Fig. 6: The persistence of developmental TC inputs disrupts exploratory behavior in adulthood. | Nature Communications

Fig. 6: The persistence of developmental TC inputs disrupts exploratory behavior in adulthood.

From: Metabotropic signaling within somatostatin interneurons controls transient thalamocortical inputs during development

Fig. 6

a Strategy for early deletion of mGluR1 in SST cINs, using CRISPR/Cas9 system (see Fig. 4a). AAVs were injected bilaterally in S1 at P0. Motion Sequencing (MoSeq) was performed around P60 (44 females (24 Ctrl and 20 KD) and 32 males (20 Ctrl and 12 KD) were analyzed). b MoSeq pipeline and examples of behavioral motifs (syllables). Modified from ref. 67. c Position occupancy heatmap. The top two rows are the min-max scaling of the occupancy percentages (max = 1, min = 0). The bottom row is the difference between occupancy percentages for mGluR1 KD and Ctrl mice (Δ index). d Top, Syllables for Ctrl and mGluR1 KD males, sorted by syllable usage differences between mGluR1 KD and Ctrl (Chi-square test on the unnormalized syllable usages p = 1e-6). The prominent syllables (see methods) are labeled with red stars (27 syllables for male, and 27 for female). The syllable behavior types are color-coded (i.e., light gray for dart, green for groom, red for rear, etc.). The inset subpanel above is aggregated usages by syllable classes and the differences between the two groups are significant (Chi-square test on the unnormalized syllable usages p = 1e-6). Bottom, Same as d but for female mice (syllable usage differences, p = 1e-6; aggregated usages by syllable classes, p = 1e-6). e Linear discriminant analysis (LDA) plot shows that Ctrl and mGluR1 KD mice are separated by the syllable ensemble (males and females combined). The top ten syllables with the highest absolute model weights consist of 7 rears, 2 grooms, and 1 scrunch. f Confusion matrices for classification accuracy of a linear classifier trained on the 19 prominent syllables identified from 76 LDA iterations by sex (see methods). The use of these syllables alone could predict Ctrl or KD mice. The classifier yields to 0.69 and 0.84 accuracy across cross-validation for females and males, respectively (top/bottom). Labels per-mouse are shuffled for control (right). Source data are provided in Zenodo repository109.

Back to article page