Dysfunctional synaptic competition at dendritic spines in Fragile X syndrome

Abstract

Dendritic spines are highly dynamic structures whose morphology and lifespan are modified in response to synaptic efficacy changes. Structural modifications following activity support the long-term encoding of information and could allow for the remodeling of neural circuits. Long-term depression (LTD) is a key mechanism for synaptic weight regulation, yet its structural correlates -- particularly for long-lasting, protein synthesis dependent forms -- remain poorly understood. Furthermore, in humans, this type of plasticity is often disrupted in neurodevelopmental disorders, correlating with cognitive dysfunction and structural abnormalities. Fragile X Syndrome (FXS) is the most common inherited form of intellectual disability and is characterized by excessive metabotropic receptor-mediated synaptic depression, excessive protein synthesis, and spine abnormalities. Here, we investigate the relationship between long lasting synaptic depression and structural plasticity, as well as the role of protein availability in determining how many spines can simultaneously undergo structural changes during LTD in both healthy and FXS mutant neurons. Using high resolution optical methods, we developed and tested a method for inducing metabotropic glutamate receptor (mGluR)-dependent depression at single spines via glutamate uncaging in mouse hippocampal neurons. We found that this form of activity leads to robust spine shrinkage, which requires new protein synthesis. However, when we induced this depression at multiple spines, they competed for structural changes and only one spine shrank. We hypothesized that this was due to limited resources, in the form of newly made proteins, and therefore, we decided to test if spine competition would be altered in the mouse model of FXS, where protein levels are abnormally elevated. Indeed, we found that competition was absent in FXS mutant neurons, and all of the stimulated spines underwent shrinkage following LTD induction. Importantly, we found that single spine structural plasticity in FXS was expressed to the same degree as in WT controls. Taken together, these findings suggest that the hallmark phenotype of excess mGluR LTD in FXS may result from a greater number of inputs undergoing synaptic depression, rather than excessive LTD at individual synapses. By probing plasticity at the level of individual inputs, our findings highlight the importance of evaluating activity across groups of synapses, in order to uncover plasticity interactions that are critical for learning. Understanding how these mechanisms are disrupted in neurodevelopmental disorders such as FXS can inform the development of effective therapeutic strategies.