postnatal development and sex difference in neurons containing estrogen receptor-? immunoreactivity...

Postnatal Development and SexDifference in Neurons Containing

Estrogen Receptor-a Immunoreactivity inthe Preoptic Brain, the Diencephalon, and

the Amygdala in the Rat

MAKOTO YOKOSUKA,1 HIROAKI OKAMURA,2 AND SHINJI HAYASHI1*1Department of Anatomy and Embryology, Tokyo Metropolitan Institute for Neuroscience,

2-6 Musashidai, Fuchu, Tokyo 183, Japan2Department of Physiology, National Institute of Animal Industry,

Kukizaki, Ibaraki 305, Japan

ABSTRACTEstrogen has been considered as a key substance that induces sexual differentiation of

the brain during fetal and neonatal life in the rat. Thus, to define the brain regions involved inthe brain sexual differentiation, we examined the regions where the estrogen receptor (ER) islocated in the developing rat brain. We examined immunohistochemical distribution of thecells containing estrogen receptor-a (ER-a) in the preoptic region, the diencephalon, and theamygdala in male and female rats on postnatal days 1–35 (PD1–PD35). The antibody usedrecognizes ER-a equally well for both occupied and unoccupied forms. ER-a immunostainingwas restricted to the cell nuclei of specific cell groups. In PD1 rats, ER-a-immunoreactive(ER-IR) signals were detected in the lateral septum, the organum vasculosum laminaterminalis, the medial preoptic nucleus (MPN), the median preoptic nucleus, the bed nucleusof the stria terminalis, the hypothalamic periventricular nucleus, the lateral habenula, theposterodorsal part of the medial amygdala nucleus, the posterior part of the cortical amygdalanucleus, the hypothalamic ventromedial nucleus (VMH), the hypothalamic arcuate nucleus,and the posterior hypothalamic periventricular nucleus. The distribution pattern of ER-IRcells in the newborn rat was much the same as that in the adult in the preoptic-hypothalamicand amygdala regions. Moreover, the signals in the MPN and the VMH were stronger in thefemale than in the male, perhaps reflecting the ability of estrogen generated by aromatizationof testosterone in the male to down-regulate the ER signal. Thus, the brain regions showingsex differences may be sites of sexual differentiation of the brain by aromatizable androgenduring the neonatal period. J. Comp. Neurol. 389:81–93, 1997. r 1997 Wiley-Liss, Inc.

Indexing terms: hypothalamus; preoptic area; immunohistochemistry; sex differentiation;

aromatization

In rodents, administration of androgen (Barraclough,1961) or estrogen (Takewaki, 1962; Gorski, 1963; Aiharaand Hayashi, 1989) to the neonatal female induces steril-ization characterized by ovulatory failure and loss offemale sexual behavior, whereas castration of neonatalmale rats evokes the capacity for sexual cyclicity andlordosis behavior that is characteristic in the female rat.These treatments are effective only during the ‘‘criticalperiod’’ of perinatal life, and the steroids given are consid-ered to masculinize or defeminize the brain. Accumulatingevidence has shown that androgen given to the newbornfemale rat becomes effective after being converted into

estrogen by aromatase present in the neonatal brain (Goyand McEwen, 1980). Thus, it has been considered that,during the neonatal period, estrogen produced from andro-

Grant sponsor: Ministry of Education, Science and Culture of Japan;Grant number: 08740644 (M.Y.); Grant number: 08640853 (S.H.); Grantnumber: 09640799 (S.H.); Grant number: 09044245 (S.H.); Grant sponsor:Japan Foundation for Aging and Health; Grant number: 92A2201; Grantsponsor: Ministry of Agriculture, Forestry and Fisheries of Japan (S.H.).

*Correspondence to: Shinji Hayashi, PhD, Department of Anatomy andEmbryology, Tokyo Metropolitan Institute for Neuroscience, 2-6 Musashidai,Fuchu, Tokyo 183-0042, Japan. E-mail: [email protected]

Received 29 November 1994; Revised 3 June 1997; Accepted 3 July 1997

THE JOURNAL OF COMPARATIVE NEUROLOGY 389:81–93 (1997)

r 1997 WILEY-LISS, INC.

gen by aromatization binds to estrogen receptors (ER) inthe target neurons and exerts a defeminizing effect (Mc-Ewen et al., 1977; MacLusky et al., 1985; Naftolin et al.,1990). Studies using binding assays (MacLusky et al.,1979a,b; Vito and Fox, 1982), autoradiography (Sheridanet al., 1974; Brown et al., 1989), and in situ hybridizationhistochemistry (DonCarlos and Handa, 1994) have re-vealed the presence of estrogen binding or ER mRNA inthe rat brain during the ‘‘critical period’’ of sexual differen-tiation of the brain.

Immunohistochemistry is better than ligand-bindingautoradiography or binding assays for localizing cellscontaining ER molecules in the brain, because it does notrequire the ligand molecules for ER detection. On the otherhand, gene expression of ER mRNA is not always parallelto the appearance of ER molecules (Maggi et al., 1989;Skipper et al., 1993; Orikasa et al., 1996). Recently, amethod for quantitative in vitro autoradiography of occu-pied ER was introduced (Walters et al., 1993), with which asex difference in the ER in the developing rat brain of theoccupied form was detected (Kuhnemann et al., 1994).However, involvement of both occupied and unoccupiedforms of ER in the sex differentiation of the brain has yet tobe analyzed.

Suppression of ER expression by estrogen given exog-enously has been reported at protein (Lustig et al., 1989;Koch, 1990; Yuri and Kawata, 1991; Blaustein, 1993;Orikasa et al., 1994, 1995) as well as gene expression(Lauber et al., 1991; Simerly and Young, 1991; Orikasa etal., 1995, 1996) levels in the rat brain. In this study, weexamined ontogeny of the cells containing estrogen recep-tor-a (ER-a) in the newborn and pubertal rat brain byimmunohistochemistry with an antiserum that recognizesER-a molecules specifically, regardless of occupancy withestrogen (Okamura et al., 1992). The present experimentshad two main aims: 1) to examine distribution of ER-a inthe forebrain by immunohistochemistry in newborn and

young rats and 2) to reveal suppression of ER-a expressionin those regions in the brain that might be attributable toendogenous estrogen converted from androgen suppliedfrom the testes in situ.

MATERIALS AND METHODS

Animals and tissue preparation

Sixty-six female and male rats of the Sprague-Dawleystrain were used, ranging in age between postnatal day 1(PD1 5 day of birth), PD10, PD21, and PD35. To minimizebatch-to-batch variations, pairs of rats of both sexes in thesame age group were collected from mothers, perfused,sectioned, and incubated for immunohistochemistry at thesame time. For PD1, PD10, PD21, and PD35, 10, 10, 10,and 4 pairs were utilized, respectively. Animals wereanesthetized on ice (for PD1) or with Nembutal (for PD10,PD21, and PD35) and perfused transcardially with 0.1 Mphosphate-buffered saline (PBS; pH 7.4), followed by ice-cold fixative solution (4% paraformaldehyde in 0.1 Mphosphate buffer; PB). Brains were removed from the skulland postfixed overnight in the same fixative solution at4°C. The materials were transferred to 30% sucrose in 0.1M PB until they sank. They were kept in a cryoprotectantsolution (0.1 M PB, pH 7.4, containing 30% sucrose, 1%polyvinyl pyrrolidone, and 30% ethylene glycol) at 220°Cuntil further processing. All experiments were performedin compliance with national regulations and the AnimalExperimentation Guidelines of the Tokyo MetropolitanInstitute for Neuroscience.

Immunohistochemistry

To minimize the variation in staining between each run,immunohistochemical procedures were carried out in par-allel in both sexes in the same age groups. Serial frontalsections of the brain were made at 40 µm thickness by a

Abbreviations

2n optic nerveaco anterior commissure, olfactory limbact anterior commissure, temporal limbAHN anterior hypothalamic nucleusAMPN anterior medial preoptic nucleusARH hypothalamic arcuate nucleusBLA basolateral amygdala nucleusBST bed nucleus of the stria terminaliscc corpus callosumCOAa cortical amygdala nucleus, anterior partCOAp cortical amygdala nucleus, posterior partCP caudate-putamencpd cerebral peduncleDG dentate gyrusDMH hypothalamic dorsomedial nucleusfr fasciculus retroflexusfx columns of the fornixGPl globus pallidus, lateral segmentINF infundibulumint internal capsuleLA lateral amygdala nucleusLH lateral habenulaLPO lateral preoptic areaLS lateral septum nucleusLSi lateral septum nucleus, intermediate partME median eminenceMEApd medial amygdala nucleus, posterodorsal partMEApv medial amygdala nucleus, posteroventral partMEPO median preoptic nucleusMH medial habenula

MM medial mammillary nucleusMPN medial preoptic nucleusMPNc medial preoptic nucleus, central partMPO medial preoptic areaMS medial septal nucleusmtt mammillothalamic tractNLOT nucleus of the lateral olfactory tractoch optic chiasmopt optic tractOVLT organum vasculosum lamina terminalisPA posterior amygdala nucleusPIR piriform cortexPMv ventral premammillary nucleusPVa hypothalamic periventricular nucleus, anterior partPVH hypothalamic paraventricular nucleusPVp hypothalamic periventricular nucleus, posterior partPVpo preoptic periventricular nucleusSCH suprachiasmatic nucleusSFO subfornical organSLSI sublenticular substantia innominataSO supraoptic nucleusst stria terminalisSUM supramammillary nucleusV3 third ventricleV3m third ventricle, mammillary recessvhc ventral hippocampal commissureVL lateral ventricleVMH hypothalamic ventromedial nucleusVMHvl hypothalamic ventromedial nucleus, ventrolateral part

82 M. YOKOSUKA ET AL.

freezing microtome between the levels of the medialseptum and the medial mammillary nucleus. Every secondsection was selected for immunohistochemistry of ER; theremaining, adjacent sections were used for control stain-ing as described below or were stained with cresyl violetfor determination of brain areas. In addition, some brains(two sex pairs for PD1–PD21) were cut parasagittally at 40µm. The latter sections allowed us to observe ER stainingat the level of the olfactory bulb.

Immunostaining for neonatal rat ER-a was performedas described previously (Yokosuka and Hayashi, 1992),with slight modifications. Briefly, after elimination ofendogenous peroxidase activity by incubation with 3%hydrogen peroxide in absolute methanol, the sections wererinsed with 0.1% Triton X-100 in 50 mM PBS (PBST; pH7.4) for 2 hours with three or four changes; then, nonspe-cific binding components were blocked with a solution of1% bovine serum albumin (BSA) in PBST (BSA-PBST)containing 10% normal goat serum and 0.05% sodiumazide for 1 hour at room temperature. The sections weresubsequently incubated with the anti-rat ER-a serum(AS409: Okamura et al., 1992; diluted to 1:20,000 withBSA-PBST) for 72 hours at 4°C. They were incubatedfurther with the biotinylated goat anti-rabbit IgG (2.0µl/ml; Vector Laboratories, Burlingame, CA) in BSA-PBSTfor 2 hours at room temperature, and then with theavidin-biotin complex solution (4.5 µl each/ml; Vector EliteKit) in BSA-PBST for 1 hour at room temperature. Eachstep was followed by four 15 minutes washes with PBST.After the last wash, sections were immersed in 50 mMTris-HCl buffer (pH 7.4) for 30 minutes with three changes,then were incubated with the 3,38-diaminobenzidine (DAB;0.2 mg/ml; Sigma Chemical Co., St. Louis, MO) and0.0025% hydrogen peroxide in the same buffer for 8minutes. The reaction was stopped by transferring thesections into the Tris-HCl buffer. After intensive washingwith 10 mM PBS, sections were mounted onto gelatin-coated glass slides. They were dehydrated in gradedalcohols, cleared in xylene, coverslipped with mountingmedium, and analyzed under a light microscope.

Characterization of AS409 has already been reported(Okamura et al., 1992). It was raised in a rabbit immu-nized with a fusion protein of b-galactosidase and an ER-amolecule corresponding to amino acid 61 to the carboxylterminus (5600). Thus, AS409 may recognize any epitopicconfigurations of the native ER-a molecule except for theone-tenth from the N-terminus (Okamura et al., 1992).

Immunohistochemical controls included substitution ofthe first antibody with preimmune rabbit serum andomission of the first and/or the second antibodies. Specific-ity of the antiserum was confirmed by preadsorption tests,in which the culture media containing the Escherichia colicells transfected with the vector with or without ER-acDNA (see Okamura et al., 1992), which had been stored at280°C, was thawed, sonicated, and centrifuged at 15,000gfor 15 minutes. The precipitate was homogenized bysonication in 10 mM PBS (pH 7.4) containing 0.01 % TritonX-100. Protein content of the resuspended solution wasmeasured by a modified Lowry method. The cell lysateequivalent to 1,850 µg protein was mixed with the antise-rum in PBS at a final dilution of 1:1,000, which contained88 µg serum protein. The mixture was kept overnight at4°C, and the supernatant was used as preadsorbed me-

dium. The brain slices were incubated in the media atdilution of 1:1,000–10,000 and processed as describedabove.

The antiserum preadsorbed with the lysate from the E.coli cells that expressed ER-a molecules failed to show ERimmunoreactivity (ER-IR) in the hypothalamus, whereasthe other antiserum preincubated with the control celllysate detected clear ER-IR. AS409 has been proved torecognize both occupied and unoccupied forms of ER-a inin vitro assays (Okamura et al., 1992). This was confirmedby immunohistochemistry in our laboratory in a systemusing neonatally castrated male pups. ER-IR in the medialpreoptic nucleus (MPN) was not suppressed until 10 hoursafter an injection of estradiol benzoate, whereas bloodestrogen reached the maximum level within 20 minutesafter the administration (data not shown). If immunoreac-tivities were suppressed shortly after ligand administra-tion, the antibodies were considered to recognize specifi-cally the unoccupied form of the ER-a. This observationwas in accordance with observations by Blaustein (1993),who examined two different antisera (ER21 and ER715),which recognize distinct epitopes of the ER-a molecule andwho reported that ER21 recognized both occupied andunoccupied forms of ER-a.

Quantification of ER-IR positive cells

To determine the brain structures, brain atlases byPaxinos et al. (1991) and Swanson (1992) were used. Toquantify the number of ER-IR cells, sections were carefullymatched across animals under a microscope at leastwithin the same age groups according to appearance of thebrain structures in sections stained with cresyl violet andby immunohistochemistry. For the MPN of PD1 and PD10rats, a single section corresponding to ‘‘P0–17 (Figure 88)’’of the Paxinos et al. (1991) atlas was selected for eachanimal because the structure is altered greatly from thatof the adjacent sections, whereas, for PD21 and PD35 rats,two adjacent sections of 120 µm apart that correspond tosections 20 and 21 of the Swanson (1992) atlas wereselected. The selected sections are shown in Figure 7. Forthe hypothalamic ventromedial (VMH) and arcuate (ARH)nuclei, eight serial sections, 120 µm apart for each, wereselected. Those sections represented the entire rostrocau-dal extent of these nuclei.

The number of ER-IR cells was measured by a computer-assisted image-analysis system (VIDAS, Kontron, Ger-many). Each pixel within the area to be measured wasscaled from 0 to 255 (0 5 black, 255 5 white). An imagebrighter than 201 (21% of the total scale) was discarded asbackground level, whereas the darker signals less than 50(20% of the total scale) were transformed to scale 0 (i.e.,pure black) as the positive signals. The scaled imagebetween 50 and 150 was then expanded to between 0 and255. Any darker signals below the threshold level of 178were considered as positive. This procedure eventuallydefined any darker pixels scaled less than 80% of thedarkness from the background-subtracted level as ER-IR-positive signals. On the image, total and median values ofthose ER-IR-positive areas were obtained. Numbers ofER-IR cells were calculated as the value of total areasdivided by the median value. Preliminary counting carriedout manually under a microscope gave values similar tothose obtained by using this procedure (data not shown).

ESTROGEN RECEPTOR IN THE NEONATAL RAT BRAIN 83

Statistical differences between sexes were examinedusing ANOVA and then paired t-tests, because the immu-nohistochemical preparations were prepared in parallelbetween the sexes as explained above. P values less than 0.05were considered significant. Differences in ER immunostain-ing across ages were not analyzed statistically; they could notbe considered sufficiently comparable (see Discussion).

RESULTS

In the preoptic region, the diencephalon, and the amyg-dala, the ER-IR signals were detected from PD1 to PD35rats of both sexes. No differences in the distributionpattern between sexes and across the different ages werefound. The distribution of ER-IR cells is represented in onebrain of a PD1 female rat and is diagrammatically shownin Figure 1. Details of the distribution of ER-IR in the PD1female pup are also shown in photomicrographs in Figures

2–5. The ER-IR signals were confined to the cell nuclei(Fig. 2B inset, 4C).

Olfactory bulb, septum, and preoptic brain

No ER-IR was found in the main and accessory olfactorybulb and in the islands of Calleja. Cells containing ER-IRwere found in the intermediate part of the lateral septum(Fig. 5A,B), the organum vasculosum lamina terminalis,and the anterior portion of the sublenticular substantiainnominata (shaded areas in Fig. 1A). The density of thesecells was low and they were scattered.

In the preoptic area, a dense population of ER-IR cellswas found in the anterior preoptic nucleus (AMPN; Fig.2A), in the preoptic periventricular nucleus (PVpo; Fig.2B), and in the MPN (Fig. 2B,D). The intensity of ER-IRand the density of the population in these nuclei were high(Fig. 2B inset). ER-IR signals were less intense in the

Fig. 1. Diagrams showing distribution of estrogen receptor immunoreactivity (ER-IR; gray area) inthe preoptic brain, the diencephalon, and the amygdala in a postnatal day (PD) 1 female rat. The mostrostral section (A 5 0 µm) is at the level of the rostral end of the third ventricle. Distances from section Aare indicated in parentheses. Black areas are ventricles. For abbreviations, see list.


Fig. 2. ER-IR in the preoptic region in the female rat within 24hours after birth (PD1). Sections A–D are aligned in rostrocaudaldirection. A,B: Sections 160 µm apart, from the same rat. AbundantER-IR signals were detected in the AMPN (A) and the MPN and thePVpo (B). Inset in B is an enlargement of the region indicated by athick arrow in B. B corresponds to a plane slightly posterior to that inFigure 1B. C,D: Adjacent sections corresponding to the plane of Figure

1C stained with cresyl violet (C) and for ER-IR (D). Arrows withasterisks indicate the same blood vessels. Note that the intensity ofER-IR signals in MPNc was less than that in the surrounding area (D).Abundant ER-IR signals were encountered in the MPN and the BST.Scale bars 5 100 µm for A–D, 10 µm for inset. For abbreviations, seelist.


central part of the medial preoptic nucleus (MPNc) than inthe adjacent regions (Figs. 2C,D, 7).

In the bed nucleus of the stria terminalis (BST), ER-IRwere also very frequently encountered. ER-IR lay in anunbroken stream from dorsal regions of the BST to ventralregions surrounding the MPN, as shown in Figure 2B,D.This is also shown diagrammatically in Figures 1C,D and7. In the median preoptic nucleus and the peripheralregion of the subfornical organ, cells with ER-IR wereclearly detected (Fig. 5D,E).

Hypothalamus

The intensity of ER-IR and the density of ER-IR cellswere high in the ventrolateral part of the VMH (VMHvl;Fig. 3A,B), the ARH (Fig. 3A,B), and the posterior periven-tricular hypothalamic nucleus (PVp; Fig. 3C,D). In particu-lar, ER-IR was observed throughout the rostrocaudalextent of the VMH. The population formed two peaks inthe rostrocaudal direction, as shown in Figure 9, thecaudal value being higher than the rostral. A group of cells

Fig. 3. ER-IR in the caudal part of the hypothalamus in the PD1female rat. A: Cells with ER-IR signals were encountered in theventrolateral part of the VMHv1 and the ARH. B: Section 640 µmposterior to that in A. ER signals were very dense in the most caudal

portion of the VMHvl. They were also observed in the ARH. Thissection corresponds to the plane in Figure 1G. C: Section 840 µmposterior to that in A. ER signals were in the PVp. D: Section 1,200 µmposterior to that in A. For abbreviations, see list. Scale bar 5 100 µm.


that contain ER-IR signals was also found in the ventralportion of the dorsomedial hypothalamic nucleus, which isadjacent to dorsomedial part of the VMH (Fig. 3B). NoER-IR signals were found in regions such as the supraop-tic, the paraventricular, the suprachiasmatic, the anteriorhypothalamic, the dorsal and ventral premammillary, andthe medial mammillary nuclei.

Amygdala

In the amygdala, ER-IR signals were found in theposterodorsal part of the medial amygdala nucleus(MEApd) and in the posterior part of the cortical amygdalanucleus (COAp), which are equivalent to layers 2 and 3 ofCOAp in the Swanson atlas (1992; Fig. 4A,B). They werealso detected in the posterior amygdala nucleus (PA; Fig.1G,H). On the other hand, no positive signals were foundin the anterior part of the cortical amygdala nucleus(COAa; Fig. 1D,E) or in the posteroventral part of themedial amygdala nucleus (MEApv; Fig. 4A,B). Moreover,in the posteromedial portion of the sublenticular substan-tia innominata, which resides lateral to the hypothalamusand adjacent to the central amygdala, a small and dis-persed population of ER-IR was found (Figs. 1E, 5F,G).

ER-IR signals were restricted to the cell nuclei asobserved in the preoptic-diencephalon region. However,the appearance of the ER-IR in the amygdala was different

from that in the preoptic-hypothalamus region. The inten-sity was always weaker in the amygdala than that in thepreoptic-hypothalamic region (cf. Fig. 2B inset and 4C).

Thalamus

In the thalamus, ER-IR signals were detected only in thelateral habenula, where they were scattered (Fig. 5C). Noother regions in the thalamus contained any specificER-IR.

Sex difference in ER-IR in MPN, VMH,and ARH

As was mentioned above, the distribution patterns ofER-IR signals in both sexes in the brain appeared constantacross the ages examined. However, the intensity of ER-IRsignals was stronger in the female than in the male insome age groups, e.g., the MPN of PD10 (cf. Fig. 6C and D)and PD21, and the VMHvl of PD10–PD35. Quantificationof ER-IR-positive cell numbers in comparable regions ofthe MPN revealed significant differences (P , 0.05) be-tween both sexes at PD10 and PD21 (Fig. 8). The numberof cells showing ER-IR signals in the ARH and the VMHvlin serial sections of the rostrocaudal direction at every 160

Fig. 4. ER-IR in the amygdala in the PD1 female rat. A: The ER-IRsignals were encountered in the amygdaloid complex. They wereabundant in the MEApd and the COAp, but were few in the MEApv. B:Section 160 µm posterior to that in A. This section corresponds to theplane in Figure 1F. C: Higher magnification of a part of the COAp,

indicated by an arrow in B. ER-IR signals were restricted to the cellnuclei. However, intensity of the signals was weaker than that in theMPN (Fig. 2B, inset). For abbreviations, see list. Scale bars 5 100 µmin A,B, 10 µm in C.


Fig. 5. ER-IR in the LSi (A,B), the LH (C), the MEPO (D), the SFO,and the SLSI (F,G) in the PD1 female rat. B is a higher magnificationof the squared area in A. In C, the signals were found in the LH but notin the MH. In F, the signals were found scattered in the SLSI. G is the

squared area in F at a higher magnification. For abbreviations, seelist. Scale bars 5 100 µm in A, 25 µm in B, 50 µm in C–E, G, 400 µmin F.

µm also revealed significant differences (P , 0.05) be-tween both sexes at PD10 (ARH) and at PD10–PD35(VMHvl), as depicted in Figure 10.

DISCUSSION

Distribution of ER-IR

The ontogeny of the cells that contain ER molecules hasbeen reported for the forebrain of some mammalian spe-cies by using different techniques (e.g., autoradiography inmice: Stumpf and Sar, 1978; Sibug et al., 1991; autoradiog-raphy in rats: Sheridan et al., 1974; Brown et al., 1989;binding assays in mice: Friedman et al., 1983; bindingassays in rats: MacLusky et al., 1979a,b; Kuhnemann et

al., 1994; immunohistochemistry in Brazilian opossum:Fox et al., 1991a; in situ hybridization histochemistry forER mRNA expression in rats: DonCarlos and Handa, 1994;O’Keefe et al., 1995). However, so far, only a few reportsare available that analyze the development of ER usingimmunohistochemistry for ER.

In the present study, we used an antibody that recog-nizes equally well both occupied and unoccupied forms ofER-a and confirmed previous reports that the cells contain-ing ER-a were already detectable at PD1 in both male andfemale rats (Pasterkamp et al., 1996). The distributionpattern of the ER-IR was similar to that reported previ-ously for animals of different species in the newborn(autoradiography in the mouse: Sibug et al., 1991; immuno-

Fig. 6. ER-IR in the medial preoptic area in the corresponding sections of PD1 (A: female; B: male) andPD10 (C: female; D: male) rats. No clear sex difference was detected at PD1, whereas at PD10 theintensity was much higher in the female (C) than in the male (D). Scale bar 5 50 µm.


histochemistry in the Brazilian opossum: Fox et al., 1991a)as well as in the adult (autoradiography in the rat: Pfaffand Keiner, 1973; mouse: Stumpf and Sar, 1975; and

hamster: Krieger et al., 1976; in situ hybridization in therat: Simerly et al., 1990).

However, the present observations did not agree indetail with previous reports from autoradiography in somerespects. Sibug et al. (1991) reported estrogen-concentrat-ing cells in the anterior olfactory nucleus, in the island ofCalleja, in the interanteromedial thalamic nucleus, and inthe hypothalamic paraventricular nucleus in the 2-day-oldmouse. Moreover, strong binding of tritiated estradiol wasreported for the ventral premammillary nucleus in thenewborn rat (Sheridan et al., 1974) and mouse (Sibug etal., 1991). On the other hand, we did not detect anypositive signals for ER-a in the regions corresponding tothose nuclei. The discrepancy between those previousreports from autoradiographic studies and the present onemay be attributable to the methods used. It is possiblethat, in the present study, ER-IR signals represent onlythose cells having strong immunoreactivity, whereas thosewith low ER concentration might not be counted aspositive cells owing to weak stains. This is a methodologi-cal limitation of immunohistochemistry. The fact that thepresent report agrees well with that of Fox et al. (1991a)from the Brazilian opossum supports this explanation. Analternative explanation for the discrepancy between theresults from the two methods is that the signals withautoradiography might indicate the presence not only ofthe ER molecules but also of other estrogen bindingactivities, or some nonspecific binding or retention. In fact,a novel ER has recently been detected in rat prostate andovary (Kuiper et al., 1996). Because the structure of thenovel ER is different from that of the ER that has alreadybeen reported (Koike et al., 1987), it was proposed to callthe novel one ER-b and the previous one ER-a. Moreover,gene expression of ER-b in the brain has been reported(Shughrue et al., 1996).

We have already reported that ER-a is expressed tran-siently in the newborn rat, i.e., in layer V of the primaryauditory cortex (Yokosuka et al., 1995) and in the ventrome-dial subnucleus of the facial nucleus (Yokosuka and Haya-

Fig. 7. Camera lucida drawings indicating distribution of ER-IR inthe MPN in PD1–PD35 female rats. Diagrams on the right indicatebrain structures judged from adjacent sections stained with cresylviolet. Quantification of ER-IR-positive cells was done on a singlesection/rat for PD1 and PD10; for PD21 and PD35, it was done on twosections, as indicated in the diagrams. In the latter, the two sectionswere 120 µm apart from each other. See Materials and Methods forfurther explanations. For abbreviations, see list.

Fig. 8. Number of ER-IR cells in the corresponding areas in theMPN. ER-IR cells were counted in the areas indicated in Figure 7.Vertical bars show SEM. Asterisks indicate significant differencesbetween sexes (P , 0.05). Note that the difference in the cell numbersacross different ages was not compared. For further explanations, seetext and the legend to Figure 7.

Fig. 9. Rostrocaudal distribution of ER-IR cells in the VMH ofnewborn (PD1) and pubertal (PD35) rat. Eight serial sections, each120 µm apart, were used. n, Number of rats used; vertical barsindicate SEM. See text for further explanation.


shi, 1992; Hayashi, 1994; Orikasa et al., 1994, 1995). TheER-IR signals found in the present study in the preoptic-hypothalamic region and the amygdala, however, were nottransient but were retained at least until PD35. Becausethe distribution pattern of ER-IR cells in the brain at PD35was similar to distribution patterns that have been ob-served in the adult by using immunohistochemistry withthe same antiserum (Okamura et al., 1992; Hayashi andOkamura, 1992), it is likely that the ER-IR found in thoseregions remains in place in the adult.

Sex differences in ER-IR in newborn rats

Classical studies by Raisman and Field (1973) andMatsumoto and Arai (1980) have revealed that androgenand/or estrogen increases the number of synapses in thepreoptic area and the hypothalamus. Furthermore, McCar-thy et al. (1993) reported that inhibiting the synthesis ofER-a by providing antisense oligonucleotides interferedwith sexual differentiation of the brain.

In this study, we found sex differences in the ER-IR inneonatal rats. In the preoptic and hypothalamic regions,the intensity of the ER-IR signals was stronger in thefemale than in the male (cf. Fig. 6C and D). The number ofER-IR cells that were stained darker than 80% of the totalgray level (see Materials and Methods) was larger in the

female than the male in the ARH at PD10 and in the MPNand the VMH at PD10–PD35 (Figs. 8, 10).

The fact that the intensity of ER-IR signals was strongerin the female than in the male strongly supports theimportance of aromatization in the sexual differentiationof the brain. Testosterone secreted from the testes in themale is converted into estradiol by aromatase in the brainand seems to suppress ER subsequently. Estrogen down-regulates ER-a at the level of gene expression (Lauber etal., 1991; Simerly and Young, 1991), and it reduces ER-IR(Koch, 1990; Yuri and Kawata, 1991) and ER bindingcapacity (Lustig et al., 1989) in adult rodent brains.Recently, we have also detected down-regulation of ER-IRand ER mRNA by exogenous estradiol in the neonatal rat(Orikasa et al., 1994, 1995, 1996). Although there weredifferences in detail in the results, suppression of ERmRNA message by exogenous estrogen in the preoptic areaof neonatal rats has also been reported by DonCarlos et al.(1995). Moreover, although the serum estradiol level isalmost the same in both sexes, testosterone is higher in themale than in the female throughout the postnatal period(PD1–PD19; Dohler and Wuttke, 1975). Furthermore, thearomatase activity in the rat brain is high during thisperiod (George and Ojeda, 1982; MacLusky et al., 1985;Michnovicz et al., 1987). Thus, as a result of aromatiza-tion, substantial amounts of estradiol may be present inthe brains of newborn males, but not in females. Thisestradiol is likely to have suppressed ER-IR expression inthe neonatal male brain.

Aromatase immunoreactivity appears in the MPN andthe VMHvl, but not in the ARH, in the neonatal rat(Yokosuka et al., 1994; Tsuruo et al., 1996). Thus, with themale brain, the estrogen locally produced by aromatase inthe MPN and the VMHvl seemed to have suppressed ER-ain these regions but not in the ARH, whereas, in thefemale, lack of the substrate for aromatization (i.e., testos-terone) failed to suppress ER-a in the former regions.Thus, it is possible that the sex difference was detected inthe MPN and the VMH at PD10 and PD21 but not in theARH at PD21. We have reported that ER-a and aromataseare found in the same neurons in certain regions such asthe preoptic area, the BNST, the MeA, and the VMH infetal and neonatal rats (Tsuruo et al., 1996). However,insofar as estradiol is lipid soluble, colocalization of thetwo in the same neurons may not be essential but theymust be localized in proximity to one another to evokeestrogenic influence (Orikasa et al., 1995). Further analy-ses on the relationship between aromatase and ER expres-sion in these regions are now in progress.

In the present study, significant difference in the ER-IRcell numbers was not detected between the sexes at PD1.This might be due to an insufficient amount of estrogenconverted from testosterone in the PD1 male. An alterna-tive explanation might be that the ER-IR neurons wereless responsive to the suppressive effect of estrogen at veryearly ages. These possibilities will be clarified by furtherexamination of the ontogeny of aromatase in these regions.

Intracellular localization of ER-IR

In the present study, ER-IR was found exclusively in thecell nuclei. This observation is in good agreement withprevious reports from the adult mouse (Koch and Ehret,1989) and the developing Brazilian opossum (Fox et al.,1991a) through immunohistochemical studies. On theother hand, several studies reported that the ER-IR mate-

Fig. 10. Numbers of ER-IR cells in the VMHvl and the ARH inPD1–PD35 rats. Open and solid columns represent females and males,respectively. The cell numbers were counted in every fourth section ofserial sections of 40 µm thickness and are indicated as sum values inthe ordinate. Vertical bars indicate SEM. Asterisks indicate significantdifferences between sexes (P , 0.05). n, Number of rats used. Seefurther explanation in the text.


rials were also detectable in the cytoplasm and even in theneuronal processes in several mammalian species, e.g., theadult Brazilian opossum (Fox et al., 1991b), the guinea pig(Blaustein et al., 1992), the rat (Blaustein, 1992), the muskshrew (Dellovade et al., 1992), the ewe (Herbison et al.,1993), and the ferret (Tobet et al., 1993a,b). Furthermore,in the adult ferret forebrain, sex differences in extra-nuclear ER immunoreactivity were reported; i.e., femaleshave significantly more ER-IR products than males in theprocesses (Tobet et al., 1993b). It is noteworthy that all ofthose reports in which immunoreactive substances weredetected outside of the cell nucleus pertained to adultanimals, except for a single case of the fetal ferret (Tobet etal., 1993a). To our knowledge, no ER-IR signals were foundin extranuclear cellular regions of very young rodents anddeveloping Brazilian opossum brains. Thus, it is possiblethat the ER-a detected in the cytoplasm and/or neuronalprocesses contributes to some as yet undefined aspects ofestrogen action in the target tissue in the adult rodentsand opossums, whereas in the newborn rodents and opos-sums the ER-a is localized mainly in the cell nuclei andcontributes to the genomic actions of estrogen, includingthe sexual differentiation of the brain.

ACKNOWLEDGMENTS

We thank Dr. Bruce S. McEwen of the RockefellerUniversity for his critical reading of and linguistic helpwith the manuscript.

LITERATURE CITED

Aihara, M., and S. Hayashi (1989) Induction of persistent diestrus followedby persistent estrus is indicative of delayed maturation of tonicgonadotropin-releasing systems in the rat. Biol. Reprod. 40:96–101.

Barraclough, C.A. (1961) Production of anovulatory, sterile rats by singleinjections of testosterone propionate. Endocrinology 68:62–67.

Blaustein, J.D. (1992) Cytoplasmic estrogen receptors in rat brain: Immu-nocytochemical evidence using three antibodies with distinct epitopes.Endocrinology 131:1336–1342.

Blaustein, J.D. (1993) Estrogen receptor immunoreactivity in rat brain:rapid effects of estradiol injection. Endocrinology 132:1218–1224.

Blaustein, J.D., M.L. Lehman, J.C. Turcotte, and G. Greene (1992) Estro-gen receptors in dendrites and axon terminals in the guinea pighypothalamus. Endocrinology 18:227–239.

Brown, T.J., N.J. MacLusky, C.D. Toran-Allerand, J.E. Zielinski, and R.B.Hochberg (1989) Characterization of 11b-methoxy-16a-[125I]iodoestra-diol binding: Neuronal localization of estrogen-binding sites in thedeveloping rat brain. Endocrinology 124:2074–2088.

Dellovade, T.L., J.D. Blaustein, and E.F. Rissman (1992) Neural distribu-tion of estrogen receptor immunoreactive cells in the female muskshrew. Brain Res. 595:189–194.

Dohler, K.D., and W. Wuttke (1975) Changes with age in levels of serumgonadotropins, prolactin, and gonadal steroids in prepubertal male andfemale rats. Endocrinology 97:898–908.

DonCarlos, L.L., and R.J. Handa (1994) Developmental profile of estrogenreceptor mRNA in the preoptic area of male and female neonatal rat.Dev. Brain Res. 79:283–289.

DonCarlos, L.L., M. McAbee, D.S. Ramer-Quinn, and D.M. Stancik (1995)Estrogen receptor mRNA levels in the preoptic area of neonatal rats areresponsive to hormone manipulation. Dev. Brain Res. 84:253–260.

Fox, C.A., L.R. Ross, and C.D. Jacobson (1991a) Ontogeny of cells contain-ing estrogen receptor-like immunoreactivity in the Brazilian opossumbrain. Dev. Brain Res. 63:209–219.

Fox, C.A., L.R. Ross, R.J. Handa, and C.D. Jacobson (1991b) Localization ofcells containing estrogen receptor-like immunoreactivity in the Brazil-ian opossum brain. Brain Res. 546:96–105.

Friedman, W.J., B.S. McEwen, C.D. Toran-Allerand, and J.L. Gerlach(1983) Perinatal development of hypothalamic and cortical estrogen

receptors in mouse brain: Methodological aspects. Dev. Brain Res.11:19–27.

George, F.W., and S.R. Ojeda (1982) Changes in aromatase activity in therat brain during embryonic, neonatal, and infantile development.Endocrinology 111:522–529.

Gorski, R.A. (1963) Modification of ovulatory mechanism by postnataladministration of estrogen to the rat. Am. J. Physiol. 205:842–844.

Goy, R.W., and B.S. McEwen (1980) Sexual Differentiation of the Brain.Cambridge, MA: MIT Press.

Hayashi, S. (1994) Immunocytochemical detection of estrogen receptor inthe facial nucleus of the newborn rat by three antibodies with distinctepitopes. Horm. Behav. 28:530–536.

Hayashi, S., and H. Okamura (1992) Factors regulating sexual differentia-tion of the brain: neonatal steroid treatment and development of thereproductive brain in the rat. In A. Yokoyama (ed): Brain Control of theReproductive System. Tokyo: Japan Scientific Societies Press, and BocaRaton, FL: CRC Press, pp. 49–68.

Herbison, A.E., J.E. Robinson, and D.C. Skinner (1993) Distribution ofestrogen receptor-immunoreactive cells in the preoptic area of the ewe:co-localization with glutamic acid decarboxylase but not luteinizinghormone-releasing hormone. Neuroendocrinology 57:751–759.

Koch, M. (1990) Effects of treatment with estradiol and parental experienceon the number and distribution of estrogen-binding neurons in theovariectomized mouse brain. Neuroendocrinology 51:505–514.

Koch, M., and G. Ehret (1989) Immunocytochemical localization andquantitation of estrogen-binding cells in the male and female (virgin,pregnant, lactating) mouse brain. Brain Res. 489:101–112.

Koike, S., M. Sakai, and M. Muramatsu (1987) Molecular cloning andcharacterization of rat estrogen receptor cDNA. Nucleic Acids Res.15:2499–2513.

Krieger, M.S., J.I. Morrell, and D.W. Pfaff (1976) Autoradiographic localiza-tion of estradiol-concentrating cells in the female hamster brain.Neuroendocrinology 22:193–205.

Kuhnemann, S., T.J. Brown, R.B. Hochberg, and N.J. MacLusky (1994) Sexdifferences in the development of estrogen receptors in the rat brain.Horm. Behav. 28:483–491.

Kuiper, G.G.J.M., E. Enmark, M. Pelto-Huikko, S. Nilsson, and J.-A.Gustaffsson (1996) Cloning of a novel estrogen receptor expressed in ratprostate and ovary. Proc. Natl. Acad. Sci. USA 93:5925–5930.

Lauber, A.H., C.V. Mobbs, M. Muramatsu, and D.W. Pfaff (1991) Estrogenreceptor messenger RNA expression in rat hypothalamus as a functionof genetic sex and estrogen dose. Endocrinology 129:3180–3186.

Lustig, R.H., C.V. Mobbs, H.L. Bradlow, B.S. McEwen, and D.W. Pfaff(1989) Differential effects of estradiol and 16a-hydroxyestrone onpituitary and preoptic estrogen receptor regulation. Endocrinology125:2701–2709.

MacLusky, N.J., C. Chaptal, and B.S. McEwen (1979a) The development ofestrogen receptor systems in the rat brain and pituitary: Perinataldevelopment. Brain Res. 178:129–142.

MacLusky, N.J., C. Chaptal, and B.S. McEwen (1979b) The development ofestrogen receptor systems in the rat brain and pituitary: Postnataldevelopment. Brain Res. 178:143–160.

MacLusky, N.J., A. Philip, C. Hurlburt, and F. Naftolin (1985) Estrogenformation in the developing rat brain: Sex differences in aromataseactivity during early post-natal life. Psychoneuroendocrinology 10:355–361.

Maggi, A., L. Susanna, E. Bettini, G. Mantero, and I. Zucchi (1989)Hippocampus: a target for estrogen action in mammalian brain. Mol.Endocrinol. 3:1165–1170.

Matsumoto, A., and Y. Arai (1980) Sex dimorphism in ‘‘wiring pattern’’ inthe hypothalamic arcuate nucleus and its modification of neonatalhormone environment. Brain Res. 190:238–242.

McCarthy, M.M., E.H. Shlenker, and D.W. Pfaff (1993) Enduring conse-quences of neonatal treatment with antisense oligodeoxynucleotides toestrogen receptor messenger ribonucleic acid on sexual differentiationof rat brain. Endocrinology 133:433–439.

McEwen, B.S., I. Lieberburg, C. Chaptal, and L.C. Krey (1977) Aromatiza-tion: important for sexual differentiation of the neonatal rat brain.Horm. Behav. 9:249–263.

Michnovicz, J.J., E.F. Hahn, and J. Fishman (1987) 19-Hydroxylation andaromatization of androgens of the developing rat brain. Endocrinology121:1209–1214.

Naftolin, F., L.M. Garcia-Segura, D. Keefer, C. Leranth, N.J. MacLusky,and J.R. Brawer (1990) Estrogen effects on the synaptology andmembranes of the rat hypothalamic arcuate nucleus. Biol. Reprod.42:21–28.


Okamura, H., K. Yamamoto, S. Hayashi, A. Kuroiwa, and A. Muramatsu(1992) A polyclonal antibody to the rat oestrogen receptor expressed inEscherichia coli: Characterization and application to immunohistochem-istry. J. Endocrinol. 135:333–341.

O’Keefe, J.A., L. Yanbing, L.H. Burgess, and R.J. Handa (1995) Estrogenreceptor mRNA alterations in the developing rat hippocampus. Mol.Brain Res. 30:115–124.

Orikasa, C., H. Okamura, and S. Hayashi (1994) Estrogen receptor found inthe facial nucleus of the newborn rat is suppressed by exogenousestrogen: Immuno- and in situ hybridization histochemical studies.Dev. Brain Res. 82:9–17.

Orikasa, C., M. Yokosuka, and S. Hayashi (1995) Expression of estrogenreceptor in the facial nucleus is suppressed by estradiol, but not bytestosterone, indicating a lack of requirement for aromatization. BrainRes. 693:112–117.

Orikasa, C., K. Mizuno, Y. Sakuma, and S. Hayashi (1996) Exogenousestrogen acts differently on production of estrogen receptor in thepreoptic area and the mediobasal hypothalamic nuclei. Neurosci. Res.25:247–254.

Pasterkamp, R.J., K. Yuri, D.T.M. Visser, S. Hayashi, and M. Kawata (1996)The perinatal ontogeny of estrogen receptor-immunoreactivity in thedeveloping male and female rat hypothalamus. Dev. Brain Res. 91:300–303.

Paxinos, G., I. Tork, L.H. Tecott, and K.L. Valentino (1991) Atlas of theDeveloping Rat Brain, San Diego: Academic Press.

Pfaff, D., and M. Keiner (1973) Atlas of estradiol-concentrating cells in thecentral nervous system of the female rat. J. Comp. Neurol. 151:121–158.

Raisman, G., and P.M. Field (1973) Sexual dimorphism in the neuropil ofthe preoptic area of the rat and its dependence on neonatal androgen.Brain Res. 54:1–29.

Sheridan, P.J., M. Sar, and W.E. Stumpf (1974) Autoradiographic localiza-tion of 3H-estradiol or its metabolites in the central nervous system ofthe developing rat. Endocrinology 94:1386–1390.

Shughrue, P.J., B. Komm, and I. Merchenthaler (1996) The distribution ofestrogen receptor-b mRNA in the rat hypothalamus. Steroids 61:678–681.

Sibug, R.M., W.E. Stumpf, P.J. Shughrue, R.B. Hochberg, and U. Drews(1991) Distribution of estrogen target sites in the 2-day-old mouseforebrain and pituitary gland during the ‘‘critical period’’ of sexualdifferentiation. Dev. Brain Res. 61:11–22.

Simerly, R.B., and B.J. Young (1991) Regulation of estrogen receptormessenger ribonucleic acid in rat hypothalamus by sex steroid hor-mones. Mol. Endocrinology 5:424–432.

Simerly, R.B., C. Chang, M. Muramatsu, and L.W. Swanson (1990) Distribu-tion of androgen and estrogen receptor mRNA-containing cells in therat brain: An in situ hybridization study. J. Comp. Neurol. 294:76–95.

Skipper, J.K., L.J. Young, J.M. Bergeron, M.T. Tetzlaff, C.T. Osborn, and D.Crews (1993) Identification of an isoform of the estrogen receptormessenger RNA lacking exon four and present in the brain. Proc. Natl.Acad. Sci. USA 90:7172–7175.

Stumpf, W.E., and M. Sar (1975) Hormone-architecture of the mouse brainwith 3H-estradiol. In W.E. Stumpf and L.D. Grant (eds): AnatomicalNeuroendocrinology. New York: Karger, pp. 82–103.

Stumpf, W.E., and M. Sar (1978) Estrogen target cells in fetal brain. In G.Dorner and M. Kawakami (eds): Hormones and Brain Development.Amsterdam: Elsevier, pp. 27–33.

Swanson, L.W. (1992) Brain Maps: Structure of the Rat Brain. Amsterdam:Elsevier.

Takewaki, K. (1962) Some aspects of hormonal mechanism involved inpersistent estrus in the rat. Experientia 18:1–6.

Tobet, S.A., M.E. Basham, and M.J. Baum (1993a) Estrogen receptorimmunoreactive neurons in the fetal ferret forebrain. Dev. Brain Res.72:167–180.

Tobet, S.A., T.W. Chickering, T.O. Fox, and M.J. Baum (1993b) Sex andregional differences in intracellular localization of estrogen receptorimmunoreactivity in adult ferret forebrain. Neuroendocrinology 58:316–324.

Tsuruo, Y., K. Ishimura, S. Hayashi, and Y. Osawa (1996) Immunohisto-chemical localization of estrogen receptors within aromatase-immuno-reactive neurons in the fetal and neonatal rat brain. Anat. Embryol.193:115–121.

Vito, C.C., and T.O. Fox (1982) Androgen and estrogen receptors inembryonic and neonatal rat brain. Dev. Brain Res. 2:97–110.

Walters, M.J., T.J. Brown, R.B. Hochberg, and N.J. MacLusky (1993) Invitro autoradiographic visualization of occupied ER in the rat brainwith an iodinated estrogen ligand. J. Histochem. Cytochem. 41:1279–1290.

Yokosuka, M., and S. Hayashi (1992) Transient expression of estrogen-receptor-like immunoreactivity (ER-LI) in the facial nucleus of theneonatal rat. Neurosci. Res. 15:90–95.

Yokosuka, M., H. Okamura, and S. Hayashi (1994) Immunohistochemicaldetection of estrogen receptor and aromatase in the developing ratbrain. Neurosci. Res. Suppl. 19:S127.

Yokosuka, M., H. Okamura, and S. Hayashi (1995) Transient expression ofestrogen receptor-immunoreactivity (ER-IR) in the layer V of thedeveloping rat cerebral cortex. Dev. Brain Res. 84:99–108.

Yuri, K., and M. Kawata (1991) The effect of estrogen on the estrogenreceptor-immunoreactive cells in the rat medial preoptic nucleus. BrainRes. 548:50–54.


postnatal development and sex difference in neurons containing estrogen receptor-? immunoreactivity...

Documents