Glutamatergic Activity Is Sufficient to Drive Striatal Spine Formation and Long-Term Plasticity Induction in Cortico-Striatal Co-Cultures

Background: Altered dendritic morphology and synaptic dysfunction are implicated in the pathophysiology of numerous neurological disorders. Clinical findings and research in genetic mouse models of autism spectrum disorder (ASD) are converging on alterations to striatal structure and function, suggesting that synaptic development and plasticity of striatal neurons may play an important role in ASD symptomatology and etiology. While the mechanisms regulating synaptic plasticity and dendritic spine development are well described at glutamatergic synapses in excitatory cortical and hippocampal neurons, less is known of how these principles translate to GABAergic striatal spiny projection neurons (SPNs), which receive glutamatergic input from cortical and thalamic neurons and form the sole output pathways to the basal ganglia.

Objectives: We sought to examine how manipulating glutamatergic activity at the cortico-striatal synapse affects spine formation and functional plasticity in SPNs. We used cortico-striatal co-cultures from wild-type male and female embryonic mice to examine the functional and structural consequences of altering the excitatory input from cortical neurons on SPNs. The aim was to develop rapid pharmacological assays that can be used in genetic mouse models, in order to study the vulnerability of striatal neurons neurological disorders such as ASD.

Methods: We examined the effects of chronic and acute action-potential silencing (by applying tetrodotoxin to the cultures), as well as the requirement for residual glutamatergic activity through AMPA and NMDA receptors, on presynaptic protein Synapsin-1 and spine/filopodia density in SPNs. To assess rapid activity-dependent plasticity, we used a 3-min glycine application in magnesium-free solution, which results in NMDAR-dependent LTP in other cell types, and measured both electrophysiological and structural outcomes. All experiments were repeated in at least 3 separate cultures, with experimenter blind to treatment condition during analysis.

Results: Chronic and medium-term silencing induced dendritic spine loss, increased filopodia density, and altered Synapsin-1 clusters in SPNs. The application of AMPA- and NMDA-type glutamate receptor antagonists prevented spine but not filopodia alterations, suggesting these are distinct processes. Glycine application was sufficient to increase spine and AMPAR subunit GluA1 cluster density and co-localization, and increased presynaptic glutamatergic release as indicated by whole-cell voltage clamp recordings.

Conclusions: These findings suggest that silencing-induced spine loss, but not filopodial proliferation, is an active process requiring residual glutamate receptor activity in SPNs. Furthermore, NMDA receptor activation in the absence of magnesium is sufficient to drive LTP-like structural and functional plasticity changes in co-cultured SPNs. Together, the data elucidate the importance of glutamate in shaping striatal structure and function even in the absence of neuromodulators such as dopamine, and how the cortico-striatal co-culture system may be useful for studying the role of striatal synapse changes in ASD.