Schizophrenia is a severe psychiatric disorder, affecting about one percent of the population worldwide. Perturbation of pre-natal development is known to contribute to the disease, although onset typically occurs between late adolescence and early adulthood. This led to formulation of the ‘two-hit’ hypothesis: the disruption of pre-natal brain development confers vulnerability (hit-1), but a second substantial impact during adolescence (hit-2) is required to trigger the disorder. 

 

Environmental factors may contribute to both early (e.g. maternal infection) and late hits (e.g. childhood adversity), but the risk of developing schizophrenia is determined to a significant extent by the constellation of genetic risk factors an individual carries. The identity of these risk variants is gradually being uncovered by genetic studies; looking at which genes the variants affect, we can start to uncover molecular and cellular processes disrupted in schizophrenia – the biology behind the two hits. 

 

In 2012, we published the first genetic evidence linking specific biology in mature brain cells (neurons) to schizophrenia. Neurons communicate with each other via specialised connections (synapses); modification of these connections (synaptic plasticity) is thought to be the basis of learning and memory. We found that rare mutations present in individuals with schizophrenia disrupted key elements of the molecular machinery regulating synaptic plasticity. This provided insight into hit-2, but the nature of hit-1 remained elusive. 

 

Our 2012 study analysed large mutations that typically span multiple genes, not all of which will be risk genes. Among the few variants we identified that disrupt single genes were two deletions in a gene called DLG2. The protein produced by this gene has long been studied as part of the synaptic machinery in mature neurons. Interestingly, DLG2 is also active in the pre-natal brain. In invertebrates, an evolutionarily related gene is known to contribute to early cell development, and we speculated that DLG2 may play a similar role during human brain development. Assuming DLG2 to be a genuine risk gene (two mutations aren’t enough to be sure), this raised the intriguing possibility that it might be involved in both early and late hits. 

 

Early stages of brain development, during which the birth of neurons (neurogenesis) occurs, can be modelled through the directed differentiation of embryonic stem cells in culture. In this process, stem cells are exposed to a sequence of chemical cues, mirroring the signals they would naturally receive in the developing brain. This causes them to transform into a proliferating pool of progenitor cells from which neurons are born. To investigate the role of DLG2 in the development of neurons, we created mutant embryonic stem cells in which the DLG2 gene was deleted. Mutant and normal cells were then differentiated to produce neurons, comparing the cultures at multiple timepoints to investigate the effects of DLG2 deletion. 

 

DLG2 loss had a profound effect, changing the activity of thousands of genes. The greatest changes coincided with the onset of neurogenesis; strikingly, the genes with decreased activity in mutant cells at this timepoint harboured many risk variants for schizophrenia. This indicated two things: DLG2 is required for the normal activation of biological pathways associated with neurogenesis, and these pathways are disrupted in schizophrenia. We had started to uncover the biology behind hit-1. 

 

Suspecting that DLG2 wasn’t the only gene involved in both early and late hits, we compared a set of mature neuronal genes previously linked to schizophrenia with the early developmental pathways we had identified. Genes common to both mature and developmental sets were enriched for schizophrenia risk, as were genes unique to the developmental pathways; those unique to the mature set were not. This suggests that hit-2 arises from molecular pathways contributing to early developmental deficits (hit-1) that remain active in post-natal life. 

 

Our data support a revised model of schizophrenia in which pathways guiding early neuron growth, migration and circuit formation are disrupted. Some of these pathways wind down as the brain matures, but those responsible for establishing synaptic connections (which leads to circuit formation) remain active into adulthood, where they contribute to learning-dependent changes (synaptic plasticity). Rather than two discrete hits, disease onset is instead due to a gradual accumulation of deficits in the structure and function of neural circuits, potentially exacerbated by external stresses. 

 

The pre-natal pathways we identified also harbour variants for other conditions: developmental delay, autism, ADHD, depression and bipolar disorder. How are these pathways disrupted in each disorder and why does this impact a different pattern of neural circuits? Many details (and surely some fundamental biology) remain to be uncovered.