Effects of prenatal infection on brain development and behavior: A review of findings from animal models

Abstract

Epidemiological studies with human populations indicate associations between maternal infection during pregnancy and increased risk in offspring for central nervous system (CNS) disorders including schizophrenia, autism and cerebral palsy. Since 2000, a large number of studies have used rodent models of systemic prenatal infection or prenatal immune activation to characterize changes in brain function and behavior caused by the prenatal insult. This review provides a comprehensive summary of these findings, and examines consistencies and trends across studies in an effort to provide a perspective on our current state of understanding from this body of work. Results from these animal modeling studies clearly indicate that prenatal immune activation can cause both acute and lasting changes in behavior and CNS structure and function in offspring. Across laboratories, studies vary with respect to the type, dose and timing of immunogen administration during gestation, species used, postnatal age examined and specific outcome measure quantified. This makes comparison across studies and assessment of replicability difficult. With regard to mechanisms, evidence for roles for several acute mediators of effects of prenatal immune activation has emerged, including circulating interleukin-6, increased placental cytokines and oxidative stress in the fetal brain. However, information required to describe the complete mechanistic pathway responsible for acute effects of prenatal immune activation on fetal brain is lacking, and no studies have yet addressed the issue of how acute prenatal exposure to an immunogen is transduced into a long-term CNS change in the postnatal animal. Directions for further research are discussed.

Introduction

Epidemiological studies in humans have provided substantial evidence that prenatal infection is associated with an increased risk for the development of several psychiatric and neurologic disorders, most prominently schizophrenia, autism and cerebral palsy. While these associations provide rationale for suggesting that prenatal infection might contribute to the cause of these disorders, they do not prove causation. Animal modeling provides an opportune approach to ask if prenatal infection can actually cause transient or lasting changes in CNS function, and what the mechanisms for this might be. Since 2000, there has been a veritable explosion of studies from various laboratories that have used in vivo animal modeling, mainly in rodents, to characterize effects of systemic prenatal infection on brain development and behavior. The aim of this review is to provide a comprehensive overview of these findings from animal models of prenatal infection, in an effort to tease out consistencies, trends and deficiencies in the literature and to suggest possibilities for future directions.

In order to understand directions these animal modeling studies have taken, I will first provide a short introduction to the epidemiological literature associating prenatal infections with increased risk for CNS disorders in humans, although it is not the purpose of this paper to review this area in depth. An extensive literature has described that increased risk for schizophrenia is associated with a history of prenatal infection with a wide variety of different infectious agents (reviewed by Brown and Derkits, 2010). First reported by Mednick et al. (1988) in the mid-1980s, numerous groups (although not all) have replicated the finding that maternal influenza during pregnancy is associated with an increased incidence of schizophrenia (Barr et al., 1990, O’Callaghan et al., 1991, Sham et al., 1992, Adams et al., 1993, Mednick et al., 1994, Takei et al., 1996, Battle et al., 1999, Munk-Jorgensen and Ewald, 2001, Limosin et al., 2003, Brown et al., 2004, Ebert and Kotler, 2005, Byrne et al., 2007, Selten et al., 2009). Increased incidence of schizophrenia has also been associated with other viral infections (measles, rubella, varicella-zoster, polio, herpes simplex virus type 2), with bacterial infections (pneumonia and other respiratory infections, pyelonephritis, diverse bacterial infections) during pregnancy, with maternal infection with the toxoplasmosis parasite and with maternal genital/reproductive infections arising from various organisms (Torrey et al., 1988, O’Callaghan et al., 1994, Suvisaari et al., 1999, Brown et al., 2000, Brown et al., 2005, Brown and Susser, 2002, Babulas et al., 2006, Mortensen et al., 2007, Buka et al., 2008, Clarke et al., 2009, Sørensen et al., 2009). While most studies have relied on maternal recall and retrospective or prospective hospital records to assess maternal infections, recent studies assessing maternal infection via analysis of archived maternal serum have confirmed the association between prenatal viral infection and increased schizophrenia risk (Brown and Susser, 2002, Brown et al., 2004, Brown et al., 2005, Mortensen et al., 2007, Buka et al., 2008, Xiao et al., 2009). In more detailed recent studies examining symptom subsets in psychosis, Zammit et al. (2009) have reported a significant association between maternal infection during pregnancy and the presence of psychosis-like symptoms in adolescents at 12 years of age. Ellman et al. (2009) have reported that persons with schizophrenia and affective psychosis showed greater deficits in verbal IQ and other aspects of cognitive functioning prior to the onset of psychosis if they had been exposed prenatally to serologically confirmed influenza B. In a similar vein, Brown et al. (2009) documented that patients with schizophrenia who were exposed to serologically confirmed influenza or toxoplasmosis in utero showed greater deficits in cognitive tests of set-shifting ability, compared to patients not exposed to prenatal infection. As an additional aspect, Clarke et al. (2009) have recently reported that maternal pyelonephritis during pregnancy is associated with increased risk for schizophrenia only in persons with a positive family history of schizophrenia, suggesting a gene x environment interaction as the mechanism for increased schizophrenia risk.

With respect to timing of the infection, numerous studies have identified the 2nd trimester of pregnancy as the critical period for exposure to influenza and other viral infections, leading to increased schizophrenia (Torrey et al., 1988, O’Callaghan et al., 1994, Suvisaari et al., 1999, Brown et al., 2000, Ebert and Kotler, 2005). However, the most pronounced association of rubella infection with schizophrenia spectrum disorders was found with 1st trimester exposure (Brown and Susser, 2002). Moreover, a recent study confirming presence of influenza antibodies in archived maternal serum has described increased schizophrenia risk with exposure during the 1st third to half of pregnancy (Brown et al., 2004), while another recent study has associated increased schizophrenia with bacterial infection during the 1st trimester (Sørensen et al., 2009). Thus it appears that animal modeling of both viral and bacterial infection during the equivalent of the 1st and 2nd trimesters of human pregnancy may be relevant to the pathophysiology of schizophrenia.

Evidence implicating a role for prenatal infection in the etiology of autism is much less extensive. Two larger studies have reported increased incidence of autism following congenital rubella or maternal viral infection during pregnancy (Chess, 1977, Wilkerson et al., 2002), however, the trimester of increased vulnerability was not identified. Beyond this, several case reports have linked autism to prenatal exposure to various viruses, such as cytomegalovirus (Ciaranello and Ciaranello, 1995, Libbey et al., 2005). Increased risk for cerebral palsy has most often been associated with chorioamnionitis (i.e. intra-uterine infection), although there is some evidence that extra-uterine maternal infections during pregnancy may also play a role (Murphy et al., 1995, Clark et al., 2008). Meta-analysis has indicated a significant association between chorioamnionitis and cerebral palsy for both preterm and term infants (Wu and Colford, 2000, Wu, 2002).