Connectomic studies of the mammalian neuromuscular system

In order to understand the properties of the mammalian nervous system, it is prerequisite to know its connectome, i.e., the full connectivity of its neuronal components. The mammalian neuromuscular circuit is a good starting point for connectomic studies due to its accessibility and simplicity in organization. Understanding the neuromuscular circuit may provide important insights to the organization of more complicated circuits such as the mammalian central nervous system (CNS). High resolution confocal fluorescent microscopy and transgenic mice that express cytoplasmic yellow fluorescent protein in all motor axons were used in this study. In order to identify individual axons from the image stacks, a semi-automated tool was developed, using the region-growing algorithm for image segmentation. Segmented contours were automatically propagated to successive sections, and the results were rendered in 3D. Segmentations from individual stacks were then assembled into montages covering entire axonal arbors. To achieve more automated segmentation, a repulsive force-based gradient vector flow (GVF) snake algorithm was developed to ensure separation of neighboring axons in image segmentation. With these imaging and image analysis methods, I addressed the question of structural stereotypy of mammalian neurons that have identical genetic background and functions. 3D reconstruction of axonal arbors from tens of thousands of confocal images revealed the connectome of a small ear muscle, the interscutularis. The connectome was compared to the contralateral connectome in the same animal and to ipsilateral connectomes in other individuals. In both cases, considerable variability was found in the connectivity and branching patterns between neurons which, according the size principle, execute identical functions. This result indicates that functional organizations of the mammalian nervous system may emerge at a macroscopic level, while the structural implementations vary. The variability in structural details may have emerged from the developmental reorganization of axonal arbors which leaves each of them unique, and this developmental strategy may be an evolutionary innovation. These findings may have significant implications for future connectomic studies of the mammalian CNS.