Genotype, uterine position, and testosterone sensitivity in older female mice

Physiology&Behavior.Vol.51, pp. 1047-1050, 1992

0031-9384/92 $5.00+ .00 Copyright© 1992PergamonPressLtd.

Printed in the USA.

Genotype, Uterine Position, and Testosterone Sensitivity in Older Female Mice A T H E N A C O L O G E R - C L I F F O R D , N E A L G. S I M O N l A N D B O Y C E M. J U B I L A N

Department of Psychology and Center for Molecular Bioscience and Biotechnology, Lehigh University, Bethlehem, PA 18015 Received 26 August 1991 COLOGER-CLIFFORD, A., N. G. SIMON AND B. M. JUBILAN. Genotype, uterineposition, and testosterone sensitivity in olderfemale mice. PHYSIOL BEHAV 51(5) 1047-1050, 1992.--CF-1 and CK (C57BL/6J X AKr) female mice that developed in utero between two males (2M), adjacent to one male (IM), or between two females (0M) were tested for their sensitivity to the aggression-promotingproperty of testosterone (T) beginning at 9 months of age. Comparisons between the strains showed that a higher proportion of CF-I females fought in response to T and that the period of hormone exposure required to induce aggression also was shorter in this strain. Within each of the genotypes, there were no systematic differencesin responsivenessto T related to contiguity to males during fetal development. While the results provide further evidence for genotype as a major influence on neural sensitivity to androgen, they do not support uterine position of females relative to males as a source of phenotypic variation in responsiveness. Genotype

Mice

Aggression

Uterine position

Testosterone

CONTIGUITY to male fetuses or the presence of caudal males within a uterine horn (closer to the cervical junction) during fetal development reportedly produces partial masculinization of the female central nervous system in several rodent species. Although there is considerable debate as to which of these models is operative [c.f., (13) vs. (18,19)], both hypothesize that differential exposure to testosterone derived from male fetuses produces varying degrees of masculinization of female littermates. Thus, females that developed between two males (2M) or with a caudal male present exhibited enhanced responsiveness to the aggression-promoting (7,8) or mount-inducing (5,11,13) properties of T in comparison to females that developed between two females (0M) or without a caudal male present. Several reports, however, suggest that genotype and age should be considered before the general applicability of uterine position effects on neural sensitivity to T can be established. With regard to genotype, for example, 70-day-old 2M and 0M CF-I female mice did not differ in the duration of androgen exposure required to induce aggression (17), as had been reported in RocklandSwiss (R-S) mice (7). Further, 2M and 0M CK females given silastic T implants exhibited similar preference profiles and ultrasonic vocalizations in response to female urine (12). Concerning age of the animals, Rines and vom Saal (15) showed that differences between 0M and 2M CF-I females in T-induced mounting at 9 months of age were no longer present at 21 months. This report raises a question as to whether there is an

interaction among uterine position, age, and effects on neural T sensitivity or whether any variation in behavioral responsiveness related to contiguity to males in utero is maintained throughout the life span. The observations described above suggest that uterine position effects on T sensitivity may be genetically constrained and that these differences, if present, may decline with age. To address these possibilities, the following study examined the interaction of uterine position, genotype, and age on neural androgen sensitivity by comparing the response of 9-1 l-month-old CF-1 and CK female mice to the aggression-promoting property of T. METHOD

Animals CK female mice were derived from a C57BL/6J X AKR/J (both lines were obtained from the Jackson Labs, Bar Harbor, MI) cross and CF-1 female mice were bred from parent stock purchased from the Charles River Breeding Farm (Wilmington, MA). This hybrid was selected for two reasons. One was that it would expand the range of genotypes examined in the context of the uterine position phenomena, and the second was that the hormonal processes regulating male-typical behaviors other than aggression had been described (12). Mice were housed in 28 X 18 x 13 cm polypropylene cages on a bedding of wood chips with food and water provided in excess. Rooms were maintained

Supported in part by grants HD-17456 from NIH to NGS and BRSG-2 S07 RR07173 to Lehigh University. Requests for reprints should be addressed to Neal G. Simon, Ph.D., CMBB, Bldg. 111, Mountaintop Campus, Lehigh University, Bethlehem, PA 18015.

1047

('OI.OGER-CLIFFORD. SIMON AN1) JI!BILAN

1048 at 23 ± 20( ` on a 12/12 h light/dark cycle, with lights on at 0800 hours. Care and maintenance procedures fully complied with Federal guidelines for the treatment of animals.

Mating and l)elivery Breeding began when the females were 60 days of age. Two male mice were introduced into the home cage of four females on a daily basis for several hours. Dams were individually housed on the day a copulatory plug was observed, which was designated day 0 of pregnancy. The offspring were delivered on the evening of the day 18 of gestation (the duration of pregnancy is 19 days in mice) by cesarean section. The dams were sacrificed by cervical dislocation, the uterus exposed, and the pups were rapidly removed. They were placed on a heating pad maintaining their relative position in the uterus and cleansed with fresh water. They were then sexed by examining anogenital distance under a dissecting microscope and the uterine position of each female with respect to male littermates was established. Female pups were classified as 0M if they were between two females, if they were located at the ovarian end of the uterine horn with a contiguous caudal female, or if they were located at the cervical end of the horn with a rostral female and another female at the cervical end of the other uterine horn. They were classified a 2M if they were between two males and 1M if they developed contiguous to l male. For the last group, it also was noted whether the male was rostral (closer to the ovary) or caudal (closer to the cervical junction) to the female. Females were toe clipped for identification and fostered to dams that had delivered within the preceding 48 hours and whose young had been removed. The females were weaned at 21 days of age and remained housed with littermates. At 60-80 days of age they were ovariectomized under nembutal anesthesia and were subsequently tested for ultrasonic vocalization production and preference for female urine [see (12) for details]. In that study, half the females were treated with testosterone via silastic implant and the remainder received oil vehicle only. Upon completion of the investigation, the capsules were removed and the females were regrouped by uterine position. Three months later, at approximately 9 months of age, the females within each strain were given a silastic implant (10 mm long, i.d. = 0.062 in., o.d. = 0.125 in.) containing 5 mg T/0.02 cc oil. This dosage has been used in several studies on the androgenic induction of male-like aggression in ovariectomized females and produces circulating T levels comparable to those found in intact males (1,7,16).

TABLE 1 S U M M A R Y O F AGGRESSIVE B E H A V I O R E X H I B I T E D BY CF-1 A N D C K F E M A L E MICE IN R E S P O N S E T O SILASTIC I M P L A N T S C O N T A I N I N G T

Genotype

Proportion Fighting(%)*

Mean Latency (Days ___SEM)*

Proportion Attacking

CF-I CK

57/65 (88) 34/56 (61)

16.0 -+ 0.58 18.5 +_0.93

2/65 (2)$ 2/56 (1.3)t-

The proportion fightingrepresentsthose femalesthat reached criterion while proportion attacking shows females that displayed less than five attacks. * Significantdifference between strains. f Mean number of attacks by females that did not reach criterion.

[ABI .I 2 ,k(i(}RESSIVE BEtIAVIOR I)ISPLA~ El) B'~ ('1.-1 I I M \ i 1 MICE 1N RESPONSE TO F E S T O S ' [ E R O N E IMPI .\N [ S [ [t.'l'iD£ Position

Proportion Fighting ('")*

,%lcall [ ¢1[clIc) (Days ' SFM t*

OM IM 2M

1 5 / 1 8 (83':11 33/35 (94';i W I 2 (75%1

16.11 , I.IO 16.36 ~ t~.37 1(~.67 ~ (I.69

* No significantdifferenceamong uterine positions.

Aggression Tests Tests for aggression began 14 days after the capsules were implanted and were conducted every other day for a maximum of 20 days ( 10 tests). Each test began by introducing an olfactory bulbectomized male into the home cage of the experimental female for |0 minutes. Olfactory bulbectomized males were used because, while neither initiating attack nor responding by fighting back, they reliably elicit aggression comparable to that directed toward an intact male (6). This allows any observed aggression to be reliably ascribed to the experimental animal. The criterion for the display of aggression was a minimum of five biting attacks toward the stimulus male. Any female reaching the criterion was classified as a fighter and was not tested again. This test system was employed because it reliably discriminates between aggressive and nonaggressive animals (10,17) while minimizing the possibility of injury to the stimulus males by limiting exposure to biting attacks. RESULTS

The aggressive behavior displayed by CF-1 and CK female mice in response to T is summarized in Table 1. An overall X2 test showed that the proportion of CF-1 females that reached criterion was greater than that seen in the CK strain, ×2(1) = 3.84, p < 0.05. Comparing data only from females that fought, a t-test showed that the duration ofT exposure required to induce aggression also was shorter in the CF-1 strain, t(89) = 2.37, p < 0.05. Interestingly, this strain difference was present even when the analyses included females that displayed any attack behavior. As seen in Table 1, only 2 females from each strain exhibited attacks hut failed to reach the criterion. Thus, the results do not appear to be due to the CK females behaving less aggressively.

lntrastrain Comparisons The aggressive behavior displayed in response to T by CF-I females from each of the three uterine positions is summarized in Table 2. Comparisons among uterine positions found no difference in the proportion of females that fought, x2(2) = 3.52, NS, or in the duration of treatment required to induce an attack, F(2, 54) = 0.15, NS. Results obtained with CK females are shown in Table 3. The proportion of 0M, I M, and 2M females that fought did not differ significantly, x2(2) = 1.99, NS. However, there was a difference in the number of days of T treatment required to activate aggression, F(2, 31) = 4.50, p < 0.05. Post hoc comparisons using Scheffe's test showed that while the 2M and 0M females did not differ from each other, their mean treatment durations were shorter in comparison to the 1M females. Given that the analyses above did not support the contiguity model of uterine position effects, additional statistical tests were

GENOTYPE, IUP, AND ANDROGEN SENSITIVITY

TABLE 3 AGGRESSIVEBEHAVIORDISPLAYEDBY CK FEMALE MICE IN RESPONSETO TESTOSTERONEIMPLANTS Uterine Position

Proportion Fighting(%)*

Mean Latency (Days + SEM)t

0M IM 2M

12/16 (78%) 15/28 (53%) 7/11 (63%)

16.67 -+ 1.27 21.33 -+ 1.53~ 15.43 +- 0.67

* No significantdifference among uterine positions. t Significantdifference among uterine positions. $ Significantlydifferent from 0M and 2M.

conducted to assess whether the presence of a caudal male in utero differentially affected the subsequent response of females to T. This model, which has been demonstrated only in rats to date (11,13), was examined by dividing 1M females into two groups, 1R and 1C. This classification system was based on available evidence showing that the number of caudal males did not differentially affect behavior or morphology ( 13,15). Therefore, the IR vs. IC classification was the simplest system for partitioning the 1M females. These data are presented in Table 4. In CF-1 females, there were no significant differences in either proportion fighting or in the number of days of T treatment to activate aggression, t(30) = 0.95, NS. In CK females, the proportion fighting did not differ between the groups, but there was a significant difference in the duration of T treatment required to induce fighting behavior, t(13) = 1.92, p < 0.05. However, the direction of the difference was not as predicted by the caudal male model because the I R females responded more rapidly. DISCUSSION The findings bear on three issues related to the development and expression of androgen-induced aggressive behavior. One is the question of prenatal influences on subsequent neural steroid sensitivity. Next, there is the issue of genetic variation in responsiveness to androgen. And finally, the results provide information concerning changes in neural sensitivity to T that may accompany advancing age. In regard to prenatal influences, the results did not support an effect of contiguous or caudal males (the former is presumably the source of T in mice (7,20) while both sources have been hypothesized in rats (5,11)) on sensitivity of females from either strain to androgen. These findings are consistent with studies of younger CF-1 females (17) but contrast with results obtained from RS mice (7). While several explanations can be advanced to account for these findings, the most likely would appear to be that uterine position effects on behavioral sensitivity to T in females are genetically constrained. This could be due to either genetic variation in the amount of T secreted by male fetuses that reaches female littermates or developmental differences in the timing of masculinization of neural substrates that mediate the expression of fighting behavior or the response to aggressioneliciting stimuli. Support for the absence of variation in T exposure can be found in a recent study that showed that whole body androgen content did not differ among female rat fetuses as a function of the presence of contiguous or caudal males (2). However, these findings contrast with those reported for mice, where a 25% elevation in T was found in 2M as compared to

1049 0M females (20), indicating that caution is in order in regard to this hypothesis until additional studies of fetal T levels in mice are conducted. In regard to possible developmental constraints, this issue was raised by the work of Bigsby et at. (3), who noted that neonatal estrogen treatments exerted differential effects on uterine epithelial tissue in BALB/C and CD-1 mice due to differences in the ontogeny of estrogen responsiveness. The latter observation is particularly interesting because it raises the possibility that the amount o f T exposure could vary among females due to uterine position without functional effects on sensitivity to androgen. The third area addressed by the study concerned changes in neural sensitivity to androgen that accompany aging. At 10 months of age, CF-1 females showed a response to the aggressionpromoting property ofT comparable to that seen in 2-3-monthold females; the latency to fight (about 14 days) and proportion responding (at least 82%) were similar (17). While data for young CK females are not available, the lesser response of this strain in comparison to the CF-I line could be due to a more rapid rate of decline in neural target tissue sensitivity or, as discussed previously, this genotype may be generally less responsive to androgen. While studies with young CK females will be needed to fully resolve this question, previous work with older mice and rats has shown that the functional response to exogenously administered steroids does not seem to decline with age. More specifically, 21-month-old T-treated CF- I females mounted and attacked at a level similar to that seen in 9-month-old females (16). And in 19-month-old female rats, the level of hypothalamic progestin receptor induction in response to estradiol was comparable to that seen in young and middle-aged females (4). These observations indicate that the response capacity of neural tissue may remain intact in older rodents if exogenous steroids are given and that the normal decline in responsiveness with aging is most likely due to the decreased availability of gonadal hormones (9). In this context, it should be noted that older nonhuman primate males do not respond with increased sexual behavior when given exogenous T (14). Therefore, it seems that there is either a species difference in the response to exogenous T in the later stages of life or that the sensitivity of neural substrates for various male-typical behaviors are differentially affected by age. In closing, the presence of contiguous or caudal males during fetal development did not alter the subsequent response of older CF-I or CK females to the aggression-promoting property of T, although the two strains differed in behavioral sensitivity to androgen. These findings indicate that uterine position effects on steroid sensitivity, if detected, may well be transient.

TABLE 4 AGGRESSIVEBEHAVIOROF IM CF-I AND CK FEMALES IN RESPONSETO TESTOSTERONE C~I Uterine Proportion Position Fighting*(%) IR IC

13/15 (87) 19/19 (100)

CK

Mean Latency Proportion (Days_+SEM) Fighting*(%)

Mean Latency (Days_+SEM)t

15.85 +_0.39 16.53__0.53

17.60_+ 1.91 23.20_+ 1.81

5/11 (45) 10/17(59)

* No significantdifference between classifications. t Significantdifference between classifications.

1050

('OI_O(3ER-CLIFFORD.

SIMON A N D J I ) B I k A N

REFERENCES 1. Barkley, M. S.; Goldman, B. D. The effects of castration and silastic implants of testosterone on intermale aggression in the mouse. Horm. Behav. 9:32-48; 1977. 2. Baum, M.; Woutersen, P.; Slob, K. Sex differences in whole body androgen content in rats on fetal days 18 and 19 without evidence that androgen passes from males to females. Biol. Repro. 44:747751; 1991. 3. Bigsby, R. M.; Aixin, L.; Luo, K.; Cunha, G. R. Strain differences in the ontogeny of estrogen receptors in murine uterine epithelium. Endocrinology 126:2592-2596; 1990. 4. Brown, T.; Maclusky, N.; Shanabrough, M.; Naftolin, F. Comparison of aged- and sex-related changes in cell nuclear estrogen-binding capacity and progestin receptor induction in the rat brain, Endocrinology 126:2965-2972; 1990. 5. Clemens, L. G. Neurohormonal control of male sexual behavior. In: Montagna, W.; Sadler, W. A., eds. Reproductive behavior. New York: Plenum Press; 1974:23-53. 6. Denenberg, V. H., Gaulin-Kremer, E., Gandelman, R., Zarrow, M. X. The development of standard stimulus animals for mouse (Mus musculus) aggression testing by means of olfactory bulbectomy. Anita. Behav, 21:590-598; 1973. 7. Gandelman, R., vom Saal, F. S., Reinisch, J. M. Contiguity to male foetuses affects morphology and behavior of female mice. Nature 266:722-724; 1977. 8. Gandelman, R. Gonadal hormones and the induction ofintraspecific fighting in mice. Neurosci. Biobehav. Rev. 4:133-1409; 1980. 9. Kaufman, J.; Deslypere, J.; Gift, M.; Vermeulen, A. Neuroendocrine regulation of pulsatile leutinizing hormone secretion in elderly men. J. Steroid Biochem. Molec. Biol. 37:421-430; 1990. 10. Klein, W.; Simon, N. Timing of neonatal testosterone exposure in the differentiation of estrogenic regulatory, systems for aggression. Physiol. Behav. 50:91-93: 1991.

1 I. Houtsmuller, E. J.; Slob, A. K. Mascutinization and defeminization of female rats by males located caudally in the uterus. Physiol. Beha~. 48:555-560: 1990. 12. Jubilan, B. M.: Nyby, J. The intrauterine position phenomenon and precopulatory behaviors of house mice. Physiol. Behav. /in press), 13. Meisel, R. L.; Ward, I. L. Fetal female rats are masculinized by male littermates located caudally in the uterus. Science 213:239-249: 1981. 14. Phoenix, C.; Chambers, K. Testosterone therapy in young and old rhesus males that display low levels of sexual activity. Physiol. Behav. 43:479-484; 1988. 15. Richmond, G.; Sachs, B. D. Further evidence tbr masculinization of female rats by males located caudally in utero. Horm. Behav. 18: 484-490; 1984. 16. Rines, J.; vom Saal, F. Fetal effects on sexual behavior and aggression in young and old mice treated with estrogen and testosterone. Horm. Behav. 18:117-129; 1984. 17. Simon, N. G.; Cologer-Clifford, A. In utero contiguity to males does not influence morphology, behavioral sensitivity to testosterone, or hypothalamic androgen binding in CF-1 female mice. Horm. Behav. 25:518-530; 1991, 18. yore Saal, F. S. The interaction of circulating oestrogens and androgens in regulating mammalian sexual differentiation. In: Balthazart, J.: Prove, E.; Giller, R., eds. Hormones and behavior in higher vertebrates. Berlin: Springer-Verlag; 1983:159-177. 19. vom Saal, F. S. Models of early hormonal effects on intrasex aggression in mice. In: Svare, B. B., ed. Hormones and aggressive behavior. New York: Plenum Press: 1984:197-222. 20. vom Saal, F.: Bronson, F.; Desjardins, C. Sexual characteristics of adult female mice are correlated with their blood testosterone levels during prenatal development. Science 208:597-599.