An investigation on the function of a rare Homer1 mutation in synaptogenesis

From its infancy to the present day one of the primary goals of neuroscience has been to understand how the neural circuits which govern the physiological processes associated with the central nervous system (CNS) are developed, maintained, and modified over time. For the CNS to appropriately perform the myriad functions essential to life which it performs, it must be ‘wired up’ in such a way that it achieves globally integrated and reproducibly accurate connections. Many neurological disorders, such as epilepsy and the autism spectrum disorders, are thought to be conditions of altered neuronal connectivity. There is an increasing recognition of the importance of rare, de novo variants as both contributors to disease risk and as targets for investigation of the disrupted cellular mechanisms that lead to neuronal connectivity. Genomic studies have identified mutations in scaffolding proteins – proteins responsible for binding and linking a variety of epitopes structurally and functionally - as key contributors underpinning these disorders. These proteins offer us valuable targets to interrogate the synaptic and developmental changes that contribute to the altered neuronal connectivity observed in these disorders.
This thesis investigated a recently described rare de novo variant (Homer1b/c^R297W) in the postsynaptic scaffolding protein, Homer1b/c. This variant features an arginine to tryptophan substitution mutation at a position within the protein’s C-terminus coiled-coil (CC) domain, a region necessary for the tetramerisation of the protein. It was hypothesised that this substitution would impair tetramerisation of the protein and consequently alter many of the functions which Homer1b/c is known to be involved in.
Homer1b/c has previously been shown to regulate aspects of axon guidance, the process by which the growth cones at the tips of extending axons are guided to form functional connections during development. Expression of Homer1b/c^R297W in DRG sensory neuron growth cones perturbed their turning responses to the Ca²⁺ -dependent guidance cue, brain derived neurotrophic factor, converting them from attractive to repulsive responses. Knockdown of Homer1b/c also converted these turning responses, an effect that could be rescued by expression of Homer1b/c^WT but not Homer1b/c^R297W. Growth cone turning to brain derived neurotrophic factor requires signal transduction involving store operated Ca²⁺ entry (SOCE), a process of secondary Ca²⁺ influx across the plasma membrane following the emptying of internal Ca²⁺ stores. Homer1b/c^R297W expression was shown to significantly blunt the amplitude of SOCE in both live DRG sensory neuron growth cones and hippocampal neuron soma, mimicking the effect of Homer1b/c knocked down with short-interfering RNA.
Homer1b/c is known to exert a powerful influence over the stabilisation and maturation of developing dendritic spines during synaptogenesis. Homer1b/c^R297W-expressing hippocampal neurons displayed consistently reduced dendritic spine densities at 4 DIV, 8 DIV and 12 DIV, and the morphology of these spines was altered to a more immature, filopodia-like phenotype as compared to Homer1b/c^WT controls. At 16 DIV, spines containing Homer1b/c^R297W presented significantly decreased levels of ER infiltration and significantly decreased expression of mGluR5, a canonical ligand of Homer1b/c. The linking of group-I mGluRs to the downstream Ca²⁺ signalling machinery is one of Homer1b/c’s best described roles, and this relationship is known to be perturbed in conditions such as fragile X. In the dendrites of Homer1b/cR^297W, cytosolic Ca²⁺ increases elicited by activation of group-I mGluRs were significantly blunted as compared to Homer1b/c^WT controls.
Emerging evidence is making increasingly clear that the clustering of receptors and their signalling effectors into nanodomains at the spine contributes strongly to the efficacy of their function. Single molecule super-resolution microscopy (SRM) was used to investigate the ability of Homer1b/c^R297W to bind and cluster mGluR5, IP3R and STIM2 into distinct nanodomains within hippocampal neuron dendritic spines heads. Using segmentation analysis based on Voronoi tessellation we demonstrate that Homer1b/c^R297W containing spines cluster each of these three epitopes into domains containing less molecules at decreased densities. SRM was further extended with DRG sensory neuron growth cones to investigate and describe the coupling of both Homer1b/c^WT and Homer1b/c^R297W with individual Homer1b/c ligands at the single molecule scale. We find that Homer1b/c^R297W is significantly uncoupled from mGluR5 and STIM1 in growth cones as compared to Homer1b/c^WT, indicating that its binding of these two ligands is altered by the CC mutation within Homer1b/c^R297W. The work presented here presents compelling evidence that the Homer1b/c^R297W variant fails to adequately perform key Homer1b/c functions relating to neuronal Ca2+ signalling and the scaffolding of key synaptic proteins. The mutation is hypothesised to disrupt the homomeric binding of Homer1b/c into homo- and heterotetramers, necessary components for the formation of higher order protein scaffolding complexes and Homer1b/c-specific functions. We propose that the impaired scaffolding of key proteins caused by the expression of Homer1b/c^R297W perturbs Ca²⁺ regulation, resulting in axon guidance defects and abnormal dendritic spine development, both of which have broad implications for the formation and maintenance of neural circuits.