Regulation of Luteinizing Hormone-Releasing Hormone (LHRH ...

6 downloads 0 Views 800KB Size Report
Simultaneous Delayed Increase in LHRH and Luteinizing Hormone ... LHRH and LH pulsatile secretion was studied in the first group during inhibition of LH ...
BIOLOGY OF REPRODUCTION 52, 1114-1120 (1995)

Regulation of Luteinizing Hormone-Releasing Hormone (LHRH) Secretion by Melatonin in the Ewe. . Simultaneous Delayed Increase in LHRH and Luteinizing Hormone Pulsatile Secretion' CATHERINE VIGUIE, 3 ALAIN CARATY, ALAIN LOCATELLI, and BENOIT MALPAUX 2

Laboratoirede NeuroendocrinologieSexuelle, INRA-PRMD, 37380 Nouzilly, France ABSTRACT This study was performed to test the hypothesis that a short-day-like melatonin treatment would induce simultaneous stimulations of LHRH and LH secretion. Simultaneous serial samples of jugular and portal blood to assay LH and LHRH, respectively, were carried out at three different periods according to the expected changes in LH secretion. Twenty-six ovariectomized ewes bearing a subcutaneous implant of estradiol and exposed to a long-day photoperiod (16L:8D) received s.c. implants of melatonin on Day 0. LHRH and LH pulsatile secretion was studied in the first group during inhibition of LH secretion by long days (on Day -1, n = 6). A second group (n = 9) was sampled on Day 39. Moreover, seven animals from this group were sampled on Day 46 to focus on the beginning of LH increase. The third group (n = 6) was sampled during maximal LH secretion (Day 74, n = 6). A fourth group (n = 5) composed of noncannulated animals was sampled for LH secretion only as a control for the effects of surgical procedures. On each occasion, blood was collected every 10 min for 6 h. Between Day -1 and Day 39, LHRH pulse frequency was low and did not differ. In the same animals, this value became 2.5-fold higher between Day 39 and Day 46. Finally, on Day 74, LHRH pulse frequency was about 6-fold higher than on Day -1. The modifications of LH pulse frequency paralleled those of LHRH with no change between Day -1 and Day 39, a 3-fold increase on Day 46 compared to Day 39, and a 4.5-fold increase on Day 74 compared to Day -1. This experiment demonstrated that the increases in LHRH and LH pulsatile secretion following a short-day-like melatonin treatment occurred simultaneously, with a 40-60-day delay.

INTRODUCTION

The sheep is a photoperiodic species in which the reproductive activity can be driven by alternating periods of exposure to inhibitory long days and inductive short days [1, 2]. Photoperiodic information is relayed to the reproductive axis by the pineal gland through its nocturnal secretion of melatonin [3,4]. Variations in day length are transduced into changes in the duration of melatonin secretion, which leads to changes in LH pulse frequency [5, 6]. In sheep, a lengthening in the duration of melatonin secretion (as during exposure to short days) is followed by an increase in LH pulse frequency [6]. Furthermore, in this species, continuous delivery of melatonin appears to convey a short-day signal [7] and thus leads to a stimulation of LH secretion [8]. The mechanisms by which melatonin modulates LH secretion are largely unknown. In particular, a critical point concerning the action of melatonin is the long latency for its effect to be observed on LH secretion. For example, in ovariectomized ewes bearing a subcutaneous implant of estradiol and subjected to long days, the insertion of melatonin implants causes a stimulation of LH secretion only after 40-60 days. Although the high frequency of LH discharges observed during the breeding season results from Accepted January 6, 1995. Received October 4, 1994. 'This work was supported in part by a grant from "Region Centre." Cv. was supported by a Ph.D. grant from "Region Centre." 2 Correspondence. FAX: 33 47 42 77 43. 3 Present address: Reproductive Sciences Program, University of Michigan, 300 North Ingalls Building, Ann Arbor, MI 49109-0404.

a high frequency of LHRH pulses [9], it is not known whether or not the delayed increase in LH release after melatonin treatment results from a synchronous stimulation of LHRH secretion. A discrepancy between LHRH and LH pulsatile secretion has already been demonstrated in other physiological situations [10, 11], in particular during the period of estradiol negative feedback preceding the preovulatory surge [12]. Hence, the objective of the present study was to determine the profile of LHRH secretion following melatonin treatment in order to test the hypothesis that LHRH and LH secretions are simultaneously stimulated by a short-day-like melatonin treatment. MATERIALS AND METHODS

General The experiment was performed with 26 sexually mature Ile de France ewes. They were maintained outdoors at the Research Center of Nouzilly, France (480N), before the study began. They were moved to a light-sealed room on 5 July and exposed to 8L:16D (lights-on at 0600 h) for 124 days and to 11L:13D for 26 days. They were then exposed to constant long days (16L:8D, lights-on at 0400 h) for 60 days before the beginning and until the end of the experiment. Ewes were allocated to four groups balanced in age and body weight. Three groups were sampled for hypophysial portal and jugular blood. Afourth group (n = 5), consisting of noncannulated animals, was sampled only for jugular blood as a control for the effects of surgery and drug treatment on the development of the LH response to melatonin. Each group was housed in a separate room. 1114

1115

MELATONIN AND LHRH PULSATILE SECRETION

All ewes were ovariectomized at least 2 mo before the beginning of the experiment and received a 20-mm subcutaneous (s.c.) silastic implant of 171-estradiol (OVX+E). These implants have been shown to maintain a physiological serum estradiol level similar to that found at the beginning of the follicular phase [13]. The major advantage of this OVX+E model is that changes in LH secretion result only from the action of melatonin; there is no interaction with changing levels of sex steroids as in intact animals.

MELATONIN LH

'I

|

-

UAYS

--

D-1

Experimental Design

On Day 0, all ewes received s.c. implants of melatonin in the left ear (Melovine; generously provided by Hoechst, Hauxton, UK) in order to increase the daytime level of melatonin near to the physiological nocturnal level [8]. It has been well established that this type of melatonin treatment in long-day-treated animals stimulates LH secretion after 4060 days, whereas in animals kept in long days, the LH levels remain low until the development of refractory conditions after about 6 mo of the long days [8, 14, 15]. The efficiency of these implants was assessed by measuring the melatonin concentration in serial plasma samples (hourly at night, every other hour during the day) obtained during a 24-h period by venipuncture of the right jugular 1 wk before and 2 wk after placement of the melatonin implants. Neuroendocrine reproductive activity was assessed from the concentration of LH in samples obtained once a week before the beginning of the experiment (Day -35 to Day -5) and then twice a week during the experiment. In addition, samples were obtained every day between Day 30 and Day 40 in order to determine the times of sampling for group 2 (see below). Serial sampling of jugular and portal blood was performed at four different times according to the expected changes in LH secretion in response to the melatonin treatment [6]. It is not possible to maintain the functionality of the portal collection model for prolonged periods, and therefore a longitudinal design could not be applied. Three different groups were thus used to characterize LHRH secretion at three different times according to the expected changes in LH secretion in response to the melatonin treatment, with only one group being used twice at a critical time. The day before the experiment, two catheters were inserted into each jugular vein, one to perfuse heparin and the other one to sample venous blood. Portal and jugular blood samples were collected continuously with peristaltic pumps as previously described [16]. The first group (n = 6, Fig. 1) was sampled during the inhibition of LH secretion by long days, on Day -1, and was used as a control for the effect of melatonin. The second group (n = 9) was sampled just prior to the onset of the LH increase on Day 39. Seven animals in this group were sampled again (third period) when LH increased, on Day 46. This group was used to measure LHRH pulsatile secretion just before and at the beginning of the increase in LH secretion. To this end, portal

D+74

I D+39 D+46 GROUP 2 (N=9;7)

GROUP 1 (N=6)

spontaneous secretionL

GROUP 3 (N=6)

RH

sampling duration (hours):

6

1.5

Interval between samples (min):

10

10

spontaneous secretion sampling duration

(hours): Intervalbetween (min): samples

LHH

6

3

1.5

10

5 5

10 10

6 10

3

1.5

FIG. 1. Experimental design. The insertion of melatonin implants in ovariectomized and estradiol-treated ewes was performed on Day 0. Top panel shows the times of collection for hypophysial portal blood according to the expected changes in mean LH levels (curve on the top) in response to melatonin treatment. Bottom panels show schedules for collection of blood from the jugular and portal vessels on each occasion. After 6 (groups 1 and 2) or 9 (group 3) h of sampling, ewes received an i.v. bolus of exogenous LHRH (arrow, 250 ng), and LHRH and LH secretion was determined for an extra 1.5 h after the injection.

and jugular serial blood sampling was initiated as soon as one ewe displayed LH levels higher than 1 ng/ml in two consecutive samples collected daily between Day 30 and Day 40. This criterion was used as a predictive index of the onset of activity for the whole group. The third group (n = 6) was sampled when LH secretion was maximum, on Day 74. In all groups, the surgery for implanting the portal sampling apparatus was performed 10-20 days before blood collection. All animals were kept alive after portal blood sampling to determine the time of the LH increase in each group. The experiment was ended on Day 86. The spontaneous secretions of LH and LHRH were studied simultaneously in samples obtained every 10 min for 6 h. In the third group, pulse frequency was expected to be high and amplitude was expected to be low, which would make pulses difficult to detect [17, 18]. Samples were therefore collected for 3 extra hours at a higher frequency (one sample every 5 min) to obtain a better definition of pulses. This change in sampling frequency could not induce any change in LHRH and LH secretion, as blood was drawn continuously at a constant flow rate by remote peristaltic pumps into a fraction collector, both of these being located in another room. After these periods during which spontaneous LH and LHRH secretions were studied, each ewe received a single

1116

VIGUIE ET AL. i.v. bolus of 250 ng of exogenous LHRH (a dose previously established to elicit a normal LH pulse in the ewe [19]) in 1 ml sterile and apyrogen saline. Samples of portal and jugular blood were then taken every 10 min for 1.5 h. This made it possible to assess pituitary responsiveness to LHRH in the experimental conditions. At this time, animals were ideally set up for another study to determine the effect of N-methyl-D,L-aspartic acid (NMDA) on LHRH secretion at these different stages. This experiment is reported in the companion to the present paper [20]. In addition, jugular blood samples were obtained by venipuncture every 10 min for 6 h on all occasions (Day - 1, Day 39, Day 46, and Day 74) in the noncannulated group.

5.0

2.0

1.0

0.5

Hormonal Assays .,

^,%.

^

,..

en

LHRH concentration was determined after extraction of 0.5 ml (or 0.25 ml when the collected volume was not large enough) of portal blood in 2 ml of methanol as described previously [16]. The extraction efficiency by this method is usually 70-80%. LHRH concentrations were measured in duplicate aliquots corresponding to 200 Ril of plasma. Sensitivity (2 SD from buffer controls) was 0.63 0.08 pg/ml. The intraassay coefficient of variation (CV) for two plasma pools was 13.8%, and the interassay CV was 10.6% (12 assays). Melatonin was assayed in duplicate 100-pl aliquots of plasma through use of the RIA method of Fraser et al. [21] with an antibody first raised by Tillet et al. [22]. Sensitivity was 4 + 0 pg/ml. Intraassay CV for two plasma pools averaged 15.7%, and interassay CV for these plasma pools averaged 11.1% (3 assays). LH was assayed in duplicate 100-1zl aliquots of plasma via the RIA of Pelletier et al. [23] as modified by Montgomery et al. [24]. Sensitivity was 0.13 ± 0.03 ng/ml of 1051CY-LH (i.e., 0.27 ng/ml of NIH-LH-S1). The intraassay CV for five plasma pools averaged 9.1%, and the interassay CV for these plasma pools averaged 11.3% (5 assays).

-

E c

-

E a:

5.0'

GROUP 3 (N=6) I

4

2.0

1.0

I

Ii,?

I

Analysis of Data

ID

D+74

0.5

v.v I I

I

-30 -15

I

I

.

I

I

I

0 15 30 45 60 Day of experiment

I

'I

75

90

FIG. 2. Mean (+ SEM) plasma concentration of LH in each group. Shaded area depicts mean values (± SEM) of LH concentrations in the control noncannulated group (n = 5). LH values were calculated after logarithmic transformation. Dashed arrows depict the time of hypophysial blood sampling, and continuous arrows depict insertion of melatonin implants on Day 0 (DO) of the experiment. N indicates the number of animals per group. All ewes were ovariectomized and treated with a subcutaneous implant of 170-estradiol. Blood samples were obtained once a week from Day -35 to Day -8 and then twice a week until the end of the experiment on Day 86.

Detection of pulses. LHRH and LH pulses were identified through use of a modification of the methods of Wallace and McNeilly [25]. Each secretory episode that was composed of two consecutive samples exhibiting a concentration greater than the mean concentration of the two previous samples was considered as a possible pulse. This secretory episode was definitively recorded as a pulse if its maximum concentration was greater than the nadir value of the two points that just preceded it by more than two standard deviations of this nadir. The standard deviation for each nadir was calculated from the regression line fitted to the quality control data (x) plotted against their standard deviations (y). The frequency was defined as the number of identified pulses per collecting period of 6 h. The amplitude was defined as the difference between peak and

1117

MELATONIN AND LHRH PULSATILE SECRETION

nadir values. Identified LHRH and LH pulses were considered as coincident if the LH pulse occurred at the same time as, or one or two samples after, the LHRH pulse. For the extra period of collection on Day 74 with sampling every 5 min for 3 h, pulse frequency was converted into pulses/ 6 h. Detection of the onset of neuroendocrinereproductive activity. For each ewe, the time when circulating LH levels started to rise in samples obtained twice weekly was determined by the first of at least three consecutive values exceeding 1 ng/ml. Statistical tests. The changes in LH levels determined in samples collected twice weekly throughout the experiment were analyzed through use of a 2-factor (within factor: time of experiment; between factor: group) repeated measures ANOVA after logarithmic transformation of data. The times of onset of the reproductive neuroendocrine activity were analyzed by a Mann-Whitney nonparametric test. For each time of portal blood collection, LH pulse frequency in groups 1, 2, and 3 was compared to that of the noncannulated group by a Mann-Whitney unilateral nonparametric test. Changes in LHRH and LH pulse frequency among groups 1, 2, and 3 were analyzed by a Mann-Whitney nonparametric test; these changes between Day 39 and Day 46 for the seven ewes of group 2 that were sampled twice were analyzed by a unilateral Wilcoxon test. Among groups 1, 2, and 3, changes in LHRH and LH pulse amplitude were analyzed by a one-way ANOVA; those between Day 39 and Day 46 for group 2 were examined by a paired t-test. The amplitude of the LH pulse induced by LHRH injection was analyzed by a one-way ANOVA. RESULTS

-15 '

12

0

c

* * *

r

I

I

9

U,

I

I

C

I

3

0 15'

·

12

II**

I

0

I

9

-

I

T

U)

(n -J

6-

-J

3-

I

I

0D-1

D+39 D+46

D+74 D+74 10min 5min

FIG. 3. Mean (+ SEM) LHRH (top) and LH (bottom) pulse frequency during inhibition of LH secretion by long days (Day -1), at two different times of the onset of LH increase after melatonin treatment (Day 39 and Day 46), and during maximal LH secretion (Day 74). For the last period, portal and jugular blood samples were obtained first every 10 min for 6 h and then every 5 min for 3 h. Both LHRH and LH pulse frequencies increased between Day 39 and Day 46. The pulse frequency was higher on Day 74 than on Day -1 and Day 39 (** p < 0.01; * p < 0.05).

6.1 days). Furthermore, the amplitude of the rise was not

PlasmaLevels of Melatonin

different between groups (Fig. 2).

Insertion of melatonin implants caused plasma levels of melatonin to be elevated throughout the 24-h period. Indeed, a large increase in the daytime plasma levels of melatonin (from 14.2 ± 2.6 pg/ml to 187.5 ± 5.7 pg/ml, mean ± SEM) was observed. Nighttime plasma levels before and after the melatonin treatment were 421.0 + 24.8 pg/ml and 473.7 18.2 pg/ml, respectively.

Effects of Melatonin on LHRH and LH Pulsatile Secretion The mean frequency of LH and LHRH pulses in the various groups is presented in Figure 3, and profiles of representative animals are described in Figure 4. In samples obtained every 10 min for 6 h, a dramatic increase in LHRH pulse frequency was observed as LH secretion was stimulated by melatonin (Fig. 3, top). LHRH pulse frequency was low on Day -1 and on Day 39. In group 2, pulse frequency increased 2.5-fold between Day 39 and Day 46 (p < 0.05). Then, a dramatic increase in LHRH pulse frequency was observed on Day 74 (6-fold compared to Day -1,p < 0.01). The changes in LH pulse frequency between the three experimental groups paralleled those of LHRH (Fig. 3, bottom; Fig. 4). LH pulse frequency was low on Day -1 and Day 39. It also increased on Day 46 in group 2 (3-fold; p < 0.05), and a dramatic change was observed on Day 74 (4.5-fold compared to Day -1,p < 0.01, period with one sample/10 min). On each occasion, pulse frequency in those groups was not different from that of the noncannulated

Effects of Melatonin on Mean LH Levels In all groups, LH levels were low at the beginning of the experiment in long days before the insertion of melatonin implants (Fig. 2). These levels remained low (< 1 ng/ml) for 43-53 days after the beginning of melatonin treatment and thereafter increased to values of 2-5 ng/ml. ANOVA did not reveal any differences among groups in the changes in LH concentration over time. The time of onset of LH increase relative to the insertion of implants did not differ among the four groups (group 1: 53.3 - 5.2; group 2: 50.5 ± 3.9; group 3: 46.0 5.8; noncannulated group: 48.6 +

1118

VIGUIE ET AL.

appeared to be coincident by visual inspection, the analysis failed to show full coincidence. Indeed, only 72.2 and 69.1% of the LHRH pulses were followed by LH pulses on Days 46 and 74, respectively. Coincidence between recorded pulses was not improved by sampling every 5 min on Day 74 (70.0%). The percentage of LH pulses preceded by LHRH pulses was also reduced (42.0, 86.4, and 81.5% on Day 46, Day 74 [every 10 min], and Day 74 [every 5 min], respectively). The amplitude of LHRH pulses was highly variable within groups (for instance, between 5.6 and 59.0 pg/ml on Day 74), and no difference was observed between the different 5.3, and 19.5 ± 8.1 stages (19.6 ± 5.6, 49.6 ± 18.3, 16.5 pg/ml on Day -1, Day 39, Day 46, and Day 74, respectively). The amplitude of LH pulses did not vary signifi0.3, 1.1 cantly among the different stages (1.6 ± 0.4, 2.1 ± 0.4, and 1.6 + 0.3 ng/ml on Day -1, Day 39, Day 46, and Day 74, respectively).

E

E2 =

1r

0

E

R I I E I

:

Stimulation by Exogenous LHRH The i.v. injection of 250 ng of exogenous LHRH caused LHRH levels in portal blood to peak to values of 19.5 ± 6.1 (Day -1), 28.8 + 3.8 (Day 39), 30.9 + 5.0 (Day 46), and 26.2 + 4.1 pg/ml (Day 74). The injection resulted in an increase in LH secretion in all the animals, whatever their group. The amplitude of the LH response did not differ between groups (3.66 ± 0.90, 4.61 ± 0.53, 3.37 0.62, 4.63 + 1.23 ng/ml on Day -1, Day 39, Day 46, and Day 74, respectively).

I

DISCUSSION

FIG. 4. Examples of individual profiles of LHRH (top profiles) and simultaneous LH (bottom profiles) secretions at different times relative to the insertion of melatonin implants on Day 0. Blood samples were obtained every 10 min for 6 h on each occasion. An extra period of 3 h (shaded area) with a sample every 5 min was added on Day 74. The closed symbols depict detected pulses.

control group (group 4; 0.40 + 0.20, 1.80 ± 1.28, 1.80 + 0.91, 7.60 ± 1.08 on Days -1, 39, 46, and 74, respectively). As expected, when both LHRH and LH pulse frequencies were high (Day 74; Fig. 4), the number of pulses detected was higher in samples collected every 5 min than in those collected every 10 min. This sampling frequency made the changes resulting from the action of melatonin even more dramatic. Frequency of detected LHRH and LH pulses increased to 13.30 ± 1.33 and 9.00 ± 1.77 pulses/6 h, respectively. Coincidence between LHRH and LH pulses was very good on Day -1 and Day 39 (only one LHRH pulse did not have a coincident identified LH pulse). In the other two situations, although obvious pulses in LH and LHRH secretion

The present experiment, in which LHRH and LH secretion was simultaneously measured, clearly indicates that the increase in LH pulse frequency observed 48 days after the beginning of a melatonin treatment results from a simultaneous increase in LHRH pulse frequency. Ovariectomized ewes treated with a constant physiological level of estradiol are characterized by a low frequency of LH pulses during the anestrous season or when exposed to long days. Lengthening the daily duration of melatonin presence by exposure to short days or by short-day-like melatonin treatment causes a stimulation of LH pulsatile secretion that is observed only after 40-60 days of treatment [6]. Our finding clearly demonstrates that the increase in LH secretion during melatonin treatment did not result from a change in pituitary responsiveness to LHRH but is a consequence of an increase in LHRH pulse frequency. This is consistent with the coincidence between LHRH and LH pulsatile secretion described in many well-established physiological situations, including anestrus and the breeding season [9, 10, 26-28]. However, these earlier studies did not determine whether or not the increase in LH secretion coincided with that of LHRH. A delay in the LHRH response could thus have been responsible, at least in part, for the

MELATONIN AND LHRH PULSATILE SECRETION

40- to 60-day latency before LH secretion increases. Our present experiment has extended these previous results, as it focused on the development of the response and particularly on the period when the initial increase in LH secretion occurred (between Day 39 and Day 46 in group 2). It clearly shows that the increase in LHRH and LH pulse frequency occurred simultaneously within the limits of our design (less than 1 wk). The functional cause of the long latency for melatonin to change LH pulse frequency lies, therefore, only in the control of LHRH secretion and not in a temporary lack of response of the pituitary to LHRH. The increases in LHRH and LH pulse frequencies after a melatonin treatment were simultaneous, but recorded LHRH and LH pulses did not always coincide. This discrepancy was noted on Day 46 and Day 74. On Day 74, pulses were more difficult to detect because of their high frequency and the elevated baseline. Specifically, although pulses were similar in amplitude to those observed on the other days, they were superimposed on an elevated baseline that was characterized by a greater variability, making the threshold for detection of a pulse more difficult to reach. The problem of detection of low-amplitude pulses was made more acute by use of the technique of portal blood sampling, which decreases the quantity of LHRH reaching the anterior pituitary and then the amplitude of LH pulses [16, 26, 27]. This problem was more acute on Day 46, since the animals were sampled for the second time for portal blood on that day. Indeed, repeated lesions of portal vessels could be responsible for the large decrease in amplitude of LHRH and LH pulses observed in some animals between Day 39 and Day 46. Importantly, the discrepancy between LHRH and LH pulses was not observed when LH secretion was still inhibited; this led us to exclude the existence of low-frequency LHRH pulses that are not followed by LH pulses at that time. Furthermore, it did not prevent demonstration of the simultaneous increases in LHRH and LH pulsatile secretion between Day 39 and Day 46. It is noteworthy that despite the lesion caused by the technique, the response of the hypothalamo-pituitary gonodal axis to melatonin was not altered, since the LH plasma levels increased at the same time in the portal-sampled groups and the noncannulated control group. In ovariectomized ewes bearing a subcutaneous implant of estradiol, sex steroid levels remained constant; this absence of variation in steroid milieu permitted us to isolate the effects of season or melatonin from those of changing steroid levels on the hypothalamo-pituitary axis. LH secretory changes in ovariectomized and estradiol-treated ewes reflect mainly a shift in estradiol negative feedback on LHRH secretion, which constitutes the key neuroendocrine mechanism responsible for seasonal reproduction [1]. It has been shown previously in castrated animals supplemented with sex steroid [5, 9, 29] that the modification of sensitivity to sex steroid is expressed through seasonal variations of LH pulse frequency. Our findings extend these previous re-

1119

sults by suggesting that the decrease in the negative feedback of estradiol on LHRH pulsatile secretion occurring at the initiation of the sexual activity is linked to a longer daily presence of high melatonin levels. The modification of LHRH pulse frequency induced by melatonin indicates that the action of this hormone involves central mechanisms. The long latency between a change in the melatonin profile and the corresponding response in LHRH or LH secretion raises many questions about the mode of action of melatonin. In particular, does a change in melatonin profile only initiate a cascade of events leading to a change in LHRH secretion after a 40-60-day latency, or does the melatonin profile need to remain present throughout that latency in order for the response to be completed? In sheep, some data suggest that the melatonin signal needs to be present during at least 30 days for a normal expression of the response and therefore tend to support the second hypothesis [30,31]. Localizing the sites of action is still a matter of conjecture. The largest density of binding sites is found in the pars tuberalis of the pituitary [32-36]. However, melatonin microimplants apposed near the pituitary stalk fail to induce short-day-like effects on LH secretion [37]. Another potential site of action could be the LHRH cell bodies; but insertion of melatonin microimplants in the preoptic area, which contains most of the LHRH cell bodies in ovine species [38,39], was found to have no effect on LH secretion [8, 40]. In contrast, the same implants placed in the mediobasal hypothalamus reproduce the effects of short days on LH secretion [8, 40]. The action of melatonin on LHRH neurons seems to involve interneurons. In particular, dopaminergic neurons seem to be implicated in this regulation. Indeed, photoperiod, most likely via melatonin, modifies dopamine content and tyrosine hydroxylase (the rate-limiting enzyme of catecholamine synthesis) activity in the median eminence [41,42]. Furthermore, systemic injection or local hypothalamic administration of a dopamine antagonist (pimozide), and neurotoxic lesions of a dopaminergic cell group (A15) in photoperiodically inhibited ewes, have been shown to stimulate LH secretion [43-45]. Neuroexcitatory amino acids could also be implicated in the regulation of LHRH secretion by melatonin. Indeed, injections of NMDA stimulate LHRH secretion and have been shown to affect LH secretion in sheep differentially according to the season [46]. Thus, this system could mediate a step in the sequence of activation of LHRH secretion by melatonin [20]. The present experiment demonstrates that the short-daylike effect of melatonin implants on LH secretion results from a stimulation of LHRH secretion. Moreover, as for LH, the action of melatonin on LHRH secretion requires a delay of 40-60 days. The identification of the central nervous mechanisms responsible for this long-term action is critical to an understanding of photoperiodic regulation of reproduction.

1120

VIGUIE ET AL.

ACKNOWLEDGMENTS We wish to thank Drs. M. Caldani, P. Chemineau, FJ. Karsch, and D.C. Skinner for help in the design of the study and comments on the manuscript; B. Delaleu, F. Maurice, and C. Fagu for their technical help; Mr. G. Durand and F. Paulmier and their team for assistance in the animal experimentation; and Mrs. A. Bouroche for revision of the English manuscript.

REFERENCES 1. Legan SJ, Karsch FJ. Photoperiodic control of seasonal breeding in the ewe, modulation of the negative feedback action of estradiol. Biol Reprod 1980; 23:10611068. 2. Lincoln GA, Short RV. Seasonal breeding: nature's contraceptive. Recent Prog Horm Res 1980; 36:1-52. 3. Bittman EL, Dempsey RJ, Karsch FJ. Pineal melatonin secretion drives the reproductive response to daylength in the ewe. Endocrinology 1983; 113:2276-2283. 4. Arendt J. Role of the pineal gland and melatonin in seasonal reproductive function in mammals. Oxf Rev Reprod Biol 1986; 8:266-320. 5. Karsch FJ, Bittman EL, Foster DL, Goodman RL, Legan SJ, Robinson JE. Neuroendocrine basis of seasonal reproduction. Recent Prog Horm Res 1984; 40:185232. 6. Bittman EL, Kaynard AH, Olster DH, Robinson JE, Yellon SM, Karsch FJ. Pineal melatonin mediates photoperiodic control of pulsatile luteinizing hormone secretion in the ewe. Neuroendocrinology 1985; 40:409-418. 7. O'Callaghan D, Karsch FJ, Bolland MP, Roche JF. What photoperiodic signal is provided by continuous-release melatonin implant? Biol Reprod 1991; 45:927933. 8. Malpaux B, Daveau A, Maurice F, Gayrard V, ThiryJC. Short-day effects of melatonin on luteinizing hormone secretion in the ewe: evidence for central sites of action in the mediobasal hypothalamus. Biol Reprod 1993; 48:752-760. 9. Barrell GR, Moenter SM, Caraty A, Karsch FJ. A marked seasonal change in GnRH secretion in the ewe. Biol Reprod 1992; 46:1130-1135. 10. ClarkeIJ, Cummins JT. The temporal relationship between gonadotropin releasing hormone (GnRH) and luteinizing hormone (LH) secretion in ovariectomized ewes. Endocrinology 1982; 111:1737-1739. 11. Clarke 1J,Thomas GB, Yao B, Cummins JT. GnRH secretion through the ovine estrous cycle. Neuroendocrinology 1987; 46:82-88. 12. Caraty A, Moenter SM, Locatelli A, Martin GB, Karsch FJ. Gonadotropin releasing hormone (GnRH) secretion during the preovulator luteinizing hormone surge in the ewe. In: Bouchard P, Haour F, Franchimont P, Schatz B (eds.), Recent Progress on GnRH and Gonadal Peptides. Paris: Elsevier; 1990: 59-70. 13. Legan SI, Karsch FJ, Foster DL. The endocrine control of seasonal reproductive function in the ewe: a marked change in response to the negative feedback action of estradiol on luteinizing hormone secretion. Endocrinoklogy 1977; 101:818824. 14. Karsch FJ, Bittman EL, Robinson JE, Yellon SM, Wayne NL, Olster DH, Kaynard AH. Melatonin and photorefractoriness: loss of response to the melatonin signal leads to seasonal reproductive transitions in the ewe. Biol Reprod 1986; 34:265274. 15. Malpaux B, Moenter SM, Wayne NL, Woodfill CJI, Karsch HJ. Reproductive refractoriness of the ewe to inductive photoperiod is not caused by an alteration of the circadian secretion of melatonin. Neuroendocrinology 1988; 48:264-270. 16. Caraty A, Locatelli A. Effect of time after castration on secretion of LHRH and LH in the ram. J Reprod Fertil 1988; 82:263-269. 17. Robinson JE, Radford HM, Karsch H.Seasonal changes in pulsatile luteinizing hormone (LH) secretion in the ewe, relationship of frequency of LH pulses to day length and response to estradiol negative feedback. Biol Reprod 1985; 33:324334. 18. Karsch F, Dahl GE, Evans NP, Manning JM, Mayfield KP, Moenter SM,Foster DL. Seasonal changes in gonadotropin releasing hormone secretion in the ewe: alteration in response to the negative feedback action of estradiol. Biol Reprod 1993; 49:1377-1383. 19. Clarke 1J,Cummins JT. Direct pituitary effects of estrogens and progesterone on gonadotropin secretion in the ovariectomized ewe. Neuroendocrinology 1984; 39:267 274. 20. Viguie C, Carat, A, Locatelli A, Malpaux B. Regulation of LHRH secretion by melatonin in the ewe. II. Changes in NMDA-induced LHRH release during the stimulation of LH secretion by melatonin. Biol Reprod 1995; 52:1156-1161. 21. Fraser S, Cowen P, Franklin M, Franey C, Arendt J. Direct radioimmunoassay for melatonin in plasma. Clin Chem 1983; 20:396-397. 22. Tillet Y, RavaultJP, Selve C, Evin G, Castro B, Dubois MP. Conditions d'utilisation d'anticorps specifiques pour la visualisation immunohistochimique de la serotonine

23.

24.

25.

26.

27. 28.

29

30.

31.

32. 33.

34. 35. 36.

37.

38.

39. 40.

41.

42.

43.

44.

45.

46.

et de la melatonine dans la glande pineale du mouton. C R Acad Sci (Paris) 1986; 303:77-82. Pelletier J, Kann J, Dolais J, Rosselin G. Dosage radioimmunologique de l'hormone lutinisante plasmatique chez le mouton Mise au point de la technique de dosage. C R Acad Sci (Paris) 1968; 266:2291-2294. Montgomery GW, Martin GB, Pelletier J. Changes in pulsatile LH secretion after ovariectomy in Ile-de France ewes in two seasons. J Reprod Fertil 1985; 73:173183. WallaceJM, McNeilly AS. Changes in FSH and the pulsatile secretion of LH during treatment of ewes with bovine follicular fluid throughout the luteal phase of the oestrous cycle. J Endocrinol 1986; 111:317-327. Levine JE, Pau KYF, Ramirez VD, Jackson GL. Simultaneous measurement of luteinizing hormone-releasing hormone and luteinizing hormone release in unanesthetized, ovariectomized sheep. Endocrinology 1982; 111:1449-1455. Clarke IJ,Thomas GB, Yao B, Cummins JT. GnRH secretion throughout the ovine estrous cycle. Neuroendocrinology 1987; 46:82-88. Caraty A, Locatelli A, Martin GB. Biphasic response in the secretion of gonadotrophin releasing hormone in ovariectomized ewes injected with oestradiol. J Endocrinol 1989; 123:375-382. D'Occhio MJ, Schanbacher BD, Kinder JE. Profiles of luteinizing hormone, follicle-stimulating hormone, testosterone and prolactin in rams of diverse breeds: effects of contrasting short (8L:16D) and long (16L:8D) photoperiods. Biol Reprod 1984; 30:1039 1054. O'Callaghan D, Karsch FJ, Bolland MP, Roche JF. Role of short days in timing the onset and duration of reproductive activity in ewes under artificial photoperiods. Biol Reprod 1991; 44:23-28. Robinson , Wallace JM, Aitken RP, Wigzell S. Effect of duration of melatonin treatment on the onset and duration of oestrous cyclicity in ewes. J Reprod Fertil 1992; 95:709-717. de Reviers MM, Ravault JP, Tillet Y, Pelletier J. Melatonin binding sites in the sheep pars tuberalis. Neurosci Lett 1989; 100:89-93. Morgan PJ, Williams LM, Davidson G, Lawson W, Howell H. Melatonin receptors on ovine pars tuberalis: characterization and autoradiographical localization. J Neuroendocrinol 1989; 1:1-4. Bittman EL, Weaver DR. The distribution of melatonin binding sites in neuroendocrine tissues of the ewe. Biol Reprod 1990; 43:986-993. de Reviers MM, Tillet Y, Pelletier J. Melatonin binding sites in the brain of sheep exposed to light or pinealectomized. Neurosci Lett 1991; 121:17-20. Stankov B, Cozzi B, Lucini V, Scaglione F, Fraschini F. Characterization and mapping of melatonin receptors in the brain of three mammalian species: rabbit, horse and sheep. Neuroendocrinology 1991; 53:214-221. Malpaux B, Daveau A, Maurice F, Locatelli A, Thi&ry JC. Evidence that melatonin binding sites in the pars tuberalis do not mediate the photoperiodic actions of melatonin on LH and prolactin secretion in ewes. J Reprod Fertil 1994; 101:625632. Lehman MN, Robinson JE, Karsch FJ, Silverman AJ. Immunocytochemical localization of luteinizing hormone-releasing hormone (LHRH) pathways in the sheep brain during anestrus and the mid-luteal phase of the estrous cycle. J Comp Neurol 1986; 244:19-35. Caldani M, Batailler M, Thiery JC, Dubois MP. LH-RH immunoreactive structures in the sheep brain. Histochemistry 1988; 89:129-139. Lincoln GA, Maeda KI. Reproductive effects of placing microimplants of melatonin in the mediobasal hypothalamus and preoptic area in rams. J Endocrinol 1992; 132:201-215. ThiryJC. Monoamine content of the stalk-median eminence and hypothalamus in adult female sheep as affected by daylength. J Neuroendocrinol 1991; 3:407411. Vigui C, Thibault J, Thiry JC, Tillet Y, Malpaux B. Increase in tyrosine hydroxylase activity in the stalk-median eminence during photoperiodic inhibition of LH secretion in the ewe. Ann Endocrinol 1993, 54:34. Meyer SL, Goodman RL. Neurotransmitters involved in mediating the steroiddependent suppression of pulsatile luteinizing hormone secretion in anestrous ewes: effects of receptor antagonists. Endocrinology 1985; 116:2054-2061. Thiry JC, Martin GB, Tillet Y, Caldani M, Quentin M,Jamain C, Ravault JP. Role of hypothalamic catecholamines in the regulation of luteinizing hormone and prolactin secretion in the ewe during seasonal anoestrus. Neuroendocrinology 1989: 49:80-87. Havern RL, Whisnant SC, Goodman RL. Hypothalamic sites of catecholamine inhibition of luteinizing hormone in the anestrous ewe. Biol Reprod 1991; 44:476482. Lincoln GA, Wu FCW. Luteinizing hormone responses to N-methyl-,i.-aspartate during a photoperiodically-induced reproductive cycle in the ram. J Neuroendocrinol 1991; 3:309-317.