Recent years have seen an impressive amount of research devoted to the developing of therapies to treat autism spectrum disorder (ASD). This work has been largely based on rodent models, employing a multitude of genetic and environmental manipulations. Unfortunately, the task of identifying suitable treatments for ASD is extremely challenging, due to a variety of problems specific to the research in this field. Here, we first discuss these problems, including (I) the presence of a large variety of rodent models (often without universal consensus on their validity), (II) the difficulties in choosing the most appropriate behavioural markers to assess the efficacy of possible treatments, (III) the limited knowledge we still have of the neurobiological bases of ASD pathology and of its aetiology, and (IV) the complexity of ASD itself, including a highly heterogeneous group of disorders sometimes with markedly different symptoms (therefore unlikely to be treated with the same approaches). Second, we give a critical overview of the most relevant advances in designing treatments for ASD, focusing on the most commonly used animal model, the laboratory mouse. We include pharmacological and non-pharmacological approaches, underlining their specific advantages, but also their current limitations especially in relation to the problems discussed before. Finally, we highlight the theoretical (e.g. the combination of multiple rather than single treatments) and methodological (e.g. use of single-gene mouse models) approaches that seem more promising to us, suggesting various strategies that can be adopted to simplify the complex field of research on treatments for ASD.
Keywords: Autism; Behavioural therapies; Developmental disorders; Gene–environment interactions; Mouse behaviour; Neuronal hyper-excitability; Social interactions.