Effects of selective and combined activation of estrogen receptor α and β on reproductive organ development and sexual behaviour in Japanese quail (Coturnix japonica)

Abstract

Excess estrogen exposure of avian embryos perturbs reproductive organ development in both sexes and demasculinizes the reproductive behaviors of adult males. We have previously shown that these characteristic effects on the reproductive organs also can be induced by exposure of Japanese quail (Coturnix japonica) embryos to selective agonists of estrogen receptor alpha (ERα). In contrast, the male copulatory behavior is only weakly affected by developmental exposure to an ERα agonist. To further elucidate the respective roles of ERα and ERβ in estrogen-induced disruption of sexual differentiation, we exposed Japanese quail embryos in ovo to the selective ERα agonist 16α-lactone-estradiol (16αLE2), the selective ERβ agonist WAY-200070, or both substances in combination. The ERα agonist feminized the testes in male embryos and reduced cloacal gland size in adult males. Furthermore, anomalous retention and malformations of the Müllerian ducts/oviducts were seen in embryos and juveniles of both sexes. The ERβ agonist did not induce any of these effects and did not influence the action of the ERα agonist. Male copulatory behavior was not affected by embryonic exposure to either the ERα- or the ERβ-selective agonist but was slightly suppressed by treatment with the two compounds combined. Our results suggest that the reproductive organs become sexually differentiated consequent to activation of ERα by endogenous estrogens; excessive activation of ERα, but not ERβ, during embryonic development may disrupt this process. Our results also suggest that the demasculinizing effect of estrogens on male copulatory behavior is only partly mediated by ERα and ERβ, and may rather involve other estrogen-responsive pathways.

Fig 1. Length of right Müllerian duct (A) and frequencies of abnormal left Müllerian duct (B) in female quail embryos.

Embryos were treated by in ovo injection on E3 with vehicle, the ERβ agonist WAY, or the ERα agonist 16αLE₂. Effects were assessed on E16. The number of examined embryos was 9, 11, 9 and 11 in the groups from left to right in the figures. The horizontal lines in A indicate the group medians. Differences in Müllerian duct lengths were analyzed using Kruskal-Wallis ANOVA (p = 0.0014 and H = 15.6), followed by Dunn's multiple comparison test. Frequencies of abnormal ducts in the treated groups were compared with the frequency in the control using one-sided Fisher's exact test. Asterisks indicate significant differences compared with the control (*p <0.05; ***p <0.001).

Fig 2. Frequencies of male quail embryos with one or both Müllerian ducts retained.

Embryos were treated by in ovo injection on E3 with vehicle, the ERβ agonist WAY, or the ERα agonist 16αLE₂. Effects were assessed on E16. The number of examined males was 13, 11, 13 and 11 in the groups from left to right in the figure. Frequencies of males with ducts in the treated groups were compared with the frequency in the control using one-sided Fisher's exact test (***p <0.001).

Fig 3. Size (length × width) of right (A) and left (B) testis in male quail embryos.

Embryos were treated by in ovo injection on E3 with vehicle, the ERβ agonist WAY, or the ERα agonist 16αLE₂. Effects were assessed on E16. 16αLE₂ significantly reduced the size of the right testis and increased the size of the left testis. This indicates a feminizing effect since in control females only the left ovary develops whereas the right ovary regresses. The horizontal lines indicate the medians. The number of examined males was 13, 11, 13 and 11 in the groups from left to right in the figures. Differences were analyzed using ANOVA (p < 0.0001, F = 40.0, total df = 47 in A; p = 0.0006, F = 7.1, total df = 46 in B). Following the ANOVA, all treated groups were compared with the control using Dunnett’s multiple comparison test.(***p <0.001).

Fig 4. Fraction cortex in histological sections of the left testis in quail embryos.

Embryos were treated by in ovo injection on E3 with vehicle, the ERβ agonist WAY, or the ERα agonist 16αLE₂. Effects were assessed on E16. 16αLE₂ caused growth of a thick ovary-like cortex containing oocyte-like germ cells. The horizontal line indicates the group mean. Differences in testis cortex area were analyzed using Kruskal-Wallis ANOVA (p < 0.0001 and H = 25.4), followed by Dunn's multiple comparison test. Asterisks indicate significant differences compared with the control (***p <0.001).

Fig 5. Frequencies of a partially developed right oviduct (A) and a malformed left oviduct (B) in juvenile female quails.

A right oviduct longer than 10 mm was defined as partially developed. The birds were treated in ovo on E3 and effects were assessed three weeks after hatching. The number of examined females was 9, 14, 10 and 14 in the control, WAY, 16αLE₂ and 16αLE₂ + WAY combination groups, respectively. Frequencies in the treated groups were compared with the frequency in the control using one-sided Fisher's exact test (*p <0.05; ***p <0.001).

Fig 6. Frequencies of juvenile (A) and adult (B) male quails with oviductal structures.

In the majority of the cases, the oviducts consisted of one or more large liquid-filled cysts. The birds were treated in ovo on E3 and effects were assessed in juveniles three weeks after hatching and in adult males nine weeks after hatching. The number of examined juvenile males was 11, 13, 12 and 7 in the control, WAY, 16αLE₂ and 16αLE₂ + WAY combination groups, respectively. Nine adult males were examined from each group except for the 16αLE₂ group from which 8 males were examined. Frequencies in the treated groups were compared with the frequency in the control using one-sided Fisher's exact test (*p <0.05; **p<0.01).

Fig 7. Testis weight ratio (left/right) in juvenile (A) and adult (B) male quails.

The birds were treated in ovo on E3 with vehicle, the ERβ agonist WAY, the ERα agonist 16αLE₂, or with both compounds in combination. Ratios were determined in juveniles three weeks after hatching and in adult males nine weeks after hatching. The horizontal lines indicate the medians. Differences in ratios were analyzed using ANOVA. (p < 0.0001, F = 16.0, total df = 42 in A; p = 0.0026, F = 5.9, total df = 34 in B). Following the ANOVA, all treated groups were compared with the control using Dunnett’s multiple comparison test (*p <0.05; **p<0.01; ***p <0.001).

Fig 8. Cloacal gland area in adult male quails.

The birds were treated in ovo on E3 with vehicle, the ERβ agonist WAY, the ERα agonist 16αLE₂, or with both compounds in combination. Cloacal gland areas were measured nine weeks after hatching. The horizontal lines indicate the means. Differences were analyzed using ANOVA (p < 0.0001, F = 36.0 and df = 32). Following the ANOVA, all treated groups were compared with the control using Dunnett’s multiple comparison test (***p <0.001).

Fig 9. Copulatory behavior displayed by adult male quails.

For each bird, the number of trials out of five in which cloacal contact movements (CCM) were displayed at least once is shown in A and the average number of times that each bird showed this behavior across the five trials is shown in B. The birds were treated in ovo on E3 with vehicle, the ERβ agonist WAY, the ERα agonist 16αLE₂, or with both compounds in combination. The behavior was assessed once a day for five consecutive days during the ninth week after hatching. The horizontal lines indicate the group medians. Differences in behavior were analyzed using Kruskal-Wallis ANOVA (p = 0.0476, H = 7.9 in A; p = 0.30, H = 3.7 in B). Following the Kruskal-Wallis ANOVA, all treated groups were compared with the control using Dunn’s multiple comparison test (*p <0.05).

Fig 10. Copulatory behavior vs plasma testosterone concentration in adult males.

The copulatory behavior was scored as the number of trials out of five in which cloacal contact movements (CCM) were displayed at least once. Linear regression analyses revealed no correlation between testosterone concentration and copulatory behavior, and ANOVA analysis showed no difference in testosterone concentration between the groups.