Cortex-restricted disruption of NMDAR1 impairs neuronal patterns in the barrel cortex

Abstract

In the rodent primary somatosensory cortex, the configuration of whiskers and sinus hairs on the snout and of receptor-dense zones on the paws is topographically represented as discrete modules of layer IV granule cells (barrels) and thalamocortical afferent terminals. The role of neural activity, particularly activity mediated by NMDARs (N-methyl-D-aspartate receptors), in patterning of the somatosensory cortex has been a subject of debate. We have generated mice in which deletion of the NMDAR1 (NR1) gene is restricted to excitatory cortical neurons, and here we show that sensory periphery-related patterns develop normally in the brainstem and thalamic somatosensory relay stations of these mice. In the somatosensory cortex, thalamocortical afferents corresponding to large whiskers form patterns and display critical period plasticity, but their patterning is not as distinct as that seen in the cortex of normal mice. Other thalamocortical patterns corresponding to sinus hairs and digits are mostly absent. The cellular aggregates known as barrels and barrel boundaries do not develop even at sites where thalamocortical afferents cluster. Our findings indicate that cortical NMDARs are essential for the aggregation of layer IV cells into barrels and for development of the full complement of thalamocortical patterns.

Figure 1

Generation of Emx1-Cre mice and their cortex-restricted recombination. a, Wild-type Emx1 allele (Emx1⁺), the targeting vector and targeted allele (Emx1^Cre). Filled boxes represent exons. The Cre recombinase gene (Cre) and the pgk-neo gene (Neo) are inserted immediately before the translation initiation codon ATG. B, BglII. b, Southern blot analysis. Genomic DNA was digested with BglII. c, Partial view of whole mount X-gal staining of an E12.5 Emx1-Cre/LacZ embryo. Blue staining (Cre-mediated recombination) is restricted to the dorsal telencephalon (arrowhead). Arrows indicate staining in non-neural tissues. d, X-gal staining of a parasagittal brain section from a P7 Emx1-Cre/LacZ mouse. Olfactory bulb (O), hippocampus (H) and neocortex (Cx) are strongly stained; thalamus (T), brainstem (B), cerebellum (C) and spinal cord (S) are not stained. Scale bar, 800 µm (c) and 1,600 µm (d).

Figure 2

Cortex-restricted NR1 disruption in CxNR1KO mice. a, Western blot analysis of NR1 protein expression in cortex (Cx) and thalamus and brainstem (TB) of flox/− control and CxNR1KO (KO) mice at P1 and P7. b, In situ hybridization of NR1 mRNA in sagittal sections of P7 brains. Arrows indicate cortex. c, In situ hybridization of barrel cortex layer IV at P7. CxNR1KO cortex has a few NR1 expressing neurons (arrows). d, Barrel cortex layer IV of P7 Emx1-Cre/LacZ mouse stained with X-gal (blue), GABA antibody (brown) and cresyl violet (purple). A GABA-containing neuron (arrow) is LacZ negative. Other neurons are LacZ positive. Scale bar, 4 mm (b), 100 µm (c) and 50 µm (d).

Figure 3

Lack of NMDAR-mediated excitation in barrel cortex of CxNR1KO mice. a, Fluorescence image of flox/− and CxNR1KO slices stained with a voltage-sensitive dye. The stimulating electrode was placed in white matter (asterisk). b, Fluorescence signals representing membrane depolarization were obtained from layer II–IV (marked red in a) in ACSF, Mg²⁺-free ACSF, Mg²⁺-free ACSF + APV (50 µM), and Mg²⁺-free ACSF + APV (50 µM) + NBQX (10 µM). Arrowheads indicate time of stimulation. Scales, 50 ms, −0.04% fluorescence change. c, Pseudocolour-coded maps of evoked depolarization 30–50 ms after stimulation (open bars in b). Scale bar, 250 µm (a, c). d, Statistical analysis of evoked responses (mean ± s.e.m.). Asterisk, P < 0.01.

Figure 4

Whisker-related patterns as revealed by cytochrome oxidase histochemistry. a, Cortical patterns for five rows (A–E) of major whiskers are present but not well defined in CxNR1KO mice. b, Subcortical patterns in the ventrobasal thalamus (VB) are indistinguishable between control and CxNR1KO samples. c, Row C whisker lesions between P0 and P3 normally lead to fusion of row C barrels (arrowheads) in the cortex and expansion of neighbouring barrels (left). Similar effects are observed in CxNR1KO mice lesioned at P1 (right). Whisker pad (wp), anterior snout (as), lower jaw (lj) and forepaw (fp) representation areas are shown. Scale bar, 800 µm (a), 500 µm (b) and 450 µm (c).

Figure 5

Partial thalamocortical axonal patterns, absence of barrels and barrel boundaries. a, DiI-labelled thalamocortical axons show patchy distribution in control layer IV, and a similar distribution with less distinct patches in the CxNR1KO cortex. b, Flattened sections stained for 5-HTT antibody also reveal distinct patterns in the control cortex and less distinct ones in the CxNR1KO cortex. In the CxNR1KO cortex, primarily the large whisker patterns are noticeable. c–e, Cytochrome-oxidase-dense patches (asterisks) in flattened cortex (c), and adjacent sections immunostained for tenascin antibody (d). The same sections shown in d were counterstained with cresyl violet (e). f, g, 5-HTT antibody (f) or cresyl violet (g) stained sections spanning equivalent regions of the whisker barrel cortex in the coronal plane. Scale bar, 200 µm (a), 650 µm (b), 250 µm (c–e) and 450 µm (f, g).