Mice

Wild-type C57BL/6NCrl mice (Charles River Laboratories, strain 027), NtsCre knock-in mice34 (Jackson Laboratory, strain 017525) and Ai9 or Ai14 tdTomato mice35 (Jackson Laboratory, strain 007909) were housed and bred in the animal facility at Stanford University in accordance with Institutional Animal Care and Use Committee (IACUC) guidance and were maintained on a 12-h light/dark cycle at temperature 70–75 °F and humidity 35–60%, with food and water provided ad libitum. Male mice of age 6–8 weeks were used for all adult social vocalization studies. Both male and female mice of age 6–8 weeks were used for all optogenetic stimulation and viral tracing studies. Neonatal mice were postnatal day 7 and were not sexed. All mouse experiments were approved by the Stanford University IACUC.

Vocalization induction for neural activity monitoring by Fos labeling

For adult social vocalization induction, adult (6–8 weeks old) male wild-type C57BL/6NCrl mice were individually housed 1 d before experiments. The next day, an adult wild-type female was placed in the male’s cage, and an ultrasonic microphone (Avisoft Bioacoustics, CM16/CMPA) was used to verify vocalization production. After 90 min, the male was immediately killed by CO2 inhalation and transcardial perfusion for smFISH studies.

For neonatal isolation cry induction, postnatal day 7 wild-type mice were removed from their home cage and placed in a large plastic box while an ultrasonic microphone was used to verify vocalization production. After 90 min, the pup was immediately killed by saturating vapors of isoflurane and transcardial perfusion for smFISH studies.

AAV injection and optical fiber implantation

Adult NtsCre mice (6–8 weeks old) were anesthetized with isoflurane (3% for induction and 1–2% for maintenance) for AAV injections. Anesthetized mice were placed in a stereotactic instrument (David Kopf Instruments, model 940) with body temperature maintained at 37 °C using a feedback-controlled heating pad (Physitemp Instruments, TCAT-2LV). Immediately before surgery, mice were given analgesic (carprofen 5 mg kg−1 and buprenorphine SR 0.5–1.0 mg kg−1, subcutaneous). The following AAV vectors were used: for caudolateral PAG labeling, AAVDJ-CAG-GFP (9.3 × 1012 genome copies per milliliter (GC ml−1), Stanford Gene Vector and Virus Core); for cell ablation, AAV8-Ef1a-FLEX-taCasp3-TEVp12 (8.8 × 1013 GC ml−1, Addgene, 45580, Janelia Viral Tools facility); for mock ablation, AAV8-CAG-FLEX-GFP (UNC Vector Core); for optogenetic stimulation, AAV8-Ef1a-DIO-bReaChES-TS-eYFP (2.9 × 1012 GC ml−1, Stanford Gene Vector and Virus Core); and for projection mapping, AAVDJ-hSyn-FLEX-mGFP-2A-Synaptophysin-mRuby36 (1.2 × 1013 GC ml−1, Stanford Gene Vector and Virus Core). To target the caudolateral PAG, 50 nl of AAV-GFP was unilaterally injected on the left side at the following stereotaxic coordinates: −4.5 mm caudal to bregma, −0.7 mm lateral and −2.0 mm ventral to the surface of the brain. To target RAm in NtsCre mice, 500–700 nl of the GFP, Casp3 or bReaChES AAV vector was bilaterally injected at the following stereotactic coordinates: 3.4 mm caudal to lambda, ±1.25 mm lateral to lambda and 6.3 mm ventral to lambda. Immediately after bilateral AAV-DIO-bReaChES injection, fiber-optic cannulas (Doric Lenses, MFC_200/230-0.37_6mm_ZF1.25_FLT) were bilaterally implanted 350 μm above the injection site and secured to the skull with dental cement (Parkell, C&B Metabond). For all experiments, mice in which each injection did not target RAm on histology were excluded from analysis. For projection mapping experiments, 100 nl of the Syp-mRuby AAV vector was injected unilaterally into the left RAm. Mice recovered for 4 weeks for ablation experiments and 3–8 weeks for optogenetic or projection mapping experiments.

Adult vocalization recordings in neural ablation experiments

For vocalization recording before ablation, adult male NtsCre mice (6–8 weeks old) were individually housed 1 d before recording. On the day of recording, the male’s cage was placed in a black plastic box with an ultrasonic microphone (Avisoft Bioacoustics, CM16/CMPA) and a video camera mounted above the cage. After 15 min of acclimation time, an adult wild-type female was placed in the male’s cage for 5 min, and the encounter was recorded. Female mice were verified to be in estrus37 on the day of recording to maintain consistency between trials. The next day, male NtsCre mice were randomized to ablation or control groups, and AAV8-Ef1a-FLEX-taCasp3-TEVp or AAV8-CAG-FLEX-GFP (mock ablation control) was bilaterally injected into RAm. After a 4-week recovery to allow for protein expression and cell ablation, vocalization was recorded as above in a 5-min encounter with a wild-type female in estrus. The male was then killed by transcardial perfusion. Vocalizations were assumed to be produced by the male because males produce the majority of vocalizations while in the presence of a female38.

Ultrasonic vocalization and behavior analysis

Adult social vocalization sound files were analyzed using MATLAB (MathWorks) with MUPET39. To extract syllables, sound files were processed using default MUPET parameters, with the exception of the following: minimum-syllable-duration, 8; minimum-syllable-total-energy, −40; minimum-syllable-peak-amplitude, −40; minimum-syllable-distance, 10; and minimum-usv-frequency, 50,000. All extracted syllables were manually examined by a blinded experimenter, and any falsely detected syllables due to noise from audible mouse movement were excluded from analysis. Peak syllable amplitude and syllable pitch (‘mean frequency’ in MUPET) were extracted directly from the MUPET output file. To calculate social interaction time, videos were manually scored by counting the number of seconds during the 5-min trial in which the male mouse’s nose or forelimb was in contact with the female.

Optogenetic stimulation with recording of vocalization

NtsCre mice injected with AAV8-Ef1a-DIO-bReaChES-TS-eYFP and recovered as above were anesthetized with isoflurane (3% induction and 1–2% maintenance), and body temperature was maintained at 37 °C. Respiration was recorded using a spirometer (ADInstruments) connected to a plastic nose cone that also delivered maintenance isoflurane. Vocalizations were recorded with an ultrasonic microphone (Avisoft Bioacoustics, 40007) attached to the nose cone and connected to a Avisoft Bioacoustics UltraSoundGate 116H recording interface. Single-lead ECG was recorded using needle electrodes (ADInstruments, MLA1203), an ADInstruments Octal Bio Amp and an ADInstruments PowerLab data acquisition system. The implanted fiber-optic cannulas were connected via fiber-optic cable (Doric Lenses, SBP(2)_200/220/900-0.37_1m_SMA-2xZF1.25) to a 577-nm laser (CNI Laser), and laser light was delivered using the following parameters: 10–15-mW power from the fiber tip, 10-ms pulse width and 5-s stimulation train duration. The interval between stimulation trains was 5 min, and two stimulation trains were performed at each pulse rate. Pulse width was not varied. Laser power and pulse frequency were varied as indicated in each figure. Because of the high spike fidelity of bReaChES40, we assume that increasing pulse frequency increased the firing rate of RAm Nts neurons.

For optogenetic stimulation in awake mice, stimulation was performed as above while mice were placed in a black plastic box with an ultrasonic microphone (Avisoft Bioacoustics, CM16/CMPA) and a video camera mounted above the cage. Lower laser powers and stimulus durations were sufficient to elicit vocalization in awake mice, so the stimulation parameters were modified to a stimulus train duration of 500 ms and laser power of 5 mW.

To calculate peak syllable amplitude and syllable pitch of optogenetically driven vocalizations, sound files were analyzed using Audacity (https://www.audacityteam.org/). The onset and offset time of each syllable was manually annotated. The ‘plot spectrum’ function was applied to each syllable to calculate the fast Fourier transform. To calculate peak syllable amplitude, the peak amplitude of each syllable in dB was subtracted from the dB of quiet background noise. The frequency of the syllable at its peak amplitude was reported as the syllable pitch. When multiple syllables were recorded at a given stimulation parameter set, the loudest syllable was used to calculate the peak syllable amplitude and syllable pitch.

Optogenetic stimulation with EO and CT muscle EMG recording

NtsCre mice injected with AAV8-Ef1a-DIO-bReaChES-TS-eYFP and recovered as above were anesthetized with isoflurane (3% induction and 1–2% maintenance), and body temperature was maintained at 37 °C. Respiration was recorded using a spirometer (ADInstruments) connected to a plastic nose cone that also delivered maintenance isoflurane. The mouse was placed in the supine position, and the skin overlying the EO muscle was aseptically prepared. A 1-cm skin incision was made to expose the EO muscle, and a two-lead needle electrode (ADInstruments, MLA1203) was inserted into the muscle. The CT muscle was similarly exposed with a 1-cm ventral neck skin incision, followed by dissection of the overlying strap muscles, and two 76.2-μm-diameter silver wires (A-M Systems) were inserted into the CT muscle. Electrodes were connected to an ADInstruments Octal Bio Amp and an ADInstruments PowerLab data acquisition system that recorded EMG and respiration at a sampling rate of 1 kHz. EMG signals were high-pass filtered at 100 Hz and then integrated using LabChart parameters: Integral, absolute value; Time constant decay, 0.2 s. Fold change of integrated EMG amplitude was calculated by dividing the peak integrated EMG amplitude during laser stimulation by the integrated EMG amplitude immediately before laser stimulation. Laser light (577 nm) was delivered as above using the following parameters: 10–15-mW power from the fiber tip and 10-ms pulse width.

CT and EO muscle injections for RAm Nts neuron projection mapping

For CT and EO muscle injections, adult (age 6–8 weeks) NtsCre mice were used that had been previously injected as described above with AAVDJ-hSyn-FLEX-mGFP-2A-Synaptophysin-mRuby into RAm 8 weeks before the muscle injections. Mice were anesthetized with isoflurane (3% for induction and 1–2% for maintenance) and then pre-treated with analgesic (carprofen 5 mg kg−1 and buprenorphine SR 0.5–1.0 mg kg−1, subcutaneous). For CT injections, a 1-cm incision was made in the ventral neck, and the CT muscles were exposed by dissection of the overlying strap muscles. A pulled glass micropipette (Drummond Scientific, 5-000-2005) was then used to inject 200–300 nl of 1% CTB solution (Sigma-Aldrich, C9903, diluted in PBS + 0.05% Fast Green dye) into the left and right CT muscles. The overlying skin was sutured, and the mouse was placed in a heated recovery cage. For EO injections, 400–1,000 nl of 1% CTB was similarly injected into the left and right EO muscles through an incision in the overlying skin that was sutured after injection. Mice recovered for 3 d (CT) or 7 d (EO) before perfusion and immunostaining.

smFISH and immunostaining

Mice were killed with CO2 and transcardially perfused with 4% paraformaldehyde (PFA), and tissues were post-fixed in 4% PFA overnight at 4 °C. Brains and spinal cords were cryoprotected in 30% sucrose at 4 °C overnight. Cryoprotected tissue was embedded in optimal cutting temperature (OCT) compound and sectioned on a Leica CM3050S cryostat at 20 μm for smFISH and 25 μm for immunostaining.

For smFISH, sections were processed with an RNAscope Multiplex Fluorescent Assay v2 kit according to manufacturer instructions and using the following probes: Mm-Fos (316921), Mm-Nts-C2 (420441-C2), Mm-Slc17a6-C3 (319171-C3) and Mm-Slc18a3-C3 (448771-C3).

For immunostaining, sections were permeabilized in PBS + 0.3% Triton X-100, blocked for 1 h in block buffer (PBS + 0.3% Triton + 10% normal donkey serum) and incubated with primary antibodies in block buffer at 4 °C overnight. Slides were washed three times, incubated in secondary antibodies in block buffer for 1 h at room temperature and washed three times, and a coverslip was applied with ProLong Gold Antifade Reagent. Primary antibodies included: chicken anti-GFP (Aves Labs, GFP-1010, 1:1,000), rabbit anti-c-Fos (Synaptic Systems, 226 003, 1:5,000), goat anti-ChAT (Millipore, AB144P, 1:100) and goat anti-CTB (List Labs, 703, 1:1,000). Species-specific donkey secondary antibodies conjugated to Alexa Fluor 488, 568 or 647 were obtained from Life Technologies or Jackson ImmunoResearch and used at a 1:500 dilution.

To determine the total number of RAm neurons, the cluster of c-Fos+ neurons after vocalization was used to define the boundaries of the region, and NeuroTrace (Invitrogen, N21479) was used to count the number of neurons within those boundaries. To count RAm Nts neurons in AAV experiments, the Nts smFISH probe or an NtsCre-driven tdTomato allele was used. NeuroTrace was used to differentiate Nts+ neurons from lineage-labeled vasculature in NtsCre;tdTomato mice. Stained neurons were counted manually from z-stacks acquired on a Zeiss LSM 780 confocal microscope. To quantify innervated RAm neurons or CT/EO motor neurons, a neuron was scored as innervated if it had at least two GFP+ (PAG experiment) or mRuby+ (CT/EO experiment) puncta directly abutting the cell soma.

To quantify RAm Nts innervation of brainstem and spinal cord nuclei, mGFP+ fibers were quantified in ImageJ (Fiji). ROIs were drawn around the brain regions, and the GFP channel was converted to a mask and then binarized. ‘Area fraction’ was then quantified for each ROI to calculate the percent area innervated.

Data collection and statistics

All results are presented as mean ± s.d. with all data points displayed. All statistical analyses were performed with GraphPad Prism. All statistical tests used are listed in the figure legends, with statistical significance set at P < 0.05. Statistical tests were not used to pre-determine sample size, but sample sizes are similar to those in previous publications3,22. Data distribution was assumed to be normal, but this was not formally tested. Fos labeling and social interaction time were quantified by a blinded experimenter. Syllable count, syllable amplitude and EMG amplitude were quantified using the same automated approaches for all mice, so blinding was not relevant. Wild-type mice were randomized into control or vocalization groups for Fos labeling. Male NtsCre mice were randomized to ablation or control groups.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.