Animals

Sprague–Dawley rats were from SLAC Laboratory Animal Co., Ltd. (Shanghai). The Vglut2-iCre line was customized by Biocytogen Co., Ltd (Beijing). Rats were housed in pairs and maintained on a 12 h/12 h light/dark schedule at a room temperature of 22–25 °C with water and food ad libitum unless otherwise noted. All experimental procedures were approved by the Institutional Animal Care and Use Committee of the Center for Excellence in Brain Science & Intelligence Technology, Chinese Academy of Sciences, Shanghai, China.

Virus Injection and Stereotaxic Surgery

Male (~8 weeks) and female (~10 weeks) rats were anesthetized using 2.5% isoflurane delivered through an R540 animal anesthesia machine (RWD, Shenzhen, China). Viral vectors were precisely introduced into targeted brain nuclei using a stereotaxic apparatus (Stoelting Co., USA). The coordinates for craniotomy—anteroposterior (AP) from bregma, mediolateral (ML) from the midline, and dorsoventral (DV) from the brain’s surface—were determined based on the sixth edition of Paxinos and Watson’s "The Rat Brain in Stereotaxic Coordinates." Adeno-associated viruses (AAVs) were sourced from Taitool Biotech Co. (Shanghai, China), while the rabies virus was obtained from BrainVTA Co. (Wuhan, China). Throughout the surgical procedures, the rat’s body temperature was regulated with a heating blanket. In addition, to mitigate the risk of post-surgical infection, the animals were given gentamicin (5 mg/kg, intraperitoneally [i.p.]) and dexamethasone (1 mg/kg, i.p.). A minimum three-week postoperative period was observed to ensure viral expression before subsequent experiments unless otherwise noted.

For fiber photometry recordings, we injected 300 nL of AAV2/8-hSyn-FLEX-GCaMP6s (5.12 × 1012 VG/mL) into the PPN (AP 7.6, ML ±1.8, DV 6.4) at 20 nL/min using a UMP3 pump coupled with a SYS-Micro4 controller (World Precision Instruments, USA). To prevent viral leakage, the glass pipette remained in place for 8 min before injection and retraction. We then implanted optical fibers (200 μm OD, 0.37 NA) 0.2 mm over the viral injection site.

For optogenetic stimulation, rats received 300 nL of either AAV2/8-hEF1α-DIO-hChR2-mCherry (3.04 × 1012 VG/mL) or AAV2/8-hEF1α-DIO-mCherry (2.96 × 1012 VG/mL) into the PPN at the same coordinates. Optical fibers (200 μm OD, 0.37 NA) were implanted 0.5 mm over the virus injection site.

For optogenetic stimulation and extracellular recording, rats were injected with 300 nL of AAV2/8-hEF1α-DIO-hChR2-mCherry (3.04 × 1012 VG/mL) into the PPN. Following a 2-week expression period, custom optrodes were implanted. Notably, we added a stainless-steel sleeve to guide the electrode interface board and threaded the ground wire (132 μm stainless steel) through protective tubing. Prior to implantation, tetrodes were gold-plated to adjust impedance to 300–500 kΩ (1 kHz, IMP-2, Bak Electronics Inc., USA). Animals were given a minimum of 7 days to recover from optrode implantation surgery before commencing recordings.

For optogenetic stimulation of PPN vGluT2 neuron axon terminals, we bilaterally injected 300 nL of AAV2/8-hEF1α-DIO-ChR2-mCherry into the PPN and implanted optical fibers 0.2 mm over the terminal nuclei of several brain regions, including the globus pallidus externa (GPe; AP 1.2, ML ± 3.4, DV-5 6), subthalamic nucleus (STN; AP 3.6, ML ± 2.5, DV 7.0), substantia nigra pars reticularis (SNr; (AP 5.4, ML ± 2.8, DV 6.5), PnC/GiA (AP-10.32, ML ± 0.6, DV-8.4), and ZI (AP-2.6, ML ± 1.9, DV-6.8).

To trace anterograde outputs from PPN-vGluT2 neurons, we administered 200 nL of AAV2/5-hSyn-FLEX-tdTomato-T2A-Synaptophysin-EGFP-WPRE-pA (2.34 × 10^12 VG/mL) into the PPN. After a 3-week expression period, rats were euthanized and prepared for immunohistochemistry.

For monosynaptic retrograde tracing inputs of PPN vGluT2 neurons, vGluT2-iCre rats received 200 nL of a helper virus, AAV2/9-hSyn-FLEX-mCherry-2A-TVA-2A-RvG-WPRE-pA (2.29 × 10^12 VG/mL), injected into the PPN. Two weeks later, 150 nL of RV-ENVA-ΔG-EGFP (3 × 108 IFU/mL) rabies virus was administered at the same site. Rats were euthanized for brain sample collection one week post-injection of the rabies virus.

To selectively stimulate PPN-vGluT2 neurons projecting to the PnC/GiA or ZI, we bilaterally administered 400 nL of AAV2/2Retro-plus-Syn-FLEX-flop (5.03 × 1012 VG/mL) into the target region and 300 nL of AAV2/9-hEF1α-fDIO-hChR2-mCherry into the PPN. Optical fibers were implanted 0.5 mm over the PPN virus injection site.

Finally, to dissect the distribution of PPN-vGluT2 neurons projecting to the PnC/GiA and ZI, we respectively injected AAV2/2Retro-EF1α-DIO-H2B-tdTomato and AAV2/2Retro-EF1α-DIO-H2B-EGFP into these regions. Rats were anesthetized and euthanized for brain sample collection after a 3-week expression period.

Fibre Photometry Recording, Behavioral Assays, and Analysis

For photometry in free-moving animals, optical fibers were interfaced with a triple-color, multi-channel fiber photometry system (ThinkerTech) with 405-nm and 470-nm LED sources. The emission intensity of the 405-nm excitation at the fiber’s tip was finely tuned to ~20 µW, and the 470-nm excitation was adjusted to ~30 µW. Both emission signals and behavioral videos were concurrently recorded at a sampling rate of 30 Hz and time-stamped with the computer’s real-time clock to facilitate synchronization of the Ca2+ signals with corresponding animal behavior in subsequent data analysis. To track velocity and head orientation, the behavioral videos were processed and quantified using the graphical user interface of DeepLabCut.

For data analysis, Ca2+ signaling was quantified using custom-written MatLab scripts. Initially, photobleaching artifacts were corrected in the raw fluorescence data acquired from both 470-nm and 405-nm channels. Subsequently, the 470-nm channel data underwent a secondary correction using the 405-nm channel as a reference. The relative changes in the Ca2+ signal (ΔF/F) were then computed using the formula (F - F0)/F0, where F represents the fluorescence intensity at a given time point and F0 denotes the baseline fluorescence intensity.

To elucidate the relationship between ΔF/F signals and both velocity and head-turning, we conducted a cross-correlation (xcorr) analysis. To assess the correlation between the Ca2+ signal and velocity, we calculated the z-scores for both the Ca2+ signal and the velocity using a bin size of 100 ms, applying the xcorr function in MatLab (M2021a). To identify head-turning events in rats, we tracked their head movements using DeepLabCut and defined an event as a head turn with an angular velocity >0.01 rad/s. We further categorized each event within a series of binned time windows: '1' for time windows containing a left or right head-turning event and '0' for time windows without any head-turning event. This binary labeling facilitated the identification of head-turning event data. In a similar manner, we calculated the z-scores for both the Ca2+ signal and the left or right head-turning events. This methodical approach ensured a systematic investigation of the temporal associations between neural activity, as reflected by ΔF/F signals, and the corresponding behavioral responses.

To clarify the relationship between the activity of rPPN-vGluT2 neurons and salient stimuli, we recorded fiber photometry in response to tactile stimulation by hand or abrupt auditory stimulation. Rats were allowed to move freely within a clear enclosure measuring 40 cm × 30 cm × 60 cm (LWD). For tactile stimulation, a gloved hand touched a rat through the open top of the chamber at intervals of 2 to 3 min, with immediate withdrawal after contact. For auditory stimulation, a 100-ms burst of sound at 100 dB was delivered once per minute. Peri-event time histograms of the Ca2+ signal were generated and analyzed in synchronization with the onset of the salient stimuli using MatLab (M2021a).

Optogenetic Stimulation, Behavioral Assays, and Analysis

For locomotor speed analysis, rats were allowed to move in a matte black open-field chamber (160 cm × 40 cm × 60 cm, LWD). Optogenetic stimulation and recording of a rat’s position and instantaneous velocity were controlled and synchronized by an Anilab instrument and software (Anilab Instrument Co, Ningbo, China). For each stimulation condition, a rat underwent 30−50 repeated trials with a 15−30 s intertrial interval. The width of the laser pulse (473 nm) was 20 ms, and the laser power at the tip of the optic fiber was ~15 mW, as measured by a light intensity meter (Thorlabs, Inc., USA) in all optogenetic experiments. We eliminated trials with speeds <3 cm/s in 500 ms before laser onset, analyzed the speed in 100-ms bins and then calculated the mean velocity changes before, during, and after the stimulus timepoint for all active trials.

For locomotion and motion analysis, rats freely moved in a transparent chamber (40 cm × 30 cm × 60 cm, LWD). Optogenetic stimulation was delivered by a 473-nm laser controlled by Anilab software. The locomotion and motion of rats were recorded by an industrial camera. Behaviors including but not limited to walking, grooming, eating, rearing, balancing on a rod, and swimming were recorded and analyzed.

For respiration rate analysis, anesthetized rats were head-fixed. An air pressure sensor was placed near one of the rat’s nostrils, not touching the nose itself, and the signals from rhythmic nasal breathing were recorded as a measure of respiration. Rats underwent 90 repeated trials with 4 s of optogenetic stimulation and a 26-s intertrial interval. The average normalized respiration rates of all rats were plotted against the time of optogenetic stimulation.

Brain Slice Recording

Rats injected with AAV2/8-EF1a-DIO-ChR2-mCherry virus were first tested to confirm that their locomotion was halted during optogenetic stimulation; they were then used for brain slice recording. Briefly, rats were anesthetized and perfused transcardially with ice-cold, oxygenated (95% O2 and 5% CO2) cutting solution (~80 mL) containing the following (in mmol/L): 105 N-methyl-D-glutamine, 105 NaCl, 2.5 KCl, 1.2 NaH2PO4, 26 NaHCO3, 25 glucose, 10 MgSO4, 0.5 CaCl2, 5 L-ascorbic acid, 3 sodium pyruvate, and 2 thiourea (pH 7.4, 295–305 mOsm). The brain was removed quickly and placed into an ice-cold, oxygenated cutting solution. Coronal slices (350 μm) were prepared using a vibratome (1200 s, Lecia, Germany) and then transferred to an incubation chamber at 32 °C with an oxygenated cutting solution and incubated for 12 min. Then, the slices were transferred into and incubated in artificial cerebrospinal fluid (aCSF) that contained the following (in mmol/L): 119 NaCl, 2.3 KCl, 1.0 NaH2PO4, 26 NaHCO3, 11 glucose, 1.3 MgSO4, 2.5 CaCl2 (pH 7.4 with HCl, 295–305 mOsm) at room temperature for at least 1 h.

After 1 h of recovery, the slices were transferred to a recording chamber under a microscope (Nikon-FN1) and continuously perfused (1 mL/min) with O2-saturated aCSF at 31 °C. Guided by the 520-nm fluorescence signals, we identified and selected PPN neurons with ChR2-mCherry expression for whole-cell patch-clamp recording. Excitation of ChR2 was achieved by 470-nm laser stimulation (25-ms-wide laser pulses, 20 Hz) and controlled by a shutter (Uniblitz, USA). Electrophysiological data were acquired using a MultiClamp 700B amplifier (Axon Instruments, USA). Borosilicate glass electrodes (tip resistance, 4−6 MΩ) were prepared using a Model P-97 Flaming/Brown micropipette puller (Sutter Instruments) and were filled with an internal solution containing (in mmol/L) 126 K-gluconate, 2 KCl, 10 HEPES, 2 MgCl2, 4 Na2ATP, 0.4 Na3GTP, 0.2 EGTA, and 10 creatine phosphate (pH 7.2 with KOH, 290 mOsm) for slice recording.

Multichannel Electrophysiological Recording and Analysis

Neural signals were recorded (digitized at 40 kHz) and bandpass filtered (300–8000 Hz) using an Omniplex data acquisition system (Plexon Inc., USA). Single units were isolated offline by MClust 3.5 software (A.D. Redish et al.; available at http://redishlab.neuroscience.umn.edu/). Single units with a refractory period ≥2 ms and a stable waveform and firing rate during recording were selected. Peri-event time histograms time-locked to laser onset were analyzed and plotted with NeuroExplorer (Version 4.126, Nex Technologies). We calculated the average firing rate of units (using 0.02 ms time bins) for 500 trials of 10 ms and 80 trials of 4-s laser stimulation to identify neurons expressing ChR2 and test the activity induced by laser stimulation.

Histology

For immunostaining, rats were sacrificed under anesthesia and transcardially perfused with PBS followed by 4% paraformaldehyde (PFA). Brains were postfixed in 4% PFA for 4 h at 4 °C and then transferred to 30% sucrose until sinking to the bottom. Coronal or sagittal sections (30 μm) were cut on a cryostat (CM1950, Leica). Goat anti-ChAT (1:1000, AB144P, Merck Millipore), rabbit anti-TH (1:1000, P40101, Pel-Freez), rabbit anti-RFP (1:1000, 600-401-379, Rockland), rabbit anti-GFP (1:1000, A11122, Thermo Fisher Scientific), and guinea pig anti-c-fos (1:1000, 226 004, Synaptic System) were used as primary antibodies, and all secondary antibodies originated from donkeys (1:500; Invitrogen). After secondary antibody incubation, sections were coverslipped using Fluomount-G mounting medium with DAPI (Southern Biotech).

For fluorescence in situ hybridization and immunostaining, rats were perfused with DEPC-treated PBS. Sections were serially cut (20 μm) and mounted on glass slides, which were stored at − 80 ℃ until the start of the multiplex fluorescent RNAscope assay (323100, ACD Biotech). RNA hybridization probes included antisense probes against genes encoding rat vGluT2 (317011-C1), CaMKIIα (317081-C2), GAD2 (435801-C3), and iCre-C2 (423321-C2). After in situ hybridization was finished, immunostaining with anti-ChAT (1:300) or anti-RFP (1:300) was applied. Finally, the slides were coverslipped with Fluomount-G mounting medium with DAPI. It’s worth noting that, during the RNAscope process, the sample slides were pre-treated with hydrogen peroxide and proteinase K III, and the fluorescent protein lost its original laser excitation ability.

In order to confirm the implantation position of the optical fiber and optrodes, we electrolytically lesioned (25 μA, 20 s) the brains with a stimulus isolator (ISO-Flex, AMPI, Israel) before perfusion. Serial sections (50 μm) were cut and coverslipped with Fluomount-G mounting medium with DAPI.

For anterograde and monosynaptic retrograde tracing experiments, 30-μm sections were collected at 240-μm intervals. It should be emphasized that the sections required for cell counting in the PPN subregion were collected extremely carefully to precisely match the rat brain atlas as much as possible by adjusting the angle of the blade while sectioning the thalamus, hippocampus, and midbrain. To assess the injection site of AAV2/5-syn-Flex-Tdtomato-P2A-Synaptophysin-EGFP for an anterograde tracking experiment, slides were pre-treated with hydrogen peroxide to eliminate the TdTomato signal and were used for immunostaining with anti-GFP and anti-ChaT (Fig. S8).

Images were captured with an Olympus VS120 or Olympus FV3000 IX83 confocal microscope (Olympus Co., Japan).

Cell Counting

We experimentally defined the PPN region by combining the expression profile of ChAT and the outline of the PPN in the rat brain atlas and then counted the cells in that region. Coronal sections covering the PPN are mainly from bregma − 6.84 mm to − 8.56 mm according to the rat brain atlas. Here, we collected sections from bregma − 7.20 to − 7.44 mm as the rostral PPN (rPPN), bregma − 7.68 to − 7.92 mm as the middle PPN (mPPN), and bregma − 8.16 to − 8.40 mm as the caudal PPN (cPPN). Based on the results of in situ hybridization and immunostaining, we counted ChAT, mCherry, vGluT2-mRNA, GAD2-mRNA, and CaMKIIα-mRNA cells. For each experimental group, cells of at least 6 sections from 3 animals were counted.

Statistics

The graphics and numerical values of the averaged speed and normalized respiration rate are presented as the mean ± SEM. The graphics and numerical values of the percentage of cell number reported are presented as the mean ± SD. All experiments were performed in multiple batches of replication at least three times per experiment. *P <0.05, **P <0.01, and ***P <0.001, two-tailed paired t-test.