- Email: [email protected]
Prenatal experience and postnatal stress modulate the adult neurosteroid and catecholaminergic stress responses
\
Pergamon
Int[ J[ Devl Neuroscience\ Vol[ 05\ No[ 2:3\ pp[ 106Ð117\ 0887 Þ 0887 ISDN[ Published by Elsevier Science Ltd All rights reserved[ Printed in Great Britain 9625Ð4637:87 ,08[99¦9[99
PII ] S9625Ð4637"87#99913Ð9
PRENATAL EXPERIENCE AND POSTNATAL STRESS MODULATE THE ADULT NEUROSTEROID AND CATECHOLAMINERGIC STRESS RESPONSES BETTY ZIMMERBERG and RACHEL C[ BROWN Department of Psychology and Program in Neuroscience\ Bronfman Science Center\ Williams College\ Williamstown\ MA 90156\ U[S[A[ "Received 19 October 0886 ^ accepted in revised form 00 March 0887 # Abstract*Allopregnanolone "2a!hydroxy!4a!regnan!19!one# is a neuroactive steroid recently shown to be involved in the neurochemical stress response via its positive modulation of the GABAA receptor complex[ This experiment investigated the e}ects of postnatal stress "daily maternal separation during the _rst week of life# on the subsequent adult response to a stressor "09 min forced swim# in LongÐEvans rats from one of three prenatal treatment groups "alcohol\ pair!fed and control#[ Indices of stress response were allopregnanolone concentrations in plasma\ cortex and hippocampus\ and dopamine and norepinephrine concentrations in prefrontal cortex\ nucleus accumbens and striatum[ Females had higher levels of allo! pregnanolone than males in both plasma and brain[ Prenatal alcohol exposure combined with early maternal separation stress resulted in an increase in the endogenous levels of allopregnanolone in the prefrontal cortex and hippocampus of adult o}spring in response to a stressor compared to subjects without a prior history of postnatal stress ^ this e}ect was greater in females[ This increased allopregnanolone was also associated with decreased dopamine and norepinephrine levels in the prefrontal cortex[ In the prenatal alcohol!exposed o}spring\ postnatal maternal separation blunted the increase in dopamine levels in the striatum seen in both control groups[ Postnatal maternal separation increased norepinephrine levels in the nucleus accumbens regardless of prenatal experience\ while in the prefrontal cortex only prenatal diet condition "pair!feeding and alcohol# resulted in lower norepinephrine levels[ The results of this experiment suggest that experience\ both pre! and postnatal\ can have long!term consequences for the developing neurochemical responses to stressors[ Þ 0887 ISDN[ Published by Elsevier Science Ltd Key words] allopregnanolone\ neuroactive steroid\ fetal alcohol syndrome\ dopamine\ sex di}erences\ maternal separation\ prenatal stress\ alcohol[
Neurochemical stress responses are the net outcome of a complex interaction between genetically determined individual di}erences and the environmental experiences that an individual encounters as the brain matures[ Understanding how the newly characterized neuroactive steroids a}ect the stress reaction may help to explain how the environment can modulate developing neural systems involved in the stress response[ These neuroactive steroids act as neuromodulators\ binding to membrane receptors and modulating ligand!gated ion channel conductance[05 The sedative and anxiolytic e}ects of neuroactive steroids appear to be due to their positive allosteric modulation of GABAA receptor[5\03\06 The GABAA receptor contains a distinct "e[g[ non!benzodiazepine or barbiturate# binding site for steroids\04\19\10 if not multiple sites[06\12 Neuroactive steroids have been found in various areas of the brain\ with relatively higher concentrations in the cerebral cortex\1\07\10 cerebellum\ hippocampus\ striatum8 and hypothalamus[15 Allopregnanolone\ a 2a!hydroxy!4a reduced metabolite of the steroid hormone progesterone\ has been shown to be a positive allosteric modulator of the GABAA receptor10 with a potency of 0999! fold higher than that of pentobarbital[5 Allopregnanolone inhibits the binding of ð24SŁ!TBPS\ which binds to and labels a site close to or on the Cl!channel04 at submicromolar concentrations[5 The potency of allopregnanolone for displacing ð24SŁ!TBPS increases in the presence of low con! centrations of GABA[5 Allopregnanolone has been shown to potentiate chloride ion currents in rat hippocampal neurons\ and has also been shown to be more e}ective than ~unitrazepam\ a benzodiazepine\ in potentiating a response in certain recombinant receptors\ speci_cally those made up of a!0\ b!0 and g!1 subunits\ suggesting a separate binding site[13 Plasma concentrations of allopregnanolone increase after exposure to a forced swim test\ and
To whom all correspondence should be addressed[ Tel[ ] 990 302 486 1335 ^ Fax ] 990 302 486 3005 ^ E!Mail ] Betty[ZimmerbergÝwilliams[edu 106
107
B[ Zimmerberg and R[ C[ Brown
this increase mirrors the increase over a period of 29Ð59 min in the cerebral cortex and hypothala! mus[15 When allopregnanolone is given to female rats\ they show an increase in open arm entries on an elevated plus maze\ indicating decreased anxiety[1\2 Similar anxiolytic e}ects of progesterone in this paradigm have been determined to be due to the hormone|s conversion into allop! regnanolone[1 In this laboratory\ intracerebroventricular injections of allopregnanolone in infant rats separated from dams and littermates were found to decrease the number of distress vocalizations made by the pups after a brief maternal separation[25 Thus there is strong evidence to support the hypothesis that allopregnanolone can act as an endogenous neurosteroid to reduce anxiety in response to stress\ in both adult rats and neonates[ Some neuroactive steroids have been shown to speci_cally in~uence the mesolimbocortical dopa! mine pathway\ a system critical in the neurochemical stress response[ Intraventricular administration of alloTHDOC\ an anxiolytic neuroactive steroid similar to allopregnanolone\ prior to exposure to 14 min of restraint stress resulted in a dose!dependent decrease in the concentration of dopamine metabolites in the prefrontal cortex ^ no signi_cant changes were seen in the nucleus accumbens or striatum[00 Other drugs a}ecting the GABAA receptor complex "e[g[ benzodiazepines and ethanol# block activation of cortical dopamine systems by footshock\6\16 suggesting that the ventral tegmental dopamine neurons projecting to the prefrontal cortex are modulated by GABA receptors[16 The role of neuroactive steroids in stress response after chronic stress has not yet been studied[ Animal models have shown that prenatal exposure to alcohol a}ects the stress response mediated by the developing hypothalamic!pituitary!adrenal "HPA# axis[29\22 As adults\ alcohol!exposed o}! spring appear to have a greater HPA activation than control o}spring in response to stress[08\18 This hyperresponsiveness is sex!dependent[ However\ although these results have typically been demonstrated only in female o}spring\ under certain stress conditions alcohol!exposed males show a higher corticosterone level for a much longer and more sustained period of time\ as compared to pair!fed and control male subjects[23 Chronic stress experienced early in life\ like prenatal alcohol exposure\ has also been demonstrated to alter the subsequent adult stress response[ Handling immediately after birth\ which involves separation from the dam\ reduces an animal|s later anxiety when placed in a stressful situation[ Rabbits handled in the _rst 09 days of life exhibit more activity in the open _eld test and spend more time examining novel objects\24 indicating reduced emotionality\ and consume less alcohol and exhibit shorter immobility time in a swim test[02 Handled rats also show a signi_cantly lowered plasma corticosterone level when exposed to stress as adults\11 as well as a faster initial response to stress[09\20\21 This experiment examined the e}ects of early stress on the subsequent adult neurosteroid and catecholaminergic stress responses to determine whether they may be di}erentially a}ected by prenatal condition or sex[ The indices of the stress response investigated were plasma and brain concentrations of allopregnanolone and regional concentrations in dopamine and norepinephrine[ The response of males and females were compared\ since the adult stress response di}ers by sex both physiologically and behaviorally[01\17\21\26 O}spring from prenatal alcohol exposed dams and their pair!fed controls were included because two previous studies in our laboratory revealed that prenatal alcohol exposure causes a decreased sensitivity to the neuromodulatory e}ects of neurosteroids[ Prenatal alcohol exposure decreased responsiveness to allopregnanolone in week!old o}spring when tested for distress ultrasonic vocalizations "USVs# after brief maternal separation[27 In addition\ prenatal alcohol exposure was found to eliminate the low dose anxiolytic e}ect and to amplify the higher dose anxiogenic e}ects of the neurosteroid pregnenolone on this same USV test[28 Since dysfunction in HPA axis stress responsiveness in alcohol!exposed o}spring is still evident as they mature\ we hypothesized that neurosteroid response to stress in prenatal alcohol exposed adult o}spring\ and its relationship to catecholaminergic systems\ might also be di}erentially a}ected by early postnatal stress experience[ EXPERIMENTAL PROCEDURES Prenatal conditions The subjects were bred in this laboratory from LongÐEvans hooded rats "Harlan SpragueÐ Dawley\ Indianapolis\ IN#[ Upon detection of a vaginal plug\ each pregnant female was assigned to
Prenatal experience and postnatal stress responses
108
one of three prenatal treatment groups ] control\ pair!fed or alcohol[ Dams in the control group had continuous access to lab chow and water throughout their pregnancies[ Dams in the pair!fed and alcohol groups were treated identically to the control groups on gestational days "GD# 0Ð4[ Starting on GD 5\ alcohol dams were given a liquid diet containing 5[6) ethanol v:v "Bioserv Liquid Diet F0154\ Bioserv Inc[\ Frenchtown\ NJ#[ This diet provided 24) of the total caloric content as ethanol[ Pair!fed dams "the nutritional control group# were given a similar liquid diet "Bioserv Liquid Diet F0153# except that the ethanol was replaced isocalorically with a maltoseÐ dextrin mixture[ Each dam in the pair!fed group was yoked to an alcohol dam ^ each pair!fed dam was paired to an alcohol dam\ and fed the same amount of diet as that consumed by the alcohol dam on a ml:kg body weight basis\ for each speci_c day of pregnancy[ Thus each yoked pair received the same relative volume of diet and the same number of calories on a body weight basis\ the only di}erence being the presence or absence of alcohol[ On GD 19\ liquid diets were replaced by continuous access to lab chow and water\ and the breeding cages were checked for births[ Postnatal conditions There were two postnatal treatment conditions ] separated and control groups[ On postnatal days "PN# 1Ð6\ two male and two female subjects from each litter were removed from the home nest and isolated for 5 h[ The isolation procedure was identical to that used in previous experiments on maternal separation ^39 subjects were placed individually in small plastic containers ~oated in a circulating\ heated water bath maintained at 23>C[ At the end of 5 h\ subjects were weighed and returned to the home nest[ Two male and two female subjects from the same litters served as intra! litter controls ^ these subjects were only weighed on PN 1Ð6[ Stress The forced swim test was used as the stressor for adult subjects[ This test has been shown to increase both the plasma corticosterone levels0 and the allopregnanolone levels in the brain and blood plasma[15 Subjects were tested at approximately 84 days of age "range from 79 to 009 days old#[ Subjects were placed in a container "27×12×23 cm# _lled with approximately 05 l of room temperature water "19>C#[ After 09 min\ the animals were removed from the water and immediately killed by decapitation[ Tissue preparation The brain was removed from the skull\ the cerebellum and brainstem removed and discarded\ and the remainder frozen in n!pentane "Fisher Scienti_c\ Cat[èUN0154# at −09>C[ The cerebral hemispheres were stored in a −79> freezer until dissection for chromatography and assay[ The plasma was obtained through the following procedure ] approximately 2 ml of trunk blood was collected immediately after decapitation in heparinized tubes "Microtainer with ammonium heparin\ Becton Dickinson\ Rutherford\ NJ#[ The blood was allowed to sit for approximately 09 min and was then centrifuged at 1999 rpm for 09 min[ The plasma from the three tubes was combined and spun again at 1999 rpm for 09 min[ Plasma was then stored at −79>C until shipped for analysis to P[ Moore\ Southwest Foundation for Biomedical Research\ San Antonio\ TX[ Samples were ana! lyzed for allopregnanolone levels with a radioimmunoassay "RIA# using ð2HŁ!allopregnanolone[ This method has been described in detail elsewhere[14 Non!stressed controls were not included in this experiment\ because baseline allopregnanolone concentrations are not detectable by RIA in the absence of a stressor[ The brains of the control and separated subjects from all three prenatal conditions were dissected on ice ^ the prefrontal cortex\ nucleus accumbens and striatum for high performance liquid chro! matography with electrochemical detection "HPLC# and the cortex and hippocampus for RIA analysis[ The prefrontal cortex\ nucleus accumbens and striatum were homogenized in 9[0 M perchloric acid containing 9[42 M sodium metabisul_te\ 1[58×09−3 Na1EDTA\ and 1[4×09−09 M dihydroxy benzylamine "DHBA# as an internal standard and frozen for future HPLC analysis[ The cortex and hippocampus were dissected and immediately frozen in liquid nitrogen and stored at −79>C until assayed by RIA with the same methods reported for the plasma above[
119
B[ Zimmerberg and R[ C[ Brown
HPLC preparation and running of samples The HPLC system consisted of a Waters 606 Autosampler\ a Waters 409 pump\ a Water 359 Electrochemical Detector and the Waters Systems Interface Module[ The column used was a Waters Plasma Catecholamine column\ containing C07 hydrocarbon chains bound to silica particles[ The prefrontal cortex\ nucleus accumbens and striatum samples were _rst homogenized in 9[0 M perchloric acid containing 9[42 M sodium metabisul_te\ 1[58×09−3 Na1EDTA\ and 1[4×09−09 M dihydroxy benzylamine "DHBA# as an internal standard using a sonic dismembranator "Fisher Scienti_c\ Model è449# set at mark 1 for 29 s "prefrontal cortex#\ 29 s "nucleus accumbens# and 0 min "striatum#[ The samples were then centrifuged at 03\999 rpm and 3>C for 29 min[ The super! natant was removed and _ltered through 9[1 mm _lters "Lida Manufacturing Corp[\ Catalogue è7492!99\ Kenosha\ WI# in the centrifuge at 1999 rpm and 3>C for 09 min[ The resulting _ltrate was placed in autosampler tubes and the samples were put in the autosampler at 3>C[ Samples were run with a mobile phase containing distilled water\ 6[4) methanol\ 1[34×09−2 M heptane sulfonic acid\ 9[94 M sodium monobasic phosphate and 1[04×09−3 M Na1EDTA at 9[4 ml:min for 34 min[ A standard containing a 0×09−6 M solution of norepinephrine\ epinephrine\ DHBA\ DOPAC\ dopamine and HVA "all purchased from Sigma Chemical Co[\ St Louis\ MO# was run between the samples from individual brains[ Individual neurotransmitter concentrations were determined using the Baseline chromatography workstation[ Data analysis Data "n − 5 per cell# were analyzed by analysis of variance with prenatal condition "alcohol\ pair! fed or lab chow#\ sex and postnatal treatment "separated or control# as the three independent variables[ Signi_cant main e}ects were subsequently analyzed by student NewmanÐKeuls tests with a P ³ 9[94 criterion[ Signi_cant interactions between factors were analyzed by post!hoc means comparison tests\ with a P ³ 9[94 criterion[ All analyses were performed using SuperAnova "Abacus Concepts Inc[\ Berkeley\ CA#[
RESULTS Allopre`nanolone The plasma data revealed signi_cant main e}ects of prenatal condition\ F"1\26# 2[66\ P ³ 9[94\ and of sex\ F"0\26#!0[60×09\1 P ³ 9[9990#[ Plasma samples were assayed only from the control animals of the three prenatal conditions due to the expense of the assay[ Pair!fed o}spring had higher blood plasma levels of allopregnanolone than prenatal alcohol!exposed or lab chow control animals\ which were not signi_cantly di}erent from each other[ Females had signi_cantly higher levels than males regardless of condition "see Fig[ 0#[ Cortical allopregnanolone concentrations were signi_cantly a}ected by all three factors\ as well as exhibiting signi_cant interactions between all of these factors[ The main e}ects were prenatal condition\ F"1\37# 5[77\ P ³ 9[90\ sex\ F"0\37# 044[19\ P ³ 9[9990\ and postnatal treatment\ F"0\37# 6[06\ P 9[90[ Females always showed signi_cantly more cortical allopregnanolone than did males[ One noteworthy two!way interaction was between prenatal condition and postnatal treatment\ F"1\37# 3[63\ P 9[91[ A prior postnatal separation history was associated with a greater allopregnanolone response to swim stress in alcohol subjects\ regardless of sex\ compared to both pair!fed and lab chow separated subjects\ which did not di}er from each other[ In contrast\ there were no di}erences among control subjects by prenatal condition when collapsed across postnatal treatment groups[ The signi_cant three!way interaction between prenatal condition\ sex and postnatal treatment\ F"1\37# 2[33\ P ³ 9[94 was further analyzed to reveal that the alcohol females had higher levels in the separated group than the pair!fed and lab chow separated females\ who did not di}er from each other[ The alcohol and lab chow control females showed similar levels\ and these were both higher than the pair!fed control females[ In both the male separated and control groups\ the pair!fed animals showed higher levels of allopregnanolone ^ the alcohol and lab chow groups were not signi_cantly di}erent for either postnatal treatment condition[ These results are shown in Fig[ 1[
Prenatal experience and postnatal stress responses
110
Fig[ 0[ Allopregnanolone concentrations "ng:ml2S[E[M[# in the plasma of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[
The results from the hippocampus show a very similar pattern of results\ with all factors showing signi_cant main e}ects ] Prenatal condition\ F"1\01# 09[57\ P ³ 9[90\ sex\ F"0\01# 039[01\ P ³ 9[9990 and postnatal treatment\ F"0\01# 02[67\ P ³ 9[90\ as well as all factors showing signi_cant interactions[ The signi_cant interaction between prenatal condition and postnatal treat! ment\ F"1\01# 6[76\ P ³ 9[90 was due to the same result as in the cortical analysis[ A prior postnatal separation history was associated with a greater allopregnanolone response to swim stress in adult alcohol subjects\ regardless of sex\ compared to both pair!fed and lab chow separated subjects\ which did not di}er from each other[ In contrast\ there were no di}erences among control subjects by prenatal condition when collapsed across postnatal treatment groups[ The signi_cant three!way interaction\ F"1\01# 4[32\ P 9[91\ was due to the _nding that while alcohol separated females were higher than both the pair!fed and lab chow separated females\ which were not signi_cantly di}erent from each other\ a parallel e}ect was not seen in alcohol males[ The distribution is similar to that seen in the cortex ^ all females again showed higher levels of allopregnanolone than the males\ and the pair!fed males in both the separated and control groups showed higher allopregnanolone than did the alcohol and lab chow males[ These results are shown in Fig[ 2[
111
B[ Zimmerberg and R[ C[ Brown
Fig[ 1[ Allopregnanolone concentrations "ng:g tissue2S[E[M[# in the prefrontal cortex of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[ Half of these subjects also had a postnatal stress experience\ daily maternal separation for 5 h from postnatal days 1 to 6[
Catecholamine concentrations Prenatal treatment had a signi_cant e}ect on the concentration of dopamine in the prefrontal cortex\ F"1\43# 2[07\ P ³ 9[94\ as shown in Fig[ 3[ The prenatal alcohol group had lower dopamine levels than either the pair!fed or control groups\ which did not di}er from each other[ There was also a signi_cant interaction between postnatal treatment and sex\ F"0\43# 5[01\ P ³ 9[91 on the concentration of dopamine in the prefrontal cortex ^ while postnatal separation did not e}ect dopamine levels in adult male subjects in the control groups\ female subjects in both the pair!fed and lab chow groups had higher levels than their non!separated comparable controls[ Although the three!way interaction did not reach signi_cance\ alcohol males from the separated condition had lower dopamine levels than any other male group\ while alcohol females did not exhibit increased dopamine with a postnatal stress history as did comparable separated female groups[ In the striatum\ there was signi_cant interaction between prenatal condition and postnatal treat! ment\ F"1\47# 5[64\ P ³ 9[90\ as seen in Fig[ 4[ While alcohol male and female subjects had higher dopamine levels in the striatum than either control group in the absence of a postnatal separation history\ this increased dopamine response to stress was not seen in the separated male and female
Prenatal experience and postnatal stress responses
112
Fig[ 2[ Allopregnanolone concentrations "ng:g tissue2S[E[M[# in the hippocampus of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[ Half of these subjects also had a postnatal stress experience\ daily maternal separation for 5 h from postnatal days 1 to 6[
groups[ There were no signi_cant e}ects of pre! or postnatal treatment or sex on the concentration of dopamine in the nucleus accumbens[ Analysis of norepinephrine concentrations in each of these same areas revealed some signi_cant e}ects[ In the prefrontal cortex\ there was a main e}ect of prenatal condition\ F"1\47# 7[48\ P ³ 9[990\ with all three groups di}ering signi_cantly from each other[ Norepinephrine con! centrations were lowest in the pair!fed groups "017[12200[87 mg:g tissue#\ next lowest in the alcohol groups "017[12200[86 mg:g tissue# and highest in the lab chow groups "048[88209[30 mg:g tissue# regardless of postnatal separation history[ In the nucleus accumbens\ in contrast\ only postnatal treatment had a signi_cant e}ect on norepinephrine levels\ F"1\47# 3[50\ P ³ 9[94[ All subjects with a separation history had higher norepinephrine concentrations "69[5322[88 mg:g tissue# that non!separated subjects "59[2721[09 mg:g tissue# regardless of prenatal treatment[ Finally\ there were no signi_cant e}ects of any factors on the concentration of norepinephrine in the striatum[ DISCUSSION Di}ering prenatal environments coupled with a postnatal experience manipulation did have an e}ect on the adult neurochemical response to an acute stressor\ although the e}ect varied among
113
B[ Zimmerberg and R[ C[ Brown
Fig[ 3[ Dopamine concentrations "mg:g tissue2S[E[M[# in prefrontal cortex of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[ Half of these subjects also had a postnatal stress experience\ daily maternal separation for 5 h from postnatal days 1 to 6[
neurochemical indices and brain areas examined[ The results from the allopregnanolone assays seem to re~ect a di}erence in the peripheral and the central nervous system response to stress[ The peripheral response "the plasma concentrations of allopregnanolone# was altered in the pair!fed group only[ This hyperresponsiveness to the swim stress may be due to the prenatal stress of the pair!feeding condition ^ since dams in the pair!fed condition have restricted feeding they tend to consume all of their non!alcohol liquid diet as soon as it is administered\ and have no food for the next 12 h\ while the alcohol dams {{sip|| their liquid diet throughout the 13 h\ and tubes are always re_lled when low[ The sex di}erence in the peripheral response "females × males# can be accounted for by the fact that females have higher circulating levels of progesterone in the blood\ due to its production in the ovaries\ and therefore may synthesize more allopregnanolone as needed in the adrenal gland\ which is the primary site of the peripheral stress response[ Enzymes responsible for the synthesis of allopregnanolone from cholesterol or progesterone have also been found in the smooth muscles\ platelets and the adrenal gland\ as well as in the testes and ovaries[ The central nervous system response\ interestingly\ showed the same distinct sex di}erence as seen in the plasma[ Female rats had higher cortical and hippocampal concentrations of allopregnanolone compared to both of the prenatal control groups[ This may signify that again that in females\ higher
Prenatal experience and postnatal stress responses
114
Fig[ 4[ Dopamine concentrations "mg:g tissue2S[E[M[# in striatum of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[ Half of these subjects also had a postnatal stress experience\ daily maternal separation for 5 h from postnatal days 1 to 6[
circulating levels of progesterone are available to the brain to be converted into allopregnanolone[ Higher levels of allopregnanolone in stressful situations would predict lower levels of anxiety in female rats compared to male rats ^ this sex di}erence has been reported in both an open _eld activity test17 and in the plus maze[26 A history of prior postnatal maternal separation was associated with increased concentrations of allopregnanolone in the cortex and hippocampus in adult o}spring prenatally exposed to alcohol\ but not to o}spring with control prenatal treatments[ This enhances allopregnanolone stress response was particularly evident in females compared to males[ The stress of maternal separation during the _rst week of life may have led to functional changes in the developing neurosteroid system that results in increased production or mobilization of allopregnanolone in the brain during the stress response later in life\ and therefore enables the animal to cope better with stress[ The change in the production of allopregnanolone in the brain should be mirrored by a decrease in prefrontal cortex dopamine concentrations\ since anxiolytics reduce the stress!induced activation of the mesolimbocortical dopamine pathway[00 The methodology used in this experiment does not determine directly the activity of catecholaminergic pathways*further experiments using mic! rodialysis or assessment of metabolites are necessary to ascertain utilization[ However\ there does
115
B[ Zimmerberg and R[ C[ Brown
appear to be evidence that pre! and postnatal treatments can have long!term consequences on the baseline level of activity "e[g[ synthesis or metabolism# in these catecholaminergic systems[ The alcohol group exhibited lower levels of dopamine in the prefrontal cortex when combined with a prior maternal separation treatment[ Dopamine neurons project from the ventral tegmental area "VTA# to the prefrontal cortex and the nucleus accumbens\ and from the substantia nigra "SN# to the striatum[4\7 These projections are innervated by inhibitory\ GABA releasing neurons[ Dopamine released in the prefrontal cortex is excitatory\ and results in the activation of more GABA neurons\ which have inhibitory e}ects on the striatum and the nucleus accumbens through two glutamate pathways projecting from the prefrontal cortex to these areas[ The dopamine released in the nucleus accumbens and the striatum can be excitatory or inhibitory\ due to the diversity of dopamine receptors[ The decrease of dopamine released in the prefrontal cortex in the alcohol animals may be due to the inhibitory e}ects of allopregnanolone[ Since allopregnanolone is acting at the GABAA receptor\ it may be acting either at the source of the pathway\ in the VTA\ or in the prefrontal cortex itself[ The decrease in prefrontal cortical levels of dopamine suggests that allopregnanolone is inhibiting activation of the pathway by acting at GABAA receptors in the VTA\ although further studies are needed to determine this hypothesis directly[ Interestingly\ there was a sex di}erence in the response to postnatal experience[ Dopamine levels in the prefrontal cortex in the two prenatal control groups were elevated in females who had been separated from their dams during the _rst week or life\ but not in the males[ This increased dopamine response to swim stress may re~ect an underlying sex di}erence in sensitization of these dopamine pathways\ and requires further investigation[ The alcohol e}ect could be due to an upregulation of dopamine receptors in response to a developmental delay in dopamine release controlled by the VTA[ There is pharmacological evidence that there are fewer GABA receptors in one!week!old rats pups prenatally exposed to alcohol than there are in controls[27\28 The pattern of dopamine concentration was di}erent in the nucleus accumbens and the striatum[ In the striatum\ prenatal and postnatal experiences again had a combined impact on dopamine concentrations[ Only in the non!separated alcohol animals\ regardless of sex\ was there an increase in dopamine compared to the two prenatal control groups[ That would indicate that although {{typical|| prenatal alcohol!exposed animals would appear to be hyper!responsive to stress when tested as adults\ this hyper!responsiveness is not detectable in subjects who have had a prior early stress experience[ In contrast\ there were no e}ects of any manipulations on dopamine concentrations in the nucleus accumbens[ The results obtained for the concentrations of norepinephrine in the brain sections examined provide some preliminary suggestions about the e}ects of pre! and postnatal treatments on the noradrenergic response to stress[ While the pair!fed o}spring\ whose dams experienced the stress of restricted feeding\ had the lowest levels of norepinephrine in the prefrontal cortex\ prenatal alcohol exposed o}spring had a partial amelioration of this e}ect\ having lower norepinephrine levels than controls\ but not as low as pair!fed[ None of the prenatal condition groups had any additional e}ects of postnatal experience on norepinephrine levels in the prefrontal cortex[ In contrast\ all groups experiencing postnatal maternal separation had higher levels of norepinephrine in the nucleus accumbens when exposed to swim stress as adults[ No e}ects of pre! or postnatal manipulations were evident in striatal norepinephrine concentrations[ Studies have shown that norepinephrine is involved in the stress response\ and that a noradrenergic pathway projecting from the locus coeruleus to the prefrontal cortex\ nucleus accumbens and striatum is responsive to stress\ particularly the projection to the prefrontal cortex[3 This would lead us to predict that prenatal alcohol exposure would interact with postnatal stressors to increase norepinephrine release in the prefrontal cortex\ and perhaps in the nucleus accumbens and striatum as well[ If increased release or synthesis of allopregnanolone modulated "via its e}ects on GABAA receptors# these noradrenergic pathways\ there would be a decrease in the amount of norepinephrine released in the stress response[ This is the trend seen in the prefrontal cortex[ The alcohol and pair! fed animals have lower levels of norepinephrine than the lab chows[ There is no e}ect of postnatal condition\ therefore it may be that the noradrenergic projections to the prefrontal cortex appear very early in brain development and are not a}ected by the postnatal treatment[ Since norepinephrine levels are a}ected in all groups by postnatal\ but not prenatal\ stress\ a later time of development of this structure may be responsible for this contrasting result[ Cenci et al[3 reported that not all
Prenatal experience and postnatal stress responses
116
stressors activated the noradrenergic pathways ] the pathway to the prefrontal cortex was the most responsive\ while those to the nucleus accumbens were less so[ At the time of testing\ there were no di}erences in body weight between pre! and postnatal treatment groups[ Body weight di}erences were apparent on postnatal day 7 ] all separated subjects\ regardless of prenatal condition\ weighed less than controls[ However\ by postnatal day 10 the percent weight gain was similar in all groups\ suggesting that there were no long!term consequences on feeding[ It is possible\ though\ that the weight loss itself\ and not the maternal separation\ was responsible for the long!term sequela reported here[ In summary\ these results suggest that a variety of stressful experiences during the period of critical brain development may a}ect how an individual deals with stressors later in life[ The role of neurosteroids in modulating the stress response may be complex and sex!dependent\ but future research appears warranted to more fully understand their role as a possible endogenous negative feedback system[ Although these results do not directly measure activity\ they do provide a sugges! tion that there may be a reciprocal relationship between neurosteroid modulation of the GABAA receptor and the catecholaminergic systems those GABA neurons inhibit in stress!responsive areas of the brain\ and that there may be long!term consequences when stress is experienced during pre! and postnatal stages of neuronal development and maturation[ Acknowled`ements*This research was supported by a grant from the National Institute on Alcohol Abuse and Alcoholism\ èR90 AA97594[ We thank Jacqueline M[ Weider for her excellent technical assistance[ We also thank Perry Moore\ Southwest Foundation for Biomedical Research\ San Antonio\ TX for his analysis of allopregnanolone concentrations[
REFERENCES 0[ Abel\ E[ L[\ Physiological correlates of the forced swim test in rats[ Physiol[ Behav[\ 0882\ 43\ 298Ð206[ 1[ Bitran\ D[\ Purdy\ R[ H[ and Kellogg\ C[ K[\ Anxiolytic e}ect of progesterone is associated with increases in cortical allopregnanolone and GABAA receptor function[ Pharmacol[ Biochem[ Behav[\ 0882\ 34\ 312Ð317[ 2[ Bitran\ D[\ Hilvers\ R[ J[ and Kellogg\ C[ K[\ Anxiolytic e}ects of 2a!hydroxy!4aðbŁ!pregnan!19!one ] endogenous metabolites of progesterone that are active at the GABAA receptor[ Brain Res[\ 0880\ 450\ 046Ð050[ 3[ Cenci\ M[\ Kalen\ A[ P[\ Mandel\ R[ J[ and Bjorklund\ A[\ Regional di}erences in the regulation of noradrenaline release in medial frontal cortex\ nucleus accumbens and caudate!putamen ] a microdialysis study in the rat[ Brain Res[\ 0881\ 470\ 106Ð117[ 4[ Cuello\ A[ C[\ Priestly\ J[ V[ and Paxinos\ G[\ Substance P and enkephalin containing pathways[ In The Rat Nervous System\ Vol[ 0\ Forebrain and Midbrain\ ed[ G[ Paxinos[ Academic Press Inc[\ Orlando\ 0874\ pp[ 330Ð353[ 5[ Deutsch\ S[ I[\ Mastropaolo\ J[ and Hitri\ A[\ GABA!active steroids ] endogenous modulators of GABA!gated chloride ion conductance[ Clin[ Neuropharmacol[\ 0881\ 04\ 241Ð253[ 6[ Fadda\ F[\ Mosca\ E[\ Ni}ol\ T[\ Colombo\ G[ and Gessa\ G[ L[\ Ethanol prevents stress!induced increase in cortical DOPAC ] reversal by RO 04!3402[ Physiol[ Behav[\ 0876\ 39\ 272Ð274[ 7[ Fallon\ J[ H[ and Loughlin\ S[ E[\ Substantia Nigra[ The Rat Nervous System\ Vol[ 0\ Forebrain and Midbrain\ ed[ G[ Paxinos[ Academic Press Inc[\ Orlando\ 0874\ pp[ 242Ð263[ 8[ Finn\ D[ A[ and Gee\ K[ W[\ The in~uence of estrus cycle on neurosteroid potency at the g!aminobutyric acidA receptor complex[ J[ Pharmacol[ Exp[ Ther[\ 0882\ 154\ 0263Ð0268[ 09[ Gandleman\ R[\ Psychobiolo`y of Behavioral Development[ Oxford University Press\ New York\ 0881\ pp[ 057Ð070[ 00[ Grobin\ A[ C[\ Roth\ R[ H[ and Deutch\ A[ Y[\ Regulation of the prefrontal cortical dopamine system by the neuroactive steroid 2a\10!dihydroxy!4a!pregnane!19!one[ Brain Res[\ 0881\ 467\ 240Ð245[ 01[ Hennessey\ M[ B[ and Weinberg\ J[\ Adrenocortical activity during conditions of brief social separation in preweaning rats[ Behav[ Neural Biol[\ 0889\ 43\ 31!44[ 02[ Hilakivi!Clarke\ L[ A[\ Turkka\ J[\ Lister\ R[ G[ and Linnoila\ M[\ E}ects of early postnatal handling on brain b! adrenoreceptors and behavior in tests related to stress[ Brain Res[\ 0880\ 431\ 175Ð181[ 03[ Lan\ N[ C[\ Gee\ K[ W[\ Bolger\ M[ B[ and Chen\ J[ S[\ Di}erential responses of expressed recombinant human g! aminobutyric acidA receptors to neurosteroids[ J[ Neurochem[\ 0880\ 46\ 0707Ð0710[ 04[ Majewska\ M[ D[\ Harrison\ N[ L[\ Schwartz\ R[ D[\ Barker\ J[ L[ and Paul\ S[ M[\ Steroid hormone metabolites are barbiturate!like modulators of the GABAA receptor[ Science\ 0875\ 121\ 0993Ð0996[ 05[ McEwen\ B[ S[\ Steroid hormones ] e}ect on brain development and function[ Hormone Res[\ 0881\ 26\ 0Ð09[ 06[ Morrow\ A[ L[\ Pace\ J[ R[\ Purdy\ R[ H[ and Paul\ S[ M[\ Characterization of steroid interaction with g!aminobutyric acid receptor!gated chloride ion channels ] evidence for multiple steroid recognition sites[ Molec[ Pharmacol[\ 0889\ 26\ 152Ð169[ 07[ Myslobodsky\ M[ S[\ Pro!and anticonvulsant e}ects of stress ] the role of neuroactive steroids[ Neurosci[ Biobehav[ Rev[\ 0882\ 06\ 018Ð028[ 08[ Nelson\ L[ R[\ Taylor\ A[ N[\ Lewis\ J[ W[\ Poland\ R[ E[\ Redei\ E[ and Branch\ B[ J[\ Pituitary!adrenal responses to morphine and footshock stress are enhanced following prenatal alcohol exposure[ Alcohol Clin[ Exp[ Res[\ 0875\ 09\ 286Ð391[ 19[ Orchinik\ M[ and McEwen\ B[\ Novel and classical actions of neuroactive steroids[ Neurotransmissions\ 0882\ 8\ 0Ð5[ 10[ Paul\ S[ M[ and Purdy\ R[ H[\ Neuroactive steroids[ The FASEB Journal\ 0881\ 5\ 1200Ð1211[ 11[ P_efer\ W[ D[\ Rotundo\ R[\ Myers\ M[ and Denenberg\ V[ H[\ Stimulation in infancy ] unique e}ects of handling[ Physiol[ Behav[\ 0865\ 06\ 670Ð673[
117
B[ Zimmerberg and R[ C[ Brown
12[ Prince\ R[ J[ and Simmonds\ M[ A[\ Di}erential antagonism by epipregnanolone of alphaxalone and pregnenolone potentiation of ð2HŁ~unitrazepam binding suggests more than one class of binding site for steroids at GABAA receptors[ Neuropharmacol[\ 0882\ 21\ 48Ð52[ 13[ Puia\ G[\ Santi\ M[\ Vicini\ S[\ Pritchett\ D[ B[\ Purdy\ R[ H[\ Paul\ S[ M[\ Seeburg\ P[ H[ and Costa\ E[\ Neurosteroids act on recombinant human GABAA receptors[ Neuron\ 0889\ 3\ 648Ð654[ 14[ Purdy\ R[ H[\ Moore\ P[ H[\ Rao\ P[ N[\ Hagino\ N[\ Yamaguchi\ T[\ Schmidt\ P[\ Rubinow\ D[ R[\ Morrow\ A[ L[ and Paul\ S[ M[\ Radioimmunoassay of 2a!hydroxy!4a!19!one in rat and human plasma[ Steroids\ 0889\ 44\ 189Ð185[ 15[ Purdy\ R[ H[\ Morrow\ A[ L[\ Moore\ P[ H[ and Paul\ S[ M[\ Stress!induced elevations of g!aminobutyric acid type A receptor!active steroids in the rat brain[ Neurobiol[\ 0880\ 77\ 3442Ð3446[ 16[ Roth\ R[ H[\ Tam\ S[!Y[\ Ida\ Y[\ Yang\ J[!X[ and Deutch\ A[ Y[\ Stress and the mesocorticolimbic dopamine systems[ Annals of the New York Academy of Science\ 0877\ 446\ 027Ð036[ 17[ Stewart\ J[ and Cyrgan\ D[\ Ovarian hormones act early in development to feminize adult open!_eld behavior in the rat[ Hormones Behav[\ 0879\ 03\ 19Ð21[ 18[ Taylor\ A[ N[\ Branch\ B[ J[\ Lin\ S[ H[ and Kokka\ N[\ Long!term e}ects of fetal ethanol exposure on pituitary!adrenal response to stress[ Pharm[ Biochem[ Beh[\ 0871\ 05\ 474Ð478[ 29[ Taylor\ A[ N[\ Branch\ B[ J[\ Nelson\ L[ R[\ Lane\ L[ A[ and Poland\ R[ E[\ Prenatal ethanol and ontogeny of pituitary! adrenal responses to ethanol and morphine[ Alcohol\ 0875\ 2\ 144Ð148[ 20[ Walshlak\ A[ and Weinstock\ M[\ Neonatal handling reverses behavioral abnormalities induced in rats by prenatal stress[ Physiol[ Behav[\ 0889\ 37\ 178Ð181[ 21[ Weinberg\ J[ and Levine\ S[\ Early handling in~uences on behavioral and physiological responses during active avoidance[ Devel[ Psychobiol[\ 0866\ 09\ 050Ð058[ 22[ Weinberg\ J[\ Prenatal ethanol exposure alters adrenocortical development of o}spring[ Alcohol Clin[ Exp[ Res[\ 0878\ 02\ 62Ð72[ 23[ Weinberg\ J[\ Prenatal ethanol e}ects ] sex di}erences in o}spring stress responsiveness[ Alcohol\ 0881\ 8\ 108Ð112[ 24[ Wyly\ M[ V[\ Denenberg\ V[ H[\ De Santis\ D[\ Burns\ J[ K[ and Zarrow\ M[ X[\ Handling rabbits in infancy ] in search of a critical period[ Developmental Psychobiol[\ 0864\ 7\ 068Ð075[ 25[ Zimmerberg\ B[\ Brunelli\ S[ A[ and Hofer\ M[ A[\ Reduction of rat pup ultrasonic vocalizations by the neuroactive steroid allopregnanolone[ Pharmacol[ Biochem[ Behav[\ 0883\ 36\ 624Ð627[ 26[ Zimmerberg\ B[ and Farley\ M[ J[\ Sex di}erences in anxiety behavior in rats ] role of gonadal hormones[ Physiol[ Behav[\ 0882\ 43\ 0008Ð0013[ 27[ Zimmerberg\ B[\ Drucker\ P[ A[ and Weider\ J[ M[\ Di}erential behavioral e}ects of the neuroactive steroid allo! pregnanolone on neonatal rats prenatally exposed to alcohol[ Pharmacol[ Biochem[ Behav[\ 0884\ 40\ 352Ð357[ 28[ Zimmerberg\ B[ and McDonald\ B[ C[\ Prenatal alcohol!exposure in~uences the e}ects of neuroactive steroids on separation!induced ultrasonic vocalizations in rat pups[ Pharmacol[ Biochem[ Behav[\ 0885\ 44\ 430Ð436[ 39[ Zimmerberg\ B[ and Shartrand\ A[ M[\ Temperature!dependent e}ects of maternal separations on growth\ activity\ and amphetamine sensitivity in the rat[ Dev[ Psycho[ Biol[\ 0881\ 14\ 102Ð115[
Pergamon
Int[ J[ Devl Neuroscience\ Vol[ 05\ No[ 2:3\ pp[ 106Ð117\ 0887 Þ 0887 ISDN[ Published by Elsevier Science Ltd All rights reserved[ Printed in Great Britain 9625Ð4637:87 ,08[99¦9[99
PII ] S9625Ð4637"87#99913Ð9
PRENATAL EXPERIENCE AND POSTNATAL STRESS MODULATE THE ADULT NEUROSTEROID AND CATECHOLAMINERGIC STRESS RESPONSES BETTY ZIMMERBERG and RACHEL C[ BROWN Department of Psychology and Program in Neuroscience\ Bronfman Science Center\ Williams College\ Williamstown\ MA 90156\ U[S[A[ "Received 19 October 0886 ^ accepted in revised form 00 March 0887 # Abstract*Allopregnanolone "2a!hydroxy!4a!regnan!19!one# is a neuroactive steroid recently shown to be involved in the neurochemical stress response via its positive modulation of the GABAA receptor complex[ This experiment investigated the e}ects of postnatal stress "daily maternal separation during the _rst week of life# on the subsequent adult response to a stressor "09 min forced swim# in LongÐEvans rats from one of three prenatal treatment groups "alcohol\ pair!fed and control#[ Indices of stress response were allopregnanolone concentrations in plasma\ cortex and hippocampus\ and dopamine and norepinephrine concentrations in prefrontal cortex\ nucleus accumbens and striatum[ Females had higher levels of allo! pregnanolone than males in both plasma and brain[ Prenatal alcohol exposure combined with early maternal separation stress resulted in an increase in the endogenous levels of allopregnanolone in the prefrontal cortex and hippocampus of adult o}spring in response to a stressor compared to subjects without a prior history of postnatal stress ^ this e}ect was greater in females[ This increased allopregnanolone was also associated with decreased dopamine and norepinephrine levels in the prefrontal cortex[ In the prenatal alcohol!exposed o}spring\ postnatal maternal separation blunted the increase in dopamine levels in the striatum seen in both control groups[ Postnatal maternal separation increased norepinephrine levels in the nucleus accumbens regardless of prenatal experience\ while in the prefrontal cortex only prenatal diet condition "pair!feeding and alcohol# resulted in lower norepinephrine levels[ The results of this experiment suggest that experience\ both pre! and postnatal\ can have long!term consequences for the developing neurochemical responses to stressors[ Þ 0887 ISDN[ Published by Elsevier Science Ltd Key words] allopregnanolone\ neuroactive steroid\ fetal alcohol syndrome\ dopamine\ sex di}erences\ maternal separation\ prenatal stress\ alcohol[
Neurochemical stress responses are the net outcome of a complex interaction between genetically determined individual di}erences and the environmental experiences that an individual encounters as the brain matures[ Understanding how the newly characterized neuroactive steroids a}ect the stress reaction may help to explain how the environment can modulate developing neural systems involved in the stress response[ These neuroactive steroids act as neuromodulators\ binding to membrane receptors and modulating ligand!gated ion channel conductance[05 The sedative and anxiolytic e}ects of neuroactive steroids appear to be due to their positive allosteric modulation of GABAA receptor[5\03\06 The GABAA receptor contains a distinct "e[g[ non!benzodiazepine or barbiturate# binding site for steroids\04\19\10 if not multiple sites[06\12 Neuroactive steroids have been found in various areas of the brain\ with relatively higher concentrations in the cerebral cortex\1\07\10 cerebellum\ hippocampus\ striatum8 and hypothalamus[15 Allopregnanolone\ a 2a!hydroxy!4a reduced metabolite of the steroid hormone progesterone\ has been shown to be a positive allosteric modulator of the GABAA receptor10 with a potency of 0999! fold higher than that of pentobarbital[5 Allopregnanolone inhibits the binding of ð24SŁ!TBPS\ which binds to and labels a site close to or on the Cl!channel04 at submicromolar concentrations[5 The potency of allopregnanolone for displacing ð24SŁ!TBPS increases in the presence of low con! centrations of GABA[5 Allopregnanolone has been shown to potentiate chloride ion currents in rat hippocampal neurons\ and has also been shown to be more e}ective than ~unitrazepam\ a benzodiazepine\ in potentiating a response in certain recombinant receptors\ speci_cally those made up of a!0\ b!0 and g!1 subunits\ suggesting a separate binding site[13 Plasma concentrations of allopregnanolone increase after exposure to a forced swim test\ and
To whom all correspondence should be addressed[ Tel[ ] 990 302 486 1335 ^ Fax ] 990 302 486 3005 ^ E!Mail ] Betty[ZimmerbergÝwilliams[edu 106
107
B[ Zimmerberg and R[ C[ Brown
this increase mirrors the increase over a period of 29Ð59 min in the cerebral cortex and hypothala! mus[15 When allopregnanolone is given to female rats\ they show an increase in open arm entries on an elevated plus maze\ indicating decreased anxiety[1\2 Similar anxiolytic e}ects of progesterone in this paradigm have been determined to be due to the hormone|s conversion into allop! regnanolone[1 In this laboratory\ intracerebroventricular injections of allopregnanolone in infant rats separated from dams and littermates were found to decrease the number of distress vocalizations made by the pups after a brief maternal separation[25 Thus there is strong evidence to support the hypothesis that allopregnanolone can act as an endogenous neurosteroid to reduce anxiety in response to stress\ in both adult rats and neonates[ Some neuroactive steroids have been shown to speci_cally in~uence the mesolimbocortical dopa! mine pathway\ a system critical in the neurochemical stress response[ Intraventricular administration of alloTHDOC\ an anxiolytic neuroactive steroid similar to allopregnanolone\ prior to exposure to 14 min of restraint stress resulted in a dose!dependent decrease in the concentration of dopamine metabolites in the prefrontal cortex ^ no signi_cant changes were seen in the nucleus accumbens or striatum[00 Other drugs a}ecting the GABAA receptor complex "e[g[ benzodiazepines and ethanol# block activation of cortical dopamine systems by footshock\6\16 suggesting that the ventral tegmental dopamine neurons projecting to the prefrontal cortex are modulated by GABA receptors[16 The role of neuroactive steroids in stress response after chronic stress has not yet been studied[ Animal models have shown that prenatal exposure to alcohol a}ects the stress response mediated by the developing hypothalamic!pituitary!adrenal "HPA# axis[29\22 As adults\ alcohol!exposed o}! spring appear to have a greater HPA activation than control o}spring in response to stress[08\18 This hyperresponsiveness is sex!dependent[ However\ although these results have typically been demonstrated only in female o}spring\ under certain stress conditions alcohol!exposed males show a higher corticosterone level for a much longer and more sustained period of time\ as compared to pair!fed and control male subjects[23 Chronic stress experienced early in life\ like prenatal alcohol exposure\ has also been demonstrated to alter the subsequent adult stress response[ Handling immediately after birth\ which involves separation from the dam\ reduces an animal|s later anxiety when placed in a stressful situation[ Rabbits handled in the _rst 09 days of life exhibit more activity in the open _eld test and spend more time examining novel objects\24 indicating reduced emotionality\ and consume less alcohol and exhibit shorter immobility time in a swim test[02 Handled rats also show a signi_cantly lowered plasma corticosterone level when exposed to stress as adults\11 as well as a faster initial response to stress[09\20\21 This experiment examined the e}ects of early stress on the subsequent adult neurosteroid and catecholaminergic stress responses to determine whether they may be di}erentially a}ected by prenatal condition or sex[ The indices of the stress response investigated were plasma and brain concentrations of allopregnanolone and regional concentrations in dopamine and norepinephrine[ The response of males and females were compared\ since the adult stress response di}ers by sex both physiologically and behaviorally[01\17\21\26 O}spring from prenatal alcohol exposed dams and their pair!fed controls were included because two previous studies in our laboratory revealed that prenatal alcohol exposure causes a decreased sensitivity to the neuromodulatory e}ects of neurosteroids[ Prenatal alcohol exposure decreased responsiveness to allopregnanolone in week!old o}spring when tested for distress ultrasonic vocalizations "USVs# after brief maternal separation[27 In addition\ prenatal alcohol exposure was found to eliminate the low dose anxiolytic e}ect and to amplify the higher dose anxiogenic e}ects of the neurosteroid pregnenolone on this same USV test[28 Since dysfunction in HPA axis stress responsiveness in alcohol!exposed o}spring is still evident as they mature\ we hypothesized that neurosteroid response to stress in prenatal alcohol exposed adult o}spring\ and its relationship to catecholaminergic systems\ might also be di}erentially a}ected by early postnatal stress experience[ EXPERIMENTAL PROCEDURES Prenatal conditions The subjects were bred in this laboratory from LongÐEvans hooded rats "Harlan SpragueÐ Dawley\ Indianapolis\ IN#[ Upon detection of a vaginal plug\ each pregnant female was assigned to
Prenatal experience and postnatal stress responses
108
one of three prenatal treatment groups ] control\ pair!fed or alcohol[ Dams in the control group had continuous access to lab chow and water throughout their pregnancies[ Dams in the pair!fed and alcohol groups were treated identically to the control groups on gestational days "GD# 0Ð4[ Starting on GD 5\ alcohol dams were given a liquid diet containing 5[6) ethanol v:v "Bioserv Liquid Diet F0154\ Bioserv Inc[\ Frenchtown\ NJ#[ This diet provided 24) of the total caloric content as ethanol[ Pair!fed dams "the nutritional control group# were given a similar liquid diet "Bioserv Liquid Diet F0153# except that the ethanol was replaced isocalorically with a maltoseÐ dextrin mixture[ Each dam in the pair!fed group was yoked to an alcohol dam ^ each pair!fed dam was paired to an alcohol dam\ and fed the same amount of diet as that consumed by the alcohol dam on a ml:kg body weight basis\ for each speci_c day of pregnancy[ Thus each yoked pair received the same relative volume of diet and the same number of calories on a body weight basis\ the only di}erence being the presence or absence of alcohol[ On GD 19\ liquid diets were replaced by continuous access to lab chow and water\ and the breeding cages were checked for births[ Postnatal conditions There were two postnatal treatment conditions ] separated and control groups[ On postnatal days "PN# 1Ð6\ two male and two female subjects from each litter were removed from the home nest and isolated for 5 h[ The isolation procedure was identical to that used in previous experiments on maternal separation ^39 subjects were placed individually in small plastic containers ~oated in a circulating\ heated water bath maintained at 23>C[ At the end of 5 h\ subjects were weighed and returned to the home nest[ Two male and two female subjects from the same litters served as intra! litter controls ^ these subjects were only weighed on PN 1Ð6[ Stress The forced swim test was used as the stressor for adult subjects[ This test has been shown to increase both the plasma corticosterone levels0 and the allopregnanolone levels in the brain and blood plasma[15 Subjects were tested at approximately 84 days of age "range from 79 to 009 days old#[ Subjects were placed in a container "27×12×23 cm# _lled with approximately 05 l of room temperature water "19>C#[ After 09 min\ the animals were removed from the water and immediately killed by decapitation[ Tissue preparation The brain was removed from the skull\ the cerebellum and brainstem removed and discarded\ and the remainder frozen in n!pentane "Fisher Scienti_c\ Cat[èUN0154# at −09>C[ The cerebral hemispheres were stored in a −79> freezer until dissection for chromatography and assay[ The plasma was obtained through the following procedure ] approximately 2 ml of trunk blood was collected immediately after decapitation in heparinized tubes "Microtainer with ammonium heparin\ Becton Dickinson\ Rutherford\ NJ#[ The blood was allowed to sit for approximately 09 min and was then centrifuged at 1999 rpm for 09 min[ The plasma from the three tubes was combined and spun again at 1999 rpm for 09 min[ Plasma was then stored at −79>C until shipped for analysis to P[ Moore\ Southwest Foundation for Biomedical Research\ San Antonio\ TX[ Samples were ana! lyzed for allopregnanolone levels with a radioimmunoassay "RIA# using ð2HŁ!allopregnanolone[ This method has been described in detail elsewhere[14 Non!stressed controls were not included in this experiment\ because baseline allopregnanolone concentrations are not detectable by RIA in the absence of a stressor[ The brains of the control and separated subjects from all three prenatal conditions were dissected on ice ^ the prefrontal cortex\ nucleus accumbens and striatum for high performance liquid chro! matography with electrochemical detection "HPLC# and the cortex and hippocampus for RIA analysis[ The prefrontal cortex\ nucleus accumbens and striatum were homogenized in 9[0 M perchloric acid containing 9[42 M sodium metabisul_te\ 1[58×09−3 Na1EDTA\ and 1[4×09−09 M dihydroxy benzylamine "DHBA# as an internal standard and frozen for future HPLC analysis[ The cortex and hippocampus were dissected and immediately frozen in liquid nitrogen and stored at −79>C until assayed by RIA with the same methods reported for the plasma above[
119
B[ Zimmerberg and R[ C[ Brown
HPLC preparation and running of samples The HPLC system consisted of a Waters 606 Autosampler\ a Waters 409 pump\ a Water 359 Electrochemical Detector and the Waters Systems Interface Module[ The column used was a Waters Plasma Catecholamine column\ containing C07 hydrocarbon chains bound to silica particles[ The prefrontal cortex\ nucleus accumbens and striatum samples were _rst homogenized in 9[0 M perchloric acid containing 9[42 M sodium metabisul_te\ 1[58×09−3 Na1EDTA\ and 1[4×09−09 M dihydroxy benzylamine "DHBA# as an internal standard using a sonic dismembranator "Fisher Scienti_c\ Model è449# set at mark 1 for 29 s "prefrontal cortex#\ 29 s "nucleus accumbens# and 0 min "striatum#[ The samples were then centrifuged at 03\999 rpm and 3>C for 29 min[ The super! natant was removed and _ltered through 9[1 mm _lters "Lida Manufacturing Corp[\ Catalogue è7492!99\ Kenosha\ WI# in the centrifuge at 1999 rpm and 3>C for 09 min[ The resulting _ltrate was placed in autosampler tubes and the samples were put in the autosampler at 3>C[ Samples were run with a mobile phase containing distilled water\ 6[4) methanol\ 1[34×09−2 M heptane sulfonic acid\ 9[94 M sodium monobasic phosphate and 1[04×09−3 M Na1EDTA at 9[4 ml:min for 34 min[ A standard containing a 0×09−6 M solution of norepinephrine\ epinephrine\ DHBA\ DOPAC\ dopamine and HVA "all purchased from Sigma Chemical Co[\ St Louis\ MO# was run between the samples from individual brains[ Individual neurotransmitter concentrations were determined using the Baseline chromatography workstation[ Data analysis Data "n − 5 per cell# were analyzed by analysis of variance with prenatal condition "alcohol\ pair! fed or lab chow#\ sex and postnatal treatment "separated or control# as the three independent variables[ Signi_cant main e}ects were subsequently analyzed by student NewmanÐKeuls tests with a P ³ 9[94 criterion[ Signi_cant interactions between factors were analyzed by post!hoc means comparison tests\ with a P ³ 9[94 criterion[ All analyses were performed using SuperAnova "Abacus Concepts Inc[\ Berkeley\ CA#[
RESULTS Allopre`nanolone The plasma data revealed signi_cant main e}ects of prenatal condition\ F"1\26# 2[66\ P ³ 9[94\ and of sex\ F"0\26#!0[60×09\1 P ³ 9[9990#[ Plasma samples were assayed only from the control animals of the three prenatal conditions due to the expense of the assay[ Pair!fed o}spring had higher blood plasma levels of allopregnanolone than prenatal alcohol!exposed or lab chow control animals\ which were not signi_cantly di}erent from each other[ Females had signi_cantly higher levels than males regardless of condition "see Fig[ 0#[ Cortical allopregnanolone concentrations were signi_cantly a}ected by all three factors\ as well as exhibiting signi_cant interactions between all of these factors[ The main e}ects were prenatal condition\ F"1\37# 5[77\ P ³ 9[90\ sex\ F"0\37# 044[19\ P ³ 9[9990\ and postnatal treatment\ F"0\37# 6[06\ P 9[90[ Females always showed signi_cantly more cortical allopregnanolone than did males[ One noteworthy two!way interaction was between prenatal condition and postnatal treatment\ F"1\37# 3[63\ P 9[91[ A prior postnatal separation history was associated with a greater allopregnanolone response to swim stress in alcohol subjects\ regardless of sex\ compared to both pair!fed and lab chow separated subjects\ which did not di}er from each other[ In contrast\ there were no di}erences among control subjects by prenatal condition when collapsed across postnatal treatment groups[ The signi_cant three!way interaction between prenatal condition\ sex and postnatal treatment\ F"1\37# 2[33\ P ³ 9[94 was further analyzed to reveal that the alcohol females had higher levels in the separated group than the pair!fed and lab chow separated females\ who did not di}er from each other[ The alcohol and lab chow control females showed similar levels\ and these were both higher than the pair!fed control females[ In both the male separated and control groups\ the pair!fed animals showed higher levels of allopregnanolone ^ the alcohol and lab chow groups were not signi_cantly di}erent for either postnatal treatment condition[ These results are shown in Fig[ 1[
Prenatal experience and postnatal stress responses
110
Fig[ 0[ Allopregnanolone concentrations "ng:ml2S[E[M[# in the plasma of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[
The results from the hippocampus show a very similar pattern of results\ with all factors showing signi_cant main e}ects ] Prenatal condition\ F"1\01# 09[57\ P ³ 9[90\ sex\ F"0\01# 039[01\ P ³ 9[9990 and postnatal treatment\ F"0\01# 02[67\ P ³ 9[90\ as well as all factors showing signi_cant interactions[ The signi_cant interaction between prenatal condition and postnatal treat! ment\ F"1\01# 6[76\ P ³ 9[90 was due to the same result as in the cortical analysis[ A prior postnatal separation history was associated with a greater allopregnanolone response to swim stress in adult alcohol subjects\ regardless of sex\ compared to both pair!fed and lab chow separated subjects\ which did not di}er from each other[ In contrast\ there were no di}erences among control subjects by prenatal condition when collapsed across postnatal treatment groups[ The signi_cant three!way interaction\ F"1\01# 4[32\ P 9[91\ was due to the _nding that while alcohol separated females were higher than both the pair!fed and lab chow separated females\ which were not signi_cantly di}erent from each other\ a parallel e}ect was not seen in alcohol males[ The distribution is similar to that seen in the cortex ^ all females again showed higher levels of allopregnanolone than the males\ and the pair!fed males in both the separated and control groups showed higher allopregnanolone than did the alcohol and lab chow males[ These results are shown in Fig[ 2[
111
B[ Zimmerberg and R[ C[ Brown
Fig[ 1[ Allopregnanolone concentrations "ng:g tissue2S[E[M[# in the prefrontal cortex of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[ Half of these subjects also had a postnatal stress experience\ daily maternal separation for 5 h from postnatal days 1 to 6[
Catecholamine concentrations Prenatal treatment had a signi_cant e}ect on the concentration of dopamine in the prefrontal cortex\ F"1\43# 2[07\ P ³ 9[94\ as shown in Fig[ 3[ The prenatal alcohol group had lower dopamine levels than either the pair!fed or control groups\ which did not di}er from each other[ There was also a signi_cant interaction between postnatal treatment and sex\ F"0\43# 5[01\ P ³ 9[91 on the concentration of dopamine in the prefrontal cortex ^ while postnatal separation did not e}ect dopamine levels in adult male subjects in the control groups\ female subjects in both the pair!fed and lab chow groups had higher levels than their non!separated comparable controls[ Although the three!way interaction did not reach signi_cance\ alcohol males from the separated condition had lower dopamine levels than any other male group\ while alcohol females did not exhibit increased dopamine with a postnatal stress history as did comparable separated female groups[ In the striatum\ there was signi_cant interaction between prenatal condition and postnatal treat! ment\ F"1\47# 5[64\ P ³ 9[90\ as seen in Fig[ 4[ While alcohol male and female subjects had higher dopamine levels in the striatum than either control group in the absence of a postnatal separation history\ this increased dopamine response to stress was not seen in the separated male and female
Prenatal experience and postnatal stress responses
112
Fig[ 2[ Allopregnanolone concentrations "ng:g tissue2S[E[M[# in the hippocampus of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[ Half of these subjects also had a postnatal stress experience\ daily maternal separation for 5 h from postnatal days 1 to 6[
groups[ There were no signi_cant e}ects of pre! or postnatal treatment or sex on the concentration of dopamine in the nucleus accumbens[ Analysis of norepinephrine concentrations in each of these same areas revealed some signi_cant e}ects[ In the prefrontal cortex\ there was a main e}ect of prenatal condition\ F"1\47# 7[48\ P ³ 9[990\ with all three groups di}ering signi_cantly from each other[ Norepinephrine con! centrations were lowest in the pair!fed groups "017[12200[87 mg:g tissue#\ next lowest in the alcohol groups "017[12200[86 mg:g tissue# and highest in the lab chow groups "048[88209[30 mg:g tissue# regardless of postnatal separation history[ In the nucleus accumbens\ in contrast\ only postnatal treatment had a signi_cant e}ect on norepinephrine levels\ F"1\47# 3[50\ P ³ 9[94[ All subjects with a separation history had higher norepinephrine concentrations "69[5322[88 mg:g tissue# that non!separated subjects "59[2721[09 mg:g tissue# regardless of prenatal treatment[ Finally\ there were no signi_cant e}ects of any factors on the concentration of norepinephrine in the striatum[ DISCUSSION Di}ering prenatal environments coupled with a postnatal experience manipulation did have an e}ect on the adult neurochemical response to an acute stressor\ although the e}ect varied among
113
B[ Zimmerberg and R[ C[ Brown
Fig[ 3[ Dopamine concentrations "mg:g tissue2S[E[M[# in prefrontal cortex of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[ Half of these subjects also had a postnatal stress experience\ daily maternal separation for 5 h from postnatal days 1 to 6[
neurochemical indices and brain areas examined[ The results from the allopregnanolone assays seem to re~ect a di}erence in the peripheral and the central nervous system response to stress[ The peripheral response "the plasma concentrations of allopregnanolone# was altered in the pair!fed group only[ This hyperresponsiveness to the swim stress may be due to the prenatal stress of the pair!feeding condition ^ since dams in the pair!fed condition have restricted feeding they tend to consume all of their non!alcohol liquid diet as soon as it is administered\ and have no food for the next 12 h\ while the alcohol dams {{sip|| their liquid diet throughout the 13 h\ and tubes are always re_lled when low[ The sex di}erence in the peripheral response "females × males# can be accounted for by the fact that females have higher circulating levels of progesterone in the blood\ due to its production in the ovaries\ and therefore may synthesize more allopregnanolone as needed in the adrenal gland\ which is the primary site of the peripheral stress response[ Enzymes responsible for the synthesis of allopregnanolone from cholesterol or progesterone have also been found in the smooth muscles\ platelets and the adrenal gland\ as well as in the testes and ovaries[ The central nervous system response\ interestingly\ showed the same distinct sex di}erence as seen in the plasma[ Female rats had higher cortical and hippocampal concentrations of allopregnanolone compared to both of the prenatal control groups[ This may signify that again that in females\ higher
Prenatal experience and postnatal stress responses
114
Fig[ 4[ Dopamine concentrations "mg:g tissue2S[E[M[# in striatum of male and female adult rats from one of three prenatal treatment groups "alcohol\ pair!fed or lab chow# after a 09 min cold water swim stressor[ Half of these subjects also had a postnatal stress experience\ daily maternal separation for 5 h from postnatal days 1 to 6[
circulating levels of progesterone are available to the brain to be converted into allopregnanolone[ Higher levels of allopregnanolone in stressful situations would predict lower levels of anxiety in female rats compared to male rats ^ this sex di}erence has been reported in both an open _eld activity test17 and in the plus maze[26 A history of prior postnatal maternal separation was associated with increased concentrations of allopregnanolone in the cortex and hippocampus in adult o}spring prenatally exposed to alcohol\ but not to o}spring with control prenatal treatments[ This enhances allopregnanolone stress response was particularly evident in females compared to males[ The stress of maternal separation during the _rst week of life may have led to functional changes in the developing neurosteroid system that results in increased production or mobilization of allopregnanolone in the brain during the stress response later in life\ and therefore enables the animal to cope better with stress[ The change in the production of allopregnanolone in the brain should be mirrored by a decrease in prefrontal cortex dopamine concentrations\ since anxiolytics reduce the stress!induced activation of the mesolimbocortical dopamine pathway[00 The methodology used in this experiment does not determine directly the activity of catecholaminergic pathways*further experiments using mic! rodialysis or assessment of metabolites are necessary to ascertain utilization[ However\ there does
115
B[ Zimmerberg and R[ C[ Brown
appear to be evidence that pre! and postnatal treatments can have long!term consequences on the baseline level of activity "e[g[ synthesis or metabolism# in these catecholaminergic systems[ The alcohol group exhibited lower levels of dopamine in the prefrontal cortex when combined with a prior maternal separation treatment[ Dopamine neurons project from the ventral tegmental area "VTA# to the prefrontal cortex and the nucleus accumbens\ and from the substantia nigra "SN# to the striatum[4\7 These projections are innervated by inhibitory\ GABA releasing neurons[ Dopamine released in the prefrontal cortex is excitatory\ and results in the activation of more GABA neurons\ which have inhibitory e}ects on the striatum and the nucleus accumbens through two glutamate pathways projecting from the prefrontal cortex to these areas[ The dopamine released in the nucleus accumbens and the striatum can be excitatory or inhibitory\ due to the diversity of dopamine receptors[ The decrease of dopamine released in the prefrontal cortex in the alcohol animals may be due to the inhibitory e}ects of allopregnanolone[ Since allopregnanolone is acting at the GABAA receptor\ it may be acting either at the source of the pathway\ in the VTA\ or in the prefrontal cortex itself[ The decrease in prefrontal cortical levels of dopamine suggests that allopregnanolone is inhibiting activation of the pathway by acting at GABAA receptors in the VTA\ although further studies are needed to determine this hypothesis directly[ Interestingly\ there was a sex di}erence in the response to postnatal experience[ Dopamine levels in the prefrontal cortex in the two prenatal control groups were elevated in females who had been separated from their dams during the _rst week or life\ but not in the males[ This increased dopamine response to swim stress may re~ect an underlying sex di}erence in sensitization of these dopamine pathways\ and requires further investigation[ The alcohol e}ect could be due to an upregulation of dopamine receptors in response to a developmental delay in dopamine release controlled by the VTA[ There is pharmacological evidence that there are fewer GABA receptors in one!week!old rats pups prenatally exposed to alcohol than there are in controls[27\28 The pattern of dopamine concentration was di}erent in the nucleus accumbens and the striatum[ In the striatum\ prenatal and postnatal experiences again had a combined impact on dopamine concentrations[ Only in the non!separated alcohol animals\ regardless of sex\ was there an increase in dopamine compared to the two prenatal control groups[ That would indicate that although {{typical|| prenatal alcohol!exposed animals would appear to be hyper!responsive to stress when tested as adults\ this hyper!responsiveness is not detectable in subjects who have had a prior early stress experience[ In contrast\ there were no e}ects of any manipulations on dopamine concentrations in the nucleus accumbens[ The results obtained for the concentrations of norepinephrine in the brain sections examined provide some preliminary suggestions about the e}ects of pre! and postnatal treatments on the noradrenergic response to stress[ While the pair!fed o}spring\ whose dams experienced the stress of restricted feeding\ had the lowest levels of norepinephrine in the prefrontal cortex\ prenatal alcohol exposed o}spring had a partial amelioration of this e}ect\ having lower norepinephrine levels than controls\ but not as low as pair!fed[ None of the prenatal condition groups had any additional e}ects of postnatal experience on norepinephrine levels in the prefrontal cortex[ In contrast\ all groups experiencing postnatal maternal separation had higher levels of norepinephrine in the nucleus accumbens when exposed to swim stress as adults[ No e}ects of pre! or postnatal manipulations were evident in striatal norepinephrine concentrations[ Studies have shown that norepinephrine is involved in the stress response\ and that a noradrenergic pathway projecting from the locus coeruleus to the prefrontal cortex\ nucleus accumbens and striatum is responsive to stress\ particularly the projection to the prefrontal cortex[3 This would lead us to predict that prenatal alcohol exposure would interact with postnatal stressors to increase norepinephrine release in the prefrontal cortex\ and perhaps in the nucleus accumbens and striatum as well[ If increased release or synthesis of allopregnanolone modulated "via its e}ects on GABAA receptors# these noradrenergic pathways\ there would be a decrease in the amount of norepinephrine released in the stress response[ This is the trend seen in the prefrontal cortex[ The alcohol and pair! fed animals have lower levels of norepinephrine than the lab chows[ There is no e}ect of postnatal condition\ therefore it may be that the noradrenergic projections to the prefrontal cortex appear very early in brain development and are not a}ected by the postnatal treatment[ Since norepinephrine levels are a}ected in all groups by postnatal\ but not prenatal\ stress\ a later time of development of this structure may be responsible for this contrasting result[ Cenci et al[3 reported that not all
Prenatal experience and postnatal stress responses
116
stressors activated the noradrenergic pathways ] the pathway to the prefrontal cortex was the most responsive\ while those to the nucleus accumbens were less so[ At the time of testing\ there were no di}erences in body weight between pre! and postnatal treatment groups[ Body weight di}erences were apparent on postnatal day 7 ] all separated subjects\ regardless of prenatal condition\ weighed less than controls[ However\ by postnatal day 10 the percent weight gain was similar in all groups\ suggesting that there were no long!term consequences on feeding[ It is possible\ though\ that the weight loss itself\ and not the maternal separation\ was responsible for the long!term sequela reported here[ In summary\ these results suggest that a variety of stressful experiences during the period of critical brain development may a}ect how an individual deals with stressors later in life[ The role of neurosteroids in modulating the stress response may be complex and sex!dependent\ but future research appears warranted to more fully understand their role as a possible endogenous negative feedback system[ Although these results do not directly measure activity\ they do provide a sugges! tion that there may be a reciprocal relationship between neurosteroid modulation of the GABAA receptor and the catecholaminergic systems those GABA neurons inhibit in stress!responsive areas of the brain\ and that there may be long!term consequences when stress is experienced during pre! and postnatal stages of neuronal development and maturation[ Acknowled`ements*This research was supported by a grant from the National Institute on Alcohol Abuse and Alcoholism\ èR90 AA97594[ We thank Jacqueline M[ Weider for her excellent technical assistance[ We also thank Perry Moore\ Southwest Foundation for Biomedical Research\ San Antonio\ TX for his analysis of allopregnanolone concentrations[
REFERENCES 0[ Abel\ E[ L[\ Physiological correlates of the forced swim test in rats[ Physiol[ Behav[\ 0882\ 43\ 298Ð206[ 1[ Bitran\ D[\ Purdy\ R[ H[ and Kellogg\ C[ K[\ Anxiolytic e}ect of progesterone is associated with increases in cortical allopregnanolone and GABAA receptor function[ Pharmacol[ Biochem[ Behav[\ 0882\ 34\ 312Ð317[ 2[ Bitran\ D[\ Hilvers\ R[ J[ and Kellogg\ C[ K[\ Anxiolytic e}ects of 2a!hydroxy!4aðbŁ!pregnan!19!one ] endogenous metabolites of progesterone that are active at the GABAA receptor[ Brain Res[\ 0880\ 450\ 046Ð050[ 3[ Cenci\ M[\ Kalen\ A[ P[\ Mandel\ R[ J[ and Bjorklund\ A[\ Regional di}erences in the regulation of noradrenaline release in medial frontal cortex\ nucleus accumbens and caudate!putamen ] a microdialysis study in the rat[ Brain Res[\ 0881\ 470\ 106Ð117[ 4[ Cuello\ A[ C[\ Priestly\ J[ V[ and Paxinos\ G[\ Substance P and enkephalin containing pathways[ In The Rat Nervous System\ Vol[ 0\ Forebrain and Midbrain\ ed[ G[ Paxinos[ Academic Press Inc[\ Orlando\ 0874\ pp[ 330Ð353[ 5[ Deutsch\ S[ I[\ Mastropaolo\ J[ and Hitri\ A[\ GABA!active steroids ] endogenous modulators of GABA!gated chloride ion conductance[ Clin[ Neuropharmacol[\ 0881\ 04\ 241Ð253[ 6[ Fadda\ F[\ Mosca\ E[\ Ni}ol\ T[\ Colombo\ G[ and Gessa\ G[ L[\ Ethanol prevents stress!induced increase in cortical DOPAC ] reversal by RO 04!3402[ Physiol[ Behav[\ 0876\ 39\ 272Ð274[ 7[ Fallon\ J[ H[ and Loughlin\ S[ E[\ Substantia Nigra[ The Rat Nervous System\ Vol[ 0\ Forebrain and Midbrain\ ed[ G[ Paxinos[ Academic Press Inc[\ Orlando\ 0874\ pp[ 242Ð263[ 8[ Finn\ D[ A[ and Gee\ K[ W[\ The in~uence of estrus cycle on neurosteroid potency at the g!aminobutyric acidA receptor complex[ J[ Pharmacol[ Exp[ Ther[\ 0882\ 154\ 0263Ð0268[ 09[ Gandleman\ R[\ Psychobiolo`y of Behavioral Development[ Oxford University Press\ New York\ 0881\ pp[ 057Ð070[ 00[ Grobin\ A[ C[\ Roth\ R[ H[ and Deutch\ A[ Y[\ Regulation of the prefrontal cortical dopamine system by the neuroactive steroid 2a\10!dihydroxy!4a!pregnane!19!one[ Brain Res[\ 0881\ 467\ 240Ð245[ 01[ Hennessey\ M[ B[ and Weinberg\ J[\ Adrenocortical activity during conditions of brief social separation in preweaning rats[ Behav[ Neural Biol[\ 0889\ 43\ 31!44[ 02[ Hilakivi!Clarke\ L[ A[\ Turkka\ J[\ Lister\ R[ G[ and Linnoila\ M[\ E}ects of early postnatal handling on brain b! adrenoreceptors and behavior in tests related to stress[ Brain Res[\ 0880\ 431\ 175Ð181[ 03[ Lan\ N[ C[\ Gee\ K[ W[\ Bolger\ M[ B[ and Chen\ J[ S[\ Di}erential responses of expressed recombinant human g! aminobutyric acidA receptors to neurosteroids[ J[ Neurochem[\ 0880\ 46\ 0707Ð0710[ 04[ Majewska\ M[ D[\ Harrison\ N[ L[\ Schwartz\ R[ D[\ Barker\ J[ L[ and Paul\ S[ M[\ Steroid hormone metabolites are barbiturate!like modulators of the GABAA receptor[ Science\ 0875\ 121\ 0993Ð0996[ 05[ McEwen\ B[ S[\ Steroid hormones ] e}ect on brain development and function[ Hormone Res[\ 0881\ 26\ 0Ð09[ 06[ Morrow\ A[ L[\ Pace\ J[ R[\ Purdy\ R[ H[ and Paul\ S[ M[\ Characterization of steroid interaction with g!aminobutyric acid receptor!gated chloride ion channels ] evidence for multiple steroid recognition sites[ Molec[ Pharmacol[\ 0889\ 26\ 152Ð169[ 07[ Myslobodsky\ M[ S[\ Pro!and anticonvulsant e}ects of stress ] the role of neuroactive steroids[ Neurosci[ Biobehav[ Rev[\ 0882\ 06\ 018Ð028[ 08[ Nelson\ L[ R[\ Taylor\ A[ N[\ Lewis\ J[ W[\ Poland\ R[ E[\ Redei\ E[ and Branch\ B[ J[\ Pituitary!adrenal responses to morphine and footshock stress are enhanced following prenatal alcohol exposure[ Alcohol Clin[ Exp[ Res[\ 0875\ 09\ 286Ð391[ 19[ Orchinik\ M[ and McEwen\ B[\ Novel and classical actions of neuroactive steroids[ Neurotransmissions\ 0882\ 8\ 0Ð5[ 10[ Paul\ S[ M[ and Purdy\ R[ H[\ Neuroactive steroids[ The FASEB Journal\ 0881\ 5\ 1200Ð1211[ 11[ P_efer\ W[ D[\ Rotundo\ R[\ Myers\ M[ and Denenberg\ V[ H[\ Stimulation in infancy ] unique e}ects of handling[ Physiol[ Behav[\ 0865\ 06\ 670Ð673[
117
B[ Zimmerberg and R[ C[ Brown
12[ Prince\ R[ J[ and Simmonds\ M[ A[\ Di}erential antagonism by epipregnanolone of alphaxalone and pregnenolone potentiation of ð2HŁ~unitrazepam binding suggests more than one class of binding site for steroids at GABAA receptors[ Neuropharmacol[\ 0882\ 21\ 48Ð52[ 13[ Puia\ G[\ Santi\ M[\ Vicini\ S[\ Pritchett\ D[ B[\ Purdy\ R[ H[\ Paul\ S[ M[\ Seeburg\ P[ H[ and Costa\ E[\ Neurosteroids act on recombinant human GABAA receptors[ Neuron\ 0889\ 3\ 648Ð654[ 14[ Purdy\ R[ H[\ Moore\ P[ H[\ Rao\ P[ N[\ Hagino\ N[\ Yamaguchi\ T[\ Schmidt\ P[\ Rubinow\ D[ R[\ Morrow\ A[ L[ and Paul\ S[ M[\ Radioimmunoassay of 2a!hydroxy!4a!19!one in rat and human plasma[ Steroids\ 0889\ 44\ 189Ð185[ 15[ Purdy\ R[ H[\ Morrow\ A[ L[\ Moore\ P[ H[ and Paul\ S[ M[\ Stress!induced elevations of g!aminobutyric acid type A receptor!active steroids in the rat brain[ Neurobiol[\ 0880\ 77\ 3442Ð3446[ 16[ Roth\ R[ H[\ Tam\ S[!Y[\ Ida\ Y[\ Yang\ J[!X[ and Deutch\ A[ Y[\ Stress and the mesocorticolimbic dopamine systems[ Annals of the New York Academy of Science\ 0877\ 446\ 027Ð036[ 17[ Stewart\ J[ and Cyrgan\ D[\ Ovarian hormones act early in development to feminize adult open!_eld behavior in the rat[ Hormones Behav[\ 0879\ 03\ 19Ð21[ 18[ Taylor\ A[ N[\ Branch\ B[ J[\ Lin\ S[ H[ and Kokka\ N[\ Long!term e}ects of fetal ethanol exposure on pituitary!adrenal response to stress[ Pharm[ Biochem[ Beh[\ 0871\ 05\ 474Ð478[ 29[ Taylor\ A[ N[\ Branch\ B[ J[\ Nelson\ L[ R[\ Lane\ L[ A[ and Poland\ R[ E[\ Prenatal ethanol and ontogeny of pituitary! adrenal responses to ethanol and morphine[ Alcohol\ 0875\ 2\ 144Ð148[ 20[ Walshlak\ A[ and Weinstock\ M[\ Neonatal handling reverses behavioral abnormalities induced in rats by prenatal stress[ Physiol[ Behav[\ 0889\ 37\ 178Ð181[ 21[ Weinberg\ J[ and Levine\ S[\ Early handling in~uences on behavioral and physiological responses during active avoidance[ Devel[ Psychobiol[\ 0866\ 09\ 050Ð058[ 22[ Weinberg\ J[\ Prenatal ethanol exposure alters adrenocortical development of o}spring[ Alcohol Clin[ Exp[ Res[\ 0878\ 02\ 62Ð72[ 23[ Weinberg\ J[\ Prenatal ethanol e}ects ] sex di}erences in o}spring stress responsiveness[ Alcohol\ 0881\ 8\ 108Ð112[ 24[ Wyly\ M[ V[\ Denenberg\ V[ H[\ De Santis\ D[\ Burns\ J[ K[ and Zarrow\ M[ X[\ Handling rabbits in infancy ] in search of a critical period[ Developmental Psychobiol[\ 0864\ 7\ 068Ð075[ 25[ Zimmerberg\ B[\ Brunelli\ S[ A[ and Hofer\ M[ A[\ Reduction of rat pup ultrasonic vocalizations by the neuroactive steroid allopregnanolone[ Pharmacol[ Biochem[ Behav[\ 0883\ 36\ 624Ð627[ 26[ Zimmerberg\ B[ and Farley\ M[ J[\ Sex di}erences in anxiety behavior in rats ] role of gonadal hormones[ Physiol[ Behav[\ 0882\ 43\ 0008Ð0013[ 27[ Zimmerberg\ B[\ Drucker\ P[ A[ and Weider\ J[ M[\ Di}erential behavioral e}ects of the neuroactive steroid allo! pregnanolone on neonatal rats prenatally exposed to alcohol[ Pharmacol[ Biochem[ Behav[\ 0884\ 40\ 352Ð357[ 28[ Zimmerberg\ B[ and McDonald\ B[ C[\ Prenatal alcohol!exposure in~uences the e}ects of neuroactive steroids on separation!induced ultrasonic vocalizations in rat pups[ Pharmacol[ Biochem[ Behav[\ 0885\ 44\ 430Ð436[ 39[ Zimmerberg\ B[ and Shartrand\ A[ M[\ Temperature!dependent e}ects of maternal separations on growth\ activity\ and amphetamine sensitivity in the rat[ Dev[ Psycho[ Biol[\ 0881\ 14\ 102Ð115[