Super-Resolution Characterization of Synaptic Neurexin-1 and Development of Highly Multiplexed RNA Imaging in the Brain

Abstract

Imaging techniques, applied at different length scales, can shed light on the complex, interconnected molecular and cellular networks that comprise the brain – how are different classes or types of neurons connected, and why are they connected that way? Variations in synaptic connections reflect not only in the type of neurons joined by a chemical synapse, but also the properties of synaptic release and plasticity. One of the most attractive candidates for explaining the high degree of variability observed in synaptic junctions is the trans-synaptic interaction between neurexins and their ligands, owing to the high degree of diversity possible due to the combinatorial nature of their alternative splicing and pairing. Understanding the synaptic localization and organization of neurexins can help elucidate the mechanism by which they function. A spatially-resolved expression profile on the transcriptomic scale of intact brain tissue can then be used to define transcriptionally distinct neuron types within a neuronal network, which can be correlated with molecular information from their synapses to examine the relationship between neurexins and synaptic identity. The first part of this thesis reports our efforts to probe the synaptic nanostructure and dynamics of neurexin-1 using multicolor 3D STORM. We found that synaptic neurexin-1 was organized into nanoclusters in a subset of Homer1(+) excitatory synapses, increasing in fraction over the course of development in cultured neurons and tissue from the mouse hypothalamus, up to ~40%. These nanoclusters also increased in size and neurexin content as the neurons mature. Furthermore, we found that the presence of synaptic neurexin-1 was correlated with an increase in markers of synaptic activity and strength. Neurexin-1 nanoclusters were dynamically regulated via ectodomain cleavage by ADAM10, and chemical inhibition of ADAM10 activity resulted in a doubling of the fraction of neurexin-1(+) synapses in cultured neurons. This cleavage is likely to be physiologically relevant, as neurexin-1 ectodomain fragments accounted for ~4-6% of total neurexin-1 levels in the mouse brain, varying across different brain regions. The remainder of this thesis describes our improvements to MERFISH towards quantifying spatially-resolved transcriptomic-scale expression profiling in intact brain tissue. By switching our method of signal removal between hybridization rounds from photobleaching to chemical reduction, restricting our readout probe sequences to a 3-letter alphabet, scaling up the field of view area, and optimizing the acquisition and analysis software, we increased MERFISH throughput by a factor of ~100, and demonstrate the ability to measure ~40,000 cells in 18 hr. The experimental changes that enabled this significant increase in measurement throughput did not negatively affect the quality of MERFISH when measured against bulk sequencing and smFISH standards. We devised a method to remove background signal in MERFISH by casting a polyacrylamide gel over the sample, imprinting the RNA signal directly into the gel matrix, and removing proteins and lipids through clearing. In cleared cultured cell samples, we demonstrated the ability to perform MERFISH measurements in up to 4 colors, and detected no loss in RNA counts despite removing cellular structural components. Additionally, we performed MERFISH measurements in tissue sections from the mouse hypothalamus medial preoptic area, and found good correlation in RNA counts measured by MERFISH and average expression levels measured by bulk sequencing of the same brain region. Understanding the mechanisms and principles that govern organization of neuronal circuits in the brain requires probing biological interactions ranging in scale from nanometers between molecules to millimeters or more between cells. STORM and MERFISH are two powerful imaging techniques applicable to these length scales, and as both techniques continue to be improved, can perhaps together clarify complex trans-synaptic code of alternatively spliced isoforms of neurexins and their ligands, and its role in shaping synaptic identity and function.