Animals and EAE induction
C57BL/6 J mice were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and housed in the animal care facility for 1–2 weeks before induction of EAE at age 8–10 weeks, following a protocol approved by the PBRC Institutional Animal Care and Use Committee. Females were used because they are more susceptible to EAE. All mice were housed in groups of four in round acrylic sleep recording cages (Pinnacle Technology, Lawrence, KS, USA) and adapted for 3 days before initiation of SF or control experiments.
EAE induction was similar to that reported recently [
7,
17,
18], with minor modifications. In brief, 80 or 100 μg of myelin oligodendrocyte glycoprotein (MOG) fragment 35–55 (MOG
35–55) was fully emulsified in 100 μl of complete Freund’s adjuvant (CFA) containing 500 μg of heat-killed
Mycobacterium tuberculosis H37RA (DIFCO Laboratories, Detroit, MI, USA). The mixture was delivered subcutaneously, divided among three flank areas. Pertussis toxin (Sigma, St. Louis, MO, USA) was injected intraperitoneally immediately after induction and then again 48 h later. The time of induction was considered EAE day 0. Symptoms were monitored daily at about noon (zeitgeber time, ZT6) by use of a standard EAE scoring system [
11,
19,
20], with 0 being symptom-free and 5 being the worst (moribund or dead).
Experimental SF maneuver
Three groups of mice were used (n = 8 /group): (1) EAE mice with SF during the light phase from day −10 to day +16 in relation to EAE induction (on day 0). SF was applied between zeitgeber time (ZT) 0–12; (2) resting EAE mice without SF intervention; (3) naïve controls with neither EAE nor SF. The dose of MOG35–55 was 100 μg/mouse. The mice were sacrificed on day 16 after EAE induction, immediately after cessation of SF.
In a second study to determine the progression of EAE symptoms, 2 groups of mice were studied (n = 8/group): SF EAE mice and resting EAE mice, as for first group above. In this case, EAE scores and body weight were monitored from day 0 till day 28 after EAE induction. The dose of MOG35–55 was reduced to 80 μg/mouse: by reducing the dose of the heptagen, we aimed to achieve a lower EAE score in this batch of mice. This would ensure that mice were not too incapacitated to have the SF procedure and that they could be monitored through the course of EAE to 28 days. This would increase the probability of identifying the effect of SF, as we hypothesized that SF worsens EAE.
The SF maneuver agitated the mice every 2 min by use of random bar rotation driven by a computer program. The mice were group-housed in round cages (3 – 4 mice/cage). At the bottom of the cage is a metal bar slightly shorter than the inner diameter of the round cage and positioned above the corncob bedding. A computer-controlled SF schedule repeats on a 120 s cycle (30 s on, 90 s off) during the light span (6 am to 6 pm, ZT0-12), with an intermediate bar rotation speed (scale of 5 out of 10). The direction of bar rotation is randomly reversed every 10–40 sec. The choice of 30 events/h of SF used was based on clinical evidence from severe sleep apnea, and shown to be effective in compromising sleep architecture in mouse and rat [
21,
22]. The effective reduction of NREM sleep and increased sleep fragmentation with this SF maneuver has been validated by sleep recording and sleep architectural analysis, as we have reported recently [
8]. In brief, SF increases Wake time by about 25% in the 12-h light span, reduces % NREM by about 20%, but decreases REM sleep only during the first night. The changes are consistent across the 10 d of SF despite the reduction of REM sleep for the first night. By contrast, control C57 mice have about 39% Wake, 56% NREM, and 5% REM. The SF maneuver also increases sleep state transitions: NREM bouts increase 2–3 fold, as do Wake bouts. In the dark span, the percent of Wake, NREM, and REM show no significant changes, and there is no increase of sleep fragmentation. This indicates a lack of circadian shift resulting from the SF maneuver. Recovery from SF is rapid, as sleep architecture returns to baseline within 24 h after the cessation of SF [
8].
Flow cytometry
Three groups of mice were studied (n = 8 /group): naïve, resting EAE, and SF EAE where SF was applied 10 days before EAE induction and terminated on day 16 of EAE, immediately before sacrifice of the mice in the afternoon (ZT6-9). To isolate leukocytes from the spinal cord and spleen following an established protocol [
23], mice were anesthetized by intraperitoneal injection of urethane (30 mg/kg). They were perfused intracardially with phosphate-buffered saline (PBS) to remove leukocytes in the spinal cord vasculature. The spinal cord was homogenized in PBS containing 0.1% fetal bovine serum. Leukocytes were recovered at the 30:70% Percoll interface after gradient centrifugation as described previously [
17]. To obtain splenocytes, spleens were ruptured, homogenized, filtered through a 40 μm nylon mesh, centrifuged, and subjected to red blood cell lysis. Cells from both tissues were washed with FACS buffer twice, immunostained, fixed with 2% paraformaldehyde (PFA), and stored in PBS until flow cytometric analysis.
All antibodies were purchased from BioLegend (San Diego, CA, USA). For cell surface staining of immune markers, leukocytes were blocked with anti-mouse CD16/32 antibody (Clone 93, catalog number 101302), incubated at 4°C with FITC-conjugated CD4 (Clone GK1.5, catalog number 100405), PE-conjugated CD8a (Clone 53–6.7, catalog number 100707), Alexa488-conjugated CD11b (Clone M1/70, catalog number 101219), APC-conjugated Gr1 (Clone RB6-8C5, catalog number 108411), or APC-conjugated CD45 antibody (Clone 30-F11, catalog number 103112). Control conditions included single staining for antibodies and non-stained cells to exclude autofluorescence. After staining, the cells were fixed with 2% PFA and stored in PBS for 1–3 days. Cell numbers were estimated by use of a hemocytometer, and the immunofluorescent intensity was analyzed by FACSCalibur (BD Pharmingen, San Diego, CA, USA). Data were analyzed with post-collection compensation by FlowJo (Tree Star, Ashland, OR, USA) software.
The gating strategy was as follows: Cells were first gated based on side scatter of CD45 immune positive cells, which differentiates plasma cells, non-lysed erythrocytes, and their precursors from the CD45+ leukocytes. Blast cells and other hematogenous cells with low granularity and low CD45 intensity were located in the left lower quadrant of the side scatter plot of the CD45+ population, whereas lymphocytes, which show high CD45 intensity and low granularity, were gated in the right lower quadrant. Granulocytes, including neutrophils of relatively lower CD45 and eosinophils with higher CD45 immunofluorescence, were gated in the upper right quadrant of the side scatter plot. Monocytes with high CD45 and intermediate granularity were separated from the small number of basophils that had lower granularity. The number of lymphocytes was then used as the denominator to determine the percentage of CD4 and CD8 T cells in double immunostaining. The percentage of CD4 and CD8 cells among all leukocytes was also determined. Although some macrophages also express CD4 and CD45 and some dendritic cells also express CD8 and CD45 besides lymphocytes, these two populations were excluded with gating for the final analysis. For Gr1 and CD11b double labeling, there were four clearly defined quadrants representing Gr1+CD11b+ myeloid-derived suppressor cells, Gr1−CD11b+ macrophages, monocytes and dendritic cells.
Statistical analysis
Means were expressed with their standard errors. Repeated measures analysis of variance (ANOVA) was used to determine the effect of EAE and SF on EAE scores. One-way ANOVA was used to determine the effect of SF and EAE on spleen weight and leukocyte populations, followed by Tukey’s post-hoc test. Student’s t-test (2-tailed) was used when only two groups were present for comparison. Prism GraphPad 5 statistical and graphic program (GraphPad, San Diego, CA, USA) was used for statistics and graphics.