| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
research-article |
UMR Physiologie de la Reproduction et des Comportements, INRA/CNRS/Université Tours/Haras Nationaux, 37380 Nouzilly, France
ABSTRACT
For a better understanding of the mechanisms that lead to the preovulatory GnRH/LH surge and estrus behavior, the minimum estradiol (E) requirements (dose and duration) to induce each of these events were determined and compared between two breeds of ewes having either single (Ile de France) or multiple (Romanov) ovulations. The ewes were initially studied during a natural estrus cycle, and were then ovariectomized and run through successive artificial estrus cycles. For these artificial cycles the duration and amplitude of the follucular phase E increase were manipulated by E implants. Under all conditions, the onset of estrus behavior was similar in the two breeds, although its duration was longer in Romanov ewes. While a moderate E signal (6 cm for 12 h) induced an LH surge in 10/10 Ile de France ewes, a larger E signal (12 cm for 12 h) was minimally effective in Romanov ewes (4/10). Additional studies revealed that a small E signal (3 cm for 6 h) induced full estrus behavior in all Romanov ewes but was completely ineffective in Ile de France animals (0/10). Higher doses and mostly longer durations of the E signal (12 cm for 24 h) were required to induce a surge in all the Romanov ewes. These results demonstrate a clear difference in the E requirement for the induction of estrus behavior and the LH surge between breeds of ewes that have different ovulation rates. These data provide compelling evidence that, in one breed, the neuronal systems that regulate both events require different estrogen signals.
ewes, hypothalamus, ovulation, pituitary, sexual behavior
It is well established that a rise in circulating estradiol (E) acts centrally upon the GnRH neurosecretory system to increase GnRH secretion. GnRH, in turn, induces the LH surge, ovulation, and the subsequent luteal phase [1, 2]. This positive feedback action of E influences both the pituitary gland, enhancing responsiveness to GnRH [3], and the hypothalamus, stimulating a large and sustained GnRH surge. The increase in the concentration of E that occurs in the midfollicular phase is also responsible for changes in sexual behavior. E induces estrus and appears to be a key factor in the coordination of endocrine and behavioral changes that ensure reproductive success.
Variations in the response to E in terms of the induction of the LH surge occur between breeds. For example, studies of the E-induced discharge of LH in ovariectomized (OVX) ewes show that breeds with high ovulation rates (3 to 5 ovulations) are less sensitive to the positive and negative feedback effects of estrogen than those with low ovulation rates (1 or 2 ovulations) [4–6]. E concentration is related to the number of growing follicles, and the maximum value of E measured before the LH peak is significantly higher in a high-fecundity breed, e.g., Romanov, than in a low-fecundity breed, e.g., Ile de France [7]. In high-fecundity breeds, although the increase in E secretion is greater, the interval between luteal regression and the LH surge is longer than in breeds with lower fecundity [5]. Breed differences are also known to exist for E-induced estrus behavior, although in this case, the differences seem to correlate with the levels of peripheral E. For instance, a positive association between prolificacy and the duration of natural estrus has been reported for the Merino, Lacaune, Finnish Landrace, and Romanov breeds [5, 8–10]. Accordingly, the duration of estrus has been positively correlated to the dose of E administered to OVX ewes [11, 12].
While there is some evidence in the literature on breed differences in sensitivity to E in the induction of the GnRH/LH surge and estrus behavior in ewes, the cell types that are responsive to E stimulation, as well as the pathways involved in these responses are still ill-defined. Previous studies from our laboratory have shown that the mediobasal hypothalamus (MBH) is the primary site at which E controls estrus behavior and the preovulatory LH surge in the ewe [13]. This indicates that within this structure, there are some cell populations that are able to read the E signal coming from the growing follicles and to activate the cascade of events involved in the activation of the surge-generating and estrus behavior-generating system.
Recent studies have examined the duration and amount of E required to induce the GnRH/LH surge in the ewe. These studies show that E is not needed for the entire presurge period [14, 15]. In fact, the period during which E is required is relatively short and localized before the onset of the GnRH/LH surge. Once the GnRH/LH surge has begun, an elevated level of E is no longer required to maintain the GnRH surge [14]. However, it must be pointed out that the E requirement for the induction of the GnRH/LH surge reported for the Suffolk ewe [14] is quite different from that reported for the Ile de France ewe [15]. This discrepancy may reflect methodological differences between laboratories or may be related to the different breeds used in these experiments. Moreover, the latter study also shows that within the same breed, the E concentration required to induce the GnRH/LH surge was higher than that needed to induce estrus behavior [15]. It is likely that within the MBH, the mechanisms that induce these two events are different. The present study indicates that the threshold doses of E needed to activate the two mechanisms are also different. Taken together, these results indicate that the differences in E sensitivity between and within a breed need to be reconsidered.
The aim of the present study was to determine if the period during which the E signal is read by the surge-generating and estrus behavior-generating systems is constant or differs within and between breeds that have different ovulation rates. To this end, the E requirement (duration and dose) to induce each of these events was tested in breeds with one (Ile de France) or multiple (Romanov) ovulations. The ewes were initially studied during a natural estrus cycle, and were then ovariectomized and run through successive artificial estrus cycles. For these artificial cycles the duration and amplitude of the follicular phase E increase were manipulated by E implants.
Experiments were conducted using mature Ile de France and Romanov ewes, which were maintained under normal husbandry conditions at INRA (Nouzilly, France) during the breeding season over a period of two years (from October to December 2003 and from November to December 2004). All experimental procedures were performed in accordance with local animal regulations (Authorization no. A38801 of the French Ministry of Agriculture).
Experiment 1. The estrus cycles of 10 Ile de France and ten Romanov ewes were synchronized using an intravaginal progesterone delivery device (CIDR; InterAg, Hamilton, New Zealand) for 12 days in October 2003. The LH concentrations were determined for plasma samples collected every 2 h from 26 to 60 h after CIDR withdrawal. The E concentration was measured in plasma collected from 2 to 74 h after CIDR withdrawal, and estrus behavior was measured from 0 to 98 h after CIDR withdrawal (Fig. 1A).
|
Experiment 2. Two weeks later, these same ewes were ovariectomized and run through three artificial cycles using peripheral implants of E and progesterone (Fig. 1B). Immediately after ovariectomy, a progesterone CIDR and a 1 cm subcutaneous Silastic E implant (Dow Corning Corp., Midland, MI) were inserted, to mimic the steroidal milieu of the midluteal phase of the estrus cycle. After 12 days, the progesterone CIDR was removed to stimulate luteal regression, followed 24 h later by subcutaneous insertion of E implants (2 x 3 cm) for 3, 6, and 12 h using a cross-over design.
The LH concentrations were determined for plasma samples collected every 30 min for 1 h before E insertion and every 3 h from 6 to 36 h after E insertion. The E concentrations were measured for plasma collected before the insertion of additional E implants, 2 h after insertion, and at the time of E implant removal (i.e., at 3, 6, and 12 h). Estrus behavior was measured from 0 to 72 h after E treatment (Fig. 1B).
Experiment 3. Since only a few LH surges were observed in the Romanov ewes in experiment 2, all of the animals were run through an additional artificial cycle, in which the E treatment was increased to induce higher E plasma concentrations using E implants (4 x 3 cm) inserted s.c. for 12 h. The LH concentrations were determined for plasma samples collected every 30 min for 1 h before E insertion and every 3 h from 6 to 36 h after E treatment. The E concentrations were measured in plasma collected before insertion of the additional E implants, 2 h after insertion, and at the time of the removal of the E implants (i.e., at hour 12).
The experiments performed in year 1 revealed that all E treatments induced estrus behavior in all OVX ewes, although estrus behavior was significantly longer in the Romanov ewes. In contrast, the Romanov ewes needed more E to induce an LH surge than the Ile de France ewes. To clarify these breed differences, a second series of studies was performed a year later (during the breeding season in November 2004) and the animals were challenged with wider ranges of E.
Experiment 4. Ten Ile de France and ten Romanov ewes were ovariectomized and run through two artificial cycles in a cross-over design (as described in experiment 2), and 24 h after CIDR withdrawal, E implants (4 x 3 cm) were inserted for 12 or 24 h. The LH concentrations were determined for plasma samples collected every 30 min for 1 h before E insertion and every 3 h from 6 to 36 h after E treatment.
Experiment 5. Animals from experiment 4 were run through an additional artificial cycle in which a very low dose of E was given to both breeds, i.e., a 3 cm E implant for 6 h. Receptivity was recorded from 0 to 144 h after E implant insertion.
Since experiment 4 revealed that only 24 h with the E implants (4 x 3 cm) resulted in all of the Romanov ewes exhibiting an LH surge, a final experiment was performed to determine if a similar duration but a lower E dose was sufficient to induce an LH surge.
Experiment 6. Romanov ewes (n = 9) were ovariectomized and immediately treated with a CIDR and a subcutaneous 1 cm Silastic E implant. The ewes were then run through two artificial cycles, and 24 h after progesterone removal, E implants (4 x 3 cm for cycle 1 and 2 x 3 cm for cycle 2) were inserted for 24 h. The LH concentrations were determined for plasma samples collected every 2 h from 13 to 41 h after E treatment.
Estrus behavior was quantified for each ewe using the standardized procedure described by Fabre-Nys and Venier [16]. Receptivity was measured using a receptivity index (RI), calculated as the number of immobilizations in response to male nudges over the total number of nudges x 100. The test lasted at least 2 min and comprised a minimum of 10 nudges. Ewes were considered to be receptive when the RI was above 80%. The receptivity duration was calculated as the interval from the first test in which a ewe was receptive until the last one plus 6 h. Latency was calculated as the interval from estrus onset minus one half of the interval of the previous test.
Luteinizing hormone. Blood samples were assayed for LH in duplicate 100 µl aliquots of plasma by RIA [17]. The intraassay and interassay coefficients of variation for three plasma pools were 8.5% and 9.5%, respectively. Assay sensitivity was 0.16 ± 0.05 ng/ml (four assays) of standard 1051-CY-LH (equivalent to 0.3 ng/ml of NIH-LH-S1).
Estradiol. The E concentrations were estimated using the 125I E2 Diasorin RIA kit (Sorin Diagnostic, Antony, France), with a slight modification for the analysis of E concentrations in ovine plasma. Duplicate 200-µl aliquots of plasma samples were extracted in 3 ml of ethyl-acetate/cyclohexane (V/V) mixed for 5 min. After 2 h, the tubes were centrifuged for 15 min and frozen in liquid nitrogen. The solvent layer was decanted, transferred into glass tubes, and evaporated under nitrogen. Standards and samples were reconstituted in 150 µl of 0.1 M PBS (pH 7.4) that contained 1 g/ml BSA and 1 mM EDTA. Samples were incubated in primary antibody (150 µl) for 24 h at 4°C and 125I estradiol (100 µl, 30 000 cpm) was added to each tube for 24 h at 4°C. The secondary antibody (500 µl of precipitating solution) was added, mixed, and incubated at 4°C for 1 h. Subsequently, 2 ml of 0.025 M Tris buffer (pH 7.4) was added, the tubes were centrifuged at 2500 x g for 30 min, the supernatant was discarded, and the radioactivity was measured.
Following these modifications, the cross-reactivities were found to be less than 0.5% for estrone and estriol and less than 0.1% for ethinylestradiol, progesterone, testosterone, androstenediol, estradiol-3-glucuronide, and estradiol-17-glucuronide. The assay sensitivity was 0.78 pg/ml. At 4.8 pg/ml (nine assays), the intraassay and interassay coefficients of variation were 10.7% and 7.1%, respectively. Recovery was determined by adding known amounts of E to four ovine plasma samples with low endogenous E concentrations. The percentage recovery, calculated as recovered/expected, was 90.5%. The linearity of the dilutions was tested by assaying four plasma samples with high E levels that were diluted with ovine plasma free of E. The observed/expected ratios were between 93% and 100% at 1:2, 1:4, and 1:8 dilutions.
Comparisons of the behavioral and LH data were performed using the StatXact 5 software (Cytel Software Corporation, Cambridge, MA). Females with a receptivity index above 80% were considered to be expressing estrus behavior. As the behavioral data did not follow a normal distribution, these data were analyzed using nonparametric tests. The latency and duration of estrus between different groups, for each E treatment in experiments 1, 2, and 5, were compared respectively using a permutation with general score test for two independent samples and the nonparametric Fisher exact test. For comparisons within the same breed, we used the Friedman test for related samples with Bonferroni correction. A permutation test for paired samples was used to compare groups two by two within the same breed. Latencies and duration of estrus behavior are given as medians [low–high quartiles].
An LH surge (experiments 1, 2, 3, 4, and 6) is defined as a sustained increase in LH concentration (twice the mean concentration preceding the administration of E and exceeding 10 ng/ml in at least one sample). The onset of the LH surge was designated as the first time-point at which the LH concentration exceeded twice the basal level.
In experiments 1, 2, 3, 4, and 6, the numbers of LH surges were compared between breeds using the Fisher exact nonparametric test. In experiment 2, for each breed, the numbers of ewes that exhibited an LH surge between the treatments (3, 6, and 12 h) were compared using a Cochran exact nonparametric test. When a significant difference was detected, groups were compared using a nonparametric Mc Nemar test with Bonferroni correction. In experiments 1, 2 (for each E treatment), 3 and 4, comparisons of amplitude and onset of the LH surge between breeds were carried out with a nonparametric exact permutation test. Estradiol concentrations produced by the 6 and 12 cm implants in Ile de France and Romanov ewes were compared using a permutation with general score test.
For the two breeds, the onset of estrus after CIDR withdrawal was not significantly different (Fig. 2). Estrus duration was longer (P < 0.001) in Romanov (42 h [39–54 h]) than in Ile de France (24 h [14–29.8 h]) ewes. All ewes of both breeds exhibited an LH surge (Fig. 3A). The onset of the surge occurred later (P < 0.01) in the Romanov (39.4 ± 1.2 h) than in the Ile de France (32.2 ± 1.4 h) ewes, and the surge amplitude was smaller (P < 0.001) in the Romanov (46.4 ± 14.5 ng/ml) than in the Ile de France (91.1 ± 5.7 ng/ml) ewes. The plasma E concentrations before the onset of the LH surge (mean of two values immediately preceding the onset) were significantly higher (P < 0.01) in the Romanov (10.39 ± 1.02 pg/ml) than in the Ile de France (5.78 ± 0.34 pg/ml) ewes (Fig. 3B).
|
|
After ovariectomy, E implants induced a rapid increase in the plasma E concentration, which reached a plateau 2 h after implant insertion and stayed at that level until removal. The E concentrations achieved by the 6 cm implant averaged 7.9 ± 0.1 pg/ml and 9.5 ± 0.7 pg/ml (P > 0.05; Fig. 3B) for the Ile de France and Romanov ewes, respectively.
All ewes showed estrus behavior (Fig. 4A), regardless of E treatment (6 cm for 3, 6, and 12 h). There was no difference between or within the groups for estrus latency, and no difference for estrus duration was found within either breed. However, Romanov ewes had a longer duration of estrus behavior than Ile de France ewes when exposed to the 6 h and 12 h E treatments (Fig. 4B).
|
While 12 h of 6-cm E induced an LH surge in all the Ile de France ewes, only 20% of the Romanov ewes showed an LH surge (Table 1). There was a significant difference (P < 0.05) between the breeds in terms of surge onset, as the Ile de France ewes showed an LH peak earlier than the Romanov females (14.4 ± 2.1 h vs. 19.5 ± 1.8 h). There was no difference in LH surge amplitude between the two breeds.
|
An LH surge was observed in all Ile de France ewes but in only 4/10 Romanov ewes (P < 0.05; Table 1). The LH surge occurred later in the Romanov ewes than in the Ile de France ewes (19.5 ± 1.2 h vs. 15.7 ± 1.2 h; P < 0.05) and with lower amplitude (17 ± 0.9 ng/ml vs. 59 ± 11.2 ng/ml; P < 0.05). The E concentration achieved with the 12 cm E implant averaged 17.6 ± 2.3 pg/ml and 20.9 ± 1.9 pg/ml (P > 0.05) for the Ile de France and Romanov ewes, respectively.
A 24 h exposure to the 12 cm E implant was essential to induce an LH surge in all the Romanov ewes, whereas only 12 h of E exposure was sufficient to induce an LH surge in all the Ile de France ewes (Fig. 3A). Although all of the Ile de France ewes showed an LH surge, only 50% of the Romanov ewes had an LH surge when treated with the 12 cm E implant for 12 h (P < 0.05). Only when the E treatment duration was increased to 24 h, was an LH surge induced in 90% of the Romanov ewes. LH surge onset was earlier in the Ile de France ewes for the 12 cm/24 h E treatment (16.4 ± 2.4 h vs. 21.1 ± 3.0 h, P < 0.05), although no significant difference was observed for the 12 cm/12 h E treatment (16 ± 1.6 h vs. 18 ± 1.9 h). For the 12 cm/12 h E treatment, a significant (P < 0.05) effect on surge amplitude was observed (52.2 ± 7.4 vs. 31.8 ± 7.7 ng/ml for Ile de France vs. Romanov). Within one breed, for the 12 h and 24 h E treatments, there were no differences in surge onset, amplitude or frequency.
E treatment (3 cm/6 h) induced estrus behavior in all the Romanov ewes but in none of the Ile de France ewes (Fig. 5). The latency of onset of estrus behavior after E treatment commenced was 13 h [7.5–33 h], and estrus behavior duration was 72 h [9–121 h].
|
Exposure to 6 cm of E implant for 24 h induced a surge in 6/9 Romanov ewes. In the three remaining ewes, a clear increase in LH secretion, which did not meet the surge criteria, was observed. As in experiment 4, 24 h of exposure to 12 cm of E induced an LH surge in all the Romanov ewes. There was no significant difference in the proportion of females showing an LH surge between the two treatments (Table 1). The LH surge onset was earlier in the 12 cm E treatment than in the 6 cm E treatment but there was no significant difference between the two groups (Table 1). However, a significant (P < 0.05) difference was observed in the amplitude of the LH surge between the two treatments (29.1 ± 6.6 vs. 61.2 ± 10.3 ng/ml for 6 cm/24 h vs. 12 cm/24 h).
The present findings clearly show a difference in the E requirement to induce estrus behavior and preovulatory LH surge between breeds that are known to have different ovulation rates. While a very small E signal (3 cm/6 h) is sufficient to induce full estrus behavior in Romanov ewes, the same treatment has no effect on Ile de France ewes. In contrast, a much higher level of E is required to induce the LH surge in the Romanov breed than in the Ile de France breed.
During the natural cycle, estrus behavior starts at the same time in intact females of both breeds but lasts longer in Romanov ewes, which supports the general view that estrus duration is positively correlated with litter size [8, 9]. Interestingly, our results show that the difference in estrus duration between breeds persists when OVX females are exposed to the same exogenous E signal. Therefore, the difference in estrus duration cannot be explained by a difference in the plasma levels of E, as previously postulated [7].
In both intact and OVX ewes, no difference was found in the onset of sexual behavioral events between these two breeds, which does not concur with the findings of an earlier study [7]. This difference may be explained by differences in the experimental protocols and conditions used. In OVX ewes of both breeds, no effect of the dose of E on the duration of estrus behavior was observed. Previous studies [12, 18] have reported a dose-dependent effect, and this difference may be due to the smaller window of time and range of E treatments used in the current study. Indeed, if high blood levels of E are maintained, estrus duration can be prolonged beyond that which occurs naturally [19]. We hypothesize that in both Romanov and Ile de France breeds, there is a threshold dose above which a ewe will express estrus behavior; this threshold is lower in Romanov ewes. This hypothesis is supported by experiment 4, in which a very low E signal induced estrus in all the Romanov ewes, whereas none of the Ile de France females responded. This lower threshold for Romanov ewes to express sexual behavior may be due to the presence of more estrogen receptors in the mediobasal hypothalamus, where E acts to stimulate expression of estrus behavior. Progesterone priming has been shown to increase the number of estrogen receptor-immunoreactive cells in this region [20]. Therefore, the difference between the breeds may also be due to a difference in sensitivity to the inductive effects of progesterone. As the onset of estrus is identical in both breeds, it appears that once E has bound to its receptors, the neuronal pathway that is activated to induce estrus behavior is identical in the two breeds.
The difference in estrus duration between the breeds cannot be explained by a difference in E persistence, as the E plasma levels in treated OVX females in experiment 2 were not significantly different between the two breeds. During the natural cycle, the peripheral E concentration decreases before or concomitant with the end of the LH surge, while GnRH secretion and sexual receptivity continue for some hours [21]. In OVX ewes treated with progesterone and E, the period of estrus behavior almost coincides with the period of increased GnRH release [1, 13, 21], and the duration of increased GnRH secretion is independent of the duration of the E signal [14]. Caraty et al. [15] have suggested that a potential role of the extended preovulatory GnRH release during the late follicular phase is to prolong receptivity after E has disappeared from the peripheral circulation. The longer estrus duration observed in Romanov ewes may be explained by longer secretion of GnRH, which would prolong estrus in this breed. However, as explained below, estrus behavior was expressed without any LH surge for the small E signals. Moreover, a direct effect of E on estrus behavior in the ewe can be dissociated from its effect via stimulation of GnRH secretion [15]. A similar GnRH-independent steroid effect on estrus behavior has also been demonstrated in other species. For example, hpg mice, which are deficient in GnRH, exhibit normal sexual behavior if given estrogen and progesterone [22]. Therefore, an alternative hypothesis to explain the longer duration of estrus behavior observed in the Romanov ewe could be that the cascade of events induced by E binding takes longer time. Further experiments involving direct measurement of GnRH release as well as GnRH antagonist blockade of the estrus behavior induced by various levels of E signals are required to clarify this point.
A marked difference in the sensitivity to E to induce an LH surge was noted between the two breeds. Contrary to what we observed for estrus behavior, Romanov females were less sensitive and required more E than Ile de France ewes. The results from experiments 2, 3, and 4 show that when all the Ile de France ewes exhibited an LH surge in response to moderate E exposure (6 cm and 12 cm E for 12 h), fewer than half of the Romanov ewes had an LH surge. The maximum E level was reached 2 h after implant insertion and remained almost constant until its removal, as previously reported [14]. For Romanov ewes, the plasma E levels achieved by 6 cm E implants were in the same range as those observed in intact females just before the LH surge. In contrast, the E signal produced by the 12 cm/12 h E treatment can be considered as supraphysiological, and yet only 40% of the Romanov ewes exhibited an LH surge. The duration of E insertion had to be increased to 24 h to observe an LH surge in all the Romanov ewes. This requirement of a longer duration of exposure to E in Romanov ewes is supported by the results of experiment 6, in which a smaller implant (6 cm E) left in for 24 h resulted in the majority of ewes having an LH surge. Therefore, it seems that the duration rather than the dose of E is critical to induce an LH surge in this breed. In Romanov ewes, the lower sensitivity to E to induce the LH surge was also marked by a longer interval between E insertion and LH release. If progesterone is present concurrently with E, the LH surge is inhibited in sheep, as in many species. The latency of the LH surge thus depends both on the stimulatory effect of E and escape from progesterone inhibitory effect. It is plausible to speculate that the difference in surge latency between the breeds is due to a different sensitivity to this inhibitory effect of progesterone.
In sheep, the MBH has been shown to be the primary central site for E to generate both the preovulatory LH surge and sexual behavior [13, 23]. Interestingly, our results show that in Romanov ewes, it is possible to dissociate entirely the induction of the preovulatory LH surge from that of estrus behavior by changing the administered dose of E. A similar dissociation probably also exists in Ile de France ewes but is much more subtle due to a very small difference in E requirement to induce both events. This breed difference in the threshold doses of E to induce both events clearly explains why it has been previously reported that during the course of the follicular phase, estrus precedes the LH surge by a few hours in prolific but not in nonprolific ewes [4, 5, 10, 24]. Moreover, the present study strongly suggests that although the primary site of E action is the MBH for both events, the neuronal systems involved require different E signals.
From the present study, several remarks can be made about the mechanisms leading to the preovulatory GnRH/LH surge. Three sequential phases have been identified in the induction of the GnRH/LH surge in the ewe: activation, transmission, and surge secretion [25]. The presence of E is only needed during the activation phase. In this species, placement of E implants into the MBH, but not into the hypothalamic preoptic area (POA) or the arcuate nucleus (ARC), induces a GnRH/LH surge [13, 23]. Therefore, it is very likely that the main mechanisms of the activation phase are restricted to the MBH, and our present results show that the duration of this activation phase differs strongly between genotypes. Following a few hours of E administration, an increase in FOS immunoreactivity has been reported in the ARC and the ventromedial nucleus (VMN) [26]. Similarly, a short E signal duration of 4 h has been shown to modulate the expression of several neuropeptides within the MBH in Ile de France ewes, including TAC3 (also known as NKB) [27], somatostatin [28], and POMC [29]. Since there are projections from the MBH to the POA in the ewe [30, 31], it has been hypothesized that some of the aforementioned neuronal populations are the primary elements in the pathways that convey the E signal from the MBH to the GnRH cells in the surge-inducing process. However, there are no definitive conclusions as to whether these early changes are related to the negative and/or positive feedback effect of E on GnRH secretion. The fact that in Romanov ewes, the E signal needs to be read for a relatively long period of time to induce a GnRH/LH surge, regardless of the E concentration applied, suggests that there is more than one step in the E activation phase. It is possible that one or several other cell populations need to be activated in the MBH after the E primary activation to induce a GnRH/LH surge. In support of this hypothesis, a decrease in NPY expression [32] and an increase in KISS1 expression [33] in the ARC occur prior to the preovulatory GnRH/LH surge in ewes. Alternatively, morphological modifications and reorganization of the neuronal systems may also explain the prolonged E requirement for the surge to occur. Indeed, E has been shown to be involved in the regulation of neural plasticity within the MBH in the rat and monkey [34, 35] and in the rat, these changes have been related to the generation of the preovulatory LH surge [36, 37]. It is possible that all of these processes take more time in the Romanov ewe than the Ile de France ewe. Clearly, between-breed comparisons of the dynamics of all these changes within the MBH following E administration will help to clarify the mechanisms involved in the surge-inducing process.
Several hypotheses, which are not mutually exclusive, can be put forward to explain the differences within a single breed in terms of the E requirement to induce the preovulatory GnRH/LH surge and estrus behavior. It is possible that E has different region- and time-specific effects on cellular activation within the MBH, which generates the preovulatory GnRH surge and estrus behavior. In support of the latter part of this hypothesis, FOS activation of cells of the VMN requires more prolonged estrogen exposure than the cells of the ARC, and it has been suggested that the FOS responses in the VMN are more likely to be involved in the behavioral response [26].
The effect of progesterone within the brain may also explain, at least in part, the dissociation of the two mechanisms. Progesterone treatment decreases the sensitivity threshold to E and is required for E to induce normal estrus behavior [23]. Conversely, pre-exposure to progesterone delays the onset of the positive feedback effects of E on gonadotropin secretion [38]. It is possible that these two different actions of progesterone are more pronounced in Romanov ewes.
Finally, a time difference in E activation of other structures involved in the cascade of events that lead to the generation of the preovulatory LH surge or estrus behavior may also explain this dissociation. In this respect, it must be pointed out that the time lag between the onset of estrus behavior and the GnRH/LH surge in the Romanov ewe (approximately 8 h) strongly suggests that there are more steps in the induction of the former than the latter. E enhances neuronal activation within the POA in time- and region-dependent manners [39]. This finding led these authors to suggest that neurons within the POA may participate in the GnRH surge-induction process even if they are not the primary site of E action but are downstream of this. These authors have also reported a significant E-induced increase in FOS expression at the time of surge onset, in the lateral septum, which is a region known to express E receptors [40]. They have further suggested that the induction of FOS in this region may be indicative of a functional role for E-receptive cells in this area in GnRH-related sexual behavior. In the rat, E-receptive neurons in the lateral septum project to the midbrain central gray matter and are thought to be involved in lordosis behavior at the time of ovulation [41].
In summary, our results demonstrate a large difference in the E requirement to induce estrus behavior and the preovulatory LH surge between ewes with different ovulation rates. The longer latency period required to generate a surge and the longer estrus duration in the Romanov ewe can be related respectively to the longer period of time required for follicle recruitment and the longer duration necessary for the ewe to be mated in high-fecundity breeds. Moreover, we have shown that estrus behavior can be induced independently of the LH surge. This is an important advance, as it will help to identify the specific neuronal systems that control each of these events.
ACKNOWLEDGMENTS
We thank Dr. F. Clement for the helpful discussions regarding the surge-generating system and Dr. D.C. Skinner for his constructive comments on the manuscript. We thank Eric Arche for the care of the animals.
FOOTNOTES
1Supported by ACI Technologie. 
Correspondence: 2FAX: 33 24 742 7743; e-mail: caraty{at}tours.inra.fr
Received: 15 September 2006.
First decision: 5 October 2006.
Accepted: 26 December 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  A. Caraty, J. T. Smith, D. Lomet, S. Ben Said, A. Morrissey, J. Cognie, B. Doughton, G. Baril, C. Briant, and I. J. Clarke Kisspeptin Synchronizes Preovulatory Surges in Cyclical Ewes and Causes Ovulation in Seasonally Acyclic Ewes Endocrinology, November 1, 2007; 148(11): 5258 - 5267. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
