Molecular genetic dissection of the development and function of spinal somatosensory circuitry
Molecular Biology and Physiology of Neurons and Glial Cells
Final Report Abstract
To start analyzing the development of spinal somatosensory circuitry, I characterized Lhx9GFP- CreER (Lhx9GCE) mice that express CreER specifically in one class of dorsal spinal cord (SC) neurons (dI1 neurons) in the embryonic mouse. In Lhx9GCE; Ai14tdTomato embryos, pulsed induction of reporter gene expression through administration of 4-hydroxy-tamoxifen (4-OHT) revealed that timed neurogenesis of dI1 neurons controls their medio-lateral settling position within the SC. While early-born neurons occupy more lateral positions, I found later-born neurons to form a more medial cluster of cells. This difference in position was accompanied by a change in projection laterality and suggested that laterally positioned, ipsilaterally projecting dI1 neurons differentiate first and are the earliest to arrive at their final settling positions before medial, contralaterally projecting neurons arise. My findings are the first indication that timing of neurogenesis can control projection laterality and therefore directly determine circuitry. Surprisingly, early-born dI1 neurons then exhibit a dramatic switch in projection position from extending to and crossing the ventral midline to extending laterally into the dorsomedial ipsilateral white matter, a phenomenon that has not been reported previously and presumably requiring complex cellular alteration. I will continue to study these effects using live imaging. I have created mouse lines that permit genetic access to dI1 neurons throughout development for anterograde and retrograde circuit tracing. The first line, TauNCLXN, extends Cre expression from embryonically transient, dI1-specific Cre driver lines into postnatal stages, enabling targeted, Cre-dependent expression of reporter genes in dI1. The second line, Ai9TVA, expresses TVA receptor and N2c glycoprotein (G) for entry and complementation of pseudotyped, G-deleted rabies virus (RV-N2c-dG) in a Cre dependent manner, eliminating the need to provide TVA and G through in utero injection of helper virus and facilitating tracing experiments in neonates. Using trans-synaptic retrograde RV-N2c-dG, I have identified several monosynaptic inputs to dI1 neurons from the brain, the SC and dorsal root ganglia (DGR). In the brain, input neurons are located in the midbrain red nucleus, layer 5 of sensory cortex, and the medullary and pontine reticulate nuclei, suggesting that both sensory and motor information impinges on dI1 neurons. In the periphery, I have identified proprioceptors and slowly-adapting low threshold mechanoreceptors providing primary sensory input to dI1s – the first evidence that dI1s serve more than one sensory modality. Together, my findings suggest that dI1 neurons integrate primary proprioceptive and mechanosensory information with central sensory signals and voluntary and automatic motor signals, placing dIs at the positional and functional interface between dorsal (sensory) and ventral (motor) spinal interneurons. Mapping input neurons within the SC is a major challenge because of a lack of methods for efficient SC-wide anatomical analysis. To facilitate my work and that of others in the SC field, I therefore created Spine Racks for high-throughput sectioning of whole SC and developed software (SpinalJ) for automated 3D reconstruction, atlas mapping and cell and projection analysis of SC sections. These tools provide the first annotated 3D common coordinate framework for adult mouse SC, a resource that enables direct comparison of datasets from different animals and across groups.
