1. Introduction
Not only is the application of assisted reproduction techniques (ARTs), such as embryo transfer (ET), artificial insemination (AI) and hormonal treatments, in donkeys possible [
1,
2,
3,
4], but it can also improve the reproductive efficiency and maximize offspring production [
5]. These goals are particularly important for the preservation of endangered breeds such as the Andalusian donkey (
Equus asinus africanus), which is listed as a breed in danger of extinction [
6] due to its small population size [
7]. In this context, manipulation of the jennies’ estrous cycle is a crucial tool to increase the breeding efficiency in donkeys. Unfortunately, the number of studies on this species is limited [
8,
9,
10,
11,
12] in comparison with the horse. In addition, the response to pharmacological therapy for the induction of luteolysis and ovulation varies among individuals of this species [
13].
Luteolytic drugs, such as prostaglandin F2alpha (PGF2α) and its synthetic analogues (PGF analogues), are commonly used for shortening the estrous cycle in mares, by decreasing diestrus and, consequently, the interovulatory interval (IOI), without affecting the fertility of the treated animals [
14]. Although the length of estrus is similar between jennies and mares, the duration of diestrus is longer in jennies [
11,
15]. Therefore, the use of luteolytic hormones becomes an essential tool to increase the number of collected embryos from a donor. Several luteolytic agents have been successfully used in jennies [
16,
17,
18]; of these, the most frequently used are dinoprost tromethamine (as PGF2α) and cloprostenol and luprostiol (as PGF analogues). The main advantages of synthetic analogues over natural prostaglandins are their longer half-life, effectiveness at lower dosages and fewer undesirable side effects [
19]. The administration of a prostaglandin (PG) to mares with a mature corpus luteum (CL) reduces the progesterone concentration within 36 h [
20] and induces estrus 2–5 days after administration [
14]. However, the interval from PG treatment to ovulation (PG–ovulation interval, POI) is greatly variable (2–16 days) [
21,
22], regardless of the day of the estrous cycle in which the treatment is administered [
20]. Hence, it is known that the main factors that influence the POI after PG administration are dosage, follicular diameter at the time of treatment and follicular status (growing vs. atresia) [
14,
23,
24]. Despite the numerous studies carried out on mares, a very limited number of studies were found in the literature on the effects of the administration of PG in jennies, and these few studies showed that the POI did not vary when PGF2α was administered at different days after ovulation (POI: 9.2–10.2 days) [
9,
18,
25].
The induction of ovulation in a short predictable time (within 48 h after treatment) becomes useful not only to reduce the duration of estrus and the number of inseminations or matings per cycle but also to have a closer synchronization between insemination and ovulation [
26,
27]. In addition, ovulation management is an important tool to synchronize the donor’s and recipient’s cycles in ET programs [
4,
28,
29]. There are several hormones that can be used to induce ovulation in equine species [
30,
31], such as human Chorionic Gonadotropin (hCG), Equine Pituitary Extract (EPE) and Gonadotropin-Releasing Hormone (GnRH). Although the efficacy of both EPE and GnRH in hastening ovulation has been widely proven in mares [
26,
28,
30,
32,
33,
34], hCG is still the most popular inductor, not only in mares [
27,
30,
31], but also in jennies [
35,
36], due to its high efficacy, easier commercial availability and lower cost. However, various studies have shown that repeated administration of this hormone can stimulate the synthesis of anti-hCG antibodies, leading to a loss of its effectiveness [
37,
38]. Hence, it is usually not recommended to administer hCG more than twice within the same breeding season [
39], which is not practical in ET programs. Luckily, GnRH agonists can be administered repeatedly [
40], becoming the preferred alternative to hCG for timed ovulation induction. In mares, the GnRH agonists buserelin acetate and deslorelin acetate were proven to be able to induce ovulation 24–48 h after administration, similarly to hCG [
41,
42,
43]. Moreover, a more reliable induction of ovulation has been suggested when deslorelin acetate is administered to mares that have large thick-walled follicles in comparison to hCG [
44,
45,
46,
47]. However, only a few studies have been performed on the use of a single administration of either hCG [
35], lecirelin acetate [
35], buserelin acetate [
3] or deslorelin acetate [
48] as ovulation-inducing agents in jennies. There are several factors related to success when using ovulation-inducing agents. The follicular diameter of the largest follicle at the moment of treatment with hCG and GnRH analogues has been previously demonstrated in mares [
27,
32] and donkeys [
49]. It is also known that the uterine edema pattern depends on the prevailing circulating levels of ovarian steroids [
50], increasing edema scores as estrogens rise. More recently, it has been suggested that the failure of LH to induce ovulation in mares might be due to the absence of estradiol-positive feedback [
51]. In mares, both the preovulatory follicular diameter and uterine edema pattern at the time of treatment may be used to predict the response to ovulation-inducing treatment [
52]. However, to the best of our knowledge, there are no studies on this issue performed on donkeys.
The main objectives of the present study were to: (i) evaluate the efficacy of two luteolytic agents, luprostiol (as a PGF analogue) and dinoprost tromethamine (as PGF2α), for luteolysis and estrus induction in donor jennies, and (ii) compare the efficacy of hCG and deslorelin acetate in inducing ovulation of embryo donor jennies. Further objectives of this study were to: (iii) examine the effects of either the prostanglandins or ovulation-inducing agents on the embryo recovery rate (ERR) and embryo morphological quality (as embryo grade score (EG)), and (iv) determine whether follicular diameter and uterine edema score at the time of treatment affect reproductive performance, ERR and EG in the treated donor jennies.
4. Discussion
To the best of our knowledge, this is the first study in which the effects of the prostaglandin agents luprostiol (LUP) and dinoprost tromethamine (DIN) are compared in jennies. According to a previous study in mares [
55], these luteolytic drugs had a similar luteolytic effect, since the interval from injection to detection of ovulation (POI) was not significantly different between PG treatments. In addition, similar mean values of IOI (interovulatory interval) were found for LUP-treated and DIN-treated jennies.
Luprostiol has previously been used in Andalusian jennies [
18]. In that study, POI (10.22 ± 0.92 days) and mainly IOI (23.07 ± 0.50 days) were slightly higher than our results (POI: 9.73 ± 0.33 days; IOI: 16.86 ± 0.34 days). As reported previously in mares [
20,
22], a high variation in the POI of jennies was observed (LUP: from a minimum of 6 days to a maximum of 20 days; DIN: from a minimum of 4 days to a maximum of 17 days), which could be caused by a dose effect [
24,
55] rather than by the day of diestrus in which the PGF analogue is administered [
23]. In contrast to a previous study on diestrus that used 7.5 mg lupostriol [
18], a lupostriol dosage of 5.25 mg on days 6–9 of diestrus was used in our study. It is also known that POI mainly depends on the dominant follicle size at the time of treatment and on the subsequent follicular development and estrogen production [
23,
24], which could be different between studies, complicating the comparison of findings [
9]. Regarding DIN, our results are in consonance with those reported by Goretti et al. [
5] in donor mares treated with dinoprost tromethamine, in which POI and IOI were 10.0 ± 0.9 days and 17.5 ± 1.1 days, respectively.
On the other hand, both PG treatments (LUP and DIN) were able to reduce the diestrus length as IOI was shortened, on average, to 17.12 ± 0.29 days (LUP: 16.86 ± 0.34 days; DIN: 17.73 ± 0.55 days), thereby improving the reproductive performance of donor jennies in ET programs. However, when the distribution of estrous cycles was examined, it was noteworthy that the induction of estrus with LUP resulted in a better synchronization of estrus and ovulation than treatment with DIN. Thus, the proportion of LUP-treated jennies with a POI of greater than 11 days was significantly lower than that of the group with a POI of 9 to 11 days (22.8% vs. 40.5%), but this finding was not observed in DIN-treated jennies. Differences in the response between PGF2α and the PGF analogue could be due to a stimulatory effect of the PGF analogue on the hypothalamus and pituitary, as has already been reported in luprostiol-treated mares during the spring transitional period [
56]. This study also reported that luprostiol causes a transient increase in FSH and LH concentrations in jugular blood, thereby resulting in direct induction of ovulation.
Similar to previous studies performed on luprostiol-treated jennies [
18] and on cloprostenol-treated mares [
57], no effect of season on both POI and IOI was observed. However, it is noteworthy that during winter, POI was significantly shorter in LUP-treated than in DIN-treated jennies. Follicular diameter at the time of treatment was higher in LUP-treated than in DIN-treated jennies during winter (38.78 ± 2.06 mm vs. 35.00 ± 0.00 mm), which could explain our results as the largest follicle at the time of treatment might have a higher daily growth rate [
23,
24].
The overall ERR obtained in this study (69.3%) was similar to that reported in the literature for donkeys (40.7–80.6%) [
4,
58,
59,
60,
61]. Our results do not show any significant differences between PG treatments for PFR, ERR and EG, which is consistent with previous reports on mares [
23]. However, in our opinion, statistical differences could not be determined because of the small subgroup sizes. Thus, the overall ERR tended to be higher (
p = 0.0701) in DIN-treated than in LUP-treated jennies (81.8% vs. 64.6%), without affecting the embryo morphological quality (i.e., EG).
The ovulation rate within 48 h is generally considered as an accurate criterion for evaluating the ability of a treatment to induce ovulation [
62]. No differences between ovulation-inducing treatments (hCG and DES) were observed for this parameter, which is consistent with previous studies in mares [
63,
64,
65,
66] and jennies [
35]. However, as previously reported [
52], it seems probable that DES has a better ability to induce ovulation of smaller follicles than hCG and thereby affect their efficacy (DES: 78.7% vs. hCG: 60.9%;
p = 0.086). DES-treated jennies ovulated with an average follicle size at the time of treatment of 36.85 ± 0.78 mm, significantly lower than non-treated jennies (39.89 ± 0.91 mm), whereas jennies that responded adequately to hCG had an average follicle size at the time of treatment not significantly different from any of the groups (37.85 ± 1.48 mm).
Our data show similar ovulation rates to previous studies using deslorelin acetate or hCG in mares [
66], but lower than those reported in previous studies using 3000 IU of hCG [
67] or 2.2 mg of deslorelin acetate [
48] in jennies. The recommended dosage for hCG (in mares) is 1500–3000 IU. Although the complete vial of Veterin Chorion contains 3000 IU of hCG, only 1500 IU was used for ovulation induction in this study. This dosage was previously described for jennies by Serres et al. [
68], and it was successfully used for inducing ovulation in jennies in later studies [
61,
69,
70,
71,
72]. However, this is the first study in which this concentration of hCG is compared to 0.75 mg of deslorelin acetate for induction of ovulation in this species. It is possible that the doses of 0.75 mg of deslorelin acetate were too low to maintain a consistent LH release and, consequently, to induce timely ovulation in jennies. However, it has been previously demonstrated that a lower dose of deslorelin acetate (i.e., 0.1–0.2 mg) might be enough to induce timely ovulation [
73]. In mares, it has been reported that 0.5 mg of deslorelin acetate is as effective as 1 mg of deslorelin acetate in inducing timely ovulation [
43]. On the other hand, a recent study suggested that individual refractoriness to GnRH analogues exists [
3]. Unfortunately, to our knowledge, no studies have been conducted yet about the efficacy of different doses of deslorelin acetate in jennies, and therefore further studies on this issue are needed in the future.
Regarding the low ovulatory response to hCG (60.9%; 14/23), this finding could be explained by differences in the size of the preovulatory follicle. Interestingly, the average size of the preovulatory follicle at the moment of hCG treatment was significantly higher in the jennies that responded to treatment (ovulation occurred within 48 h, 40.09 ± 1.48 mm) than in those that did not ovulate after 48 h of treatment (35.11 ± 1.86 mm). It is known that mares respond most consistently to hCG if the follicular size is ≥ 35 mm [
38], a time at which the granulosa cell receptors respond to LH [
74]. A reduced efficacy of hCG after repeated use in the same breeding season has also been described previously [
31,
39]. In our study, 6 of the 13 hCG-treated jennies received more than two consecutive treatment sessions, which might affect its efficacy. However, this is somewhat difficult to interpret in this study because only a limited number of jennies were treated more than twice. Altogether, both products used to induce ovulation resulted in acceptable response rates for routine use in donkey ET programs.
It has been suggested that in mares that ovulated within 24 h after treatment, either they responded exceptionally fast to the ovulation-inducing agent or, more probably, the dominant follicle may have already been under the influence of endogenous LH [
41]. The rate of ovulation between 0 and 24 h was numerically higher but not significantly different in DES-treated (24.6%) compared to hCG-treated jennies (17.4%), which is consistent with that reported previously in mares [
28]. Moreover, the percentage of hCG-treated jennies that ovulated within 24 h after treatment was similar to that described in other studies in mares (12.7–18.5%) [
41,
75].
It is noteworthy that the overall interval from hCG to ovulation (ITO) was longer than that observed in DES-treated jennies (62.61 ± 7.20 h vs. 48.79 ± 2.69 h). Moreover, we observed a peak in ovulations 24–48 h after DES treatment, similarly to previous studies [
30,
66,
76]. Overall, our results suggest that both treatments have similar efficacy in inducing ovulation in treated jennies, but it seems that DES treatment could induce ovulation in a shorter and more predictable time window.
The overall incidence of double ovulations in induced estrus was surprisingly low (28.6%) compared to that reported in the literature [
18,
77], where estrus induction using PGF2α was associated with a higher occurrence of double ovulations or twin pregnancy. This difference could be explained by the influence of other factors such as age [
18], breed [
11] or body condition score [
55]. The incidence of double ovulations in hCG-treated jennies was similar to that of DES-treated jennies (39.1% vs. 24.6%), which is in agreement with findings reported in mares [
22]. Unfortunately, the increased ovulation rate for hCG-treated jennies did not result in a greater number of embryos recovered (hCG: 60.9% vs. DES: 72.1%).
The current study shows a lack of seasonal effect on ITO for both ovulation-inducing treatments (hCG and DES). This contrasts with reports in previous studies on mares, where ovulation occurred sooner after induction during the breeding season by administration of 1500–2000 IU of hCG [
27,
78] or 0.5 mg of buserelin acetate [
79]. It is widely known that the reproductive cyclicity of jennies is less affected by season than that of horses and ponies [
8,
9,
15,
80], a fact which could explain our results. In contrast to previous studies in mares [
79,
81], the ovulation distribution was clearly different in DES-treated jennies, being lower in winter. This finding could be explained by the fact that nonresponder jennies took longer to ovulate (spontaneously), as reflected by the ITO.
Although no significant differences between ovulation-inducing agents were found for PFR, ERR and EG, the ERR obtained in DES-treated jennies was numerically higher than that of hCG-treated jennies (72.1% vs. 60.9%). This finding could be related to the ovulation rate observed in our study (DES: 78.7% vs. hCG: 60.9%). Moreover, ITO was significantly higher in hCG-treated than in DES-treated jennies (62.61 ± 7.20 h vs. 48.79 ± 2.69 h), and the phenomenon of aged oocytes from persistent follicles is widely recognized [
82]. The present study, and previous studies on mares [
83] have demonstrated that EG is similar between ovulation-inducing treatments. Together, our findings could suggest that the administration of DES to donor jennies may result in a greater number of recovered embryos.
The present study also demonstrates that follicle size at the time of treatment influenced the ITO, thereby supporting the hypothesis that smaller follicles (≤35 mm) require a longer time interval for ovulation induction, as they have fewer receptors for FSH and LH [
74]. This finding is in consonance with that reported by Carluccio et al. [
35,
49] in Martina Franca donkeys, where a longer ITO was found when the follicular diameter at the time of treatment was ≤35 mm as compared to larger follicles (36–40 mm and >40 mm), and it was supported by the correlation analysis. On the other hand, we observed that PFR and ERR were numerically higher (
p > 0.05) when ovulation was induced in larger follicles (>40 mm), whereas the opposite was found for EG. These findings could be due to the fact that the bigger the follicle, the closer the ovulation and, therefore, the higher the pregnancy rate [
84]. However, further experiments are needed to confirm these results.
Abnormalities in the uterine edema pattern could help to predict the reproductive performance of treated donor jennies. Accordingly, EG significantly decreased with an increase in the uterine edema grade (≥3, range 0–4) at the time of treatment. Similarly, both PFR and ERR were numerically higher, but not significant, in jennies with a uterine edema score of 1. A low edema score probably indicates that luteinization is starting [
50], shortening the time between breeding and fertilization, increasing the pregnancy rates. Although, to our knowledge, there are no previous studies evidencing the relationship between the EG and uterine edema patterns in jennies, a significant decrease in both embryo recovery and pregnancy rates was recorded in mares with excessive edema, pre- and post-mating [
85,
86]. In summary, our results could suggest that the presence of a uterine edema score of 3 at the time of treatment is not a good indication of induction of ovulation in jennies as it affects the EG.