Reprod Dom Anim 49, 420–426 (2014); doi: 10.1111/rda.12290 ISSN 0936–6768
Luteoprotective Role of Equine Chorionic Gonadotropin (eCG) During Pregnancy in the Mare G Flores-Flores*, E Velazquez-Canton, M Boeta and L Zarco Facultad de Medicina Veterinaria y Zootecnia, Universidad Nacional Aut onoma de M exico,Ciudad Universitaria, M exico City, M exico
Contents The effects of repeated cloprostenol administration were compared in mares impregnated by horses and mares impregnated by donkeys in order to assess the role of eCG on the development of pregnancy-associated resistance to the luteolytic and abortifacient effects of PGF2a. Eleven mares impregnated by donkey (mule pregnancy) and 9 mares impregnated by horse (horse pregnancy) were used. Six mares with mule pregnancy and four with horse pregnancy were injected with cloprostenol (0.25 mg) when they were between day 65 and day 75 of pregnancy, and the treatment was repeated 48, 72 and 96 h latter. The rest of the mares remained as controls. Concentrations of eCG were 10 times higher (p < 0.001) in mares impregnated by horses than in mares impregnated by donkeys, and they were not affected by cloprostenol treatment. Luteolysis was completed 30 h after the first cloprostenol injection in mule pregnancies, while mares with horse pregnancies required 96 h and three cloprostenol injections to complete luteolysis. Regression analysis revealed significant associations between eCG concentrations at time 0 and the time required for completion of luteolysis (p < 0.001), foetal death (p < 0.01) and foetal expulsion (p < 0.05). It is concluded that high eCG concentrations in mares impregnated by horses protect the corpora lutea of pregnancy against the luteolytic effects of PGF2a. Low eCG concentrations in mares carrying mule foetuses afford them less protection against the luteolytic effect of PGF2a, and this may be a cause of the increased foetal mortality that occurs between days 60 and 90 of pregnancy in these mares.
Introduction The role of equine chorionic gonadotropin (eCG) is not completely understood. It acts on the same receptors as LH (LH/eCG receptors) in the luteal cells of the mare (Saint-Dizier et al. 2003), where it has important luteotropic effects (Bergfelt et al.1989), stimulating luteal synthesis of progesterone, androgens and oestrogens (Daels et al. 1991a; Albrecht et al. 2001). It also has luteogenic effects, stimulating the formation of supplementary CL during the first third of pregnancy (Allen 1984; Boeta and Zarco 2012). However, these effects of eCG are not indispensable for pregnancy success, as a large proportion of mares impregnated by donkeys carry their mule foetuses to term even though they produce very little or no eCG (Allen 1975; Boeta and Zarco 2005) and have progesterone concentrations significantly lower than those of mares impregnated by horses (Boeta and Zarco 2012). Thus, it has been suggested that pituitary LH could generally suffice to maintain luteal function during pregnancy in the mare and that the role of eCG might be to protect pregnancy in cases of special need (Boeta and Zarco 2010, 2012).
One such possibility is that eCG could protect the CL of pregnancy against endogenous release of PGF2a. In this respect, PGF2a release followed by active luteolysis has been found to occur during colic, laminitis, endotoxemia and other pathological processes (Daels et al. 1989), and PGF2a release followed by luteolysis can be induced by the administration of Salmonella typhimurium endotoxin to non-pregnant mares (Fredriksson et al. 1986). Administration of endotoxin to mares that are between days 23 and 36 of pregnancy also results in luteolysis, which is soon followed by foetal death (Daels et al. 1987). However, it is unlikely that the endogenous release of PGF2a associated with the above-mentioned pathologies would result in luteal regression and abortion after days 40–50 of pregnancy, since after that time, the primary and supplementary CL of pregnancy are quite resistant to the luteolytic action of PGF2a, and multiple doses of exogenous PGF2a are required to induce abortion (Douglas et al. 1974; Squires et al. 1980; Van Leewen et al. 1983; Daels et al. 1995). Also, most mares treated with endotoxin after day 40 of pregnancy, and all of those treated after day 60, remained pregnant and maintained progesterone concentrations above 2 ng/ml, even though PGF2a was released in response to the endotoxin (Daels et al. 1987) The CL’s resistance to PGF2a appears at the time when eCG concentrations reach peak levels during pregnancy. Thus, resistance has been assumed to be due to a luteoprotective effect provided by high concentrations of eCG (Daels et al. 1995; Allen 2001), which could protect the CL of the pregnant mare in a manner similar to the way in which exogenous hCG has been shown to protect the ovine CL against PGF2a administration (Bolt 1979). This concept is supported by the observation that mares that produce very little or no eCG, such as mares carrying mule pregnancies or xenogeneic donkey pregnancies, appear to be more vulnerable to endogenous PGF2a, as suggested by the high incidence of abortions in which foetal death is immediately preceded by an abrupt fall in progesterone concentrations in mule pregnancies (Boeta and Zarco 2005, 2010) and in donkey-in-horse pregnancies (Allen 2001). However, no study has been conducted to assess whether luteal resistance to PGF2a in pregnant mares is actually related to high circulating eCG concentrations, and not simply a function of pregnancy stage. Thus, the objective of this study was to use mares impregnated by donkey, a natural model of eCG deficiency, to assess the role of this hormone on the development of resistance to PGF2a in pregnant mares. To do this, the effects of repeated injections of a PGF2a analogue on luteal © 2014 Blackwell Verlag GmbH
Luteoprotective Role of eCG in Pregnant Mares
function and foetal viability were compared in mares impregnated by horses and mares impregnated by donkeys, and the relation between individual eCG concentrations and several indicators of resistance to exogenous PGF2a were evaluated.
Materials and Methods Animals and treatments The study was carried out at an experimental farm located near Mexico City, at 19° 09′20″ north. The experimental protocol was approved by the Institutional Committee for the Use and Care of Experimental Animals according to Mexican Official Norms relative to experimentation on animals. Twenty adult pregnant mares were used. The mares were fed according to requirements with oat hay, alfalfa hay and commercial concentrate. Eleven of them carried mule foetuses, having being inseminated with donkey semen. The other nine mares had been inseminated with horse semen and thus carried horse foetuses. When the animals were between 65 and 75 days pregnant, six mares carrying mule foetuses and four mares carrying horse foetuses (groups mule-PGF2a and horse-PGF2a, respectively) were injected with 0.25 mg of the synthetic analogue of PGF2a, cloprostenol (Celosil; MSD Animal Health, Huixquilucan, México). Five mares with mule pregnancy and five with horse pregnancy were administered 1 ml of physiological saline solution (mulecontrol and horse-control groups, respectively). The treatments were repeated at 48, 72 and 96 h after the first injection.
421
death was considered to have occurred when no cardiac beat or foetal movements were found during an examination, and abortion was considered complete upon foetal expulsion (Boeta and Zarco 2005, 2010). Statistical analysis The effects of time, type of pregnancy (mule or horse) and treatment (cloprostenol or saline) on hormone concentrations were evaluated by analysis of variance for repeated measures. Tukey’s multiple comparison tests were used to compare differences between groups at different times. An analysis of variance for repeated measures was also used to compare the percentage decline in progesterone concentrations at each time with respect to the initial values for each experimental group (mule-PGF2a or horse-PGF2a). The intervals from the first injection to foetal death and to foetal expulsion were compared between the two cloprostenol-treated groups by Student′s t-test. To evaluate the effects of eCG concentrations on luteal function and on resistance to PGF2a, data for all the cloprostenol-treated mares (independently of type of pregnancy) were pooled, and four lineal regression analyses were performed using the concentrations of eCG at time 0 as the independent variable, and progesterone concentrations at time 0, the interval from the first injection to completion of luteolysis, the interval from the first injection to foetal death and the interval from the first injection to foetal expulsion as the dependent variables.
Results Samples and hormone assays Blood samples were obtained by jugular vein puncture in vacuum tubes with serum activator gel (BD Diagnostic systems, Mexico City, Mexico) immediately before the first treatment (time 0), every 6 h for the first 36 h after the first injection, immediately before each subsequent injection (48, 72 and 96 h) and every 24 h thereafter until 168 h had elapsed from the time of the first injection. The samples were centrifuged immediately after extraction, and serum was separated and kept frozen at 20°C until assayed. Progesterone concentrations were determined in all samples in a single assay run. The sensitivity of the solid-phase radioimmunoassay (Boeta and Zarco 2005) was 0.01 ng/ml, and the variation coefficient was 10.6%. Luteolysis was considered to be complete when progesterone concentrations first decreased below 1 ng/ml (Boeta and Zarco 2005). Concentrations of eCG were measured in a single assay run in samples obtained at 0, 24, 48 and 72 h. The sensitivity of the enzyme immunoassay (PMSG-ELISA; DRG Instruments, Marburg, Germany) was 3.0 IU/ml, and the variation coefficient was 7.5%. Ultrasonographic evaluations Ultrasonographic observations of the foetus were conducted immediately before the first injection and every 12 h thereafter until the product was expelled. Foetal © 2014 Blackwell Verlag GmbH
The concentrations of eCG were at all times significantly higher (p < 0.001) in mares with horse pregnancies than in those with mule pregnancies (Fig. 1). Within each type of gestation, there were no differences between the mares receiving cloprostenol and the control mares (p > 0.05). Treatments with cloprostenol did not affect subsequent eCG production, since at all times, the concentrations of eCG were similar to those present in the same group before the first injection (time 0) and to
Fig. 1. Concentrations of eCG immediately before (0) and at different intervals after the first treatment with 0.25 mg of a synthetic PGF2a analogue (cloprostenol) or with 1 ml of physiological saline solution in mares impregnated by a horse stallion and mares impregnated by a jack donkey. Treatments were repeated at 48 and 72 h after the first injection. a,b Different letters indicate significant differences (p < 0.001)
422
G Flores-Flores, E Vel azquez-Cant on, M Boeta and L Zarco
those present in the control group with the same type of pregnancy (p > 0.05). As shown in Table 1, progesterone concentrations were higher at all times in horse-control pregnancies than in mule-control pregnancies (p < 0.01). No significant temporal variations were observed within any of the control groups. In contrast, in both groups of mares treated with cloprostenol, the concentrations of progesterone started to decline immediately after the first injection, and in each case, the progesterone concentrations became significantly lower (p < 0.05) than the initial value for the group within 6 h after the first injection of cloprostenol. Also, starting 6 h after the first injection, the concentrations in each cloprostenol-treated group were always significantly lower (p < 0.05) than those in the control group for the same type of pregnancy. Concentrations in the mule-PGF2a group decreased below 1 ng/ml 30 h after the first injection, when only one injection had been administered, and remained below 1 ng/ml during the rest of the study. In contrast, in the horse-PGF2a group three injections and an interval of 96 h from the first injection were needed before progesterone concentrations decreased below 1 ng/ml. Figure 2 shows that in the mares assigned to cloprostenol treatment, the concentrations of progesterone immediately before administering the first injection were significantly correlated (p < 0.01) with the concentrations of eCG present at that time. Independently of type of pregnancy, the interval from the first injection to completion of luteolysis (progesterone 0.05)
expulsion. Differences between groups were not significant (p > 0.5). However, regression analysis of the individual data for all the cloprostenol-treated mares, independently of their type of pregnancy, showed that the intervals from the first injection to foetal death (Fig. 5a) and to foetal expulsion (Fig. 5b) were significantly correlated with the concentrations of eCG that were present immediately before the first injection (p < 0.001 and p < 0.05, respectively). © 2014 Blackwell Verlag GmbH
Luteoprotective Role of eCG in Pregnant Mares (a)
(b)
Fig. 5. Relationship between the concentration of eCG present immediately before the first cloprostenol injection and the interval from that injection to foetal death (a) and to foetal expulsion (b) in pregnant mares, independently of type of pregnancy
Discussion As expected from previous reports (Allen 1975; Boeta and Zarco 2005, 2010), in this study, the concentrations of eCG were much lower in mule pregnancies than in horse pregnancies (Fig. 1). Thus, results enabled evaluation of the susceptibility to a PGF2a analogue in mares with highly divergent eCG concentrations even though they were in the same stage of pregnancy. In addition, variation in eCG concentrations within each type of pregnancy allowed the use of regression analysis for further evaluation of the effects of eCG on luteal resistance to cloprostenol. The lower concentrations of progesterone in mares with mule pregnancies than in mares impregnated by horses have been attributed to differences in luteotropic support due to the low concentrations of eCG during mule pregnancies (Boeta and Zarco 2005, 2012), which in turn are due to premature and exacerbated rejection of the endometrial cups by the immune system of the mare reacting against paternally inherited donkey histocompatibility antigens (Antczak et al. 1985). Evidence of the luteotropic role of eCG is provided in the present study by the significant (p < 0.01) correlation between progesterone and eCG concentrations at time 0, regardless of type of pregnancy (Fig. 2). Boeta and Zarco (2005, 2010) found higher incidence of pregnancy loss in mares impregnated by donkeys than in mares impregnated by horses and identified several instances in which foetal death was immediately © 2014 Blackwell Verlag GmbH
423
preceded by a sudden drop in progesterone concentrations that suggested an active luteolytic process due to endogenous release of PGF2a. Of a total of 49 mule pregnancies monitored in those two studies, 7 (14.2%) terminated in abortions preceded by sudden luteal regression after day 60 of pregnancy. This type of abortion accounted for 44% (7/16) of all the abortions recorded in the mule pregnancies. The authors suggested that the very low levels of eCG during mule pregnancies would have rendered the CLs of pregnancy vulnerable to the action of endogenous PGF2a that could have been released during diverse pathological processes,. This concept was supported by the fact that in mares impregnated by horses only 4% (2/50) of pregnancies terminated in foetal death preceded by sudden luteolysis, and those cases always occurred before day 40 of pregnancy, when eCG secretion had not yet commenced (Boeta and Zarco 2005, 2010). The results of the present study support the hypotheses of a luteoprotective role of eCG during horse pregnancy and of insufficient eCG to protect the CLs in mule pregnancies. Thus, in mares impregnated by donkeys, the time required for a reduction in progesterone concentrations to
Luteoprotective Role of Equine Chorionic Gonadotropin (eCG) During Pregnancy in the Mare G Flores-Flores*, E Velazquez-Canton, M Boeta and L Zarco Facultad de Medicina Veterinaria y Zootecnia, Universidad Nacional Aut onoma de M exico,Ciudad Universitaria, M exico City, M exico
Contents The effects of repeated cloprostenol administration were compared in mares impregnated by horses and mares impregnated by donkeys in order to assess the role of eCG on the development of pregnancy-associated resistance to the luteolytic and abortifacient effects of PGF2a. Eleven mares impregnated by donkey (mule pregnancy) and 9 mares impregnated by horse (horse pregnancy) were used. Six mares with mule pregnancy and four with horse pregnancy were injected with cloprostenol (0.25 mg) when they were between day 65 and day 75 of pregnancy, and the treatment was repeated 48, 72 and 96 h latter. The rest of the mares remained as controls. Concentrations of eCG were 10 times higher (p < 0.001) in mares impregnated by horses than in mares impregnated by donkeys, and they were not affected by cloprostenol treatment. Luteolysis was completed 30 h after the first cloprostenol injection in mule pregnancies, while mares with horse pregnancies required 96 h and three cloprostenol injections to complete luteolysis. Regression analysis revealed significant associations between eCG concentrations at time 0 and the time required for completion of luteolysis (p < 0.001), foetal death (p < 0.01) and foetal expulsion (p < 0.05). It is concluded that high eCG concentrations in mares impregnated by horses protect the corpora lutea of pregnancy against the luteolytic effects of PGF2a. Low eCG concentrations in mares carrying mule foetuses afford them less protection against the luteolytic effect of PGF2a, and this may be a cause of the increased foetal mortality that occurs between days 60 and 90 of pregnancy in these mares.
Introduction The role of equine chorionic gonadotropin (eCG) is not completely understood. It acts on the same receptors as LH (LH/eCG receptors) in the luteal cells of the mare (Saint-Dizier et al. 2003), where it has important luteotropic effects (Bergfelt et al.1989), stimulating luteal synthesis of progesterone, androgens and oestrogens (Daels et al. 1991a; Albrecht et al. 2001). It also has luteogenic effects, stimulating the formation of supplementary CL during the first third of pregnancy (Allen 1984; Boeta and Zarco 2012). However, these effects of eCG are not indispensable for pregnancy success, as a large proportion of mares impregnated by donkeys carry their mule foetuses to term even though they produce very little or no eCG (Allen 1975; Boeta and Zarco 2005) and have progesterone concentrations significantly lower than those of mares impregnated by horses (Boeta and Zarco 2012). Thus, it has been suggested that pituitary LH could generally suffice to maintain luteal function during pregnancy in the mare and that the role of eCG might be to protect pregnancy in cases of special need (Boeta and Zarco 2010, 2012).
One such possibility is that eCG could protect the CL of pregnancy against endogenous release of PGF2a. In this respect, PGF2a release followed by active luteolysis has been found to occur during colic, laminitis, endotoxemia and other pathological processes (Daels et al. 1989), and PGF2a release followed by luteolysis can be induced by the administration of Salmonella typhimurium endotoxin to non-pregnant mares (Fredriksson et al. 1986). Administration of endotoxin to mares that are between days 23 and 36 of pregnancy also results in luteolysis, which is soon followed by foetal death (Daels et al. 1987). However, it is unlikely that the endogenous release of PGF2a associated with the above-mentioned pathologies would result in luteal regression and abortion after days 40–50 of pregnancy, since after that time, the primary and supplementary CL of pregnancy are quite resistant to the luteolytic action of PGF2a, and multiple doses of exogenous PGF2a are required to induce abortion (Douglas et al. 1974; Squires et al. 1980; Van Leewen et al. 1983; Daels et al. 1995). Also, most mares treated with endotoxin after day 40 of pregnancy, and all of those treated after day 60, remained pregnant and maintained progesterone concentrations above 2 ng/ml, even though PGF2a was released in response to the endotoxin (Daels et al. 1987) The CL’s resistance to PGF2a appears at the time when eCG concentrations reach peak levels during pregnancy. Thus, resistance has been assumed to be due to a luteoprotective effect provided by high concentrations of eCG (Daels et al. 1995; Allen 2001), which could protect the CL of the pregnant mare in a manner similar to the way in which exogenous hCG has been shown to protect the ovine CL against PGF2a administration (Bolt 1979). This concept is supported by the observation that mares that produce very little or no eCG, such as mares carrying mule pregnancies or xenogeneic donkey pregnancies, appear to be more vulnerable to endogenous PGF2a, as suggested by the high incidence of abortions in which foetal death is immediately preceded by an abrupt fall in progesterone concentrations in mule pregnancies (Boeta and Zarco 2005, 2010) and in donkey-in-horse pregnancies (Allen 2001). However, no study has been conducted to assess whether luteal resistance to PGF2a in pregnant mares is actually related to high circulating eCG concentrations, and not simply a function of pregnancy stage. Thus, the objective of this study was to use mares impregnated by donkey, a natural model of eCG deficiency, to assess the role of this hormone on the development of resistance to PGF2a in pregnant mares. To do this, the effects of repeated injections of a PGF2a analogue on luteal © 2014 Blackwell Verlag GmbH
Luteoprotective Role of eCG in Pregnant Mares
function and foetal viability were compared in mares impregnated by horses and mares impregnated by donkeys, and the relation between individual eCG concentrations and several indicators of resistance to exogenous PGF2a were evaluated.
Materials and Methods Animals and treatments The study was carried out at an experimental farm located near Mexico City, at 19° 09′20″ north. The experimental protocol was approved by the Institutional Committee for the Use and Care of Experimental Animals according to Mexican Official Norms relative to experimentation on animals. Twenty adult pregnant mares were used. The mares were fed according to requirements with oat hay, alfalfa hay and commercial concentrate. Eleven of them carried mule foetuses, having being inseminated with donkey semen. The other nine mares had been inseminated with horse semen and thus carried horse foetuses. When the animals were between 65 and 75 days pregnant, six mares carrying mule foetuses and four mares carrying horse foetuses (groups mule-PGF2a and horse-PGF2a, respectively) were injected with 0.25 mg of the synthetic analogue of PGF2a, cloprostenol (Celosil; MSD Animal Health, Huixquilucan, México). Five mares with mule pregnancy and five with horse pregnancy were administered 1 ml of physiological saline solution (mulecontrol and horse-control groups, respectively). The treatments were repeated at 48, 72 and 96 h after the first injection.
421
death was considered to have occurred when no cardiac beat or foetal movements were found during an examination, and abortion was considered complete upon foetal expulsion (Boeta and Zarco 2005, 2010). Statistical analysis The effects of time, type of pregnancy (mule or horse) and treatment (cloprostenol or saline) on hormone concentrations were evaluated by analysis of variance for repeated measures. Tukey’s multiple comparison tests were used to compare differences between groups at different times. An analysis of variance for repeated measures was also used to compare the percentage decline in progesterone concentrations at each time with respect to the initial values for each experimental group (mule-PGF2a or horse-PGF2a). The intervals from the first injection to foetal death and to foetal expulsion were compared between the two cloprostenol-treated groups by Student′s t-test. To evaluate the effects of eCG concentrations on luteal function and on resistance to PGF2a, data for all the cloprostenol-treated mares (independently of type of pregnancy) were pooled, and four lineal regression analyses were performed using the concentrations of eCG at time 0 as the independent variable, and progesterone concentrations at time 0, the interval from the first injection to completion of luteolysis, the interval from the first injection to foetal death and the interval from the first injection to foetal expulsion as the dependent variables.
Results Samples and hormone assays Blood samples were obtained by jugular vein puncture in vacuum tubes with serum activator gel (BD Diagnostic systems, Mexico City, Mexico) immediately before the first treatment (time 0), every 6 h for the first 36 h after the first injection, immediately before each subsequent injection (48, 72 and 96 h) and every 24 h thereafter until 168 h had elapsed from the time of the first injection. The samples were centrifuged immediately after extraction, and serum was separated and kept frozen at 20°C until assayed. Progesterone concentrations were determined in all samples in a single assay run. The sensitivity of the solid-phase radioimmunoassay (Boeta and Zarco 2005) was 0.01 ng/ml, and the variation coefficient was 10.6%. Luteolysis was considered to be complete when progesterone concentrations first decreased below 1 ng/ml (Boeta and Zarco 2005). Concentrations of eCG were measured in a single assay run in samples obtained at 0, 24, 48 and 72 h. The sensitivity of the enzyme immunoassay (PMSG-ELISA; DRG Instruments, Marburg, Germany) was 3.0 IU/ml, and the variation coefficient was 7.5%. Ultrasonographic evaluations Ultrasonographic observations of the foetus were conducted immediately before the first injection and every 12 h thereafter until the product was expelled. Foetal © 2014 Blackwell Verlag GmbH
The concentrations of eCG were at all times significantly higher (p < 0.001) in mares with horse pregnancies than in those with mule pregnancies (Fig. 1). Within each type of gestation, there were no differences between the mares receiving cloprostenol and the control mares (p > 0.05). Treatments with cloprostenol did not affect subsequent eCG production, since at all times, the concentrations of eCG were similar to those present in the same group before the first injection (time 0) and to
Fig. 1. Concentrations of eCG immediately before (0) and at different intervals after the first treatment with 0.25 mg of a synthetic PGF2a analogue (cloprostenol) or with 1 ml of physiological saline solution in mares impregnated by a horse stallion and mares impregnated by a jack donkey. Treatments were repeated at 48 and 72 h after the first injection. a,b Different letters indicate significant differences (p < 0.001)
422
G Flores-Flores, E Vel azquez-Cant on, M Boeta and L Zarco
those present in the control group with the same type of pregnancy (p > 0.05). As shown in Table 1, progesterone concentrations were higher at all times in horse-control pregnancies than in mule-control pregnancies (p < 0.01). No significant temporal variations were observed within any of the control groups. In contrast, in both groups of mares treated with cloprostenol, the concentrations of progesterone started to decline immediately after the first injection, and in each case, the progesterone concentrations became significantly lower (p < 0.05) than the initial value for the group within 6 h after the first injection of cloprostenol. Also, starting 6 h after the first injection, the concentrations in each cloprostenol-treated group were always significantly lower (p < 0.05) than those in the control group for the same type of pregnancy. Concentrations in the mule-PGF2a group decreased below 1 ng/ml 30 h after the first injection, when only one injection had been administered, and remained below 1 ng/ml during the rest of the study. In contrast, in the horse-PGF2a group three injections and an interval of 96 h from the first injection were needed before progesterone concentrations decreased below 1 ng/ml. Figure 2 shows that in the mares assigned to cloprostenol treatment, the concentrations of progesterone immediately before administering the first injection were significantly correlated (p < 0.01) with the concentrations of eCG present at that time. Independently of type of pregnancy, the interval from the first injection to completion of luteolysis (progesterone 0.05)
expulsion. Differences between groups were not significant (p > 0.5). However, regression analysis of the individual data for all the cloprostenol-treated mares, independently of their type of pregnancy, showed that the intervals from the first injection to foetal death (Fig. 5a) and to foetal expulsion (Fig. 5b) were significantly correlated with the concentrations of eCG that were present immediately before the first injection (p < 0.001 and p < 0.05, respectively). © 2014 Blackwell Verlag GmbH
Luteoprotective Role of eCG in Pregnant Mares (a)
(b)
Fig. 5. Relationship between the concentration of eCG present immediately before the first cloprostenol injection and the interval from that injection to foetal death (a) and to foetal expulsion (b) in pregnant mares, independently of type of pregnancy
Discussion As expected from previous reports (Allen 1975; Boeta and Zarco 2005, 2010), in this study, the concentrations of eCG were much lower in mule pregnancies than in horse pregnancies (Fig. 1). Thus, results enabled evaluation of the susceptibility to a PGF2a analogue in mares with highly divergent eCG concentrations even though they were in the same stage of pregnancy. In addition, variation in eCG concentrations within each type of pregnancy allowed the use of regression analysis for further evaluation of the effects of eCG on luteal resistance to cloprostenol. The lower concentrations of progesterone in mares with mule pregnancies than in mares impregnated by horses have been attributed to differences in luteotropic support due to the low concentrations of eCG during mule pregnancies (Boeta and Zarco 2005, 2012), which in turn are due to premature and exacerbated rejection of the endometrial cups by the immune system of the mare reacting against paternally inherited donkey histocompatibility antigens (Antczak et al. 1985). Evidence of the luteotropic role of eCG is provided in the present study by the significant (p < 0.01) correlation between progesterone and eCG concentrations at time 0, regardless of type of pregnancy (Fig. 2). Boeta and Zarco (2005, 2010) found higher incidence of pregnancy loss in mares impregnated by donkeys than in mares impregnated by horses and identified several instances in which foetal death was immediately © 2014 Blackwell Verlag GmbH
423
preceded by a sudden drop in progesterone concentrations that suggested an active luteolytic process due to endogenous release of PGF2a. Of a total of 49 mule pregnancies monitored in those two studies, 7 (14.2%) terminated in abortions preceded by sudden luteal regression after day 60 of pregnancy. This type of abortion accounted for 44% (7/16) of all the abortions recorded in the mule pregnancies. The authors suggested that the very low levels of eCG during mule pregnancies would have rendered the CLs of pregnancy vulnerable to the action of endogenous PGF2a that could have been released during diverse pathological processes,. This concept was supported by the fact that in mares impregnated by horses only 4% (2/50) of pregnancies terminated in foetal death preceded by sudden luteolysis, and those cases always occurred before day 40 of pregnancy, when eCG secretion had not yet commenced (Boeta and Zarco 2005, 2010). The results of the present study support the hypotheses of a luteoprotective role of eCG during horse pregnancy and of insufficient eCG to protect the CLs in mule pregnancies. Thus, in mares impregnated by donkeys, the time required for a reduction in progesterone concentrations to
