Extended Data Fig. 6: Optogenetic stimulation of LH^Vglut2 neurons augments walking after contusion SCI via vGi neurons.

a, Schematic of the experimental scheme demonstrating contusion at T9 in the mouse. Cleared spinal cord demonstrating a representative lesion with cross-sectional histology of the lesion epicenter (preserved tissue stained with GFAP) and associated lesion reconstruction. b, Kinematic analysis of mice following contusion SCI. Walking was quantified using principal component analysis as described in Fig. 1c and Extended Data Fig. 1c (n ≥ 10 gait cycles per mouse, n = 5 mice in the uninjured group, n = 8 in the contusion SCI group). Statistics are provided in Supplementary Data 1. c, LH connections with the lumbar spinal cord are not reliably preserved after contusion SCI. Bilateral G-deleted rabies were injected into the lumbar spinal cord at 8 weeks post-SCI (n = 3 mice). After 4 days, mice were perfused and the brains extracted, cleared, imaged, and registered to the Allen Brain Atlas. Neurons were segmented and quantified. Left, Representative cleared whole mouse brain and spinal cord demonstrating injection site, contusion, and labeling of brain neurons. Middle, 3D brain representations demonstrating labeling in the vGi but not the lateral hypothalamus with associated optical sections (50 µm projection). Right, Bar plots demonstrating the percentage of all neurons with direct connections to the lumbar spinal cord after contusion for several regions with the highest proportion of neurons. vGi neurons (gigantocellular nucleus, magnocellular nucleus) with direct projections past the contusion SCI were prominent. We noted an absence of direct projections from the LH to the lumbar spinal cord. Lesion reconstructions and quantifications for the animals analyzed are also provided. Mean sparing = 29.0%. 3 V: third ventricle. d, Schematic of the experimental scheme to understand efferent projections from LH^Vglut2 neurons to various motor centers in the brainstem. Vglut2^Cre mice underwent a left lateral thoracic hemisection and were allowed to spontaneously recover for 6 weeks (n = 3 mice). To understand the efferent projections of recovered mice, AAV5-DIO-mGFP was injected into the right LH of these mice, which was allowed to express for 4 weeks. Photomicrograph demonstrating a representative injection site with labeled LH^Vglut2 neurons. Representative images demonstrating axon projections from LH^Vglut2 neurons to various brainstem motor centers. Top panel demonstrates projections to the PAG (periaqueductal gray), CnF (cuneiform nucleus), and PTg (pedunculopontine tegmental nucleus). Middle panels demonstrate projections to the PnC (pontine reticular nucleus) and RMg (raphe magnus). Bottom panels demonstrate projections to the medullary reticular formation including the Gi (gigantocellular reticular nucleus) and vGi, and LPGi (lateral paragigantocellular reticular nucleus). Insets demonstrate magnified views of the axons. Density quantification of fibers to each of the brainstem motor centers divided into ipsilateral, contralateral, and midline (for RMg) quantifications. 3 V: third ventricle.