Heterogeneity in attenuation and timing of simulated excitatory synaptic potentials following propagation from distal dendritic locations to the soma.

a,b, 2-photon fluorescence-guided dual dendritic and somatic patch recordings from two MSO neurons of comparable distances (78 and 73 µm, respectively). Example neuron in a1 shows stronger attenuation of somatic propagation of a simulated EPSP (a2 vs. b2) and longer peak delay (a3 vs. b3). EPSC amplitudes: cell a, 600 pA; cell b, 400 pA. c,d, Group data for attenuation and peak delay for all paired MSO neuron recordings. There is a poor correlation of both measures as a function of distance.

EPSP propagation delays are not correlated with local morphology but rather more global physiological properties.

a, Closeups of distal dendritic morphology during paired dendritic and somatic recordings. b,c, Delay of peak EPSPs during propagation from dendrite to soma shows no significant correlation with local diameter or tapering rate of local dendrite. (dendritic EPSCs: 400-1000 pA). d, Peak EPSP delay from dendrite to soma was weakly correlated with the membrane time constant recorded at the dendritic recording site. e, Dendritic membrane time constant increased with distance of the recording site from the soma.

MSO neuron morphology is inhomogeneous and asymmetric, regardless of tonotopy.

a, Reconstructed morphologies rank ordered by tonotopy (left) and degree of asymmetry (right; average path length). Orderings span from low frequency (top) to high frequency (bottom) and medial (top) and lateral (bottom) asymmetry, respectively. Colored gradients added to visualize ordering. Uniquely colored dots placed beneath each morphology for differentiation. b, Fold difference between average path length of medial and lateral dendrites for all reconstructed morphologies (n=40) with vertical axes ordered by tonotopic location (top) or degree of asymmetry (bottom). Gradients added as in a. Colored points correspond to a. c, Same as b, except comparing total surface area of medial and lateral branches. Average medio-lateral path length and surface area fold differences (± SEM) were 1.080 ± 0.057 and 1.112 ± 0.133 (biased laterally) and tonotopic R2 values were -0.42 and -0.23.

MSO morphology nonlinearly filters and attenuates propagating EPSPs through its passive cable properties.

a, Visualization EPSP filtering at different dendritic sites. Cell morphology (top) with similarly distant example synaptic sites indicated with green circles and corresponding numbers. Red triangle indicates the somatic compartment. Synaptic conductance indicated as gsyn. Four sets of paired dendritic (green) and somatic (red) traces (bottom) for each example synaptic site. Numbering of traces corresponds to labels on morphology above. Dendritic distance from soma for each site indicated below traces. b: Same procedure as in a, with an adjusted unitary synaptic conductance (see Methods), with measurements from each computational compartment. Model morphology and orientation same as in a. Comparison between active channel (left) and passive channel model (right). Recorded EPSP amplitude at synaptic (top, green) and somatic (bottom, red) compartment as a function of distance from soma. c, Actual (top) and electrotonically transformed (bottom) morphologies. Transformation visualizes degree of EPSP attenuation from each compartment (see Methods). d, Group data (n=40) of average somatic depolarization asymmetry between mediolateral sides. Gradient as in Fig 3. Dotted line indicates equal medial and lateral average somatic depolarizations. Cell from a-c indicated with a triangle.

EPSP timing significantly varies within and between cells, with morphological asymmetry informing overall cell timing.

a, EPSP delay visualization in two exemplar MSO neurons. Coloring of each cell corresponds to the propagation delay of EPSPs to the soma from its respective compartment. Legend of propagation delay coloring in top right. Cells are oriented mediolaterally according to the scale bar. Outline added to increase gradient readability. b, Same data as in a, except as a function of distance from the soma. Orientation of cells same as in a. c, Group data (n=40) of the difference in average EPSP propagation delay between medial and lateral dendrites. Cells 1 & 2 are indicated with same color as in b. Dotted line at 0 ms indicates an equal average EPSP delay between sides. Blue gradient along horizontal axis (as in Fig. 3) to indicate spectrum of asymmetry. d, Morphological transforms according to EPSP delays. Morphological centroids of actual cell morphologies (top) and transformed morphologies (bottom). Length corresponds to each sections EPSP delay contribution. Coloring of cell sections is consistent between actual and transformed morphologies for easier differentiation and comparison. Dotted line added to mark the midline.

Simulated ITDs demonstrate conservation of morphological impact on coincident detection.

a, Schematic visualization of ITD innervation. Somatic compartment and dendrites are shown in gray and black, respectively. Green bulbs represent excitatory synapses, with corresponding arrows to illustrate axonal input velocity (νinput). Dotted lines illustrate point at which there is considered to be no axonal input delay (see Methods). Axon attached to soma, with initial segment, nodes, and internodes. b, Example somatic (top) and axonal (bottom) traces from ITD simulations. ITDs of -0.2 ms (left), 0.1 ms (middle, optimal coincidence), 0.3 ms (right) shown. Traces colored only for differentiation. Dotted line (Vthreshold) indicates depolarization threshold for action potential consideration. Morphology of model cell indicated in upper right. c, Three example ITD firing curves from varied morphologies, medially biased (left), symmetrical (middle), and laterally biased (right). Corresponding morphologies imaged above. Light orange span represents range of ecologically possible ITDs for model species. Dashed line placed at ITD of 0 ms. Dot indicates centroid ITD value. d, Sorted group data (n=39) of centroid ITDs, with νinput = 1 ms-1. Cells from c indicated with filled colored points. e,f, Sorted group data (n=39) of absolute centroid ITDs, varying axonal input velocity (e) and peak firing probability (f). Cells from c indicated with correspondingly filled colored points.

Inhibition may serve to modulate a pre-established morphological baseline ITD response.

a, Schematic of innervation pattern with excitatory (green) and inhibitory (red) inputs. b, Somatic (top) and axonal (bottom) traces from ITD stimulation (ITD: 0.1 ms; IPSP timing: -0.5 ms) for each tested IPSP amplitude. Tested amplitude values and their unique colors above corresponding traces. Morphology of the exemplar neuron shown in the upper right. c, Exemplar cell’s ITD response curves colored corresponding to IPSP amplitude in b. Centroid (square) and maximum slope (circle) are indicated for each curve. Dotted line at an ITD of zero. d, ITD centroid and maximum slope as a function of IPSP amplitude (colors corresponding to b). e,f, Group data (n=39) for ITD centroids (e) and steepest sections (f) for each IPSP amplitude. Data in e are largely overlapping. Shifting of population data in f corresponds to trend seen in the exemplar cell (d). g-k, Same as b-f, instead varying the timing of the inhibition (IPSP amplitude: 3 mV). Absolute firing probability, in h, is lower than c, resulting in lower signal to noise ratio. Centroids in i show more variation than d, though still minimal. Shift of exemplar cell’s maximum slope ITD section (i) corresponds to degree of reduction in absolute firing probability (h).

Maximal channel conductances (µS/cm2).

Tapered axon initial segment (TAIS) and constant disameter axon initial segment (CAIS).