Lippincott Williams & Wilkins, Inc., Philadelphia
Developmental Neurotoxicity of Phenytoin on Granule Cells and Purkinje Cells in Mouse Cerebellum
Article first published online: 25 DEC 2001
Journal of Neurochemistry
Volume 72, Issue 4, pages 1497–1506, April 1999
How to Cite
Ohmori, H., Ogura, H., Yasuda, M., Nakamura, S., Hatta, T., Kawano, K., Michikawa, T., Yamashita, K. and Mikoshiba, K. (1999), Developmental Neurotoxicity of Phenytoin on Granule Cells and Purkinje Cells in Mouse Cerebellum. Journal of Neurochemistry, 72: 1497–1506. doi: 10.1046/j.1471-4159.1999.721497.x
Abbreviations used: ABC, avidin-biotin-peroxidase complex; BrdU, 5-bromo-2′-deoxyuridine; EGL, external granular layer; IGL, internal granular layer; IP3R, inositol 1,4,5-trisphosphate receptor; IP3R1, inositol 1,4,5-trisphosphate receptor type 1; ML, molecular layer; PHT, phenytoin; PN, postnatal day(s).
- Issue published online: 25 DEC 2001
- Article first published online: 25 DEC 2001
- Developmental neurotoxicity;
- Neonatal period;
- Mouse cerebellum;
- Inositol 1,4,5-trisphosphate receptor type 1;
- Motor coordination
Abstract: Phenytoin (PHT) is a primary antiepileptic drug. Cerebellar malformations in human neonates have been described following intrauterine exposure to PHT. The neonatal period of development in the cerebellum in mice corresponds to the last trimester in humans. To examine the neurotoxic effects of PHT in the developing cerebellum, we administered PHT orally to newborn mice once a day during postnatal days 2-4. We observed many apoptotic cells in the external granular layer (EGL) on postnatal day 5, labeled cells in the EGL still remaining 72 h after labeling with 5-bromo-2′-deoxyuridine, and EGL thicker than that in the control on postnatal day 14. These results showed that PHT induced cell death of external granule cells and inhibited migration of granule cells in cerebella. In specimens immunostained with antibody against inositol 1,4,5-trisphosphate receptor type 1, Purkinje cells in the treated group had poor and immature arbors, and partially showed an irregular arrangement. The motor performance of the treated mice in a rotating rod test was impaired, although there were no changes in muscular strength or in walking pattern at the period of maturity. These findings indicate that PHT induces neurotoxic damage to granule cells and Purkinje cells in the developing cerebellum and impairs selected aspects of motor coordination ability.
Phenytoin (PHT) is a commonly used drug for all types of epilepsy except absence seizures (McNamara, 1996). PHT is a weak acid with a pKa of ∼8.3, and its aqueous solubility is limited, even in the intestine (McNamara, 1996). Significant differences in bioavailability of oral pharmaceutical preparations have been manifested (McNamara, 1996). Acute or chronic PHT administration to epileptic patients may cause cerebellar dysfunction and cerebellar degeneration (McClain et al., 1980); Baier et al., 1984; Luef et al., 1994). Gadisseaux et al. (1984) reported a case with pontocerebellar hypoplasia following intrauterine exposure to PHT in humans. Prenatal PHT exposure has been demonstrated to induce various neurobehavioral effects in experimental animals (Vorhees et al., 1988; Adams et al., 1990; Pizzi and Jersey, 1992). Recently, we have shown developmental neurotoxicity of PHT in the cerebellum during the neonatal period (Ohmori et al., 1992, 1997).
The neonatal period of development in the CNS in mice corresponds to the last trimester in humans (Dobbing and Sands, 1973; Jacobson, 1991b). In the mouse cerebellum, the cortex undergoes complex ontogenesis during the first 3 postnatal weeks (Altman, 1972a,b,c, 1982; Rodier, 1980; Jacobson, 1991b; Hatten and Heintz, 1995). Very active cell proliferation and differentiation occur in the external granular layer (EGL) after birth (Jacobson, 1991b; Hatten and Heintz, 1995). From postnatal days (PN) 3 to 10, granule cells migrate along the Bergmann glial fibers from the EGL to the internal granular layer (IGL), and axons of granule cells (parallel fibers) grow to form connections with extending Purkinje cell dendrites. The dendritic arborization of Purkinje cells is fully developed by PN 21 (Altman, 1972a,b,c, 1982; Jacobson, 1991b; Hatten and Heintz, 1995).
Cerebellar damage induced by PHT administered to newborn mice may thus be a useful model for the investigation of the mechanisms of cerebellar malformations associated with PHT exposure to pregnant epileptics.
Traditionally, the study of cell proliferation has involved the use of [3H]thymidine to enable monitoring of DNA synthesis in individual cells by autoradiography, and hence identify proliferating cells or cell populations. More recently, an alternative rapid nonradioactive technique has been developed by use of a thymidine analogue, 5-bromo-2′-deoxyuridine (BrdU), which is incorporated into replicating DNA (S phase). BrdU in cell nuclei can be revealed subsequently by a specific monoclonal antibody (Gratzner, 1982).
Apoptosis is associated with physiological or programed cell death, in contrast with necrosis, which is associated with cell injury. The in situ presence of numerous DNA strand breaks is a typical feature of apoptotic cells. Selective DNA strand break induction by photolysis at sites that contain incorporated halogenated DNA precursors has been proposed recently as a method of analyzing DNA replication. Detection of DNA strand breaks thus enables one to identify apoptotic and/or DNA replicating cells (Li and Darzynkiewicz, 1995).
The inositol 1,4,5-trisphosphate receptor (IP3R) is a specific intracellular calcium release channel protein, which has been demonstrated in several cell types, such as neurons and secretory cells (Berridge, 1993; Furuichi and Mikoshiba, 1995; Mikoshiba, 1997). IP3R type 1 (IP3R1) is a major neuronal member of the IP3R family in the CNS, predominantly enriched in cerebellar Purkinje cells (Maeda et al., 1988, 1989; Furuichi et al., 1989; Ross et al., 1989). The developmental expression pattern of IP3R1 in the mouse cerebellum has been consistent with the maturation of Purkinje cells (Maeda et al., 1989; Worley et al., 1989; Nakanishi et al., 1991; Ryo et al., 1993).
In this study, we administered a sesame oil suspension of low-dose PHT orally to newborn mice, and examined neurotoxic effects on the apoptosis, proliferation, and migration of external granule cells and the dendritic development of Purkinje cells in the developing cerebellum by histological, especially immunohistochemical studies, using monoclonal antibodies against BrdU and IP3R1 and the method for detection of apoptotic cells, as well as effects on the motor function by a battery of behavioral tests.
MATERIALS AND METHODS
Jcl:ICR mice, purchased from Japan Clea Co., Ltd. (Tokyo, Japan), were bred in our laboratory at 22 ± 2°C with 50 ± 10% humidity. The mice were mated overnight, and the day on which a vaginal plug was found was designated as gestational day 0. Pregnant mice were housed separately in a plastic cage. We used only mice delivered spontaneously on gestational day 19. The day of birth was designated as PN 0. Pups were marked directly by branding on their bodies and were marked again by coloring with picric acid on PN 10. The pups of each litter, without regard to sex, were divided randomly into treated and control groups. In total, 25 litters were used for morphological, immunohistochemical, and behavioral studies and determination of PHT concentrations.
A fine powder of PHT with a mean particle size of 17 μm was specially prepared from commercial PHT free acid (Dainippon Pharmaceutical Co., Ltd., Tokyo, Japan) by filtration through a fine mesh. PHT was suspended in sesame oil (Maruishi Pharmaceutical Co., Ltd., Osaka, Japan) by an ultrasonic homogenizer (model UR-200P, Tomy Seiko Co., Ltd., Tokyo, Japan) to make PHT suspensions. To the pups assigned to the treated group, the sesame oil suspension of PHT was administered orally through a polyethylene tube (0.28 mm i.d.) by a hypodermic syringe with a 30-gauge needle at a dose volume of 10 ml/kg of body weight (corresponding to 35 mg of PHT/kg of body weight) once a day during PN 2-4. In the control group, sesame oil alone was administered at a volume of 10 ml/kg of body weight once a day for the corresponding period. The pups were weighed every day from birth to PN 4, and thereafter weighed weekly.
Proliferation and migration of cerebellar granule cells were studied after a single intraperitoneal injection of BrdU using the Amersham cell proliferation kit (Amersham International plc, Amersham Place, Little Chalfont, Buckinghamshire, U.K.) (60 mg/kg of body weight) 3 h after the last PHT treatment on Pn 4. Some of the animals in the control and treated groups were killed with overdoses of diethyl ether. Brains were removed 1, 24, or 72 h. after BrdU labeling and fixed with Bouin’s solution for 2 h. Dewaxed sagittal sections (8 μm thick) were treated with 1 mol/L hydrochloric acid at 37°C for 20 min. These sections were then incubated for 1 h in anti-BrdU monoclonal antibody and subsequently incubated for 30 min in peroxidaselabeled goat anti-mouse IgG. The immunoreacted section was stained by incubating with diaminobenzidine. The section was counterstained with eosin.
The presence of numerous DNA strand breaks appears to be a very specific marker of apoptosis. A common method of detection for apoptosis is based on specific labeling of DNA strand breaks (Li and Darzynkiewcz, 1995). The TUNEL (terminal deoxytransferase-mediated dUTP-biotin nick end-labeling) method is based on the specific binding of terminal deoxynucleotidyl transferase to 3′-OH ends of DNA, resulting in a polydeoxynucleotidase polymer (Gavrieli et al., 1992). Recently, commercial in situ end-labeling kits have become available (Naruse and Keino, 1995). The samples were processed using the terminal deoxynucleotidyl transferase kit (Apop Tag, ONCOR Inc., Gaithersburg, MD, U.S.A.).
Neurotoxic effects of PHT on the development of Purkinje cells were studied spatiotemporally by immunostaining with an anti-IP3R1 antibody. Dissected brains (PN 7, 10, 14, or 21) were fixed in Bouin’s solution for 2 h. Paraffin sections (8 μm thick) in the parasagittal plane were processed sequentially as follows (Maeda et al., 1989; Ryo et al., 1993): (a) in 0.3% hydrogen peroxide/methanol for 30 min to block endogenous peroxidase; (b) in 2% normal rabbit serum/phosphate-buffered saline for 1 h; (c) in the monoclonal antibody against IP3R1 (18A10) (Maeda et al., 1988) for 2 h; (d) in a biotinylated anti-rat IgG solution for 1 h; and (e) in an avidin-biotin-peroxidase complex (ABC) solution for 1 h. The immunoreacted section was stained by incubating with diaminobenzidine solution. The sections were counterstained with hematoxylin. Biotinylated anti-IgG and the ABC solution were from the Vectastain Elite ABC kit (Vector Laboratories, Inc., Burlingame, CA, U.S.A.). The other paraffin sections were stained with hematoxylin and eosin.
PHT concentrations in the plasma and brain were determined by HPLC (Ohmori et al., 1992). In our previous study, plasma PHT levels 3 h after administration on the third day of PHT treatment reached a steady state (Ohmori et al., 1992). Therefore, PHT levels in the plasma and brain were determined 3 h after the last oral administration of PHT. The mice were anesthetized with diethyl ether, and ∼ 100-150 μl of blood was collected into heparinized tubes. The blood was centrifuged at 3,000 rpm for 5 min with a microcentrifuge to obtain the plasma fraction. The brains were removed and also frozen at -30°C immediately until analysis. The brain homogenates were prepared using a glass homogenizer. An internal standard of 5-(4-methylphenyl)-5-phenylhydantoin (Aldrich Chemical Co., Milwaukee, WI, U.S.A.) was added to give a concentration of 10 μg/ml of the plasma and brain homogenates. The plasma and brain samples were then extracted with 7 ml of ethyl acetate by vigorously shaking for 20 min. The extract was evaporated to dryness under reduced pressure and was reconstituted in methanol. The solvent system for HPLC was acetonitrile/water/acetic acid (2:3:0.05, by volume). The flow rate was 1.0 ml/min. Fifty microliters of the sample was injected into an HPLC column. A column of TSKgel ODS-80TMCTR (100 × 4.6 mm) (Tosoh Manufacturing Co., Ltd., Tokyo, Japan) was used. The compounds were detected at 235 nm with a UV-970 wavelength detector (Jasco, Tokyo, Japan). PHT was eluted from the column in 5.2 min. The retention time of the internal standard was 8.3 min.
On PN 56, brains were removed, and total brain weight and cerebellar weight were determined. Statistical comparisons of differences between the treated and control groups were made by Student’s t test.
Thirteen PHT-treated and 11 control female mice were subjected to a battery of behavioral tasks for evaluating motor function 8-9 weeks after birth. Motor coordination was assessed with a rotating rod apparatus (KN-75, Natsume Seisakujo Co., Ltd., Tokyo, Japan), which consisted of a plastic rod (3 cm diameter, 8 cm long) with a gritted surface flanked by two large discs (40 cm diameter). A mouse was placed on the rod, and the rod was rotated at a speed of 0 (stationary), 5, 10, and 20 rpm. Latency until a fall occurred was recorded for four trials at each speed. Spontaneous locomotor activity in a Plexiglas box (30 × 20 × 13 cm) was monitored for 60 min with a SCANET (Tokyo Sangyo Co., Ltd., Toyama, Japan), which detected the movement of an animal in terms of the number of times an infrared beam was broken. The grip strength of a mouse was measured with a traction apparatus (FU-1, Muromachi Kikai Co., Ltd., Tokyo, Japan) to which a horizontal stainless steel bar (2 mm diameter) was attached for holding. A mouse was lifted by the tail and made to grasp the horizontal bar with the forepaws. The experimenter slowly pulled the mouse back by the tail and the maximum tension in the cable was recorded. To obtain footprints, black ink was applied to the hindpaws of each mouse, and the mouse was observed to walk forward spontaneously in a narrow alley (9 × 25 × 10 cm) on white paper. All behavioral experiments were done between 10 a.m. and 3 p.m. by a well trained experimenter who had no knowledge of the treatment of the mice. Behavioral data were analyzed by one- or two-way ANOVA with repeated measures.
Some of the pups treated with PHT showed clinical symptoms, such as anorexia, motor hypoactivity, and motor incoordination. Total mortality was 30.4% in males and 33.3% in females in the treated group and none in the control group (Table 1). At the termination of PHT administration (PN 4), pups treated with PHT exhibited significant lower body weight, but the treated pups subsequently gained weight and recovered the control body weight level by PN 56 (Table 1).
|No. of mice examined||16||16||13||16|
|Body weight (g)||46.0 ± 3.2||40.7 ± 2.5||35.0 ± 3.0||32.6 ± 3.2|
|Total brain weight (mg)||524.4 ± 15.1||458.3 ± 21.1a||521.7 ± 11.3||454.8 ± 23.3a|
|Cerebral weight (mg)||378.7 ± 10.8||335.4 ± 12.4a||379.8 ± 9.4||331.4 ± 17.8a|
|Cerebellar weight (mg)||73.7 ± 3.6||58.3 ± 2.8a||71.2 ± 4.5||56.9 ± 3.8a|
Total brain weight and cerebellar weight were reduced significantly in the treated group compared with controls on PN 56 (Table 1).
The plasma PHT level was 17.7 ± 2.0 μg/ml 3 h after administration on the third day of PHT treatment. The PHT level in the brain was 31.4 ± 4.3 μg/g, significantly higher than that in the plasma (Fig. 1). No difference in PHT levels in the plasma and brain was seen between males and females.
Pyknotic cells in the EGL increased in the treated group compared with the control (Fig. 2A and B), and by using the Apop Tag kit for detection of apoptosis, we observed many apoptotic cells in the treated group compared with the control (Fig. 2C and D), especially in the vermis on PN 5 (24 h after the last PHT administration). Labeled cells in the EGL 1 h after BrdU labeling were almost the same in the control and the treated groups. Cells in the EGL were heavily labeled with BrdU in the control group 24 h after BrdU treatment, but labeled cells decreased in the EGL in the treated group (Fig. 3A and B). In the control group 72 h after the BrdU injection, the labeled cells were not seen in the EGL and detected only in the molecular layer (ML) and the IGL (Fig. 3C). In contrast, the labeled cells were still observed in the EGL in the treated group 72 h after the treatment with BrdU (Fig. 3D). In addition, the EGL in the vermis in the treated group was thicker than that of the control group, and the ML in the vermis in the treated group was thinner than that of the control group on PN 14 (Fig. 4A and B).
Cell bodies and dendrites of Purkinje cells were specifically immunostained with IP3R1 in the control cerebellum. On PN 7, dendritic arbors of Purkinje cells in the ML were well immunostained, and elongated dendrites were evident from PN 7 in the control group (Fig. 5A). Dendritic arbors of Purkinje cells were also clearly observed by IP3R1 immunostaining in the control group on PN 10, 14, or 21 (Fig. 5B, C, and D). In the treated group, dendrites of Purkinje cells, however, were poorly developed and showed a dispersed and indistinct staining pattern. Dendritic arbors in the treated group appeared poorly stained and immature on PN 7 or 10 (Fig. 5E and F), and some Purkinje cells were aligned irregularly on either PN 14 or 21 (Fig. 5G and H).
At 8-9 weeks after birth, motor coordination in the treated group was impaired in the rotating rod test. The retention time of PHT-treated mice on the rod was significantly shorter than that of control mice (stationary: F1,22 = 8.0, p = 0.01; 5 rpm: F1,22 = 7.2, p = 0.01; 20 rpm: F1,22 = 5.2, p = 0.03) (Fig. 6A). In this task, learning effects were statistically significant at these speeds (stationary: F3,66 = 8.0, p = 0.01; 5 rpm: F3,66 = 11.9, p = 0.01; 20 rpm: F3,66 = 4.1, p = 0.01), whereas treatment × trials interaction in the ANOVA showed that learning effects were independent of treatment (stationary: F3,66 = 0.85, p = 0.47; 5 rpm: F3,66 = 0.61, p = 0.61; 20 rpm: F3,66 = 0.27, p = 0.85). In a novel environment, PHT-treated mice showed decreased locomotor activity (F1,22 = 7.3, p = 0.01) (Fig. 6B). Grip strength of the forepaws of PHT-treated mice was comparable to that of control mice (F1,22 = 0.15, p = 0.70) (Fig. 6C). No significant treatment effects were revealed in stride length and width of hindlimbs during walking (stride length: F1,22 = 0.0003, p = 0.99; step width: F1,22 = 0.067, p = 0.80) (Fig. 6D).
PHT is one of the most widely used antiepileptic agents, and effective in most forms of epilepsy except absence seizures (McNamara, 1996). Many epileptic women take PHT during pregnancy (Kelly, 1984; Kelly et al., 1984; Lander and Eadie, 1991). Clinicoepidemiologic studies suggest an approximate two- to threefold increase in malformations in neonates exposed to PHT during the prenatal period (Janz, 1982; Kaneko et al., 1988). Cerebellar malformations in human neonates have been reported following intrauterine exposure to anticonvulsant drugs, including PHT (Mallow et al., 1980; Gadisseaux et al., 1984; Squier et al., 1990). Squier et al. (1990) have demonstrated no Purkinje cells in any part of the cerebellar hemispheres and cerebellar hypoplasia accompanied by extensive gliosis. Novel cases of pontoneocerebellar hypoplasia have been shown in children born to mothers treated with PHT during pregnancy (Gadisseaux et al., 1984; Squier et al., 1990). These have been classified clinically and pathologically as a subtype of congenital pontocerebellar hypoplasia (Barth, 1993).
The perinatal period of cerebellar development in mice corresponds to the last trimester in humans (Kameyama, 1985; Jacobson, 1991b). In the mouse cerebellum, granule cells are generated in the EGL after birth and migrate through the ML and Purkinje cell layer to form the IGL (Altman, 1972a,b,c, 1982; Jacobson, 1991b). In all vertebrates, the EGL increases in thickness from a single layer of cells to a layer six to eight cells deep as a result of proliferation of external granule cells (Jacobson, 1991b). Mitotic figures are scattered throughout the EGL during the period of its increase in thickness (Jacobson, 1991b). Proliferation and migration of cerebellar granule cells have been studied in the newborn mouse by means of [3H]thymidine autoradiography (Fujita et al., 1966). During the initial period of increase in thickness of the EGL, cumulative labeling with [3H]thymidine results in labeling of 100% of the external granule cells. The EGL grows to a maximal thickness during the second postnatal week, then gradually regresses (Rodier, 1980). Proliferation in the EGL ceases after PN 10 (Fujita et al., 1966). The external granule cells can produce at least 14 generations in the first 14 days after birth with generation times of 16-20 h, whereas the total number of cells (DNA content of the cerebellum) actually increases six times. The evidence indicates that many external granule cells cease their DNA synthesis and mitosis before PN 14 (Jacobson, 1991b). The maximal transit time of external granule cells is estimated at 42 h, and the speed of migration of cerebellar granule cells is ∼100 μm/day in the mouse (4.2 μm/h) (Fujita et al., 1966). The EGL disappears at about PN 20 in the mouse (Jacobson, 1991b; Hatten and Heintz, 1995).
Dendritic growth of the Purkinje cells is shown by the increase in width of the ML of the cerebellar cortex, which occurs from PN 4 to 14 in the mouse (Jacobson, 1991b). The final stage in Purkinje cell differentiation occurs during the second and third postnatal weeks, when Purkinje cells undergo a period of rapid growth that results in the elaboration of their characteristic dendritic arbor and formation of synaptic contacts at specific sites along these arbors, i.e., dendritic spines (Hatten and Heintz, 1995).
Agents acting in the perinatal or early postnatal periods are thus likely to interfere with the development of cerebellar granule cells, and other cerebellar cells may be affected as a result of the loss of granule cells (Altman et al., 1969; Yamano et al., 1983; Bejar et al., 1985). Cytotoxic depletion of Purkinje cells results in a proportional reduction in granule cells (Jacobson, 1991b). After destruction of the cerebellar granule cells, a variety of changes have been seen in the other cells of the cerebellum (Jacobson, 1991b). Some of the changes, for example, the stunting of Purkinje cell dendrites, may be ascribed to the removal of normal afferents to the dendrites of Purkinje cells (Jacobson, 1991b). Granule cell-deprived cerebellar hypoplasia can result from different causes that principally affect proliferation, migration, and early synaptogenesis of granule cells (Barth, 1993).
We have shown previously that oral administration of PHT in the neonatal period induces neurotoxic effects on the developing mouse cerebellum (Ohmori et al., 1992, 1997). In one study (Ohmori et al., 1992), we administered 50 mg/kg PHT suspended in sesame oil orally once a day during PN 2-14, and we observed pyknotic cells in the EGL, wide EGL, and reduction in cerebellar weight, together with retardation of motor and behavioral development. The plasma PHT levels were 34-36 μg/ml on the third day of PHT treatment. These plasma levels correspond to the toxic ranges in humans. Therefore, it is no wonder that neurobehavioral effects of PHT on the cerebellum were seen. We also have reported that oral administration of low-dose PHT to newborn mice once a day during PN 2-4 produced pyknotic cells in the EGL, wide EGL, and reduction of cerebellar weight in the 25 and 35 mg/kg PHT-treated groups (Ohmori et al., 1997). These plasma levels correspond to the therapeutic ranges in pregnant epileptics (Lander and Eadie, 1991). In the present study, we administered 35 mg/kg PHT orally to newborn mice and examined neurotoxic effects on developing cerebellum by morphological and immunohistochemical studies, and a battery of behavioral tests.
PHT has been shown to induce apoptotic cell death of cultured rat cerebellar granule neurons (Yan et al., 1995), and PHT applied to developing mouse cerebellum in tissue culture induces Purkinje cell degeneration (Blank et al., 1982).
In the present histological and immunohistochemical studies, we found many apoptotic cells in the EGL on PN 5 (24 h after the last PHT treatment), labeled cells in the EGL 72 h after BrdU labeling (PN 7), and EGL thicker than in the control and ML thinner than in the control on PN 14. These results indicate that PHT induces cell death of the external granule cells and inhibits proliferation and migration of granule cells in newborn mouse cerebella.
Inositol 1,4,5-trisphosphate, which mediates the release of intracellular calcium ions, is generated from phosphatidyl inositol 4,5-bisphosphate in response to stimulation of a variety of neurotransmitter receptors (Berridge, 1993; Furuichi and Mikoshiba, 1995; Mikoshiba, 1997). IP3R1 in the CNS is highly expressed and is located mostly on the smooth endoplasmic reticulum in cerebellar Purkinje cells (Mikoshiba et al., 1979; Worley et al., 1987; Maeda et al., 1988, 1989; Supattapone et al., 1988; Furuichi et al., 1989; Ross et al., 1989). The developmental expression of IP3R1 in the mouse cerebellum has been closely associated with the growth of dendritic arbors in Purkinje cells (Maeda et al., 1989; Worley et al., 1989; Nakanishi et al., 1991; Ryo et al., 1993). We observed immature dendritic development of the Purkinje cells following PHT administration to newborn mice by immunostaining with an antibody against IP3R1. To explain this immaturity, two alternative possibilities can be raised: (a) delayed migration of granule cells to the IGL may deteriorate the synaptic connection of parallel fibers with dendrites of Purkinje cells, resulting in immature development; or (b) PHT may directly damage Purkinje cells and their dendritic extensions.
The mechanism of teratogenicity by PHT is still under investigation, but an epoxide intermediate or other toxic oxidative metabolites of PHT are postulated to bind to macromolecules (DNA, RNA, proteins, etc.) in mouse fetuses, and this binding may be associated with an increased teratogenic action of PHT (Kaneko and Kondo, 1995). Proliferation and migration of granule cells, and dendritic growth of Purkinje cells in the developing mouse cerebellum, would be susceptible to neurotoxic agents like PHT, because EGL of the cerebellum is the site with the most intense proliferative and migratory activity in newborn mammals and the dendritic growth of Purkinje cells develops during the first 3 weeks after birth (Jacobson, 1991b; Hatten and Heintz, 1995).
In this study, total brain weight, cerebral weight, and cerebellar weight were reduced in the treated group compared with controls on PN 56. PHT administration to newborn pups resulted in a substantial mortality compared with vehicle-treated controls. Surviving pups treated with PHT had lower body weight during neonatal PHT treatment. It has been reported that PHT induces reduction of thyroid hormone thyroxin levels (Kaneko, 1991). Thyroid hormone is a major physiological regulator of mammalian brain development, cerebral cortex, and cerebellum (Jacobson, 1991a; Bernal and Nunez, 1995). Neurotrophins are responsible for the survival and differentiation of defined neuronal populations during CNS development and thus contribute to the formation of complex cellular networks (Bernal and Nunez, 1995). Their biological properties make them good candidates to be studied as possible targets of thyroid hormone action (Bernal and Nunez, 1995). It has been reported that control of neurotrophin expression appears to be a major mechanism in the regulation of Purkinje cell differentiation by thyroid hormone (Bernal and Nunez, 1995). Both in vitro and in vivo, active thyroid hormone 3,3′,5-triiodothyronine is able to include an increased production of neurotrophin-3 by cerebellar granule cells (Bernal and Nunez, 1995). 3,3′,5-Triiodothyronine may induce Purkinje cell differentiation by acting indirectly to increase neurotrophin-3 production by granule cells (Bernal and Nunez, 1995). We observed many apoptotic cells in the EGL immediately after the last PHT administration, immature dendritic development of Purkinje cells, and higher PHT level in the brain, corresponding to the toxic range in humans (McNamara, 1996). Thus, these findings indicate that neurotoxic damage to the developing mouse cerebellum is due to PHT’s direct neurotoxic effects. It has also been shown that malnutrition reduces brain weight and cerebellar weight in developing brain (De Guglielmone et al., 1974; Hillman and Chen, 1981). Therefore, the effects may be due to undernutrition of pups in addition to PHT’s effects.
The cerebellum is concerned with the coordination of somatic motor activity, the regulation of muscle tone, and mechanisms that influence and maintain equilibrium (Parent, 1996). The cerebellum has been involved in two important distinct functions, learning associated with component movement (discrete motor learning) and smooth performance of compound movements (motor coordination) (Chen and Tonegawa, 1997). The rotating rod test, developed many years ago to evaluate motor coordination ability (Dunham and Miya, 1957), was applied in this study. The results showed that PHT-treated mice fell earlier than control mice, whereas learning effects (lengthening the latency until fall) in this test were not different between the two groups. The motor learning ability of the PHT-treated mice was unlikely to be impaired in this study. Furthermore, muscular strength and walking pattern were found to be normal, as revealed by the traction and footprint tests. It is possible that the motor dysfunction seen in this study can be accounted for by impairment of motor coordination. Spontaneous motor activity was also found to be decreased in the treated group. The periods when hypoactivity can be induced coincide with the periods when neurons of the cerebellum are proliferating, and hypoactivity is only one of the behavioral effects that seem to be closely related to cerebellar malfunction (Rodier, 1980). It has been reported that impairment of motor coordination and decreased spontaneous activities were shown in cerebellar mutant mice (Aiba et al., 1994; Kashiwabuchi et al., 1995). In the present study, impairment of the motor coordination and decreased spontaneous locomotor activity in the treated group were observed. These findings indicate that PHT induces persistent neurobehavioral effects on the mouse cerebellum even at maturity (8-9 weeks after birth). Oral administration of PHT to newborn mice at a plasma level corresponding to the therapeutic range in humans interferes with proliferation and migration of granule cells and with dendritic arborization of Purkinje cells and impairs cerebellar motor coordination. Immunostaining with antibodies against BrdU and IP3R1 produces important markers for assessing developmental neurotoxicity of PHT in the mouse cerebellum.
In a recent report (Lander and Eadie, 1991), the dose of PHT increased in 85% of 106 pregnant epileptics, and the maximal plasma PHT concentration recorded during pregnancy was 7-25 μg/ml in 81 pregnant patients. On the basis of available data in humans, it has been shown that for most antiepileptic drugs, including PHT, plasma concentrations in the umbilical vein and in maternal venous blood are nearly identical (Kaneko and Kondo, 1995). The neonatal period of cerebellar development in mice corresponds to the last trimester in humans (Jacobson, 1991b). Animal studies indicate that PHT dosing that produces serum levels lower than 25 μg/ml in pregnant mice, rats, and rhesus monkeys can have adverse morphological or behavioral consequences for the offspring (Adams et al., 1990). We administered PHT to newborn mice and demonstrated neurotoxic effects on the developing mouse cerebellum at a plasma level of <25 μg/ml in the present study.
PHT is extensively (∼90%) bound to plasma proteins, mainly albumin (McNamara, 1996). The brain uptake of some compounds, such as phenobarbital and PHT, is lower than predicted from their lipid solubility as a result of binding to plasma proteins (Betz et al., 1994). The concentration of PHT in the CSF is equal to the unbound fraction in plasma (McNamara, 1996). It has been reported that a greater fraction remains unbound in the neonate, in patients with hypoalbuminemia, and in uremic patients (McNamara, 1996). In this study, it is highly important that the PHT level in the brain was significantly higher than that in the plasma. This may be related to the lower plasma binding capacity to PHT in the neonatal period (McNamara, 1996). The developing mouse cerebellum in the neonatal period is more vulnerable to the neurotoxic effects of PHT because of the higher brain concentration of PHT.
To prevent birth defects in offspring of women treated with antiepileptic drugs, the use of the lowest dosage that effectively controls seizures and a change from polypharmacy to monotherapy have been recommended before conception (Kaneko and Kondo, 1995). Extrapolation from our findings in mice to humans may be difficult because of possible differences between species and between gestational and neonatal exposure in susceptibility to PHT. Nevertheless, our present data would provide useful information on the management of pregnant epileptics taking PHT, and emphasize the importance of giving PHT at the lowest effective dose and monitoring maternal PHT plasma levels regularly in the evaluation of developmental neurotoxicity of PHT.
This research was supported by grants from Epilepsy Research Foundation (Japan) and the Science and Technology Agency of the Japanese Government. The technical assistance of Mr. H. Ishihara, Mr. N. Shimizu, and Ms. S. Okamura (Department of Anatomy, Hiroshima University School of Medicine) and Ms. A. Nobukiyo (Research Institute for Laboratory Animal Science, Hiroshima University School of Medicine) is gratefully acknowledged. We also appreciate Prof. M. Takano and Dr. Murakami (Institute of Pharmaceutical Sciences, Hiroshima University School of Medicine) for the determination of PHT concentrations in the plasma and brain. The authors wish to thank Prof. P. D. Andrew (Institute of Health Sciences, Hiroshima University School of Medicine) for helpful comments on the manuscript.
- 61982) Morphological development of the rat cerebellum and some of its mechanisms. Exp. Brain Res. 6 (Suppl.),8–46.(
- 131994) Blood-brain-cerebrospinal fluid barriers, in Basic Neurochemistry (Siegel G. J., Agranoff B. W., Albers R. W., and Molinoff P. B., eds), pp. 681–699. Raven Press, New York., , (
- 271991a) Role of thyroid hormones in mammalian brain development, in Developmental Neurobiology, 3rd edit. (Jacobson M., ed), pp. 294–296. Plenum Press, New York.(
- 281991b) Histogenesis and morphogenesis of cortical structures, in Developmental Neurobiology, 3rd edit. (Jacobson M., ed), pp. 401–451. Plenum Press, New York.(
- 291982) On major malformations and minor anomalies in the offspring of parents with epilepsy, in Epilepsy, Pregnancy and the Child (Janz D., Dam M., Richens A., Bossi L., Helge H., and Schmidt D., eds), pp. 211–222. Plenum Press, New York.(
- 301985) Comparative developmental pathology of malformations of the central nervous system, in Prevention of Physical and Mental Congenital Defects, Part A: The Scope of the Problem (Marois M., ed), pp. 143–156. Alan R. Liss, Inc., New York.(
- 441996) Hydantoins, in Goodman and Gilman’s Pharmacological Basis of Therapeutics, 9th edit. (Hardman J., Limbird L. E., Molinoff P. B., and Ruddon R. W., eds), pp. 468–471. Pergamon Press, New York.(
- 511996) Cerebellum, in Carpenter’s Human Neuroanatomy, 4th edit. (Parent A., ed), pp. 583–629. Williams and Wilkins, Philadelphia.(