Fetal ethanol exposure alters pituitary-adrenal sensitivity to dexamethasone suppression

Pergamon

Psychoneuroendocrinology, Vol. 21, No. 2, pp. 127-143, l!196 Copyright © 1996 Elsevier Science Ltd. All rights reserved Printed in Great Brilain 0306-4530/96 $15.00 + .00

0306-4530(95) 00037-2

FETAL ETHANOL EXPOSURE ALTERS PITUITARYA D R E N A L S E N S I T I V I T Y TO D E X A M E T H A S O N E SUPPRESSION J. A. Osborn, C. K. Kim, W. Yu, L. Herbert and J. Weinberg Department of Anatomy, Faculty of Medicine, University of British Columbia, Vancouver, B.C., Canada

SUMMARY The present study investigated the hypothesis that a deficit in feedback inhibition of the hypothalamic-pituitary-adrenal (HPA) axis may underlie the hormonal hyperresponsiveness seen in fetal ethanol-exposed rats. Male and female Sprague-Dawley rats from prenatal ethanol (E), pairfed (PF) and ad lib-fed control (C) treatment groups were tested in adulthood. The effects of dexamethasone (DEX) blockade on basal and stress corticosterone (CORT) levels and stress adrenocorticotropin (ACTH) levels were examined over a 36-h period. Stress CORT and ACTH levels after DEX administration at the trough (AM) and peak (PM) of the CORT circadian rhythm were compared. DEX administration significantly suppressed both resting and stress levels of CORT and ACTH in all animals, regardless of prenatal treatment. Importantly, E animals did not differ from PF and C animals in basal CORT. However, E males and females had significantly higher stress levels of CORT and/or ACTH than PF and C animals, and further, showed differential responsiveness following DEX administration depending on the time of day when testing occurred. At the trough of the CORT circadian rhythm, E males did not differ from PF and C males, whereas E females had increased stress levels of CORT compared to PF and C females. In contrast, at the peak of the circadian rhythm, E males showed increased stress levels of CORT but not ACTH, whereas E females showed increased stress levels of both CORT and ACTH compared to males and females in respective control groups. These data support the hypothesis that E animals may exhibit deficits in HPA feedback inhibition compared to controls and suggest a sex-specific difference in sensitivity of the mechanism underlying HPA hyperresponsiveness. Copyright © 1996 Elsevier Science Ltd. Keywords--Prenatal ethanol; Hypothalamic-pituitary-adrenal (HPA) axis; Feedback inhibition; Dexamethasone; Corticosterone; Circadian rhythm.

INTRODUCTION

Prenatal exposure to ethanol (E) alters development and responsiveness of the offspring hypothalamic-pituitary-adrenal (HPA) axis. During the early neonatal period, rodents prenatally exposed to ethanol have elevated basal levels of brain, plasma and adrenal corticosterone (CORT), and decreased corticosterone binding globulin (CBG) binding capacity (Kakihana et al., 1980; Taylor et al., 1983; Weinberg et al., 1986; Weinberg, 1989) as well as elevated plasma and decreased pituitary fl-endorphin (/3-EP) levels (Angelogianni Address correspondence and reprint requests to: Dr Joanne Weinberg, Department of Anatomy, Faculty of Medicine, 2177 W e s b r o o k Mall, Vancouver, B.C., Canada V 6 T 1Z3 (Tel.: 604 822 6214, Fax: 604 822 2316; Email: j o a n n e w @ u n i x g . u b c . c a ) . 127

128

J.A. Osborn et al.

& Gianoulakis, 1989) compared to control animals. By 3-5 days of age, basal CORT levels in E animals return to normal. However, the HPA and/~-EP responses to stressors appear to be suppressed or blunted throughout the preweaning period compared to that in control animals (Angelogianni & Gianoulakis, 1989; Taylor et al., 1986a; Weinberg et al., 1986; Weinberg, 1989). Importantly, this blunted stress responsiveness in E animals is a transient phenomenon. By 15-21 days of age, the HPA response to stress appears to normalize, and from weaning on E offspring typically demonstrate HPA hyperresponsiveness to stressors which persists throughout life (Nelson et al., 1986; Taylor et al., 1981; Weinberg, 1988, 1992a; Weinberg & Gallo, 1982). Adult E offspring are hyperresponsive to many physical and neurogenic stressors including cardiac puncture (Taylor et al., 1982), restraint (Taylor et al., 1982; Weinberg, 1988, 1992a), noise and shaking (Taylor et al., 1982), novel environments (Weinberg, 1988), intermittent shock (Nelson et al., 1984, 1986), ether (Angelogianni & Gianoulakis, 1989; Weinberg & Gallo, 1982), and cold (Angelogianni & Gianoulakis, 1989) In addition to HPA hyperresponsiveness, E animals demonstrate deficits in pituitaryadrenal response inhibition or recovery from stress. For example, E animals show prolonged CORT, adrenocorticotropin (ACTH) and /~-EP elevations during and following restraint stress (Weinberg, 1988, 1992a; Weinberg et al., 1996) and also show smaller CORT decreases when allowed access to water in a novel environment (Weinberg, 1988), as compared to control animals. Similarly, E offspring show more prolonged ACTH elevations than control animals following 10 min footshock stress (Taylor et al., 1986b). Interestingly, HPA hyperresponsiveness and/or the deficits in response inhibition may be manifested differently in males and females depending on the nature and intensity of the stressor, the time course measured, and the hormonal endpoint examined. A possible mechanism underlying the HPA hyperresponsiveness seen in E offspring is a deficit in feedback inhibition of the HPA axis. As the hippocampus is a principal target site for glucocorticoid feedback in the brain (McEwen et al., 1986; Sapolsky et al., 1984), a recent study (Weinberg & Petersen, 1991) investigated the possibility that an ethanolinduced decrease in hippocampal glucocorticoid receptor concentration might, in part, mediate this altered HPA responsiveness. The data demonstrated that there were no significant differences in specific binding density or binding affinity for either Type I (mineralocorticoid) or Type II (glucocorticoid) receptors in the hippocampus of E compared to control animals, indicating that feedback deficits in E animals do not occur at the level of the hippocampal receptors, at least under basal or nonstressed conditions. In contrast, support for a deficit in feedback inhibition in E animals comes from the work of Nelson et al. (1985) demonstrating that E animals appear to have an accelerated rebound of basal CORT levels following a high dose of the synthetic glucocorticoid, dexamethasone-21-phosphate (DEX). Clinically, the DEX suppression test has been used to evaluate HPA axis function in a number of psychiatric conditions and it appears that feedback inhibition of CORT is altered in a number of affective disorders (Nemeroff et al., 1988). The present study utilized DEX suppression of the HPA axis to explore further the hypothesis that the HPA hyperresponsiveness and/or delays in recovery from stressors that occur in E offspring result from deficits in feedback inhibition of the HPA axis induced by prenatal ethanol exposure. The effects of DEX blockade on basal and stress CORT levels and stress ACTH levels were examined over a 36-h period. In addition, stress CORT and ACTH levels were examined after administration of DEX, during both the trough (AM) and peak (PM) of the CORT circadian rhythm.

Prenatal Ethanol Exposure and HPA Feedback

t29

MATERIALS AND METHODS

Animals and Mating Sprague-Dawley male (n = 25) and female (n = 145) rats were obtained from Canadian Breeding Farms, St. Constant, Quebec. Both males and females were group housed for 1-2 weeks prior to breeding to allow recovery from transportation and adaptation to the colony room. Males were then singly housed in stainless steel mesh hanging cages (25 x 18 x 18 cm), and were maintained on standard laboratory chow (Ralston Purina of Canada, Woodstock, Ontario) and water. The colony room had controlled temperature (21°C) and lighting, with lights on from 0600h to 1800h. Females were placed singly with males and cage papers were checked daily for vaginal plugs. Day 1 (D1) of gestation was considered the day the plug was found. Three replicate breedings, with 48--49 females in each, were done. All animal use procedures were in accordance with NIH guidelines and were approved by the University of British Columbia Animal Care Committee.

Diets and Feeding On D 1 of gestation, females were rehoused into polycarbonate cages (24 x 16 x 46 cm) and randomly assigned to one of three groups: (1) ethanol (E), liquid ethanol diet (36% ethanolderived calories), ad lib (n = 45); (2) pair-fed (PF), liquid control diet (maltose-dextrin isocalorically substituted for ethanol), with each animal pair-fed the amount consumed by a female in the ethanol group (g/kg body weight) on the same day of gestation (n = 45); and (3) control (C), laboratory chow and water, ad lib (n = 45). The diets used were previously developed by our laboratory to provide adequate nutrition to pregnant females regardless of ethanol intake (Weinberg, 1985) and were prepared by Bioserv Inc., Frenchtown, NJ, USA. Fresh diet was placed on the cages daily just prior to lights off to avoid a shift in the CORT circadian rhythm. It has been demonstrated that if animals receive a restricted amount of food (such as that received by the PF group), circadian rhythms will re-entrain to the feeding time thus shifting the CORT rhythm (Gallo & Weinberg, 1981). Bottles from the previous day were removed and weighed at this time to determine the amount of diet consumed, Experimental diets were continued until gestation D22 when they were replaced with laboratory chow and water, ad lib. Females were undisturbed except for weighing and cage cleaning on DI, 7, 14 and 21 of gestation. At birth, designated D1 of lactation, dam and pups were weighed and all litters culled to 10 (five males and five females). Dam and pups were weighed and cages cleaned on D1, 8, 15, 22 of lactation. On D22, pups were weaned and housed by sex and by litter until testing at 90-110 days of age. On D12-14 of gestation, blood samples (0.4-0.6 ml) were obtained from the tail of nine unanesthetized females at 1900h for determination of blood ethanol levels (Sigma Diagnostic Kit 332-UV, based on Bonnischsen & Theorell, 1951).

Testing and Blood Sampling One week prior to testing, animals were singly housed, divided into four subsets, and randomly assigned to IP injection dose and test time. Testing order was counterbalanced across prenatal treatment, sex and injection dose (n = 6-9 for each of E, PF and C, males and females at each dose and time). DEX injections for the first three subsets of animals occurred at 0730h-0900h on the initial test day (AM groups). Doses of DEX were based on pilot studies which indicated that females required higher doses of DEX than males to show

130

J, A. Osborn et al.

suppression of the CORT response to stress. In addition, to prevent unnecessary sacrfice of animals, only higher doses of DEX were used at the 10, 26, and 36 h time points as pilot data indicated that animals given the lower doses no longer showed suppressed CORT responses to stress at these times. Animals in the first subset received one of four doses of DEX [males, 0 (saline), 1.0, 5.0 or 15.0 ~g/100 g body wt (bw); females, 0, 1.0, 10.0 or 30.0/~g/100g bw] and were tested at 3 or 6 h post-injection. Animals in the second subset received one of three doses of DEX [males, 0, 5.0 or 15.0 ~g/100g bw; females, 0, 10.0 or 30.0 #g/100g bw] and were tested at 10 or 26 h post-injection. Animals in the third subset received one of two doses of DEX [males, 0 or 15.0 ~g/100g bw; females, 0 or 30.0 ~g/100g bw] and were tested at 36 h post-injection. DEX injections for the fourth subset of animals occurred at 1500h1700h on the initial test day (PM group). Animals received one of three doses of DEX [males, 0, 5.0 or 15.0/~g/100g bw; females, 0, 10.0 or 30.0 ~g/100g bw], and were tested 3 h post-injection under red light. All animals were returned to the colony room between injection and blood sampling. At the designated sampling time, animals were taken from the colony room to an adjacent laboratory, quickly and lightly anesthetized with ethyl ether, and blood samples (0.5 cc) obtained by cardiac puncture using heparinized syringes (the fourth subset of animals was exposed to ether only; no basal samples were drawn). The entire sampling procedure was completed within 2 min of removing the animal from the colony room, which is rapid enough to obtain a reliable measure of CORT, without any effects of disturbance or etherization (Davidson et al., 1968). Basal or resting levels of ACTH could not be obtained with this procedure as it is too slow; samples for basal ACTH must be obtained within seconds of removing the animal from the colony room. The procedure of taking the resting sample served as the stressor. Twenty min later, animals were rapidly decapitated (within 10-15 s of touching the cage; Rivier et al., 1982) and trunk blood collected on ice in 12 x 75 mm plastic test tubes containing 7.5 mg EDTA and 1000 KIU aprotinen (0.2 ml/5 cc blood). The blood was centrifuged at 2200 xg for 10 min at 4°C and plasma transferred with plastic pipettes to microcentrifuge tubes for storage at - 7 0 ° C until assayed for CORT and ACTH.

Radioimmunoassays Corticosterone. Total CORT (bound plus free) was measured by radioimmunoassay in plasma extracted in absolute ethanol (1:10 v/v), using our adaptation (Weinberg & Bezio, 1987) of the method of Kaneko et al. (1981). Antiserum was obtained from Immunocorp, Montreal, PQ: tracer, [1,2,6,7-3H]-corticosterone, was obtained from Dupont, New England Nuclear, Mississauga, ON, Canada; unlabelled corticosterone for standards was obtained from Sigma Chemical Co, St. Louis, MO, USA. Dextran coated charcoal was used to absorb and precipitate free steroids after incubation. Samples were counted in Formula 989 (Dupont). The intra and interassay coefficients of variation were 3 and 3.9%, respectively. ACTH. Plasma ACTH was assayed using the Incstar ACTH Equilibrium RIA kit (Incstar Inc, Stillwater, Minnesota, USA) with all reagent volumes halved and 50 #1 plasma per tube. The antiserum has 100% cross-reactivity with porcine ACTHl_39 and human ACTHl_24 but less than 0.001% cross-reactivity with ctMSH, fl-endorphin and fl-lipotropin (Orth, 1979). The midrange intra and interassay coefficients of variation were 3.9 and 6.5%, respectively.

Prenatal Ethanol Exposure and HPA Feedback

131

Statistical Analyses All data were analyzed by appropriate analyses of variance (ANOVA) for factors of prenatal treatment, sex and dose of DEX. Significant main and interactions effects were further analyzed by Tukey's paired comparisons.

RESULTS

Developmental Data Ethanol intake of the pregnant females was consistently high throughout gestation in all three breedings, averaging 9.7 ± 1.4, 11.15 ± 1.0 and 10.7 _+0.8 g/kg bw for weeks 1, 2 and 3 of gestation, respectively. Blood alcohol levels were consistent with those reported previously (Weinberg, 1985), averaging 145.4 ± 10.9 mg/dl. Repeated measures ANOVAs on maternal weight gain during pregnancy and pup weight gain during lactation revealed significant main effects of group (p<.001 and p < . 0 1 , respectively) and days (ps<.001), as well as group × days interactions (ps<.01). Body weights of E and PF dams were significantly less than those of C dams on gestation D7-21 (ps < .001). There were no significant differences among groups for litter size. However, E and PF pups weighed significantly less than C pups on all days measured during lactation (ps< .001). There were no significant differences in body weight among E, PF and C offspring at the time of testing.

CORT and ACTH Levels (AM Groups) Both male and female offspring tested in adulthood showed dose-response relationships for both resting CORT levels and for stress CORT and ACTH levels following DEX blockade; the higher the dose of DEX, the greater the CORT and ACTH suppression and the longer the time taken to return to basal levels. Undisturbed CORT Levels. Significant main effects of dose of DEX were obtained at 3, 6 and 10 h post-injection (ps < .01). Both males and females injected with all doses of DEX had significantly lower undisturbed CORT levels than their saline-injected counterparts Table I. Undisturbed CORT levels (/~g/100 ml; mean + SEM) DEX (/~g/100g bw)

3h

0 1.0 5.0 15.0

1.7 + 0.4 0.7_+ 0.04* 0.7+-0.03 * 0.7+-0.2*

0 1.0 10.0 30.0

3.6 +-0.7 0.9 __.0.04* 0,9+0.04 * 0,9 4-_0.03"

6h

Time Post-DEX Injection 10 h 26 h

Males 2.8 +-0.5 13.2 + 1.8 0.9 + 0.2" -0.7+0.03 * 2.7+0.7* 0.7 +-0.02" 0.7+_0.1* Females 2.8 +-0.6 29.7 + 3.0 1.5 + 0.5 -1.4+-0.2" 3.0+-1.0" 1.4 +_0.2* 1.3 +_0.2*

*Main effectof dose,p< .01:0 (saline)>DEXp< .01.

36 h

1.0 + 0.1 -0.6+-0.1 0.8+-0.2

11.4 +_0.7 --10.3+-0.9

3.0 +-0.7 -1.3+_.0.3 1.4 +-0.2

29.2 _ 2.6 --26.9 _ 2.5

132

J . A . Osborn et al.

E

100 j

~

90

100

3 hr

80

}

7o

90

26 hr

80 70

~

60

~ k) ~

50

50

40

40

~

30

30

60

20 10

|0 0 0

--"

5.0

I 0

50

15.0

150

100

90

I00 r

6hr

}~o

90 t-

36 hr

80 b

70

70 1-

~

60

~ ks

50

~

40

~-,

20

2O

10

1

60 150 ~40 b

30

0

I0

5.0

0

150

ll)0

~

oo

~

so

}

~o

~

150

DE)[ (ug/lO0 g bw) I

CONTROL

~

PAIRFED

~

EI'HAN()I

10 hr

5o 40

~

30

~',

20 I0

50

15.0

DE)( (ugh OOgbw)

Fig. 1. Stress CORT levels in males following DEX injection in the AM. *Main effect of dose, p < .01: 5.0, 15.0 < 0 #g/100 g bw at 3, 6, 10 and 26 h, ps <.01.

(Table I). The only exception was at the 6 h sampling time when females injected with the lowest dose of DEX (1.0/lg/100 g body wt) did not differ from saline injected females. By 26 h, there were no significant differences between DEX- and saline-injected groups. There were no significant differences among E, PF and C animals in undisturbed CORT levels at any dose of DEX or at any of the sampling times. Data in Table I were combined across E, PF and C groups. Stress CORT Levels. Significant main effects of dose were obtained for both males (Fig. 1) at 3, 6, 10 and 26 h post-injection (ps < .01). At all of these times, males injected with 5.0 or 15.0 #g/100 g bw DEX had lower CORT levels than males injected with saline (ps < .01). There were no significant differences in CORT levels among E, PF and C males at any dose of DEX or at any of the times tested. At 3, 6 and 26 h post-injection, females injected with 10.0 or 30.0/zg/100 g bw DEX had lower CORT levels than females injected with saline

Prenatal Ethanol Exposure and HPA Feedback

10o

133

I00

90I 70 60~

3 hr

~o

~~/~ , ~ .

40

Z

# ~

I0

26 hr

~o

60 50

50

i 30 20 10 0 0 10,9I ~ q070 ~ 60 ~ 50 i 4030

90 70

I00

40

30 20 10

0

300

0

3O0

100

100

6 hr

,o

~

~

90 70

36 hr

~o

60 50 40 30

20

:

I0

. ~

0

10

I00

300

20 I0

0

0

300

DE)( (ug/lO0 g bw)

90 80

10 hr

CONTROl,~

PAIRFED

~ FIH,V,()I

60

~

50 40 30 20

I0 I)

lOI)

300

DE)( (ug/'l OOg b w)

Fig. 2. Stress CORT levels in females following DEX injection in the AM. *Main effect of dose, p < .01: 10.0, 30.0 < 0 #g/100 g bw at 3, 6 and 26 h, ps < .01; 30.0 #g/100 g bw at 10 h, p < .0l. ~'Main effect of group, p < .05: E > PF = C, ps < .05.

(ps < .01). At 10 h post-injection, only females injected with 30.0 pg/100 g bw DEX had lower CORT levels than saline-injected females (p < .01). Importantly, at 3 h post-injection, analysis of female CORT levels revealed a significant main effect of group (p < .05). E females had higher CORT levels than PF and C females (ps < .05). E females injected with 30/ag/100 g bw DEX also had higher CORT levels than PF and C females at 6 h post-injection (ps < .05). There were no significant differences in CORT levels among E, PF and C females at 10, 26 or 36 h post-injection. Stress ACTH Levels. Significant main effects of dose were obtained at 3, 6 and 10 h postinjection for both males (Fig. 3) and females (Fig. 4) (ps < .001). At 3 h post-injection, males and females injected with all doses of DEX had lower ACTH levels than their saline-injected counterparts. At 6 and 10 h post-injection, males injected with the two highest doses and females injected with the highest dose of DEX, had lower ACTH levels than those injected

134

J.A. Osborn et al.

~o

30C

3O0

3 hr

25C

200

201]

"~

150

15C

i

100

i0~

5o

0

1.0

50

{~50 ~~ ~

15.0

300

"~

5.0

15.0

300

6hr

36 hr

250

200

200 I

150

150 I-

I00

100 50

50

oL 0

10

50

15.0

15.0

300

DE)( (ug/lO0 g bw) CONTROL

~

PAIR-FED

~

El HANOI

10 hr

250 200 '~

150

~ a.,

100 50 0 0

5.0

15.0

DE)( (ug/l O0g/bw)

Fig. 3. Stress ACTH levels in males following DEX injection in the AM. *Main effect of dose, p<.001: 1.0, 5.0, 15.0<0 #g/100 g bw at 3 h, ps<.01; 5.0, 15.0<0 /~g/100 g bw at 6 h, p<.01; 15.0<5.0 = 0 at 10 h, ps< .01. #Main effect of group, p < .05: E = C > PF at 26 h, ps< .05. with saline (ps < .01). At 26 and 36 h post-injection, there were no significant differences for either males or females in ACTH levels between DEX- and saline-injected animals. Significant main effects of group were also observed (ps < .05). At 26 h post-injection, E and C males had significantly higher ACTH levels than PF males (ps < .05), and E and PF females had significantly higher ACTH levels than C females (ps < .05).

CORT and ACTH Levels (PM Group) Stress CORT Levels. Significant main effects of dose were obtained for both males and females (ps < .001) (Fig. 5); animals injected with DEX had lower CORT levels than animals injected with saline. For males, there was a significant main effect of group (p < .05) and a group × dose interaction (p < .01). At 5.0 ~g/100 g bw DEX, E males had higher CORT levels than C males; at 15.0/~g/100 g bw DEX, E males had higher CORT levels than PF

Prenatal Ethanol Exposure and H P A Feedback

400

3hr

.~ 350 3(10

1 35

400 350 300

,~

250

250

~r~

200

200

150

150 100 50 0 0

I00

I 0

0

300

I00

~0

400

6hr

3511 300

~'~

i

400

h r 36

350 300

250

250

200

200

150

150

I00

100 50

0 0

10.0

10

4(X)

10hr

35(1

0

30.0

300

DE)( (ug/lO0 g bw) I

CONTROL

~

PAIR-FED

~

FILtAN()I

311(I 25(1 2(10 ~

15(1

50 11

100

30.0

DE)( (ug/lO0 g bw)

Fig. 4. Stress ACTH levels in females following DEX injection in the AM. *Main effect of dose, ~< .001: 1.0, 10.0, 30.0 < 0/~g/100 g bw at 3 h, p < .01; 30.0 < 0/~g/100 g bw at 6 and 10 h, ps < .01. Main effect of group, p< .05: E = PF > C at 26 h, ps< .05.

males (ps< .05). Similarly, a significant main effect of group was found for females (/9 < .001). E females injected with 10.0 and 30.0/~g/100 g bw DEX had significantly higher CORT levels than C females (ps < .01).

Stress ACTH Levels. Significant main effects of dose were obtained for both males and females (ps < .001) (Fig. 6); animals injected with DEX had significantly lower ACTH levels than saline-injected animals. For males, there were no significant differences in ACTH levels among E, PF and C animals. In contrast, there was a significant main effect of group for females (p < .05); at 30 #g/100 g bw DEX, E females had significantly higher ACTH levels than PF and C females (ps < .05).

136

J.A. Osborn et al.

1

CONTROL

~

PAIRFED

~

ETHANOl,

100 90

Male

80 70 60 50 40 30 20, 10 0 0 ~'

5.0

150

I00 90

~ ~,J

i

6080 5070

Female #

0

~- 20 r 0 0

I0.0

30.0

DEX (ugh O0 bw)

Fig. 5. Stress CORT levels following DEX injection in the PM. *Main effect of dose, p<.001: DEX< saline, ps<.001. #Main effect of group, p<.05: males E > PF = C, ps < .05; females E > C, ps < .01. +Group × dose interaction, p < .05: males at 5.0/xg/100 g bw, E > C, p < .05; at 15.0 pg/100 g bw, E > PF, p < .05. DISCUSSION The results from these experiments support our hypothesis that HPA hyperresponsiveness and/or delays in recovery from stressors that occur in E offspring may result, at least in part, from deficits in feedback inhibition of the HPA axis induced by prenatal ethanol exposure. Administration of DEX to block HPA activity significantly suppressed both resting levels of plasma CORT and stress levels of plasma CORT and ACTH in all animals, regardless of prenatal treatment. Importantly, E animals did not differ from PF and C animals in basal CORT levels but exhibited significantly higher stress levels of CORT and/or ACTH than PF and C animals at 3 and 6 h following DEX blockade. Furthermore, males and females exhibited differential responsiveness depending on the time of day when testing occurred. When tested at the trough of the CORT circadian rhythm, reduced sensitivity to DEX suppression of stress hormone levels was observed only in E females; E males were similar to PF and C males in responsiveness. In contrast, when tested at the peak of the CORT circadian rhythm, both E males and E females exhibited reduced sensitivity to DEX suppression of stress hormone levels. Sex differences in HPA responsiveness following DEX were further demonstrated by the finding that E males showed increased stress levels of CORT but not ACTH, whereas E females showed increased stress levels of both CORT and ACTH compared to their respective controls. These data suggest that the insult of prenatal ethanol exposure affects both male and female offspring, but that there may be a sex-specific

Prenatal Ethanol Exposure and HPA Feedback

1

CONTROL

~

I'AIR-F[D

137

[-JHANOL

300 250 200 150 .~

100 50

0

¢ )

150

400 350

Female

300

250 200 150 I00

o

I 0.o

30 o

DEX (ug/lO0 g hw)

Fig, 6. Stress ACTH levels following DEX in the PM. *Main effect of dose, p < .001: DEX < saline, ps < .001. #Main effect of group, p < .05: females at 30.0 pg/100 g bw, E > PF = C, ps < .05.

difference in sensitivity of the mechanism(s) underlying HPA hyperresponsiveness. Moreover, consistent with previous studies (Taylor et al., 1983; Weinberg, 1992a, b), it appears that E offspring may not differ from controls under nonstressed conditions, but exhibit significant deficits and/or alterations in responsiveness when challenged with stressors, hormones or pharmacological agents, or when placed in behaviorally aversive or challenging situations. At this time we cannot rule out the possibility that the pharmacokinetics of DEX were altered in E compared to PF and C animals. However, this appears unlikely as undisturbed levels of CORT following DEX administration did not differ in E, PF and C animals. Basal levels of ACTH could not be measured in the present study as the method of blood sampling used was too slow to obtain a reliable measure of undisturbed ACTH levels. However, previous data from our lab (Weinberg et al., 1996) and others (Taylor et al., 1986b; Lee et al., 1990) have shown that E animals do not differ from PF and C animals in basal ACFH levels. In contrast to previous work from our lab (Weinberg, 1988, 1992a; Weinberg & Gallo, 1982) and others (Lee et al., 1990; Nelson et al., 1986; Taylor et al., 1982), we did not observe increased CORT or ACTH responses to stress in E compared to PF and C animals in the nontreated (i.e. saline-injected) condition. One possible reason is that the potent physiological stressor used in the present study, i.e. ether stress, probably elicited a maximal response in all animals. Furthermore, CORT and ACTH were measured at only a single time point, i.e. 20 min post-stress. Our previous data suggest that the parameters of the test

138

J.A. Osborn et al.

situation, the nature and intensity of the stressor, the time course measured, and the level of the stress axis examined all play a role in determining whether E animals differ from controls in stress responsiveness and whether differential effects of fetal ethanol exposure are observed in males and females. In addition to prenatal ethanol effects, we also noted prenatal nutritional effects as well as an effect of pair-feeding itself. At 26 h post-DEX injection, PF males showed suppressed ACTH levels compared to E and C males whereas E and PF females both showed increased ACTH levels compared to C females. Previous data demonstrate that although pair-feeding provides an essential nutritional control group, pair-feeding itself is a type of experimental treatment (Weinberg, 1984). For example, pair-feeding can produce a shift in the circadian rhythm of a number of physiological variables as well as alter body and organ weights and behavior of both the maternal female and the offspring (Gallo & Weinberg, 1981; Weinberg, 1989; Weinberg & Gallo, 1982). The present data further demonstrate long-term effects of pair-feeding and highlight the importance of including an ad lib-fed control group in prenatal alcohol studies. The control of the HPA stress response occurs through multiple feedback loops occurring during three different time domains and at several different levels. CORT feedback inhibition of ACTH and CRF occurs within seconds (fast rate sensitive feedback), over 2-10 h (intermediate feedback), and over hours to days (slow feedback) (Jones & Gillham, 1988 Keller-Wood & Dallman, 1984). Fast rate sensitive feedback is thought to inhibit release of ACTH and CRF but not affect synthesis, whereas intermediate feedback is thought to decrease release of both ACTH and CRF and to decrease CRF synthesis (Keller-Wood & Dallman, 1984). Slow feedback which occurs only in pathological conditions, where CORT is elevated for days, has been shown to decrease not only ACTH release but also ACTH synthesis (Schacter et al., 1982). DEX has been shown to bind preferentially to the anterior pituitary (DeKloet et al., 1975). Thus, when DEX is given at high doses (300/~g/100 g bw), ACTH content is affected to a much greater extent than CRF content (Carnes et al., 1987). Our finding that E animals show less suppression than controls at 3 and 6 h post-DEX injection suggests that alterations in HPA responsiveness to stressors in E animals may be mediated through a defect in feedback inhibition at the level of the anterior pituitary during the intermediate feedback time domain. In contrast, Taylor et al. (1986b), found that at 10 min post-foot shock stress (during the fast feedback time domain), ACTH levels in E animals remain elevated compared to those in controls. Together, these data suggest that deficits in feedback inhibition may occur during both the fast and intermediate feedback time domains and that it is release and not synthesis of ACTH that contributes to the elevated CORT and ACTH levels seen in E animals. A diurnal variation in both basal and stress-induced HPA hormone release has been demonstrated (Bradbury et al., 1991; Kant et al., 1986). CORT, ACTH, fl-EP and fl-lipotropin release are greater in the AM than in the PM following a variety of stressors (Bradbury et al., 1991; Kant et al., 1986). The mechanism for this diurnal variation is not clearly understood at present. Recent data from Bradbury et al. (1991) indicate that circadian variations in stress-induced CORT and ACTH release are independent of basal CORT levels. However, a number of other factors may be involved. First, the rate of CORT elevation required for fast feedback inhibition (greater than 1.3 /tg/dl/min) in the PM is much faster (3-15 min) than in the AM (15-30 min). Thus, although maximal CORT levels after restraint stress in the AM and the PM may be similar, maximal CORT levels are reached approximately 12 min earlier in the PM (Bradbury et al., 1991). Second, feedback

Prenatal Ethanol Exposure and HPA Feedback

139

inhibition in the PM is less sensitive than in the AM; CORT and ACTH levels are higher in the PM than the AM after the same amount of DEX or CORT is administered (Gibbs, 1970; Wilson et al., 1983). Third, there is an increased sensitivity of the adrenal to ACTH in the PM compared to the A1V~(Ungar, 1964; Haus, 1964; Dallman et al., 1978); CORT release in response to exogenous ACTH is 2.5 times greater in the PM than in the AM (Dallman et al., 1978). Fourth, there appears to be a decrease in tissue absorption, distribution and/or metabolism of CORT in the PM compared to the AM (Gibbs, 1970; Wilson et al., 1983: Saba et al., 1963; Wilkinson et al., 1979). Plasma CORT concentrations 5 rain after a CORT injection in adrenalectomized rats are significantly higher in the PM than the AM (Wilson et al., 1983) and the half-life of CORT is 9% greater in the PM than the AM (Gibbs, 1970). Together these data suggest that the 'resetting' of feedback inhibition of the HPA axis in the PM was sufficient to unmask the altered sensitivity in E animals to the inhibitory effects of DEX. That is, in the PM, when the HPA axis is less sensitive to glucocorticoid feedback inhibition, stress CORT levels of both E males and E females, as well as stress ACTH levels of E females, were not effectively suppressed by DEX. It is possible that the sexual dimorphism of the HPA stress response also underlies the differences in sensitivity to DEX suppression seen in males and females in the AM vs. PM. Females have greater diurnal variation in plasma CORT (Ottenweller et al., 1979), higher basal CORT and transcortin levels, show greater CORT responses to stress (Kitay, 1961; Critchlow et al., 1963; Weinberg, 1988) and to ACTH (Osborn et al., 1994), and require higher doses of DEX to produce HPA suppression than males. Furthermore, hippocampal glucocorticoid receptor concentration is higher and binding affinity is lower in females than males (Weinberg & Petersen, 1991; Turner & Weaver, 1985). The sex hormones are thought to influence the HPA axis indirectly through effects on hepatic enzyme systems that inactivate CORT (Glenister & Yates, 1961; Kitay, 1961) and binding proteins (Sandberg & Slaunwhite, 1959; Slaunwhite et al., 1962), as well as through noncompetitive binding to glucocorticoid receptors to affect the rate of CORT dissociation (Chou & Luttge, 1988, Suthers et al., 1976; Svec et al., 1980). Burgess and Handa (1992) demonstrated that estrogen elevates and prolongs activation of the HPA axis after stress and interferes with Type II receptor down-regulation in the hippocampus after 4 days of administration of RU 28362, a Type II receptor-specific agonist. This sexual dimorphism may help to explain the differential responses of E males in the AM vs the PM. First, it is possible that in the present study, peak levels of CORT in males were missed in the AM but not in the PM. As noted, peak CORT levels following stress are reached more quickly in the PM than in the AM. Differences between E and control females, on the other hand, may have been seen in the AM because the rate of rise is faster in females, resulting in the detection of differences even if the peak was missed. Second, the adrenal cortex responds linearly to a log dose of ACTH (Keller-Wood & Dallman, 1984). Thus, differences in ACTH which are not statistically significant could result in significant CORT level differences. As the CORT response 1o a specific dose of ACTH is greater in females than in males (Osborn et al., 1994), small differences in ACTH release could result in larger differences in CORT levels. Furthermore, in the PM, an increased adrenal sensitivity to ACTH could have resulted in higher stress CORT levels in both E males and E females. This latter possibility is supported by our finding that differences in ACTH levels among E, PF and C animals were less robust than differences in CORT levels following stress. It is also possible that altered neurotransmitter release may be involved in mediating the HPA hyperresponsiveness of E animals. For example, following restraint stress, cortical and

140

J.A. Osborn et al.

hypothalamic norepinephrine (NE) content is lower in E animals compared to controls (Rudeen & Weinberg, 1993). NE and epinephrine have been shown to stimulate CRF release in a dose-dependent manner (Plotsky, 1987). In addition, it has been shown that depletion of hypothalamic NE and serotonin enhances the inhibitory effects of DEX on the CORT response to ether stress (Feldman & Weidenfeld, 1991). If lower hypothalamic NE levels in E animals are indicative of increased NE turnover post-stress, it is possible that prenatal ethanol effects on NE regulation of CRF secretion may play a role in HPA hyperactivity in E offspring. Consistent with this hypothesis, Lee et al. (1990) demonstrated increased CRF biosynthesis and expression along with an increased ACTH response to stressors in E animals compared to controls. Thus, altered feedback inhibition of neurotransmitterstimulated CRF secretion may also play a role in the stress hyperresponsiveness of E animals. However, the finding that E animals demonstrate altered responses to physiological, physical and neurogenic stressors suggests that more than one neural pathway may be affected by prenatal ethanol exposure (Nelson et al., 1984; Taylor et al., 1981, 1987; Weinberg & Galio, 1982; Weinberg, 1988, 1992a). Finally, these data may be of clinical importance. Children prenatally exposed to alcohol are hyperactive, uninhibited and impulsive in behavior, and have attention deficits which may reflect an inability to inhibit responses (Streissguth et al., 1983, 1985; Streissguth, 1986). These behavioral deficits are particularly noticeable in stressful situations (Streissguth, 1986). Recently, it has been shown that maternal drinking during pregnancy is associated with higher post-stress cortisol levels in infants (Jacobson et al., 1993). CRF, ACTH and glucocorticoids are known to modulate behavior during stress (McEwen et al., 1986). Thus, it is possible that sustained increases in hormones of the HPA axis could play a role in mediating the increased hyperactivity and behavioral arousal that are observed in fetal alcohol-exposed children.

Acknowledgements: This research was funded by grant AA07789 from the National Institute on Alcohol Abuse and Alcoholism. J. A. Osborn was supported by a Medical Research Council of Canada Studentship.

REFERENCES Angelogianni P, Gianoulakis C (1989) Prenatal exposure to ethanol alters the ontogeny of the flendorphin response to stress. Alcohol Clin Exp Res 13:564-571. Bonnischsen RK, Theorell H (1951) An enzymatic method for the micro-determination of ethanol. Scand J Clin Lab Invest 3:58-62. Bradbury MJ, Cascio CS, Scribner KA, Dallman MF (1991) Stress-induced adrenocorticotropin secretion: Diurnal responses and decreases during stress in the evening are not dependent on corticosterone. Endocrinology 128:680--688. Burgess LH, Handa RJ (1992) Chronic estrogen-induced alterations in adrenocorticotropin and corticosterone secretion, and glucocorticoid receptor-mediated functions in female rats. Endocrinology 131:1261-1269. Carnes M, Barksdale CM, Kalin NH, Brownfield MS, Lent SJ (1987) Effects of dexamethasone on central and peripheral ACTH systems in the rat. Neuroendocrinology 42:160-164. Chou YC, Luttge WC (1988) Activated type-II receptors in the brain cannot rebind glucocorticoids: Relationship to progesterone's antiglucocorticoid actions. Brain Res 440:67-78. Critchlow V, Liebelt RA, Bar-Sela M, Mountcastle W, Lipscomb HS (1963) Sex difference in resting pituitary-adrenal function in the rat. Am J Physiol 205:807-815. Dallman MF, Engeland WC, Rose JC, Wilkinson CW, Shinsako J, Siedenburg F (1978) Nycthemeral rhythm in adrenal responsiveness to ACTH. Am J Physiol 235: R210-R218. Davidson JM, Jones LE, Levine S (1968) Feedback regulation of adrenocorticotropin secretion in

Prenatal Ethanol Exposure and HPA Feedback

141

"basal' and 'stress' conditions: Acute and chronic effects of intrahypothalamic corticoid implantation. Endocrinology 82:655-663. DeKloet ER, Wallach G, McEwen BS (1975) Differences in corticosterone and dexamethasone binding to rat brain and pituitary. Endocrinology 96:598-609. Feidman S, Weidenfeld JW (1991) Depletion of hypothalamic norepinephrine and serotonin enhances the inhibitory effects of dexamethasone on the adrenocortical response to ether stress. Psychoneuroendocrinology 16:397-405. Gallo PV, Weinberg J (1981) Corticosterone rhythmicity in rat: Interactive effects of dietary restriction and schedule of feeding. J Nutr 111:208-218. Gibbs FP (1970) Circadian variation of ether-induced corticosterone secretion in the rat. Am J Physiol 219:288-292. Glenister DW, Yates FE (1961) Sex difference in the rate of disappearance of corticosterone-4-Cl4 from plasma of intact rats: Further evidence for influence of hepatic delta-4-steroid hydrogenase activity on adrenalcorticoid function. Endocrinology 68:747-758. Haus E (1964) Periodicity in response and susceptibility to environmental stimuli. Ann NY Acad Sci 117:293-315. Jacobson SW, Jacobson JL, Bihun JT, Chiodo L, Sokol RJ (1993) Effects of prenatal alcohol exposure on poststress cortisol levels in infants. Alcohol Clin Exp Res 17:456 Jones MT, Gillham B (1988) Factors involved in the regulation of adrenocorticotropin hormone//3lipotropin hormone. Physiol Rev 68:743-818. Kakihana R, Butte JC, Moore JA (1980) Endocrine effects of maternal aicoholization: Plasma and brain testosterone, dihydrotestersone, estradiol, and corticosterone. Alcohol Clin Exp Res 1:57-61. Kaneko M, Kaneko K, Shinsako J, Dallman MF (1981) Adrenal sensitivity to adrenocorticotropin varies diurnally. Endocrinology 109:70-75. Kant G J, Mougey EH, Meyerhoff JL (1986) Diurnal variation of neuroendocrine response to stress in rats: Plasma ACTH, [:LLPH, corticosterone, prolactin, and pituitary cyclic AMP responses. Neuroendocrinology 43:383-390. Keller-Wood ME, Dallman MF (1984) Corticosteroid inhibition of ACTH secretion. Physiol Rev 5:124. Kitay JL (1961) Sex differences in adrenal cortical secretion in the rat. Endocrinology 68:818-824. Lee S, Imaki T, Vale W, Rivier C (1990) Effects of prenatal exposure to ethanol on the activity of the hypothalamic-pituitary-adrenal axis of the offspring: Importance of the time of exposure to ethanol and possible modulating mechanisms. Mol Cell Neurosci 1:168-177. McEwen BS, DeKloet ER, Rostene W (1986) Adrenal steroid receptors and actions in the nervous system. Physiol Rev 66:1121-1188. Nelson LR, Redei E, Liebeskind JC, Branch BJ, Taylor AN (1985) Corticosterone response to dexamethasone in fetal ethanol exposed rats. Proc West Pharmacol Soc 28:299-302. Nelson LR, Taylor AN, Lewis JW, Poland RE, Redei E, Branch BJ (1986) Pituitary-adrenal responses to morphine and footshock stress are enhanced following prenatal alcohol exposure. Alcohol Clin Exp Res 10:397-402. Nelson LR, Taylor AN, Redei E, Branch BJ, Lewis JW (1984) Fetal exposure to ethanol enhances corticosterone release to footshock stress. Alcohol Clin Exp Res 8:109-114. Nemeroff CB, Ottenweller JE, Cook JA, Pitman D, McCarty R, Tapp WN (1988) The role of corticotropin-releasing factor in the pathogenesis of major depression. Pharmacopsychiatry 21:7682. Orth DN (1979) Adrenocorticotropin hormone (ACTH). Anonymous Methods of Hormone Radioimmunoassay, 2nd Edn. Academic Press, New York, pp 245-250. Osborn JA, Stelzl GE, Weinberg J (1994) Fetal ethanol effects on adrenal sensitivity to adrenocorticotropic hormone (ACTH). Alcohol Clin Exp Res 18:407 Ottenweller JE, Mier AH, Russo AC, Frenzlee ME (1979) Circadian rhythms of plasma corticosterone binding activity in rat and mouse. Acta Endocrinol 91:150-157. Plotsky PM (1987) Regulation of hypophysiotropic factor mediating ACTH secretion. Ann NY Acad Sci 512:205-217. Rivier C, Brownstein M, Spiess J, Rivier J, Vale W (1982) In vivo corticotropin-releasing factorinduced secretion of adrenocorticotropin, fl-endorphin, and corticosterone. Endocrinology 110:272278.

142

J.A. Osborn et al.

Rudeen PK, Weinberg J (1993) Prenatal ethanol exposure: Changes in regional brain catecholamine content following stress. J Neurochem 61:1907-1915. Saba GC, Saba A, Carnicelli A, Marescotti V (1963) Diurnal rhythm in the adrenal cortical secretion and in the rate of metabolism of corticosterone in the rat. Acta Endocrinol Copenh 44:409-412. Sandberg AA, Slaunwhite WRJ (1959) Transcortin: A corticosteroid-binding protein of plasma II levels in various conditions and effects of estrogens. J Clin Invest 38:1290-1307. Sapolsky RM, Krey LC, McEwen BS (1984) Glucocorticoid-sensitive hippocampal neurons are involved in terminating the adrenocortical stress response. Proc Natl Acad Sci 81:6174--6177. Schacter BS, Johnson LK, Baxter JD, Roberts JL (1982) Differential regulation by glucocorticoids of proopiomelanocortin mRNA levels in the anterior and intermediate lobes of the rat's pituitary. Endocrinology 110:1442-1447. Slaunwhite WRJ, Lockie GN, Sandberg AA (1962) Inactivity in vivo of transcortin-bound cortisol. Science 135:1062-1063. Streissguth AP (1986) The behavioral teratology of alcohol: Performance, behavioral, and intellectual deficits in prenatally exposed children. In: West JR (Ed.) Alcohol and Brain Development. Oxford University Press, New York, pp 3-44. Streissguth A_P, Barr HM, Martin DC (1983) Maternal alcohol use and neonatal habituation assessed with the Brazelton scale. Child Dev 54:1109-1118. Streissguth AP, Clarren SK, Jones KL (1985) Natural history of the fetal alcohol syndrome: A 10-year follow-up of eleven patients. Lancet 10:85-92. Suthers MB, Pressley LA, Funder JW (1976) Glucocorticoid receptors: Evidence for a second nonglucocorticoid binding site. Endocrinology 99:260--269. Svec F, Yeakley J, Harrison RWl (1980) Progesterone enhances glucocorticoid dissociation from AtT20 glucocorticoid receptor. Endocrinology 107:566--572. Taylor AN, Branch BJ, Kokka N, Poland RE (1983) Neonatal and long-term neuroendocrine effects of fetal alcohol exposure. Monogr Neural Sci 9:140-152. Taylor AN, Branch BJ, Liu SH, Kokka N (1982) Long-term effects of fetal ethanol exposure on pituitary-adrenal response to stress. Pharmacol Biochem Behav 16:585-589. Taylor AN, Branch B J, Liu SH, Weichmann AF, Hill MA, Kokka N (1981) Fetal exposure to ethanol enhances pituitary-adrenal and temperature responses to ethanol in adult rats. Alcohol Clin Exp Res 5:237-246. Taylor AN, Branch BJ, Nelson LR, Lane LA, Poland RE (1986a) Prenatal ethanol and ontogeny of pituitary-adrenal responses to ethanol and morphine. Alcohol 3:255-259. Taylor AN, Branch BJ, Randolph D, Hill MA, Kokka N (1987) Prenatal ethanol exposure affects temperature responses of adult rats to pentobarbital and diazepam alone and in combination with ethanol. Alcohol Clin Exp Res 11:254--260. Taylor AN, Branch B J, Van Zuylen JE, Redei E (1986b) Prenatal ethanol exposure alters ACTH stress responsiveness in adult rats. Alcohol Clin Exp Res 10:120. Turner BB, Weaver DA (1985) Sexual dimorphism of glucocorticoid binding in rat brain. Brain Res 343:16-23. Ungar F (1964) In vitro studies of adrenal-pituitary circadian rhythm in the mouse. Ann NY Acad Sci 117:374-385. Weinberg J (1984) Nutritional issues in perinatal alcohol exposure. Neurobehav Toxicol Teratol 6:261-269 Weinberg J (1985) Effects of ethanol and maternal nutritional status on fetal development. Alcohol Clin Exp Res 9:49-55. Weinberg J (1988) Hyperresponsiveness to stress: Differential effects of prenatal ethanol on males and females. Alcohol Clin Exp Res 12:647--652. Weinberg J (1989) Prenatal ethanol exposure alters adrenocortical development of offspring. Alcohol Clin Exp Res 13:73--83. Weinberg J (1992a) Prenatal ethanol effects: Sex differences in offspring stress responsiveness. Alcohol 9:219-223. Weinberg J (1992b) Prenatal ethanol exposure alters adrenocortical response to predictable and unpredictable stressors. Alcohol 9:427-432. Weinberg J, Bezio S (1987) Alcohol-induced changes in pituitary-adrenal activity during pregnancy. Alcohol Clin Exp Res 11:274-280.

Prenatal Ethanol Exposure and HPA Feedback

143

Weinberg J, Gallo PV (1982) Prenatal ethanol exposure: Pituitary-adrenal activity in pregnant dams and offspring. Neurobehav Toxicol Teratol 4:515-520. Weinberg J, Nelson LR, Taylor AN (1986) Hormonal effects of fetal alcohol exposure. In: West JR (Ed.) Alcohol and Brain Development. Oxford University Press, New York, pp 310-342. Weinberg J, Petersen TD (1991) Effects of prenatal ethanol exposure on glucocorticiod receptors in rat hippocampus. Alcohol Clin Exp Res 15:711-716. Weinberg J, Taylor AN, Gianoulakis C (1996) Fetal ethanol exposure: hypothalamic-pituitaryadrenal and fl-endorphin responses to repeated stress. Alcohol Clin Exp Res 20:122-131. Wilkinson CW, Shinsako J, Dallman MF (1979) Daily rhythm in adrenal responsiveness to adrenocorticotropin are determined primarily by the time of feeding in the rat. Endocrinology 104:350-359. Wilson MM, Greer SE, Greer MA (1983) Nycterohemeral difference in inhibition of stress-induced ACI~H in adrenalectomized rats. Am J Physiol 244:E186-189.