TERATOLOGY 4 3 % - 7 0 (1991)
Prenatal Exposure to Gamma/Neutron Irradiation: Sensorimotor Alterations and Paradoxical Effects on Learning DANIELA DI CICCO, SARA ANTAL, AND MARTINE AMMASSARI-TEULE Istituto di Psicobiologia e Psicofarmacologia del CNR, Via Reno, 1 00198 Rome, Italy (D.D.C., M.A.-T.);'%rideric Joliot-Curie" National Research Institute for Radiobiotogy and Radiohygiene, Budapest, 1775, Hungary (S.A.)
ABSTRACT The effects of prenatal exposure on gamma/neutron radiations (0.5 Gy at about the 18th day of fetal life) were studied in a hybrid strain of mice (DBA/Cne males x C57BL/Cne females). During ontogeny, measurements of sensorimotor reflexes revealed in prenatally irradiated mice 1)a delay in sensorial development, 2) deficits in tests involving body motor control, and 3) a reduction of both motility and locomotor activity scores. In adulthood, the behaviour of prenatally irradiated and control mice was examined in the open field test and in reactivity to novelty. Moreover, their learning performance was compared in several situations. The results show that, in the open field test, only rearings were more frequent in irradiated mice. In the presence of a novel object, significant sex x treatment interactions were observed since ambulation and leaning against the novel object increased in irradiated females but decreased in irradiated males. Finally, when submitted to different learning tasks, irradiated mice were impaired in the radial maze, but paradoxically exhibited higher avoidance scores than control mice, possibly because of their low pain thresholds. Taken together, these observations indicate that late prenatal gamma/ neutron irradiation induces long lasting alterations at the sensorimotor level which, in turn, can influence learning abilities of adult mice. Teratogenic effects of ionizing (X- or neu- distribution and differentiation of neurons tron) radiations on development have been (Hayashi et al., '78; Kameyama et al., '78), widely described after either incidental or as well as cell piknosis and neuroblast deexperimental exposure of mammalian em- struction (Cowen and Geller, '60; Altman bryos and fetuses (Hicks and D'Amato, '66; and Nicholson, '71; Antal and Fulop, '86) Cairnie et al., '74; Brent, '77; Vogel and have been regularly observed in embryos Antal, '84; Fonagy et al., '85; Jensh and and fetuses of rodents dying some hours afBrent, '88a-c). The extent of the damage ter the irradiation or in young animals sachas been found to depend on several factors, rificed during postnatal development. In the such as dose and duration of exposure, pe- case of X-irradiation, these cellular alterriod of gestation, and species or strain con- ations have been found to be related particsidered. Since microcephaly associated with ularly to early-developing structures such mental retardation is the predominant fea- as the cerebellum and spinal cord (Inouye, ture of prenatally irradiated human sub- '79; Cerda, ?33), and, to a lesser extent, the jects (Mole, '82; Otake and Schull, '84), nu- neostriatum and cortex (Kameyama et al., merous investigations in the rat and the '78; Hoshino et al., '80). These morphologic observations have led mouse have attempted to identify alterations of brain structures underlying behav- several authors to focus, at the behavioral ioral modifications. Histological data have level, on possible change in activities inyielded a general framework of the brain sensitivity to X- or neutron rays. Disturbed Received June 7,1988; accepted August 15,1990 cell proliferation and migration, abnormal 0 1991 WILEY-LISS, INC.
62
D.DI CICCO ET AL.
havioral testing. These pups were considered as irradiated at about the 18th day of fetal life. Eight other hybrid females from the remaining 12 served as controls. Those females were sham irradiated, i.e., placed in the rotating chamber of the reactor but without irradiation. Experimental design The following physiological and behavioral measurements were performed in pups exposed to each treatment. Body weights and reflexes Three litters from irradiated mothers involving 4, 5, and 6 pups and five control litters of 4-6 pups each were selected and tested every other day from postnatal day 2 to postnatal day 16. The following items were recorded for the pups: body weights, rooting, righting, grasping, forelimb placing, cliff aversion, hair growth, opening of ears, and eyes, and auditory startle response, according to a previously described method (Castellano and Ammassari-Teule, '84). The age at which a given response was present in all pups of each litter indicated METHODS the score. Finally, body weights were Animals checked in adult irradiated and control anThe mice used in these experiments were imals before learning tests. F1 hybrids produced by mating male DBAI2 Motility with females C57BL/6 strains. Adult (2-3 Motility rates were examined in the same months) nulliparous hybrid females (N = 20) animals on postnatal days 7, 17, and subsewere individually housed with food and wa- quently at 3 months of age. For each mouse, ter ad lib. Males (2-3 months) were intro- the motility score was measured during 5 duced into the cages (ratio 1:l)and left for 1 minutes by a selective activity meter (Anweek. The day after the males had been in- imex). troduced in the cages was considered as day Locomotor activity 1 of pregnancy. The females were weighed regularly and isolated when determined to At 1month of age, the same animals were be pregnant. Eighteen days after mating, also tested for locomotor activity in Plexithe females were irradiated according to the glas toggle-floor boxes. Each box was difollowing procedure. Each female was put vided into two 20 x 10 compartments sepinto a plastic animal chamber rotating axi- arated with a black partition having a 3 x ally and inserted in the cavity of the 3 cm opening at the floor level to connect research reactor Tapiro (from the Italian the compartments. Each crossing through National Center for Alternative Energy- the opening tilted the grid floor, thus actiENEA-Rome) at a distance of 140 cm from vating a microswitch connected with a the core. This reactor produced variable counter. neutrodgamma (nly) dose rates. The single Reactivity to open field and novelty whole body dose was approximately 0.5 Gy Irradiated and control 4-month-old naive neutrons, this neutron dose being accompanied by approximately one-third of gamma animals (6 males and 6 females randomly rays, i.e., 0.18 Gy. Since in CD1 mice deliv- selected from three litters for each treatery generally occurs after 19 days of preg- ment) were examined in the open field sitnancy, only pups born between 24 and 36 uation and, successively, tested for reactivhours after irradiation were retained for be- ity to novelty after the introduction of a
volving motor control. In fact, motor coordination (Sharp, '61; Werboff et al., '61), locomotor activity (Lipton, '661, and, in some cases, maze learning (Furchtgoff et al., '58; Werboff et al., '62; Shibagaki et al., '81; Seo et al., '82), were profoundly disturbed in developing and adult animals prenatally exposed to X-rays. However, in the case of neutron radiation-which appears to produce more deleterious biological effects than X-irradiation for equivalent doses in mouse embryos (Di Majo et al., '81)-a more general pattern of early morphological alterations has been detected. This makes the search for specific functional deficits rather difficult. The present study has been designed to investigate the behavioral alterations due to prenatal gammaheutron irradiation in mice. A longitudinal study has been conducted in order to identify modifications of several physiological and behavioral parameters from birth to adulthood, with a particular attention devoted to learning abilities of animals irradiated during their fetal life.
PRENATAL EXPOSURE TO GAMMAINEURON IRRADIATION
63
novel object in the apparatus. The open field was a grey Plexiglas box consisting of a square floor (30 x 30 cm) surrounded by 15 cm high walls. The floor was painted white and divided into 36 squares by drawing black lines. Each mouse was introduced in the center of the open field and retained for 15 sec in a transparent Plexiglas cylinder. After the removal of the cylinder, behavioral recordings were performed for 5 minutes. At the end of this period, a transparent Plexiglas box (8.5 x 8.5 x 11 cm) with holes irregularly distributed on each side was placed in the center of the open field. After a free observation period of 15 sec, behavioral recordings were performed for 3 min. The mouse was then removed from the open field and the novel object cleaned with alcohol. The open field was placed in a soundproof chamber and the behavior of each animal was recorded by a videotaping system. The videotapes were then examined and the following behavioral items were selected by means of a keyboard connected to an Apple IIe P.C.: number of crossed squares, number of rearing responses (against the walls or not), time spent freezing and grooming, and finally time spent sniffing and rearing against the novel object.
free order. The animals went through one learning trial per day during ten consecutive days. Several dependent variables were recorded: the number of runs made within each trial, the number of correct trials throughout the experiment and, among those, the number of correct trials made by running the maze in a clockwise or counterclockwise fashion.
Radial maze Spatial learning performance of the same animals was estimated in a radial six-arm maze. This apparatus was made of a grey plastic material and consisted of six arms (7 x 35 cm) radiating from a hexagonal platform. Identical food cups were placed at the end of each arm. The maze was elevated 60 cm above the floor, uniformly illuminated, maintained in a constant orientation during the whole experiment, and surrounded by four cardboard walls forming a cubic enclosure. Three days before the pretraining began, the animals were slightly food deprived in order to bring their weights t o 85%of the initial level. For three consecutive days, each mouse was placed on the central platform of the maze and allowed to explore the apparatus on which a large amount of food had been placed. On the fourth day, the actual testing procedure started. Each mouse was put on the central platform of the maze, all the arms being baited by a 25 mg piece of food, and allowed to collect food with a maximum of 15 runs. A correct trial consisted in visiting all the six paths in six runs, in a
Pain thresholds For each animal, the pain threshold was determined in a cubic box (35 x 35 x 35 cm) made of grey plastic material with a grid floor connected t o a scrambled shock producer. After the animal had been introduced in the box, the potentiometer was manipulated in order to progressively increase the intensity of the current delivered through the grid floor. The pain threshold expressed in pA corresponded to the minimal intensity eliciting vocalization in the mouse.
Active avoidance Active avoidance performance of four month-old naive irradiated and control mice (6 males and 6 females from three litters for each treatment) were measured in the previously described apparatus used for testing locomotor activity. In the shuttle box training, a light (10 W) was switched on alternatively in the two compartments and used as the conditioned stimulus (CS). The CS preceded the onset of the unconditioned stimulus (US) by 5 sec and overlapped it for 25 sec. The US was an electric shock (0.2 mA) applied continuously to the grid floor. The intertrial interval was 30 sec. An avoidance response was recorded when the animal avoided the US by running into the dark compartment within 5 sec after the onset of the CS. The subjects went through five consecutive daily 100-trial sessions.
Statistical analysis The data concerning the mean body weights from birth until postnatal day 16, the mean ages of occurrence of each reflex, and the motility scores at 7 and 17 days of age were statistically analyzed by a Mann & Whitney U-test, each litter being the unit of analysis. Postweaning measures were compared by parametric tests. Locomotor activity rates were analyzed by a one factor (treatment: irradiation vs. sham irradiation) ANOVA, open field, and novelty scores
64
D. DI CICCO ET AL.
TABLE 1. Weight gain during Dostnatal development Postnatal dav 3
Irradiated
5
1.66 (3) 2.33 (3)* k. 16 k.17
10
12
3.83 (3)* k.33
4.16 (3)** k.66
Mean ( + S.E.) body weights per litter in grams. In parentheses No. of litters. * P = .05. **P=.02.
as well as pain thresholds by a two factor (treatment x sex) ANOVA, radial maze, and active avoidance performance by a three factor (treatment x sex x repeated measures) ANOVA. The frequencies of correct trials and of clockwise correct sequences of runs were compared by a x2 test. RESULTS
Body weight and reflexes Mean body weights and the values of the statistical comparisons are reported on Table 1. Measurements performed during development in the eight selected litters (three irradiated and five control) indicated no difference in body weights at birth according t o the treatment but a reduced weight gain in irradiated pups during the first 12 days of postnatal life. The occurrence of reflexes involving body motor control such as cliff aversion (U = 0, P = .02), bar holding (U = 0, P = .02) and, to a lesser extent, righting (U = 2, P = .07),was delayed in irradiated litters. Conversely, reflexes involving fine motor behavior (grasping, U = 4 , P=.19, ns) and forelimb placing, (U = 3, P = .12, ns) were present at about the same age in irradiated and control litters. Other parameters related to sensorial development (ears-, U=O, P = .02, and eyes-, U = 0 , P = .02, opening) or maturation (hair growth, U = 2, P = .07) were also delayed in irradiated litters. Finally, in adulthood, body weights of 6 males and 6 females from three irradiated and three control litters, were lower in females than in males (significant sex effect, F(1,20) = 77.01, P
Prenatal Exposure to Gamma/Neutron Irradiation: Sensorimotor Alterations and Paradoxical Effects on Learning DANIELA DI CICCO, SARA ANTAL, AND MARTINE AMMASSARI-TEULE Istituto di Psicobiologia e Psicofarmacologia del CNR, Via Reno, 1 00198 Rome, Italy (D.D.C., M.A.-T.);'%rideric Joliot-Curie" National Research Institute for Radiobiotogy and Radiohygiene, Budapest, 1775, Hungary (S.A.)
ABSTRACT The effects of prenatal exposure on gamma/neutron radiations (0.5 Gy at about the 18th day of fetal life) were studied in a hybrid strain of mice (DBA/Cne males x C57BL/Cne females). During ontogeny, measurements of sensorimotor reflexes revealed in prenatally irradiated mice 1)a delay in sensorial development, 2) deficits in tests involving body motor control, and 3) a reduction of both motility and locomotor activity scores. In adulthood, the behaviour of prenatally irradiated and control mice was examined in the open field test and in reactivity to novelty. Moreover, their learning performance was compared in several situations. The results show that, in the open field test, only rearings were more frequent in irradiated mice. In the presence of a novel object, significant sex x treatment interactions were observed since ambulation and leaning against the novel object increased in irradiated females but decreased in irradiated males. Finally, when submitted to different learning tasks, irradiated mice were impaired in the radial maze, but paradoxically exhibited higher avoidance scores than control mice, possibly because of their low pain thresholds. Taken together, these observations indicate that late prenatal gamma/ neutron irradiation induces long lasting alterations at the sensorimotor level which, in turn, can influence learning abilities of adult mice. Teratogenic effects of ionizing (X- or neu- distribution and differentiation of neurons tron) radiations on development have been (Hayashi et al., '78; Kameyama et al., '78), widely described after either incidental or as well as cell piknosis and neuroblast deexperimental exposure of mammalian em- struction (Cowen and Geller, '60; Altman bryos and fetuses (Hicks and D'Amato, '66; and Nicholson, '71; Antal and Fulop, '86) Cairnie et al., '74; Brent, '77; Vogel and have been regularly observed in embryos Antal, '84; Fonagy et al., '85; Jensh and and fetuses of rodents dying some hours afBrent, '88a-c). The extent of the damage ter the irradiation or in young animals sachas been found to depend on several factors, rificed during postnatal development. In the such as dose and duration of exposure, pe- case of X-irradiation, these cellular alterriod of gestation, and species or strain con- ations have been found to be related particsidered. Since microcephaly associated with ularly to early-developing structures such mental retardation is the predominant fea- as the cerebellum and spinal cord (Inouye, ture of prenatally irradiated human sub- '79; Cerda, ?33), and, to a lesser extent, the jects (Mole, '82; Otake and Schull, '84), nu- neostriatum and cortex (Kameyama et al., merous investigations in the rat and the '78; Hoshino et al., '80). These morphologic observations have led mouse have attempted to identify alterations of brain structures underlying behav- several authors to focus, at the behavioral ioral modifications. Histological data have level, on possible change in activities inyielded a general framework of the brain sensitivity to X- or neutron rays. Disturbed Received June 7,1988; accepted August 15,1990 cell proliferation and migration, abnormal 0 1991 WILEY-LISS, INC.
62
D.DI CICCO ET AL.
havioral testing. These pups were considered as irradiated at about the 18th day of fetal life. Eight other hybrid females from the remaining 12 served as controls. Those females were sham irradiated, i.e., placed in the rotating chamber of the reactor but without irradiation. Experimental design The following physiological and behavioral measurements were performed in pups exposed to each treatment. Body weights and reflexes Three litters from irradiated mothers involving 4, 5, and 6 pups and five control litters of 4-6 pups each were selected and tested every other day from postnatal day 2 to postnatal day 16. The following items were recorded for the pups: body weights, rooting, righting, grasping, forelimb placing, cliff aversion, hair growth, opening of ears, and eyes, and auditory startle response, according to a previously described method (Castellano and Ammassari-Teule, '84). The age at which a given response was present in all pups of each litter indicated METHODS the score. Finally, body weights were Animals checked in adult irradiated and control anThe mice used in these experiments were imals before learning tests. F1 hybrids produced by mating male DBAI2 Motility with females C57BL/6 strains. Adult (2-3 Motility rates were examined in the same months) nulliparous hybrid females (N = 20) animals on postnatal days 7, 17, and subsewere individually housed with food and wa- quently at 3 months of age. For each mouse, ter ad lib. Males (2-3 months) were intro- the motility score was measured during 5 duced into the cages (ratio 1:l)and left for 1 minutes by a selective activity meter (Anweek. The day after the males had been in- imex). troduced in the cages was considered as day Locomotor activity 1 of pregnancy. The females were weighed regularly and isolated when determined to At 1month of age, the same animals were be pregnant. Eighteen days after mating, also tested for locomotor activity in Plexithe females were irradiated according to the glas toggle-floor boxes. Each box was difollowing procedure. Each female was put vided into two 20 x 10 compartments sepinto a plastic animal chamber rotating axi- arated with a black partition having a 3 x ally and inserted in the cavity of the 3 cm opening at the floor level to connect research reactor Tapiro (from the Italian the compartments. Each crossing through National Center for Alternative Energy- the opening tilted the grid floor, thus actiENEA-Rome) at a distance of 140 cm from vating a microswitch connected with a the core. This reactor produced variable counter. neutrodgamma (nly) dose rates. The single Reactivity to open field and novelty whole body dose was approximately 0.5 Gy Irradiated and control 4-month-old naive neutrons, this neutron dose being accompanied by approximately one-third of gamma animals (6 males and 6 females randomly rays, i.e., 0.18 Gy. Since in CD1 mice deliv- selected from three litters for each treatery generally occurs after 19 days of preg- ment) were examined in the open field sitnancy, only pups born between 24 and 36 uation and, successively, tested for reactivhours after irradiation were retained for be- ity to novelty after the introduction of a
volving motor control. In fact, motor coordination (Sharp, '61; Werboff et al., '61), locomotor activity (Lipton, '661, and, in some cases, maze learning (Furchtgoff et al., '58; Werboff et al., '62; Shibagaki et al., '81; Seo et al., '82), were profoundly disturbed in developing and adult animals prenatally exposed to X-rays. However, in the case of neutron radiation-which appears to produce more deleterious biological effects than X-irradiation for equivalent doses in mouse embryos (Di Majo et al., '81)-a more general pattern of early morphological alterations has been detected. This makes the search for specific functional deficits rather difficult. The present study has been designed to investigate the behavioral alterations due to prenatal gammaheutron irradiation in mice. A longitudinal study has been conducted in order to identify modifications of several physiological and behavioral parameters from birth to adulthood, with a particular attention devoted to learning abilities of animals irradiated during their fetal life.
PRENATAL EXPOSURE TO GAMMAINEURON IRRADIATION
63
novel object in the apparatus. The open field was a grey Plexiglas box consisting of a square floor (30 x 30 cm) surrounded by 15 cm high walls. The floor was painted white and divided into 36 squares by drawing black lines. Each mouse was introduced in the center of the open field and retained for 15 sec in a transparent Plexiglas cylinder. After the removal of the cylinder, behavioral recordings were performed for 5 minutes. At the end of this period, a transparent Plexiglas box (8.5 x 8.5 x 11 cm) with holes irregularly distributed on each side was placed in the center of the open field. After a free observation period of 15 sec, behavioral recordings were performed for 3 min. The mouse was then removed from the open field and the novel object cleaned with alcohol. The open field was placed in a soundproof chamber and the behavior of each animal was recorded by a videotaping system. The videotapes were then examined and the following behavioral items were selected by means of a keyboard connected to an Apple IIe P.C.: number of crossed squares, number of rearing responses (against the walls or not), time spent freezing and grooming, and finally time spent sniffing and rearing against the novel object.
free order. The animals went through one learning trial per day during ten consecutive days. Several dependent variables were recorded: the number of runs made within each trial, the number of correct trials throughout the experiment and, among those, the number of correct trials made by running the maze in a clockwise or counterclockwise fashion.
Radial maze Spatial learning performance of the same animals was estimated in a radial six-arm maze. This apparatus was made of a grey plastic material and consisted of six arms (7 x 35 cm) radiating from a hexagonal platform. Identical food cups were placed at the end of each arm. The maze was elevated 60 cm above the floor, uniformly illuminated, maintained in a constant orientation during the whole experiment, and surrounded by four cardboard walls forming a cubic enclosure. Three days before the pretraining began, the animals were slightly food deprived in order to bring their weights t o 85%of the initial level. For three consecutive days, each mouse was placed on the central platform of the maze and allowed to explore the apparatus on which a large amount of food had been placed. On the fourth day, the actual testing procedure started. Each mouse was put on the central platform of the maze, all the arms being baited by a 25 mg piece of food, and allowed to collect food with a maximum of 15 runs. A correct trial consisted in visiting all the six paths in six runs, in a
Pain thresholds For each animal, the pain threshold was determined in a cubic box (35 x 35 x 35 cm) made of grey plastic material with a grid floor connected t o a scrambled shock producer. After the animal had been introduced in the box, the potentiometer was manipulated in order to progressively increase the intensity of the current delivered through the grid floor. The pain threshold expressed in pA corresponded to the minimal intensity eliciting vocalization in the mouse.
Active avoidance Active avoidance performance of four month-old naive irradiated and control mice (6 males and 6 females from three litters for each treatment) were measured in the previously described apparatus used for testing locomotor activity. In the shuttle box training, a light (10 W) was switched on alternatively in the two compartments and used as the conditioned stimulus (CS). The CS preceded the onset of the unconditioned stimulus (US) by 5 sec and overlapped it for 25 sec. The US was an electric shock (0.2 mA) applied continuously to the grid floor. The intertrial interval was 30 sec. An avoidance response was recorded when the animal avoided the US by running into the dark compartment within 5 sec after the onset of the CS. The subjects went through five consecutive daily 100-trial sessions.
Statistical analysis The data concerning the mean body weights from birth until postnatal day 16, the mean ages of occurrence of each reflex, and the motility scores at 7 and 17 days of age were statistically analyzed by a Mann & Whitney U-test, each litter being the unit of analysis. Postweaning measures were compared by parametric tests. Locomotor activity rates were analyzed by a one factor (treatment: irradiation vs. sham irradiation) ANOVA, open field, and novelty scores
64
D. DI CICCO ET AL.
TABLE 1. Weight gain during Dostnatal development Postnatal dav 3
Irradiated
5
1.66 (3) 2.33 (3)* k. 16 k.17
10
12
3.83 (3)* k.33
4.16 (3)** k.66
Mean ( + S.E.) body weights per litter in grams. In parentheses No. of litters. * P = .05. **P=.02.
as well as pain thresholds by a two factor (treatment x sex) ANOVA, radial maze, and active avoidance performance by a three factor (treatment x sex x repeated measures) ANOVA. The frequencies of correct trials and of clockwise correct sequences of runs were compared by a x2 test. RESULTS
Body weight and reflexes Mean body weights and the values of the statistical comparisons are reported on Table 1. Measurements performed during development in the eight selected litters (three irradiated and five control) indicated no difference in body weights at birth according t o the treatment but a reduced weight gain in irradiated pups during the first 12 days of postnatal life. The occurrence of reflexes involving body motor control such as cliff aversion (U = 0, P = .02), bar holding (U = 0, P = .02) and, to a lesser extent, righting (U = 2, P = .07),was delayed in irradiated litters. Conversely, reflexes involving fine motor behavior (grasping, U = 4 , P=.19, ns) and forelimb placing, (U = 3, P = .12, ns) were present at about the same age in irradiated and control litters. Other parameters related to sensorial development (ears-, U=O, P = .02, and eyes-, U = 0 , P = .02, opening) or maturation (hair growth, U = 2, P = .07) were also delayed in irradiated litters. Finally, in adulthood, body weights of 6 males and 6 females from three irradiated and three control litters, were lower in females than in males (significant sex effect, F(1,20) = 77.01, P
