Synapse Development in Autism

This article outlines some of our current research into the mechanisms underlying autism that we are exploring in our lab at the Centre for Developmental Neurobiology, King’s College London. In the lab, we are trying to understand what is different about the development of the autistic brain at the level of brain cells, or ‘neurons’, and the connections between them – ‘synapses’.

Genetic studies of autistic individuals have shown that a large group of affected ‘autism risk’ genes code for proteins that are known to be involved in synaptic structure, function or formation. This suggests that alterations in synapse development may play a key role in the development of autism, but we have very little idea of what exactly these alterations are, and how they vary between different individuals. We study neuronal and synapse development in mouse neurons which have been modified to have similar genetic changes to those seen in human individuals with very severe autism (and with associated intellectual disability, seizures, etc.). In our lab we have focused our research on a specific region of the brain: the prefrontal cortex. This region is known to be important for many aspects of social behaviour as well as executive cognitive functioning, and has been consistently implicated as affected in autism.

Electrodes can be used to record the activity of individual neurons and the tiny electrical signals that reflect synaptic inputs from other neurons onto a recorded neuron. In a recent study, we targeted the deeper layers of the prefrontal cortex, where the output neurons that project to other parts of the brain are found. When we recorded from neurons that were missing a copy of a gene called ‘Chd8’ – one of the highest confidence risk genes for autism – we found that there were changes to both excitatory and inhibitory synapses on to these output neurons (Figure 1), shifting the balance between excitation and inhibition. This shift was most pronounced at a particular stage of development, roughly corresponding to childhood. These changes to the synaptic inputs meant that the neurons didn’t fire as often.

Of course, the brain is very good at keeping excitation and inhibition in balance, as too much excitation might lead to seizures, while too little could affect normal functioning. Synapses are highly plastic, meaning that their strength can be adjusted depending on the circumstances. The brain uses a process called homeostatic plasticity to tweak synaptic strength (among other things) and so maintain the balance between excitation and inhibition in a ‘safe’ range, allowing it to compensate for changes in global neuronal activity levels. Interestingly, we also found that the normal homeostatic plasticity mechanisms that keep synapses in balance were altered in neurons missing a copy of Chd8, giving us a clue as to how this shift in synaptic balance may have happened (Figure 2, Ellingford et al., 2021, link: here). We are now trying to understand how the genetic change – in this case, missing a copy of Chd8 – actually links to these cellular changes, and how this might affect behaviour.

Figure 1. Examples of excitatory synapses (blue) and inhibitory synapses (red) along a neuronal dendrite (green) in wild type (WT) mice (top) and mice lacking a copy of the gene Chd8 (bottom) at postnatal day 14.

Figure 2 . Summary diagram comparing wild type (WT) control neurons (left panel, light green excitatory neurons and light red inhibitory neurons) with Chd8+/-neurons (right panel, dark green excitatory neurons and dark red inhibitory neurons). Reduced Chd8 expression leads to reduced excitation:inhibition balance through decreased functional excitatory transmission (symbolized by smaller excitatory synapses) and increased numbers of inhibitory synapses onto basal dendrites, overall resulting in diminished neuronal output (dashed blue arrow). Altered homeostatic responses to drugs that change activity (TTX – tetrodotoxin, blocks action potential firing; APV blocks NMDA receptors, GZ- gabazine, blocks GABAA receptors) are seen in Chd8+/- neurons (shaded areas on right).

We are also able to test whether there are similar synaptic changes when different autism risk genes are deleted in mice neurons. We are beginning to discover that while there are some similar changes, there are also differences in terms of how genetic alterations affect cellular and synaptic function. This could be important, as we try to develop more personalised approaches to neurodevelopmental disorders, and realise that autism in particular is different in every individual. Finally, we are very interested in whether these changes are sex-specific. Autism is more common in males than females, but why this is the case is poorly understood. Unfortunately, until recently the majority of research was carried out in males, meaning that autism in females was even less well characterised and understood. This is now changing, and we have found that even very early on in development, there are strong sex differences in how risk gene mutations affect synapse development.

References :

Ellingford RA, Panasiuk MJ, Rabesahala de Meritens E, Shaunak R, Naybour L, Browne L, Basson MA, Andreae LC. Molecular Psychiatry 2021; doi: 10.1038/s41380-021-01070-9 [find here]