Extended Data Fig. 10: A model for emergence of tonotopic representation in the hindlimb dorsal column–medial lemniscal pathway.

a, Top, experimental schematic for blocking the signal transmission by the indirect dorsal column pathway through the spinal cord dorsal horn neurons (pink). Bottom, the frequency preference of neurons are colour-coded before (left) and after (right) the application of synaptic inhibitors (MK-801 and NBQX) of excitatory synaptic transmission to the spinal cord. b, Example tuning curves of nine DCN neurons showing various changes in response amplitude after application of blocking the signal transmission by the indirect dorsal column pathway. c, Left, the correlation of the frequency preference of individual neurons measured before and after the inactivation of the synaptic activity within the spinal cord (linear regression, P = 0.00007). Right, the change of the calcium response amplitude of individual neurons in three different groups, categorized by their preferred frequency, before and after the inactivation (paired-sample t-test, two-sided, *P < 0.001). d, An example photo of the imaging window over the spinal cord. e, Left, schematic of the clusters of neuronal cell bodies (top) and example fields-of-view under two-photon microscope (bottom). Right, each neuron's preference for location and frequency is colour-coded. f, The relationship of horizontal distances between neurons in a pair and their difference in location (top) and frequency (bottom) is calculated (*P < 0.01, permutation test). The slope of the linear fit (SLF, italic) of the first 5 data points from the original data (black line) is shown. Error bars represent S.E.M. g, An example photo of the imaging window over the inferior colliculus. h, Left, schematic of the clusters of neuronal cell bodies (top) and example fields-of-view (bottom). Right, each neuron's preference for location and frequency is colour-coded. i, The relationship of horizontal distance between neurons in a pair and their difference in location (top) and frequency (bottom) is calculated (*P < 0.01, permutation test). The slope of the linear fit (SLF, italic) of the first 5 data points from the original data (black line) is shown. Error bars represent S.E.M. All the neurons shown here did not respond to auditory stimulation with the hindlimb off the stimulator. This suggests they are a specific population in inferior colliculus that only responds to tactile stimulation. n/N = number of neurons and mice. j, Mechanosensitive end-organs and their LTMR afferents - Merkel cells (blue), Meissner corpuscles (yellow) and Pacinian corpuscles (red) - detect a wide frequency spectrum of vibrations (0.1 – 2000 Hz) in the periphery. The end-organs are distributed throughout the hindlimb, but each is most densely found in one area: Merkel cells in the toes, Meissner corpuscles in the foot pads and Pacinian corpuscles along the fibula. At the cell bodies of the LTMRs in the dorsal root ganglia no tonotopic organisation is observed. However, at their central projections in the gracile nucleus of the brainstem, but not in the spinal cord dorsal horn, fine-scale topographic organisation arises. Furthermore, inputs from the same end-organ group are found projecting onto the gracile nucleus with a rostrocaudal bias. Each colour-coded circle illustrates a synaptic connection. The selective dendritic sampling of the gracile nucleus neurons (black) is the basis for the emergence of a topographic map in the brainstem. Along the ascending pathway, the tonotopic map found in the gracile nucleus is partially preserved in the midbrain, thalamus and hindlimb somatosensory cortex.

Source data