Dynamic Pathway Crosstalk Networks Reveal Distinct Core Mechanism Underlying Autism And Schizophrenia

Autism Spectrum Disorders (ASD) and schizophrenia are common neuropsychiatric disorders with high prevalence in population, categorized into two disease types according to the current diagnosis criteria. ASD usually occurs in babyhood with three core symptoms later, such as impaired sociality, disabled communication and stereotyped behavior, whereas schizophrenia generally emerges in adolescence, mainly displaying delusion or fantasy. Although both disoders have explicit descriptive characteristics, these characteristics in reality often mix into one individual. And moreover, ASD and schizophrenia have higher probability of clustering in one family, implicating their link at the molecular level. Therefore, distinguishing the two diseases at the molecular level is significantly meaningful for capturing the specific nature of the illnesses and further developing specific treatment strategies.Here, we proposed a systematic framework that integrated biological pathways and unbiased gene expression information to infer the dynamic changes of crosstalks among different biological processes in the state of illnesses by a two-step randomization test. The significantly altered crosstalks collectively form the disturbance network, which probably represents the pathological disruption pattern, and the hub nodes in the network may reflect the core mechanism of the illnesses. Following this framework, we constructed the disturbance networks for autism and schizophrenia. These networks are significantly enriched with neurodevelopmental and neurophysiological pathways as well as those pathways frequently implicated in these disorders. In contrast, asthma disturbance network constructed as a negative control in the current context only contains sparse neural pathways. The correlation between the two psychotic disturbance networks provides the interpretation for the strong symptomatic similarity between autism and schizophrenia at the pathway level. However, our result highlights synaptic vesicle cycle as the key pathway in schizophrenia for its high node degree in the disturbance network, whereas cell cycle dysregulation is suggested as the dominant element in the pathophysiology of autism and probably interprets brain overgrowth in the autism cases. There exists comorbidity between schizophrenia and diabetes, however, the mechanism of the link is still obscure. Another hub pathway in the schizophrenia disturbance network, integration of energy metabolism, may establish the connection between the two disorders.Overall, our study proposed the concept of core mechanism and developed a framework to explore it for two neuropsychiatric disorders. The resulting conclusions provided rational explanations for the disorder features and clues for medication strategies.