Left part: Schematic drawing of a horizontal section of the hypothalamus in which the Foxb1-neurons of the parvafox and of the PMd are located (modified after (Alvarez-Bolado et al. 2000). Their chemogenetic (DREADD) stimulation leads to an increase in breaths / minutes.

Right part: dorsolateral PAG (dlPAG), seen in a cross-section of the midbrain. Axon terminals of parvafoxFoxb1 and PMdFoxb1 converge in the rostral part of the dlPAG. Optogenetic activation of these terminals lead to immobility and bradycardia.

a) Data across all DREADD WBP recordings was plotted as averaged line plots for each animal group (DREADD_neg, hM3Dq and hM4Di) and condition (Saline (red), Clozapine (green) or CNO (blue) i.p. injection). Each animal was recorded three times within each condition (nine recordings in total). Respiratory parameters do not differ between condition in DREADD_neg animals. Injection of CNO or Clozapine in hM3Dq mice significantly alters several respiratory parameters. The effects of Clozapine and CNO injections on respiratory parameters largely overlap and clearly separate from the effects saline injections. In hM4Di animals, more variability within respiratory parameters compared to DREADD_neg controls, however, no statistically significant effect is detected. b) Violin plots with integrated boxplots for each group and condition. The width of the violin plot represents the data distribution density. The boxplot ‘s lower and upper limits represent the 25 % quantile and 75 % quantile, respectively. The horizontal bar inside the boxplot represents the median. The whiskers of the boxplot display 1.5x the interquartile range. c) Comparison of gross locomotion as assessed by an open field test. hM3Dq animals show statistically significant reduction in distance moved, while there are no differences observed in the DREADD_neg and hM4Di animals. It is important to note, that the effect size for this reduction in distance moved is small. d) A statistically significant reduction in time spent in immobility is observed in DREADD_neg animals, however, the effect size is negligible. Time spent in an immobile state does not differ in hM3Dq and hM4Di animals. Track visualization with underlying density maps and zone visit diagrams during open field tests for all DREADD mice can be found in supplementary file S1 and S2. I = BPM (Breaths per minute); II = IT (Inspiratory time); III = ET( Expiratory time); IV = TT (Total respiratory cycle time); V = TVadjPerGram (Tidal volume normalized to bodyweight in microliters per gram); VI = MVadjPerGram( Minute volume normalized to bodyweight in milliliters per gram); VII = PIFadj (Peak inspiratory flow); VIII = PEFadj (Peak expiratory flow); CNO: clozapine-N-Oxide.

Representative pseudo-colored immunofluorescent wide field images of DREADD-expressing ParvafoxFoxb1 neurons (green) and c-Fos immunoreactive nuclei (in magenta). Please note, that in the activating DREADD animal (hM3Dq), there is a marked increase of c-Fos signal colocalized to the DREADD expression. Contrary to the activating DREADD mouse, there is complete absence of c-Fos+ nuclei in the area of DREADD expression in hM4Di expressing mice. Since these are no confocal images, double labeling in the merged images may theoretically result from out-of-focus immunofluorescence. Scale bar = 250 μm.

a) A representative example of track visualizations with underlying density maps for a ChR2-expressing Foxb1-Cre mouse belonging to the “OnTarget_antPAG” group during 3 minutes of baseline (left) and 3 minutes of optogenetic stimulation (right). Note that the mouse remained immobile for the entire duration of the stimulation period. The color coding scales are not fixed between the two conditions. Track visualization with underlying density maps for all other mice can be found in supplementary file S3. b) A representative example of a zone visit diagram taken from the same recording as the track visualizations in a). Note, that the zone transitions stop completely after the onset of optogenetic stimulation. Zone visit diagrams for all other mice can be found in supplementary file S4 c) A visual representation of the arena partitioning into the different zones (i.e. arena, periphery, center, and 4 corners). d-h) Comparison of multiple parameters extracted from the open field experiments. Optogenetic activation of parvafoxFoxb1 terminals in the PAG reduces distance moved (d), speed.moving (e), and time in center (h), while time spent immobile (f) and time in periphery (g) increases. Optogenetic silencing of parvafoxFoxb1 terminals in the PAG induces opposing effects.

a) Single cell transcriptomic analysis of murine ventral-posterior hypothalami groups PMd cells into a distinct cluster (Cluster 9). b) Cluster identity of the PMd cluster is confirmed by expression pattern analysis of genes known to be upregulated in the PMd (Cck, Foxb1, Synpr, Ebf3, Dlk1 and Stxpb6). c) In situ hybridization photomicrographs from the Allen Brain Atlas show the localization of Cck and Foxb1 transcripts in the PMd. d) A magnified UMAP plot representation of the PMd cluster highlights the differential expression profiles of Cck and Foxb1 within the PMd cluster. While cells expressing high transcript levels of Cck (middle column, red) preferentially localize to the left side of the PMd cluster, cells with high levels of Foxb1 transcripts (middle column, green) preferentially localize to the opposite (i.e. right) side of the PMd cluster. Analysis of co-expression of Cck and Foxb1 transcripts identifies only few cells as strongly double positive (yellow; see color coding representation), while most cells with high expression levels for one of the two genes have very low to non-existing expression levels of the other gene.

Results from hot place experiments performed on Foxb1-Cre animals does not reveal statistically significant differences in a) latency until endpoint behavior (i.e. hindpaw lick(-attempt) or jumping) nor in b) number of shakes before endpoint behavior in any of the three tested groups. These results contradict the hypothesis of a reciprocal effect of the parvafoxPV and parvafoxFoxb1 on pain sensation.

a: Telemetric recording of heart rate of mouse 34-21/10 in an open field. Average every 10 seconds, 40 minutes long recording. Heart rate increases from 700 to 740 during the first few minutes and remain at this level until the start of the optogenetic activation of the Foxb1 endings in the dlPAG. During the 2 minutes of flashing blue light, heart rate drops to under 640 but has an initial, short recovery to 690. During the following 10 minutes of baseline recording, the cardiovascular system is slightly dysregulated, with an increased variability in heart rate. During the second photo-stimulation, the drop in heartbeats has an amplitude comparable to the one during the first stimulation.

b: Telemetric recording of heart rate (blue curve) and movements (red curve) of mouse 106-21/10 in an open field. Average every 10 seconds, 20 minutes long recording. Heart rate values are around 740 during the first minutes (baseline condition), while the mouse is continuously moving around in the open field. Shortly after the beginning of the optogenetic activation of the Foxb1 endings in the dlPAG, the heartbeat drops massively (by ~40 %) from 740 to 450 bpm and remains around this value during the 2 minutes of blue light flashing (photo-stimulation). Notice the complete immobility of the mouse during this period. At the end of the optogenetic activation, heart rate rapidly returns to the baseline level of approximately 740 bpm.

Localization of the glass fibers during optogenetic activation of axon terminals in the PAG leading to immotility (a.), respectively bradycardia (b.). The effects were achieved by inserting the fibers in the rostral part of the dlPAG (Bregma −3.28, −4.04) Glass fibers colored in yellow, with their tip located over the dlPAG (a.), provoked immobility. Glass fibers colored in black had their tip below the dlPAG and did not evoke changes in mobility. Glass fibers indicated by a green arrow, with their tips over the dlPAG, provoked bradycardia. Glass fibers indicated by black arrows, with their tip below dlPAG or distal to bregma −4.04., did not affect the cardiovascular system (b.).

c.) Cross sections of the rostral PAG with EGFP-labelled Foxb1-terminals located in the dlPAG (mouse 106-21/10). The position of the obliquely inserted glass cannula is indicated with a large white arrow and the flat tip of the cannula is positioned over the dlPAG on the right at bregma −3.40 (left image), and on the left at bregma −3.64 (right image). Aq: Aqueductus cerebri; Dk: Darkschewitsch Nucleus; dlPAG: dorsolateral periaqueductal gray; EW: Edigener-Westphal nucleus; lPAG: lateral PAG; Ma3: medial accessory oculomotor nucleus; PrEW: pre-Edinger-Westphal nucleus.

Heart rate variability (HRV) during the optogenetic activation of the Foxb1-terminal endings in the dlPAG of mouse 106-21/10. Comparison between the one-minute period before (green rectangle) and after (red rectangle) the start of the optogenetic activation. While all 4 parameters changed significantly during optogenetic stimulation (heart rate, systolic BP, diastolic BP, mean BP), the SD of heart cycle duration changed by a factor of more than five, from 3.8 ms during the baseline period to 20.6 ms during optogenetic activation.