- Email: [email protected]
Endogenous levels of 5 alpha-reduced progestins and androgens in fetal vs. adult rat brains
Developmental Brain Research 115 Ž1999. 17–24
Research report
Endogenous levels of 5 alpha-reduced progestins and androgens in fetal vs. adult rat brains Carol K. Kellogg a
a,)
, Cheryl A. Frye
b
Department of Brain and CognitiÕe Sciences, Room 186 Meliora Hall, UniÕersity of Rochester, Rochester, NY 14627, USA b Department of Psychology, The UniÕersity at Albany-SUNY, Albany, NY 12222, USA Accepted 2 March 1999
Abstract 5 Alpha-reduced metabolites of certain steroids have been shown to have important functions in adult brains and may play a role in brain development. To assess which 5 alpha-reduced steroid metabolites may have an impact during development, endogenous levels of 5 alpha-reduced androgens and progestins and their parent hormones were measured in male and female fetal brains over the last 5 days of gestation. These levels were compared to levels measured in adult male and female brains Ževaluated at different stages of the estrous cycle.. Neither the brain levels of parent hormones nor of their 5 alpha-reduced metabolites varied as a function of fetal sex or of gestational age. Therefore, the data from the two sexes were combined. In fetal brains, the levels of the progesterone reduced metabolites were 20-fold higher than levels of progesterone itself whereas levels of testosterone reduced metabolites were 10-fold lower than testosterone levels. In contrast to fetal brain, conversion of progesterone to reduced metabolites was much lower in adult brain, but the level of 5 alpha-reduced androgens was 3–10-fold higher than the level of testosterone in all adult tissue, indicating more conversion of androgen to 5 alpha-reduced metabolites in adult than in fetal brains. These results imply that the reduction of progesterone to reduced metabolites may play a critical role in brain development. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Neurosteroid; Brain development; Progesterone; Testosterone; Non-genomic
1. Introduction Modulation of GABA A receptors in the brain by 5 alpha-reduced metabolites of several steroids was demonstrated several years ago w18,29,34x. Since then, considerable evidence has accumulated showing that reduced metabolites, particularly of the steroid progesterone, may have important functions in adult brain. Anxiolytic w2,4,39x and anticonvulsant w9,10,14x activity has been attributed to endogenous as well as exogenous reduced steroids. Furthermore, 5 alpha-reduced steroid metabolites may play a role in brain development, perhaps via modulation of GABA A receptor function. In vitro studies have demonstrated that prolonged exposure of embryonic neural cultures to neurosteroids functionally uncouples the GABA A receptor w43x, and the addition of 5 alpha-reduced metabolites of progesterone and androgens to cortical synaptoneu) Corresponding author. [email protected]
Fax:
q 1-716-442-9216;
E-mail:
rosomes prepared from fetal brain potentiates the action of GABA w23x. These studies indicate that GABA A receptors in fetal brain are potential targets of endogenous reduced steroids. Other studies indicate the potential for production of reduced steroids in the developing brain. The activity of the enzyme 5 alpha-reductase in hypothalamus is measurable by gestation day 15 and increases over threefold to reach peak levels at gestation day 18. Activity then declines to moderate levels by birth w26x. Thus, the brain has a marked capacity to produce reduced steroids during fetal development. In situ analysis of mRNA for 5 alpha-reductase over pre- and postnatal development indicates that during rat fetal development the enzyme is present in germinal zones and differentiating fields in the brain w25x. Postnatally, low levels of the mRNA were primarily observed in white matter. Considered together these studies strongly suggest that reduced steroids may indeed play a defined role in prenatal brain development in the rat. However, while there is ample information on the expression of and rate of activity of 5 alpha-reductase during
0165-3806r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 3 8 0 6 Ž 9 9 . 0 0 0 4 1 - 3
18
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
fetal development, information on the actual endogenous levels of any of the potentially important 5 alpha-reduced steroids in fetal brain is clearly lacking. There are several possible sources of substrate for 5 alpha-reductase in the brains of developing organisms. During in utero development, fetuses of both sexes could be exposed to reduced metabolites of progesterone, with the parent hormone derived either from the maternal compartment or produced in the fetal placenta w30x. While testosterone is also produced in the fetal placenta w30x, reduced androgen metabolites in male fetuses also might originate from androgens produced in the fetal testes during the period of the prenatal androgen surge in male rats w7,42x. Furthermore, there is evidence that the brain can synthesize steroid hormones de novo w1x, even in the developing brain w6x. While de novo synthesis is particularly evident in adults when glial cells are segregated w19x and has been demonstrated in glial cell cultures from newborn rats w22x, recent observations indicate the potential for de novo synthesis in developing neurons w40x. We reasoned that in the presence of high levels of activity of 5 alpha-reductase as well as available sources of progesterone and testosterone, 5 alpha-reduced metabolites of these parent compounds should be detectable in fetal brain, and their production from the parent hormone should be pronounced. The objective of the present study, therefore, was to measure endogenous levels of reduced metabolites of progesterone and testosterone in rat brain during late fetal development and compare the levels of the metabolites to that of the parent hormones. By elucidating which endogenous steroid hormones represent substrate for 5 alpha-reductase in fetal brains, the potential roles for specific 5 alpha-reduced steroids can be more clearly addressed. For comparison, the levels of reduced steroid metabolites measured in fetal brains were compared to levels measured in brains of adult males and females Žat different stages of the estrous cycle..
2. Methods 2.1. Animals In the fetal study, adult females ŽLong–Evans wBlue Sprucex strain, Harlan Sprague–Dawley, Altamont, NY. were mated overnight with male rats of the same strain Žbeginning at 1600 h.. Vaginal lavage was used to determine the presence of sperm on the day after mating: the presence of sperm designated gestation day 0, thereby defining parturition as gestation day 22. Pregnant females were maintained under a 12–12 h light–dark cycle Žlights on at 0500 h. with ad lib access to food and water. Pregnant dams were assigned randomly to be killed by rapid cervical luxation followed by decapitation on gestational days 17, 18, 19, 20 or 21. All animals were killed between 1000 and 1200 h. The uterus was quickly re-
moved by abdominal incision and buried in ice. Fetuses were removed individually from the uterus, decapitated and the brains removed and frozen on dry ice. Beginning on gestation day ŽGD. 18, we determined sex by measuring ano-genital ŽAG. distance Žusing calipers and a magnifying lamp.. In a previous study w23x, we had verified the accuracy of AG measurements in fetuses at GD 20 by internal examination of the pups for testes. In removing tissue for endogenous neurochemical measurements, quick freezing of the tissue is critical for accurate measurements. In the present study, internal examination was performed if the AG measurement was indeterminate. No attempt was made to determine sex of the fetuses at GD 17: at that age the brains of all fetuses from one uterine horn were combined to yield one sample. Since quick freezing of the tissue is of paramount importance, and since AG distance is not a reliable measurement at GD 17, we chose not to jeopardize chemical measurements by taking the time for internal examination of each fetus at that age. Thus, at GD 17, two brain samples were obtained per litter and three litters were used, thereby yielding six samples at this age. From GD 18 through GD 21, male and female fetuses from a single uterine horn were separated and the brains from each sex combined to produce a sample, yielding a male and a female sample from each uterine horn in a litter. Three litters were analyzed at each gestational age, providing a total of six samples per age per sex. At day 21, only two litters were used, as one litter was resorbed, thereby resulting in four samples of each sex. Litter size ranged from 11 to 16 fetuses. For analysis of adults, female and male Long–Evans rats were obtained from Charles River Laboratories ŽKingston, NY. at approximately 55 days of age and housed individually in hanging stainless steel cages Ž24 = 18 = 19 cm. in a temperature-controlled room Ž21 " 18C.. Rats were maintained on a 12–12 h light–dark cycle Žlights off at 0800 h. with continuous access to Purina Rat Chow and water. All rats were allowed to acclimate after arrival for four weeks, then vaginal smears were obtained daily from females to determine the day of the estrous cycle w28x. After two normal cycles Ž4–5 daysrcycle., females were randomly assigned to one of the following endocrine groups: diestrus 1, diestrus 2, proestrus, estrus, metestrus. Animals at each estrous stage were killed by rapid decapitation. Whole brains from three females at each stage of the estrous cycle and three males were rapidly removed from the skull and tissue was frozen and stored at y708C until assayed. 2.2. Analysis of parent hormones and reduced metabolites The levels of progesterone, 5a-pregnan-3-20-dione Ždihydroprogesterone, DHP., 3a-hydroxy-5a-pregnan-20-one Ž3a ,5a-THP., testosterone, 5a-androstan-17b-ol-3-one Ždihydrotestosterone, DHT., and 5a-androstan-3a ,17b diol Ž3a-diol. were measured in whole brains of both adults
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
19
steroid. The total assay volumes were as follows: 800 ml for progesterone, 950 ml for DHP and 3a ,5a-THP, 1200 ml for testosterone, DHT, 3a-diol. All assays were incubated overnight at 48C except for 3a-diol which was incubated at room temperature. Separation of bound and free steroid was done by the rapid addition of dextran-coated charcoal. Following incubation with charcoal, samples were centrifuged at 1200 = g and the supernatant was pipetted into a glass scintillation vial with scintillation cocktail. Sample tube concentrations were calculated using the logit–log method of Rodbard and Hutt w35x, interpolation of the standards, and correction for recovery. The intra-assay and inter-assay coefficients of variance for progesterone were 10 and 9%; DHP and 3a ,5a-THP, 12 and 15%; testosterone, 5 and 5%; DHT, 2 and 15%; 3a-diol, 12 and 9%.
and fetuses. This approach was taken to provide an overall profile of the reduction of brain steroids in the fetal vs. adult brains, allowing for more precise probing of different regions in future studies. Tritiated progesterone ŽNET-208: specific activitys 47.5 Cirmmol., 3a ,5a-THP Žused for 3a ,5a-THP and DHP; NET-1047: specific activitys 65.0 Cirmmol., testosterone ŽNET-387: specific activitys 51.0 Cirmmol., DHT ŽNET-302: specific activitys 43.5 Cirmmol., and 3a-diol ŽNET-806: specific activitys 41.00 Cirmmol. used for radioimmunoassays ŽRIA. were purchased from New England Nuclear ŽBoston, MA.. The progesterone antibody Ža337. was obtained from Dr. G.D. Niswender ŽColorado State University., DHP ŽX-947. and 3a ,5a-THP Ž921412-5. antibodies were obtained from Dr. Robert Purdy ŽVeteran’s Medical Center, La Jolla, CA., testosterone antibody ŽT3-125. and DHT antibody ŽDT3-351. were purchased from Endocrine Science ŽCalabasas Hills, CA., and 3a-diol antibody ŽX-144. was purchased from Dr. P.N. Rao ŽSouthwest Foundation for Biomedical Research, San Antonio, TX.. Brain tissue was homogenized with a glassrteflon homogenizer in 2 ml of distilled water and 800 cpm of tritiated compound. Steroids were extracted from the homogenate with diethyl ether, and dried down in an evaporator–drier. The pellet was reconstituted in Trimethyl Pentane ŽTMP. to half the homogenate volume. Reconstituted extracts were separated using Celite column chromatography. Solvents of increasing polarity were used to elute the steroid: progestins Ž100% TMP., DHT Ž5% ethyl acetaterTMP., testosterone and 3a-Diol Ž15% ethyl acetaterTMP.. Fractions were dried using a Savant evaporator drier and then reconstituted in phosphate assay buffer. All steroids were measured according to previously published methods: progesterone w15x; DHP w11,12x; 3a ,5a-THP w14,39x; testosterone, DHT and 3a-diol w8,13x. Antibody concentrations were as follows: progesterone: 1:30,000; 3a ,5a-THP and DHP: 1:5000, testosterone and 3a-diol: 1:20,000; DHT: 1:10,000. All standard curves were prepared in duplicate Žprogesterone, range s 5–800 pg; 3a ,5a-THP, range s 10–4000 pg; DHP, range s 5– 800 pg; testosterone, DHT and 3a-diol, range s 50–2000 pg.. The standards were added to BSA assay buffer, followed by addition of the appropriate antibody and w3 Hx
2.3. Statistical analysis All data were analyzed by a commercially available program ŽStatView 512 q e. using Analysis of Variance ŽANOVA.. Significant effects were probed with post-hoc testing ŽFisher PLSD.. Statistical significance was noted when the probability of a Type I error was - 0.05. 3. Results 3.1. Parent hormone and metabolite leÕels in fetal brain Analysis of endogenous levels of the parent hormones and the reduced metabolites indicated no significant differences between male and female samples over the ages that the sexes were separated for any of the compounds. Therefore, the values for the male and female fetuses in a single uterine horn were averaged to yield a single sample at each gestational age, thereby yielding two samples per litter for a total of six samples at each age. The levels in fetal brain from GD 17 through 21, expressed as ngrg of tissue, are shown in Table 1. Neither the levels of the parent hormones nor of their reduced metabolites varied significantly as a function of age over the last week of gestation. Observation of the levels of progesterone-related compounds indicates that the level of the parent hormone was very low throughout the last week of gestation whereas the
Table 1 Brain steroid levels Žngrg. as a function of gestation age Ždays. a Steroid
Progesterone DHP 3a ,5a-THP Testosterone DHT 3a-diol a
Gestation age 17
18
19
20
21
0.70 " 1.40 18.48 " 3.66 5.48 " 1.19 4.58 " 1.76 0.27 " 0.06 0.71 " 0.26
0.80 " 0.09 12.59 " 0.99 4.18 " 0.45 1.91 " 0.27 0.35 " 0.05 0.40 " 0.05
0.73 " 0.09 10.88 " 1.20 3.55 " 0.46 1.58 " 0.34 0.41 " 0.13 0.32 " 0.08
0.50 " 0.09 10.38 " 1.74 4.08 " 0.51 3.56 " 0.84 0.60 " 0.10 0.33 " 0.11
0.89 " 0.23 14.08 " 1.88 6.04 " 2.34 2.18 " 0.68 0.42 " 0.11 0.50 " 0.14
Data for the two sexes combined.
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
20 Table 2 Steroid levels Žngrg. in adult rat brain Sex
Female Diestrus 1 Diestrus 2 Proestrus Estrus Metestrus Male U
Steroid Progesterone
DHP
3a ,5a-THP
16.59 " 0.54 11.98 " 0.60 U 46.48 " 3.04 U 36.00 " 2.75 16.36 " 4.28
19.56 " 3.46 16.22 " 5.39 30.56 " 0.87 23.57 " 2.49 21.69 " 1.14
1.17 " 0.18 1.80 " 0.52 U 7.76 " 1.21 U 5.87 " 0.80 1.97 " 0.56
15.06 " 0.75
19.16 " 5.60
3.28 " 0.06
Testosterone
DHT
0.042 " 0.006 0.098 " 0.006 0.017 " 0.006 0.111 " 0.006 0.014 " 0.006 U
1.279 q 0.246
3a-diol
0.098 " 0.006 0.091 " 0.006 0.142 " 0.006 0.100 " 0.006 0.025 " 0.006 U
5.885 q 0.003
0.076 " 0.006 0.149 " 0.006 0.198 " 0.006 0.328 " 0.006 0.096 " 0.006 U
3.184 " 1.02
Indicates significant difference from all other groups.
levels of the reduced progesterone metabolites were several-fold higher. In contrast, the levels of testosterone were considerably higher than the levels of its reduced metabolites. These observations implicate a role for reduced progesterone metabolites in fetal brain during the last week of gestation, irrespective of sex or gestational age. 3.2. Parent hormone and metabolite leÕels in adult brain The endogenous levels of progesterone and its reduced metabolites in adult brain are shown in Table 2. ANOVA indicated a significant effect of hormonal state on steroid levels. Post-hoc analysis revealed that the levels of progesterone and 3a ,5a-THP were significantly higher in female brains at proestrus and estrus than in any of the other groups. The levels of progesterone and its metabolites in male brain did not differ from those in female brains at any stage of estrous except at proestrus and estrus. Analysis of the levels of testosterone and its reduced metabolites Ždata shown in Table 2. indicated a significant effect of hormonal status, with levels of all androgens in male brains being considerably higher than in female brains at any stage of the estrous cycle. However, despite the very
low levels of androgens in female brains, analysis of levels in females alone as a function of estrous cycle did indicate that testosterone and both reduced metabolites varied significantly with estrous cycle. Based upon post-hoc analysis, testosterone appeared to reach two peaks over the cycle, at diestrus 2 and estrus, with troughs at proestrus and metestrus. The levels of DHT increased significantly with each successive stage from diestrus 1 to proestrus then decreased significantly at each stage until diestrus 1. Similarly, the levels of 3a-diol increased significantly with each stage from diestrus 1 to estrus and decreased dramatically following estrus. Therefore, the levels of reduced androgen metabolites are higher in female brains at proestrus and estrus than at other stages of the estrous cycle. While the levels of androgens in adult female brains overall are quite low, the changes that take place in androgen metabolism over the estrous cycle may have functional implications. 3.3. Comparison of steroid leÕels in fetal Õs. adult brains Comparing the hormone levels in fetal and adult brains, it is clear that progesterone levels are much higher in adult
Fig. 1. Utilization of progesterone and testosterone during late gestation in the rat brain, averaged across sexes. Left panel depicts the ratio of the progesterone metabolites, DHP and 3a ,5a-THP, to the parent hormone, progesterone ŽP.. Right panel depicts the ratio of the androgen metabolites, DHT and 3a-diol, to the parent hormone, testosterone ŽT..
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
21
Fig. 2. Utilization of progesterone and testosterone in adult rat female brains at different stages of the estrous cycle and in adult rat male brains. Left panel depicts the ratio of the progesterone metabolites, DHP and 3a ,5a-THP, to the parent hormone, progesterone ŽP.. Right panel depicts the ratio of the androgen metabolites, DHT and 3a-diol, to the parent hormone, testosterone ŽT.. Stages of the female estrous cycle: D1 s diestrus 1, D2 s diestrus 2, P s proestrus, E s estrus, ME s metestrus. U Indicates that females at proestrus are significantly different from all other groups.
brains Žregardless of sex. than fetal brains, whereas in fetal brains, the levels of reduced progesterone metabolites are comparable to levels in adult brain and considerably higher than that of their parent hormone. Fetal levels of testosterone are comparable to those measured in adult male brain and considerably higher than in adult female brain. On the other hand, the levels of reduced androgen metabolites are higher than those of testosterone in adult male brain but not in fetal brain. These observations suggest that turnover of the parent hormones differs between fetal and adult brain. 3.4. Analysis of the ratio of reduced metabolites to parent hormone in fetal brain To provide an index of the conversion of parent hormone to its reduced metabolites, the ratio of the reduced metabolites to parent hormone was calculated for each sample. As illustrated in Fig. 1A, the ratio of the reduced progesterone metabolites ŽDHP plus 3a ,5a-THP. to progesterone in fetal brain did not vary significantly over gestational age. However, at all ages the ratio was very high Ž20–30., indicating considerable production of reduced progesterone metabolites from the parent hormone in fetal brain. In contrast, the ratio of reduced testosterone metabolites ŽDHT plus 3a-diol. to testosterone ŽFig. 1B. was below 1 at all ages and did not vary significantly across gestational age. Therefore, very little testosterone in fetal brain appears to be converted to reduced metabolites. 3.5. Analysis of the ratio of reduced metabolites to parent hormone in adult brain Analysis of the conversion of parent hormone to metabolites in adult brains revealed a very different profile
from that observed in fetal brain. The ratio of reduced progesterone metabolites to progesterone was around 1 and did not vary significantly with hormonal status ŽFig. 2A.. Thus, in comparison to fetal brains, the conversion of progesterone to reduced metabolites appears to be considerably less in adult brain and does not differ between the two sexes. In contrast, the ratio of reduced testosterone metabolites to testosterone ŽFig. 2B. was higher in all adult hormonal groups than in fetal tissue, even in adult female brain where the overall levels of androgens were much lower than in fetal brain. Furthermore, analysis of this ratio in adult brain indicated that the ratio varied significantly with hormonal group, with the ratio measured in female brains at proestrus considerably higher than in any other groups Žincluding adult males.. Thus, the reduction of androgens may play a different role in adult brain as compared to fetal brain.
4. Discussion Endogenous levels of reduced metabolites of steroid hormones were detected in the brain at all gestational ages evaluated; however, neither the levels of parent hormones nor of reduced metabolites appeared to differ as a function of fetal sex. While it could be argued that the accuracy of determination of fetal sex was suspect due to the reliance primarily on AG distance, the lack of an impact of sex on endogenous levels of reduced steroid metabolites in fetal brain is consistent with the observation that the activity of 5 alpha-reductase, measured in vitro, in the brain does not differ between sexes at any developmental stage analyzed w26x. Furthermore, none of the hormones measured in the present study varied as a function of gestational age, neither when analyzed by sex nor when the data for the
22
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
sexes were pooled. Compared to literature reports on the levels of progesterone and testosterone in fetal circulation during late gestation, it appears unlikely, therefore, that reduced metabolites in fetal brain originate solely from parent hormones taken up from the fetal circulation. Progesterone levels in fetal circulation reportedly decrease consistently throughout the last week of gestation from high levels measured at GD 17 Žwith day 0 of gestation designated as the day of mating., and the circulating levels do not differ between male and female fetuses w42x. In contrast to this profile, reduced progesterone metabolites remained high in fetal brain from GD 17 through 21. Circulating testosterone levels in male fetuses are higher than in female fetuses at GD 17 and increase markedly at GD 18, decreasing thereafter. Circulating levels of testosterone in female fetuses remain fairly constant throughout the last week of gestation w42x. Consistent adult-like Ži.e., adult male-like. levels of testosterone were observed in fetal brain throughout this period. The lack of any impact from sex or gestational age on fetal brain levels of progesterone, testosterone and their reduced metabolites raises the question of whether the source of the parent hormones may have been de novo. The results obtained Ždata pooled across sex. suggest that the reduction of progesterone may be important during fetal brain development: the levels of reduced progesterone metabolites were 10- to 30-fold higher than the levels of the parent hormone. In contrast, the levels of reduced testosterone metabolites were around 10-fold lower than that of the parent hormone. Considering the critical importance of the aromatization of testosterone to estrogens for sexual differentiation of the brain w20,32x, the low levels of 5 alpha-reduced androgen metabolites observed during late gestation even in male brains is not surprising. Additionally, reduced androgens are endogenous inhibitors of aromatase w21x, therefore, low levels of reduced androgen metabolites in fetal brain may be protective. The results of the present study suggest that even in fetal male brain the reduction of androgens is of lesser importance than other routes of androgen metabolism during development. The high endogenous levels of reduced progesterone metabolites in fetal brains relative to the level of parent hormone raises the question of what function the reduction of progesterone might play in fetal brain development. Reduced metabolites of several steroid hormones interact with GABA A and glutamate receptors w24,31,34x. During early brain development, GABA may act as a trophic factor w27x. And while GABA acts as an inhibitory substance in adult brain w37x, it can exert depolarizing actions in developing brain w5x. Reduced progesterone metabolites may participate in these actions of GABA in the developing brain. In support of this mechanism, the addition of 3a ,5a-THP to hippocampal cultures prepared from 18 day rat fetuses induced neurite retraction, and modulation of chloride channels was considered important to this effect w3x. However, certain reduced steroids also have been
shown to act intracellulary at the progesterone receptor w36x. The high rate of conversion of progesterone to its reduced metabolites in fetal brain may provide a mechanism for cross-talk between membrane and genomic hormonal events. Recent evidence suggests that progesterone potentially could be involved in sexual differentiation of the brain: progestin receptor immunoreactivity was observed in specific hypothalamic nuclei in male rats at fetal and neonatal ages, but was absent in female rats brains w41x. While the results of the present study indicated no sex differences in 5 alpha-reduced metabolites of progesterone in the brain during the prenatal period of sexual differentiation, the reduced progestins produced during fetal development could still affect sexual differentiation because of other sex specific factors. The role of progesterone in late fetal brain development in the rat warrants further study. Clearly, the reduction of progesterone and testosterone in vivo differed between fetal and adult brain. Whereas progesterone was readily converted to its reduced metabolites in fetal brain, the reduction of progesterone in adult brain Žmale and female. was considerably less. The absolute levels of progesterone and 3a ,5a-THP in adult female brain appear to follow changes in circulating progesterone levels over the estrous cycle. The levels of both steroids were significantly higher in female brain at proestrus and estrus, and circulating progesterone levels reportedly increase at proestrus and remain elevated into estrus w11,12,38x. The progesterone substrate for 5 alpha-reductase in adult female brain, therefore, may be circulating progesterone, whereas circulating progesterone may not be the substrate for 5 alpha-reductase in fetal brain. In contrast to the reduction of progesterone, testosterone was more actively converted to reduced metabolites in the adult brain than in fetal brain. While the endogenous levels of all of the androgen hormones were markedly higher in adult male brains than in adult female brains, changes in the reduction of testosterone did take place in female brains over the estrous cycle. The marked turnover of testosterone to reduced androgen metabolites observed in the female brains at proestrus is consistent with the observations that 5 alpha-reduced androgens synergize with 5 alpha-reduced progestins to initially facilitate sexual receptivity and then to terminate mating in the female rat w13,16,17x. More relevant, however, to the focus of the present study is the question of what determines the change in the reduction of brain steroid hormones that must take place across postnatal development. Progesterone clearly appears to be the primary substrate for 5 alpha-reductase in the fetal brain, whereas in the adult brain testosterone may have a slight edge: the ratio of metabolites to parent hormone was between 1–2 for progesterone and 3–10 for testosterone in the adult, excepting the female at proestrus, when this ratio was even greater. Progesterone has a higher affinity for 5 alpha-reductase than does testosterone w26x, but there has been no indication that substrate affinity differs between fetal and adult brain. The observation in
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
the present study that progesterone turnover to its reduced metabolites decreases as testosterone turnover increases between fetal ages and adulthood would suggest that substrate affinity alone probably is not responsible for the apparent shift in preferred substrate. The enzyme 5 alphareductase can exist in more than one isoform, and in rat brain, the predominant form appears to be 5 alpha-reductase type 1 w33x. There has been no indication, though, that the preferred isoform is different in the brain during early development than during adulthood. It is critical, therefore, to evaluate the reduction of steroids during early postnatal and adolescent development in order to determine the developmental stageŽs. where the reduction of progesterone becomes diminished and the reduction of androgens increases, as this information may help elucidate reasons for the apparent shift in preferred hormone substrate by brain 5 alpha-reductase from late fetal ages to adulthood. In summary, the measurement of endogenous levels of 5 alpha-reduced steroids and their parent hormones has indicated that the reduction of progesterone by 5 alpha-reductase in fetal brain rather than the reduction of androgens may play a critical role in early brain development. Further work is necessary to localize specific regions of the brain where this process may be most active.
Acknowledgements This work was supported in part by PHS grant no. DA 07080 to C.K. Kellogg and NSF CAREER grant IBN9514463, Whitehall Foundation grant F96-10, and Donaghue Foundation grant 96-001 to C.A. Frye.
References w1x E-E. Baulieu, P. Robel, Neurosteroids: a new brain function, J. Steroid Biochem. Mol. Biol. 37 Ž1990. 395–403. w2x D. Bitran, R.J. Hilvers, C.K. Kellogg, Anxiolytic effects of 3a-hydroxy-5a wb x-pregnan-20-one: endogenous metabolites of progesterone that are active at the GABA A receptor, Brain Res. 561 Ž1991. 157–161. w3x R.D. Brinton, The neurosteroid 3a-hydroxy-5a-pregnane-20-one induces cytoarchitectural regression in cultured fetal hippocampal neurons. J. Neurosci. 14 Ž1994. 2763–2774. w4x M.D. Brot, Y. Akwa, R.H. Purdy, G.F. Koob, G.F. Britton, The anxiolytic-like effects of the neurosteroid allopregnanolone: interactions with GABA A receptors, Eur. J. Pharmacol. 325 Ž1997. 1–7. w5x E. Cherbini, J.L. Gaiarsa, Y. Ben-Ari, GABA: an excitatory transmitter in early postnatal life, Trends Neurosci. 14 Ž1991. 515–519. w6x N.A. Compagnone, A. Bulfone, J.L.R. Rubenstein, S.H. Mellon, Expression of the steroidogenic enzyme P450scc in the central and peripheral nervous systems during rodent embryogenesis, Endocrinology 136 Ž1995. 2689–2696. w7x P. Corbier, B. Kerdelhue, R. Picon, J. Roffi, Changes in testicular weight and serum gonadotropin and testosterone levels before, during, and after birth in the perinatal rat, Endocrinology 103 Ž1978. 985–992.
23
w8x M. Erskine, N. Hippensteil, E. Kornberg, Metabolism of dihydrotestosterone to 3a-androstanediol in brain and plasma: effect on behavioural activity in female rats, J. Endocrinol. 134 Ž1992. 183– 195. w9x D.A. Finn, K.W. Gee, The estrus cycle, sensitivity to convulsants and the anticonvulsant effect of a neuroactive steroid, J. Pharmacol. Exp. Ther. 271 Ž1994. 164–170. w10x C.A. Frye, The neurosteroid 3a ,5a-THP has anti-seizure and possible neuroprotective effects in an animal model of epilepsy, Brain Res. 696 Ž1995. 113–120. w11x C.A. Frye, L.E. Bayon, Seizure activity is increased in endocrine states characterized by decline in endogenous 3a ,5a-THP, Neuroendocrinology 68 Ž1998. 272–280. w12x C.A. Frye, L.E. Bayon, Pregnane and androstane neurosteroids are increased in mated rats and across various hormonal and reproductive milieru, J. Neuroendocrinol. Žin press.. w13x C.A. Frye, K.R. van Keuran, M.S. Erskine, Behavioral effects of 3a-androstandiol: I. Modulation of sexual receptivity and promotion of GABA-stimulated chloride flux, Behav. Brain Res. 79 Ž1996. 109–118. w14x C.A. Frye, T.J. Scalise, L.E. Bayon, Finasteride blocks the reduction in ictal activity produced by exogenous estrous cyclicity, J. Neuroendocrinol. 10 Ž1998. 291–296. w15x C.A. Frye, C.M. McCormick, C. Coopersmith, M.S. Erskine, Effects of paced and non-paced mating stimulation on plasma progesterone, 3a-diol and corticosterone, Psychoneuroendocrinology 21 Ž1996. 431–439. w16x C.A. Frye, J.E. Duncan, M. Basham, M.E. Erskine, Behavioral effects of 3a-androstanediol: II. Hypothalamic and preoptic area actions via GABAergic mechanism, Behav. Brain Res. 79 Ž1996. 119–130. w17x C.A. Frye, K.R. van Keuran, P.N. Rao, M.S. Erskine, Progesterone and 3a-androstanediol conjugated to bovine serum albumin affects estrous behavior when applied to the MBH and POA, Behav. Neurosci. 110 Ž1996. 603–612. w18x K.W. Gee, M.B. Bolger, R.D. Brinton, H. Coirini, B.S. McEwen, Steroid modulation of the chloride ionophore in rat brain: structure–activity requirements, regional dependence and mechanism of action, J. Pharmacol. Exp. Ther. 246 Ž1988. 803–812. w19x Z.Y. Hu, E. Bourreau, I. Jungo-Testas, P. Robel, E-E. Baulieu, Oligodendrocyte mitochondria convert cholesterol to pregnenolone, Proc. Natl. Acad. Sci. U.S.A. 84 Ž1987. 8129–8215. w20x J.B. Hutchison, T. Steimer, Androgen metabolism in the brain: behavioural correlates, Prog. Brain Res. 61 Ž1984. 23–51. w21x J.B. Hutchison, C. Beyer, R.E. Hutchison, A. Wozniak, Sex differences in the regulation of embryonic brain aromatase, J. Steroid Biochem. Mol. Biol. 61 Ž1997. 315–322. w22x I. Jung-Testas, Z.Y. Hu, P. Robel, E-E. Baulieu, Biosynthesis of pregnanolone and progesterone in primary cultures of rat glial cells, Endocrinology 125 Ž1989. 2083–2091. w23x C.K. Kellogg, V.G. Olson, G.L. Pleger, Neurosteroid action at the GABA A receptor in fetal rat forebrain, Dev. Brain Res. 108 Ž1998. 131–137. w24x J.J. Lambert, D. Belelli, C. Hill-Venning, J.A. Peters, Neurosteroids and GABA A receptor function, Trends Pharmacol. Sci. 16 Ž1995. 295–303. w25x M.E. Lauber, W. Lichtensteiger, Ontogeny of 5a-reductase Žtype 1. messenger ribonucleic acid expression in rat brain: early presence in germinal zones, Endocrinology 137 Ž1996. 2718–2730. w26x E.D. Lephart, Brain 5a-reductase: cellular, enzymatic, and molecular perspectives and implications for biological function, Mol. Cell. Neurosci. 4 Ž1993. 473–484. w27x F.M. Leslie, Neurotransmitters as neurotrophic factors, in: S.E. Loughlin, J.H. Fallon ŽEds.., Neurotrophic Factors, Academic Press, San Diego, 1993, pp. 565–598. w28x J.A. Long, H.M. Evans, Oestrus cycle in the rat and its associated phenomena, Memoirs Univ. California 6 Ž1922. 1–146.
24
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
w29x M.D. Majewska, Neurosteroids: endogenous bimodal modulators of the GABA A receptor. Mechanism of action and physiological significance, Prog. Neurobiol. 38 Ž1992. 379–395. w30x D.W. Matt, G.J. McDonald, In vitro progesterone and testosterone production by the rat placenta during pregnancy, Endocrinology 115 Ž1984. 741–747. w31x B.S. McEwen, Non-genomic and genomic effects of steroids on neural activity, Trends Pharmacol. Sci. 12 Ž1991. 141–147. w32x R. Naftolin, Brain aromatization of androgens, J. Reprod. Med. 39 Ž1994. 257–261. w33x K. Normington, D.W. Russell, Tissue distribution and kinetic characteristics of rat steroid 5a-reductase isozymes, J. Biol. Chem. 267 19548–19554. w34x S.M. Paul, R.H. Purdy, Neuroactive steroids, FASEB J. 6 Ž1992. 51–63. w35x D. Rodbard, D. M. Hutt, Statistical analysis of radioimmunoassay and immunoradiometric assays: a generalized, weighted iterative, least squares method for logistic curve fitting, in: International Atomic Energy Agency ŽEd.., Symposium on Radioimmunoassay and Related Procedures in Medicine, Uniput, New York, 1974. w36x R. Rupprecht, C.A.E. Hauser, T. Trapp, F. Holsboer, Neurosteroids: molecular mechanisms of action and psychopharmacological significance, J. Steroid Biochem. Mol. Biol. 56 Ž1996. 163–168.
w37x W. Singer, Neurophysiology: the changing face of inhibition, Curr. Biol. 6 Ž1996. 395–397. w38x S.S. Smith, D.J. Woodward, J.K. Chapin, Sex steroids modulate motor-correlated increases in cerebellar discharge, Brain Res. 476 Ž1989. 307–316. w39x S.S. Smith, Q.H. Gong, X. Li, M. Moran, D. Bitran, C.A. Frye, F.C. Hsu, Withdrawal from 3a-OH-5a-pregnan-20-one using a pseudopregnancy model alters the kinetics of hippocampal GABA A -gated current and increases the GABA A receptor a4 subunit in association with increased anxiety, J. Neurosci. 18 Ž1998. 5275–5284. w40x K. Ukena, M. Usui, C. Kohchi, K. Tsutsui, Cytochrome P450 side-chain cleavage enzyme in the cerebellar purkinje neuron and its neonatal change in rats, Endocrinology 139 Ž1998. 137–147. w41x C.K. Wagner, A.T. Nakayama, G.J. De Vries, Potential role of maternal progesterone in the sexual differentiation of the brain, Endocrinology 139 Ž1998. 3658–3661. w42x J. Weisz, I.L. Ward, Plasma testosterone and progesterone titers of pregnant rats, their male and female fetuses, and neonatal offspring, Endocrinology 106 Ž1980. 306–316. w43x R. Yu, M.K. Ticku, Chronic neurosteroid treatment produces functional heterologous uncoupling at the g-aminobutyric acid type Arbenzodiazepine receptor complex in mammalian cortical neurons, Mol. Pharmacol. 47 Ž1995. 603–610.
Research report
Endogenous levels of 5 alpha-reduced progestins and androgens in fetal vs. adult rat brains Carol K. Kellogg a
a,)
, Cheryl A. Frye
b
Department of Brain and CognitiÕe Sciences, Room 186 Meliora Hall, UniÕersity of Rochester, Rochester, NY 14627, USA b Department of Psychology, The UniÕersity at Albany-SUNY, Albany, NY 12222, USA Accepted 2 March 1999
Abstract 5 Alpha-reduced metabolites of certain steroids have been shown to have important functions in adult brains and may play a role in brain development. To assess which 5 alpha-reduced steroid metabolites may have an impact during development, endogenous levels of 5 alpha-reduced androgens and progestins and their parent hormones were measured in male and female fetal brains over the last 5 days of gestation. These levels were compared to levels measured in adult male and female brains Ževaluated at different stages of the estrous cycle.. Neither the brain levels of parent hormones nor of their 5 alpha-reduced metabolites varied as a function of fetal sex or of gestational age. Therefore, the data from the two sexes were combined. In fetal brains, the levels of the progesterone reduced metabolites were 20-fold higher than levels of progesterone itself whereas levels of testosterone reduced metabolites were 10-fold lower than testosterone levels. In contrast to fetal brain, conversion of progesterone to reduced metabolites was much lower in adult brain, but the level of 5 alpha-reduced androgens was 3–10-fold higher than the level of testosterone in all adult tissue, indicating more conversion of androgen to 5 alpha-reduced metabolites in adult than in fetal brains. These results imply that the reduction of progesterone to reduced metabolites may play a critical role in brain development. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Neurosteroid; Brain development; Progesterone; Testosterone; Non-genomic
1. Introduction Modulation of GABA A receptors in the brain by 5 alpha-reduced metabolites of several steroids was demonstrated several years ago w18,29,34x. Since then, considerable evidence has accumulated showing that reduced metabolites, particularly of the steroid progesterone, may have important functions in adult brain. Anxiolytic w2,4,39x and anticonvulsant w9,10,14x activity has been attributed to endogenous as well as exogenous reduced steroids. Furthermore, 5 alpha-reduced steroid metabolites may play a role in brain development, perhaps via modulation of GABA A receptor function. In vitro studies have demonstrated that prolonged exposure of embryonic neural cultures to neurosteroids functionally uncouples the GABA A receptor w43x, and the addition of 5 alpha-reduced metabolites of progesterone and androgens to cortical synaptoneu) Corresponding author. [email protected]
Fax:
q 1-716-442-9216;
E-mail:
rosomes prepared from fetal brain potentiates the action of GABA w23x. These studies indicate that GABA A receptors in fetal brain are potential targets of endogenous reduced steroids. Other studies indicate the potential for production of reduced steroids in the developing brain. The activity of the enzyme 5 alpha-reductase in hypothalamus is measurable by gestation day 15 and increases over threefold to reach peak levels at gestation day 18. Activity then declines to moderate levels by birth w26x. Thus, the brain has a marked capacity to produce reduced steroids during fetal development. In situ analysis of mRNA for 5 alpha-reductase over pre- and postnatal development indicates that during rat fetal development the enzyme is present in germinal zones and differentiating fields in the brain w25x. Postnatally, low levels of the mRNA were primarily observed in white matter. Considered together these studies strongly suggest that reduced steroids may indeed play a defined role in prenatal brain development in the rat. However, while there is ample information on the expression of and rate of activity of 5 alpha-reductase during
0165-3806r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 3 8 0 6 Ž 9 9 . 0 0 0 4 1 - 3
18
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
fetal development, information on the actual endogenous levels of any of the potentially important 5 alpha-reduced steroids in fetal brain is clearly lacking. There are several possible sources of substrate for 5 alpha-reductase in the brains of developing organisms. During in utero development, fetuses of both sexes could be exposed to reduced metabolites of progesterone, with the parent hormone derived either from the maternal compartment or produced in the fetal placenta w30x. While testosterone is also produced in the fetal placenta w30x, reduced androgen metabolites in male fetuses also might originate from androgens produced in the fetal testes during the period of the prenatal androgen surge in male rats w7,42x. Furthermore, there is evidence that the brain can synthesize steroid hormones de novo w1x, even in the developing brain w6x. While de novo synthesis is particularly evident in adults when glial cells are segregated w19x and has been demonstrated in glial cell cultures from newborn rats w22x, recent observations indicate the potential for de novo synthesis in developing neurons w40x. We reasoned that in the presence of high levels of activity of 5 alpha-reductase as well as available sources of progesterone and testosterone, 5 alpha-reduced metabolites of these parent compounds should be detectable in fetal brain, and their production from the parent hormone should be pronounced. The objective of the present study, therefore, was to measure endogenous levels of reduced metabolites of progesterone and testosterone in rat brain during late fetal development and compare the levels of the metabolites to that of the parent hormones. By elucidating which endogenous steroid hormones represent substrate for 5 alpha-reductase in fetal brains, the potential roles for specific 5 alpha-reduced steroids can be more clearly addressed. For comparison, the levels of reduced steroid metabolites measured in fetal brains were compared to levels measured in brains of adult males and females Žat different stages of the estrous cycle..
2. Methods 2.1. Animals In the fetal study, adult females ŽLong–Evans wBlue Sprucex strain, Harlan Sprague–Dawley, Altamont, NY. were mated overnight with male rats of the same strain Žbeginning at 1600 h.. Vaginal lavage was used to determine the presence of sperm on the day after mating: the presence of sperm designated gestation day 0, thereby defining parturition as gestation day 22. Pregnant females were maintained under a 12–12 h light–dark cycle Žlights on at 0500 h. with ad lib access to food and water. Pregnant dams were assigned randomly to be killed by rapid cervical luxation followed by decapitation on gestational days 17, 18, 19, 20 or 21. All animals were killed between 1000 and 1200 h. The uterus was quickly re-
moved by abdominal incision and buried in ice. Fetuses were removed individually from the uterus, decapitated and the brains removed and frozen on dry ice. Beginning on gestation day ŽGD. 18, we determined sex by measuring ano-genital ŽAG. distance Žusing calipers and a magnifying lamp.. In a previous study w23x, we had verified the accuracy of AG measurements in fetuses at GD 20 by internal examination of the pups for testes. In removing tissue for endogenous neurochemical measurements, quick freezing of the tissue is critical for accurate measurements. In the present study, internal examination was performed if the AG measurement was indeterminate. No attempt was made to determine sex of the fetuses at GD 17: at that age the brains of all fetuses from one uterine horn were combined to yield one sample. Since quick freezing of the tissue is of paramount importance, and since AG distance is not a reliable measurement at GD 17, we chose not to jeopardize chemical measurements by taking the time for internal examination of each fetus at that age. Thus, at GD 17, two brain samples were obtained per litter and three litters were used, thereby yielding six samples at this age. From GD 18 through GD 21, male and female fetuses from a single uterine horn were separated and the brains from each sex combined to produce a sample, yielding a male and a female sample from each uterine horn in a litter. Three litters were analyzed at each gestational age, providing a total of six samples per age per sex. At day 21, only two litters were used, as one litter was resorbed, thereby resulting in four samples of each sex. Litter size ranged from 11 to 16 fetuses. For analysis of adults, female and male Long–Evans rats were obtained from Charles River Laboratories ŽKingston, NY. at approximately 55 days of age and housed individually in hanging stainless steel cages Ž24 = 18 = 19 cm. in a temperature-controlled room Ž21 " 18C.. Rats were maintained on a 12–12 h light–dark cycle Žlights off at 0800 h. with continuous access to Purina Rat Chow and water. All rats were allowed to acclimate after arrival for four weeks, then vaginal smears were obtained daily from females to determine the day of the estrous cycle w28x. After two normal cycles Ž4–5 daysrcycle., females were randomly assigned to one of the following endocrine groups: diestrus 1, diestrus 2, proestrus, estrus, metestrus. Animals at each estrous stage were killed by rapid decapitation. Whole brains from three females at each stage of the estrous cycle and three males were rapidly removed from the skull and tissue was frozen and stored at y708C until assayed. 2.2. Analysis of parent hormones and reduced metabolites The levels of progesterone, 5a-pregnan-3-20-dione Ždihydroprogesterone, DHP., 3a-hydroxy-5a-pregnan-20-one Ž3a ,5a-THP., testosterone, 5a-androstan-17b-ol-3-one Ždihydrotestosterone, DHT., and 5a-androstan-3a ,17b diol Ž3a-diol. were measured in whole brains of both adults
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
19
steroid. The total assay volumes were as follows: 800 ml for progesterone, 950 ml for DHP and 3a ,5a-THP, 1200 ml for testosterone, DHT, 3a-diol. All assays were incubated overnight at 48C except for 3a-diol which was incubated at room temperature. Separation of bound and free steroid was done by the rapid addition of dextran-coated charcoal. Following incubation with charcoal, samples were centrifuged at 1200 = g and the supernatant was pipetted into a glass scintillation vial with scintillation cocktail. Sample tube concentrations were calculated using the logit–log method of Rodbard and Hutt w35x, interpolation of the standards, and correction for recovery. The intra-assay and inter-assay coefficients of variance for progesterone were 10 and 9%; DHP and 3a ,5a-THP, 12 and 15%; testosterone, 5 and 5%; DHT, 2 and 15%; 3a-diol, 12 and 9%.
and fetuses. This approach was taken to provide an overall profile of the reduction of brain steroids in the fetal vs. adult brains, allowing for more precise probing of different regions in future studies. Tritiated progesterone ŽNET-208: specific activitys 47.5 Cirmmol., 3a ,5a-THP Žused for 3a ,5a-THP and DHP; NET-1047: specific activitys 65.0 Cirmmol., testosterone ŽNET-387: specific activitys 51.0 Cirmmol., DHT ŽNET-302: specific activitys 43.5 Cirmmol., and 3a-diol ŽNET-806: specific activitys 41.00 Cirmmol. used for radioimmunoassays ŽRIA. were purchased from New England Nuclear ŽBoston, MA.. The progesterone antibody Ža337. was obtained from Dr. G.D. Niswender ŽColorado State University., DHP ŽX-947. and 3a ,5a-THP Ž921412-5. antibodies were obtained from Dr. Robert Purdy ŽVeteran’s Medical Center, La Jolla, CA., testosterone antibody ŽT3-125. and DHT antibody ŽDT3-351. were purchased from Endocrine Science ŽCalabasas Hills, CA., and 3a-diol antibody ŽX-144. was purchased from Dr. P.N. Rao ŽSouthwest Foundation for Biomedical Research, San Antonio, TX.. Brain tissue was homogenized with a glassrteflon homogenizer in 2 ml of distilled water and 800 cpm of tritiated compound. Steroids were extracted from the homogenate with diethyl ether, and dried down in an evaporator–drier. The pellet was reconstituted in Trimethyl Pentane ŽTMP. to half the homogenate volume. Reconstituted extracts were separated using Celite column chromatography. Solvents of increasing polarity were used to elute the steroid: progestins Ž100% TMP., DHT Ž5% ethyl acetaterTMP., testosterone and 3a-Diol Ž15% ethyl acetaterTMP.. Fractions were dried using a Savant evaporator drier and then reconstituted in phosphate assay buffer. All steroids were measured according to previously published methods: progesterone w15x; DHP w11,12x; 3a ,5a-THP w14,39x; testosterone, DHT and 3a-diol w8,13x. Antibody concentrations were as follows: progesterone: 1:30,000; 3a ,5a-THP and DHP: 1:5000, testosterone and 3a-diol: 1:20,000; DHT: 1:10,000. All standard curves were prepared in duplicate Žprogesterone, range s 5–800 pg; 3a ,5a-THP, range s 10–4000 pg; DHP, range s 5– 800 pg; testosterone, DHT and 3a-diol, range s 50–2000 pg.. The standards were added to BSA assay buffer, followed by addition of the appropriate antibody and w3 Hx
2.3. Statistical analysis All data were analyzed by a commercially available program ŽStatView 512 q e. using Analysis of Variance ŽANOVA.. Significant effects were probed with post-hoc testing ŽFisher PLSD.. Statistical significance was noted when the probability of a Type I error was - 0.05. 3. Results 3.1. Parent hormone and metabolite leÕels in fetal brain Analysis of endogenous levels of the parent hormones and the reduced metabolites indicated no significant differences between male and female samples over the ages that the sexes were separated for any of the compounds. Therefore, the values for the male and female fetuses in a single uterine horn were averaged to yield a single sample at each gestational age, thereby yielding two samples per litter for a total of six samples at each age. The levels in fetal brain from GD 17 through 21, expressed as ngrg of tissue, are shown in Table 1. Neither the levels of the parent hormones nor of their reduced metabolites varied significantly as a function of age over the last week of gestation. Observation of the levels of progesterone-related compounds indicates that the level of the parent hormone was very low throughout the last week of gestation whereas the
Table 1 Brain steroid levels Žngrg. as a function of gestation age Ždays. a Steroid
Progesterone DHP 3a ,5a-THP Testosterone DHT 3a-diol a
Gestation age 17
18
19
20
21
0.70 " 1.40 18.48 " 3.66 5.48 " 1.19 4.58 " 1.76 0.27 " 0.06 0.71 " 0.26
0.80 " 0.09 12.59 " 0.99 4.18 " 0.45 1.91 " 0.27 0.35 " 0.05 0.40 " 0.05
0.73 " 0.09 10.88 " 1.20 3.55 " 0.46 1.58 " 0.34 0.41 " 0.13 0.32 " 0.08
0.50 " 0.09 10.38 " 1.74 4.08 " 0.51 3.56 " 0.84 0.60 " 0.10 0.33 " 0.11
0.89 " 0.23 14.08 " 1.88 6.04 " 2.34 2.18 " 0.68 0.42 " 0.11 0.50 " 0.14
Data for the two sexes combined.
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
20 Table 2 Steroid levels Žngrg. in adult rat brain Sex
Female Diestrus 1 Diestrus 2 Proestrus Estrus Metestrus Male U
Steroid Progesterone
DHP
3a ,5a-THP
16.59 " 0.54 11.98 " 0.60 U 46.48 " 3.04 U 36.00 " 2.75 16.36 " 4.28
19.56 " 3.46 16.22 " 5.39 30.56 " 0.87 23.57 " 2.49 21.69 " 1.14
1.17 " 0.18 1.80 " 0.52 U 7.76 " 1.21 U 5.87 " 0.80 1.97 " 0.56
15.06 " 0.75
19.16 " 5.60
3.28 " 0.06
Testosterone
DHT
0.042 " 0.006 0.098 " 0.006 0.017 " 0.006 0.111 " 0.006 0.014 " 0.006 U
1.279 q 0.246
3a-diol
0.098 " 0.006 0.091 " 0.006 0.142 " 0.006 0.100 " 0.006 0.025 " 0.006 U
5.885 q 0.003
0.076 " 0.006 0.149 " 0.006 0.198 " 0.006 0.328 " 0.006 0.096 " 0.006 U
3.184 " 1.02
Indicates significant difference from all other groups.
levels of the reduced progesterone metabolites were several-fold higher. In contrast, the levels of testosterone were considerably higher than the levels of its reduced metabolites. These observations implicate a role for reduced progesterone metabolites in fetal brain during the last week of gestation, irrespective of sex or gestational age. 3.2. Parent hormone and metabolite leÕels in adult brain The endogenous levels of progesterone and its reduced metabolites in adult brain are shown in Table 2. ANOVA indicated a significant effect of hormonal state on steroid levels. Post-hoc analysis revealed that the levels of progesterone and 3a ,5a-THP were significantly higher in female brains at proestrus and estrus than in any of the other groups. The levels of progesterone and its metabolites in male brain did not differ from those in female brains at any stage of estrous except at proestrus and estrus. Analysis of the levels of testosterone and its reduced metabolites Ždata shown in Table 2. indicated a significant effect of hormonal status, with levels of all androgens in male brains being considerably higher than in female brains at any stage of the estrous cycle. However, despite the very
low levels of androgens in female brains, analysis of levels in females alone as a function of estrous cycle did indicate that testosterone and both reduced metabolites varied significantly with estrous cycle. Based upon post-hoc analysis, testosterone appeared to reach two peaks over the cycle, at diestrus 2 and estrus, with troughs at proestrus and metestrus. The levels of DHT increased significantly with each successive stage from diestrus 1 to proestrus then decreased significantly at each stage until diestrus 1. Similarly, the levels of 3a-diol increased significantly with each stage from diestrus 1 to estrus and decreased dramatically following estrus. Therefore, the levels of reduced androgen metabolites are higher in female brains at proestrus and estrus than at other stages of the estrous cycle. While the levels of androgens in adult female brains overall are quite low, the changes that take place in androgen metabolism over the estrous cycle may have functional implications. 3.3. Comparison of steroid leÕels in fetal Õs. adult brains Comparing the hormone levels in fetal and adult brains, it is clear that progesterone levels are much higher in adult
Fig. 1. Utilization of progesterone and testosterone during late gestation in the rat brain, averaged across sexes. Left panel depicts the ratio of the progesterone metabolites, DHP and 3a ,5a-THP, to the parent hormone, progesterone ŽP.. Right panel depicts the ratio of the androgen metabolites, DHT and 3a-diol, to the parent hormone, testosterone ŽT..
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
21
Fig. 2. Utilization of progesterone and testosterone in adult rat female brains at different stages of the estrous cycle and in adult rat male brains. Left panel depicts the ratio of the progesterone metabolites, DHP and 3a ,5a-THP, to the parent hormone, progesterone ŽP.. Right panel depicts the ratio of the androgen metabolites, DHT and 3a-diol, to the parent hormone, testosterone ŽT.. Stages of the female estrous cycle: D1 s diestrus 1, D2 s diestrus 2, P s proestrus, E s estrus, ME s metestrus. U Indicates that females at proestrus are significantly different from all other groups.
brains Žregardless of sex. than fetal brains, whereas in fetal brains, the levels of reduced progesterone metabolites are comparable to levels in adult brain and considerably higher than that of their parent hormone. Fetal levels of testosterone are comparable to those measured in adult male brain and considerably higher than in adult female brain. On the other hand, the levels of reduced androgen metabolites are higher than those of testosterone in adult male brain but not in fetal brain. These observations suggest that turnover of the parent hormones differs between fetal and adult brain. 3.4. Analysis of the ratio of reduced metabolites to parent hormone in fetal brain To provide an index of the conversion of parent hormone to its reduced metabolites, the ratio of the reduced metabolites to parent hormone was calculated for each sample. As illustrated in Fig. 1A, the ratio of the reduced progesterone metabolites ŽDHP plus 3a ,5a-THP. to progesterone in fetal brain did not vary significantly over gestational age. However, at all ages the ratio was very high Ž20–30., indicating considerable production of reduced progesterone metabolites from the parent hormone in fetal brain. In contrast, the ratio of reduced testosterone metabolites ŽDHT plus 3a-diol. to testosterone ŽFig. 1B. was below 1 at all ages and did not vary significantly across gestational age. Therefore, very little testosterone in fetal brain appears to be converted to reduced metabolites. 3.5. Analysis of the ratio of reduced metabolites to parent hormone in adult brain Analysis of the conversion of parent hormone to metabolites in adult brains revealed a very different profile
from that observed in fetal brain. The ratio of reduced progesterone metabolites to progesterone was around 1 and did not vary significantly with hormonal status ŽFig. 2A.. Thus, in comparison to fetal brains, the conversion of progesterone to reduced metabolites appears to be considerably less in adult brain and does not differ between the two sexes. In contrast, the ratio of reduced testosterone metabolites to testosterone ŽFig. 2B. was higher in all adult hormonal groups than in fetal tissue, even in adult female brain where the overall levels of androgens were much lower than in fetal brain. Furthermore, analysis of this ratio in adult brain indicated that the ratio varied significantly with hormonal group, with the ratio measured in female brains at proestrus considerably higher than in any other groups Žincluding adult males.. Thus, the reduction of androgens may play a different role in adult brain as compared to fetal brain.
4. Discussion Endogenous levels of reduced metabolites of steroid hormones were detected in the brain at all gestational ages evaluated; however, neither the levels of parent hormones nor of reduced metabolites appeared to differ as a function of fetal sex. While it could be argued that the accuracy of determination of fetal sex was suspect due to the reliance primarily on AG distance, the lack of an impact of sex on endogenous levels of reduced steroid metabolites in fetal brain is consistent with the observation that the activity of 5 alpha-reductase, measured in vitro, in the brain does not differ between sexes at any developmental stage analyzed w26x. Furthermore, none of the hormones measured in the present study varied as a function of gestational age, neither when analyzed by sex nor when the data for the
22
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
sexes were pooled. Compared to literature reports on the levels of progesterone and testosterone in fetal circulation during late gestation, it appears unlikely, therefore, that reduced metabolites in fetal brain originate solely from parent hormones taken up from the fetal circulation. Progesterone levels in fetal circulation reportedly decrease consistently throughout the last week of gestation from high levels measured at GD 17 Žwith day 0 of gestation designated as the day of mating., and the circulating levels do not differ between male and female fetuses w42x. In contrast to this profile, reduced progesterone metabolites remained high in fetal brain from GD 17 through 21. Circulating testosterone levels in male fetuses are higher than in female fetuses at GD 17 and increase markedly at GD 18, decreasing thereafter. Circulating levels of testosterone in female fetuses remain fairly constant throughout the last week of gestation w42x. Consistent adult-like Ži.e., adult male-like. levels of testosterone were observed in fetal brain throughout this period. The lack of any impact from sex or gestational age on fetal brain levels of progesterone, testosterone and their reduced metabolites raises the question of whether the source of the parent hormones may have been de novo. The results obtained Ždata pooled across sex. suggest that the reduction of progesterone may be important during fetal brain development: the levels of reduced progesterone metabolites were 10- to 30-fold higher than the levels of the parent hormone. In contrast, the levels of reduced testosterone metabolites were around 10-fold lower than that of the parent hormone. Considering the critical importance of the aromatization of testosterone to estrogens for sexual differentiation of the brain w20,32x, the low levels of 5 alpha-reduced androgen metabolites observed during late gestation even in male brains is not surprising. Additionally, reduced androgens are endogenous inhibitors of aromatase w21x, therefore, low levels of reduced androgen metabolites in fetal brain may be protective. The results of the present study suggest that even in fetal male brain the reduction of androgens is of lesser importance than other routes of androgen metabolism during development. The high endogenous levels of reduced progesterone metabolites in fetal brains relative to the level of parent hormone raises the question of what function the reduction of progesterone might play in fetal brain development. Reduced metabolites of several steroid hormones interact with GABA A and glutamate receptors w24,31,34x. During early brain development, GABA may act as a trophic factor w27x. And while GABA acts as an inhibitory substance in adult brain w37x, it can exert depolarizing actions in developing brain w5x. Reduced progesterone metabolites may participate in these actions of GABA in the developing brain. In support of this mechanism, the addition of 3a ,5a-THP to hippocampal cultures prepared from 18 day rat fetuses induced neurite retraction, and modulation of chloride channels was considered important to this effect w3x. However, certain reduced steroids also have been
shown to act intracellulary at the progesterone receptor w36x. The high rate of conversion of progesterone to its reduced metabolites in fetal brain may provide a mechanism for cross-talk between membrane and genomic hormonal events. Recent evidence suggests that progesterone potentially could be involved in sexual differentiation of the brain: progestin receptor immunoreactivity was observed in specific hypothalamic nuclei in male rats at fetal and neonatal ages, but was absent in female rats brains w41x. While the results of the present study indicated no sex differences in 5 alpha-reduced metabolites of progesterone in the brain during the prenatal period of sexual differentiation, the reduced progestins produced during fetal development could still affect sexual differentiation because of other sex specific factors. The role of progesterone in late fetal brain development in the rat warrants further study. Clearly, the reduction of progesterone and testosterone in vivo differed between fetal and adult brain. Whereas progesterone was readily converted to its reduced metabolites in fetal brain, the reduction of progesterone in adult brain Žmale and female. was considerably less. The absolute levels of progesterone and 3a ,5a-THP in adult female brain appear to follow changes in circulating progesterone levels over the estrous cycle. The levels of both steroids were significantly higher in female brain at proestrus and estrus, and circulating progesterone levels reportedly increase at proestrus and remain elevated into estrus w11,12,38x. The progesterone substrate for 5 alpha-reductase in adult female brain, therefore, may be circulating progesterone, whereas circulating progesterone may not be the substrate for 5 alpha-reductase in fetal brain. In contrast to the reduction of progesterone, testosterone was more actively converted to reduced metabolites in the adult brain than in fetal brain. While the endogenous levels of all of the androgen hormones were markedly higher in adult male brains than in adult female brains, changes in the reduction of testosterone did take place in female brains over the estrous cycle. The marked turnover of testosterone to reduced androgen metabolites observed in the female brains at proestrus is consistent with the observations that 5 alpha-reduced androgens synergize with 5 alpha-reduced progestins to initially facilitate sexual receptivity and then to terminate mating in the female rat w13,16,17x. More relevant, however, to the focus of the present study is the question of what determines the change in the reduction of brain steroid hormones that must take place across postnatal development. Progesterone clearly appears to be the primary substrate for 5 alpha-reductase in the fetal brain, whereas in the adult brain testosterone may have a slight edge: the ratio of metabolites to parent hormone was between 1–2 for progesterone and 3–10 for testosterone in the adult, excepting the female at proestrus, when this ratio was even greater. Progesterone has a higher affinity for 5 alpha-reductase than does testosterone w26x, but there has been no indication that substrate affinity differs between fetal and adult brain. The observation in
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
the present study that progesterone turnover to its reduced metabolites decreases as testosterone turnover increases between fetal ages and adulthood would suggest that substrate affinity alone probably is not responsible for the apparent shift in preferred substrate. The enzyme 5 alphareductase can exist in more than one isoform, and in rat brain, the predominant form appears to be 5 alpha-reductase type 1 w33x. There has been no indication, though, that the preferred isoform is different in the brain during early development than during adulthood. It is critical, therefore, to evaluate the reduction of steroids during early postnatal and adolescent development in order to determine the developmental stageŽs. where the reduction of progesterone becomes diminished and the reduction of androgens increases, as this information may help elucidate reasons for the apparent shift in preferred hormone substrate by brain 5 alpha-reductase from late fetal ages to adulthood. In summary, the measurement of endogenous levels of 5 alpha-reduced steroids and their parent hormones has indicated that the reduction of progesterone by 5 alpha-reductase in fetal brain rather than the reduction of androgens may play a critical role in early brain development. Further work is necessary to localize specific regions of the brain where this process may be most active.
Acknowledgements This work was supported in part by PHS grant no. DA 07080 to C.K. Kellogg and NSF CAREER grant IBN9514463, Whitehall Foundation grant F96-10, and Donaghue Foundation grant 96-001 to C.A. Frye.
References w1x E-E. Baulieu, P. Robel, Neurosteroids: a new brain function, J. Steroid Biochem. Mol. Biol. 37 Ž1990. 395–403. w2x D. Bitran, R.J. Hilvers, C.K. Kellogg, Anxiolytic effects of 3a-hydroxy-5a wb x-pregnan-20-one: endogenous metabolites of progesterone that are active at the GABA A receptor, Brain Res. 561 Ž1991. 157–161. w3x R.D. Brinton, The neurosteroid 3a-hydroxy-5a-pregnane-20-one induces cytoarchitectural regression in cultured fetal hippocampal neurons. J. Neurosci. 14 Ž1994. 2763–2774. w4x M.D. Brot, Y. Akwa, R.H. Purdy, G.F. Koob, G.F. Britton, The anxiolytic-like effects of the neurosteroid allopregnanolone: interactions with GABA A receptors, Eur. J. Pharmacol. 325 Ž1997. 1–7. w5x E. Cherbini, J.L. Gaiarsa, Y. Ben-Ari, GABA: an excitatory transmitter in early postnatal life, Trends Neurosci. 14 Ž1991. 515–519. w6x N.A. Compagnone, A. Bulfone, J.L.R. Rubenstein, S.H. Mellon, Expression of the steroidogenic enzyme P450scc in the central and peripheral nervous systems during rodent embryogenesis, Endocrinology 136 Ž1995. 2689–2696. w7x P. Corbier, B. Kerdelhue, R. Picon, J. Roffi, Changes in testicular weight and serum gonadotropin and testosterone levels before, during, and after birth in the perinatal rat, Endocrinology 103 Ž1978. 985–992.
23
w8x M. Erskine, N. Hippensteil, E. Kornberg, Metabolism of dihydrotestosterone to 3a-androstanediol in brain and plasma: effect on behavioural activity in female rats, J. Endocrinol. 134 Ž1992. 183– 195. w9x D.A. Finn, K.W. Gee, The estrus cycle, sensitivity to convulsants and the anticonvulsant effect of a neuroactive steroid, J. Pharmacol. Exp. Ther. 271 Ž1994. 164–170. w10x C.A. Frye, The neurosteroid 3a ,5a-THP has anti-seizure and possible neuroprotective effects in an animal model of epilepsy, Brain Res. 696 Ž1995. 113–120. w11x C.A. Frye, L.E. Bayon, Seizure activity is increased in endocrine states characterized by decline in endogenous 3a ,5a-THP, Neuroendocrinology 68 Ž1998. 272–280. w12x C.A. Frye, L.E. Bayon, Pregnane and androstane neurosteroids are increased in mated rats and across various hormonal and reproductive milieru, J. Neuroendocrinol. Žin press.. w13x C.A. Frye, K.R. van Keuran, M.S. Erskine, Behavioral effects of 3a-androstandiol: I. Modulation of sexual receptivity and promotion of GABA-stimulated chloride flux, Behav. Brain Res. 79 Ž1996. 109–118. w14x C.A. Frye, T.J. Scalise, L.E. Bayon, Finasteride blocks the reduction in ictal activity produced by exogenous estrous cyclicity, J. Neuroendocrinol. 10 Ž1998. 291–296. w15x C.A. Frye, C.M. McCormick, C. Coopersmith, M.S. Erskine, Effects of paced and non-paced mating stimulation on plasma progesterone, 3a-diol and corticosterone, Psychoneuroendocrinology 21 Ž1996. 431–439. w16x C.A. Frye, J.E. Duncan, M. Basham, M.E. Erskine, Behavioral effects of 3a-androstanediol: II. Hypothalamic and preoptic area actions via GABAergic mechanism, Behav. Brain Res. 79 Ž1996. 119–130. w17x C.A. Frye, K.R. van Keuran, P.N. Rao, M.S. Erskine, Progesterone and 3a-androstanediol conjugated to bovine serum albumin affects estrous behavior when applied to the MBH and POA, Behav. Neurosci. 110 Ž1996. 603–612. w18x K.W. Gee, M.B. Bolger, R.D. Brinton, H. Coirini, B.S. McEwen, Steroid modulation of the chloride ionophore in rat brain: structure–activity requirements, regional dependence and mechanism of action, J. Pharmacol. Exp. Ther. 246 Ž1988. 803–812. w19x Z.Y. Hu, E. Bourreau, I. Jungo-Testas, P. Robel, E-E. Baulieu, Oligodendrocyte mitochondria convert cholesterol to pregnenolone, Proc. Natl. Acad. Sci. U.S.A. 84 Ž1987. 8129–8215. w20x J.B. Hutchison, T. Steimer, Androgen metabolism in the brain: behavioural correlates, Prog. Brain Res. 61 Ž1984. 23–51. w21x J.B. Hutchison, C. Beyer, R.E. Hutchison, A. Wozniak, Sex differences in the regulation of embryonic brain aromatase, J. Steroid Biochem. Mol. Biol. 61 Ž1997. 315–322. w22x I. Jung-Testas, Z.Y. Hu, P. Robel, E-E. Baulieu, Biosynthesis of pregnanolone and progesterone in primary cultures of rat glial cells, Endocrinology 125 Ž1989. 2083–2091. w23x C.K. Kellogg, V.G. Olson, G.L. Pleger, Neurosteroid action at the GABA A receptor in fetal rat forebrain, Dev. Brain Res. 108 Ž1998. 131–137. w24x J.J. Lambert, D. Belelli, C. Hill-Venning, J.A. Peters, Neurosteroids and GABA A receptor function, Trends Pharmacol. Sci. 16 Ž1995. 295–303. w25x M.E. Lauber, W. Lichtensteiger, Ontogeny of 5a-reductase Žtype 1. messenger ribonucleic acid expression in rat brain: early presence in germinal zones, Endocrinology 137 Ž1996. 2718–2730. w26x E.D. Lephart, Brain 5a-reductase: cellular, enzymatic, and molecular perspectives and implications for biological function, Mol. Cell. Neurosci. 4 Ž1993. 473–484. w27x F.M. Leslie, Neurotransmitters as neurotrophic factors, in: S.E. Loughlin, J.H. Fallon ŽEds.., Neurotrophic Factors, Academic Press, San Diego, 1993, pp. 565–598. w28x J.A. Long, H.M. Evans, Oestrus cycle in the rat and its associated phenomena, Memoirs Univ. California 6 Ž1922. 1–146.
24
C.K. Kellogg, C.A. Frye r DeÕelopmental Brain Research 115 (1999) 17–24
w29x M.D. Majewska, Neurosteroids: endogenous bimodal modulators of the GABA A receptor. Mechanism of action and physiological significance, Prog. Neurobiol. 38 Ž1992. 379–395. w30x D.W. Matt, G.J. McDonald, In vitro progesterone and testosterone production by the rat placenta during pregnancy, Endocrinology 115 Ž1984. 741–747. w31x B.S. McEwen, Non-genomic and genomic effects of steroids on neural activity, Trends Pharmacol. Sci. 12 Ž1991. 141–147. w32x R. Naftolin, Brain aromatization of androgens, J. Reprod. Med. 39 Ž1994. 257–261. w33x K. Normington, D.W. Russell, Tissue distribution and kinetic characteristics of rat steroid 5a-reductase isozymes, J. Biol. Chem. 267 19548–19554. w34x S.M. Paul, R.H. Purdy, Neuroactive steroids, FASEB J. 6 Ž1992. 51–63. w35x D. Rodbard, D. M. Hutt, Statistical analysis of radioimmunoassay and immunoradiometric assays: a generalized, weighted iterative, least squares method for logistic curve fitting, in: International Atomic Energy Agency ŽEd.., Symposium on Radioimmunoassay and Related Procedures in Medicine, Uniput, New York, 1974. w36x R. Rupprecht, C.A.E. Hauser, T. Trapp, F. Holsboer, Neurosteroids: molecular mechanisms of action and psychopharmacological significance, J. Steroid Biochem. Mol. Biol. 56 Ž1996. 163–168.
w37x W. Singer, Neurophysiology: the changing face of inhibition, Curr. Biol. 6 Ž1996. 395–397. w38x S.S. Smith, D.J. Woodward, J.K. Chapin, Sex steroids modulate motor-correlated increases in cerebellar discharge, Brain Res. 476 Ž1989. 307–316. w39x S.S. Smith, Q.H. Gong, X. Li, M. Moran, D. Bitran, C.A. Frye, F.C. Hsu, Withdrawal from 3a-OH-5a-pregnan-20-one using a pseudopregnancy model alters the kinetics of hippocampal GABA A -gated current and increases the GABA A receptor a4 subunit in association with increased anxiety, J. Neurosci. 18 Ž1998. 5275–5284. w40x K. Ukena, M. Usui, C. Kohchi, K. Tsutsui, Cytochrome P450 side-chain cleavage enzyme in the cerebellar purkinje neuron and its neonatal change in rats, Endocrinology 139 Ž1998. 137–147. w41x C.K. Wagner, A.T. Nakayama, G.J. De Vries, Potential role of maternal progesterone in the sexual differentiation of the brain, Endocrinology 139 Ž1998. 3658–3661. w42x J. Weisz, I.L. Ward, Plasma testosterone and progesterone titers of pregnant rats, their male and female fetuses, and neonatal offspring, Endocrinology 106 Ž1980. 306–316. w43x R. Yu, M.K. Ticku, Chronic neurosteroid treatment produces functional heterologous uncoupling at the g-aminobutyric acid type Arbenzodiazepine receptor complex in mammalian cortical neurons, Mol. Pharmacol. 47 Ž1995. 603–610.