Fig 1.
Overview and confocal photomicrographs of human and mouse striatal projection neurons (SPNs) injected with Lucifer Yellow (LY) in the human putamen and the mouse caudoputamen (CPu).
(A) Macroscopic image of a thick coronal human brain slice. The rectangle surrounds the caudate and putamen regions. The asterisk indicates the location where brain tissue was removed for further processing and analysis. (B) Schematic drawing from the human brain reference atlas of the Allen Institute at a similar level of the image in A (https://atlas.brain-map.org/). (C) Image of a coronal section of the human brain containing the caudate and putamen regions. Note that the internal capsule is visible. Green dots point out the approximate location where SPNs were intracellularly injected with LY in the putamen region. (D) Nissl staining and schematic drawing from the mouse brain reference atlas of the Allen Institute at a similar level of the image in E (https://atlas.brain-map.org/). CPu is shown in light blue colour. (E) Coronal mouse brain section containing the CPu region. Green dots point to the approximate location where SPNs were intracellularly injected with LY in the CPu region. (F, G) Confocal photomicrographs showing an injected human (F) and mouse (G) SPN. Scale bar is 1.56 cm in A; 1.36 cm in B; 3 mm in C; 1.06 mm in D; 0.77 mm in E; and 66 µm in F and G.
Fig 2.
Repairing morphological reconstructions (soma centre as black dot, dendrite centre lines in blue, diameter not up to scale).
(A) Cut points are identified in the (x, y)-plane as dendritic terminal nodes in the vicinity of the maximum value of the z-coordinate of the reconstruction (red dots). Scale bars are 100 μm. (B) Complete dendrites are selected as dendritic branches not containing cut points (shown in blue). (C) The predicted morphology is composed of half of the original morphology artificially cut at the level of the soma and its mirror image flipped along the z-axis (grey lines). (D) Incomplete dendrites are extended using parts of the complete dendrites of the same topological order as the corresponding cut points (shown in red). (E) Validation of the repair process for human SPNs. Distributions of morphometric features of the repaired reconstructions (dark blue, 147 dendrites in 27 reconstructions) is compared to the morphometry of the complete dendrites (light blue, 35 complete dendrites), Z-scores are shown in bold for each morphometric feature.
Fig 3.
Morphometry of the dendritic branches.
(A) Example dendrograms of the dendritic branches with average number of terminal segments (mouse in red, human in blue). (B) Boxplot diagrams for morphometric features of individual dendritic branches (red mouse, blue human, white line for median values, outliers shown with dots). Statistical significance p < 0.001 (***) and p > 0.05 (n/s). Total dendritic length and total terminal length are significantly longer in human SPNs (p = 0.00028 and p = 0.00008, respectively), while number of terminals and maximal order are similar (p = 0.74889 and p = 0.83002, respectively). (C) Mean length of the dendritic segment per topological order (order 1 for primary dendrites, etc.). Whisker lines for STD, unfilled boxes for single dendrites. (D) Relative length of the terminal segments normalised to the total length of the dendritic branch. The number of complete dendrites analyzed in (B-D) is 47 for mouse and 35 for human.
Fig 4.
Morphometry of repaired SPN reconstructions.
(A) Violin plot diagrams for neuronal morphometric features (dots for mean values). (B) Sholl diagram of SPN reconstructions. Solid lines for mean values, shaded areas for standard deviation and dots for individual reconstructions. Step of the Sholl radii is 10 μm (see inset). The number of reconstrcutions analyzed in (A-B) is 31 for mouse and 27 for human.
Fig 5.
Dendritic diameter as function of distance to the soma.
(A) Best fit for primary, intermediate and terminal dendritic segments, boxplots for segments lengths. Mean values are shown in the insets for base and end diameters, segment lengths, exponential decay or linear taper rates. Dots correspond to the reconstructed points of the complete dendrites (47 complete mouse dendrites and 35 complete human dendrites). (B) Mean diameter and mean length of the primary, intermediate and terminal segments of the complete dendrites (left to right, respectively). (C) Diameter of a given dendritic segment d is linearly proportional to the total dendritic length L of the distal part of a dendritic branch originating from that segment (see inset; arrows point out a dendritic node and corresponding dendritic subtree in red).
Fig 6.
Reconstruction of the dendritic spines of human SPNs.
(A) Confocal photomicrograph of a horizontally projecting dendrite from an intracellular injected human SPN. Soma is shown with an arrow, and the rectangle depicts the region of interest. (B) Magnified ROI. (C) Same image as in B to illustrate the position of spines. (D) 3D reconstructions of spines showing a clear head. Spine head area is shown in red, spine neck length in white numbers (see S1 Table for full statistics). Scale bar (in D) is 16 μm in A and 5 μm in B-D.
Fig 7.
Characterisation of the dendritic spines.
(A) Total spine area calculated in four reconstructed dendrites of mouse (red) and human (blue) neurons. Numbers correspond to the total number of spines per dendrite. (B) Total membrane area of dendritic spines and corresponding dendritic shafts in four reconstructed dendrites of mouse (red) and human (blue) neurons. (C) Dendritic spine density with distance from the soma. Parameters of the sigmoidal distribution (a, b, c) are given in the insets. (D) Proportion of spines located at the terminal dendritic sections (p = 0.000001). (E) Total number of spines per neuron (p = 0.67292) estimated for the repaired morphological reconstructions using spine density distributions from (C).
Fig 8.
Simulation of the detailed biophysical models of SPNs for mouse (red) and human (blue).
(A) Experimental current clamp recording from the mouse D1-positive SPNs. (B-C) Simulated neuron models optimised to the experimental data (A) using mouse (B) and human (C) morphologies. (D-E) Ensembles of validated models optimised to three selected SPNs, experimental data plotted in grey. Subthreshold current-voltage responses (D) and suprathreshold discharge rates (E) are shown. Shaded areas display the range of the values within each ensemble. (F) Magnified current-frequency relation for one set of models (from the middle of (E)). Black color for experimental data, red for the mouse and blue for the human morphologies. (G) Current-frequency relation with exchanged morphologies, mouse SPN reconstruction is scaled to human proportions (retains red color) and human reconstruction is downscaled to mouse dimensions (retains blue color). Rescaled mouse and human SPN reconstructions behave as valid replacements of each other in electrophysiological simulation.
Fig 9.
Electric properties of the dendritic terminal segments of the mouse (red, 35 terminals) and human (blue, 37 terminals) neuron morphologies in corresponding computational neuron models.
(A) Electrotonic length of the terminal segments. (B) Shape of EPSP in the soma (dark line) and in the terminal dendritic segment (light lines) at different locations on the terminal dendritic segment (points 1-5). The glutamatergic synapse is located at the tip (point 1). (C-D) Response to simulated glutamatergic synaptic inputs, AMPA and NMDA. (C) Distributed asynchronous synaptic drive. Left: 10 synapses placed uniformly along the terminal segment; right: membrane depolarisation in the dendritic segment (light lines) and soma (dark lines), error bars for STD, variability for all terminals. (D) Clustered synchronous synaptic drive. Left: synapses are placed in the middle of a terminal segment and stimulated in sequence with 1 ms interval; right: EPSPs in the dendritic segment (light lines) and soma (dark lines). Prolonged NMDA-spike in a dendrite may trigger action potential in the soma (truncated). Inset, shows the effect of setting the NMDA part to 0. Scale bars in (B-D) are 100 μm.