| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Behavior; |
Departments of Psychology,3 Pediatrics,4 and Neuroscience;5 and Reproductive Sciences Programs;6 and School of Natural Resources,7 University of Michigan, Ann Arbor, Michigan 48109
ABSTRACT
The process of sexual differentiation leaves genetically female individuals at risk of being masculinized by exogenous androgens. Previous research with sheep indicates that exposure to excess testosterone from Gestational Day (GD) 30 to GD 90 of the 147-day gestation masculinizes and defeminizes behavior as well as genitalia. Lower doses and shorter durations produce animals with varying degrees of genital virilization and alterations of the hypothalamic-pituitary-gonadal axis, but to our knowledge, the effects on complex behavior and its prediction by the amount of external virilization have not been explored. Previous research in rodents has suggested that sexual differentiation of the central nervous system and the external genitalia can be dissociated. Therefore, we hypothesized that the extent of virilization of external genitalia would not be predictive of the lack of female-typical, or the presence of male-typical, mating behavior. To test this hypothesis, we compared control females, females exposed to exogenous testosterone from GD 30 to GD 90 (T60 females) that have virilized genitalia, and females exposed to testosterone from GD 60 to GD 90 (T30 females) that have female-typical genitalia. Both natural behavioral estrus in the flock and hormonally controlled behavioral tests were used to explore reproductive behavior. The T60 and T30 females exhibited more masculinized reproductive behavior than the controls; however, the T30 females also exhibited feminine behavior. Neither testosterone-treated group was receptive or was mounted at rates comparable to those of controls. These data illustrate that variation in the timing or duration of exposure to prenatal testosterone during a critical period for masculinization can have variable effects on defeminization and that the effects of testosterone on genitalia are not entirely predictive of behavior.
behavior, differentiation, early development, organization
In humans and other mammals, sexual differentiation is initiated by genetic sex, which determines the gonadal sex of the individual. Phenotypic expression of sex develops according to the presence or absence of testosterone. If development occurs under normal conditions, the genetic, gonadal, and phenotypic sex will be congruent (for review, see Breedlove and Hampson [1]). This process of sexual differentiation, however, allows disruptions that may lead to varying amounts of incongruence between the genetic, gonadal, phenotypic, and resulting behavioral sex. Male fetuses are exposed to their own gonadal testosterone, which acts throughout the body to both masculinize and defeminize structures as well as processes. In contrast, females are particularly vulnerable to perturbations by both endogenous and exogenous hormones during the critical period for sexual differentiation; female tissues, including the brain and genital tract, also are capable of being masculinized and defeminized during development by appropriately timed exposure to exogenous testosterone [1–3].
The disruption of reproductive physiology by excess prenatal testosterone during a critical period of fetal or neonatal life has been demonstrated in several species, including guinea pig [4], rat [5], sheep [6–9], rhesus macaque [10, 11], and human [12]. For sheep, the critical developmental period for sexual differentiation of the hypothalamic-pituitary-gonadal (HPG) axis spans Gestational Days (GD) 30–90 of their 147-day gestational period [6–9]. Females treated with testosterone for 60 days during GD 30–90 (T60 females) are born with completely virilized genitalia—that is, with a functioning pseudopenis and an empty scrotal sac in place of a vaginal opening [7]. This outward masculinization is accompanied by multiple neuroendocrine and ovarian disruptions, such as LH hypersecretion, disrupted estradiol negative and positive feedback, and the development of multifollicular ovaries [7–9].
In contrast, female sheep treated for a shorter duration later in gestation develop less severe perturbations [7, 13–15]. For instance, administration of testosterone from GD 50 to GD 100 resulted in females with a masculinized urinary stance and restricted vaginal openings [13]. A shorter treatment from GD 65 to GD 86 produced females with female-typical external genitalia as measured by comparative anogenital ratios, whereas early exposure of the same duration (GD 30–51) masculinized genitalia [7]. Both groups produced partially masculinized patterns of tonic LH secretion, suggesting that the critical timing of organization of the brain differs from that of the external genitalia. Treatment lasting 30 days between GD 60 and GD 90 (T30 females), spanning the second half of the critical period, yielded females born with essentially normal external genitalia, with one of the 12 females studied having a noticeably constricted vaginal opening [16]. These phenotypically normal females displayed a masculinized urinary posture and altered neuroendocrine function, which manifested as delayed estradiol positive-feedback response [14, 15], suggesting that the brain is either defeminized or masculinized.
The T60 females that were treated from GD 30 to GD 90 with testosterone were fully virilized [7, 9] and expressed more male-typical behavior than control females as both lambs and adults [17, 18]. To our knowledge, such detailed behavioral investigations have not been undertaken with T30 females (i.e., the nonvirilizing model). Our recent studies found that despite the presence of a vaginal opening and an apparently minor alteration in the HPG axis [14, 15], T30 females failed to mate or successfully conceive after being bred [16]. These data suggest that behavioral patterns of T30 females may be impacted more negatively by the prenatal testosterone exposure compared with the external phenotype.
Because of the long duration of the critical period (i.e., 60 days) and the variable timing of central nervous system and peripheral masculinization and defeminization, we hypothesized that genital masculinization would not predict behavioral masculinization and/or defeminization. Specifically, we predicted that excess prenatal testosterone would alter behavior significantly, even in the absence of phenotypic virilization. In the present study, we compared reproductive-related behaviors of T30 females, the nonvirilized model, and T60 females, the fully virilized model, to those of unexposed control females under natural, free-living and hormone-induced estrus.
Animal Generation and Maintenance
Breeding. All animal procedures were approved by the University of Michigan Committee for the Use and Care of Animals. Prenatal testosterone-treated ewes were generated over 2 yr, with the T60 females in the first year and the T30 females in the second. For breeding purposes, adult Suffolk ewes were purchased from local farmers and moved to a farm inspected and approved by both the U.S. Department of Agriculture and the University of Michigan Department of Laboratory Animal Medicine. To determine the date of mating during breeding, a mixture of paint and grease was applied daily to the chests of rams in the flock (raddled rams). Daily monitoring allowed visual confirmation of marks left by the rams, and the date of mating was recorded. Once bred, ewes were maintained on pasture and supplemented with 1.25 kg of alfalfa/brome mix hay per ewe.
Prenatal hormone treatment. For generation of T60 females, pregnant ewes received twice-weekly, 2-ml i.m. injections of 100 mg of testosterone propionate (Sigma-Aldrich Corp.) in cottonseed oil between GD 30 and GD 90 of the 147-day gestational period. Control females included those that were bred in our facility with the T60 females and those that were purchased from the local farmers, who provided the ewes for generating the experimental females. The purchased controls were brought to the facility at 10 wk of age (i.e., weaning). The T30 females were generated the following breeding season by treating pregnant ewes with testosterone from GD 60 to GD 90. Because no significant differences were observed in the growth trajectory, cyclicity, or behaviors of control females born at the facility and those purchased during the first year, all control females used during the second year (n = 12) were purchased from the same local farmers and brought to the facility after weaning (i.e., 10 wk of age). Anogenital distances were measured in all females born at the facility (T60 females [19], T30 females [16]). The T60 females had anogenital distances equivalent to those of males, whereas the T30 females had anogenital distances that did not differ from those of control females.
Diet. Six weeks before expected lambing, each pregnant ewe was fed an enriched diet consisting of 0.5 kg of shelled corn, 2 kg of alfalfa hay, and 250 mg of chlortetracycline (Aureomycin, Fort Dodge Animal Health)/day. Lambing occurred each year between early March and early April. Throughout this time, each lactating ewe was provided a ration of 1 kg of shelled corn and 2–2.5 kg of alfalfa hay per day. Lambs were weaned at approximately 8 wk of age. Postweaning, female lambs were moved to the University of Michigan Sheep Research Facility, where they were kept in their same-age cohort, always able to freely interact with each other. The lambs were given commercial sheep food pellets (Shur-Gain) and alfalfa hay, available ad libitum, until they reached a body weight of 40 kg. At that time, they were provided a diet containing 15% crude protein until 6 mo of age. All lambs and ewes were provided water and minerals ad libitum and were treated regularly with anthelmintics to minimize parasitic infection. Throughout the testing period, all lambs were kept at the Sheep Research Facility.
During the first year, data were collected from 18 T60 females and 22 control females (11 locally produced and 11 purchased). During the second year, 12 T30 females and 12 control females were studied. Free-living behaviors of all lambs in both years were observed during their first breeding season. In the first year, both T60 and control female lambs were housed with males that were vasectomized at 3–3.5 mo of age. The T30 and corresponding control females from the second year were housed without males until 24 wk of age. At that time, males were added to the group for estrous detections (see below).
Detection of natural behavioral estrus. Throughout their first mating season (i.e., 6–10 mo of age), all ewes were observed daily to detect behavioral estrous cycles. Raddled rams were used to monitor behavioral estrus in the group. A female was considered to be in estrus if a new, heavy, and evenly distributed paint mark was found. In addition, for the T60 cohort, blood samples were collected from the females by venipuncture twice weekly throughout the breeding season and were assayed for plasma progesterone (P4) concentrations to monitor reproductive cyclicity. Behavioral estrus was compared with P4 cycles, allowing determination of whether behavioral estrus preceded the P4 rises of the estrous cycle and, thereby, discrimination between mountings related to rank or other social behaviors rather than copulation. This combined data collection also allowed us to determine the regularity of behavioral estrous cycles. Blood samples were collected on the T30 cohort both before and at the onset of puberty; blood samples were not collected during the next 2 mo of the breeding season. Thus, cycles were determined only by raddle marks. The T30 females had one additional month of biweekly blood samples during the third month of the breeding season to verify ovarian cyclicity. Natural behavioral estrous collection was followed by synchronized mating and hormone-induced mating tests that were performed on a subset of postpubertal females.
Synchronized group test.
At approximately 8 mo of age, lambs were primed with 20 mg of prostaglandin F2
(PGF2; 5 mg/ml of Lutalyse; Pfizer Animal Health) to synchronize estrus and facilitate observation of reproductive behavior. Synchronized mating tests were performed on nine T60 females and 10 control females of the first cohort and on 12 T30 females with 12 control females in the second cohort. A PGF2 injection was administered twice, 11 days apart, to induce luteolysis and initiate natural estrus. In both years, interactions with the males and other females in the group were observed and recorded using a mating ethogram developed from Banks [20] and our own observations (Table 1). Behavior was recorded between 0730 and 1730 h on three consecutive days after the second injection. For purposes of analysis, "investigation" behaviors (i.e., sniff and nudge) were combined as one behavior because of the difficulty encountered in discriminating between them. Similarly, two aggressive behaviors (i.e., head butt and body slam) were combined for analysis as one behavior (i.e., aggression.)
|
Induced individual test.
Because ovarian function may be compromised in testosterone-treated ewes during the natural estrous synchronization by PGF2, estrus was induced with steroid hormones at doses that reliably induce estrus in controls. Mating behavior of females toward males was examined at approximately 10 mo of age for the first cohort (n = 7 control and 6 T60 females chosen randomly) and at approximately 34 mo of age for the second cohort (n = 4 control and 8 T30 females). An additional four control females were tested in parallel with the T30 females to increase the number of control females (total n = 8 control and 8 T30 females).
All ewes were primed with two controlled, internal, drug-releasing implants containing P4 (CIDR; device type-G, CIDR-G; InterAg; implanted subcutaneously) for 10 days, followed by estradiol implants (four 30-mm, estradiol-filled Silastic capsules implanted subcutaneously) 16 h after P4 removal. Approximately 20 h after the estradiol was implanted, all control females reliably displayed estrus. Each ewe was placed in a 30-m2 pen for the test. In the first cohort, females were tested with familiar, same-age cohort alpha (highest ranking) and beta (second-highest ranking) males. Only one male, familiar and alpha ranked, was used for testing the second cohort. For both cohorts, interactions were videotaped for 5 min and then analyzed for the frequencies of the various male- and female-typical mating behaviors exhibited by the test ewe (Table 1).
Detection of natural behavioral estrus. Because the natural, free-living conditions differed in the two cohorts relative to both year of birth and any potential random environmental events (e.g., weather and food), data were analyzed separately for the T60 and T30 studies. For each cohort, the average date of first behavioral estrus, number of behavioral cycles, and cycle interval were analyzed using ANOVA. The number of animals exhibiting behavioral estrus was compared using chi-square analysis.
Mating behavior. To compare the T60 and T30 treatments directly, an initial comparison of the control females was conducted to ensure that no effect of year, birthplace, or free-living conditions existed. For each test, behavioral patterns were compared across control females born in the facility during each year and those that were purchased. No differences were found in behavioral attributes of in-house and purchased control females in either mating test. Therefore, all control females were consolidated into one group for direct comparison with T60 and T30 females. For each of these tests, the frequency of each behavior was recorded for each animal and then averaged for each group. Multivariate ANOVA was used to compare the average number of displays in each category of behavior and each behavior separately (Table 1). Additionally, for all data, chi-square analysis was used to compare the percentage of each group of females displaying each behavior.
T60 females.
Significantly more control females (95%) exhibited behavioral estrous cycles than did T60 females (15.8%; 
2 = 26.66, P < 0.001) (Table 2). All control and T60 ewes that exhibited behavioral estrus subsequently produced normal P4 rises (data not shown), confirming a normal relationship between behavioral estrus and ovulation. No difference was found in timing of pubertal onset as determined by the initiation of behavioral estrus or, if behavioral estrus was not observed, P4 cycles. Significant differences, however, were found between the groups of females that demonstrated estrus in the number of behavioral cycles (t = 3.048, P = 0.006) and the interval between cycles (t = 2.402, P = 0.026) over the three monitored months of the breeding season. Thus, the few T60 females that produced behavioral estrus had fewer and longer cycles than the control females.
|
T30 females.
No T30 female exhibited a complete behavioral estrous cycle as measured by raddle marks. All control females exhibited cycles of behavioral estrus (2 = 24, P < 0.001) (Table 3). All control females were marked by the raddled ram 2.75 ± 0.351 times during the period of estrous detection. Significantly fewer T30 females were marked during this time, and each only once. Control females cycled at regular intervals; no T30 female cycled. Similar to the T60 females, the T30 females did not produce patterns of behavioral estrous cycles typical of control females.
|
Synchronized group test. A significant effect of treatment on male-typical mating behaviors exhibited during the group mating test was found (Wilks' lambda: F = 2.354, P = 0.006) (Table 4). Specifically, both T60 and T30 females produced significantly higher frequencies of male-typical mating behaviors compared with control females (Wilks' lambda: T60 vs. control, F = 5.007, P = 0.001; T30 vs. control: F = 2.921, P = 0.017) (Table 4). No significant difference was found in the frequency of male-typical behaviors between the T30 and T60 groups (Wilks' lambda: T60 vs. T30, F = 1.194, P = 0.384) (Table 4). All male-typical mating behaviors were directed toward the estrous control females in the group. The T60 females displayed no female-typical behaviors (and were excluded from multivariate ANOVA). Both T30 and control females exhibited female-typical behaviors, and no differences were found between these two groups in the frequency of behavior produced (Wilks' lambda: F = 1.012, P = 0.442) (Table 4).
|
In terms of the percentage of animals that displayed each behavior, chi-square analysis revealed that significantly more T60 and T30 females produced male-typical courting and mating behaviors compared with control females (see Table 4 for percentages; follow: 2 = 17.34, P < 0.001; call: 2 = 17.17, P < 0.001; nudge/sniff: 2 = 13.31, P < 0.005; flehmen: 2 = 8.76, P < 0.025; paw: 2 = 11.07, P < 0.005; headover: 2 = 13.47, P < 0.005; prepare to mount: 2 = 18.49, P < 0.001; mount: 2 = 14.94, P < 0.001). The numbers of females producing several female-typical behaviors also were affected by treatment. More control females produced the lookback and stand behaviors compared with T30 and T60 females, although the number of T30 females was intermediate between those of the control and the T60 females (lookback: 2 = 8.24, P < 0.025; stand: 2 = 6.15, P < 0.05). The same pattern was found in the number of animals mounted by a male (2 = 13.20, P < 0.005) (Fig. 1): Control females were readily mounted, but only one T60 female was mounted. The numbers of T30 and T60 females that were mounted did not significantly differ.
|
Induced individual test. Following P4 and estradiol treatment, all T60 females exhibited two or more male- and female-typical behaviors (Table 5) in the presence of a male. All control and T30 females also exhibited female-typical behavior but no male-typical behavior. Overall, a significant effect of treatment on female-typical behavior was found (Wilks' lambda: F = 15.516, P < 0.001) (Table 5). Specifically, T60 females produced higher frequencies of approach compared to both control and T30 females, and control females produced higher frequencies of lookback and stand behaviors compared to both testosterone-treated groups (Table 5). Too few control and T30 females displayed these behaviors to perform multivariate analysis on male-typical behaviors.
|
Chi-square analysis of the percentage of animals displaying each behavior revealed that more T60 females produced the follow, sniff, and mount behaviors directed toward the male, whereas no control or T30 female produced any of these behaviors (see Table 5 for percentages; follow: 2 = 12.83, P < 0.005; sniff: 2 = 23.90, P < 0.001; mount: 2 = 8.24, P < 0.025). No significant differences were found in the percentage of animals that produced any of the proceptive female behaviors, but more control females than T30 or T60 females were receptive, standing for the male to mount (2 = 21.04, P < 0.001). Males were more attracted to control females than to T30 or T60 females, as demonstrated by a much higher percentage of control females being mounted (2 = 15.22, P < 0.001) (Fig. 1) despite appropriate male-directed proceptive behaviors by testosterone-treated females.
Our findings clearly document that the reproductive behavioral patterns of the T60 and T30 females differ from that of control females. Our results support the hypothesis that prenatal testosterone exposure can significantly alter behavior in the presence of very minor or no other physiological or phenotypic anomalies in an animal with a midgestation critical period for phenotypic and central nervous system differentiation. This is best illustrated by the altered patterns in mating behavior of the nonvirilized T30 females. Throughout the present study, control females retained their female-typical patterns of mating behavior, whereas the behavioral patterns of the T60 and T30 females were dependent on hormonal and social context.
In the synchronized group test following PGF2 synchronization, we found that T60 females were both masculinized and defeminized. They presented high numbers of male-typical mating behaviors directed at estrous, control females and aggression directed at males, but they produced no female-typical behavior. The T30 females showed greater variation in behavioral patterns than the T60 females within this test. In contrast to the T60 females, the T30 females, overall, retained female-typical behaviors in addition to their masculinized patterns during the synchronized group test. When examined individually, several distinct patterns emerge. Some T30 females exhibited only male-typical behaviors, whereas others exhibited only female-typical behaviors. The most interesting cases, however, were two T30 females that alternated between male- and female-typical behaviors during this test. While in estrus, these females moved back and forth between soliciting the ram and courting the estrous, control ewes.
During the induced individual test, the T60 females produced some female-typical behavior in addition to low levels of male-typical behavior, including aggression toward the males. They also approached the males more often than the control and T30 females did. Whereas all approaches to the males were recorded by observers as a female reproductive behavior (i.e., approach), one cannot rule out the possibility that such approaches may be more related to competition with the male. If so, it should be considered an aggressive behavior, reinforcing the idea that the T60 females are more aggressive than the T30 females. This test, however, does not allow us to interpret the motives of the T60 ewes. The T30 females that manifested male-typical behavior during the synchronized group test failed to produce the male-typical patterns of behavior during the induced individual tests but did display female-typical behaviors toward the male. The results of the induced individual test suggest that all genetically female individuals will produce some female behavior under steroid hormone manipulation when in the presence of a male. The marked decrease in the receptive, stand behavior exhibited by both the T60 and T30 groups indicates that affected individuals would not reproduce successfully. This notion is supported by the decreased fertility of T30 females during repeated attempts to impregnate them [16]. Because T60 females are phenotypically male (i.e., no vaginal opening), fertility could not be tested directly.
The overall behavioral differences seen between the T60 and T30 groups may relate to the timing (GD 30–90 vs. GD 60–90) or duration (60 days in T60 females vs. 30 days in T30 females) of prenatal testosterone treatment. Alternatively, other factors may have contributed to theses differences. For instance, the differential rearing contexts of the T60 and T30 groups may have affected the display of sexual behavior. The T60 females were reared from birth with 10 males in their flock, whereas the T30 females lived in an all-female group until the beginning of the breeding season. Earlier studies report that heterosexual experience was vital to the development of sexual performance of rams, which includes aggressive guarding of estrous ewes from other animals exhibiting sexual interest [21]. To our knowledge, the present study is the first report of such data pertaining to female sexual behavior.
Within each treatment group, the differences in manifestation of male- and female-typical behaviors during the synchronized group and induced individual tests may reflect differences in hormonal state. The timing, pattern, and/or concentration of endogenous P4 and estradiol that accompanied the group tests may have differed from the exogenously administered P4 and estradiol in the individual tests. Alternatively, these behavioral differences may reflect differences in social context. The group test provided the females with multiple other animals, females as well as males, for interaction; the individual test presented the females with only male counterparts. Additionally, the number and origin of the males used in the behavioral tests differed between the T60 and T30 cohorts. Two males that had been living with the T60 cohort since birth were used while testing the T60 females. In contrast, only one male was used while testing the T30 females. This male was introduced at the beginning of the breeding season, was 1 yr older, and was larger than the test females. We may have inadvertently masked aggression exhibited by the T30 animals toward the male during the test because of the size and experience advantage of the male used and/or because the T30 females were not raised with males from 10 to 24 wk of age.
The difference in behavioral patterns between the T60 and T30 groups suggests that exposure for the full 60 days of the critical period or from GD 30 to GD 60 is necessary for defeminization, whereas exposure during only the second half of the critical period is sufficient for varying degrees of masculinization and loss of receptivity (while retaining much of the proceptive behavioral patterns). Our results suggest that early or longer exposure to testosterone is essential for defeminization, as seen in the T60 treatment group, whereas exposure during the second half of the critical period is sufficient for masculinization, as illustrated in the T30 females. Conversely, it is possible that masculine reorganization is more sensitive to testosterone than defeminization is throughout the critical period. Ideally, to test this, we could examine the behavior of animals exposed to testosterone between GD 30 and GD 60. Previous work with females exposed to testosterone from GD 30 to GD 50 found that although outwardly masculinized, the neuroendocrine system of these animals was no more altered than that in the animals exposed from GD 65 to GD 85 [22], but the behavior of these animals was not examined. Together with our results, this suggests that the differential programming of behavior most likely is related to timing rather than to duration of testosterone exposure.
Previous studies with different species have demonstrated the separation of the critical periods for the effects of androgens on separate aspects of behavioral and genital differentiation, but these studies have largely failed to measure all processes (e.g., rhesus macaque [10], ferret [23], guinea pig [24], and rat [25, 26]). For example, when Goy et al. [10] manipulated the time period of prenatal testosterone exposure in rhesus macaques, the early and late groups differed from each other in both genital virilization and male-typical behavior. Unfortunately, female-typical behavior was not closely examined. Female rats are behaviorally defeminized by early postnatal exposure to testosterone, but they only become anatomically masculinized with both prenatal and postnatal testosterone exposure [26]. In that study, female-typical, but not male-typical, behaviors were examined. Our data suggest that a study of the differentiation of sexual behavior, either alone or in addition to morphological effects, should examine both female- and male-typical behavior to illuminate the treatment effects on both defeminization and masculinization, which can be temporally separate.
As a way of identifying estrous individuals in both mating tests, we relied on the mount behavior of the males. One limitation in the use of measurements such as mounting to determine sexual roles of the test females is that factors outside of behavior can affect them. For example, sexual performance can be affected by genital abnormalities, ambulatory difficulties, smooth muscular coordination, and sexual inexperience [27]. Thus, the low percentage of virilized T60 females that were mounted successfully during the mating tests may have resulted from lack of interest by the male because the females lacked vaginal openings rather than from a lack of proper female-typical behavior. Genitalia, however, are not the only important physical factor, as suggested by one of the 12 nonvirilized females in the T30 group being ignored by the male despite her robust (to our analysis) display of all proceptive and receptive behaviors following estrous synchronization with PGF2. Because T30 females have female-typical external genitalia, neither the genitalia nor the behavior were preventing interest from the male. Indeed, the next year, the aforementioned ignored T30 female was impregnated by a different ram during a fertility trial [16]. This unexplainable male preference for particular females, regardless of the female's behavior, has been documented in sheep and may affect mounting patterns [28]. Another possibility is that the very minor changes in ovarian cycles of T30 females are sufficient to alter their attractiveness [14]. For example, vaginal odors that play a key role in female attractiveness are pH dependent [29]; perhaps pH is altered in these animals. Whether alteration in the vaginal flora of T30 females occurs remains to be investigated.
Although we can only speculate as to the exact hormonal mediation of the decreased receptive behavior in the T30 females, the documented delay in onset of the LH surge [14, 15] may play a role. The delayed LH surge could be the result of a perturbation in the function of GnRH, and this would have direct implications for behavior. Whereas hormonal communication between the hypothalamus, pituitary, and ovaries coordinate ovulation and sexual behavior [30], distinct roles for GnRH and estradiol in female sexual behavior seem to exist. Estradiol from the mature ovarian follicle is responsible for the initiation of estrous behavior. The GnRH secretion triggered by the same rise in estradiol maintains behavioral estrus for a longer period of time [30]. These data illustrate the essential role of endocrine hormones in reproductive behavior and the importance of coordinated development of ovaries, HPG axis, and brain areas that control behavior (e.g., ventromedial hypothalamus, preoptic area, and amygdala) to ensure correct synchronization of behavior and ovulation. It also is conceivable that different brain regions are responsible for proceptive and receptive behaviors. If so, these regions may be masculinized and/or defeminized at different stages of sexual differentiation such that the T30 treatment did not coincide with differentiation of the brain region responsible for proceptive behavior but did alter the region responsible for receptive behavior. Further research is necessary to identify clearly the brain regions responsible for the behavioral disruption.
In summary, the present study provides a detailed analysis regarding the programming of male- and female-typical behaviors as a function of prenatal testosterone treatment using sheep as a model system. The variation in time and duration of exposure resulting in significant behavioral differences in testosterone-treated groups indicates important subdivisions of events within the prenatal period of sexual differentiation. Although virilization of the external genitalia appeared to correlate with complete defeminization of behavior, masculinization of behavior occurred independently of genital virilization. Despite their female-typical genitalia, the T30 females fail to reproduce successfully because of behavioral perturbations [16]. Our data do not clarify the role of the testosterone metabolites, estradiol and dihydrotestosterone, on the organization of sheep reproductive behaviors. Previous work, however, strongly suggests that masculinization and/or defeminization of reproductive behaviors is the consequence of the estrogenic action of testosterone [6, 31].
These results are important for greater understanding the process of sexual differentiation of reproductive behavior, but they also provide an improved understanding of the risks associated with exposure to excess androgenic substances. Multiple potential sources of androgens, both natural and unnatural, exist that put females at risk for masculinization and/or defeminization. Cases of natural exposure to testosterone-like androgens occur in species with multiple young (e.g., rat [32]) and via genetic disorders that increase endogenous androgens (e.g., congenital adrenal hyperplasia in humans [12, 33]). Unnatural masculinization and defeminization also can result from exposure to environmental sources of exogenous androgen, such as pulp and paper mill effluent [34] or trebolone acetate, a cattle-growth stimulator [35]. Considering these findings and the increasing risk of exposure to steroid mimics that can act through steroid receptors during development because of increased environmental contamination, the need for continued research is clear.
ACKNOWLEDGMENTS
We wish to thank Doug Doop for excellent care of the sheep and many lessons on how to handle sheep, insert implants, and avoid rambunctious rams. In addition, University of Michigan undergraduate students assisted in collecting and analyzing these data; these students include Tiffany Chao, Erica LaVire, Michael Zakalik, and Allie Spencer. We also wish to thank Carol Herkimer, James Lee, Teresa Steckler, and Pam Olton for help with prenatal testosterone treatment and/or determination of P4 levels in the sheep. Additionally, we thank Dr. Fred Karsch for considerable advice at the outset of our studies on how to observe sheep reproductive behavior under controlled conditions and Dr. Kim Wallen for assistance in establishing methods for collecting, summarizing, and analyzing natural behavioral estrus data.
FOOTNOTES
1Supported by U.S. Public Health Service grants R01 P01-HD44232 (to V.P. and T.M.L.) and R01 HD41098 (to V.P.). 
Correspondence: 2Theresa M. Lee, Department of Psychology, University of Michigan, 530 Church St., Ann Arbor, MI 48109-1043. FAX: 734 763 7480; e-mail: terrilee{at}umich.edu
Received: 10 December 2007.
First decision: 9 January 2008.
Accepted: 21 March 2008.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  L. M. Gorton, M. M. Mahoney, J. E. Magorien, T. M. Lee, and R. I. Wood Estrogen Receptor Immunoreactivity in Late-Gestation Fetal Lambs Biol Reprod, June 1, 2009; 80(6): 1152 - 1159. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. K. Roberts, J. N. Flak, W. Ye, V. Padmanabhan, and T. M. Lee Juvenile Rank Can Predict Male-Typical Adult Mating Behavior in Female Sheep Treated Prenatally with Testosterone Biol Reprod, April 1, 2009; 80(4): 737 - 742. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
