Altered profiles of spontaneous novelty seeking, impulsive behavior, and response to D-amphetamine in rats perinatally exposed to bisphenol A.

Bisphenol A (BPA) is an environmental estrogen with potentially averse effects on public health. We studied the long-term effects of perinatal exposure to BPA on later behavior in adult rats of both sexes. BPA or vehicle was administered orally to mother rats from mating to pups' weaning, at a concentration (0.040 mg/kg) within the range of human exposure. The offspring of both sexes were tested at adolescence (postnatal days 35-45) for novelty preference (experiment 1). After a 3-day familiarization to one side of a two-chamber apparatus, on day 4 rats were allowed to freely explore the whole apparatus. BPA-exposed females spent significantly less time than did controls in exploration of the novel side (i.e., increased neophobia), whereas no effect was found in the male group. At adulthood, the same animals were food deprived and tested for profiles of impulsive behavior (experiment 2), in operant chambers provided with two nose-poking holes (delivering either five or one food pellet). After the establishment of a baseline preference for the large reinforcer, a delay was introduced before the delivery of the five food pellets, which was progressively increased each day (10, 20, 30, 45, 60, 80, 100 sec). As expected, all animals exhibited a progressive shift toward the immediate but smaller reinforcer. A reduced level of impulsive behavior (i.e., a shift to the right in the intolerance-delay curve) was evidenced in BPA-treated rats. The frequency of inadequate responding (during the length of the delay) also provided a measure of restless behavior. Interestingly, the profile of BPA-treated males was feminized, strongly resembling that of control females. Animals were then tested (experiment 3) for the response to an amphetamine challenge (1 mg/kg, subcutaneously). The drug-induced increment activity was significantly less marked in BPA-treated male rats compared with controls. These findings provide clear indirect evidence of long-term alterations in brain monoaminergic function after perinatal BPA exposure. This may be a cause for concern for public health, confirming that exposure to a weak environmental estrogen in the period of sexual differentiation of the brain can influence adult behavior.

There is increasing concern about the negative impact on public health of environmental chemicals with estrogenic activity (Carlsen et al. 1992;Guillette et al. 1996;Wolff et al. 1993). The ability of estrogenic hormones to affect sexual differentiation of the brain during a critical period of perinatal life is well known (Arnold and Gorski 1984;Hutchison 1997). In addition to reproductive and sexual behavior, a variety of behavioral patterns are organized and sexually differentiated in rodents under the influence of perinatal gonadal hormones (Beatty 1979;McClusky 1981;McEwen 1992). Recent advances in the neurosciences have shown that estrogens interact with the dopaminergic (Alderson and Baum 1981;Becker 1999;Euvrard et al. 1980;Hruska and Pitman 1982;Menniti and Baum 1981;Peris et al. 1991) and the serotonergic (Osterlund et al. 2000; Hurd 1998) brain systems. Perinatal exposure to estrogenic pollutants could hence alter development of these major neurochemical pathways (see, e.g., Christian and Gillies 1999;Lilienthal et al. 1997), leading to permanent neurobehavioral alterations in the offspring.
Bisphenol A (BPA) is a particularly important environmental estrogen. It is not only widespread but also potentially ingested by humans, being released by polycarbonate plastics, the lining of food cans, and dental sealants (Brotons et al. 1995;Olea et al. 1996). Prenatal exposure to BPA can affect the development and function of reproductive organs as well as adult sexual behavior, especially in male rodents and in their offspring (Atanassova et al. 2000;Dessì-Fulgheri et al. 2002;Farabollini et al. 2002;Fisher et al. 1999;Howdeshell et al. 1999;Rubin et al. 2001;Vom Saal et al. 1995Williams et al. 2001aWilliams et al. , 2001b. Perinatal exposure to BPA has also been implicated in altered profiles of nonsocial behaviors, resulting in a reduced motivation to explore and a reduced anxiety in the male offspring (Farabollini et al. 1999). In the present work, we wanted to extend the analysis of the consequences of early BPA exposure on nonsexual behaviors, investigating behaviors that rely upon central serotonergic and dopaminergic brain systems. Specifically, the serotonergic system is thought to be important for impulse control across a wide range of behaviors (Soubrié 1986), whereas the dopaminergic system is strongly involved in mediating motivation and reward (Robbins and Everitt 1996;Wise 1996).
The dopaminergic system is particularly important for the expression of novelty-seeking behavior (Bardo et al. 1996;Pierce et al. 1990). Both humans and animals have a natural need to search for novel and rewarding stimuli (Renner 1990;Zuckerman 1994), and the experience of novelty is rewarding via the activation of the mesolimbic dopaminergic system (Rebec et al. 1997a(Rebec et al. , 1997b. This behavioral trait is particularly expressed during adolescence in both humans (Arnett 1992;Zuckerman 1994) and animal models (Adriani et al. 1998; Bardo et al. 1996;Macrì et al. 2002). In rodents, adolescence is classically defined as the ontogenetic period including the week preceding the onset of puberty and the first few days thereafter (Spear and Brake 1983). We tested the hypothesis that possible alterations in coping with novelty, deriving from perinatal exposure to an estrogenic pollutant, could be easily detectable around puberty. The latter is indeed characterized by the onset of prominent hormonal regulation. For this reason, BPA-exposed rats of both sexes were assessed in a novelty preference test (Bardo et al. 1988;Misslin and Ropartz 1981) during adolescence (Spear and Brake 1983).
The role of the serotonergic system in modulating premature and impulsive responding is widely recognized on both clinical (Linnoila et al. 1983) and preclinical literature (Soubrié 1986). Impulsivity can be defined in several ways, including a) the "failure to resist Bisphenol A (BPA) is an environmental estrogen with potentially averse effects on public health. We studied the long-term effects of perinatal exposure to BPA on later behavior in adult rats of both sexes. BPA or vehicle was administered orally to mother rats from mating to pups' weaning, at a concentration (0.040 mg/kg) within the range of human exposure. The offspring of both sexes were tested at adolescence (postnatal days 35-45) for novelty preference (experiment 1). After a 3-day familiarization to one side of a two-chamber apparatus, on day 4 rats were allowed to freely explore the whole apparatus. BPA-exposed females spent significantly less time than did controls in exploration of the novel side (i.e., increased neophobia), whereas no effect was found in the male group. At adulthood, the same animals were food deprived and tested for profiles of impulsive behavior (experiment 2), in operant chambers provided with two nose-poking holes (delivering either five or one food pellet). After the establishment of a baseline preference for the large reinforcer, a delay was introduced before the delivery of the five food pellets, which was progressively increased each day (10, 20, 30, 45, 60, 80, 100 sec). As expected, all animals exhibited a progressive shift toward the immediate but smaller reinforcer. A reduced level of impulsive behavior (i.e., a shift to the right in the intolerance-delay curve) was evidenced in BPA-treated rats. The frequency of inadequate responding (during the length of the delay) also provided a measure of restless behavior. Interestingly, the profile of BPA-treated males was feminized, strongly resembling that of control females. Animals were then tested (experiment 3) for the response to an amphetamine challenge (1 mg/kg, subcutaneously). The drug-induced increment activity was significantly less marked in BPA-treated male rats compared with controls. These findings provide clear indirect evidence of long-term alterations in brain monoaminergic function after perinatal BPA exposure. This may be a cause for concern for public health, confirming that exposure to a weak environmental estrogen in the period of sexual differentiation of the brain can influence adult behavior. Key words: amphetamine, behavior, bisphenol A, environment, impulsivity, novelty-seeking, pollutants, rats. an impulse, drive, or temptation" (Evenden 1999); b) responding without consideration of alternatives and/or consequences; or c) behaving in a way that is adequate to the environmental contingency. Many different aspects of impulsivity have been studied with operant paradigms in laboratory settings (Bradshaw and Szabadi 1992;Evenden 1999;Richards et al. 1997Richards et al. , 1999. One of the most widely adopted paradigms assumes that impulsive subjects are intolerant to situations when reward is delayed. Smaller immediate reinforcers are preferred to larger rewards, which come only after a delay (Bizot et al. 1999;Evenden andRyan 1996, 1999;Logue 1988;Thiebot et al. 1985). In the present study, we evaluated the possibility that perinatal exposure to BPA may influence the development of the serotonergic system in adult animals through the impulsivity test.
The specificity of the developmental changes affecting a central neurochemical system can be evaluated by assessing the effects of a psychoactive agent targeting that system upon the behavioral responses known to be modulated by that system. For this reason, it seemed appropriate to evaluate the increase of locomotion and rearing behavior that follow amphetamine administration (Kelly et al. 1975), because it is well known that release of dopamine within the dorsal and ventral striatum is involved in such a behavioral change (Staton and Solomon 1984). For this study, we considered a potential alteration in the behavioral effects of amphetamine administration an index of BPA-induced long-term effects on the function of the brain dopaminergic system.
We studied the effects of precocious exposure to BPA [at concentrations within the range of human exposure and not teratogenic (Brotons et al. 1995;Olea et al. 1996)]. To this purpose, we administered BPA orally to pregnant females from mating to weaning day. The offspring were then tested for novelty preference during adolescence. When adult, the same subjects were also tested for intolerance to delay and for amphetamineinduced behaviors in an open field.

Subjects, breeding, and rearing conditions.
Sprague-Dawley rats were housed in an airconditioned room (temperature, 21 ± 1°C; relative humidity, 60 ± 10%), with a 12-hr light/dark cycle (lights off from 2100 to 0900 hr). Water and food (Enriched Standard Diet; Mucedola, Settimo Milanese, Italy) were available ad libitum. Breeding pairs were formed and housed in Plexiglas cages, with metal tops and sawdust bedding. After detection of the vaginal plug, the male was removed, and the females were housed individually. The day of delivery was considered postnatal day (PND) 0; pups were weaned on PND 25 and housed in groups of three, according to sex. One male and one female per litter were observed in the present study. Animals were tested during adolescence (PND 30-45) for novelty seeking and when adult (PNDs > 70) for impulsivity and response to amphetamine.
The estrogenic pollutant BPA (Fluka Chemie Ag, Buchs, Switzerland) was administered daily to females (n = 9) from mating day to weaning day. The substance was dissolved in arachis oil at a concentration of 0.04 mg/kg, which was administered orally by micropipette, the volume administered depending on body weight. Control females (n = 9) received arachis oil without BPA. Because animals were trained to receive the oil before mating, this procedure was not stressful.
Experiment 1: novelty preference test. Animals of both sexes were tested for levels of novelty seeking during adolescence. The experimental apparatus consisted of an opaque Plexiglas box with smooth walls (70 × 30 × 35 cm), subdivided into two compartments. The connecting door between the two compartments could be closed by means of a temporary partition. One compartment had a wide-mesh floor, whereas the other had narrow mesh. Animals were video recorded and later scored for measures of time spent in each compartment and activity rate in each compartment. To evaluate the activity rate, the floor of each compartment was subdivided into three sections by lines placed on the video screen at the time of video-recording analysis, and the number of line crossings (with both forepaws) was scored.
The whole experimental schedule took 5 days, each subject from both age groups being tested between 1000 and 1800 hr. Testing of different experimental groups was counterbalanced across time. The test was carried out under dim illumination. The floor of the apparatus was cleaned after each animal was tested. During the familiarization phase (days 1-3), animals were gently placed for 20 min in one compartment of the apparatus. During the novelty preference test (on day 4), animals were placed in the familiar compartment for a 5-min session. The partition separating the two compartments of the apparatus was then removed, and rats were thus allowed to freely explore the whole apparatus (both the familiar and the novel sides) for 24 min.
Experiment 2: impulsivity test. When adult, the same animals were tested for levels of impulsivity. Before the schedule started, animals were food deprived (80% of free-feeding weight; see Table 1) to increase their motivation to work for food delivery. Each animal was then placed daily in a computer-controlled operant chamber (Coulbourn Instruments, Allentown, PA, USA), provided with two nosepoking holes, a chamber light, a feeder device, a magazine where pellets (45 mg; BioServ, Frenchtown, NJ, USA) were dropped, and a magazine light. The nose poking in either hole was detected by a photocell and was recorded by a computer, which also controlled food delivery. After a 30-min session, animals were returned to their home cages, where they were given standard chow (~8 g each), to keep animals at 80-85% of their free-feeding weight.
During the training phase (1 week), nose poking in one of the two holes [called the "immediate and small" (IAS) hole] resulted in the delivery of one pellet of food, whereas nose poking in the other hole ["large and delayed" (LAD) hole] resulted in the delivery of five pellets of food. After nose poking and before food delivery, the chamber light was turned on for 1 sec. After the food delivery, the magazine light was turned on for 25 sec, during which additional nose poking was recorded but was without any scheduled consequence (time out).
During the testing phase (1 week), a delay was inserted between nose poking in the LAD hole and the delivery of the five pellets. The chamber light was turned on during the length of this delay. Any additional nose poking taking place during this time interval was recorded but was without any consequence ["inadequate responding" (Sagvolden and Sergeant 1998;Sagvolden 2000)]. The delay was kept fixed for each daily session and was increased progressively over subsequent days (0, 10, 20, 40, 60, 80, 100 sec). The dependent variables were the percentage of choice between the LAD and IAS holes and the frequency of inadequate nose poking.
Experiment 3: open field with amphetamine. One week after the impulsivity test, all animals were tested for response to amphetamine in an open-field apparatus. This consisted of an opaque Plexiglas rectangular box with smooth gray walls and floor (70 × 30 × 35 cm). D-Amphetamine (AMPH; 1 mg/kg) was dissolved in saline (SAL; NaCl 0.9%) and injected subcutaneously in a volume of 1 mL/kg body weight. Approximately 15 min after the injection with either SAL or AMPH, 396 VOLUME 111 | NUMBER 4 | April 2003 • Environmental Health Perspectives Articles | Adriani et al. Table 1. Mean (± SE) body weight before food deprivation (day 0) and during the schedule (days 3-12) in experiment 3.

Males Females Day
Control BPA-treated Control BPA-treated 0 478 ± 11 482 ± 1 284 ± 10 302 ± 9 3 441 ± 9 449 ± 11 265 ± 7 278 ± 8 6 410 ± 9 432 ± 10 258 ± 8 269 ± 9 9 410 ± 8 414 ± 11 251 ± 6 262 ± 8 12 383 ± 9 389 ± 11 235 ± 6 246 ± 8 animals were placed in the open field for a single 30-min session. The behavioral profile expressed by each animal was video recorded and later scored by a treatment-blinded individual, using a computer and specific software (The Observer, version 2.0 for DOS; Noldus Information Technology, Wageningen, The Netherlands). This allowed a detailed analysis of several parameters, including latency, frequency, and duration of each behavior. Three behaviors were scored: rearing (body in vertical position), grooming (mouth or paws on body), and crossing (the floor of each compartment was subdivided into three sections by lines placed on the video screen at the time of video-recording analysis, and the number of line crossings with both forepaws was scored).
Design and data analysis. Data were analyzed by multifactorial analysis of variance (ANOVA). The general design of all experiments was two sex (male vs. female) × two treatment (BPA vs. oil) × subject. For the novelty-seeking paradigm (experiment 1), a side (familiar vs. novel) and a time factor were added. For the impulsivity paradigm (experiment 2), a delay factor (0, 10, 20, 40, 60, 80, 100 sec) was added. In the openfield test (experiment 3), a drug factor (SAL vs. AMPH) was added. Multiple comparisons within significant interactions were performed with the Tukey HSD test.

Results
Experiment 1: novelty preference test. Activity rate. The four-way ANOVA yielded significance for the time effect [F(5,160) = 4.96, p < 0.001] and for the sex by time interaction [F(5,160) = 2.49, p < 0.05], indicating that the time-course profile of activity during the test was markedly different in the two sexes. On this basis, and to analyze more specifically the effects of BPA exposure, the two sexes were analyzed separately by a three-way ANOVA.
For males ( Figure 1C), the ANOVA yielded significance for the time by side interaction [F(2,32) = 3.99, p < 0.05]. Specifically, when animals were in the novel compartment, the activity rate was particularly elevated in the first part of the session, decreasing thereafter. Conversely, in the familiar compartment, the activity profile was flat during the whole session (data not shown). Moreover, a treatment × side × time interaction [F(2,32) = 9.28, p < 0.05] emerged. The prenatal exposure to BPA resulted in higher activity levels than for controls in the novel environment, especially at the end of the session. In other words, the habituation profile was less pronounced in BPA-exposed rats. Conversely, levels of activity in the familiar were not affected (data not shown).
For females ( Figure 1D) the ANOVA yielded significance for the side by treatment interaction [F(1,16) = 10.53, p < 0.01]. As in males, the perinatal exposure to BPA resulted in higher levels of activity than for controls in the novel environment. Conversely, levels of activity in the familiar compartment were not affected (data not shown).
Novelty preference. In the three-way ANOVA, the main effect of time was significant [F(5,160) = 12.26, p < 0.01]. The novelty preference increased as session progressed. The ANOVA yielded significance for the sex by treatment interaction [F(1,32) = 4.40, p < 0.05]. As a whole, early exposure to BPA produced a marked reduction of time spent in the novel environment in females ( Figure 1B), whereas the group of males was not affected ( Figure 1A).
Separate analyses confirmed this picture. For females, but not males, the two-way ANOVA yielded significance for the main effect of treatment [F(1,16) = 10.44, p < 0.01]. Multiple comparisons performed within the female group revealed that, as a consequence of perinatal BPA exposure, a reduction of time spent in the novel environment was found at the beginning and at the end of the session.
Experiment 2: impulsivity test. Choice between reinforcers. As expected, after the training period, animals of both sexes developed a significant preference for the LAD hole, delivering the large reinforcer ( Figure 2). The preference also progressively shifted toward the hole delivering the immediate reinforcer as the length of the delay was increased [delay, F(6,192) = 31.6, p < 0.01], but no evidence of a sex difference was found in the ANOVA. Interestingly, a main effect of treatment [F(1,32) = 4.28, p < 0.05] revealed that adult rats exposed perinatally to BPA were associated with a more marked preference for the LAD reinforcer during the whole experiment. As a whole, this profile suggests a shift to the right in the delay-response curve (i.e., reduced impulsivity) in rats of both sexes.
Inadequate responding. Because the nose poking in either hole during the course of the delay had no scheduled consequences, it was considered an "inadequate response" (Sagvolden 2000;Sagvolden and Sergeant 1998). Such a measure provides an index of inability to inhibit an unnecessary response. As the length of the delay was increased, the inadequate nose poking in the LAD hole was progressively reduced, whereas the inadequate nose poking in the IAS hole increased progressively Environmental Health Perspectives • VOLUME 111 | NUMBER 4 | April 2003 397 Articles | Endocrine disruptors and neurobehavioral alterations activity rate, measured as number of line crossings per minute, shown by subjects of both sexes in the novel compartment on testing day. During the pretreatment period (days 1-3), subjects were familiarized to one compartment. On testing day, animals were placed in the familiar compartment. After 5 min, a partition was removed and subjects were allowed free access to a novel compartment of the apparatus for a 24-min session. *p < 0.05 in comparisons between BPA and control perinatal treatments (n = 9).

Females (min)
[delay, F(6,192) = 14.9, p < 0.01; delay × hole, F(6,192) = 49.1, p < 0.01]. Such a finding suggests that, during the length of the delay, when they had to wait for the large reinforcer, rats were demanding more and more the small and immediate one. Interestingly, a significant main effect of sex [F(1,32) = 4.44, p < 0.05] and significant sex by treatment interaction [F(1,32) = 4.25, p < 0.05] were found. To better depict the profile, data from the two sexes were analyzed separately.
For males, a main effect of treatment [F(1,16) = 8.29, p < 0.05] as well as a delay by treatment interaction [F(6,96) = 2.12, p < 0.05] emerged. Multiple comparisons revealed that, as the delay increased, adult rats were associated with elevated nose poking. Interestingly, a significant delay by hole by treatment interaction [F(6,96) = 3.24, p < 0.01] appeared. Multiple comparisons revealed that, when the length of the delay was set to 1 min or more, BPA-exposed rats were specifically associated with a significantly lower frequency of nose poking in the IAS hole ( Figure 3A). It is interesting to note that all these interactions were not significant within the female group; that is, female subjects were apparently not affected by BPA exposure ( Figure 3B). Furthermore, early BPA exposure results in males whose profile is comparable with that expressed by females.
To better depict the effects, we analyzed the data from the two sexes separately (see Figure 4C,D). For males, the ANOVA evidenced a main effect of drug [F(1,14) = 37.7, p < 0.01], with AMPH resulting in elevation of crossing frequency. Moreover, a main effect of treatment [F(1,14) = 10.7, p < 0.01] and a drug by treatment interaction [F(1,14) = 6.34, p < 0.05] emerged. Specifically, multiple comparisons revealed that AMPH administration resulted in elevation of crossing in control but not in treated subjects. Conversely, for females, only a main effect of drug [F(1,14) = 20.5, p < 0.01] appeared. AMPH administration resulted in elevation of crossing in both control and BPA-treated subjects. As a whole, these results suggest that BPA exposure impaired the response to AMPH only in male subjects.
Rearing. The ANOVA yielded a main effect of drug [F(1,28) = 22.5, p < 0.01], AMPH resulting in elevation of rearing. Interestingly, a main effect of sex just missed significance [F(1,28) = 3.27, p < 0.081] and the sex by treatment interaction was significant [F(1,28) = 6.25, p < 0.05]. To better depict this effect, the two sexes were analyzed separately ( Figure 4A,B). For males, the ANOVA evidenced a main effect of drug [F(1,14) = 18.5, p < 0.01] and treatment [F(1,14) = 7.03, p < 0.05]. Specifically, as is evident from Figure 4A, AMPH-induced elevation of rearing was less marked in BPAtreated than in control subjects. Conversely, for females, only a main effect of drug [F(1,14) = 8.19, p < 0.05] appeared. Specifically, AMPH administration resulted in elevation of rearing in both control and BPA-treated subjects. As a whole, these results suggest that BPA exposure impaired the response to AMPH in male subjects.

Discussion
As a whole, the present results can be summarized as follows: a) Rats of both sexes, perinatally exposed to BPA and tested during adolescence for novelty seeking, were associated with more marked levels of noveltyinduced hyperactivity, compared with controls. However, BPA-exposed females spent a lower percentage of time in the novel environment (an index of neophobia). b) BPAexposed rats of both sexes were associated with a more marked preference for the LAD reinforcer during the whole experiment (an index of decreased impulsivity). Compared with controls, BPA-exposed males exhibited a feminization in the frequency of inadequate nose poking at the IAS hole during the length of the delay. c) As expected, AMPH injection induced an elevation of crossing and rearing in control male subjects and in both groups of females. Perinatal BPA exposure was able to impair the classical response to AMPH in male subjects.
Novelty seeking in adolescent rats. Periadolescent rats and mice express elevated levels of behavioral activation in specific forms. For instance, they show elevated levels of social play and affiliative behaviors (Meaney and Stewart 1981;Panksepp 1981) that progressively shift toward aggressive and competitive behaviors (Terranova et al. 1993(Terranova et al. , 1998. Moreover, rodents at this age exhibit a marked peak in novelty-seeking behavior (Adriani et al. 1998;Bardo et al. 1996) and low levels of exploration-induced anxiety (Macrì et al. 2002). The psychobiology of novelty-seeking behavior has been studied in mice and rats. Specifically, the dopaminergic system has been widely implicated in mediating the 398 VOLUME 111 | NUMBER 4 | April 2003 • Environmental Health Perspectives Articles | Adriani et al. . Mean (± SE) choice (%) of the large reinforcer, demanded by nose poking at the LAD hole, shown by rats during the test for impulsivity (experiment 2). These data reveal that, as the length of the delay increased, animals increased demanding the small but immediate reinforcement and decreased demanding the larger but delayed one. A shift to the right of the whole curve (i.e., a profile of reduced impulsivity) was evident in BPA-exposed rats compared with controls. In the absence of significant differences, data from the two sexes were collapsed (n = 18). . Mean (± SE) frequency of inadequate responding at the IAS hole (i.e., nose poking during the length of the delay, when it was without any consequence) shown by rats during the test for impulsivity (experiment 2). These data reveal that, when animals were waiting for the delivery of the large reinforcer, they failed to rest and were demanding the immediate one. A clear-cut demasculinization in the restlessness profile was evident.
In the present experiment, female rats spent less time than did males in the new environment at the beginning of the freechoice exploration, suggesting a lower interest of females in exploring the novel side. These findings are consistent with other previous results, suggesting that females show lower levels of novelty seeking than do males in both rats (Hughes 1968) and mice (Palanza et al. 2001). Compared with controls, BPAtreated females spent a minor percentage of time in the novel compartment of the apparatus, remaining most of the time in the familiar compartment. These data indicate that, rather than being attractive as is normally reported for rats (Bardo et al. 1988), the experience of novelty was avoided by females after maternal BPA exposure. In other words, prenatal exposure to BPA was apparently responsible for an increased neophobia in adolescent female rats. In a previous study, parameters of motor activity and motivation to explore were depressed in adult female rats after maternal exposure to BPA (Farabollini et al. 1999). These findings suggest that, compared with control subjects, BPA-treated females were less prone to explore a novel environment.
Regarding locomotion in the novel compartment, rats treated with BPA expressed elevated levels of activity and a less marked profile of habituation, an effect that was evident in both sexes. In the novelty-seeking test, exploratory activity and emotional reactivity represent two different dimensions based on different mechanisms (Zimmermann et al. 2001). A profile of hyperlocomotion during exploration of novel environments has been proposed as an index of novelty-induced stress (Exner and Clark 1993;Misslin and Ropartz 1981;Misslin et al. 1982). Present data may suggest that prenatal treatment with BPA produced rats that were more likely to experience noveltyinduced stress during adolescence or, alternatively, a slowing down of the process of habituation. This behavioral profile could be related to alterations in the function of brain neurochemical systems involved in the locomotor response to novelty-induced stress and/or in the locomotor habituation to novelty.
Impulsive behavior. Impulsivity, defined as a reduced ability to tolerate a delay of gratification (Evenden 1999), has been studied in rats by means of various procedures, providing a choice between a large but delayed food reinforcement versus a smaller and immediate one (Bizot et al. 1999;Evenden andRyan 1996, 1999). Delay has actually been shown to have a discounting effect on the subjective value of a given reinforcement (Bradshaw and Szabadi 1992;Richards et al. 1997).
In the present study, food-restricted animals were trained in operant chambers, where nose poking resulted in food delivery. As expected, all animals significantly preferred the hole associated with the large reinforcement (LAD hole) and also exhibited a shift toward the small, immediate reinforcement (IAS hole) as the length of the delay was increased. Rats exposed perinatally to BPA were associated with a more marked preference for the LAD reinforcer during the whole experiment-that is, with a rightward shift of the delay-preference curve-suggesting a reduction of impulsive behavior. In previous studies involving a similar paradigm, a marked increment in the preference for the large-but-delayed reward was induced by serotonin uptake inhibitors such as indalpine, zimelidine (Thiebot et al. 1985), fluoxetine, and fluvoxamine (Bizot et al. 1999). These data support the idea that serotonergic mechanisms are involved in the regulation of impulsive behavior, suggesting that an elevated serotonergic tone may result in elevated tolerance to reward delays. On this basis, it may be supposed that perinatal BPA exposure affected the ontogenesis of this central neurochemical system. Nose poking in either hole during the length of the delay had no scheduled consequences. However, in the course of the present experiment, animals kept on demanding the food reinforcement even during the signaled nonreinforced component of the schedule. This might happen because animals were unable to modify response patterns with changes in the experimental contingency, being under the behavioral urge of doing something and unable to simply wait. This kind of behavior has been defined as "inadequate responding" (Sagvolden 2000;Sagvolden and Sergeant 1998;, and its measure might hence provide an index of restlessness and reduced ability to wait. Interestingly, a sexual difference emerged, both at basal level and in the response to BPA: a) Females showed lower levels of inadequate responding than did the corresponding group of males; b) no effect of BPA exposure was found in females; and c) levels shown by BPA-exposed males resembled those shown by both groups of females. Results in control subjects suggest that males have a stronger preference for the immediate reinforcer than do females, which is progressively more expressed during the length of the delay. Alternatively, these results suggest that male subjects are less able than females to inhibit nose poking behavior during the delay. Interestingly, BPA-exposed males were specifically associated with lower levels of inadequate nose poking in the inactive hole, compared with controls. This suggests that BPA-treated males are less restless and more tolerant to the delay and/or more able to inhibit the inadequate behavior. Interestingly, early BPA exposure results in males whose profile is comparable with that expressed by females, suggesting a demasculinization for this measure. Consistently, modifications of sociosexual behavior in the direction of a demasculinization have been observed in adult male rats perinatally exposed to BPA .
Open-field test and response to amphetamine. As expected, the AMPH-induced elevation of both crossing and rearing was significantly reduced in BPA-treated male subjects. Such a picture suggests that early BPA exposure impaired the function of central neurochemical systems targeted by AMPH in the male offspring. A reduced dopaminergic function can be hypothesized for BPA-exposed males, which may also partially account for the particular hypoactivity shown by these subjects during the length of the delay in experiment 2 (discussed above). We may suppose that perinatal BPA exposure interacted with some steps in the development and organization of the dopaminergic system during the perinatal period of male offspring.
Environmental Health Perspectives • VOLUME 111 | NUMBER 4 | April 2003 399 Articles | Endocrine disruptors and neurobehavioral alterations Regarding possible mechanisms, BPA exhibits weak estrogenic activity in adult rats of both sexes. Specifically, BPA administration causes a significant increase in uterus and vagina weights in ovariectomized females (Kim et al. 2001), whereas it directly inhibits testicular functions and produces a reduction in the negative feedback of testosterone (Tohei et al. 2001). Long-term exposure of adult female rats to BPA induces modifications in β-estrogen receptor immunoreactivity in various brain areas regulating reproductive and maternal behavior (Aloisi et al. 2001). Many studies have addressed the adverse effects of perinatal exposure to BPA on various indexes of sexual development and maturation (Atanassova et al. 2000;Fisher et al. 1999;Rubin et al. 2001;Williams et al. 2001aWilliams et al. , 2001b. Unfortunately, little is known about the effect of estrogen-like compounds on developing monoamine systems. One paper reported, however, that intrauterine exposure to estradiol has a significant effect on the organization of monoamine systems within the fetal hypothalamus (Kaylor et al. 1984). More is known about interactions of estrogens with the adult dopaminergic and serotonergic systems (for a review, see, e.g., Cyr et al. 2002;Dluzen 2000;Fink et al. 1996Fink et al. , 1999Rubinow et al. 1998).

Conclusions
As a whole, perinatal treatment with the estrogenic pollutant BPA resulted in marked alterations in rats' behavioral repertoire. Specifically, an increased novelty-induced stress and/or a reduced habituation to novelty was found during adolescence, as well as reduced levels of impulsivity at adulthood. Both findings may well be seen as indexes of a reduced reactivity or readiness to react to environmental challenges. It could be argued that BPA exposure resulted in individuals that do not easily adapt to environmental changes (see Benus et al. 1987).
Interestingly, some of these effects were sex dependent. The perinatal treatment with BPA affected the restlessness profile in male rats, with BPA-treated animals becoming undistinguishable from females. This finding, together with the reduced sensitivity of BPAtreated adult males to AMPH, suggests that the perinatal exposure to BPA interacts with some steps in the organization of the serotonergic and dopaminergic neural systems in the male offspring. On the contrary, perinatal BPA exposure produced neophobia in adolescent females but not in males. This effect was possibly determined via a different mechanism from that controlling impulsivity, because BPA exposure had no effect upon behavior of adult females in the impulsivity test or in the open-field test. On the basis of the scientific literature, the various behavioral alterations reported in the present study could be ascribed to an altered development of dopaminergic and/or serotonergic pathways. It can be hypothesized that both these systems were affected by perinatal BPA treatment, but further work is needed to clarify the neural basis of long-term neurobehavioral deficits induced by BPA.
The present results acquire even more importance on the basis of recent reports, indicating that performance in operant tasks is used also in children to evaluate adverse consequences of exposure to polychlorinated biphenyls (Stewart et al. 2001). As a general conclusion, the present findings provide indirect evidence of long-term consequences of perinatal BPA exposure at the level of neurobehavioral development. These alterations should be further investigated by means of biochemical testing. However, our results might be a cause of concern for public health, indicating that exposure to a weak environmental estrogen in the period of sexual differentiation of the brain may influence adult behavior. Further research is needed to better understand which exposure levels would not be potentially dangerous for human health.