Brainstem ADCYAP1+ neurons control multiple aspects of sickness behaviour – Nature

Mice

Animal experiments were approved by The Rockefeller University Animal Care and Use Committee and were carried out in accordance with the National Institutes of Health guidelines. All experiments were conducted on male mice older than 8 weeks of age housed in a 12-h light/12-h dark cycle at 22 °C and 30–60% humidity. We used the following genotypes of mice: C57BL/6J (wild type; number 000664, The Jackson Laboratory), FOS^2A-iCreER (TRAP2) mice (The Jackson Laboratory, stock number 030323), CAG–Sun1/sfGFP (INTACT) mice (The Jackson Laboratory, stock number 021039), VGLUT2–IRES–Cre mice (The Jackson Laboratory, stock no. 028863), VGAT–IRES–Cre mice (The Jackson Laboratory, stock no. 028862), ADCYAP1–2A–Cre mice (The Jackson Laboratory, stock number 030155), PHOX2B–Cre (The Jackson Laboratory, Stock 016223), DBH–Cre (The Jackson Laboratory, stock number 033951), TH–Cre (The Jackson Laboratory, stock number 008601), TAC1–Cre (The Jackson Laboratory, stock number 021877), PDYN–Cre (The Jackson Laboratory, stock number 027958), CCK–Cre (The Jackson Laboratory, stock number 012706), NTS–Cre (The Jackson Laboratory, stock number 017525). For all Cre mouse line experiments, only heterozygous animals were used. Sample sizes were decided based on experiments from similar studies. Animals were randomly assigned to groups when possible.

Viral vectors

For chemogenetic studies, AAV5-hSyn-DIO-hM3D(Gq)–mCherry (activating), AAV5-hSyn-DIO-hM4Di(Gi)–mCherry (inhibitory) or AAV5-hSyn-DIO–mCherry (control) was used. AAV5 serotype was used as we observed minimal tropism for dorsal motor vagus neurons. All viruses were obtained from Addgene.

Stereotactic surgery

Animals were anaesthetized using isofluorane anaesthesia (induction 3%, maintenance 1.5–2%). For the NTS–AP region, the following coordinates relative to the bregma were used: anterior–posterior, −7.7 mm to −7.8 mm; medial–lateral, ± 0.1 mm; and dorsal–ventral, −4.8 mm. For chemogenetics, 25–50 nl of virus was injected bilaterally at a rate of 1 nl s⁻¹ using a Drummond Scientific Nanoject III Programmable Nanoliter Injector.

LPS administration

LPS from Escherichia coli (O55:B5; Sigma-Aldrich L2880) was resuspended in saline (3 mg ml⁻¹) and frozen into single-use aliquots. Further dilutions in saline to the appropriate experimental dose and a standardized volume (about 100 µl per animal) were further generated before each experiment. All doses were delivered with a single i.p. injection.

Phenotyping measurements and analysis

Animals were phenotyped in custom home-cages consisting of a modified Feeding Experimentation Device 2.0 (refs. ^44,45) for single-pellet-based food intake measurement, a capacitive lickometer for lick-based water intake measurement and an infrared-compatible camera for movement measurement. Measurements were continuously collected at about 100 Hz for food and water intake and eight frames per second for movement. All data were collected using Bonsai⁴⁶. Mice were housed singly and allowed at least 1 week to habituate to the phenotyping cages before any data were acquired. Data were post hoc analysed using custom MATLAB code to identify pellet withdrawal events and licking events, and with ToxTrac⁴⁷ and MATLAB code for movement. For normalized plots of food intake, water intake and movement, data for each animal were analysed with respect to cumulative measurements of each respective parameter for 24 h for that respective animal without any treatment (baseline behaviour).

Thermal measurements

Measurements of core body temperature were obtained using an anal probe (Braintree Scientific). BAT temperature measurements were obtained using wireless implantable temperature probes (IPTT-300, Bio Medic Data Systems). Further temperature characterization was visualized with a thermal camera (FLIR Systems) taking at least three images at each respective time point reported.

AdipoClear whole-brain clearing and imaging

Mice were euthanized by intracardial perfusion with 1× PBS followed by 4% paraformaldehyde. Two saline-treated groups were collected at 0.5 h and 1 h post injection of saline as previous experiments have shown that FOS patterns for saline treatment do not change beyond the first hour after injection. Six LPS-treated groups were collected at 0.5 h, 1 h, 3 h, 6 h, 12 h and 24 h post injection of 0.5 mg kg⁻¹ LPS. Brains were carefully dissected and post-fixed in 4% paraformaldehyde for 16 h at 4 °C before proceeding with the AdipoClear protocol as previously described¹⁸ with the exception of primary and secondary antibody incubations conducted at 37 °C and for 6 days per incubation. Synaptic Systems rabbit c-FOS antibody (number 226003) or a custom guinea pig c-FOS antibody (courtesy of Zuckerman Institute Antibody Core) was used for primary staining at 1:2,500 and 1:25,000 respectively, and a donkey anti-rabbit Alexa Fluor 647 (Thermo Fisher Scientific A-31573) secondary antibody at 1:2,500 was used. Samples were stored in dibenzyl ether at room temperature until imaging. Samples were imaged in a dibenzyl ether reservoir using a light-sheet microscope (Ultramicroscope II, LaVision Biotec) equipped with a 4× objective lens for FOS imaging and 1.3× lens for autofluorescence imaging.

ClearMap analysis of FOS data

FOS and autofluorescence data were analysed and mapped to the Allen Brain Reference Atlas using the ClearMap pipeline as described previously¹⁷. The 0.5 h LPS time point was normalized to the 0.5 h saline-treated group, whereas all other LPS time points (1 h–24 h) were normalized to the 1 h saline-treated group.

TRAP2 labelling for chemogenetics

AAV5-hSyn-DIO-Gq–mCherry (Addgene) or AAV5-hSyn-DIO-Gi–mCherry (Addgene) virus was delivered to TRAP2 mice of >8 weeks of age through stereotactic injection into the NTS–AP region. Mice were given at least 3 weeks for recovery post surgery. Habituation for i.p. injection was carried out daily for 1 week using injections of 100 µl saline. Mice were intraperitoneally injected with 4HT at 20 mg kg⁻¹ concurrently with saline or LPS (0.5 mg kg⁻¹). Mice were allowed to recover and express DREADDs for at least 3 weeks before animals were injected with 1 mg kg⁻¹ CNO or saline (control) to assess behaviour. For inhibition experiments, mice were concurrently injected with LPS during CNO delivery. Exposure to LPS, which can reduce subsequent responses, was controlled between saline- and LPS-labelled groups by injecting saline-labelled animals with LPS at least 3 days after 4HT injection.

TRAP2 labelling for TRAP2-seq

TRAP2 animals were bred to INTACT (Cre-dependent Sun1–GFP) animals to produce heterozygous animals for both alleles. Animals of about 8 weeks of age were habituated to i.p. injections daily for 1 week using injections of 100 µl of saline before labelling with 4HT and LPS similarly to TRAP2 labelling for chemogenetics.

Isolation of nuclei for snRNA-seq

Animals were euthanized through transcardial perfusion using a HEPES–sucrose solution (NaCl 110 mM, HEPES 10 mM, glucose 25 mM, sucrose 75 mM, MgCl₂ 6H₂O 7.5 mM, KCl 2.5 mM)⁴⁸. Brains were dissected on ice, frozen using LN₂ and stored at −80 °C until isolation of nuclei. Samples were then processed according to a modified version of a previously described protocol⁴⁹. On the day nuclei were extracted, samples were thawed on ice, resuspended in HD buffer (tricine KOH 10 mM, KCl 25 mM, MgCl₂ 5 mM, sucrose 250 mM, 0.1% Triton X-100, SuperRNaseIn 0.5 U ml⁻¹, RNase Inhibitor 0.5 U ml⁻¹), and homogenized using a 1-ml dounce homogenizer. Homogenates were filtered using a 40-μM filter, centrifuged at 600g for 10 min and resuspended in nucleus storage buffer (sucrose 166.5 mM, MgCl₂ 10 mM, Tris buffer pH 8.0 10 mM, SuperRNaseIn 0.05 U ml⁻¹, RNase Inhibitor 0.05 U ml⁻¹) for staining. Nucleus quality and number were assessed using an automated cell counter (Countess II, Thermo Fisher). Nuclei were stained with combinations of Hoechst 33342 (Thermo Scientific H3570; 1:500 per volume), anti-NeuN Alexa Fluor 647-conjugated antibody (Abcam ab190565) (1:500 per volume) and TotalSeq anti-Nuclear Pore Complex Proteins Hashtag antibody (BioLegend 682205) (0.5 mg per million nuclei) for 30 min at 4 °C. Samples were then washed with 10 ml nucleus storage buffer and centrifuged at 600g for 10 min. Nuclei were resuspended in PBS with 2% BSA and RNase inhibitors (SuperRNaseIn 0.5 U ml⁻¹, RNase Inhibitor 0.5 U ml⁻¹) for subsequent fluorescence-activated cell sorting. Samples were gated on Hoechst to identify nuclei followed by a subsequent gate on high Alexa Fluor 647 expression level for NeuN⁺ nuclei designating neurons.

snRNA-seq of the NTS–AP

Nuclei were captured and barcoded using 10x Genomics Chromium v3 according to the manufacturer’s protocol. Samples were processed and libraries were prepared by The Rockefeller Genomics Core. Sequencing was performed by Genewiz using Illumina sequencers.

snRNA-seq analysis

The FASTQ file for each sample was analysed with Cell Ranger version 5.0. The following analysis was based on Orchestrating Single-Cell Analysis from Bioconductor (v1.0.6; https://bioconductor.org/books/release/OSCA/) and vignettes from Seurat (v4.0; https://satijalab.org/seurat/)⁵⁰. The ambient RNA contamination and doublets were estimated by using DropletUtils (https://bioconductor.org/packages/release/bioc/html/DropletUtils.html) and scDblFinder (https://bioconductor.org/packages/release/bioc/html/scDblFinder.html), respectively. After removing ambient RNA, doublets and nuclei with >1% mitochondrial reads, the data from individual samples were integrated using canonical correlation analysis from Seurat using the first 2,000 variable features of each sample. The counting matrices of the samples were then merged and rescaled by using the common variable features of the samples. Finally, a single-cell dataset with 29,329 (features) × 19,918 (cell barcodes) was generated. To optimize the clustering and projection, we tested the principal component analysis up to 40 principal components and evaluated the optimal cutoff by using an elbow plot. In this case, we took the first 25 principal components for clustering and projection with both (UMAP) and t-distributed stochastic neighbour embedding. We also tested the resolution of clustering from 0.1 to 1.0. The results were evaluated with clustree (https://github.com/lazappi/clustree). In this dataset, the resolution was set to 1.0, resulting in 44 clusters in this dataset. The clustering and projection information was also applied to generate a customized LOUPE file for further analysis and visualization.

Chemogenetics for activation or inhibition

AAV5-DIO-Gq (Addgene), AAV5-DIO-Gi (Addgene) or AAV5-DIO–mCherry (Addgene) virus was delivered to respective Cre line mice of >8 weeks of age through stereotactic injection into the NTS–AP region. Mice were allowed to recover and express DREADDs for at least 4 weeks. For activation, animals were injected with 1 mg kg⁻¹ of CNO or saline (control). For inhibition, animals were injected with 1 mg kg⁻¹ of CNO or saline (control) concurrent with LPS (0.5 mg kg⁻¹).

Immunofluorescence

Animals were transcardially perfused with PBS followed by 4% paraformaldehyde. Brains were dissected and post-fixed for 16–24 h at 4 °C. Brains were sectioned using a vibratome (Leica) to a thickness of 50 µm. Samples were then stained with the following primary antibodies: anti-GFP (Aves Labs GFP-1020) at 1:1,000, anti-mCherry (Rockland 600-401-379) at 1:1,000 or anti-FOS antibody (Synaptic Systems 226 003) at 1:2,000 in PTxwH buffer for 24 h at 4 °C. After four 15-min washes in PTxwH buffer, samples were incubated in Alexa Fluor-conjugated secondary antibodies (Thermo Fisher) at 1:2,000 for 90 min at room temperature, followed by four further 15-min washes before mounting using 4′,6-diamidino-2-phenylindole Fluoromount-G (SouthernBiotech 0100-20 OB010020).

Statistical analysis

Statistical parameters for individual experiments are reported as sample size (n represents the number of animals per group), statistical test used and statistical significance. ANOVA denotes two-way ANOVA. All data and graphs are shown as mean ± s.e.m. unless noted otherwise. Significance is defined as P < 0.05, with significance annotations of *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. All statistical analysis was performed using GraphPad Prism, MATLAB, R, Python or ClearMap.

Reporting summary

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