Estradiol and chlordecone interactions with the estradiol receptor

TOXlCOLoGY

AND APPLIED

PHARMACOLWY

98,4

13-42 I(1 989)

Estradiol and Chlordecone Interactions with the Estradiol Receptor’ JUDYWILLIAMS,KRISECKOLS,ANDLYNDAUPHOUSE* Department ofBiology, Texas Woman’s University, Demon, Texas 76204

Received July 25* 1988: accepted December 18, 1988 Estradiol and Chlordecone Interactions with the Estradiol Receptor. WILLIAMS, J., ECKOLS, and UPHOUSE, L. (1989). Toxicoi. Appl. Pharmacol. 98,4 13-42 I. The in vivo and in vitro effects of the chlorinated pesticide, chlordecone, on the estradiol receptor of adult ovariectomized CDF-344 rats were examined. Chlordecone’s competition with [3H]estradiol for binding to the estradiol receptor in vitro was similar whether receptors were derived from neural or uterine tissue. Zn viva, chlordecone led to nuclear retention of estradiol receptors in both tissues, but in comparison to estradiol, the latency for such retention was slower following chlordecone. Nevertheless, chlordecone- and estradiol-treated animaIs eventually achieved the same degree of nuclear retention in uterine tissue. In brain, estradiol-like levels of nuclear retention were never found in chlordecone-treated females. Estradiol receptor replenishment following chlordecone or estradiol was similar in uterine tissue but not in brain. Chlordecone’s lesser ability to mimic estradiol’s receptor dynamics in the brain may contribute to the apparent tissue differences in chlordecone’s estrogenicity. 0 1989 Academic Press, IK K.,

tive dysfunctions. Since estradiol decreases LH secretion (Karsch, 1987), probably via an estradiol receptor-mediated mechanism (Walters, 1985), these findings are consistent with chlordecone’s ability to bind to the intracellular estradiol receptor. There are many reproductively relevant effects of estradiol which are not mimicked by chlordecone exposure. For example, estradiol has both positive and negative feedback on pituitary secretion of LH, but chlordecone produces only the negative feedback (Uphouse et ul., 1984). Moreover, chlordecone does not substitute for estradiol in facilitating sexual receptivity of ovariectomized rats (Uphouse et ul., 1986). Furthermore, chlordecone reduces sexual behavior of estrogenprimed, ovariectomized rats (Uphouse et ui., 1986) and of intact, cycling female rats (Uphouse, 1986). Inhibition, rather than facilitation, of pituitary gonadotroph’s secretion of LH and follicle stimulating hormone (FISH) in response to gonadotropin releasing hor-

Chlordecone (Kepone) is a chlorinated pesticide which, although best known for its neurological effects such as severe tremor, also affects reproductive capacity. Since the Palmiter and Mulvihill (1978) and Hammond et ul. (1979) reports that chlordecone could compete with estradiol for binding to the uterine and oviduct estradiol receptor, interaction with the estradiol receptor has been postulated as a mechanism for reproductive deficits (Eroschenko and Palmiter, 1980; Uphouse, 1986). Chronic treatment with chlordecone produces oligospermia and hypomotile sperm in human males (Taylor et ul., 1978) and a failure to ovulate in female rodents (Huber, 1965: Gellert et ul., 1978). In females, the pesticide can reduce serum levels of luteinizing hormone (LH) (Uphouse etul., 1984) which may account for these reproduc’ This work was supported by ES0335 1 and in part by a TWU Institutional grant to L.U. ’ To whom correspondence should be addressed.

..-

0041-008x/89

$3.00

414

WILLIAMS,

ECKOLS.

mone (GnRH) and antagonism of estrogen’s facilitation have led to the suggestion that chlordecone may function as an estrogen antagonist in the brain and pituitary (Huang and Nelson, 1986) despite its apparent “estrogenicity” in uterine tissue. Tissue differences in the consequences of estradiol receptor binding compounds have been observed for several estrogen target tissues. Tamoxifen and CI 628 have both estrogenic and antiestrogenic properties when rat uterotrophic effects are examined (Jordan et d., 1978; Sutherland and Foe, 1979). Only antiestrogenie effects are seen in human breast cancer cells (Jordan and Robinson, 1987) or in rat pituitary cells, when GnRH-stimulated release of LH is considered (Adashi et al., 198 1). Estrogen antagonists resemble estradiol in binding to the intracellular receptor and retaining it in the nucleus (Kurl and Morris, 1978). However, they generally attenuate the magnitude but prolong the duration of nuclear retention (Roy et al., 1979b) and show minimal, if any, evidence of receptor replenishment (Kurl and Morris, 1978). Although it has been generally assumed that chlordecone’s interaction with and nuclear retention of estradiol receptors in brain is similar to that reported for uterine tissue, the validity of this assumption has not been evaluated. The emerging evidence for tissue differences in chlordecone’s apparent “estrogenicity” following systemic treatment with the pesticide has made examination of this assumption essential. In the following report, chlordecone’s in vitro interactions with central nervous system (CNS) and uterine estradiol receptors were examined. In addition, the degree of nuclear retention of estradiol receptors was examined in CNS and in uterus following in vivo treatment of ovariectomized rats with estradiol or chlordecone. The time course, extent, and duration of nuclear retention and the degree of receptor replenishment were examined. MATERIALS

AND

METHODS

Materials Bisbenzimidazole (Hoechst Dye 33258) was obtained from Polysciences (Washington, PA). Chlordecone (85-

AND UPHOUSE 95% pure, remainder mirex and water) was obtained from Radian Laboratories (Austin, TX). EcoLite scintillation cocktail came from WestChem (San Diego, CA). [3H]Estradiol (101 Ci/mmol) was from DuPont/New England Nuclear (Wilmington, DE). Glycerol, estradiol 170, estradiol benzoate, Sephadex LH 20-100, sesame oil, and calf thymus DNA were purchased from Sigma (St. Louis, MO). All other supplies were purchased from Fisher Scientific (Fair Lawn, NJ). Animals Adult ovariectomized female CDF-344 rats were used in these studies. Animals were either purchased from Charles River Laboratories (Wilmington, MA) or bred in our colony from stock obtained from Charles River Laboratories. At 60 days of age rats were bilaterally ovariectomized under sodium pentobarbital anesthesia. Animals were housed four or five per cage, with ad lib. access to food and water, in a colony room with a 12-l 2 hr lightdark cycle (lights on between 6 AM and 6 PM local time). Treatment ofAnimals A. In vitro studies. When chlordecone’s competition with [3H]estradiol for binding to the estradiol receptor was examined, animals were injected with 5 pg estradiol benzoate (sc) 1 week post-ovariectomy and were sacrificed 5 days after injection. Tissue samples from several animals (eight per brain-pituitary; four per uterus) were pooled and used for examination of chlordecone’s inhibition of [3H]estradiol binding. B. In vivo studies. When in vivo effects of estradiol or chlordecone on the estradiol receptor were compared, animals were injected with 10 fig estradiol benzoate (SC) 1 week post-ovariectomy to assure high levels of unoccupied receptors. Two weeks after ovariectomy, animals were injected with 25 pg estradiol benzoate (sc), 75 mg/ kg chlordecone (ip), or oil and sacrificed 2,4, 12, 20, or 36 hr later. The dose of chlordecone chosen for these studies represents a dose reliably capable of producing reproductive effects and tremor activity within 6 to 8 hr. Lower doses produce greater variability of response, and doses greater than 100 mg/kg approach the LD50 for female rats. Specific binding of [3H]estradiol was determined separately for nuclear and cytosol fractions. Preparation ofNuclear and Cytosol Fractions Estradiol receptors were assayed with a modification of the procedure of Clark and Roy ( 1983). Animals were sacrificed by decapitation at 0900 hr, and the brain, pituitary, and uterus were rapidly removed. All following procedures were performed at CC. Brains were dissected

ESTRADIOL

RECEPTOR

according to McEwen and Pfaff ( 1970) into preoptic area and hypothalamus. Pituitary, preoptic area, and hypothalamus were combined and homogenized together in buffer I (0.32 IVIsucrose, 3 mM MgC&, 10 mM KHIPOd, and 10% glycerol; pH 7.4) with a motor-driven Teflon pestle. Uteri were stripped of connective tissue and fat, minced, and homogenized in buffer 1 using a Brinkman polytron. Homogenate aliquots were taken for determination of DNA content according to the method of Labarca and Paigen (1980). Homogenates were centrifuged at 8OOg in a Beckman Model J-2 1 centrifuge for 5 min to separate cytosol fraction (supernatant) and nuclear fraction (pellet). The supernatant was then centrifuged at 105,OOOgfor 40 min in a Beckman L5-65 ultracentrifuge to obtain the high-speed cytosol fraction which was used immediately for estradiol receptor analysis. The nuclear fraction was homogenized consecutively in buffer 1 containing 0.2% Triton and in buffer 2 (1.2 M sucrose, I mM MgClz, IO mM KH2POb, and 10% glycerol; pH 7.4) and then centrifuged at 25,OOOgfor 20 min. Tbe supematant was discarded and the pellet rehomogemzed consecutively in equal volumes of buffer 3 (10 rnM This and 10% glycerol: pH 7.6) and buffer 3 plus 0.8 M KC1 to yield a final concentration of 0.4 M KCl. After rehomogenization, samples were incubated on ice for 10 min and then centrifuged at 18,OOOgfor 10 min. The supernatant containing the KC1 extracted receptors was used for the binding assay.The pellet was rehomogenized in buffer 1 and aliquots removed for determination of DNA content (Labarca and Paigen, 1980).

E.ctradiol Receptor ,4ssaJ Estradiol receptors were assayed in 0.5 ml of buffer 4 (0.4 M KCl, 10 rnM Tris, and 10% glycerol; pH 7.4) containing 1 nM [3H]estradiol, plus or minus 10 nM unlabeled estradiol as competitor. Quantity and cellular location of receptors were determined in females treated in vivo with estradiol or with chlordecone. When chlordecone and estradiof were compared as competitors with [‘H]estradiol, incubation took place in buffer 4 containing 1 nM [3H]estradiol plus or minus unlabeled estradiol or chlordecone of various concentrations ( 10e4 to lo-” M). After a 4-hr incubation at room temperature, the contents ofthe assay tubes were filtered on Sephadex LH20 columns. An incubation time of 4 hr was chosen because this permits equilibrium binding of [3H]estradiol. In preliminary studies with chlordecone as competitor, binding was examined at various incubation times from I to 12 hr. Chlordecone’s inhibition of [3H]estradiol binding was similar at 4 and at 12 hr, so the 4-hr incubation allowed for the best comparison between estradiol and chlordecone as competitors. Following incubation, radioactivity in the eluant, containing [3H]estradiol bound to receptor, was quantified by liquid scintillation counting. For determination of specific [3H]estradiol

41s

INTERACTIONS

binding, cpm in the presence of the 10 nM unlabeled estradiol competitor (nonspecific binding) was subtracted from that in the absence of competitor (total binding). Cytosol data are expressed as femtomoles [ ‘Hlestradiol bound per 100 fig homogenate DNA. Nuclear data are expressed as femtomoles [3H]estradiol bound per 100 pg nuclear pellet DNA. Statistics When more than two groups were compared. data were evaluated by analyses of variance followed by Dunnett’s test for comparison of individual means to the appropriate control. When only two groups were compared, Student’s t test was used (Zar, 1984). An cxlevel of 0.05 was required for rejection of the null hypothesis.

RESULTS In Vitro E#ects In agreement with Hammond et ul. ( 1979), chlordecone competed in vitro with [3HJestradiol for binding to the estradiol receptor. Similar competition was seen in uterus and in brain. The IC50 for chlordecone (approximately 1.O X 10e5~) was about lOOO-fold greater than that of estradiol (Fig. 1). In Vivo E&cts in Brain Zn vivo effects of chlordecone and estradiol on the location of brain estradiol receptors are shown in Figs. 2 and 3. Two hours after estradiol injection, there was a sharp increase in nuclear binding and a decrease in cytosol binding. The location of binding then remained relatively constant through 36 hr. Analysis of variance indicated a significant effect of cellular location (F,,, I2 = 57.7 1, p G 0.001) and time (Fs,l,z = 5.13, p G 0.001) following injection. Relative to the 0 control, nuclear concentrations of estradiol receptor were significantly elevated at each time point investigated (all q 2 6.35, p G 0.05). Cytosol receptors showed the inverse profile with estradiol receptor concentrations significantly lower than that of the control at all time

416

WILLIAMS,

ECKOLS,

points (all q 2 2.74, p < 0.05). Because of the change in location between the 0 time and all other time points, the interaction term was also significant (F5,, ,z = 36.60, p G 0.001). For chlordecone-treated animals, analysis of variance also demonstrated a significant time by location interaction (R’5,10z= 8.28, p G 0.001). The main effect of location narrowly escaped statistical significance (F,,,Oz = 3.81, p = 0.059) (Fig. 3). The profile for nuclear binding was similar to that of estradiol in that once binding significantly increased in the nucleus (about 12 hr following treatment), it remained elevated thereafter (all q > 3.32, p G 0.05). Cytosol showed the inverse picture, but there was never a significant depletion of receptors from the cytosol of chlordecone-treated females (all q < 2.20, p > 0.05). In contrast to the estradiol treatment, there was no overall effect of time Vu02 = 1.06, p = 0,38). In Vivo Efects in Uterus The effect of estradiol on [3H]estradiol binding in uterine tissue is shown in Fig. 4. Analysis of variance demonstrated a significant effect of time (FS,,oJ = 3.38, p G 0.01) and time by location interaction (F,,,od = 24,74, p G 0.00 1). The main effect of location was not significant (F,,,oa = 2.09, p 3 0.05) even though nuclear binding increased and cytosol binding decreased at all times after 0 (respectively q > 3.60; q 2 1.97; p G 0.05). There appeared to be a secondary change in cellular location of the receptors between 20 and 36 hr. The relatively constant nuclear binding between 2 and 20 hr showed a further increase at 36 hr. Cytosol binding, rapidly depleted by 2 hr, appeared to return toward control levels by 20 hr. Thereafter, a secondary decline in cytosol receptors accompanied the increase in nuclear binding. Uterus from chlordecone-treated animals also showed significant effects of time (F5,~d = 3.24, p = 0.01) and time by location interaction (F5,qd = 7.34, p s 0.001) (Fig. 5). In

AND UPHOUSE

04

, IO

-

, 6

Log

.

, 6

Competitor

.

, 7

, 6

, 5

.

, 4

T

Concentration

FIG. I. Inhibition of cytosol [3H]estradiol binding by estradiol and chlordecone. Animals were injected with 5 pg estradio1 benzoate I week post-ovariectomy and sacrificed 5 days later. Aliquots of the cytosol fraction were incubated with [3H]estradiol plus or minus varying concentrations of unlabeled estradiol or chlordecone. Binding in the absence of competitor was taken as 100% Binding in the presence of competitor was expressed as a percentage ofthis maximum binding. Data represent the mean percentage of two to four assays per competitor concentration. (A) Brain: Hypothalami, preoptic areas, and pituitaries from eight animals were pooled. (B) Uterus: Uteri from four animals were pooled.

contrast to the results following estradiol treatment, there was a significant effect of location (F,,gd = 11.99, p = 0.001) on [3H]estradiol binding. This difference between estradiol and chlordecone probably reflects the slower time course and lesser amount of cytosol depletion in chlordecone-treated females. Nuclear binding was not significantly increased by chlordecone until 12 hr post-injection (qgdqS= 3.08, p 6 0.05) but remained elevated thereafter (qgd,Sa 3.44, p < 0.05). Similar to the profile observed in estradiol-treated animals, chlordecone treatment appeared to

ESTRADIOL

Tima After Estradiol

RECEPTOR

(hrs.)

FIG. 2. Specific [3H]estradiol binding to brain nuclear and cytosol factions following estradio1 treatment. Animals were injected with IO pg estradiol benzoate I week post-ovariectomy and sacrificed 2 weeks post-ovariectomy. Animals were injected SCwith 25 pg estradiol benzoate or with oil vehicle 2, 4, 12, 20, or 36 hr prior to sacrifice. The 0 time point represents the pooled means of all oil-injected controls. Hypothalamus, preoptic area, and pituitary from individual animals were used for [3H]estradioi binding. Data represent the means ? SE of four to eight animals at the 2- to 36-hr time points.

417

INTERACTIONS

from chlordecone-treated females was less than that for estradiol treated animals (F,.5, = 29.74, p < 0.001). This was evident at all time points, so neither time (F4,S1 = 0.34, j? aO.05) nor the time by treatment interaction In V4s, = 1.286, p > 0.05) was significant. uterus both main effects of time and treatment were significant (respectively, F4,43 = 4.17, p =z 0.006; and F,,43 = 31.97. p < 0.001). Higher binding was present at 2 hr in the estradiol-treated animals relative to that in chiordecone-treated animals (qd3.> = 3.35, p =S0.05) but by 4 hr, no differences were present between the two treatment conditions b > 0.05). This differential change over time produced a significant time by treatment interaction (F4,43= 3.19,~ < 0.05). At 36 hr, the longest time point examined, the sums of nuclear and cytosol (Total) bind-

4f IY

produce a secondary increase in nuclear receptors between 20 and 36 hr. Cytosol binding decreased between 0 and 12 hr, but at no time was cytosol depletion significant b > 0.05). The prior cellular location analyses did not enable a comparison between chlordecone and estradiol in the number of receptors retained in the nucleus. A two-way analysis of variance (time by treatment) was performed on the fmol [3H]estradiol bound/l00 pg DNA in nuclei from estradiol and chlordecone-treated animals. The amount of specific [3HJestradiol binding in nuclei from chlordecone-treated females was divided by the amount of specific [3H]estradiol binding in nuclei from estradiol-treated females. The resulting ratios are pIotted in Fig. 6. In brain, binding of [3H]estradiol to nuclear fractions

-t3-

0

10

20

Time After Chlordecone

30

nucleue

40

(hre.)

FIG. 3. Specific [3H]estradiol binding to brain nuclear and cytosol fractions following chlordecone treatment. Animals were injected with IO pg estradiol benzoate 1 week post-ovariectomy and sacrificed 2 weeks postovariectomy. Animals were injected sc with 75 mg/kg chiordecone or with oil vehicIe 2,4, 12,20. or 36 hr prior to sacrifice. The 0 time point represents the pooled means of all oil-injected controls. Hypothalamus, preoptic area, and pituitary from individual animals were used for [3H]estradiol binding. Data represent the means 2 SE of four to eight animals at the 2- to 36-hr time points.

418

WILLIAMS.

0

10 Time After

20 Estradiol

30

ECKOLS,

40

(hrs.)

FIG. 4. Specific [3H]estradiol binding to uterine nuclear and cytosol fractions following estradiol treatment. Animals were injected with 10 pg estradiol benzoate 1 week post-ovariectomy and sacrificed 2 weeks postovariectomy. Animals were injected with 25 pg estradiol benzoate or with oil vehicle 2.4, 12, 20, or 36 hr prior to sacrifice. The 0 time point represents the pooled means of all oil-injected controls. Data represent the means ? SE of four to seven animals at the 2- to 36-hr time points.

ing were compared as an assessment of the effectiveness of estradiol and chlordecone in increasing total estradiol receptors of brain and uterus. Treatment effects on Total estradial receptors in brain tissue were not statistically significant (Fig. 7A) (Fz,lg = 3.12, p = 0.07). The treatment effect in uterus was considerably more robust (Fig. 7B) (F*,,s = 5.80; p G 0.02) with both estradiol- and chlordecone-treated animals different from the oil-injected controls (respectively, t = 4.589 and 3.147; p G 0.03). Chlordeconeand estradiol-treated animals did not differ from each other ($ > 0.05). DISCUSSION A major objective of the present study was to compare the effects of chlordecone on estradiol receptors in brain and uterine tissue. The majority of laboratory investigations

AND UPHOIJSE

have suggested that chlordecone acts as a “weak” estrogen (Huber, 1965; Eroschenko and Palmiter, 1980) binding to the estradiol receptor, but with relatively low affinity (Palmiter and Mulvihill, 1978; Hammond et al., 1979). More recent studies, designed to demonstrate this estrogenicity in neural tissue, have suggested that chlordecone’s putative “estrogenicity” might be less apparent in neural than in peripheral tissue (Uphouse et al., 1986). The present studies suggest a possible explanation for the tissue differences. In vitro effects of chlordecone were assessed by its competition with [3H]estradiol for binding to nuclear and cytosol estradiol receptors. These studies confirm prior reports that chlordecone interacts with the estradiol receptor in the uterus (Hammond et al., 1979) and extend this observation to neural tissue. Chlordecone’s ability to compete in vitro with [3H]estradiol for binding to estradiol receptors was similar whether receptors were obtained from brain or uterine tissue.

o+

. 0

’ 10

-

’

*

20

Time After Chlordecone

’

30

-

’

40

(hrs.)

FIG. 5. Specific [3HJestradiol binding to uterine nuclear and cytosol fractions following chlordecone treatment. Animals were injected with 10 pg estradiol benzoate 1 week post-ovariectomy and sacrificed 2 weeks post-ovariectomy. Animals were injected with 75 mg/kg chlordecone or with oil vehicle 2,4, 12,20, or 36 hr prior to sacrifice. The 0 time point represents the pooled means of all oil-injected controls. Data represent the means * SE of four to seven animals at each time point.

ESTRADIOL

El

0.0 a7 *

IO

20 Tim

After Tremnw~t

SO

RECEPTOR

40

(hrs.)

FIG. 6. Relative nuclear [3H]estradiol binding following chlordecone or estradiol treatment. The data plotted in Figs. 2-S were used to directly compare estradiol- and chlordecone-treated females. Nuclear fmol [3H]estradiol bound/ 100 pg DNA of chlordecone-treated rats was divided by nuclear fmol ]3H]estradiol bound/ 100 pg DNA of estradiol-treated rats,

The in vivu effects of chlordecone showed both similarities and differences between uterine and neural tissue. In each tissue, chlordecone retained estradiol receptors in the nucleus, but the latency of onset of this retention was longer and the amount of retention was lower for chlordecone than forestradiol. The slower nuclear retention seen in both tissues is reminiscent of the effects of antiestrogens (Kurl and Morris, 1978; Roy et ul., 1979b). However, for brain tissue this may reflect a longer latency for chlordecone to reach neural tissue. Three to five hours are required for brain concentrations of chlordecone to plateau (unpublished observations). In contrast with uterine tissue, chlordeconetreated animals never achieved estradiol-like levels of nuclear retention in brain tissue and chlordecone’s biological effects are considerably different from those of estradiol in brain tissue. Since nuclear retention of estradiol receptors is important for the modulation of estrogen-mediated activities (Clark and Peck. 1974). these differences between estradiol and chlordecone may be important to the differences in chlordecone’s apparent estrogenicity in brain and uterus. When chlordecone’s Gi vjvo effects on uterine or oviduct tis-

419

1NTERACTfONS

sue have been compared to estradiol. the pesticide closely resembles estradiol. Chlordecone leads to an increase in ovalbumin and conalbumin mRNA and protein synthesis in chick oviduct (Palmiter and Mulvihill, 1978), and increases uterine progesterone receptors and uterine weight in rats (Hammond et d.. 1979). When neural effects are evaluated, chlordecone’s estrogenicity is less evident. Chlordecone fails to mimic estradiol’s positive feedback on LH release (Uphouse et ~1.. 1984) and does not substitute for estradiol in facilitating sexual behavior in ovariectomized female rats (Uphouse et ~1.. 1986). Estradiol-like priming of ovariectomized females is essential for lordosis behavior (Edwards et al., 1968) and may require induction of CNS progesterone receptors which are then available for the facilitating action of

Tr*atm*nt

FIG. 7. Total [3H]estradiol binding in brain and uterus 36 hr after estradiol or chlordecone treatment. The data plotted in Figs. 2-S were used to compare estradiof receptor replenishment at 36 hr post-injection. Data are Total (nucleus + cytosol) [‘Hlestradiol bound/l00 fig DNA. (A) Brain: Estradiol, chlordecone, and control data represent the means 2 SE of seven, six, and seven ammals respectively. (B) Uterus: Estradiol. chlordecone, and control data represent the means k SE of six. seven, and six animals, respectively.*Significantly different from control, p G 0.03.

420

WILLIAMS,

ECKOLS,

progesterone (Roy et al., 1979a). Recent data showing that chlordecone does not increase CNS progesterone receptors (Eckols et al., 1989) may account for chlordecone’s inability to prime the ovariectomized rat for sexual receptivity. Hammond et aI., (1979) speculated that chlordecone’s delayed retention of nuclear receptors was responsible for its reproductive effects, While the present findings are consistent with that suggestion, they offer an additional explanation for chlordecone’s failure to mimic estradiol-like action. The present results suggest that chlordecone’s failure in viva to exhibit substantial “estrogenicity,” as measured by CNS indices, may depend on its retention of too few estradiol receptors in the nucleus. Ifa minimum density ofestradiol receptors must be tightly bound in the nucleus before cellular events responsible for the hormone specific action can occur, chlordecone achieves this in uterus but not in brain. Estrogen exerts both positive and negative feedback effects on neuroendocrine function (e.g., LH secretion). Chlordecone’s failure to mimic these positive feedback actions could result from its lower retention of CNS estradiol receptors. However, chlordecone does mimic the negative feedback effects of estrogen on LH secretion (Uphouse et aI., 1984) and this negative feedback is also thought to require intracellular estradiol receptors (McEwen et al., 1982). These observations suggest the speculation that a greater number of tightly bound nuclear receptors may be required for the positive than for the negative feedback effects of estrogen on LH secretion. It is also possible that the positive feedback of estradiol requires both the nuclear binding of and the removal of estradiol. Since chlordecone has a prolonged half-life in the rat (Egle et al., 1978) chlordecone’s persistence in brain tissue may maintain the nuclear retention of estradiol receptors and provide continued negative feedback. Alternatively, the positive and negative feedback actions may be anatomically distinct or the chlordecone receptor complex may bind to some, but not

AND UPHOUSE

all, of the chromatin sites bound by the estradiol receptor complex. Further evidence for a functional importance to the differential nuclear retention following chlordecone and estradiol treatment is indicated by their estradiol receptor replenishment. Chlordecone’s effect on estradiol receptor replenishment, similar to estradiol in the uterus but not in the brain, suggests a difference in the nature and/or degree of the molecular responses to chlordecone in the two tissues. Since receptor replenishment is required for facilitation of some estrogen-mediated events (Kassis and Gorski, 1983), failure to produce such replenishment in brain could also be a contributing factor in chlordecone’s apparent tissue difference in estrogenicity. In summary, chlordecone’s effectiveness as a competitor for binding to the estradiol receptor in vitro was similar in neural and uterine tissues. Similarities between estradiol and chlordecone in vim, however, were much more apparent for uterine than for neural tissue, Relative to the estradiol-treated animals, nuclear retention of estradiol receptors was delayed in both tissues following chlordecone treatment. However, in uterine tissue irz viva treatment with estradiol or chlordecone eventuahy produced similar 1eveI.sof nucIear retention and receptor replenishment. In contrast, relative to estradiol a reduced nuclear retention and a lesser degree of receptor replenishment occurred in brain tissue of chlordecone-treated females. Although these findings extend the evidence for chlordecone’s interaction with the estradiol receptor to neural tissue, it is clear that the pesticide’s resemblance to estradiol is less apparent for neural than for uterine tissue. Chlordecone’s failure to mimic estradiol’s receptor dynamics in vivu may partially account for the pesticide’s failure to exhibit substantial CNS estrogenicity and may also play a role in chlordecone-induced reproductive dysfunction. ACKNOWLEDGMENTS The authors thank Dr. Louise Higgins for photography, Mr. Martin Cudd for animal care, and Ms. Karen

ESTRADIOL

RECEPTOR

Nettles for manuscript preparation. The authors appreciate the helpful suggestions of Dr. George Stewart who read earlier versions of the manuscript.

INTERACTIONS

421

KASSIS, J., AND GORSKI, J. (1983). On the mechanism of estrogen receptor replenishment: Recycling, resyn thesis and/or processing. Mol. Cell. Biochem. 52, 2.736.

REFERENCES ADASHI, E. Y., HSUEH, A. J., BAMBINO, T. H., AND YEN, S. S. C. (1981). Disparate effects of clomiphene and tamoxifen on pituitary gonadotropin release irr vitro. Amer. J. Physiol. Endocrinol. Metub. 3, 125-l 30. CLARK, A., AND Roy, E. (1983). Behavioral and celhdar responses to pulses of low doses of estradiol- 17 beta. Physiot. Behuv. 30,561-565. CLARK, J., AND PECK, E., JR. (1976). Nuclear retention of receptor-oestrogen complex and nuclear acceptor sites. Nature (London) 260,635-637. ECKOL~. K., WILLIAMS, J., AND UPHOUSE, L. (1989). E&X% of chiordecone on progesterone receptors in immature and adult rats. Submitted. EDWARDS, D., WHALEN, R., AND NADLER, R. (1968). Induction of estrus. Estrogen-progesterone interactions. Physiol. Behuv. 3,22-33. EGLE, J., FERNANDEZ, S., GUZELAN, P., AND BORZELLENA, J. (1978). Distribution and excretion ofchlordecone (Kepone) in the rat. Drug Metab. Dispos 6,9195.

EROSCHENKO, V. P., AND PALMITER, R. D. (1980). Estrogenicity ofkepone in birds and mammals. In Estrogens in the Environment (J. McLachlan, Ed.), Elsevier/ North-Holland Biomedical, Amsterdam. GELLERT, R. J. (1978). Kepone, mirex, dieldrin, and aldrin. Estrogenic activity and the induction of persistent vaginal estrus and anovulation in rats following neonatal treatment. Environ. Res. 16,13 1-l 38. HAMMOND, B., KATZENELLENBCK~EN,B., KRAUTHAMMER. N., AND MCCONNELL, J. (1979). Estrogenic activity of the insecticide chlordecone (Kepone) and interaction with uterine estrogen receptor. Proc. Nati. &ad, Sci. 76,6641-6645. HUANG, E., AND NELSON. F. (1986). Anti-estrogenic action of chlordecone in rat pituitary gonadotrophs in vitro. Toxicol. Appl. Pharmacol. 82,62-69. HIJEER, J . ( 1965). Some physiological effects of the insecticide Kepone in the laboratory mouse. Toxicol. Appt. Fharmacol. 7,5 16-524. JORDAN, V. C., DIX, C. J., NAYLOR, K. E., PRESTWICH, G.. AND ROWSBY, L. (1978). Nonsteroidal antiestrogens; their biological effects and potential mechanisms of action. J. Toxicol. Environ. Health 4, 363-390. JORDAN, V., AND ROBINSON, S. (1987). Species-specific pharmacology of antiestrogens: Role of metabolism. Fed. Proc. 46* 1870-I 874. KARSCH. F. (1987). Central actions of ovarian steriods in the feedback regulation of pulsatile secretion of luteinizing hormone. Annu. Rev. Physiol, 49, 365-382.

KURL, R., AND MORRIS, I. (1978). I. Differential depletion of cytoplasmic high affinity oestrogen receptors after the in vivo administration of the antiestrogens clomiphene, MER-25 and tamoxifen. Brit. J. Pharmacol.62,487-493. LABARCA, C., AND PAIGEN,

K. (1980). A simple, rapid, and sensitive DNA assay procedure. A&. Biochem. 102,344-352.

MCEWEN, B. S., BIEGON, A., DAVIS, P. G., KREY, L. C.. LUINE, V. M., MCGINNIS, M. Y., PAYDEN, C. M., PARSONS, B., AND RAINBOW, T. C. (1982). Steroid hormones: Humoral signals which alter brain cell properties and functions. Rec. Prog. Harm. Res. 38, 41-83.

MCEWEN. B. S., AND PFAFF, D. fluencing sex hormone uptake Effects of neonatal treatment, competing steroid on estradiol

W. (1970). Factors inby rat brain regions. I. hypophysectomy. and uptake. Bruin Res. 21,

l-16.

PALMITER, R., AND MULVIHILL, Ed ( 1978) Estrogenic activity of the insecticide Kepone on the chicken oviduct. Science 20& 356-3S8. ROY, E., MACLUSKY, N., AND MCEWEN, B. (1979a). Antiestrogen inhibits the induction of progestin receptors by estradiol in the hypothaIamus-preoptic area and pituitary. Endocrinology 104,13331336. ROY, E., SCHMITT, E., MCEWEN, B., AND WADE, G. (1979b). Antiestrogens in the central nervous system. in Armhormones (M. Agrawal, Ed.), Elsevier/NorthHolland Biomedical, Amsterdam, SUTHERLAND. R.. AND Foe, M. (1979). Differential binding of antiestrogens by rat uterine and chick oviduct cytosol. Biochem. Biophys. Re.s. Commun. 91, 183-191. TAYLOR, R., SWANECK, J., HOLJFF,S., AND MAR~IXE~, A. (19%). Chlordecone intoxication in man. Part 1. Clinical observations. Neurology 28626-630. UPHOUSE. L. (1986). Single injection with chlordecone reduces behavioral receptivity and fertility of adult rats. Neurobehav. ToGcot. Teratol, 8, 12 1- 126, UPHOUSE, L.. ECKOLS, K., SIERRA. V., KOLODZI~J. M., AND BROWN, H. ( 1986). Failure of chlordecone (Kepane) to induce behavioral estrus in adult ovariectomized rats. NeuroTo.xicoiogv 7, 1?I- 142, UPHO~JSE.L.. MASON, G.. AND HUNTER. V, ( 1984). Persistent vaginal estrus and serum hormones after chlordecone (Kepone) treatment ofadult female rats, FuY;c,ol.Appl. Phurmacol. 12, 177-l Ub, WALTERS. M. (1985). Steroid hormone receptors and the nucleus. Endow Rev, 6, 5 I?-S43. ZAR. J. ( 1984). Biostuti.stic~t/ .+lnu/~si.~ Prentice-Hall. Englewood Cliffs. NJ.