VPA-induced apoptosis and behavioral deficits in neonatal mice

Abstract

Sodium valproate (VPA) administered to neonatal mice causes cognitive and motor deficits similar to those observed in humans with autism. In an effort to further evaluate similarities between early VPA exposure and autism, the present study examined treated mice for deficits in social behavior and neuronal damage. BALB/c mice injected on P14 with 400 mg/kg VPA engaged in fewer social interactions (including ano-genital sniffs, allogrooming, and crawl-under/over behaviors) than control mice. Treated mice also exhibited reduced motor activity in a social context but were not significantly different from controls when motor activity was assessed in non-social settings. A second set of BALB/c mice were treated with VPA on P14 and sacrificed at different times thereafter for histopathological analysis. At time-points 12 and 24 h following VPA, treated mice had up to a 30-fold increase in the number of TUNEL-positive cells in the external granule cell layer of the cerebellum and a 10-fold increase in TUNEL-positive cells in the dentate gyrus of the hippocampus. These observations may provide a histopathological correlate for the social deficits observed following post-natal VPA exposure and supports the use of early VPA administration as an animal model for the study of autism.

Introduction

Autism is a complex developmental disorder with core symptoms that include impaired social interactions, deficits in verbal and non-verbal communication, and the appearance of stereotypic and sometimes self-injurious behaviors. Post-mortem histological analysis of the brains of individuals with autism revealed a reduction in the number of cerebellar Purkinje cells (Bauman and Kemper, 1985, Bauman and Kemper, 1998, Bauman and Kemper, 2005, Ingram et al., 2000). Bauman and Kemper (1985) also found a decrease in cerebellar granule cell density in two of their case studies. In addition to these observations that cerebellar cell numbers are reduced, Fatemi (2002) showed that the size of cerebellar Purkinje cells is also reduced. Taken together, these studies have been interpreted to indicate that permanent alterations in cerebellar Purkinje and granule cells may be neuroanatomical markers found in adult individuals with autism. However, in a study by Bailey et al. (1998), there appeared to be no localized pathology underlying autism, including no change in cerebellar Purkinje cell number or size. Given that these studies were post-mortem, they have several inherent limitations that may account for this discrepancy including a lack of controls for institutionalization, medication history, and co-morbid diagnosis of mental retardation. Studies using animal models of autism can control these types of variables and, perhaps, link anatomical and behavioral deficits. One important animal model of autism involves early administration of sodium valproate (VPA) to rodents.

Clinical studies of children exposed to VPA in utero have characterized a fetal valproate syndrome with symptoms similar to autism, including deficits in language and communication, the appearance of stereotypic behavior, hyperexcitability and global delays in behavioral development (Ardinger et al., 1998, Koch et al., 1996, Mawer et al., 2002, Moore et al., 2000, Williams et al., 2001). Clinical case studies of infants with fetal valproate syndrome also report physical abnormalities including low myelomeningocele lesion, minor abnormalities of the face and ear, and microcephaly (Ardinger et al., 1998, Moore et al., 2000). These clinical similarities have led Rodier et al. (1997) to propose that prenatal VPA exposure in rodents might serve as an animal model of autism.

Prenatal VPA treatment has been shown to affect behavioral development with treated rats exhibiting repetitive/stereotypic-like behavior, decreased exploratory activity, and a decreased number of social behaviors (Voorhees, 1986, Schneider and Przewtocki, 2004). Likewise, VPA exposure in rodents at the time of neural tube closure has been shown to reduce the number of cerebellar Purkinje cells (Ingram et al., 2000, Sobaniec-Lotowweska, 2001). These observations lend credibility to the suggestion that early exposure to VPA in rodents leads to behavioral and neurological deficits that resemble autism.

Previously, we demonstrated that both pre- and early postnatal exposure to VPA lead to neurodevelopmental deficits that are similar to the motor and cognitive deficits seen in humans diagnosed with autism (Wagner et al., 2006). In that study, the neurobehavioral deficits induced by early VPA exposure were classified as retardations (i.e., a behavioral skill that matured later in the treated mice as compared to controls), regressions (i.e., a behavioral skill that matured on schedule but was then lost subsequent to VPA exposure), or intrusions (i.e., the VPA exposure induced behaviors aberrant in intensity or frequency that overshadowed the normal maturation of behavioral skills). The behavioral tests assessed broad categories of motor and cognitive performance but, in this first study, the maturation of social skills following VPA exposure was not addressed. In addition, there was no assessment of the VPA-induced neural damage.

Accordingly, the objectives of the present study were to advance the VPA model of autism in mice by evaluating social behaviors and neuronal apoptosis following exposure on postnatal day 14 (P14). Wagner et al. (2006) demonstrated that critical cerebellar-mediated behaviors first appear on P14. Furthermore, a single 400 mg/kg postnatal exposure on P14 was effective in producing VPA-induced motor and cognitive deficits that mimicked autistic regression. Therefore, the current study used an acute dose of VPA at P14 in order to determine if a single exposure would also result in social deficits in the developing mice. The second objective was to determine if the P14 VPA treatment affected the integrity of the cerebellum and/or hippocampus, two regions associated with autism in humans. Postnatal day 14 roughly corresponds to the third trimester in human development (Rice and Barone, 2000), a period when hippocampal and cerebellar granule cells are undergoing migration and differentiation (Bachevalier and Beauregard, 1993, Rice and Barone, 2000, Voorhees, 1986). Accordingly, we examined these regions using the TUNEL stain for apoptosis.