The neural substrates of insomnia/hyperarousal induced by stress remain unknown. Here, we show that restraint stress leads to hyperarousal associated with strong activation of corticotropin-releasing hormone neurons in the paraventricular nucleus of hypothalamus (CRH^PVN) and hypocretin neurons in the lateral hypothalamus (Hcrt^LH). CRH^PVN neurons directly innervate Hcrt^LH neurons and optogenetic stimulation of LH-projecting CRH^PVN neurons elicits hyperarousal. CRISPR-Cas9-mediated knockdown of the crh gene in CRH^PVN neurons abolishes hyperarousal induced by stimulating LH-projecting CRH^PVN neurons. Genetic ablation of Hcrt neurons or crh gene knockdown significantly counteracts restraint stress-induced hyperarousal. As stress is further associated with changes in immune function, we used single-cell mass cytometry by time of flight (CyTOF) to analyze peripheral blood and found extensive changes to immune cell distribution and functional responses during wakefulness upon optogenetic stimulation of CRH^PVN neurons. Our findings suggest both central and peripheral systems are synergistically engaged in the response to stress via CRH^PVN circuitry.

CRH::Cre mice were equipped with EEG/EMG electrodes as described previously. AAVs encoding the optogenetic actuator ChR2 (AAV-DJ-DIO-EF1α-ChR2-eYFP; Lot # 5227; 4.12 × 10¹² gc/ml) were infused into the paraventricular nucleus (unilateral; 0.4 μl; -0.90 mm AP, ±0.25 mm ML, -4.85 mm DV (injection), -4.6 mm DV (optic fiber)) over the course of 3 min followed by 10 min for viral distribution to the tissue. Then, a fiber optic implant (200 µm diameter, 5.5 mm length, Doric Lenses, Inc.) was slowly lowered and cemented in place above the PVN. Mice were allowed to recover for at least 3 weeks prior to starting the experiment.

Following recovery, 15 ms 10 mW (LaserGlow, calibrated with Thorlabs light meter) 473 nm light pulses at 10 Hz (10 sec ON, 20 sec OFF) were delivered to the PVN for 1 hour, and then whole blood was collected 3 hours following the start of stimulation using microcapillary tubes (retro-orbital bleed; approximately 350 ml blood). Blood was immediately stabilized using a protein stabilization reagent, allowed to sit for 10 minutes, and then placed at -80 ℃ in cryovials until further processing. Red blood cells (RBCs) were lysed using 1× Thaw-lyse buffer (Smart Tube Inc., San Carlos, CA) in ddH₂O. After lysing, cells were pelleted at 600× g for 5 min at room temperature. RBC depleted samples were barcoded using palladium isotopes. In brief, cells were washed with cell staining media (CSM), 1× PBS, and 1× PBS containing 0.02% saponin and then placed into a deep well block. Unique barcoding reagents were added to each well of the block containing cells from each mouse and incubated for 15 minutes at room temperature. After incubation, cells were washed with CSM and then resuspended via brief vortexing. Then, all samples were pooled into a single fluorescence-activated cell sorting (FACS) tube for the next staining step. After washing and pelleting the pooled cells, 150 ml of Fc block was added and cells were incubated at room temperature (RT) with gentle shaking (600 rpm) for 10 minutes. Then, 300 ml of the surface antibody mix (see Table. S1) was added and cells were incubated at RT for 30 minutes with gentle shaking. Then, cells were washed, pelleted, and the supernatant was collected and discarded. To permeabilize cells for intra-cellular staining, 700 ml of MeOH was added to the tube and cells were incubated at 4 ℃ for 10 minutes. Then, cells were washed with 1× PBS and CSM, followed by incubation with the intracellular antibody mix (see Table. S1) for 30 minutes at RT with gentle shaking (600 rpm). Cells were washed and then a DNA intercalator (DVS # Inter-1X-natir) was added in a mixture of 16% paraformaldehyde/PBS. This mixture was incubated at 4 °C overnight.

On the following day, the pooled sample was washed and then 1× normalization “beta” beads were added and cells were strained into a new FACS tube. Cells were run on a Helios (Fluidigm) machine using the SuperSampler attachment. Following the run, the data were normalized to the normalization bead signal. Then, data was run through de-barcoder software (MATLAB; bead distance 0.2) and individual .FCS files were gated and analyzed manually using ImmuneAtlas and Cytobank platforms. tSNE plots were generated for filtered singlet/leukocyte (i.e., red blood cells excluded) populations for each file and formed into a composite plot. In brief, .FCS files were arcsinh transformed (a = 0, b = 1/5, c = 0) and 3000 randomly selected events were sampled for each file. tSNE was computed on the aggregated matrix with all events (3000 × 12 = 36,000). The following parameters were used for tSNE in the RTsne package (https://CRAN.R-project.org/package = Rtsne): perplexity = 30, max_iter = 500. Then, the transformed coordinate events were mapped back to the sample they come from, as well as composite maps for each condition and the total population.

The files are in .fcs (flow cytometry standard) format. The files contain discrete events (each one is a cell or bead) detected by the cytof machine. Each event can be associated with any number of antibodies in the panel, a palladium barcode, and DNA intercalator staining.

Control group files are labelled with CRH_Cre. Stimulated group files are labelled with CRH_Cre_10Hz.