This study shows for the first time that: (1) AMH is present in the plasma of prepubertal ewe lambs; (2) plasma AMH concentration is related to the occurrence of ovulation in ewe lambs after administration of an ovarian stimulation treatment, possibly reflecting that a population of gonadotropin-responsive follicles is already present at this age in most ewes; and (3) plasma AMH concentration before puberty could be used as a predictor of the fertility of adult ewes at first mating. This test would allow farmers to perform a precocious selection of replacement ewe lambs with the highest expected fertility at first mating. In the conditions of the present study, plasma AMH concentrations did not depend on age and weight, being detectable in most ewe lambs. Nevertheless, longitudinal studies would be necessary to accurately determine the age and weight at which AMH becomes detectable as well as the precise relationships of AMH with age and live weight.
In ewe lambs at around 3.6
months of age, plasma AMH concentration was about 3–4-fold higher in ovulating than in non-ovulating ewe lambs after eCG treatment. This difference could be due to the presence of a different numbers of antral gonadotropin-responsive follicles between animals at this age. In this way, the mean age of the ewe lambs of this study (15.6
weeks) would be coincident with a period where a peak in the total number of follicles
mm has been observed in other breeds of sheep, varying between 14
] and 16
]. Therefore, the relationship found between the ovulatory response to eCG and plasma AMH concentrations could be explained by the fact that they likely both reflect the presence of antral gonadotropin-responsive follicles in the ovaries of ewe lambs. In adult cows, a high correlation of plasma AMH concentrations with the number of corpora lutea after superovulation was found [15
]. However, in the present study a relationship between AMH and the number of ovulations in ovulating ewe lambs was not found. This discrepancy could be attributed to physiological differences between prepubertal and adult females and/or to differences in the ovarian stimulation treatments. In this way, the ovarian response to a single eCG treatment without previous progestogen priming might be different to the response to a superovulation protocol consisting in multiple FSH decreasing doses with previous progestogen synchronization.
In the present work, neither plasma AMH concentration before puberty nor fertility at first mating was related with age or live weight. Moreover, all ewes received the same feeding plane during the whole experiment. Therefore, the observed differences in plasma AMH concentrations amongst ewes of similar age and similar live weight may be due to inherent variations in the follicular population, possibly influencing fertility at first mating. Genetic factors or maternal nutrition during pregnancy could therefore be responsible for this asynchronous follicular development; as in sheep, alterations in the maternal diet have been demonstrated to affect fetal ovarian development [5
]. Nevertheless, further studies are needed to elucidate the causes of this asynchronous development of the ovary between the same aged and weighted ewes, as well as to determine whether it could affect ovarian function and reproductive performance during the adult life.
Apart from age, live weight and nutritional status, it is widely known that successful pregnancy is affected by many other factors, such as season, endocrine status or uterine conditions [20
]. In the present work, healthy ewe lambs were reared under the same conditions to minimize the impact of such factors, although possible unknown individual disorders can not be discarded. The results of the present study showed that ewes with higher prepubertal plasma AMH concentrations displayed a higher probability of becoming pregnant at first mating than those with lower plasma AMH concentrations. Moreover, those ewes which failed to conceive after two consecutive mating opportunities had the lowest prepubertal plasma AMH concentrations. As far as we know, this is the first time that a relationship between AMH and early fertility has been established in a domestic species. Our study reinforces the prognostic value of AMH, even when the physiological status is not fully known. Both in the human and bovine, AMH levels are almost independent of the phase of the ovarian cycle, explaining why a single AMH measurement is usually sufficient [21
]. This could be due to the fact that AMH is not involved in feed-back mechanisms of the hypothalamo-pituitary-gonadal axis [21
]. Previous studies in the human have demonstrated a relationship between AMH and fertility. Very low concentrations of AMH are associated with infertility in the case of pre-menopause and premature ovarian failure, and very high AMH concentrations with an absence of ovulation in women with polycystic ovarian syndrome. Low or undetectable levels of AMH is explained by decrease or absence of follicular growth due to depletion of the ovarian reserve of primordial follicles. Abnormally high levels of AMH are explained by an excess of antral follicles whose growth is blocked prior to ovulation [21
]. Recent results have also evidenced that AMH seems to be an excellent marker of ovarian function in girls and adolescents. In this way, elevated serum AMH levels have been found in girls with a precocious maturation of the adrenal cortex, suggesting that follicular development has proceeded to a more advanced developmental stage than expected for chronological age [23
]. It has been demonstrated in adolescent patients with inherited reproductive problems (Turner syndrome) that their AMH concentrations correlate significantly with ovarian function at time of AMH measurement, showing an excellent sensitivity and specificity as a screening test of premature ovarian failure. Furthermore, it has been suggested that variations in AMH concentrations during childhood may theoretically predict the duration of any given girl’s reproductive lifespan [24
], as AMH seems to reflect the continuous decline of the follicle pool with age [25
In the case of ewe lambs, the observed AMH values would be rather a reflection of the beginning of follicular growth in prepubertal animals, showing an ovarian activity quantitatively different between animals of the same age. It is possible that ewe lambs in our study with the highest ovarian activity at 3.6
months were also the most precocious in terms of early puberty, resulting in the improved fertility observed when testing fertility of different ewes at the same age. The lambs with the earliest follicular development would be fertile at an early age. All these findings highlight the importance of individual differences in chronological follicular development and in the follicular reserve as being responsible for future reproductive life.
In sheep, recent in vitro studies have demonstrated that prepubertal oocytes isolated from ovaries with a high number of follicles larger than 2
mm are more competent for in vitro development, presenting a higher cleavage rate, developmental kinetics and cell number at the blastocyst stage [26
]. As it has been established that the granulosa cells of large preantral and small antral healthy follicles produce the highest AMH amounts in rat [27
] and cow [15
], prepubertal ewe lambs with higher plasma AMH concentrations are supposed to have a higher number of healthy follicles in these stages, suggesting that the quality of the oocytes and embryos coming from these ewes may be higher. Although further studies are needed to confirm this hypothesis, it could partially explain the increased fertility at first mating of ewe lambs with the higher concentrations in plasma AMH during their prepubertal phase.
Concerning the ability of AMH to be used as a screening test to select the most precocious ewes in terms of fertility at first mating, in the present study the AUC value for AMH (0.698) was very close to 0.7, the limit value to classify tests as "moderately accurate" according to an arbitrary guideline. This guideline distinguishes between non-informative (AUC
0.5), less accurate (0.5
0.7), moderately accurate (0.7
0.9), highly accurate (0.9
AUC <1) and perfect tests (AUC
]. The values obtained by cross-validation methods were somewhat lower and would correspond to a “less accurate” test. This is not strange as it its well known that the establishment and maintenance of pregnancy depends upon several genetic and environmental factors. Previous reports in the human have shown that while high AMH levels prior to the initiation of IVF treatments is a good predictor of the ovarian response, it has not proven to be as predictive of successful pregnancy. This indicates that factors other than the quantitative aspects of ovarian reserve may influence the fate of pregnancy [29
]. Nevertheless, although not being highly accurate, we have shown that fertility of replacement ewes at first mating can be increased up to about 20 percentage points by using plasma AMH concentration as a screening test. It remains to be assessed at which age and weight the diagnostic test based on the measurement of AMH concentrations on ewe lambs is the most efficient for accurate prediction. Concerning fertility at the second mating and fertility after two consecutive service periods, we have shown that the differences between ewes with AMH concentration above and below the cutoff point are also significant and of great practical importance (+21.5 and +20.6 percentage points, respectively). This would indicate that AMH could also be a predictor of fertility in adulthood, but this needs further confirmation.