Fig. 5: Inputs in the cortices of SS, AI, and PIR have a cortical layer-specific property and are clustered into three groups according to their morphological characteristics.

a Representative image of tile-scanned coronal sections of inputs in the SS, AI, and PIR. (Scale bar: 500 μm and 150 μm, respectively). b All reconstructed neurons were clustered into three groups by their 11 morphological characteristics. The representative neurons displayed in each group illustrated that the complexity of neuronal structure occupied a larger proportion in the clustering weight and that there were no significant differences in the sizes of neuronal somata or their fibers. The relative scale data were normalized by the z score method for each morphological parameter. c Correlation between 11 morphological parameters and somatic depth. Strong correlations were found between parameters demonstrating the structural complexity of neurons (i.e., number of stems, bifurcations, branches, and tips), while parameters showing the sizes of neurons also exhibited strong correlations (i.e., Euclidean distance, path distance, and branch path length). d–h The complexity of the inputs—numbers of stems (d, r² = 0.1080, p = 0.0226), bifurcations (e, r² = 0.2812, p = 0.0001), branches (f, r² = 0.2922, p < 0.0001), tips (g, r² = 0.2923, p < 0.0001), and branch path length (h, r² = 0.1920, p = 0.0018)—in these three regions were negatively correlated with their somatic depths, suggesting that the cytoarchitecture of the input neurons tended to have cortical depth-specific organization, and that inputs with deeper distributions along the depth axis were larger but simpler in structure (N = 48). i Immunofluorescent-labeled NECAB1 (green), a marker of layer 4, illustrating that inputs (red) were densely distributed in layer 4 on the dorsal side of the SS, while they were more dispersed toward the ventral side in the cortex (Scale bar: 500 μm and 100 μm, respectively). j–m The numbers of bifurcations (j, t [22] = 4.650, p < 0.001), branches (k, t [22] = 4.487, p < 0.001), tips (l, t [22] = 4.223, p < 0.001) and branch path length (m, t [22] = 3.036, p = 0.0061) of the input neurons in layer 2/3 were markedly greater than those in layer 4 of SS and AI with respect to layers and somatic depth. Data are mean ± s.e.m., two-tailed unpaired t test, layer 2/3: N = 14, layer 4: N = 10.