The paper by Meyer and colleagues is of great interest and adds to a growing body of preclinical and epidemiologic literature implicating exposure to in utero infection as a risk factor for schizophrenia. One especially important feature of this paper is that poly I:C was administered during the equivalent of early to mid-gestation, the gestational period during which our group observed an association between serologically documented influenza exposure and risk of schizophrenia in an epidemiologic study. Furthermore, these investigators have examined both dopaminergic and glutamatergic effects of in utero infection. The differential effects of prenatal poly I:C on these two neurochemical systems during pre-and post-adolescence is intriguing. Hence, this work complements previous studies by several investigators in the field, including the Patterson, Zuckerman, and Fatemi labs, in demonstrating behavioral, pharmacologic, and neuropathological abnormalities analogous to those seen in schizophrenia following prenatal immune activation, and further supports the biological plausibility of early to mid-gestational prenatal infection in the etiopathogenesis of schizophrenia.

View all comments by Alan Brown

  I recommend this paper

The prenatal infection model, whereby pregnant mice are exposed at various gestational times to a synthetic RNA (poly[I:C]) to induce immune activation stands as one of the best developmental animal models for schizophrenia. Firstly, the model is based in the epidemiology of the disease, with good evidence that maternal infection is a valid non-genetic risk factor for this disease. The model causes hyper-responsiveness to psychotomimetic agents, consistent with abnormal dopamine/glutamate signaling, as well as inducing deficits in sensory motor gating and attention. Many of these changes are reversed via the use of antipsychotic drugs. Therefore, this animal model is rare in that it possesses relatively good construct, face, and predictive validities.

In this study, Meyer’s group has varied the prenatal exposure window by making it much earlier—prenatal day (PND) 9—a time when there is perhaps less overt cellular differentiation in the brain compared to previous work where the exposure was at PND 15. Not only this, the authors go on to consider the effect of the post-utero environment with a cross-fostering experiment in using poly(I:C) exposed dams.

For me the most dramatic finding from this latest study is that postnatal infection is sufficient for certain phenotypic effects in the non-infected offspring, e.g., amphetamine-induced hyperlocomotion. This indicates that the post-utero environment is also part of the critical window for infection/inflammation modulation of DA signaling in the developing brain.

The data showing increased striatal tyrosine hydroxylase immunoreactivity in PND 9 poly(I:C) exposed males raises interesting questions about sex differences in the metabolism of DA. In limbic areas, the effect is magnified in the shell rather than the core of the nucleus accumbens. Incidentally, this study also reveals evidence that Glu R1 sites are also reduced in the shell. This has implications for microdialysis studies that claim certain differences in DA/Glu release from these nearby subregions (Pacchioni et al., 2007; Corda et al., 2006). The tyrosine hydroxylase immunohistochemical data prompt the conclusion that either DA innervation of these regions or the ability to synthesize DA in these regions has been altered by this early prenatal exposure. Alternately, the enhanced glutamate signaling that would result from existing glutamate binding at sites other than NMDA sites (when blocked by MK-801) may contribute to the hyperlocomotion phenotype. Presumably, selective dialysis studies in these exact regions are planned to address this critical issue.

Finally, the data describing a qualitative reduction in DA1 and DA2 receptors in the poly(I:C)-exposed males would tend to suggest a reduction in both prefrontal DA innervation and postsynaptic targets. This needs to be confirmed by quantitative binding studies where binding site densities for these two receptors could be determined. This study suggests a selective vulnerability of male rather than female neonates to infection, consistent with a preponderance of schizophrenia in male patients.

One wonders about the direct mechanisms in the brain linking this prenatal exposure and the adult behaviors in this animal model. For instance, it would be of great use to know whether prenatal poly(I:C) exposure had long-term effects on the production of certain cytokines in the adult offspring. Certainly the ability of cytokines such as IL-2 to promote DA release and induce a host of hyperlocomotor behaviors (Petitto et al., 1997) in exposed animals could be one candidate mechanism.

View all comments by Darryl Eyles