Characterization of Type I and Type II nNOS-Expressing Interneurons in the Barrel Cortex of Mouse

Abstract

IN THE NEOCORTEX, NEURONAL NITRIC OXIDE (NO) SYNTHASE (NNOS) IS ESSENTIALLY EXPRESSED IN TWO CLASSES OF GABAERGIC NEURONS: type I neurons displaying high levels of expression and type II neurons displaying weaker expression. Using immunocytochemistry in mice expressing GFP under the control of the glutamic acid decarboxylase 67k (GAD67) promoter, we studied the distribution of type I and type II neurons in the barrel cortex and their expression of parvalbumin (PV), somatostatin (SOM), and vasoactive intestinal peptide (VIP). We found that type I neurons were predominantly located in deeper layers and expressed SOM (91.5%) while type II neurons were concentrated in layer II/III and VI and expressed PV (17.7%), SOM (18.7%), and VIP (10.2%). We then characterized neurons expressing nNOS mRNA (n = 42 cells) ex vivo, using whole-cell recordings coupled to single-cell reverse transcription-PCR and biocytin labeling. Unsupervised cluster analysis of this sample disclosed four classes. One cluster (n = 7) corresponded to large, deep layer neurons, displaying a high expression of SOM (85.7%) and was thus very likely to correspond to type I neurons. The three other clusters were identified as putative type II cells and corresponded to neurogliaform-like interneurons (n = 19), deep layer neurons expressing PV or SOM (n = 9), and neurons expressing VIP (n = 7). Finally, we performed nNOS immunohistochemistry on mouse lines in which GFP labeling revealed the expression of two specific developmental genes (Lhx6 and 5-HT(3A)). We found that type I neurons expressed Lhx6 but never 5-HT(3A), indicating that they originate in the medial ganglionic eminence (MGE). Type II neurons expressed Lhx6 (63%) and 5-HT(3A) (34.4%) supporting their derivation either from the MGE or from the caudal ganglionic eminence (CGE) and the entopeduncular and dorsal preoptic areas. Together, our results in the barrel cortex of mouse support the view that type I neurons form a specific class of SOM-expressing neurons while type II neurons comprise at least three classes.

Keywords: Immunohistochemistry; development; neuropeptide Y; nitric oxide; parvalbumin; patch-clamp; somatostatin; vasointestinal peptide.

Figure 1

Distribution of type I and type II nNOS-expressing GABAergic interneurons in the mouse barrel cortex. (A) Colocalization of GFP (1) with nNOS immunolabeling (2) in GAD67:GFP mice. Arrows point to examples of type I (heavily labeled) nNOS-expressing GABAergic neurons and arrowheads point to examples of type II (weakly labeled) nNOS-expressing GABAergic cells. (B) Histogram of the density of type I (1) and type II (2) nNOS immunolabeled GFP-expressing cells in cortical layers I–VI (n = 3 mice; error bars: standard deviation; *, **, and *** indicate statistically significant differences with p ≤ 0.05, p ≤ 0.01, and p ≤ 0.001, respectively; roman numerals refer to neocortical layers).

Figure 2

Distributions of SOM, PV, and VIP expressing cells within type I and type II nNOS immunolabeled GABAergic interneurons. (A–C) Colocalization of SOM, PV, and VIP (1) with nNOS immunolabeling (2) in GAD67:GFP knock-in mice. Arrows point to examples of type I (heavily labeled) nNOS-expressing neurons, arrowheads point to examples of type II nNOS (weakly labeled) expressing neurons. Double-labeled cells are marked by an asterisk. (D–F) Histograms reflecting the densities of SOM (green), PV (red), and VIP (blue) expressing neurons within type I (1) and type II (2) nNOS-GFP-expressing interneuron classes in cortical layers I–VI. The densities of doubled labeled cells are superimposed on the overall densities of nNOS type I (white) and nNOS type II (gray) interneurons (n = 3 mice; error bars: standard deviation; *, **, and *** indicate significant differences with p ≤ 0.05, p ≤ 0.01, and p ≤ 0.001, respectively; roman numerals refer to neocortical layers).

Figure 3

Clustering of interneurons expressing nNOS mRNA. (A) Ward’s clustering of 42 nNOS interneurons sampled from slices of the barrel cortex of juvenile (P14–P17) mice. Individual cells are represented along the x-axis. The y-axis represents the distance of aggregation in a space of 17 electrophysiological variables and 6 molecular markers. Four clusters, termed Adapt-NPY (cluster 1, yellow), Adapt-SOM (cluster 2, green), non-Ad-PV/SOM (cluster 3, red), and Adapt-VIP/CR (cluster 4, blue) were identified. (B) Table corresponding to Ward’s clustering in (A) and a clustering output generated by the K-means algorithm with the same sample and parameters with non-overlapping components in gray font. Note that the clusters are mostly overlapping. (C) Laminar distribution of cluster 2. Note that the distribution is consistent with the distribution of type I nNOS immunolabeled neurons (Figure 1). (D) Cumulated laminar distribution of clusters 1 (yellow), 3 (red), and 4 (blue). Note that the distribution is consistent with the distribution of type II nNOS immunolabeled neurons (Figure 1). (E) Histograms of the expression of Vglut1, GAD, CB, PV, CR, NPY, VIP, SOM, and CCK in clusters 1, 2, 3, and 4. Note that the expression of SOM mRNA in cluster 2 closely resembles the profile of SOM expression in type I nNOS immunolabeled neurons (Figure 2) while this is not the case for other clusters (green contour on histogram bars; C1: cluster 1; C2 cluster 2; C3: cluster 3; C4: cluster 4; *, **, and *** indicate significant differences with p ≤ 0.05, p ≤ 0.01, and p ≤ 0.001, respectively). (F) Representative examples of electrophysiological properties (left and right panels) and morphology (central panel) from neurons from clusters 1 (yellow, upper left), 2 (green, lower left), 3 (red, lower right), and 4 (blue, upper right). Electrophysiological traces correspond to voltage responses induced by current injections (bottom traces: −100 to 0 pA by increments of 10 pA and rheobase; top trace: last step of current before saturation). For the morphological reconstructions of corresponding neurons, somatodendritic trees are illustrated in black, axons in red. Neurons are represented at their respective laminar positions.

Figure 4

Distributions of GFP-expressing cells among type I and type II nNOS immunolabeled neurons in Lhx6-Cre/R26R-YFP and 5-HT_3A:GFP mice. (A) Colocalization of GFP (1) with nNOS immunolabeling (2) in Lhx6-Cre/R26R-YFP. (B) Colocalization of GFP (1) with nNOS immunolabeling (2) in 5-HT_3A:GFP mice. Arrows point to examples of type I (heavily labeled) nNOS-expressing neurons, arrowheads point to examples of type II nNOS (weakly labeled) expressing neurons. Double-labeled cells are marked by an asterisk. (C) Histograms reflecting the density of GFP-expressing type I (1) and type II (2) nNOS immunolabeled cells in cortical layer I–VI of Lhx6-Cre/R26R-YFP mice (yellow). (D) Histograms reflecting the density of GFP-expressing type I (1) and type II (2) nNOS immunolabeled cells in cortical layer I–VI of 5-HT_3A:GFP mice (blue). In (C) and (D) the densities of nNOS type I (white) and nNOS type II (gray) neurons are superimposed (n = 3 mice; error bars: standard deviation; *, **, and *** indicate significant differences with p ≤ 0.05, p ≤ 0.01, and p ≤ 0.001, respectively; roman numerals refer to neocortical layers).