- Email: [email protected]
Perinatal development of estrogen receptors in mouse brain assessed by radioautography, nuclear isolation and receptor assay
Developmental Brain Research, 11 (1983) 7-18 Elsevier
7
Perinatal Development of Estrogen Receptors in Mouse Brain Assessed by Radioautography, Nuclear Isolation and Receptor Assay J. L. GERLACH, B. S. McEWEN, C. D. TORAN-ALLERAND1and W. J. FRIEDMAN
The Rockefeller University, 1230 York Avenue, New York, NY 10021 and 1Centerfor Reproductive Sciences (I1SHR) and Department of Neurology, Columbia University, College of Physicians and Surgeons, New York, NY (U.S.A.) (Accepted June 14th, 1983)
Key words: estrogen receptors - - cell nuclei - - radioautography - - hypothalamus - - preoptic area - - cingulate cortex - neural development - - mouse - - sexual differentiation
The development of estrogen receptors was investigated in vivo in the brains of fetal and neonatal mice 2 h after administering [3H]moxestrol to the pregnant mothers or neonates. Moxestrol bypasses the alpha-fetoprotein 'protective barrier' and gains access to estrogen receptors. Analysis of [3H]moxestrol uptake by radioautography and by cell nuclear isolation and counting of radioactivity revealed a marked increase in the number of estrogen receptors and estrophilic cells in the brain during late fetal and early postnatal development. Assays of cytosol estrogen receptors were conducted in parallel and revealed a comparable pattern of development. The increase in estrogen receptors and labeling was especially great from embryonic day (E)15 to El8. Cytosol assays revealed a low level of receptors in the whole brain on E13. Radioautography revealed that clearly labeled cells in the hypothalamus and preoptic area were virtually absent on E13 but were evident on El5, with marked increases occurring between El5 and EIS, both in number of labeled cells and in intensity of labeling per cell. Within the cerebral cortex the dorsal cingulate cortex was the most extensively labeled area; however, clearcut labeling was not evident on El3 or El5. Thus, the development of cortical estrogen receptors occurs somewhat later than that in the hypothalamus and preoptic area. The perinatal increase in estrogen receptors usually begins several days after the birthdates of neurons in these estrophilic regions of the brain, and corresponds to the early responsiveness of these neurons to the organizational and activational influences of estrogen. INTRODUCTION
tured explants of the neonatal hypothalamus/preoptic area estrogen enhances neuritic development29,33
Receptors for gonadal steroid hormones are present in the rodent brain and mediate the activity of the
in regions possessing estrogen receptors. Both radioaut0graphy and binding assays have found estrogen
steroid hormones influencing certain reproductive behaviors and n e u r o e n d o c r i n e functions. In adults estrogens and androgens activate sex-specific neuronal functions that differentiated early in development ~. The perinatal organization of these functions is also controlled by the gonadal steroid hormones and involves two processes: defeminization, thought to be mediated by the enzymatic conversion (aromatization) of androgen to estrogen, and its subsequent binding to estrogen receptors; and masculinization, which may be mediated by the binding of hormones to both androgen and estrogen receptors 3.9.
receptors in various areas of the developing brain, including the cerebral cortex and limbic structures such as the hypothalamus, preoptic area and amygdala 4. 17,24,25,28. Studies on the postnatal ontogeny of these receptors 2.11.14.2°.27,38,39 include those demonstrating
In addition to influencing sexual differentiation of the brain, estrogens may participate in the growth and maintenance of developing neurons 22,29. In cul0165-3806/83/$03.00 © 1983 Elsevier Science Publishers B.V.
that estrogen receptors in the male hypothalamus and limbic brain are occupied by estrogen derived from testosteronel2. 37. A m o n g studies on the prenatal development of estrogen receptors, few have examined these receptors from the time of their initial appearance13.35.36. In this study we examined the ontogeny of estrogen receptors in the mouse brain from the time of their prenatal appearance, by radioautography and cell nuclear isolation following in vivo labeling, and by cytosol receptor assays following in vitro labeling
with [3H]moxestrol, a synthetic estrogen which does not bind to alpha-fetoprotein (AFP). A companion paper 6 presents a complementary study using in vitro assays of cytosols from the hypothalamus and cerebral cortex of the pre- and postnatal mouse. MATERIALS AND METHODS Animals The ontogeny of neuronal estrogen receptors and the ability of neurons in the brain to bind estrogens was studied by radioautography, nuclear isolation techniques, and receptor assay in mice of the R i l l strain from embryonic day ( E ) l l to postnatal day (P)9. The fetal and postnatal mice utilized in this study were obtained by timed mating from the breeding colony of one of the authors (C.D.T.-A.). The mice were maintained in their own room kept at a constant temperature of 70-72 °F on a 14:10 light/ dark cycle; they were fed mouse breeder chow and water ad libitum, as well as oats and whole wheat bread twice weekly. All mice were handled daily. The males were placed with females (ratio of 2:3) for a single 45-min period daily (from 07.30 to 08.15 h) and checked immediately thereafter for the presence of vaginal plugs, evidence of sperm transfer. The day a vaginal plug was found (the first 24 h) was considered day 0 (E0) of gestation. Offspring were virtually always born in the early morning of E l 9 (P1). In vivo studies Binding of [3H]moxestrol and/or its metabolites to receptors in cell nuclei of brains of fetal mice, ages E l l (number of experiments, n = 1), El3 (n = 2), E14 (n = 2), El5 (n = 3), El6 (n = 2), El7 (n = 3) and E18 (n = 5), was studied by the cell nuclear isolation technique 16. Pregnant mice were injected subcutaneously with = 400 pmoles of [3H]moxestrol (= 14 nmol/kg) (New England Nuclear; spec. act. 88 Ci/mmol) and decapitated 2 h later, a survival time which maximizes the ratio of bound to unbound radioactive estrogen in brain cells 41. Levels of radioactivity in 25 Izl aliquots of maternal blood serum were determined by scintillation counting. The fetuses were removed, weighed, and placed on ice before the heads were homogenized in a solution of sucrose and Triton X-100. After the homogenates were filtered through nylon mesh to remove connective tissue, the
cell nuclei were isolated from the filtrates, the radioactive moxestrol and/or its metabolites were extracted from the nuclear pellets with ethanol, and the extracts were counted in Liquiscint (National Diagnostics, Somerville, N J) to determine levels of radioactivity. These levels were expressed as fmol/mg DNA, based on determinations of nuclear DNA in the pellets by the method of Burton 5. In vitro studies The ontogeny of estrogen receptors in the brains of fetal mice, ages El3 (n = 6), El5 (n = 6) and El8 (n = 6), was examined by exchange assays for cytosol receptors 6. Cytosols were prepared as outlined in the companion paper 6, passed through LH-20 columns to remove unlabeled endogenous estrogens which may have bound to receptors when the brains were homogenized, thereby avoiding subsequent competition with [3H]moxestrol, then incubated with 10-7 M [3H]moxestrol for 4 h, half of each sample in the presence of 10-6 M diethylstilbestrol to determine nonspecific binding. Nuclear D N A was determined and specific binding of radioactivity from [3H]moxestrol to estrogen receptors in the cytosol was expressed as fmol/mg DNA. Radioautographic studies The uptake and retention of radioactivity from [3H]moxestrol by estrogen receptor-containing cells in the brains of fetal mice, ages El3, (n = 4), El5 (n = 4) and El8 (n = 4), and postnatal mice, ages P1 (n = 8) and P9 (n = 6) were studied using thawmount radioautography which is designed to prevent or minimize translocation of soluble substances 7. Pregnant mice received subcutaneous injections in the back of the neck with ~ 400 pmol of [3H]moxestrol (= 14 nmol/kg) and were decapitated 2 h later. Levels of radioactivity in 25 ~1 aliquots of maternal blood serum were established by scintillation counting. The fetuses were immediately removed and placed on ice until the head of each was quickly frozen onto a brass cryostat chuck with powdered dry-ice, then stored in liquid nitrogen. The heads of the P1 mice and the brains of the P9 mice were similarly frozen. Unfixed, unembedded, coronal frozen sections, 2-4/~m thick and 30-50/~m apart, were cut at --20 °C in a cryostat, thaw mounted onto glass slides precoated with Kodak NTB-3 radioauto-
to the PO'A and hypothalamus. A cell with a density of silver grains at least 5 times that of an equal volume of adjacent extracellular tissue was considered labeledl9. Control radioautograms provided no evidence for either negative chemography (false negatives) or positive chemography (false positives).
graphic emulsion, exposed at 4 °C with Drierite from 675 to 738 days, developed in Dektol and stained with cresyl violet acetate. For each age the total number of labeled cells in the cingulate/frontal cortex (CTX), preoptic area (POA), ventromedial nucleus (VM) and arcuate nucleus ( A R C ) was counted in radioautograms spanning the anterior border of the P O A and the posterior border of the A R C . The data for each region are presented as an average per coronal section, based on the total number of labeled cells in all sections which included that region. The anterior cingulate cortex was selected, because earlier observations (C.D. Toran-Allerand, unpublished observations) suggested that in the cerebral cortex of the neonatal rat most of the neurons containing estrogen receptors appear to be localized in the deepest layers (V and VI) of the anterior cingulate cortex ]() with some extension laterally into the frontal cortex as also shown subsequently by Sheridan 24. These regions lie dorsal
RESULTS The estrogen receptors in the brains of fetal and neonatal rats and mice are mostly unoccupied by maternal estrogens, because alpha-fetoprotein (AFP), a protein in the serum produced by the fetal yolk sac and liver and the neonatal liver, is believed to sequester extracellulary the high levels of maternal estradiol that cross the placenta LS. This scavenging action of A F P presumably protects estrogen sensitive tissues from the masculinization and defeminization that such high levels of estradiol could otherwise produce 2]. Since [3H]moxestroi, a synthet-
24
m A.
•
B.
22
" 6 00~
:t
2O
X I i
40C o
2oc A
,~
16
z Q
A
u
z E)
Z
14
14
E
E -6 //
IO E
"-: 0.8 •
•/ '~
•
8
/
0.6 04 0.2
U -I
z
C.
1.2 1.0
0
6
•
/// /
4
2
; i II
I
,,,
1
,'6
,;
ID 0 E I-1
~
if°i
,'8 ii G e s t a t i o n a l Age (Days)
I
z 12
13
. I
•
.
•
i
15
16
17
I
Fig. 1. Growth in the population of estrogen receptors in the brains of mice during the gestational days 11-18, revealed (A) through increased binding of the synthetic estrogen [3H]moxestrol and/or metabolites (fmol/mg DNA) to estrogen receptors in cytosol (. . . . . ) and to receptors transiocated from cytosol to nuclei of cells from whole brains ( ). Moxestrol was used because, unlike estradiol, it does not bind to alpha-fetoprotein, a protein in the serum produced by the liver of perinatal mice. The population of receptors in nuclei and cytosol began to increase markedly around fetal day 15. Binding to estrogen receptors, corrected for nuclear DNA (A), increased with age, as total radioactivity in serum remained level (D), as both cell nuclear DNA content (B) and fetal body weight (C) rose.
Ill
ic estrogen does not bind to AFP, it was used to label estrogen receptors in the brains of fetal mice, in order to chart the regions in which they appear, as well as the course of their development. Pregnant mice were injected with ~ 14 nmol/kg of [3H]moxestrol, a dose estimated to saturate the population of estrogen receptors 14, and the fetuses were removed after 2 h of exposure to the hormone. Since the average amounts of radioactivity recovered in the serum of pregnant mice appeared to be level across gestational age (Fig. 10), approximately the same dose of [3H]moxestrol probably reached each fetus. Brains of fetal mice contained estrogen receptors which translocated from the cytoplasm to nuclei and were capable of taking up and retaining the non-steroidal estrogen [3H]moxestrol and/or its metabolites (Fig. 1A, solid line). The level of this uptake was extremely small on E l l (1.4 fmol/mg DNA), then increased markedly between E14 and El8 (from 1.8 to 18.0 fmol/mg DNA). Although the fetal body weight (Fig. 1C) and the content of nuclear DNA in the brain (Fig. 1B) both increased with age, the increase in nuclear binding was corrected for DNA content (Fig. 1A, solid line) and therefore reflected growth in the population of nuclear estrogen receptors per cell, rather than increases based simply on more cells. By measuring cytosol receptors, we ascertained further whether the increase in nuclear incorporation of radioactivity from [3H]moxestrol resulted from a parallel change in the cellular density of receptors. The population of receptors in the cytosol of brains of fetal mice was small on El3, increased slightly between El3 and El5, from 2.0 to 3.3 fm/mg DNA, then rose sharply by El8 to 10.8 fm/mg DNA (Fig. 1A. dashed line). These concentrations, which are adjusted for DNA content, represent growth in the population of cytosol receptors per cell, although the total number of cells in the brain did increase during the same interval, as shown by increased levels of nuclear DNA (Fig. 1B). The result of our radioautographic experiments corresponds to and complements the results of our cell fractionation and binding studies. In the CTX, POA, VM and ARC of fetal mice, the population of radioactively labeled cells, as well as the intensity of cellular labeling, increased dramatically with age. The density of radioautographic silver grains is an in-
dex of the concentration of receptors. These increases, therefore, reflect on the one hand, an increase in the number of cells with estrogen receptors and, on the other, a proliferation of receptors in individual cells. The radioautographs illustrate both types of changes (Figs. 2-5). In the CTX, no labeled cells were evident on El5 (Fig. 2a): some cells were labeled by El8 (Fig. 2b-d); and by P9 this number had increased, as had the density of labeling (Fig. 2e and f). In radioautographs of the POA and hypothalamus, virtually no cells were labeled on El3 (hypothalamus, Fig. 3a; POA, Fig. 3b); many were labeled by El5 (POA, Fig. 3c and d; VM, Fig. 4a-c; ARC, Fig. 5a-c); and by El8 both the population of labeled cells and intensity of labeling had further increased (POA, Fig. 3e and f: VM, Fig. 4d-f: ARC, Fig. 5d-f). The results from counting labeled cells confirm the changes in cell population already illustrated (Figs. 2-5), and suggest several additional developmental differences. However, the quantitative value of these data is limited, since the various regions of the brain were not sampled systematically among individuals. In the cingulate cortex, binding to estrogen receptors was evident in radioautograms by El8, but not El5. The average number of labeled cells rose between El5 and El8, from zero to 58, and again between P1 and P9, from 59 to 115 (Fig. 2). in the deeper layers of this cortical region the postnatal increase from P1 to P9 was 300%, and in its more superficial layers the increase was 150%. In the remaining (mostly frontal) cortex the average number of labeled cells rose between El5 and El8, from zero to 79, then remained at this level through P9. The population of labeled cells in the frontal cortex appeared earlier and reached a plateau earlier (by El8) than that of the cingulate which did not increase significantly until after P1, growing sharply by P9 to exceed that in the frontal cortex. The different patterns by which the cingulate and frontal cortex each acquired estrophilic cells may reflect the migratory patterns of these cells as they form the various layers of these two regions. In the POA or the hypothalamic ARC and VM almost no cells were labeled on El3; however, by E15 the average number had risen to 81 in the POA, to 23 in the ARC and to 35 in the VM. Each of these areas showed labeling at least 2-3 days sooner than the
11
Fig. 2-5. The ontogeny of estrogen receptors in selected regions of the brain in prenatal and neonatal mice of the R i l l strain is illustrated by radioautograms depicting increases in both the amount of binding per cell and the number of cells binding [3H]moxestrol and/or its metabolites over a 2-week period of development (E15-P9). Radioautograms were prepared by thawing unfixed frozen sections, 2--4/~m thick, onto glass slides precoated with Kodak NTB-3 radioautographic emulsion. They were exposed at 4 °C from 675 to 738 days, developed in Dektol and stained lightly with cresyl violet acetate 7. Magnification: 495 diameters. Fig. 2. Cingulate/frontal cortex, a: age, embryonic day El5. b-d: El8, e and f: postnatal day 9.
12 CTX. By E18 the population of labeled cells had grown by 3 to 4-fold to 290 in the POA, by 3-fold to 69 in the ARC and by 2-fold to 70 in the VM. By P9 the size of the labeled population had not changed in the ARC (n = 67) and had decreased slightly in the VM (n = 50). We do not have data for the POA on P1 or P9. Since the sparsely labeled, as well as the densely labeled, cells were counted, the quantitative data underestimate the differences among regions and ages realized by visual inspection of the radioautograms, whereby the densely labeled cells were especially noticeable. DISCUSSION The studies presented here and in the accompanying paper 6 on the ontogeny of estrogen receptors in the brains of mice reveal that a variety of brain regions containing estrogen receptors in adulthood initiate the synthesis of receptors prenatally. Our experiments employing either in vitro exchange assays of cytosols (this paper ref. 6) or the isolation of cell nuclei after in vivo labeling showed that the populations of receptors in the whole brains of fetuses were very low and at the limit of detection on E l l and E13, increased sharply between E l 4 and E18 and reached or exceeded adult levels within the following two weeks. We have shown in this and the accompanying paper 6 that this change in the population of receptors determined by in vivo and in vitro techniques reflects increases between E13 and E l 8 within specific brain regions. Radioautograms showed that these regions included the CTX, POA and the hypothalamic VM and ARC. Within the POA,VM and ARC the extent of radioautographic labeling, established both by counting the number of labeled cells and observing the density of labeling within cells, increased markedly between El5 and El8, with the POA showing the most labeling. After this time and until P9, binding of radioactivity derived from [3H]moxestrol remained high in both the ARC and VM. Receptors in the CTX appeared approximately 2-3 days later than those in the POA and hypothalamus, by El8 but not El5. They appeared at first in more cells of the frontal cortex than the cingulate, then increased by P9 in the cingulate cortex to reach the highest level of binding anywhere in the cortex.
Within each region of the brain where estrogen receptors appeared, both the number of cells with receptors, as well as the apparent density of receptors within individual cells increased with age. Although counting all labeled cells (whatever the density of their labeling) revealed both types of increases, visual inspection of the radioautograms disclosed these changes most dramatically, the increasing numbers of densely labeled cells standing out conspicuously. The in vivo and in vitro biochemical experiments, which recorded increases in binding relative to DNA content, also revealed a proliferation in the cellular population of receptors. Thus, the ontogenic profiles were concordant among the radioautographic and biochemical experiments. Friedman et al. 6 confirmed this developmental pattern and extended observations to an older age in the accompanying report on the regional distribution of estrogen receptors in cytosols prepared from the brains of perinatal mice. The ability to detect estrogen receptors earlier with biochemical methods than with radioautographic methods, El3 by binding assay and E15 by autoradiography, is in agreement with observations from studies of corticoid receptors in gliat cells18, wherein the threshold for detection of labeling by radioautography exceeded the threshold for detection by nuclear isolation or cytosol assays. On El3 only a very few cells in radioautograms were unequivocally labeled; instead, a low density of silver grains appeared evenly distributed within and among cell bodies in all regions examined. Since these radioautograms were exposed for the relatively long period of 675-738 days, grain densities probably approached a maximum and would not have increased with more exposure. The detection of estrogen receptors by El3 using binding assays and the subsequent dramatic increase in their concentration between El5 and E18 corresponds to certain observations made in other studies. Using DNA-cellulose cytosol assays Vito and Fox 36 found that the population of estrogen receptors in the combined hypothalamus and POA of mice was relatively high by El5; but, contrary to our observation of a marked increase by ES, they saw no increase until after P10. Stumpf et al. 28 reported radioautographic labeling in the hypothalamus of mice after injecting [3H]diethylstiibestrol on E16; however, they did not specify their criterion for assigning embryonic age.
© >
r~
~D
© >
© >
0
~r
0
#
r~
o
c~j
Q
t~
#
lO
In addition, the appearance and developmental increase of estrogen receptors in the regions studied closely follow the birthdates of the majority of neurons in these regions of the mouse as reported by others 1-26 including the same strain used in our studies (C.D. Toran-Allerand, unpublished observations). These birthdates range from El() to El7 and peak between E l l and El5. The temporal correspondence between this appearance and increase of limbic and cortical estrogen receptors demonstrated in this and the other studies cited, and the birthdates of neurons in the same regions of the brain supports the suggestion of Vito and Fox 36 that neuroblasts may become responsive to estrogen when they cease dividing and begin to differentiate, and soon thereafter the critical or sensitive period for sexual differentiation of the brain may start. In male rodents estradiol is derived from aromatizable androgens in neurons of the hypothalamus and POA 12, and serves to masculinize and defeminize these regions and their associated reproductive physiology and behavior3.9. Thus it is particularly interesting that the appearance and further development of estrogen receptors in the hypothalamus and POA observed here correspond temporally to the appearance of aromatase activity in the diencephalon of male and female rats on El5, peaking in the hypothalamus and POA between El7 and E19s. Birth, as well as many other developmental stages in rats occur approximately 2 days later than in mice. This relationship between neuronal birthdates and aromatase activity does not apply, however, to the CTX, because in rodents it not only lacks aromatase, but its estrogen receptors do not appear to be occupied by endogenous estrogens 12. These observations on the CTX of perinatal mice, together with the fact that AFP may sequester maternal estradiol extracellularly and that neurons in only the hypothalamus and POA of rodents convert testosterone to estradiol, raise the question of a role for cortical estrogen receptors. The presence of intraneuronal AFP only during the ontogeny of the brain, including the cerebral cortex 31, however, suggests a mechanism by which low levels of estrogens might have access to the cortical receptors despite the presence of extracellular AFP. Since intraneuronal AFP is derived from extracellular sources and not synthesized locally23, and is taken up specifically by neurons, at least in culture
(C. D. Toran-Allerand, unpublished observations; ref. 34), neuronal uptake of AFP (as the AFP/estradiol complex) would bring estradiol into the cell. Such a mechanism might provide target neurons of both sexes with the intracellular source of estrogen of non-androgenic origin which was postulated earlier 29-31. That the cortical estrogen receptors may be functional, like those of the hypothalamus and POA, and may similarly be implicated in the ontogeny of cortical regions is suggested by preliminary studies showing that cortical cultures, which contain estrogen receptors as in vivo (C. D. Toran-Allerand, J. L. Gerlach, and B. S. McEwen, unpublished observation), respond directly to exogenous estradiol by a marked enhancement of neuritic outgrowth (C. D Toran-Allerand, unpublished observation) and by the enhanced incorporation of [3H]leucine into the TCA (trichloracetic acid)-precipitable proteins of the cultures (N. J. MacLusky and C. D. Toran-Allerand, unpublished observation). In summary, this study demonstrated that the brains of perinatal mice produce estrogen receptors as early as El3, and that both the population of cells with receptors, as well as the density of receptors within cells, increases dramatically over the ensuing perinatal period in the hypothalamus, POA and CTX. The coincidence of this ontogeny with the appearance of aromatase activity and the presence of circulating and intraneuronal AFP, closely following the birthdates of neurons in these regions, suggests that maternal estradiol as well as fetal estradiol derived from testosterone influence development and sexual differentiation of the brain. ACKNOWLEDGEMENTS Supported by Grants NS07080 (to B. S. M.) and by HD-08364, NSF BNS77-0859 (to C. D. T.-A.), as well as by an institutional grant (RF81062) from the Rockefeller Foundation for research in reproductive biology. C. D. T.-A. also received support from the March of Dimes Birth Defects Foundation, the Whitehall Foundation and NIMH Research Scientist Development Award MH00192. The authors wish to acknowledge the devoted assistance of Mrs. Lily Skaredoff in the breeding of mice, as well as the excellent editorial work of Mrs. Oksana Wengerchuk and Ms. Pattey Fong.
17 REFERENCES 1 Angevine, J. B. and Sidman, R. L., Autoradiographic study of cell migration during histogenesis of cerebral cortex in the mouse, Nature (Lond.), 192 (1961) 766--768. 2 Attardi, B. and Ohno, S., Androgen and estrogen receptors in the developing mouse brain, Endocrinology, 99 (1976) 127%1290. 3 Baum, M. J., Differentiation of coital behavior in mammals: a comparative analysis, Neurosci. Biobehav. Rev., 3 (1979) 265-284. 4 Barley, J., Ginsburg, M., Greenstein, B. D., MacLusky, N. J. and Thomas, P. J., A receptor mediating sexual differentiation, Nature (Lond.), 252 (1974) 25%260. 5 Burton, K., A study of the conditions and mechanisms of the diphenylamine reaction for the colorimetric estimation of DNA, Biochem. J., 62 (1956) 315-323. 6 Friedman, W. C., McEwen, B. S., Toran-Allerand, C. D. and Gerlach, J. L., Perinatal development of hypothalamic and cortical estrogen receptors in mouse brain: methodological aspects, Develop. Brain Res., 11 (1983) 1%27. 7 Gerlach, J. L. and McEwen, B. S., Rat brain binds adrenal steroid hormone: Radioautography of hippocampus with corticosterone, Science, 175 (1972) 1133-1136. 8 George, F. W. and Ojeda, S. R., Changes in aromatase activity in the rat brain during embryonic, neonatal, and infantile development, Endocrinology, 111 (1982) 522-529. 9 Goy, R. W. and McEwen, B. S. (Eds.), Sexual Differentiation of the Brain, based on a Work Session of the Neurosciences Research Program, MIT Press, Cambridge, 1980, 211 pp. 10 Hamilton, L. W., Basic Limbic System Anatomy of the Rat, Plenum Press, New York and London, 1976, pp. 95-100, 11 Kato, J., Atsumi, Y. and Inaba, M., Estradiol receptors in female rat hypothalamus in the developmental stages and during pubescence, Endocrinology, 94 (1974) 30%317. 12 Lieberburg, I., Krey, L. C. and McEwen, B. S., Sex differences in serum testosterone and in exchangeable brain cell nuclear estradiol during the neonatal period in rats, Brain Res., 178 (1979) 207-212. 13 MacLusky, N. J., Lieberburg, I. and McEwen, B. S., The development of estrogen receptor systems in the rat brain: perinatal development, Brain Res., 178 (1979) 12%142. 14 MacLusky, N. J., Chaptal, C. and McEwen, B. S., The development of estrogen systems in the rat brain and pituitary: postnatal development, Brain Res., 178 (1979) 143-160, 15 McEwen, B. S., Gonadal steroid influences on brain development and sexual differentiation. In R. O. Greep (Ed.), Reproductive Physiology, Vol. IV, University Park Press, Baltimore, 1983, pp. 9%145. 16 McEwen, B. S. and Zigmond, R, E., Isolation of brain cell nuclei. In N. Marks and R. Rodnight (Eds.), Methods in Neurochemistry, Plenum Press, New York, 1972, pp. 140-161. 17 McEwen, B. S., Plapinger, L., Chaptal, C., Gerlach, J. and Wallach, G., Role of fetoneonatal estrogen binding proteins in the association of estrogen with neonatal brain cell nuclear receptors, Brain Res., 96 (1975) 400-406. 18 Meyer, J. S., Leveille, P. J., de Vellis, J., Gerlach, J. L. and McEwen, B. S., Evidence for glucocorticoid target cells in rat optic nerve. Hormone binding and glycerolphosphate dehydrogenase induction, J. Neurochem., 39 (1982)
436-442. 19 Morrell, J. I. and Pfaff, D. W., Autoradiographic technique for steroid hormone localization: application to the vertebrate brain. In N. T. Adler (Ed.), Neuroendocrinology of Reproduction: Physiology and Behavior, Plenum Press, New York, 1981, pp. 51%531. 20 Plapinger, L. and McEwen, B. S., Ontogeny of estradiolbinding sites in rat brain. I. Appearance of presumptive adult receptors in cytosol and nuclei, Endocrinology, 93 (1973) 1119-1128. 21 Plapinger, L. and McEwen, B. S., Gonadal steroid-brain interactions in sexual differentiation. In J. Hutchinson (Ed.), Biological Determinants of Sexual Behavior, J. Wiley, New York, pp. 193-218. 22 Puymirat, J., Loudes, C., Faivre-Bauman, A., Bourre, J. M. and Tixier-Vidal, A., Expression of neuronal functions by mouse fetal hypothalamic cells cultured in hormonally defined medium. In G. Sato, A. B. Pardee, D. A. Sirbasku (Eds.), Growth of Cells in Hormonally Defined Media, Cold Spring Harbor Laboratories, Cold Spring Harbor, NY, 1982, pp. 1033-1052. 23 Schachter, B. S. and Toran-Allerand, C. D., Intraneuronal alpha-fetoprotein and albumin are not synthesized locally in developing brain, Develop. brain Res., 5 (1982) 93-98. 24 Sheridan, P. J., Estrogen binding in the neonatal cortex, Brain Res., 178 (1979) 201-206. 25 Sheridan, P. J., Sar, M. and Stumpf, W. E., Autoradiographic localization of 3H-estradiol or its metabolites in the central nervous system of the developing rat, Endocrinology, 94 (1974) 1386--1390. 26 Shimada, M. and Nakamura, T., Time of neuron origin in mouse hypothalamic nuclei, Exp. Neurol., 41 (1973) 163173. 27 Spona, J., Bieglmayer, C., Adamiker, C. and Jettmar, W., Ontogeny of 17fl-estradiol-binding protein in the female rat hypothalamus and anterior pituitary, FEBS Lett., 76 (1977) 306--310. 28 Stumpf, W. E., Narbaitz, R. and Sar, M., Estrogen receptors in the fetal mouse, J. Steroid Biochem., 12 (1980) 55-63. 29 Toran-Allerand, C. D., Sex steroids and the development of the newborn mouse hypothalamus and preoptic area in vitro: implications for sexual differentiation, Brain Res., 106 (1976) 407--412. 30 Toran-Allerand, C. D., Sex steroids and the development of the newborn mouse hypothalamus and preoptic area in vitro: II. Morphological correlates and hormonal specificity, Brain Res., 189 (1980) 413--427. 31 Toran-Allerand, C. D., Coexistence of alpha-fetoprotein albumin and transferrin immunoreactivity in neurons of the developing mouse brain, Nature (Lond.), 286 (1980) 733--735. 32 Toran-Allerand, C. D., Gerlach, J. L. and McEwen, B. S., Autoradiographic localization of [3H]estradiol releated to steroid responsiveness in cultures of the newborn mouse hypothalamus and preoptic area, Brain Res., 184 (1980) 517-522. 33 Toran-Allerand, C. D., Hashimoto, K., Greenough, W. T. and Saltarelli, M., Sex steroids and the development of the newborn mouse hypothalamus and preoptic area in vitro: III. Effects of estrogen on dendritic differentiation, Brain Res., in press. 34 Uriel, J., Faivre-Bauman, A., Trojan, J. and Foiret, D., Immunocytochemical demonstration of alpha-fetoprotein
18
35 36
37
38
uptake by primary cultures of fetal hemisphere cells from mouse brain, Neurosci. Lett., 27 (198l) 171-175. Vito, C. C. and Fox, T. O., Embryonic rodent brain contains estrogen receptors, Science, 204 (1979) 517-519. Vito, C. C. and Fox, T. O., Androgen and estrogen receptors in embryonic and neonatal rat brain, Develop. Brain Res., 2 (1982) 9%110. Westley, B. R. and Salaman, D. F., Role of oestrogen receptor in androgen-induced sexual differentiation of the brain, Nature (Lond.). 262 (1976) 407-408. Westley, B. R., Thomas, P. J., Salaman, D. F., Knight, A. and Barley, J., Properties and partial purification of an oes-
trogen receptor from neonatal rat brain, Brain Res., 113 (1976) 441-447. 39 White, J. O., Hall, C. and Lim. L., Developmental changes in the content of oestrogen receptors in the hypothalamus of the female rat, Biochem. J., 184 (1979) 465-468. 40 Young, W. C., Goy, R. W. and Phoenix, C. H.. Hormones and sexual behavior. Broad relationships exist between the gonadal hormones and behavior, Science, 143 (1964) 212-218. 41 Zigmond, R. E. and McEwen, B. S., Selective retention of oestradiol by cell nuclei in specific brain regions of the ovariectomized rat, J. Neurochem., 17 (1970) 889-899.
Note added in proof Recently, an autoradiographic study with [3H]diethylstilbestrol has provided independent confirmation of the prenatal detectability of estrogen receptors in mouse hypothalamus and preoptic area around fetal day 15 (Keefer, D. A. and Holderegger, C., Anat. Rec., 205 (1983) 96A).
7
Perinatal Development of Estrogen Receptors in Mouse Brain Assessed by Radioautography, Nuclear Isolation and Receptor Assay J. L. GERLACH, B. S. McEWEN, C. D. TORAN-ALLERAND1and W. J. FRIEDMAN
The Rockefeller University, 1230 York Avenue, New York, NY 10021 and 1Centerfor Reproductive Sciences (I1SHR) and Department of Neurology, Columbia University, College of Physicians and Surgeons, New York, NY (U.S.A.) (Accepted June 14th, 1983)
Key words: estrogen receptors - - cell nuclei - - radioautography - - hypothalamus - - preoptic area - - cingulate cortex - neural development - - mouse - - sexual differentiation
The development of estrogen receptors was investigated in vivo in the brains of fetal and neonatal mice 2 h after administering [3H]moxestrol to the pregnant mothers or neonates. Moxestrol bypasses the alpha-fetoprotein 'protective barrier' and gains access to estrogen receptors. Analysis of [3H]moxestrol uptake by radioautography and by cell nuclear isolation and counting of radioactivity revealed a marked increase in the number of estrogen receptors and estrophilic cells in the brain during late fetal and early postnatal development. Assays of cytosol estrogen receptors were conducted in parallel and revealed a comparable pattern of development. The increase in estrogen receptors and labeling was especially great from embryonic day (E)15 to El8. Cytosol assays revealed a low level of receptors in the whole brain on E13. Radioautography revealed that clearly labeled cells in the hypothalamus and preoptic area were virtually absent on E13 but were evident on El5, with marked increases occurring between El5 and EIS, both in number of labeled cells and in intensity of labeling per cell. Within the cerebral cortex the dorsal cingulate cortex was the most extensively labeled area; however, clearcut labeling was not evident on El3 or El5. Thus, the development of cortical estrogen receptors occurs somewhat later than that in the hypothalamus and preoptic area. The perinatal increase in estrogen receptors usually begins several days after the birthdates of neurons in these estrophilic regions of the brain, and corresponds to the early responsiveness of these neurons to the organizational and activational influences of estrogen. INTRODUCTION
tured explants of the neonatal hypothalamus/preoptic area estrogen enhances neuritic development29,33
Receptors for gonadal steroid hormones are present in the rodent brain and mediate the activity of the
in regions possessing estrogen receptors. Both radioaut0graphy and binding assays have found estrogen
steroid hormones influencing certain reproductive behaviors and n e u r o e n d o c r i n e functions. In adults estrogens and androgens activate sex-specific neuronal functions that differentiated early in development ~. The perinatal organization of these functions is also controlled by the gonadal steroid hormones and involves two processes: defeminization, thought to be mediated by the enzymatic conversion (aromatization) of androgen to estrogen, and its subsequent binding to estrogen receptors; and masculinization, which may be mediated by the binding of hormones to both androgen and estrogen receptors 3.9.
receptors in various areas of the developing brain, including the cerebral cortex and limbic structures such as the hypothalamus, preoptic area and amygdala 4. 17,24,25,28. Studies on the postnatal ontogeny of these receptors 2.11.14.2°.27,38,39 include those demonstrating
In addition to influencing sexual differentiation of the brain, estrogens may participate in the growth and maintenance of developing neurons 22,29. In cul0165-3806/83/$03.00 © 1983 Elsevier Science Publishers B.V.
that estrogen receptors in the male hypothalamus and limbic brain are occupied by estrogen derived from testosteronel2. 37. A m o n g studies on the prenatal development of estrogen receptors, few have examined these receptors from the time of their initial appearance13.35.36. In this study we examined the ontogeny of estrogen receptors in the mouse brain from the time of their prenatal appearance, by radioautography and cell nuclear isolation following in vivo labeling, and by cytosol receptor assays following in vitro labeling
with [3H]moxestrol, a synthetic estrogen which does not bind to alpha-fetoprotein (AFP). A companion paper 6 presents a complementary study using in vitro assays of cytosols from the hypothalamus and cerebral cortex of the pre- and postnatal mouse. MATERIALS AND METHODS Animals The ontogeny of neuronal estrogen receptors and the ability of neurons in the brain to bind estrogens was studied by radioautography, nuclear isolation techniques, and receptor assay in mice of the R i l l strain from embryonic day ( E ) l l to postnatal day (P)9. The fetal and postnatal mice utilized in this study were obtained by timed mating from the breeding colony of one of the authors (C.D.T.-A.). The mice were maintained in their own room kept at a constant temperature of 70-72 °F on a 14:10 light/ dark cycle; they were fed mouse breeder chow and water ad libitum, as well as oats and whole wheat bread twice weekly. All mice were handled daily. The males were placed with females (ratio of 2:3) for a single 45-min period daily (from 07.30 to 08.15 h) and checked immediately thereafter for the presence of vaginal plugs, evidence of sperm transfer. The day a vaginal plug was found (the first 24 h) was considered day 0 (E0) of gestation. Offspring were virtually always born in the early morning of E l 9 (P1). In vivo studies Binding of [3H]moxestrol and/or its metabolites to receptors in cell nuclei of brains of fetal mice, ages E l l (number of experiments, n = 1), El3 (n = 2), E14 (n = 2), El5 (n = 3), El6 (n = 2), El7 (n = 3) and E18 (n = 5), was studied by the cell nuclear isolation technique 16. Pregnant mice were injected subcutaneously with = 400 pmoles of [3H]moxestrol (= 14 nmol/kg) (New England Nuclear; spec. act. 88 Ci/mmol) and decapitated 2 h later, a survival time which maximizes the ratio of bound to unbound radioactive estrogen in brain cells 41. Levels of radioactivity in 25 Izl aliquots of maternal blood serum were determined by scintillation counting. The fetuses were removed, weighed, and placed on ice before the heads were homogenized in a solution of sucrose and Triton X-100. After the homogenates were filtered through nylon mesh to remove connective tissue, the
cell nuclei were isolated from the filtrates, the radioactive moxestrol and/or its metabolites were extracted from the nuclear pellets with ethanol, and the extracts were counted in Liquiscint (National Diagnostics, Somerville, N J) to determine levels of radioactivity. These levels were expressed as fmol/mg DNA, based on determinations of nuclear DNA in the pellets by the method of Burton 5. In vitro studies The ontogeny of estrogen receptors in the brains of fetal mice, ages El3 (n = 6), El5 (n = 6) and El8 (n = 6), was examined by exchange assays for cytosol receptors 6. Cytosols were prepared as outlined in the companion paper 6, passed through LH-20 columns to remove unlabeled endogenous estrogens which may have bound to receptors when the brains were homogenized, thereby avoiding subsequent competition with [3H]moxestrol, then incubated with 10-7 M [3H]moxestrol for 4 h, half of each sample in the presence of 10-6 M diethylstilbestrol to determine nonspecific binding. Nuclear D N A was determined and specific binding of radioactivity from [3H]moxestrol to estrogen receptors in the cytosol was expressed as fmol/mg DNA. Radioautographic studies The uptake and retention of radioactivity from [3H]moxestrol by estrogen receptor-containing cells in the brains of fetal mice, ages El3, (n = 4), El5 (n = 4) and El8 (n = 4), and postnatal mice, ages P1 (n = 8) and P9 (n = 6) were studied using thawmount radioautography which is designed to prevent or minimize translocation of soluble substances 7. Pregnant mice received subcutaneous injections in the back of the neck with ~ 400 pmol of [3H]moxestrol (= 14 nmol/kg) and were decapitated 2 h later. Levels of radioactivity in 25 ~1 aliquots of maternal blood serum were established by scintillation counting. The fetuses were immediately removed and placed on ice until the head of each was quickly frozen onto a brass cryostat chuck with powdered dry-ice, then stored in liquid nitrogen. The heads of the P1 mice and the brains of the P9 mice were similarly frozen. Unfixed, unembedded, coronal frozen sections, 2-4/~m thick and 30-50/~m apart, were cut at --20 °C in a cryostat, thaw mounted onto glass slides precoated with Kodak NTB-3 radioauto-
to the PO'A and hypothalamus. A cell with a density of silver grains at least 5 times that of an equal volume of adjacent extracellular tissue was considered labeledl9. Control radioautograms provided no evidence for either negative chemography (false negatives) or positive chemography (false positives).
graphic emulsion, exposed at 4 °C with Drierite from 675 to 738 days, developed in Dektol and stained with cresyl violet acetate. For each age the total number of labeled cells in the cingulate/frontal cortex (CTX), preoptic area (POA), ventromedial nucleus (VM) and arcuate nucleus ( A R C ) was counted in radioautograms spanning the anterior border of the P O A and the posterior border of the A R C . The data for each region are presented as an average per coronal section, based on the total number of labeled cells in all sections which included that region. The anterior cingulate cortex was selected, because earlier observations (C.D. Toran-Allerand, unpublished observations) suggested that in the cerebral cortex of the neonatal rat most of the neurons containing estrogen receptors appear to be localized in the deepest layers (V and VI) of the anterior cingulate cortex ]() with some extension laterally into the frontal cortex as also shown subsequently by Sheridan 24. These regions lie dorsal
RESULTS The estrogen receptors in the brains of fetal and neonatal rats and mice are mostly unoccupied by maternal estrogens, because alpha-fetoprotein (AFP), a protein in the serum produced by the fetal yolk sac and liver and the neonatal liver, is believed to sequester extracellulary the high levels of maternal estradiol that cross the placenta LS. This scavenging action of A F P presumably protects estrogen sensitive tissues from the masculinization and defeminization that such high levels of estradiol could otherwise produce 2]. Since [3H]moxestroi, a synthet-
24
m A.
•
B.
22
" 6 00~
:t
2O
X I i
40C o
2oc A
,~
16
z Q
A
u
z E)
Z
14
14
E
E -6 //
IO E
"-: 0.8 •
•/ '~
•
8
/
0.6 04 0.2
U -I
z
C.
1.2 1.0
0
6
•
/// /
4
2
; i II
I
,,,
1
,'6
,;
ID 0 E I-1
~
if°i
,'8 ii G e s t a t i o n a l Age (Days)
I
z 12
13
. I
•
.
•
i
15
16
17
I
Fig. 1. Growth in the population of estrogen receptors in the brains of mice during the gestational days 11-18, revealed (A) through increased binding of the synthetic estrogen [3H]moxestrol and/or metabolites (fmol/mg DNA) to estrogen receptors in cytosol (. . . . . ) and to receptors transiocated from cytosol to nuclei of cells from whole brains ( ). Moxestrol was used because, unlike estradiol, it does not bind to alpha-fetoprotein, a protein in the serum produced by the liver of perinatal mice. The population of receptors in nuclei and cytosol began to increase markedly around fetal day 15. Binding to estrogen receptors, corrected for nuclear DNA (A), increased with age, as total radioactivity in serum remained level (D), as both cell nuclear DNA content (B) and fetal body weight (C) rose.
Ill
ic estrogen does not bind to AFP, it was used to label estrogen receptors in the brains of fetal mice, in order to chart the regions in which they appear, as well as the course of their development. Pregnant mice were injected with ~ 14 nmol/kg of [3H]moxestrol, a dose estimated to saturate the population of estrogen receptors 14, and the fetuses were removed after 2 h of exposure to the hormone. Since the average amounts of radioactivity recovered in the serum of pregnant mice appeared to be level across gestational age (Fig. 10), approximately the same dose of [3H]moxestrol probably reached each fetus. Brains of fetal mice contained estrogen receptors which translocated from the cytoplasm to nuclei and were capable of taking up and retaining the non-steroidal estrogen [3H]moxestrol and/or its metabolites (Fig. 1A, solid line). The level of this uptake was extremely small on E l l (1.4 fmol/mg DNA), then increased markedly between E14 and El8 (from 1.8 to 18.0 fmol/mg DNA). Although the fetal body weight (Fig. 1C) and the content of nuclear DNA in the brain (Fig. 1B) both increased with age, the increase in nuclear binding was corrected for DNA content (Fig. 1A, solid line) and therefore reflected growth in the population of nuclear estrogen receptors per cell, rather than increases based simply on more cells. By measuring cytosol receptors, we ascertained further whether the increase in nuclear incorporation of radioactivity from [3H]moxestrol resulted from a parallel change in the cellular density of receptors. The population of receptors in the cytosol of brains of fetal mice was small on El3, increased slightly between El3 and El5, from 2.0 to 3.3 fm/mg DNA, then rose sharply by El8 to 10.8 fm/mg DNA (Fig. 1A. dashed line). These concentrations, which are adjusted for DNA content, represent growth in the population of cytosol receptors per cell, although the total number of cells in the brain did increase during the same interval, as shown by increased levels of nuclear DNA (Fig. 1B). The result of our radioautographic experiments corresponds to and complements the results of our cell fractionation and binding studies. In the CTX, POA, VM and ARC of fetal mice, the population of radioactively labeled cells, as well as the intensity of cellular labeling, increased dramatically with age. The density of radioautographic silver grains is an in-
dex of the concentration of receptors. These increases, therefore, reflect on the one hand, an increase in the number of cells with estrogen receptors and, on the other, a proliferation of receptors in individual cells. The radioautographs illustrate both types of changes (Figs. 2-5). In the CTX, no labeled cells were evident on El5 (Fig. 2a): some cells were labeled by El8 (Fig. 2b-d); and by P9 this number had increased, as had the density of labeling (Fig. 2e and f). In radioautographs of the POA and hypothalamus, virtually no cells were labeled on El3 (hypothalamus, Fig. 3a; POA, Fig. 3b); many were labeled by El5 (POA, Fig. 3c and d; VM, Fig. 4a-c; ARC, Fig. 5a-c); and by El8 both the population of labeled cells and intensity of labeling had further increased (POA, Fig. 3e and f: VM, Fig. 4d-f: ARC, Fig. 5d-f). The results from counting labeled cells confirm the changes in cell population already illustrated (Figs. 2-5), and suggest several additional developmental differences. However, the quantitative value of these data is limited, since the various regions of the brain were not sampled systematically among individuals. In the cingulate cortex, binding to estrogen receptors was evident in radioautograms by El8, but not El5. The average number of labeled cells rose between El5 and El8, from zero to 58, and again between P1 and P9, from 59 to 115 (Fig. 2). in the deeper layers of this cortical region the postnatal increase from P1 to P9 was 300%, and in its more superficial layers the increase was 150%. In the remaining (mostly frontal) cortex the average number of labeled cells rose between El5 and El8, from zero to 79, then remained at this level through P9. The population of labeled cells in the frontal cortex appeared earlier and reached a plateau earlier (by El8) than that of the cingulate which did not increase significantly until after P1, growing sharply by P9 to exceed that in the frontal cortex. The different patterns by which the cingulate and frontal cortex each acquired estrophilic cells may reflect the migratory patterns of these cells as they form the various layers of these two regions. In the POA or the hypothalamic ARC and VM almost no cells were labeled on El3; however, by E15 the average number had risen to 81 in the POA, to 23 in the ARC and to 35 in the VM. Each of these areas showed labeling at least 2-3 days sooner than the
11
Fig. 2-5. The ontogeny of estrogen receptors in selected regions of the brain in prenatal and neonatal mice of the R i l l strain is illustrated by radioautograms depicting increases in both the amount of binding per cell and the number of cells binding [3H]moxestrol and/or its metabolites over a 2-week period of development (E15-P9). Radioautograms were prepared by thawing unfixed frozen sections, 2--4/~m thick, onto glass slides precoated with Kodak NTB-3 radioautographic emulsion. They were exposed at 4 °C from 675 to 738 days, developed in Dektol and stained lightly with cresyl violet acetate 7. Magnification: 495 diameters. Fig. 2. Cingulate/frontal cortex, a: age, embryonic day El5. b-d: El8, e and f: postnatal day 9.
12 CTX. By E18 the population of labeled cells had grown by 3 to 4-fold to 290 in the POA, by 3-fold to 69 in the ARC and by 2-fold to 70 in the VM. By P9 the size of the labeled population had not changed in the ARC (n = 67) and had decreased slightly in the VM (n = 50). We do not have data for the POA on P1 or P9. Since the sparsely labeled, as well as the densely labeled, cells were counted, the quantitative data underestimate the differences among regions and ages realized by visual inspection of the radioautograms, whereby the densely labeled cells were especially noticeable. DISCUSSION The studies presented here and in the accompanying paper 6 on the ontogeny of estrogen receptors in the brains of mice reveal that a variety of brain regions containing estrogen receptors in adulthood initiate the synthesis of receptors prenatally. Our experiments employing either in vitro exchange assays of cytosols (this paper ref. 6) or the isolation of cell nuclei after in vivo labeling showed that the populations of receptors in the whole brains of fetuses were very low and at the limit of detection on E l l and E13, increased sharply between E l 4 and E18 and reached or exceeded adult levels within the following two weeks. We have shown in this and the accompanying paper 6 that this change in the population of receptors determined by in vivo and in vitro techniques reflects increases between E13 and E l 8 within specific brain regions. Radioautograms showed that these regions included the CTX, POA and the hypothalamic VM and ARC. Within the POA,VM and ARC the extent of radioautographic labeling, established both by counting the number of labeled cells and observing the density of labeling within cells, increased markedly between El5 and El8, with the POA showing the most labeling. After this time and until P9, binding of radioactivity derived from [3H]moxestrol remained high in both the ARC and VM. Receptors in the CTX appeared approximately 2-3 days later than those in the POA and hypothalamus, by El8 but not El5. They appeared at first in more cells of the frontal cortex than the cingulate, then increased by P9 in the cingulate cortex to reach the highest level of binding anywhere in the cortex.
Within each region of the brain where estrogen receptors appeared, both the number of cells with receptors, as well as the apparent density of receptors within individual cells increased with age. Although counting all labeled cells (whatever the density of their labeling) revealed both types of increases, visual inspection of the radioautograms disclosed these changes most dramatically, the increasing numbers of densely labeled cells standing out conspicuously. The in vivo and in vitro biochemical experiments, which recorded increases in binding relative to DNA content, also revealed a proliferation in the cellular population of receptors. Thus, the ontogenic profiles were concordant among the radioautographic and biochemical experiments. Friedman et al. 6 confirmed this developmental pattern and extended observations to an older age in the accompanying report on the regional distribution of estrogen receptors in cytosols prepared from the brains of perinatal mice. The ability to detect estrogen receptors earlier with biochemical methods than with radioautographic methods, El3 by binding assay and E15 by autoradiography, is in agreement with observations from studies of corticoid receptors in gliat cells18, wherein the threshold for detection of labeling by radioautography exceeded the threshold for detection by nuclear isolation or cytosol assays. On El3 only a very few cells in radioautograms were unequivocally labeled; instead, a low density of silver grains appeared evenly distributed within and among cell bodies in all regions examined. Since these radioautograms were exposed for the relatively long period of 675-738 days, grain densities probably approached a maximum and would not have increased with more exposure. The detection of estrogen receptors by El3 using binding assays and the subsequent dramatic increase in their concentration between El5 and E18 corresponds to certain observations made in other studies. Using DNA-cellulose cytosol assays Vito and Fox 36 found that the population of estrogen receptors in the combined hypothalamus and POA of mice was relatively high by El5; but, contrary to our observation of a marked increase by ES, they saw no increase until after P10. Stumpf et al. 28 reported radioautographic labeling in the hypothalamus of mice after injecting [3H]diethylstiibestrol on E16; however, they did not specify their criterion for assigning embryonic age.
© >
r~
~D
© >
© >
0
~r
0
#
r~
o
c~j
Q
t~
#
lO
In addition, the appearance and developmental increase of estrogen receptors in the regions studied closely follow the birthdates of the majority of neurons in these regions of the mouse as reported by others 1-26 including the same strain used in our studies (C.D. Toran-Allerand, unpublished observations). These birthdates range from El() to El7 and peak between E l l and El5. The temporal correspondence between this appearance and increase of limbic and cortical estrogen receptors demonstrated in this and the other studies cited, and the birthdates of neurons in the same regions of the brain supports the suggestion of Vito and Fox 36 that neuroblasts may become responsive to estrogen when they cease dividing and begin to differentiate, and soon thereafter the critical or sensitive period for sexual differentiation of the brain may start. In male rodents estradiol is derived from aromatizable androgens in neurons of the hypothalamus and POA 12, and serves to masculinize and defeminize these regions and their associated reproductive physiology and behavior3.9. Thus it is particularly interesting that the appearance and further development of estrogen receptors in the hypothalamus and POA observed here correspond temporally to the appearance of aromatase activity in the diencephalon of male and female rats on El5, peaking in the hypothalamus and POA between El7 and E19s. Birth, as well as many other developmental stages in rats occur approximately 2 days later than in mice. This relationship between neuronal birthdates and aromatase activity does not apply, however, to the CTX, because in rodents it not only lacks aromatase, but its estrogen receptors do not appear to be occupied by endogenous estrogens 12. These observations on the CTX of perinatal mice, together with the fact that AFP may sequester maternal estradiol extracellularly and that neurons in only the hypothalamus and POA of rodents convert testosterone to estradiol, raise the question of a role for cortical estrogen receptors. The presence of intraneuronal AFP only during the ontogeny of the brain, including the cerebral cortex 31, however, suggests a mechanism by which low levels of estrogens might have access to the cortical receptors despite the presence of extracellular AFP. Since intraneuronal AFP is derived from extracellular sources and not synthesized locally23, and is taken up specifically by neurons, at least in culture
(C. D. Toran-Allerand, unpublished observations; ref. 34), neuronal uptake of AFP (as the AFP/estradiol complex) would bring estradiol into the cell. Such a mechanism might provide target neurons of both sexes with the intracellular source of estrogen of non-androgenic origin which was postulated earlier 29-31. That the cortical estrogen receptors may be functional, like those of the hypothalamus and POA, and may similarly be implicated in the ontogeny of cortical regions is suggested by preliminary studies showing that cortical cultures, which contain estrogen receptors as in vivo (C. D. Toran-Allerand, J. L. Gerlach, and B. S. McEwen, unpublished observation), respond directly to exogenous estradiol by a marked enhancement of neuritic outgrowth (C. D Toran-Allerand, unpublished observation) and by the enhanced incorporation of [3H]leucine into the TCA (trichloracetic acid)-precipitable proteins of the cultures (N. J. MacLusky and C. D. Toran-Allerand, unpublished observation). In summary, this study demonstrated that the brains of perinatal mice produce estrogen receptors as early as El3, and that both the population of cells with receptors, as well as the density of receptors within cells, increases dramatically over the ensuing perinatal period in the hypothalamus, POA and CTX. The coincidence of this ontogeny with the appearance of aromatase activity and the presence of circulating and intraneuronal AFP, closely following the birthdates of neurons in these regions, suggests that maternal estradiol as well as fetal estradiol derived from testosterone influence development and sexual differentiation of the brain. ACKNOWLEDGEMENTS Supported by Grants NS07080 (to B. S. M.) and by HD-08364, NSF BNS77-0859 (to C. D. T.-A.), as well as by an institutional grant (RF81062) from the Rockefeller Foundation for research in reproductive biology. C. D. T.-A. also received support from the March of Dimes Birth Defects Foundation, the Whitehall Foundation and NIMH Research Scientist Development Award MH00192. The authors wish to acknowledge the devoted assistance of Mrs. Lily Skaredoff in the breeding of mice, as well as the excellent editorial work of Mrs. Oksana Wengerchuk and Ms. Pattey Fong.
17 REFERENCES 1 Angevine, J. B. and Sidman, R. L., Autoradiographic study of cell migration during histogenesis of cerebral cortex in the mouse, Nature (Lond.), 192 (1961) 766--768. 2 Attardi, B. and Ohno, S., Androgen and estrogen receptors in the developing mouse brain, Endocrinology, 99 (1976) 127%1290. 3 Baum, M. J., Differentiation of coital behavior in mammals: a comparative analysis, Neurosci. Biobehav. Rev., 3 (1979) 265-284. 4 Barley, J., Ginsburg, M., Greenstein, B. D., MacLusky, N. J. and Thomas, P. J., A receptor mediating sexual differentiation, Nature (Lond.), 252 (1974) 25%260. 5 Burton, K., A study of the conditions and mechanisms of the diphenylamine reaction for the colorimetric estimation of DNA, Biochem. J., 62 (1956) 315-323. 6 Friedman, W. C., McEwen, B. S., Toran-Allerand, C. D. and Gerlach, J. L., Perinatal development of hypothalamic and cortical estrogen receptors in mouse brain: methodological aspects, Develop. Brain Res., 11 (1983) 1%27. 7 Gerlach, J. L. and McEwen, B. S., Rat brain binds adrenal steroid hormone: Radioautography of hippocampus with corticosterone, Science, 175 (1972) 1133-1136. 8 George, F. W. and Ojeda, S. R., Changes in aromatase activity in the rat brain during embryonic, neonatal, and infantile development, Endocrinology, 111 (1982) 522-529. 9 Goy, R. W. and McEwen, B. S. (Eds.), Sexual Differentiation of the Brain, based on a Work Session of the Neurosciences Research Program, MIT Press, Cambridge, 1980, 211 pp. 10 Hamilton, L. W., Basic Limbic System Anatomy of the Rat, Plenum Press, New York and London, 1976, pp. 95-100, 11 Kato, J., Atsumi, Y. and Inaba, M., Estradiol receptors in female rat hypothalamus in the developmental stages and during pubescence, Endocrinology, 94 (1974) 30%317. 12 Lieberburg, I., Krey, L. C. and McEwen, B. S., Sex differences in serum testosterone and in exchangeable brain cell nuclear estradiol during the neonatal period in rats, Brain Res., 178 (1979) 207-212. 13 MacLusky, N. J., Lieberburg, I. and McEwen, B. S., The development of estrogen receptor systems in the rat brain: perinatal development, Brain Res., 178 (1979) 12%142. 14 MacLusky, N. J., Chaptal, C. and McEwen, B. S., The development of estrogen systems in the rat brain and pituitary: postnatal development, Brain Res., 178 (1979) 143-160, 15 McEwen, B. S., Gonadal steroid influences on brain development and sexual differentiation. In R. O. Greep (Ed.), Reproductive Physiology, Vol. IV, University Park Press, Baltimore, 1983, pp. 9%145. 16 McEwen, B. S. and Zigmond, R, E., Isolation of brain cell nuclei. In N. Marks and R. Rodnight (Eds.), Methods in Neurochemistry, Plenum Press, New York, 1972, pp. 140-161. 17 McEwen, B. S., Plapinger, L., Chaptal, C., Gerlach, J. and Wallach, G., Role of fetoneonatal estrogen binding proteins in the association of estrogen with neonatal brain cell nuclear receptors, Brain Res., 96 (1975) 400-406. 18 Meyer, J. S., Leveille, P. J., de Vellis, J., Gerlach, J. L. and McEwen, B. S., Evidence for glucocorticoid target cells in rat optic nerve. Hormone binding and glycerolphosphate dehydrogenase induction, J. Neurochem., 39 (1982)
436-442. 19 Morrell, J. I. and Pfaff, D. W., Autoradiographic technique for steroid hormone localization: application to the vertebrate brain. In N. T. Adler (Ed.), Neuroendocrinology of Reproduction: Physiology and Behavior, Plenum Press, New York, 1981, pp. 51%531. 20 Plapinger, L. and McEwen, B. S., Ontogeny of estradiolbinding sites in rat brain. I. Appearance of presumptive adult receptors in cytosol and nuclei, Endocrinology, 93 (1973) 1119-1128. 21 Plapinger, L. and McEwen, B. S., Gonadal steroid-brain interactions in sexual differentiation. In J. Hutchinson (Ed.), Biological Determinants of Sexual Behavior, J. Wiley, New York, pp. 193-218. 22 Puymirat, J., Loudes, C., Faivre-Bauman, A., Bourre, J. M. and Tixier-Vidal, A., Expression of neuronal functions by mouse fetal hypothalamic cells cultured in hormonally defined medium. In G. Sato, A. B. Pardee, D. A. Sirbasku (Eds.), Growth of Cells in Hormonally Defined Media, Cold Spring Harbor Laboratories, Cold Spring Harbor, NY, 1982, pp. 1033-1052. 23 Schachter, B. S. and Toran-Allerand, C. D., Intraneuronal alpha-fetoprotein and albumin are not synthesized locally in developing brain, Develop. brain Res., 5 (1982) 93-98. 24 Sheridan, P. J., Estrogen binding in the neonatal cortex, Brain Res., 178 (1979) 201-206. 25 Sheridan, P. J., Sar, M. and Stumpf, W. E., Autoradiographic localization of 3H-estradiol or its metabolites in the central nervous system of the developing rat, Endocrinology, 94 (1974) 1386--1390. 26 Shimada, M. and Nakamura, T., Time of neuron origin in mouse hypothalamic nuclei, Exp. Neurol., 41 (1973) 163173. 27 Spona, J., Bieglmayer, C., Adamiker, C. and Jettmar, W., Ontogeny of 17fl-estradiol-binding protein in the female rat hypothalamus and anterior pituitary, FEBS Lett., 76 (1977) 306--310. 28 Stumpf, W. E., Narbaitz, R. and Sar, M., Estrogen receptors in the fetal mouse, J. Steroid Biochem., 12 (1980) 55-63. 29 Toran-Allerand, C. D., Sex steroids and the development of the newborn mouse hypothalamus and preoptic area in vitro: implications for sexual differentiation, Brain Res., 106 (1976) 407--412. 30 Toran-Allerand, C. D., Sex steroids and the development of the newborn mouse hypothalamus and preoptic area in vitro: II. Morphological correlates and hormonal specificity, Brain Res., 189 (1980) 413--427. 31 Toran-Allerand, C. D., Coexistence of alpha-fetoprotein albumin and transferrin immunoreactivity in neurons of the developing mouse brain, Nature (Lond.), 286 (1980) 733--735. 32 Toran-Allerand, C. D., Gerlach, J. L. and McEwen, B. S., Autoradiographic localization of [3H]estradiol releated to steroid responsiveness in cultures of the newborn mouse hypothalamus and preoptic area, Brain Res., 184 (1980) 517-522. 33 Toran-Allerand, C. D., Hashimoto, K., Greenough, W. T. and Saltarelli, M., Sex steroids and the development of the newborn mouse hypothalamus and preoptic area in vitro: III. Effects of estrogen on dendritic differentiation, Brain Res., in press. 34 Uriel, J., Faivre-Bauman, A., Trojan, J. and Foiret, D., Immunocytochemical demonstration of alpha-fetoprotein
18
35 36
37
38
uptake by primary cultures of fetal hemisphere cells from mouse brain, Neurosci. Lett., 27 (198l) 171-175. Vito, C. C. and Fox, T. O., Embryonic rodent brain contains estrogen receptors, Science, 204 (1979) 517-519. Vito, C. C. and Fox, T. O., Androgen and estrogen receptors in embryonic and neonatal rat brain, Develop. Brain Res., 2 (1982) 9%110. Westley, B. R. and Salaman, D. F., Role of oestrogen receptor in androgen-induced sexual differentiation of the brain, Nature (Lond.). 262 (1976) 407-408. Westley, B. R., Thomas, P. J., Salaman, D. F., Knight, A. and Barley, J., Properties and partial purification of an oes-
trogen receptor from neonatal rat brain, Brain Res., 113 (1976) 441-447. 39 White, J. O., Hall, C. and Lim. L., Developmental changes in the content of oestrogen receptors in the hypothalamus of the female rat, Biochem. J., 184 (1979) 465-468. 40 Young, W. C., Goy, R. W. and Phoenix, C. H.. Hormones and sexual behavior. Broad relationships exist between the gonadal hormones and behavior, Science, 143 (1964) 212-218. 41 Zigmond, R. E. and McEwen, B. S., Selective retention of oestradiol by cell nuclei in specific brain regions of the ovariectomized rat, J. Neurochem., 17 (1970) 889-899.
Note added in proof Recently, an autoradiographic study with [3H]diethylstilbestrol has provided independent confirmation of the prenatal detectability of estrogen receptors in mouse hypothalamus and preoptic area around fetal day 15 (Keefer, D. A. and Holderegger, C., Anat. Rec., 205 (1983) 96A).