Fig 2: Density of VIP+ INs and boutons in the mouse BLA. A, Magnetic resonance images in the coronal plane of a mouse brain before (left) and after (right) transcardial perfusion obtained using a 3 tesla MRI device. B, 3D reconstruction of the distribution of VIP+ INs (green dots) within the BLA. Note the higher density of VIP+ INs in the more caudal portion of the LA nucleus. C, D, Rostrocaudal density (C; cells/mm3) and number (D) of VIP+ cells in the mouse BLA. x-axis represents all bregma levels analyzed and containing the BLA at 50 µm intervals. N = 4 mice. Lineplots depict the mean and the filled area represents SEM. E, Correlation of the density of VIP+ somata and boutons in three different nuclear subdivisions of the BLA. Each dot in the graphs represents the counts from one amygdala in the given subdivision. BA, Basal nucleus of the amygdala; LA, lateral nucleus of the amygdala. Scale bars: A, 5 mm; B, 300 µm.

Fig 3: VIP+ INs innervate distinct populations of cells labeled in GAD65-EGFP knock-in mice. A, An image taken from the amygdala region of a GAD65-EGFP mouse. There is a clear difference in the quantity of EGFP-expressing cells between the BLA and striatal-like amygdala regions (CeA, central amygdala; Astr, amygdalostriatal area). B, CCK, NPY, VIP, CR, and CaMKII, but not PV, were found to be expressed in EGFP+ neurons. White arrows show green cells that express the given neurochemical markers, white arrowheads indicate green cells that do not contain the given neurochemical markers, and open arrowheads point to PV-containing cells that were not EGFP+. C, Pie charts show the percentage of EGFP+ neurons containing the given neurochemical markers in both LA and BA nuclei. The “Others” category comprises all green cells that lack the immunoreactivity for the neurochemical markers tested and the three SOM+/EGFP+ neurons. As VIP and CR form partially overlapping groups of INs, they were pooled together. D, Maximum projection intensity images of representative biocytin-filled neurons from four cell types expressing EGFP. Insets show their characteristic responses upon step current injections. CCKBC, CCK+ basket cell; NGF, neurogliaform cell; PN, principal neuron. Below each image, the postsynaptic currents evoked by whole-field blue-light illumination with high chloride intrapipette solution are shown. Blue arrows show the beginning of a 10-ms-long light illumination. Gray traces indicate five consecutive events, while the black traces are their averages. E, Bar graph shows the percentage of different types of EGFP+ neurons in which blue light-evoked postsynaptic events could be detected. F, Plot shows the magnitude of inhibitory postsynaptic responses recorded in different types of EGFP+ neurons (CCK-BC, 0.77 ± 0.21 nS, n = 6; NGF, 0.41 ± 0.11 nS, n = 7; IS-IN, 0.66 ± 0.14 nS, n = 12; PN, 0.89 ± 0.54 nS, n = 7). Each dot represents the average peak conductance (IPSG) obtained in individual cells, while horizontal bars indicate the mean. Scale bars: A, 150 µm; B, 20 µm; D, top, 40 µm (inset: x = 100 ms, y = 20 mV); D, bottom, x = 5 ms, y = 10 pA).

Fig 4: Genetic labeling reveals absence of cortical interneuron cell loss in R6/2 mice. (A,C,E,G) Representative images of the primary motor cortex layer 2/3 from 12-week-old mice with the indicated interneuron populations labeled with tdTomato. The sections were stained with NeuroTrace. (B,D,F,H) Quantifications of the genetically labeled cells. Values were normalized to controls. N = 4–5 mice. Unpaired two-sided t-test. (I,K,M) Representative images of the primary motor cortex layer 2/3 from 12-week-old mice. Genetic tracing was combined with immunostaining for the indicated markers. Arrowheads in panels (K,M) point to tdTomato-labeled cells showing no expression of the respective molecular marker. (J,L,N) Quantifications of the fraction of genetically labeled cells immunopositive for the respective marker. Values were normalized to 12-week-old controls. PV: N = 3 mice. Unpaired two-sided t-test, n.s. SST: N = 3–4 mice. Two-way ANOVA with Bonferroni’s multiple comparisons test. ANOVA: Genotype, *?*?**p < 0.0001; Age, **p = 0.0012; Genotype × age, ***p = 0.0004. Significant pairwise multiple comparisons are indicated on the graph. VIP: N = 4–5 mice. Two-way ANOVA with Bonferroni’s multiple comparisons test. ANOVA: Genotype, **p = 0.0083; Age, ***p = 0.0002; Genotype × age, ***p = 0.0005. Significant pairwise multiple comparisons are indicated on the graph. *p < 0.05; **p < 0.01; *?*?**p < 0.0001. Scale bars in panels (A,C,E,G,I,K,M) are 50 µm.

Fig 5: Selective reduction in SST and VIP-positive interneurons in the cingulate cortex of R6/2 mice. (A,C,E) Representative images of the cingulate cortex immunostained for the indicated molecular markers. Cortical layers are marked with dashed lines and labeled on the left. (B,D,F) Quantifications of cell densities per layer. N = 3 mice for all the groups. RM two-way ANOVA with Bonferroni’s multiple comparisons test. PV cells: Genotype, n.s.; Layer, ***p = 0.0003; Genotype × layer, n.s. SST cells: Genotype, **p = 0.0034; Layer, n.s.; Genotype × layer, *p = 0.0285. VIP cells: Genotype, *p = 0.0272; Layer, *?*?**p < 0.0001; Genotype × layer, *p = 0.0176. Significant pairwise comparisons are indicated on the graphs. **p < 0.01; ***p < 0.001. Scale bars in panels (A,C,E) are 100 µm.