Mapping the cell-surface proteome underlying synapse type-specific identity

Neural circuits in the brain are composed of distinct neuronal cell types connected in highly specific patterns. Precise connectivity between neuronal cell types in a neural circuit is instrumental in generating specific patterns of output activity, ultimately supporting motor and cognitive functions. Within a neural circuit, different neuronal cell types communicate through synaptic contacts that vary in their subcellular location, architecture, and functional properties. These features directly impact information processing in neural circuits. Elucidating the molecular mechanisms that underlie the emergence of specific types of synapses is thus important for understanding circuit function in physiology and disease.

The morphological and functional diversity of synapses is thought to be specified by their molecular composition. Multiple studies have described differential expression of synaptic proteins between synapse types, followed by functional characterization of their role in specifying defined synaptic properties. Differences in protein composition between synapse types are most apparent when comparing synapses of different neurotransmitter type, but further diversity in morphology, function, and protein composition can be found within these broad synapse classes. These studies have confirmed the potential for specific proteins to regulate defined synaptic properties, but have focused on a small subset of proteins. Other studies have unbiasedly identified synaptic proteins differentially distributed at excitatory versus inhibitory synapse types, but have mostly relied on in vitro and/or protein overexpression strategies. As a first step towards understanding how synaptic protein composition shapes the properties of specific synapse types, new methods are necessary to systematically dissect their molecular composition in vivo.

Here I developed an approach to unbiasedly profile the proteome of a specific hippocampal excitatory synapse type: the mossy fiber-CA3 synapse. This synapse, established between dentate gyrus (DG) granule cell (GC) axons, or mossy fibers (MFs), and CA3 pyramidal neurons, is unusually large, contains multiple release sites and is capable of powerfully depolarizing the postsynaptic CA3 pyramidal neuron. Filopodia extending from the main bouton form excitatory synapses on nearby interneurons (INs) that provide feedforward inhibition to CA3 neurons to keep the powerful MF-CA3 synapse in check. The large size of MF-CA3 synapses facilitated initial biochemical isolation from other synapses in the mouse hippocampus. Immunolabeling of MF-CA3 synapses using a known surface marker and subsequent fluorescent sorting further improved purification of this synapse type. Mass spectrometry (MS) analysis of isolated MF-CA3 synapses identified a large set of proteins enriched at this synapse compared to other, smaller hippocampal synapses.

Cell-surface proteins (CSPs), including transmembrane, membrane-anchored, and secreted proteins, play a key role in the formation of precise connectivity patterns and regulating input-specific properties. CSPs are expressed in cell type-specific combinations and form protein-protein interaction networks that regulate circuit assembly. Comprehensive characterization of the CSP composition of specific connections has remained challenging however, and the networks of cell-surface interactions specifying distinct synapses types, such as the MF-CA3 synapse, are not understood. Using MS to profile the proteome of isolated MF synaptosomes, I dissected the CSP composition of the MF-CA3 synapse. This effort uncovered a rich MF-CA3 synapse CSP repertoire that includes adhesion proteins, guidance cue receptors, extracellular matrix (ECM) proteins, and several CSPs of uncharacterized function. To map ligand-receptor relationships among the MF-CA3 synapse CSPs, I used automated interactome screening, revealing several novel interactions. Combining proteome and CSP interactome screening with extensive validation identified IgSF8 as an uncharacterized neuronal receptor strongly enriched in the MF pathway. Genetic analysis showed that IgSF8 is required in GCs to regulate MF-CA3 synapse architecture and function, and control feedforward inhibition of CA3 pyramidal neurons.

In conclusion, I developed a novel strategy to isolate a specific type of hippocampal excitatory synapse. Synapse type-specific proteomic dissection revealed a diverse set of proteins at MF-CA3 synapses including multiple uncharacterized synaptic CSPs. Interactome analysis further identified several novel modules of CSP interactions. Finally, the CSP IgSF8 was determined to be a novel regulator of MF-CA3 synapse connectivity and function. Together, the work in this thesis provides the first insight into the CSP landscape and cell-surface interactome of a specific excitatory synapse type. Hence, this work provides a first step towards understanding how synaptic protein composition shapes the properties of specific synapse types that are essential for neural circuit function.