The supramammillary nucleus and the claustrum activate the cortex during REM sleep

Abstract

Evidence in humans suggests that limbic cortices are more active during rapid eye movement (REM or paradoxical) sleep than during waking, a phenomenon fitting with the presence of vivid dreaming during this state. In that context, it seemed essential to determine which populations of cortical neurons are activated during REM sleep. Our aim in the present study is to fill this gap by combining gene expression analysis, functional neuroanatomy, and neurochemical lesions in rats. We find in rats that, during REM sleep hypersomnia compared to control and REM sleep deprivation, the dentate gyrus, claustrum, cortical amygdaloid nucleus, and medial entorhinal and retrosplenial cortices are the only cortical structures containing neurons with an increased expression of Bdnf, FOS, and ARC, known markers of activation and/or synaptic plasticity. Further, the dentate gyrus is the only cortical structure containing more FOS-labeled neurons during REM sleep hypersomnia than during waking. Combining FOS staining, retrograde labeling, and neurochemical lesion, we then provide evidence that FOS overexpression occurring in the cortex during REM sleep hypersomnia is due to projections from the supramammillary nucleus and the claustrum. Our results strongly suggest that only a subset of cortical and hippocampal neurons are activated and display plasticity during REM sleep by means of ascending projections from the claustrum and the supramammillary nucleus. Our results pave the way for future studies to identify the function of REM sleep with regard to dreaming and emotional memory processing.

Keywords: REM sleep; dentate gyrus; hippocampus; immediate early genes; supramammillary nucleus.

Fig. 1. Expression level of plasticity-related genes and REM sleep amounts as quantified in the microarrays.

(A to G) Histograms showing Bdnf, Arc, Fos, and Cox2 expression levels in the hippocampus (A to D) and the cortex (E to G) in RSC (blue), RSD (red), and RSR (green) rats. For all conditions, the bar on the left shows the mean (±SEM) expression level of each gene measured by qPCR relative to the RSC condition (six rats per group). The other two bars show the expression level relative to the RSC condition measured by the two microarrays in two independent groups (six rats per condition). Kruskal-Wallis followed by Mann-Whitney, RSD versus RSC or RSR versus RSC: **P < 0.01, *P < 0.05; RSD versus RSR: ^##P < 0.01, ^#P < 0.05. The graphs show the correlations between the REM sleep amounts (in minutes) during the last 6 hours of the protocol and the expression levels of Bdnf, Arc, Fos, and Cox2 measured by qPCR (six rats per condition; Spearman test: *P < 0.05, **P < 0.01). (H) Histogram showing the vigilance state amount in minutes during the last 6 hours of the protocol (H) in RSC, RSD, and RSR animals (12 rats per condition). Kruskal-Wallis followed by Mann-Whitney, RSD versus RSR versus RSC: ***P < 0.0001, **P < 0.001, *P < 0.05; RSR versus RSD: ^###P < 0.0001, ^##P < 0.001, ^#P < 0.5. Bdnf (A), Arc (B), Fos (C), and Cox2 (D) mRNA expression levels are significantly increased in the hippocampus in RSR rats compared to RSC and RSD rats, and Bdnf, Arc, and Cox2 expression is positively correlated with REM sleep quantities. In the cortex, Bdnf (E) mRNA expression is increased in RSR rats compared to the other two conditions and is positively correlated with REM sleep amounts, whereas Arc (F) and Fos (G) mRNA levels are increased after REM sleep deprivation and Arc expression level is negatively correlated with REM sleep amounts.

Fig. 2. The expression of plasticity-related markers is increased in the granule layer of the dentate gyrus during REM sleep hypersomnia.

(A to L) Photomicrographs and histograms showing the in situ hybridization of Bdnf (A to C) and the immunohistochemistry of ARC (D to F), FOS (G to I), and COX2 (J to L) in the dentate gyrus (DG), CA3, and CA1 in RSD (A, D, G, and J) and RSR (B, E, H, and K) rats. The photomicrographs show that the numbers of Bdnf-, ARC-, FOS-, and COX2-stained neurons are increased in the dentate gyrus in RSR rats compared to RSD rats. The histograms also show an increase of Bdnf and COX2 staining in CA3. In contrast, there is no change in CA1 among conditions for any marker. Scale bar, 50 μm. Kruskal-Wallis followed by Mann-Whitney: **P < 0.02, *P < 0.05, RSD versus RSC or RSC versus RSR; ^##P < 0.02, ^#P < 0.05, RSD versus RSR.

Fig. 3. The expression of plasticity-related markers is increased in a subset of cortical neurons during REM sleep hypersomnia.

(A to L) Photomicrographs and histograms showing the in situ hybridization of Bdnf (A to D) and the immunostaining of ARC (E to H) and FOS (I to K) in the anterior cingulate (ACA) (A, E, and I) and retrosplenial cortices (RSP) (B, F, and J) and in the claustrum (CLA) (C, G, and K) of RSD and RSR representative rats. In (D), the histograms show the mean intensity (±SEM) of Bdnf labeling in six cortical areas in the RSD and RSR groups relative to the RSC group. The other two histograms show the mean number (±SEM) of ARC⁺ (H) and FOS⁺ (L) cells in the same cortical areas and same experimental conditions. Bdnf labeling (A to D) is significantly increased in RSR rats compared to the other two groups in the anterior cingulate (A) and the medial entorhinal and retrosplenial (B) cortices, the claustrum (C), and the cortical amygdaloid nucleus. The retrosplenial cortex (F) and the claustrum (G) contained a significantly higher number of ARC⁺ cells not only in RSR but also in RSD rats compared to the RSC group. (H) The cortical amygdaloid nucleus and, to a minor extent, the medial entorhinal cortex contained a significantly increased number of ARC⁺ neurons only during REM sleep hypersomnia compared to the control group. The auditory cortex contained a significantly higher number of ARC⁺ cells in RSD rats. The number of FOS⁺ cells is increased in the medial entorhinal and retrosplenial cortices (J), the claustrum (K), and the cortical amygdaloid nucleus in the RSR group compared to the other two groups. Scale bars, 200 μm (A, B, E, F, I, and J); 25 μm (C, G, and K). Kruskal-Wallis followed by Mann-Whitney: **P < 0.02, *P < 0.05, RSD versus RSC or RSC versus RSR; ^##P < 0.02, ^#P < 0.05, RSD versus RSR (see also fig. S2 and table S3 for the complete quantification).

Fig. 4. A restricted number of cortical structures are active during REM sleep hypersomnia compared to W.

(A to C) Photomicrographs showing FOS immunostaining in the hippocampus (A), auditory cortex (B), and cortical amygdaloid nucleus (C) of RSR and W rats. Scale bars, 200 μm. (D and E) The histograms show the mean number (±SEM) of FOS⁺ cells in the hippocampal formation and in six cortical areas (n = 4 rats per condition). (F) The drawings illustrate the distribution of FOS⁺ cells on coronal sections located at −7.10 mm relative to bregma in RSR and W rats. The dentate gyrus (DG) (A and D), medial entorhinal cortex (ENTm) (E and F), claustrum (CLA), and cortical amygdaloid nucleus (COA) (C and E) are the only structures in which the number of FOS⁺ cells is similar in RSR and W rats. The dorsal CA1, known to contain place cells, displays a large number of FOS-labeled cells only after W (A and D). The auditory (AUD) (B and E) and visual (VIS) (F) cortices contain a very large number of FOS-labeled neurons in W and, in contrast, only a small number of FOS-labeled cells during REM sleep recovery. The medial entorhinal and retrosplenial (RSP) cortices (E and F) and the anterior cingulate (ACA) (E) cortices are activated during REM sleep hypersomnia, but to a lower extent than during W. Scale bars, 200 μm (A to C). Kruskal-Wallis followed by Mann-Whitney: **P < 0.02, *P < 0.05, RSR versus W (see also fig. S2 and table S4). ECT, ectorhinal area; ENTl, entorhinal area, lateral part; PAR, parasubiculum; POST, postsubiculum; PRE, presubiculum; SUB, subiculum; TEA, temporal association areas.

Fig. 5. The dentate gyrus and the anterior cingulate and retrosplenial cortices receive projections from the SUML and the claustrum active during REM sleep hypersomnia.

(A to I) Photomicrographs showing double-labeled cells in RSR rats with a black FOS-labeled nucleus and a brown cytoplasm (retrogradely labeled). Scale bars, 200 μm; 25 μm (inset). (J) The histogram shows the mean number (±SEM) of FOS/CTb⁺ or FOS/FG⁺ cells contained in the major areas projecting to the dentate gyrus (DG), anterior cingulate cortex (ACA), and retrosplenial cortex (RSP) (four rats per structure, three sections quantified for the SUM; details given in tables S4 and S5). After CTb injection in the dentate gyrus (A), the SUML (B) is the only region containing FOS/CTb⁺ cells; the large number of CTb⁺ neurons in the medial septum (C) does not express the FOS protein. After FG injection in the anterior cingulate cortex (D), a large number of double-labeled cells are seen in the claustrum (CLA) (F and J), a small number are observed in the SUML (E and J), and only a few in the thalamic nuclei (TH) (H and J). After CTb injection in the retrosplenial cortex (G), FOS/CTb⁺ cells are observed only in the claustrum (I and J). The large number of CTb⁺ cells located in the thalamic nuclei (H) does not express the FOS protein (J). (K) A large number of anterogradely labeled fibers (in brown) are seen in close proximity to FOS⁺ neurons (in black) in the dentate gyrus of an RSR rat after a CTb injection in the SUML (inset). Scale bars, 50 μm; 500 μm (inset). AC, anterior commissure; CP, caudate putamen; f, fornix; PAG, periaqueductal gray; pm, principal mammillary tract; VIS, visual cortex.

Fig. 6. Lesion of the SUM alters dentate gyrus activation during REM sleep hypersomnia.

(A) Series of schematic transverse sections modified from the Swanson atlas depicting the extent of each lesion of the SUM (one color per animal). (B) Illustration of a representative neurotoxic lesion of the SUML in a section immunostained with NeuN. (C) Histogram illustrating the quantities of each vigilance state in lesioned and sham rats (n = 4 per group) during the 6 hours preceding euthanasia (Kruskal-Wallis followed by Mann-Whitney). (D) Graph showing the mean spectral power of the EEG during REM sleep hypersomnia in the lesioned rats (±SEM) in relative value compared to the sham rats. The horizontal black points and lines below the graph show the frequency bins significantly different between the two conditions according to the Kruskal-Wallis and Mann-Whitney tests (P < 0.05). (E and F) Photomicrographs and histogram showing that the lesion of the SUML induces a strong decrease in the number of FOS⁺ neurons in the dentate gyrus compared to sham. Scale bars, 50 μm. Kruskal-Wallis followed by Mann-Whitney: **P < 0.02, *P < 0.05, RSR-sham versus RSR-lesion (see also fig. S3 and table S7).

Fig. 7. Pathways and structures activated during REM sleep according to FOS imaging and retrograde tracing studies.

The solid line shows the pathway demonstrated by the retrograde tracing studies and confirmed by the lesion of the SUML. The dashed lines show the other pathways revealed by retrograde tracing. ACA, anterior cingulate cortex; CLA, claustrum; DG, dentate gyrus; RSP, retrosplenial cortex.