All procedures were performed according to the guidelines of the Canadian Council on Animal Care and approved by the animal care committee of McGill University.
Sleep deprivation and recovery experimental procedures
A total number of 12 adult male mice (C57BL/6, 20-25g) were received from the supplier (Charles River) and housed individually under 12-h light : 12-h dark schedule (lights on from 7:00 to 19:00) at 22 °C ambient temperature and with unlimited access to food and water at all time. Animals were maintained in their home cages for the duration of the experiment and recorded by video and telemetric electroencephalogram (EEG) using HomeCageScan software (HomeCageScan™ 3.0; Clever Systems) (del Cid-Pellitero, Plavski and Jones, unpublished results). For telemetric recording of the EEG, two electrodes were placed symmetrically over parietal cortex along with two for reference over cerebellum and were connected by wires to a transmitter (F20-EET, Data Sciences International, DSI) implanted subcutaneously along the flank. Following surgery, the mice were allowed 1 week to recover.
The three experimental groups were composed of: (1) sleep control (SC) mice allowed to sleep undisturbed for 2 h from ~14:00 to ~16:00 (~ZT 7–9) (n = 3), (2) sleep deprived mice (SD) maintained awake for 2 h (n = 3) or 4 h (n = 3) from ~12:00 to ~16:00 (~ZT 5–9) and (3) sleep recovery (SR) mice allowed to sleep for 2 h from ~14:00 to ~16:00 (~ZT 7–9) after being maintained awake for 4 h prior to euthanasia (n = 3). The mice were maintained awake by gentle stimulation with a soft paintbrush of the whiskers each time the mouse appeared to be preparing to sleep. Mice were immediately anaesthetized after the experimental period at ~16:00 (~ZT 9) with sodium pentobarbital (Euthanyl, 100 mg ⁄ kg; Bimeda-MTC) and perfused transcardially with 30 ml of cold saline followed by 200 ml of 3% paraformaldehyde solution. Brains were removed, post-fixed in 3% paraformaldehyde for 1 h at 4°C, then placed in 30% sucrose solution at 4 °C for 2 days, frozen to −50 °C and stored at −80 °C.
Sleep and waking were scored by behavior and EEG using HomeCageScan software.
Immunohistochemistry
Brains were cut and processed for fluorescent staining in batches of 2–4 that included mice from SC, SD and/or SR groups of the same experimental session or period. Coronal sections were cut through the brainstem on a freezing microtome at a 20 µm thickness and collected in 5 adjacent series, such that sections were separated by 100 µm intervals in each series. Free floating sections were rinsed in 0.1 M trizma saline buffer (pH 7.4), then incubated in 6% normal donkey serum buffer for 30 min and subsequently incubated overnight at room temperature in a buffer containing 1% normal donkey serum with one of the primary antibodies. The following antibodies were employed: mouse anti-GABAAR β2-3-chain (clone BD17, 1:100, Millipore (Chemicon), CAT# MAB 341, RRID: AB_2109419), guinea pig anti-GABABR1 (1:2500, Millipore (Chemicon), CAT# AB1531, RRID: AB_2314472) or rabbit anti-AChM2R (1:600, Sigma, CAT# M9558, RRID: AB_260727). Subsequently, sections were incubated at room temperature for 2 h in Cyanine-conjugated (Cy3) secondary antibodies from donkey (Jackson ImmunoResearch Laboratories): Cy3-conjugated anti-mouse (1:1000, CAT# 715-165-150, RRID: AB_2340813), Cy3-conjugated anti-guinea pig (1:1000, CAT# 706-165-148, RRID: AB_2340460) or Cy3-conjugated anti-rabbit (1:1000, CAT# 711-165-152, RRID: AB_2307443). Sections were subsequently stained with green fluorescent Nissl stain (FNS) (1:2000, CAT # N-21,480, Molecular Probes) for 20 min. Finally, sections were rinsed, mounted and coverslipped with glycerol (Fisher).
All the receptor antibodies employed were produced and characterized years ago and have since been in use over many years (as cited herewith). For the GABA
AR, the antibody against the β2-3-chain was employed because it stains the most prevalent types of GABA
A heterodimeric receptors on neurons in the brain and stains clusters of the GABA
AR on the plasma membrane, which are associated with functional inhibitory post-synaptic currents (IPSCs) (Fritschy and Mohler
1995; Wan et al.
1997; Nusser et al.
1998). For the GABA
BR, the antibody against the R1 subunit was employed because it also stains vast populations of neurons in the brain and is visible over cytoplasmic organelles and the plasma membrane, where it forms together with the R2 subunit the heteromeric functional receptor (Margeta-Mitrovic et al.
1999; Filippov et al.
2000; Straessle et al.
2003). For the AChM2R, the antibody employed was shown to be highly specific and to stain most prominently the surface membranes of all cholinergic as well as diverse noncholinergic neurons in the brain (Levey et al.
1991).
Immunohistochemical image analysis
Stained sections were viewed using a Leica DMLB microscope equipped with x/y/z motorized stage, a digital camera (Orca-R2, C10600-10B, Hamamatsu photonics K.K) and fluorescence filters for excitation and emission of Cy2 and Cy3 dyes. Images were acquired and analyzed using the Optical Fractionator Probe of StereoInvestigator (MicroBrightField, MBF), which allows unbiased, systematic random sampling of a region of interest for cell number estimation or measurement of specific parameters, including luminance. Given application of systematic random sampling for examining and marking cells at high magnification and the subsequent measurement of fluorescence intensity of the receptor staining in the marked receptor + cells employed here, double blind procedures were not applied in this process. In each series, three sections (at 100 µm intervals) were taken through the Mo5 nucleus. In each section, a contour was traced around the Mo5 nucleus under a 5× objective. Multi-channel image stacks with 0.5 μm thickness of optical sections were then acquired under a 40× objective through the mounted histological section of approximately 15 µm thickness. For the image acquisition, the exposure time and contrast were set for each receptor series according to the parameters which provided suitable images of the brightest to the dimmest fluorescence. These parameters were maintained for all sections and mice across all groups for each receptor series. In the Optical Fractionator Probe, a grid size of 200 × 200 μm2 and a counting frame of 120 × 120 μm2 were used in the image acquisition and assessment with marking of positively labeled cells. Across the three sections, approximately 21 counting frames for Mo5 neurons were acquired and analyzed per series. Within these images, cell somata with visible nuclei located > 1 µm below the surface of the section were marked for counting, thus through 14 µm of the section. The motor neurons were identified in the FNS stained sections by their distinct morphology and thus referred to as MoFNS-positive (+). The average number of MoFNS + cells counted across series on one side was 32.86 ± 1.16 (mean ± SEM). For marking of cells positively stained for receptors, the immunostaining for the GABAA, GABAB and AChM2 receptors on the membrane or over the cytoplasm of the MoFNS + somata was assessed by moving through the z stack of 0.5 μm thick optical sections of each MoFNS + cell soma. Estimated total numbers of double-labeled cells were computed for each series (GABAAR-FNS, GABABR-FNS, and AChM2R-FNS) and expressed as % of MoFNS + cell population per series through the Mo5 nucleus.
Luminance measurements of the receptor immunofluorescence were performed in the marked double-labeled cells, comprising 5–10 cells per animal for GABA
AR+ and 10 cells per animal for GABA
BR+ and AChM2R+. The images had been acquired with the 8-bit setting of the digital camera, which thus provides converted gray scale images of the fluorescence with arbitrary units of 0–256 for luminance measures. Image acquisition was made as rapidly as possible for each cell so as to avoid bleaching of the fluorescence. As it was previously described in (Toossi et al.
2016b), two approaches were used to measure the intensity of receptor immunofluorescence over the membrane vs. that over the cytoplasm plus membrane of the soma. A rectangular box sized at 1.5 × 0.3 μm
2 was placed over the plasma membrane for luminance measurement of the membrane fluorescence and another box over the nucleus for luminance measurement and subsequent subtraction of the background fluorescence in each cell. A donut-shaped contour was drawn around the perikaryon to include the cytoplasm and plasma membrane for luminance measurement of the cytoplasm plus membrane and another spherical contour drawn around the nucleus for measurement and subtraction of background fluorescence in each cell.
Cell counts and luminance measurements were analyzed across experimental groups for each receptor series (GABAA, GABAB or AChM2R) using one-way analysis of variance (ANOVA) followed by post-hoc paired comparisons with Fisher’s LSD using SYSTAT (SYSTAT Software Inc., version13). In an initial analysis, the proportion of GABAAR+/MoFNS + neurons was found to differ significantly across the original 4 groups (F(3,8) = 4.27, p = 0.045), however to not differ significantly between the SD2 and SD4 groups (post-hoc paired comparisons, p = 0.387), for which reason they were subsequently combined into one SD group for subsequent analysis and presentation of results. In addition, the proportion of GABABR+ and AChM2R + of the MoFNS + neurons was 100% in all groups.
For higher resolution, images were also acquired using an LSM 710 confocal laser scanning microscope equipped with Ar 488-nm and He–Ne 543-nm lasers for excitation and emission of Cy2 and Cy3 dyes. Images were acquired under 63× oil objectives with a 1.0 airy unit pinhole size for each channel and 0.5 μm thick optical section.
Image plates were prepared and composed using Adobe Creative Suite (CS4, Adobe System) from fluorescent microscopic images which were used for quantitative measurements and from confocal images which were used for qualitative assessment with higher resolution of the receptor immunostaining. In all cases as stated above, the parameters of acquisition including exposure time and contrast, were set at the beginning for each receptor series and maintained the same across sections, mice and groups. In no case were brightness or contrast adjusted on individual images. However, for the fluorescent microscopic images of the GABAAR immunofluorescence, which was quite dim in the SC brains, the brightness and contrast were enhanced uniformly across the three images of the three mice for better visibility of the receptor immunostaining in the image plate.