Crespi & Badcock (C&B) suggest that two poorly understood neuropsychiatric syndromes, autism and schizophrenia, can be conceptualized as diametrical, opposed disorders. They suggest the root of the divergence is a disruption of maternally versus paternally imprinted genes, which serves as a switch point, after which the affected individual is shunted down either the schizophrenia or the autism tracks. C&B array a wide range of evidence to support this ambitious hypothesis. Hypotheses like these are intrinsically appealing – we all desperately wish the complexity of neurobiology could be so easily condensed!

What type of results, from what type of experiment, would allow us to confidently reject the hypothesis outlined by C&B? Hypotheses linking early life exposures (genetic and non-genetic) and later neuropsychiatric syndromes are notoriously difficult to test. This is particularly the case for syndromes such as schizophrenia where the disorder may not be apparent until the second or third decade of life (McGrath et al. Reference McGrath, Feron, Burne, Mackay-Sim and Eyles2003).

Although clinical research is clearly important, we suggest that animal models remain the only practical tool for unravelling the mechanisms behind early life disruptions that may lead to adult neuropsychiatric disorders. The developing human brain is not open to ready observation, and experimental manipulations of normal brain developmental are clearly neither ethical nor feasible. Rats and mice do not “get” schizophrenia or autism – so animal models will never recapitulate the full phenotype of disorders involving higher cognitive function. However, they provide an experimental platform that allows the researchers to focus on more substrate-pure neurobiological correlates of clinical syndromes (e.g., brain structure or animal behaviour) (Arguello & Gogos Reference Arguello and Gogos2006).

Animal studies allow the examination of both candidate genes and environmental exposures on brain development either in vitro or at the whole animal level. In such investigations outcomes as narrow as regional gene expression right up to social behaviour can be studied. Animal models in schizophrenia research have explored the impact of early life exposure to various candidate risk factors. For example, animal models in schizophrenia research have included postnatal lesioning of selected brain areas (Lipska et al. Reference Lipska, Swerdlow, Geyer, Jaskiw, Braff and Weinberger1995), prenatal exposure to specific viruses like influenza (Fatemi et al. Reference Fatemi, Cuadra, El Fakahany, Sidwell and Thuras2000), and prenatal nutritional deficiencies (Eyles et al. Reference Eyles, Brown, Mackay-Sim, McGrath and Feron2003; Kesby et al. Reference Kesby, Burne, McGrath and Eyles2006). Various animal models of autism are being similarly explored (Crawley Reference Crawley2007; Moy et al. Reference Moy, Nadler, Young, Perez, Holloway, Barbaro, Wilson, Threadgill, Lauder, Magnuson and Crawley2007).

To make their case, C&B have focused on divergence between schizophrenia and autism. We would like to draw attention to two interesting areas of convergence between schizophrenia and autism: namely, prenatal infection, and advanced paternal age. Early life exposure to infection has been associated with an increased risk of autism (Libbey et al. Reference Libbey, Sweeten, McMahon and Fujinami2005) and schizophrenia (Brown Reference Brown2006). Prenatal infection models in animals are proving informative for both autism and schizophrenia research (Fatemi et al. Reference Fatemi, Reutiman, Folsom and Sidwell2007; Hornig et al. Reference Hornig, Weissenbock, Horscroft and Lipkin1999). The evidence linking prenatal infection has led to rodent models of maternal immune activation. These models, which use non-infective viral (Poly I:C) or bacterial-like components (lipopolysaccharide; LPS) to trigger innate immune systems, have produced informative behavioural phenotypes (Meyer et al. Reference Meyer, Feldon, Schedlowski and Yee2006a; Reference Meyer, Nyffeler, Engler, Urwyler, Schedlowski, Knuesel, Yee and Feldon2006b).

Advanced paternal age has also been linked with an increased risk of both schizophrenia and autism. Malaspina et al. (Reference Malaspina, Harlap, Fennig, Heiman, Nahon, Feldman and Susser2001) reported that, compared with offspring of fathers younger than age 25 years, the relative risk of schizophrenia was 2.96 in offspring of men aged 50 years or more. Several other groups have replicated this association (El-Saadi et al. Reference El-Saadi, Pedersen, McNeil, Saha, Welham, O'Callaghan, Cantor-Graae, Chant, Mortensen and McGrath2004; Sipos et al. Reference Sipos, Rasmussen, Harrison, Tynelius, Lewis, Leon and Gunnell2004). It has also been shown that the offspring of men aged 40 or above have a nearly six-fold increase risk of autism spectrum disorders compared to the offspring of men younger than age 30 (Reichenberg et al. Reference Reichenberg, Gross, Weiser, Bresnahan, Silverman, Harlap, Rabinowitz, Shulman, Malaspina, Lubin, Knobler, Davidson and Susser2006). Advanced paternal age may be contributing to the transgenerational accumulation of copy error mutations (Crow Reference Crow2000; Keller & Miller Reference Keller and Miller2006). Alternately, epigenetic processes may be compromised in the sperm of older fathers. The resultant alterations in developmental gene structure and/or expression could both contribute to the increased risk of schizophrenia in the offspring of older fathers (Perrin et al. Reference Perrin, Brown and Malaspina2007). Once again, studying the impact of advance paternal age on genomic fidelity and structural and behavioural phenotypes is feasible in inbred animal strains.

Many different factors could contribute to the differential trajectory of brain development found in schizophrenia and autism – this is acknowledged by C&B in their introduction. In the light of this commentary, we would like to propose an alternative hypothesis. Advanced paternal age may lead to both point mutations (Crow Reference Crow2000) and small de novo chromosomal rearrangements (e.g., copy number variations) and DNA repeat expansions (Pearson et al. Reference Pearson, Nichol Edamura and Cleary2005; Perrin et al. Reference Perrin, Brown and Malaspina2007). There is increasing evidence that many neurological disorders, including autism, are associated with such small genomic rearrangements (Lee & Lupski Reference Lee and Lupski2006; Lupski Reference Lupski2007; Sebat et al. Reference Sebat, Lakshmi, Malhotra, Troge, Lese-Martin, Walsh, Yamrom, Yoon, Krasnitz, Kendall, Leotta, Pai, Zhang, Lee, Hicks, Spence, Lee, Puura, Lehtimäki, Ledbetter, Gregersen, Bregman, Sutcliffe, Jobanputra, Chung, Warburton, King, Skuse, Geschwind, Gilliam, Ye and Wigler2007). Indeed, Prader-Willi and Angelman syndromes, which are discussed in detail by C&B, are classic examples of small chromosomal rearrangements (Emanuel & Saitta Reference Emanuel and Saitta2007). Velocardiofacial syndrome, a deletion syndrome involving Chr 22q11.2, is strongly associated with both schizophrenia and autistic spectrum disorders (Vorstman et al. Reference Vorstman, Morcus, Duijff, Klaassen, Heineman-de Boer, Beemer, Swaab, Kahn and van Engeland2006). Certain types of genomic rearrangements are particularly likely to occur in regions of the genome enriched with short nucleotide repeats (e.g., trinucleotide repeats), expansions of which are already linked to a range of neurological disorders (Nithianantharajah & Hannan Reference Nithianantharajah and Hannan2007). The fact that disease-associated repeat expansion is a phenomenon that seems unique to humans (Pearson et al. Reference Pearson, Nichol Edamura and Cleary2005) may appeal to those interested in theories linking human neuropsychiatric disorder to recent evolutionary forces. We propose that the one general mechanism could contribute to susceptibility to both schizophrenia and autism (and probably a wide range of other disorders). However, variations in the location and “dose” of the rearrangements could then be translated to relative brain overgrowth versus undergrowth, as emergent properties of the complex systems governing brain development. These rearrangements may or may not include genomic segments maternally or paternally imprinted.

We hope that the hypothesis outlined by C&B will stimulate research that links schizophrenia and autism. We argue that animal models provide the most efficient platform to explore the hidden layers of complexity underlying normal and perturbed brain development. In particular, animal models may help fractionate which neurobiological mechanisms are shared between schizophrenia and autism, versus which are discrete for the two disorders.

ACKNOWLEDGMENTS

We acknowledge the support of the National Health and Medical Research Council of Australia.