of antiserum to bovine LH. Excerpta Med. Int. Cong. Ser. 161, 398–400. Turek, F.W., Van Cauter, E., 1994. Rhythms in reproduction. In: Knobil, E. (Ed.), The Physiology of Reproduction, 2nd Edition, Vol. 2. Raven Press, New York, USA, pp. 487–540. Turzillo, A.M., Clapper, J.A., Moss, G.E., Nett, T.M., 1998a. Regulation of ovine GnRH receptor gene expression by progesterone and oestradiol. J. Reprod. Fertil. 113, 251–256. Turzillo, A.M., Nolan, T.E., Nett, T.M., 1998b. Regulation of gonadotropin-releasing hormone (GnRH) receptor gene expression in sheep: interaction of GnRH and estradiol. Endocrinology 139, 4890–4894. Wright, P.J., Geytenbeek, P.E., Clarke, I.J., Findlay, J.K., 1980. Pituitary responsiveness to LH-RH, the occurrence of oestradiol-17 beta induced LH-positive feedback and the resumption of oestrous cycles in ewes post partum. J. Reprod. Fertil. 60, 171–176. Youngquist, R.S., 1997. Surgical correction of abnormalities of genital organs of cows. In: Youngquist, R.S. (Ed.), Current Therapy in Large Animal Theriogenology, 1st Edition. Saunders, Philadelphia, USA, pp. 429–440.