Satellite glial cells modulate proprioceptive neuron activity in dorsal root ganglia

Proprioception, the sense of body and limb position, is mediated by proprioceptors and is crucial for important motor functions such as standing and walking. Proprioceptor cell bodies reside within the peripheral dorsal root ganglia (DRG) and are tightly e nveloped by satellite glial cells (SGCs). SCGs express a number of Gq protein-coupled receptors (Gq GPCRs), but their functional consequences on proprioceptor activity is unknown . Using a combination of chemogenetics, genetics, Ca 2+ imaging, pharmacology, immunohistochemistry, and biochemistry, we provide evidence that SGC Gq GPCR signaling is sufficient to drive purinergic receptor-mediated Ca2+ responses in proprioceptor cell bodies. Our findings suggest a potential role for SGC Gq GPCR signaling in shaping proprioceptor information processing. Furthermore, t his demonstration of SGC-induced proprioceptor activation has profound implications with SGC Gq GPCR signaling and purinergic receptors representing potential therapeutic targets for alleviating some propriocept or and sensorimotor impairment s associated with spinal muscular atrophy or Friedreich’s ataxia.

Proprioceptors innervate muscle spindles and tendons, and synapse onto spinal ventral horn lower motor neurons (MNs) to provide reflexive information about the length and contraction of muscles. They are crucial for coordinating the activity of MNs and skeletal muscles to achieve essential motor tasks, such as posture, locomotion, manoeuvring one’s way around obstacles or reacting rapidly to external perturbations1,2. When receiving sensory inputs at their peripheral nerve endings, DRG sensory neurons fire action potentials, eliciting the release of neurotransmitters from their cell bodies within DRGs 3 as well as from their axonal terminals in the spinal cord . There is emerging evidence that transmitters released from sensory neuron cell bodies activate receptors at the surface of SGCs4,5 and that SGC-to- preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 2 neuron purinergic communication takes place in DRG 11,12. Both SGCs and somata of sensory neurons express a plethora of receptors, including Gq GPCR purinergic receptors6–10, however, only a few studies have examined the involvement of SGC Gq GPCR signaling in SGC-to-sensory neuron interactions and no study has specifically examined SGC-to-proprioceptor interactions. Therefore, in this study, we asked whether selective activation of SGC Gq GPCR signaling is sufficient to elicit proprioceptor responses in DRGs. We further clarified whether purinergic receptors contributed to such proprioceptor responses. Our findings have identified, for the first time, a SGC-to- proprioceptor communication involving purinergic receptor-mediated Ca2+ activity in proprioceptors.

Chemogenetic and genetic strategies for selectively activating SGC Gq GPCR signaling in DRGs To define the role of SGC G q GPCR signaling in proprioceptor activity, we used the GFAP-hM3Dq transgenic m ouse model expressing a chemogenetic hM3Dq under the control of the glial GFAP promoter13. In DRGs , we found that hM3Dq is expressed in ~88% of SGCs with no detectable expression in neurons , indicating that this mouse line represent s a valuable model to active GPCR signaling selectively in the vast majority of SGCs (Fig. 1a, Fig. S1, Supplementary Movie 1 and Table 1). Then, t o assess hM3Dq functionality in SGCs , we performed 2-photon Ca 2+ imaging experiments in ex vivo intact DRGs from GFAP-hM3Dq::Cx43-CreERT2::GCaMP6 triple transgenic mice. These mice express hM3Dq and the genetically encoded Ca 2+ indicator GCaMP6f ( fast) selectively in SGCs (Fig. 1a and Fig. S2), enabling both hM3Dq-mediated Ca2+ elevations to be induced and detected selectively in SGCs . Indeed, bath application of 10 M CNO elicited Ca2+ increases in ~94% of SGCs, showing that, as expected, hM3Dq couples to Gq in these glial cells (amplitude: 16.62 ± 1.5 ΔF/F0; n = 119 SGCs from 9 DRGs and 3 mice; Fig. 1b,d and Supplementary Table 2). Importantly, no Ca2+ increase and no change in the frequency of spontaneous events were induced by CNO in DRGs from control Cx43-CreERT2::GCaMP6 double-transgenic mice14 (i.e. lacking hM3Dq expression in SGCs). However, a Gq GPCR agonist cocktail reliably triggered Ca2+ elevations in SGCs from these mice (n = 149 SGCs from 9 DRGs and 3 mice; Fig. 1c, Fig. S3 and Supplementary Table 2). Together, these data demonstrate that CNO has no non-specific effects in itself in our experimental conditions. We next asked whether MAPK/ERK pathway was also activated downstream of hM3Dq stimulation in SGCs . Compared to control groups, DRGs from CNO-treated GFAP-hM3Dq mice (1mg/kg CNO intraperitoneal) exhibited a ~76% increase in activated ERK1/2 (pERK) expression levels (Fig. 1e, Fig. S4 and Supplementary Table 3), which was found to be selective to SGCs and not to sensory neurons (Fig. 1f, Fig. S5 and Supplementary Table 3). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 3 Together, these results validate the use of GFAP -hM3Dq mice for stimulating canonical Gq GPCR signaling cascades (i.e. Ca2+ and MAPK/ERK pathways) selectively in SGCs of DRGs. SGC Gq GPCR activation induces Ca2+ responses in proprioceptor cell bodies To determine whether activating SGC Gq GPCR signaling is sufficient to modulate proprioceptor activity, we used ex vivo DRGs from another GFAP-hM3Dq::PV-Cre::GCaMP6 triple transgenic mouse model. In this model hM3Dq expression is under the control of GFAP promoter control (Fig. 1a), while GCaMP6f expression is under the control of parvalbumin (PV) promoter. In DRGs, parvalbumin is primarily found in proprioceptors . This model therefore e nables the vast majority of SGC s to be activated by the application of CNO (Fig. 1b) while monitoring Ca 2+ responses in ~97% of proprioceptors expressing GCaMP6f (Fig. 2a,c,d; Fig. S6 and Supplementary Table 1). Notably, while proprioceptors exhibited no spontaneous Ca2+ transients in DRG preparations , 10µM CNO bath application elicited Ca2+ increases in a subpopulation of proprioceptors (amplitude: 1.26 ± 0.13 ΔF/F0; rise time: 23.6 ± 2.2 s; duration: 93.1 ± 3.5 s; Fig. 2a,c,d and Supplementary Movie 2 & Table 5). Furthermore, upon 10 M CNO application, we observed that the onset of these proprioceptor Ca2+ responses (in GFAP-hM3Dq::PV-Cre::GCaMP6 mice) occurred 25.6 ± 2.5 s after the onset of Ca2+ elevations elicited in SGCs (in GFAP-hM3Dq::Cx43-CreERT2::GCaMP6 mice) (2-tailed unpaired t - test, P = 0.0002; Fig. 2a-d and Supplementary Movie 3 & Table 4). This result was confirmed in DRGs obtained from a fourth mouse model expressing hM3Dq in SGCs as well as GCaMP6f in both SGCs and proprioceptors (GFAP-hM3Dq::Cx43-CreERT2::PV-Cre::GCaMP6 quadruple transgenic; Fig. 3a). Calcium invariably increased in SGCs first , followed by proprioceptors with a similar delay , demonstrating that SGC Gq GPCR activation drives (or alters) the activity of a subpopulation of proprioceptors (2-tailed unpaired t-test, P < 0.0001; Fig. 3b-e; Fig. S6 and Supplementary Movie 4 & Table 4). In support of this conclusion, CNO never induced Ca2+ elevations in proprioceptors of control PV-Cre::GCaMP6 double transgenic mice, demonstrating that proprioceptor Ca2+ responses are not due to a non-specific effect of CNO directly on neurons (Fig. S7 and Supplementary Table 2). SGC-induced proprioceptor Ca2+ responses are mediated by purinergic receptors We next aimed to determine which receptor s were involved in this SGC-to-proprioceptor communication. Because functional ionotropic and metabotropic glutamate, GABA and ATP receptors are expressed at the plasma membrane of the soma of small -diameter DRG nociceptors7,15,16, we hypothesized that soma of large proprioceptive neurons may express similar receptors. Similar 2-photon Ca 2+ imaging experiments as previously described were performed using DRGs from GFAP-hM3Dq::PV-Cre::GCaMP6 mice. We found that concomitant bath application of 10 M CNO and inhibitors of glutamatergic (AMPAR, NMDAR, groups I, II, III mGluRs ), GABAergic (GABAAR, GABA BR) or purinergic (P2X3R and P2Y 1R) receptors onto DRGs did not substantially preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 4 depress the number of proprioceptors responding to CNO-induced SGC activation. Furthermore, proprioceptor Ca2+ response amplitude, rise time and duration were not significantly attenuated (Fig. 4a,b, Fig. S8 and Supplementary Table 5). However, co-applying 10 M CNO with the broad-spectrum purinergic receptor antagonist PPADS (100 M) reduced proprioceptor Ca2+ increase amplitude and rise time by ~34% and ~49% as shown in the left shifted cumulative probability distributions (0.83 ± 0.16 ΔF/F 0; 12.1 ± 1.3 s; Kolmogorov-Smirnov test: P = 0.0348, P = 0.0007; Fig. 4d,e), with no change in Ca2+ response duration (Fig. 4c, Fig. S8 and Supplementary Table 5). A decrease (~57% drop) in the total number of responsive proprioceptors was also observed (Fig. 4b). Taken together, these results suggest that proprioceptor purinergic receptors - other than P2X 3R and P2Y 1R previously reported in nociceptor s11,12 - and presumably ATP released by SGCs, mediate Ca2+ increases in proprioceptors. We therefore examined whether proprioceptor cell bodies could respond to ATP through functional purinergic receptors . Adding 50 M non-hydrolyzable ATPγS on DRGs from PV- Cre::GCaMP6 mice elicited Ca2+ responses in ~33% of proprioceptors as compared to CNO application (Fig. 4f and Supplementary Table 6 ). Thus, a subpopulation of proprioceptors express purinergic receptors at their cell body plasma membrane. The fact that nonhydrolyzable ATPγS did not activate as many proprioceptors as CNO did, may suggest that another transmitter (e.g. ADP via ATP hydrolysis) is required for the full effectiveness in SGC -to-proprioceptor activation . In agreement with this hypothesis, 30 M nonhydrolyzable ADPS induced Ca2+ responses in ~ 46% of proprioceptors as compared to CNO (Fig. S9 and Supplementary Table 6). Proprioceptor Ca2+ responses require both extracellular and intracellular Ca2+ We next investigated the possible contribution of extracellular Ca2+ to the CNO-induced/SGC-mediated proprioceptor responses. To this end, a Ca2+ free extracellular solution was used to perform 2 -photon imaging experiments, always using DRGs from GFAP-hM3Dq::PV-Cre::GCaMP6. We observed that the number of responsive proprioceptors w as not appreciably modified upon 10 M CNO application in an Ca2+ free extracellular solution (Fig. 4a,b), indicating that extracellular Ca2+ is not required for the induction of proprioceptor responses. However, the response amplitude, rise time and duration were markedly attenuated by 25%, 38% and 43%, respectively, suggesting that extracellular Ca2+ is required for reaching the largest proprioceptor responses (0.94 ± 0.12 ΔF/F 0; 14.6 ± 1.8 s; 53.0 ± 5.4 s; Kolmogorov-Smirnov test: P = 0.0047, P = 0.007 and Kruskal Wallis test: P < 0.0001; Figure 4a,c-e, Fig. S8 and Supplementary Table 5). The contribution of ext racellular Ca2+ was further substantiated by the observation of a ~41s delay in the proprioceptor response onset compared to the response onset obtained in regular extracellular solution condition (i.e. containing Ca2+) (Fig. S10 and Supplementary Table 4). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 5 Overall, our data are consistent with a model in which activation of both ionotropic P2XR (mediating Ca2+ entry from extracellular space) and metabotropic P2YR (mediating Ca2+ release from internal stores) contribute to the full proprioceptor cytosolic Ca2+ responses. The P2XR-mediated Ca2+ increase preceding a subsequent delayed metabotropic P2YR-mediated Ca2+ increase. Such type of synergistic interactions between P2X R and P2XR has been previously described in marrow megakaryocytes and blood platelets17.

The primary goal of this study was to address the role of SGC G q GPCR signaling in proprioceptor activity. Collectively, our findings rev eal a new mechanism of interaction between SGCs and proprioceptors involving, at least partially, purinergic P2XR and P2YR as well as ATP and/or ADP transmitters. Although scarce, studies have suggested that ATP release from SGCs results in the co-activation of neuronal P2Y1R and P2X3R and Ca2+ rises in nociceptive neurons, with P2Y1R exerting an inhibitory action on the function of P2X 3R4,5,10. However, our data show that these purinergic receptors are not involved in SGC -mediated proprioceptor Ca 2+ responses. Indeed, selective antagonists of these two receptors do not prevent or alter Ca2+ responses in proprioceptors (Fig. 4). Instead, our results suggest that SGCs exert control on proprioceptor activity via other purinergic receptors (blocked by pan PPADS broad-spectrum purinergic antagonist). Thus, the purinergic system appears to be a conserved mechanism for SGC communication with both nociceptors and proprioceptors, but to involve different purinergic signaling modalities depending on the type of sensory neurons. To the best of our knowledge, differential communication between SGCs and different types of sensory neurons h as never been documented before. Nevertheless, it represents a form of communication whereby SGCs could discern specific populations of sensory neurons and induce distinct control/alteration of the Ca2+ homeostasis of these neurons depending on the involved sensory modality or the physiopathological state. Additionally, the partial reduction of SGC-induced proprioceptor Ca2+ responses in the presence of PPADs can be explained by the fact that PPADS is a general inhibitor of purinergic receptors that does not inhibit all purinergic receptors. Therefore, it remains possible that residual proprioceptor Ca2+ activity is due to the lack of inhibition of certain purinergic receptors, or possibly to the indirect opening of plasma membrane Ca 2+ channels downstream of purinergic receptor activation. Elucidating these questions and mechanisms is beyond the scope of the present study. It will require using a combination of available purinergic receptor knockout animals and purinergic receptor and Ca2+ channel antagonists, which represent a considerable endeavor. Clear proof of expression of functional purinergic receptors on proprioceptor cell bodies has not yet been described. Our data showing that non-hydrolyzable ATPS and ADPS induce Ca2+ elevations preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 6 in proprioceptors indicate that proprioceptors express functional purinergic receptors at their soma level, although additional indirect scenario involving another cell type cannot be totally ruled out . Furthermore, a recent transcriptional profiling study has uncovered that mRNAs of purinergic ionotropic P2X5R and P2X 6R as well as metabotropic P2Y 14R are enriched in proprioceptors 18. Although the presence of these transcripts does not indicate that the corresponding receptors are expressed at the level of proprioceptor soma, P2X5R, P2X6R and P2Y14R represent interesting candidates potentially involved in SGC-to-proprioceptor communication. As mentioned above, further studies are necessary to address this hypothesis. Not only proprioceptors, but also low-threshold mechanoreceptors , express parvalbumin. However, our PV-Cre::GCaMP6 mouse model (in which parvalbumin promoter controls GCaMP6f expression) exhibits the presence of the Ca2+ indicator primarily in proprioceptors (Fig. S6), suggesting that it is a suitable model to detect intracellular Ca2+ homeostasis changes specifically in proprioceptors. Yet, the possibility that a few low-threshold mechanoreceptors also express GCaMP6f and their activity is included in our data sets cannot be excluded. Finally, irrespective of mechanism (s), our results raise the possibility that SGC Gq GPCR signaling-induced disruption of activity in a subpopulation of proprioceptors might be sufficient to modulate some type of proprioceptive information processing within DRGs and produce changes in sensorimotor behavior. In conclusion, our study is relevant to proprioceptive impairments as well as sensorimotor deficits and ataxia associated with spinal muscular atrophy or Friedreich’s ataxi a, respectively19–21

Animals Experiments were conducted on 2 - to 3 -month old male and female mice from the C57BL/6N background. Mice were grouped housed (5 mice/cage) and fed ad libitum. Illumination was controlled automatically with a 12/12h light-dark schedule. All experiments were conducted during the dark phase. Wildtype (WT) littermates were used as controls in experiments involving transgenic mice. The following mouse lines were used and/or generated (the mouse lines used in experiments appear in italic; Supplementary table 7): (1) PV-Cre22 (Jackson Laboratories, stock #017320) and Cx43 -CreERT223 mice were crossed with CAG-lox-STOP-lox-GCaMP624 (Jackson Laboratories, stock #024105) in order to obtain two new double transgenic mouse lines that we called PV-Cre::GCaMP6 and Cx43- CreERT2::GCaMP6, respectively . Th ese two lines were used for running control Ca 2+ imaging experiments to test the internees of CNO (e.g. to test whether or not CNO in itself evokes Ca2+ elevations preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 7 or modulates spontaneous Ca 2+ transients in these two mouse lines). GFAP-hM3Dq13 transgenic mice were used for Western blot, immunohistochemical and behavioral experiments. These mice were crossed with both PV-Cre::GCaMP6 and Cx43-CreERT2::GCaMP614 transgenic mice in order to obtain two new triple transgenic mouse line that we named GFAP-hM3Dq::PV-Cre::GCaMP6 and GFAP- hM3Dq::Cx43-CreERT2::GCaMP6, respectively. These two triple transgenic mouse lines were used for Ca2+ imaging experiments. Finally, we crossed GFAP-hM3Dq::Cx43-CreERT2::GCaMP6 with PV- Cre::GCaMP6 in order to obtain a new quadruple transgenic mouse line that we called GFAP- hM3Dq::Cx43-CreERT2::PV-Cre::GCaMP6. To induce GCaMP6f expression in Cx43- CreERT2::GCaMP614 and in GFAP-hM3Dq::Cx43-CreERT2::GCaMP6, mice were treated i.p. with tamoxifen (1mg/day, Sigma) diluted in corn oil (Sigma) during 10 days and used 15 days after the first day of treatment. Animal care and procedures were carried out according to the guidelines set out in the European Community Council Directives. Immunohistochemistry Animals were sacrificed 2min after treatment (CNO 1mg/kg or 0.9% NaCl, i.p.) and DRGs were immediately harvested and drop fixed in 4% paraformaldehyde for 4h prior to cryoprotection in 0.02M PBS containing 20% sucrose overnight at 4°C . For other immunohistochemistry experiments (e.g. transgenic mice characterization), animals were transcardiacally perfused with 4% paraformaldehyde under ketamine/xylazine (100mg/kg / 10mg/kg respectively, i.p.) anesthesia. L umbar L3, L4 and L5 DRGs were harvested, postfixed 2 h in 4% paraformaldehyde, and cryoprotected in 0.02M PBS containing 20% sucrose overnight at 4°C. Then, tissues were frozen in optimal cutting temperature compound (OCT). Fourteen µm sections were cut using a cryostat (Leica), mounted on Superfrost glass slides and stored at -80°C. The day of the experiment, sections were washed two times for 15 min each in 0.02M PBS. Sections were incubated overnight in 0.02M PBS (pH 7.4) containing 0.3% Triton X100, 0.02% sodium azide and primary antibodies (Supplementary table 8) at room temperature in a humid chamber. The following day, sections were washed 3 times in 0.02M PBS, and incubated for 2h at room temperature with secondary antibodies diluted in 0.02M PBS (pH 7.4) containing 0.3% Triton X100 and 0.02% sodium azide. Then, sections were washed 3 times for 15min in 0.02M PBS and mounted between slide and coverslip with Vectashield medium containing DAPI (Vector Laboratories). Negative controls, i.e. slices incubated with secondary antibodies only, were used to set criteria (gain, exposure time) for image acquisition for each experiment. Images acquisition was performed with an Axio Observer Z1 epifluorescence Zeiss microscope , an ORCA Flash 2.8 million pixel came ra, a PlanNeoFluar 20x/0.5NA objective, a LED COLIBRI 2 light source with 4 narrowband LED (365nm, 470nm,590nm, 625nm), as well as filters [DAPI (49), eGFP (38HE), Cy3 (43HE) and Cy5 (50)]. A Zeiss LSM710 confocal microscope with a Plan -apochromat 63x oil immersion objective (umerical aperture of 1.4) was also used. The same image acquisition settings were used for the negative control preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 8 slides and the experimental slides. Images were extracted using the ZEN 2011 blue edition software (Zeiss). Cell c ounting and mean grey value measurements were performed using ImageJ software software from the National Institute of Health (USA). Because it was difficult to discriminate individual SGCs within all SGCs surrounding a single neuronal cell body, “rings” surrounding neuronal cell bodies were quantified. For confocal imaging of GFAP -hM3Dq DRG sections, 12µm thick z -stacks were acquired with 0.3µm steps (40 z-stacks in total). 3D visualization was done using the 3D viewer plugin of Fiji. Raw data were analyzed and quantified. Western Blots Animals were sacrificed 2min, 30min and 4h after treatment (CNO 1mg/kg i.p.) and L3, L4, L5 DRGs were dissected, frozen and stored at -80°C. The tissues were homogenized in 150μL RIPA buffer (50mM Tris pH 8, 150mM NaCl, 1% NP-40, 0,5% deoxycholate, 0,1% SDS, 1 mM sodium orthovanadate) with 1X protease inhibitor cocktail cOmplete (Roche) and 1X Halt phosphatase inhibitor cocktail (Thermo Scientific). Tissue lysate was obtained using a Bioruptor sonication system (Diageno de) and then centrifugated at 2000 RPM for 5min. S upernatant was kept and protein concentration was determined using a BCA assay (Bio-Rad). Aliquots of 25µg of protein for each mouse were deposited and run on 10% acrylamide SDS -page and transferred to nitr ocellulose membranes. Membranes were cut (according to protein weight), and the pieces of membranes were then saturated in TBS -Tween (TBS- T) containing 5% fat-free milk for 30min. Membranes were then incubated overnight at 4°C in TBS -T containing primary antibodies (Supplementary table 8). The day after, membranes were washed in TBS- T and then incubated with horse radish peroxidase ( HRP)-conjugated secondary antibodies diluted in TBS-T ( Supplementary table 8) for 1h30 at room temperature. Clarity ECL chemiluminescence detection (Biorad) and ImageQuant LAS4000 (GE Healthcare Life Sciences) were used to reveal and visualize the proteins. The average exposure time was 30s and images were taken with 1s increments. The images obtained before saturation of the signal were used for quantification. The mean gray values corresponding to the signal were measured using ImageJ software. Western blot experiments were replicated 4 times to minimize the technique variability and the 4 mean grey values obtained per sample were then averaged. Then each average value was normalized to the values obtained for the control group. Furthermore, to average away any possible position effect, the order of the deposits was different in each replicate. Two photon Ca2+ imaging Acute intact DRG preparations were prepared from GFAP -hM3Dq::PV-Cre::GCaMP6, GFAP - hM3Dq::Cx43CreERT2::GCaMP6, and GFAP -hM3Dq::PV-Cre::Cx43-CreERT2::GCaMP6 mouse lines. PV-Cre::GCaMP6 and Cx43CreERT2::GCaMP6 mice were also used to perform prior control experiments to test the inertness of CNO. Vertebras and dura mater were removed and L4 and L5 were preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 9 exposed under a dissecting stereomicroscope SZ61 -45TR (Olympus) and immediately covered by ice cold (slushy) incubation ACSF solution containing (in mM): 95 NaCl, 1.8 KCl, 1.2 KH2PO4, 0.5 CaCl2, 7 MgSO4, 26 NaHCO3, 15 glucose, and 50 sucrose, oxygenated with 95% O2 and 5% CO2. DRGs were harvested and incubated at 35°C for 30min in incubation solution and then left to recover for 1h30 at room temperature. A single DRG was placed in the recording chamber of a custom -built two-photon laser-scanning microscope with a 20x water immersion objective (x2 0/0.95w XLMPlanFluor, Olympus). GCaMP6f was excited at 920nm with a Ti:Sapphire laser (Mai Tai HP; Spectra -Physics). DRGs were continuously superfused with oxygenated recording solution identical to the incubation solution except for the following (in mM): 127 NaCl; 2.4 CaCl2; 1.3 MgSO4; and 0 sucrose at a rate of 4 ml/min. For experiments in Ca 2+-free ACSF, the recording solution was identical to incubation solution except the following (in mM) : 127 NaCl; 0 CaCl2; 1.3 MgSO4; and 0 sucrose. Image acquisition was performed at a rate of 1 image/second (1Hz). Drugs applied are detailed in Supplementary table 8. To determine the viability of proprioceptors GFAP-hM3Dq::PV-Cre::GCaMP6 mice, KCl (50mM) was applied as a positive control at the end of every experiment ( Supplementary Movie 5); the number of proprioceptors responding to KCl represented the t otal number of proprioceptors indicated in Supplementary Tables 4, 5, 6 . To determine the viability of SGCs in GFAP- hM3Dq::Cx43CreERT2::GCaMP6 and Cx43CreERT2::GCaMP6 mice, we considered SGCs responding to CNO and/or agonist cocktail to endogenous Gq GPCRs (50 M DHPG, 10 M histamine, 10 M carbachol, 50 M ATP-S, 1 M adenosine, 200 M glutamate) as the total number of alive SGCs. Agonist cocktail was systematically applied at the end of every experiment. ImageJ and Metamorph (Molecular Devices) so ftwares were used to analyze the data. Regions of interest were determined on GCaMP6f -expressing cells and the relative changes in fluorescence ( F/F0) were calculated as the ratio of the fluorescence intensity to that recorded before any drug application . In GFAP-hM3Dq::Cx43-CreERT2::PV-Cre::GCaMP6 mice, in which GCaMP6f expression is driven both in SGCs and proprioceptors, the two cell types were discriminated by their distinct morphological features and GCaMP6f brightness differences (indeed SGCs express ed more GCaMP6f than proprioceptors and thus were brighter than proprioceptors in basal conditions). Positive responses were defined as those that exceeded 3 standard deviations (SD) above the baseline level. To determine the onset of Ca2+ responses, the time for the drug-containing ACSF to reach the recording chamber through the tubing was subtracted . Rise times were calculated as the 10 -90% rise time. The designs of the different Ca2+ imaging experiments are presented in Supplementary Fig.11. Statistics Data were acquired and analyzed blind of genotype and treatment. Data are shown as mean ± S.E.M. n values correspond to the number of cells, DGR, or mice. All statistical tests were performed after verification of normal data distribution using D’Agostino and Pearson omnibus normality test and, when applicable, equality of variances with F -test (Fisher). If normality assumptions were not met, we used preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 10 non-parametric tests (two-tailed Mann-Whitney for analyzing 2 groups and Kruskal-Wallis followed by Dunn's multiple comparison test for analyzing more than 2 groups). If normality was met, we used parametric tests (two -tailed Student t -test for analyzing 2 groups and one -way ANOVA followed by Tukey post -hoc for analyzing more than 2 groups). Kolmogorov -Smirnov t est was used to analyze cumulative distributions of pairs of groups. All statistical tests were performed using GraphPad Prism 6. No statistical methods were used to pre-determine sample sizes. Author disclosure statement The authors declare that they have no conflict of interest. Author contributions C.A. and Y.R. designed and interpreted experiments . Y.R. performed experiments and data analysis. Y.R. wrote the first draft of the manuscript. C.A. edited and wrote the manuscript. C.A. conceived and supervised the project. Funding Research in the authors’ laboratories is supported by grants from Fonds de dotation Neuroglia , NeurATRIS Innovation for Translational Neuroscience , French Friedreich’s Ataxia Association (A.F.A.F), and Ile-de-France Regional Council to C.A. Initial experiments were supported by a starting grant (Chair of Excellence) from the Foundation Ecole des Neurosciences de Paris (ENP) to C.A. Y.R. was a recipient of a master 2 and a PhD fellowships from the Institute of Neuroscience and Cognition (Université Paris Descartes) and ED158 doctoral school, respectively. Acknowledgments We gratefully acknowledge C. Steinhäuser for providing the Cx43-CreERT2 mouse line; S. Antoine and S. Guinoiseau for animal care; J.M. Andrieu, F. Charbonnier, B. Delhomme, P. Djian, C. Levenes, C. Meunier and M. Oheim for sharing pieces of equipment and laboratory spaces; O. Biondi and E. Schmidt for advice and support on image acquisition; A. Bessis and S. Dieudonné for valuable discussions and feedback; T. Fiacco and J. Stinnakre for critical reading and I. Melnychuk for editing the manuscript; and both the imaging and mouse core facilities, which are supported and funded by CNRS, INSERM and Université Paris Cité. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 11

1. Proske, U. & Gandevia, S. C. The Proprioceptive Senses: Their Roles in Signaling Body Shape, Body Position and Movement, and Muscle Force. Physiological Reviews 92, 1651–1697 (2012). 2. Riemann, B. L. & Lephart, S. M. The Sensorimotor System, Part I: The Physiologic Basis of Functional Joint Stability. J Athl Train 37, 71–79 (2002). 3. Huang, L.-Y. M. & Neher, E. Ca2+-Dependent Exocytosis in the Somata of Dorsal Root Ganglion Neurons. Neuron 17, 135–145 (1996). 4. Zhang, X., Chen, Y., Wang, C. & Huang, L.-Y. M. Neuronal somatic ATP release triggers neuron–satellite glial cell communication in dorsal root ganglia. Proc. Natl. Acad. Sci. U.S.A. 104, 9864–9869 (2007). 5. Suadicani, S. O. et al. Bidirectional calcium signaling between satellite glial cells and neurons in cultured mouse trigeminal ganglia. Neuron Glia Biol 6, 43–51 (2010). 6. Hanani, M. Satellite glial cells in sensory ganglia: from form to function. Brain Research Reviews 48, 457–476 (2005). 7. Kung, L.-H. et al. Evidence for Glutamate as a Neuroglial Transmitter within Sensory Ganglia. PLoS ONE 8, e68312 (2013). 8. Gerevich, Z., Müller, C. & Illes, P. Metabotropic P2Y1 receptors inhibit P2X3 receptor- channels in rat dorsal root ganglion neurons. European Journal of Pharmacology 521, 34–38 (2005). 9. Carlton, S. M. & Hargett, G. L. Colocalization of metabotropic glutamate receptors in rat dorsal root ganglion cells. J of Comparative Neurology 501, 780–789 (2007). 10. Yokoyama, T., Fukuzumi, S., Hayashi, H., Nakamuta, N. & Yamamoto, Y. GABA-mediated modulation of ATP-induced intracellular calcium responses in nodose ganglion neurons of the rat. Neuroscience Letters 584, 168–172 (2015). 11. Huang, L.-Y. M., Gu, Y. & Chen, Y. Communication between neuronal somata and satellite glial cells in sensory ganglia: Soma-Satellite Cell Communication. Glia 61, 1571–1581 (2013). 12. Takeda, M. et al. Enhanced excitability of nociceptive trigeminal ganglion neurons by satellite glial cytokine following peripheral inflammation. Pain 129, 155–166 (2007). 13. Agulhon, C. et al. Modulation of the autonomic nervous system and behaviour by acute glial cell Gq protein-coupled receptor activation in vivo. J Physiol 591, 5599–5609 (2013). 14. Rabah, Y., Rubino, B., Moukarzel, E. & Agulhon, C. Characterization of transgenic mouse lines for selectively targeting satellite glial cells and macrophages in dorsal root ganglia. PLoS One 15, e0229475 (2020). 15. Chen, Y. et al. Activation of P2X7 receptors in glial satellite cells reduces pain through downregulation of P2X3 receptors in nociceptive neurons. Proc. Natl. Acad. Sci. U.S.A. 105, 16773– 16778 (2008). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 12 16. Sung, K.-W., Kirby, M., McDonald, M. P., Lovinger, D. M. & Delpire, E. Abnormal GABA A Receptor-Mediated Currents in Dorsal Root Ganglion Neurons Isolated from Na–K–2Cl Cotransporter Null Mice. J. Neurosci. 20, 7531–7538 (2000). 17. Vial, C., Rolf, M. G., Mahaut‐Smith, M. P. & Evans, R. J. A study of P2X 1 receptor function in murine megakaryocytes and human platelets reveals synergy with P2Y receptors. British J Pharmacology 135, 363–372 (2002). 18. Chiu, I. M. et al. Transcriptional profiling at whole population and single cell levels reveals somatosensory neuron molecular diversity. eLife 3, e04660 (2014). 19. Mentis, G. Z. et al. Early Functional Impairment of Sensory-Motor Connectivity in a Mouse Model of Spinal Muscular Atrophy. Neuron 69, 453–467 (2011). 20. Fletcher, E. V. et al. Reduced sensory synaptic excitation impairs motor neuron function via Kv2.1 in spinal muscular atrophy. Nat Neurosci 20, 905–916 (2017). 21. Lynch, D. R., Deutsch, E. C., Wilson, R. B. & Tennekoon, G. Unanswered Questions in Friedreich Ataxia. J Child Neurol 27, 1223–1229 (2012). 22. Hippenmeyer, S. et al. A Developmental Switch in the Response of DRG Neurons to ETS Transcription Factor Signaling. PLoS Biol 3, e159 (2005). 23. Eckardt, D. Functional role of connexin43 gap junction channels in adult mouse heart assessed by inducible gene deletion. Journal of Molecular and Cellular Cardiology 36, 101–110 (2004). 24. Madisen, L. et al. Transgenic Mice for Intersectional Targeting of Neural Sensors and Effectors with High Specificity and Performance. Neuron 85, 942–958 (2015). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Legends Figure 1. hM3Dq allows selective activation of Gq GPCR signaling cascades in DRG SGCs. a, Immunohistochemistry (IHC) in DRGs of GFAP-hM3Dq mice and confocal fluorescent images showing HA -tagged hM3Dq protein (green), sensory neurons labeled with neurofilament antibody (NF, red, top panel , asterisks ) and SGCs labeled with GLAST antibody (red, bottom panel , arrowheads). Nuclei are stained with DAPI (blue). b, Two -photon Ca 2+ imaging o f ex vivo intact DRGs from GFAP-hM3Dq::Cx43-CreERT2::GCaMP6 mouse model . Left pane l showing SGCs expressing GCaMP6f, outlined areas of interest (I-III) and Ca 2+ increases during baseline , CNO application , and wash . Asterisk denotes neuronal cell soma. Right panel showing time course of CNO-mediated Ca 2+ increases in SGCs (I-III). c, Calcium imaging of DRGs from Cx43 - CreERT2::GCaMP6 mice. Left panel showing cells expressing GCaMP6f, outlined areas of interest (i-iii), and Ca 2+ increases during baseline , CNO application , wash  and cocktail application . Right panel showing time course of Ca 2+ traces in SGCs (i-iii). A cocktail of ligands to endogenous G q GPCR (1µM a denosine, 50µM ATP-S, 10µM c arbachol, 50µM DHPG, 200µM glutamate, 10µM h istamine) has been applied after CNO washing to ensure SCG viability. d, Schematic summarizing hM3Dq functionality in DRG SGCs. e,f, Western blot (WB) (e) and IHC (f) showing an increase of 76% ( e) and 60% ( f) of pERK expression level in GFAP -hM3Dq mice vs. control groups, 2min after in vivo CNO treatment (1mg/kg i.p.) (n = 9-10 animals/group for WB; n = 4 animals/group for IHC; two-tailed unpaired t-test for WB: P = 0.0006; Kruskal-Wallis test followed by Dunn’s multiple comparison test for IHC : P = 0.0231). GAPDH was used as a loading control in WB. *P < 0.05, ***P < 0.001; error bars indicate mean ± SEM. Figure 2. CNO/hM3Dq-mediated SGC activation induces proprioceptive neuron Ca2+ responses. a, Two-photon Ca 2+ imaging of ex vivo DRGs from GFAP-hM3Dq::Cx43-CreERT2::GCaMP6 (top panel) and GFAP-hM3Dq::PV-Cre::GCaMP6 mouse lines (bottom panel) during baseline , CNO application , and wash , and outlined areas of interest (i-iii for SGCs and I-III for proprioceptors). b,c, Time course of CNO -mediated Ca 2+ increases in SGCs from GFAP-hM3Dq::Cx43- CreERT2::GCaMP6 mice (b, i-iii) and proprioceptors from GFAP-hM3Dq::PV-Cre::GCaMP6 mice (c, I -III). d, Quantification of the onset of Ca2+ transients in SGCs and proprioceptors after bath application of CNO (n = 86 SGCs from 6 DRGs and 3 GFAP -hM3Dq::Cx43-CreERT2::GCaMP6 mice; n = 49 proprioceptors from 23 DRGs and 14 GFAP-hM3Dq::PV-Cre::GCaMP6 mice; Kruskal- Wallis test followed by Dunn’s multiple comparison test : P<0.0001; See Supplementary Table 4 for details). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Figure 3 . Direct d emonstration that CNO/hM3Dq-induced activation of SGCs drives proprioceptor Ca 2+ responses occurring ~20 s later within same DRGs . a,b, Two-photon Ca2+ imaging of ex vivo DRGs from GFAP -hM3Dq::Cx43-CreERT2::PV-Cre::GCaMP6 quadruple transgenic mouse model, expressing hM3Dq in SGCs and GCaMP6f Ca2+ indicator in both SGCs and proprioceptors (a), during baseline , CNO application , and wash , and outlined areas of interest (SCG & proprioceptor) (b). The discrimination of SGCs and neurons is based on both morphological features and intensity of GCaMP 6f brightness (left, black & white ). c, Time course of CNO-mediated Ca2+ increases in SGC (purple) and neighboring proprioceptor (orange), and time axis enlargement. d, Quantification of the data showing that in pairs of SGCs and associated proprioreceptors, Ca2+ transients occuring in proprioreceptors invariably appears ~20 s after CNO/hM3Dq-induced Ca2+ elevations in the surrounding SCGs (n=9 pairs each consisting of 1 SGC and 1 proprioceptor soma, n=5 DRGs, n=3 GFAP-hM3Dq::Cx43-CreERT2::PV-Cre::GCaMP6 mice. Two-tailed unpaired t -test: P = 0.0002. See Supplementary Table 4 for details ). e, Schematic summarizing the hM3Dq-mediated and SGC-induced Ca2+ responses in neighboring proprioceptors. ***P < 0.001; error bars indicate mean ± SEM. Figure 4. SGC Gq GPCR signaling induces Ca2+ responses in proprioceptors through purinergic receptors. a, Time course of CNO-induced Ca2+ increases in proprioceptors from GFAP-hM3Dq::PV- Cre::GCaMP6 mice following concomitant bath applications of CNO and antagonists to endogenous glutamatergic, GABAergic or purinergic receptors or to an application of CNO in extracellular Ca 2+ free solution. b, Number of responding proprioceptors; the marked 57% decrease in the number of responsive proprioceptors when CNO is applied in presence of PPADS is not statistically significant (2-photon Ca2+ imaging data normalized to CNO condition; n = 49 DRGs from19 mice for CNO group and n = 16 -19 DRGs from 5-6 mice for the other groups; Kruskal -Wallis test followed by Dunn’s multiple comparison test: P = 0.43; See Supplementary Table 5 for details). c, Duration of CNO/SGC- mediated Ca2+ increases in proprioceptors showing a significant decrease when CNO is applied in extracellular Ca 2+ free solution (Kruskal-Wallis test followed by Dunn’s multiple comparison test : P<0.0001). d,e, Cumulative frequency distribution of proprioceptor Ca2+ response amplitudes and rise times (Kolmogorov-Smirnov tests; Amplitude: P = 0.0097 for CNO vs. “ CNO + PPADS” and P = 0.0014 for CNO vs. “CNO in Ca2+ free ACSF”; Rise time: P < 0.0001 for CNO vs. “ CNO + PPADS” and P = 0.0016 for CNO vs. “CNO in Ca 2+ free ACSF”). f, Two-photon Ca2+ imaging of DRGs from GFAP-hM3Dq::PV-Cre::GCaMP6, outlined areas of interest (I -III, right panel) , and time course of ATP-S (50µM) -mediated Ca 2+ increases in proprioceptors during baseline , ATP-S application , and wash . **P < 0.01, ***P < 0.001; error bars indicate mean ± SEM . Antagonist cocktails used and corresponding target receptors: [200 µM PPADS (P2R)];[100 M AIDA (mGluR I) + 10 M NBQX (AMPAR) + 50 M AP5 (NMDAR) + 0.5 M MRS2279 (P2Y 1R)]; [50 M Bicuculline preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint (GABAAR) + 10 M RO-3 (P2X3R)]; [100 M MSOP (mGluR II) + 1 M LY341495 (mGluR III)] ; and [100 M CGP35348 (GABABR)]. **P < 0.01, ***P < 0.001; error bars indicate mean ± SEM. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Figure 1 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Figure 2 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Figure 3 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Figure 4 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 1 Supplementary information Supplementary Figure 1 (related to Fig. 1a). Cellular characterization of GFAP-hM3Dq mouse line. Immunohistochemical visualization of HA- tagged hM3Dq protein (HA, red) and Na/K-ATPase neuronal plasma membrane marker (green) in DRGs from GFAP-hM3Dq mouse. Top panel, Single plan imaged with a confocal microscope. Bottom panel, Since the distance between SGC and neuronal plasma membranes is about 20nm, i.e. beyond confocal microscope resolution , z-optical sections were stacked to explore possible colocalization between HA and Na/K-ATPase staining: 40 z-optical sections, 12 µm each, with a step of 0.3 µm were stacked to obtain a maximum intensity projection. Areas expressing HA and Na/K-ATPase are different, showing that HA-tagged hM3Dq is not expressed at the plasma membrane of sensory neurons. Nuclei of HA-expressing SGCs (stained in blue with DAPI , arrowheads) appear small with a bright DAPI labeling. This is different from neuronal nuclei, which are large with a dim DAPI staining (asterisks). Supplementary Figure 2 (related to Fig. 1b,c) Cellular characterization of Cx43-CreERT2::GCaMP6 double transgenic mouse line. Top panel, Confocal images of i mmunohistochemical experiments showing GCaMP6f (green) overlapping with SGC GLAST marker (red). Bottom panel, GCaMP6f was not detected in neurofilament-expressing cell bodies of sensory neurons (red). Arrowheads and asterisks denote GLAST -expressing SGCs and neurofilament-expressing neurons, respectively. SGC and neuronal cell bodies are stained with DAPI (blue). Supplementary Figure 3 (related to Fig. 1b,c) Inertness of CNO on the frequency of spontaneous Ca2+ events in SGCs and absence of hM3Dq receptor constitutive activity . In Cx43-CreERT2::GCaMP6 double transgenic mouse line, spontaneous Ca 2+ elevations occurred in ~12% of SGCs. a, Schematic of experimental design. b, Calcium traces showing the pattern of expression of spontaneous Ca2+ events compared to CNO-evoked Ca2+ elevations in SGCs . c, Bath application of 10 M CNO to ex vivo DRGs from Cx43- CreERT2::GCaMP6 mice does not modulate (increase or decrease) the frequency of spontaneous Ca2+ elevations compared to baseline (2min before CNO application). This result indicates that CNO has no non-specific effect in itself in agreement with the data reported in Fig. 1c. Both GFAP-hM3Dq::Cx43- CreERT2::GCaMP6 and Cx43 -CreERT2::GCaMP6 control mice, exhibit similar frequency in SGC preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 2 spontaneous Ca2+ elevations during baseline, suggesting that hM3Dq receptor has no constitutive activity in itself. Data quantification: Cx43-CreERT2::GCaMP6 mice, during CNO applicati on: 0.67 event/min; Cx43-CreERT2::GCaMP6 mice, during baseline: 0.63 events/min ; GFAP-hM3Dq::Cx43- CreERT2::GCaMP6 mice, during baseline: 0.66 event/min. n=24 cells from Cx43-CreERT2::GCaMP6 mice, n=31 cells from GFAP-hM3Dq::Cx43-CreERT2::GCaMP6 mice. Kruskal-Wallis test followed by Dunn’s Multiple Comparison Test: P= 0.8503. Supplementary Figure 4 (related to Fig. 1e) Western blot showing the timeline of MAPK/ERK pathway activation downstream of hM3Dq stimulation in DRGs in vivo . Top panel , E xperimental design. GFAP-hM3Dq and WT littermate control mice were treated intraperitoneally with 1mg/kg CNO, sacrificed 2min, 30min or 4h after treatment, and the L3, L4 and L5 DRGs (in total 6 DRGs per mouse) were dissected and processed for Western blot experiments. Note that the average time intervals between the dissection and freezing procedure of the 1st DRG and the 6th DRG were ~1min 30s and ~3min 40s, respectively. Middle panel, Western blot showing the pattern of expression of activated ERK1/2 (pERK). Bottom panel , Quantification of Western blot data coming from 4 replicates. CNO induces a ~7 6% increase of pERK1/2 in DRGs 2min after treatment of GFAP-hM3Dq mice as compared to WT mice (n=10 GFAP- hM3Dq mice, n=10 WT mice, Two -tailed unpaired t test: P = 0.0006). The level of pERK1/2 is not different from control levels 30min and 4h after treatment. Corresponding detailed quantification of the data is found in Supplementary Table 3. Supplementary Figure 5 (related to Fig. 1f) Immunofluorescence and confocal microscopy to study the expression of pERK in SCGs 2 min after 1 mg/kg CNO or saline i.p. treatment. Three groups of mice were used: CNO-treated WT mice, saline-treated-GFAP-hM3Dq mice and CNO -treated GFAP-hM3Dq mice. Mice were sacrificed 2min after treatment and their DRGs (L3, L4 and L5) were dissected, drop fixed in 4% paraformaldehyde and processed for immunohistochemistry. Top panel, confocal images showing pERK expression pattern. Note that pERK immunoreactivity is very low in SGCs from CNO-treated WT mice and saline-treated GFAP-hM3Dq mice as well in sensory neurons from the 3 groups of mice, contrary to SGCs from CNO- treatedGFAP-hM3Dq mice in which pERK immunoreactivity is high (green circles surrounding neuron soma, right). Bottom panel, Quantification of the data in large (~40-50 m diameter) sensory neurons (as delineated by dashed lines in top panel). CNO treatment does not induces an increase of pERK1/2 in sensory neuron soma 2min after CNO treatment in GFAP-hM3Dq mice as compared to saline-treated GFAP-hM3Dq and CNO-treated WT mice (n=4 CNO-treated GFAP-hM3Dq mice, n=4 saline-treated preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 3 GFAP-hM3Dq mice, n=4 CNO-treated WT mice; Kruskal Wallis test: P = 0.5836). Corresponding detailed quantification of cellular characterization is found in Supplementary Table 3. Supplementary Figure 6 (related to Fig. 2a) Characterization of the PV-Cre::GCaMP6 double transgenic mouse line showing that GCaMP6f protein (green) is expressed in ~97% of parvalbumine (PV)-expressing proprioceptors (red) in DRGs (n=3 mice, n=6 slices 2 slices/mouse, n=37 PV-expressing proprioceptors). Cell nuclei are stained with DAPI (blue) . One n euronal cell body is marked with asterisks. Images were acquired using an epifluorescence microscope. Corresponding detailed quantification of cellular characterization is found in Supplementary Table 1. Supplementary Figure 7 (related to Fig. 2a, c) Inertness of CNO on proprioceptor activity. a, Experimental design. Two-photon Ca2+ imaging of ex vivo intact DRGs from PV -Cre::GCaMP6 double transgenic mice. b, Proprioceptor expressing GCaMP6f, outlined areas of interest (I, II) and Ca2+ increases during baseline , CNO application , wash  and cocktail application . c, Time course of Ca2+ traces in proprioceptors (I, II). A cocktail of ligands to endogenous G q GPCR (1µM adenosine, 50µM ATP -S, 10µM carbachol, 50µM DHPG, 200µM glutamate, 10µM histamine) has been applied after CNO washing to ensure proprioceptor viability. Corresponding detailed quantification of the data is found in Supplementary Table 2. Supplementary Figure 8 (related to Fig. 3a-d) Pharmacological characterization of the receptors involved in CNO/hM3Dq/SGC-induced Ca 2+ elevations in proprioceptors. Relative amplitude (top panel), rise time ( middle panel) and duration (bottom panel) of the proprioceptor Ca2+ elevations in presence of CNO only or CNO combined with various transmitter receptor inhibitors alone or mixed: (1) 10 M CNO (white), (2) 10 M CNO + 100 M PPADS (broad spectrum blocker of P2XRs & P2YRs) (orange), (3) 10 M CNO + 100 M AIDA (blocker of mGluR I) + 10 M NBQX (blocker of AMPAR) + 50 M AP5 (blocker of NMDAR) + 0.5 M MRS2279 ( blocker of P2Y1R) (dark blue) ; (4) 10 M CNO + 50 M Bicuculline ( blocker of GABAAR) + 10 M RO-3 (blocker of P2X3R) (blue); (5) 10 M CNO + 100 M MSOP (blocker of mGluR II) + 1 M LY341495 ( blocker of mGluR III) (red); (6) 10 M CNO + 100 M CGP35348 (blocker of GABABR) (yellow); and (7) 10 M CNO applied in Ca 2+ free extracellular solution (light preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 4 blue). Left column, histograms showing the quantification of the data normalized to CNO condition. Right columns, cumulative frequency distribution of proprioceptor response amplitude, rise time and duration (raw data), showing that the presence of PPADS or the absence of extracellular Ca 2+ leads to decreases in the amplitude, rise-time and duration of the proprioceptor Ca2+ transients. Kruskal-Wallis and Kolmogorov-Smirnov tests were used: **P < 0.01, ***P < 0.001; error bars indicate mean ± SEM. Corresponding detailed quantification and statistics are found in Supplementary Table 5. Supplementary Figure 9 (related to Fig. 4f) Proprioceptors express functional ADP receptors at their soma level. a, Experimental design. Two- photon Ca2+ imaging of ex vivo DRGs from PV-Cre::GCaMP6 double transgenic mice. b, Left panel, Cells expressing GCaMP6f, outlined areas of interest (I, II) and Ca2+ increases during baseline , ADP- S application  and wash . Right panel showing time course of Ca 2+ traces in proprioceptors (I, II). Corresponding detailed quantification of the data is found in Table 6. Supplementary Figure 10 (related to Figs. 2a-d and 4a-e) CNO/hM3Dq/SGC-induced Ca 2+ responses in proprioceptors are delayed in absence of Ca 2+ in the extracellular artificial cerebral fluid solution (ACSF). a, Experimental design. Two-photon Ca2+ imaging of ex vivo DRGs from GFAP -hM3Dq::Cx43-CreERT2::GCaMP6 and GFAP -hM3Dq::PV- Cre::GCaMP6 transgenic mouse lines expressing hM3Dq in SGCs and GCaMP6f either in SGCs or in proprioceptors, respectively. b, quantification of the data showing that Ca2+ elevations in SCGs (light blue circles, GFAP-hM3Dq::Cx43-CreERT2::GCaMP6 mice) occurs first, ~25s before the induction of Ca2+ responses in proprioceptors (orange circles, GFAP-hM3Dq::PV-Cre::GCaMP6 mice) when CNO is applied in Ca2+-containing ACSF (see Fig. 2d for details). However, the onset of proprioceptor Ca2+ responses is delayed by ~41 s (orange triangles , GFAP-hM3Dq::PV-Cre::GCaMP6 mice mice) when 10 M CNO is applied in Ca2+ free ACSF as compared to the onset of the proprioceptor responses when CNO is applied in Ca2+-containing ACSF (Ca2+-containing ACSF: n=49 proprioceptors, n=23 DRGs, n=14 GFAP-hM3Dq::PV-Cre::GCaMP6 mice; Ca2+ free ACSF: =14 proprioceptors, n=15 DRGs, n=6 GFAP-hM3Dq::PV-Cre::GCaMP6 mice; Kruskal-Wallis test followed by Dunn’s multiple comparison test: P < 0.0001). Corresponding detailed quantification and statistics are found in Supplementary Table 4. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 5 Supplementary Figure 11 (related to Figs. 1-4 and Material & Methods) Designs of the different Ca2+ imaging experiments reported in the main manuscript and supplementary information document. Supplementary Table 1 Detailed data and statistics relative to cellular characterization of GFAP-hM3Dq and PV-Cre::GCaMP6 mouse lines (immunohistochemistry experiments related to Fig. 1, Fig. 2 and Supplementary Fig. 6). Supplementary Table 2 Detailed data and statistics relative to functional characterization of GFAP-hM3Dq::Cx43- CreERT2::GCaMP6, Cx43-CreERT2::GCaMP6 and PV-Cre::GCaMP6 mouse lines (2 -photon Ca 2+ imaging experiments related to Fig. 1, Fig. 2 and Supplementary Fig. 8). Supplementary Table 3 Detailed data and statistics relative to functional characterization of GFAP-hM3Dq mouse line (Western blot and immunohistochemistry experiments to reveal pERK expression levels; related to Fig. 1 and Supplementary Fig. 4). Supplementary Table 4 Detailed data and statistics relative to the discovery that t he onset of CNO/hM3Dq-induced Ca 2+ elevations in SGCs occur invariably ~ 20-25 s before proprioceptor Ca 2+ responses using GFAP- hM3Dq::Cx43-CreERT2::GCaMP6, GFAP-hM3Dq::PV-Cre::GCaMP6, and GFAP -hM3Dq::Cx43- CreERT2::PV-Cre::GCaMP6 transgenic mouse lines (2-photon Ca 2+ imaging experiments related to Fig. 2, Fig. 3 and Supplementary Fig. 8 & Fig. 10). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 6 Supplementary Table 5 Detailed data and statistics relative to the discovery that purinergic P2XR and P2YR underlie CNO/SGC-mediated proprioceptor Ca2+ responses using GFAP-hM3Dq::PV-Cre::GCaMP6 mouse line (2-photon Ca2+ imaging experiments related to Fig. 4 and Supplementary Fig. 8). Supplementary Table 6 Detailed data and statistics relative to the discovery that AT PS and ADPS induce Ca2+ increases in proprioceptor using PV-Cre::GCaMP6 mouse line (2-photon Ca2+ imaging experiments related to Fig. 4 and Supplementary Fig. 9). Supplementary Table 7 List of single, double, triple and quadruple transgenic mouse lines used in the study. Supplementary Table 8 List of antibodies , drugs and reagents as well as corresponding concentrations used in the study. Abbreviations: IHC, immunohistochemistry; WB, Western blot. Supplementary Movie 1 (related to Fig. 1a and Supplementary Fig.1) 360° rotation of z-stack acquisition corresponding to Supplementary Fig. 1. HA in red, Na/K-ATPase in green. Supplementary Movie 2 (related to Fig. 2) 10 M CNO application to ex vivo DRG from GFAP-hM3Dq::PV-Cre::GCaMP6 triple transgenic mice leads to Ca2+ responses in proprioceptors. CNO is applied at frame 120 until frame 240 (2-photon data acquisition rate: 1frame/s). Movie speed: 100 frames per second. AVI files were made with ImageJ. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint 7 Supplementary Movie 3 (related to Fig. 2) 10 M CNO application to ex vivo DRG from GFAP -hM3Dq::Cx43-CreERT2::GCaMP6 triple transgenic mice induces Ca 2+ elevations in SGCs. CNO is applied at frame 120 until frame 240 (2 - photon data acquisition rate: 1frame/s). Movie speed: 100 frames per second. AVI files were made with ImageJ. Supplementary Movie 4 (related to Fig. 3) 10 M CNO application to DRG from GFAP-hM3Dq::Cx43-CreERT2::PV-Cre::GCaMP6 quadruple transgenic mice shows Ca2+ responses occurring first in SGCs and then in neighboring proprioceptors. CNO is applied at frame 120 until frame 240 (2 -photon data acquisition rate: 1frame/s). Movie speed: 100 frames per second. AVI files were made with ImageJ. Supplementary Movie 5 (related to Material & Methods – Calcium imaging section) 50mM KCl application to DRG from GFAP-hM3Dq::PV-Cre::GCaMP6 triple transgenic mice was used systematically as a positive control to determine proprioceptor viability (i.e. proprioceptors responding to KCl are considered alive). Movie speed: 100 frames per second. AVI files were made with ImageJ. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 1preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 2preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 3preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 4preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 5preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 6preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 7preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 8preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 9preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 10preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Supplementary figure 11preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint n=mice n=slice n=neurons n=SGCs (circle) n= HA-positive neurons % of HA-positive neurons n= HA-positive SGCs % of HA-positive SGCs GFAP-hM3Dq 3 12 278 208 0 0 184 88,5% n=mice n=slice n = PV-positive neurons n = G6-positive PV-positive neurons % of G6-positive PV-positive neurons PV-Cre::GCaMP6 3 6 37 36 97,3% Supplementary table 1. IMMUNOHISTOCHEMISTRY - CELLULAR CHARACTERIZATION OF GFAP-hM3Dq AND PV-CRE::GCaMP6 TRANSGENIC LINES (related to Fig. 1, Fig. 2 and Supplementary Fig. 6) Supplementary table 1preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint CNO Agonist cocktail GFAP-hM3Dq::Cx43-CreERT2::GCaMP6 3 9 119 16.62 ± 1.5 Cx43-CreERT2::GCaMP6 3 9 149 no Ca2+ events 3.14 ± 0.44 PV-Cre::GCaMP6 5 15 23 no Ca2+ events 2.48 ± 0.48 Supplementary table 2. CALCIUM IMAGING - FUNCTIONAL CHARACTERIZATION OF GFAP-hM3Dq::Cx43-CreERT2::GCaMP6, PV- Cre::GCaMP6 & CNO INERTNESS (related to Fig. 1 and Supplementary Fig. 8) Amplitude ± SEM (DF/F0)n=mice n=DRGs n=cells Supplementary table 2preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint pERK WB 2min n=mice Mean density ± SEM Normality Homosced. WT / CNO 10 1 .00 ± 0.10 p=0.2743 GFAP-hM3Dq / CNO 10 1.76 ± 0.15 0.6235 pERK WB 30min n=mice Mean density ± SEM Normality Homosced. WT / CNO 9 1 .00 ± 0.10 0.4052 GFAP-hM3Dq / CNO 11 1.13 ± 0.13 0.003 pERK WB 4h n=mice Mean density ± SEM Normality Homosced. WT / CNO 9 1 .00 ± 0.09 0.1321 GFAP-hM3Dq / CNO 11 0.91 ± 0.08 0.0811 pERK IHC 2min - mean density of DRG (Epifluorescence) n=mice n=slice Mean density ± SEM Homosced. Normality WT / CNO 4 16 1 .00 ± 0.05 n too small WT / CNO vs GFAP-hM3Dq / saline 1 ns GFAP-hM3Dq / saline 4 16 0.9611 ± 0.05 n too small WT / CNO vs GFAP-hM3Dq / CNO 5.5 ns GFAP-hM3Dq / CNO 4 16 1.602 ± 0.13 n too small GFAP-hM3Dq / saline vs GFAP-hM3Dq / CNO 6.5 * pERK IHC neurons 2min - mean density of sensory neuron (confocal) n=mice n=slice Mean density in neurons ± SEM Homosced. Normality WT / CNO 4 16 1.00 ± 0.2 n too small WT / CNO vs GFAP-hM3Dq / saline 2 ns GFAP-hM3Dq / saline 4 16 0.72 ± 0.10 n too small WT / CNO vs GFAP-hM3Dq / CNO 2.5 ns GFAP-hM3Dq / CNO 4 16 0.79 ± 0.19 n too small GFAP-hM3Dq / saline vs GFAP-hM3Dq / CNO 0.5 ns Significant results Dunn's multiple comparison test (rank) Dunn's multiple comparison test (rank) Kruskal-Wallis P=0.5836 Statistics Statistics Kruskal-Wallis P=0.0231 Statistics nst=0.7446F test p=0.8406 Normalized to WT / CNO condition (fold change) IMMUNOHISTOCHEMISTRY - pERK QUANTIFICATION (related to Fig. 1 and Supplementary Fig. 5) Two-tailed Unpaired t test p=0.4661 Supplementary table 3. WESTERN BLOT - pERK QUANTIFICATION (related to Fig. 1 and Supplementary Fig. 4) *** ns Statistics Statistics p=0.0006 Normalized to WT / CNO condition (fold change) t=4.127 rank= 92 rank = 118 Two-tailed Unpaired t testF test p=0.2335 Two-tailed Mann- Whitney p=0.8792 Supplementary table 3preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint n= mice n=DRGs n=cells Onset (s) Normality Proprioceptors (GFAP-hM3Dq::PV-Cre::GCaMP6) 14 23 49 58.96 ± 2.2 p=0.3206 rank1 SGCs (GFAP-hM3Dq::Cx43-CreERT2::GCaMP6) 3 6 86 33.38 ± 1.4 p=0.2129 -55.06 *** Proprioceptors / Ca2+-free ACSF (GFAP-hM3Dq::PV-Cre::GCaMP6) 6 15 14 99.57 ± 7.3 p=0.0182 -36.51 * n= mice n=DRGs n=cells Onset (s) Normality Homosced. Proprioceptors 42.8 ± 3.4 p=0.4820 SGCs 23.11 ± 2.3 p=0.6704 1 compared with data of proprioceptors Significant results "n" used for the statistical comparison Supplementary table 4. CALCIUM IMAGING - ONSET OF Ca2+ ELEVATIONS IN SGCs OR PROPRIOCEPTORS (related to Fig. 2, Supplementary Fig. 7 and Fig. 11) GFAP-hM3Dq::Cx43-CreERT2::PV-Cre::GCaMP6 Statistics *** Statistics F test p=0.31439 Two-tailed Unpaired t-test p=0.00023 5 t=4.757 Dunn's multiple comparison testp<0.0001Kruskal-Wallis Supplementary table 4preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint Mean ± SEM Mean ± SEM CNO 19 49 535 68 p<0.0001 0.13 ± 0.03 rank2 1.00 ± 0.25 rank2 CNO + PPADS 6 19 233 14 n too small 0.05 ± 0.02 10.40 ns 0.43 ± 0.16 10.32 ns CNO + AIDA + NBQX + AP5 + MRS2279 6 17 180 17 n too small 0.11 ± 0.03 -3.263 ns 0.88 ± 0.26 -3.263 ns CNO + Bicuculine + RO3 6 19 255 26 n too small 0.10 ± 0.03 -2.763 ns 0.80 ± 0.21 -2.763 ns CNO + MSOP + LY341495 6 16 188 28 n too small 0.12 ± 0.03 -4.513 ns 0.98 ± 0.27 -4.430 ns CNO + CGP35348 5 15 276 28 n too small 0.12 ± 0.02 -6.763 ns 0.92 ± 0.14 -6.763 ns CNO in Ca2+ free ACSF 6 15 173 14 n too small 0.10 ± 0.05 7.904 ns 0.78 ± 0.42 7.904 ns Mean ± SEM (DF/F0) Mean ± SEM CNO 19 49 68 p<0.0001 1.26 ± 0.13 rank2 1.00 ± 0.11 rank2 CNO + PPADS 6 19 14 0.0113 0.83 ± 0.16 22.74 ns 0.66 ± 0.12 22.74 ns CNO + AIDA + NBQX + AP5 + MRS2279 6 17 17 0.0048 1.15 ± 0.29 14.50 ns 0.91 ± 0.23 14.50 ns CNO + Bicuculine + RO3 6 19 26 0.0014 1.43 ± 0.18 -21.53 ns 1.14 ± 0.14 -21.53 ns CNO + MSOP + LY341495 6 16 28 0.0075 1.78 ± 0.24 -33.64 ns 1.40 ± 0.18 -33.64 ns CNO + CGP35348 5 15 28 0.0001 1.44 ± 0.19 -19.82 ns 1.14 ± 0.15 -19.82 ns CNO in Ca2+ free ACSF 6 15 14 0.6748 0.94 ± 0.12 4.898 ns 0.75 ± 0.09 4.898 ns p value2 p value2 p= 0.0097 ** p= 0.0348 * p=0.9284 ns p=0.9780 ns p=0.7435 ns p=0.8186 ns p=0.3243 ns p=0.5596 ns p=0.5176 ns p=0.5596 ns p=0.0014 ** p=0.0047 * Mean ± SEM (s) Mean ± SEM CNO 19 49 68 P<0.0001 23.6 ± 2.2 rank2 1.00 ± 0.09 rank2 CNO + PPADS 6 19 14 0.9604 12.1 ± 1.3 43.42 ns 0.51 ± 0.06 43.42 ns CNO + AIDA + NBQX + AP5 + MRS2279 6 17 17 0.0369 20.4 ± 2.7 1.406 ns 0.87 ± 0.12 1.440 ns CNO + Bicuculine + RO3 6 19 26 0.5392 26.0 ± 2.1 -28.81 ns 1.10 ± 0.09 -28.76 ns CNO + MSOP + LY341495 6 16 28 0.0049 18.8 ± 1.5 0.8750 ns 0.80 ± 0.06 0.8875 ns CNO + CGP35348 5 15 28 0.0129 22.5 ± 2.4 -8.278 ns 0.95 ± 0.10 -8.207 ns CNO in Ca2+ free ACSF 6 15 14 0.0004 14.60 ± 1.8 30.38 ns 0.62 ± 0.07 30.53 ns p value2 p value2 <0.0001 *** 0.0007 *** 0.112 ns 0.1314 ns 0.112 ns 0.27 ns 0.0463 * 0.0222 * 0.2402 ns 0.491 ns 0.0016 ** 0.007 ** Mean ± SEM (s) Mean ± SEM CNO 19 49 68 p=0.2499 93.1 ± 3.5 rank2 1.00 ± 0.04 rank2 CNO + PPADS 6 19 14 p=0.9711 80.0 ± 6.7 21.42 ns 0.86 ± 0.07 21.34 ns CNO + AIDA + NBQX + AP5 + MRS2279 6 17 17 p=0.3411 80.2 ± 8.9 21.40 ns 0.86 ± 0.10 21.43 ns CNO + Bicuculine + RO3 6 19 26 p=0.8287 97.9 ± 4.3 -10.43 ns 1.05 ± 0.05 -10.39 ns CNO + MSOP + LY341495 6 16 28 p=0.8892 101.4 ± 7.3 -13.91 ns 1.09 ± 0.08 -13.91 ns CNO + CGP35348 5 15 28 p=0.3582 110.1 ± 6.7 -27.98 ns 1.18 ± 0.07 -27.99 ns CNO in Ca2+ free ACSF 6 15 14 p=0.0119 53.0 ± 5.4 63.84 ** 0.57 ± 0.06 63.84 ** p value2 p value2 0.978 ns 0.9963 ns 0.978 ns 0.9963 ns 0.978 ns >0.9999 ns 0.8186 ns 0.9963 ns 0.5596 ns 0.9963 ns 0.1725 ns 0.5173 ns 1 Proprioceptors alive are determined by the response to the positive control application of KCl 50mM 2 Each test (Kruskal-Wallis followed by Dunn's multiple comparison test or Kolmogorov-Smirnov) is the result of the comparison between CNO condition and (CNO+antagonist) or (CNO in calcium free ACSF) conditions. Significant results "n" used for the statistical comparison Supplementary table 5. CALCIUM IMAGING - MECHANISM UNERLYING CNO/SGC-MEDIATED Ca 2+ RESPONSES IN PROPRIOCEPTORS (related to Fig. 3 and Supplementary Fig. 9) NUMBER OF PROPRIOCEPTORS RESPONDING TO CNO GFAP-hM3Dq::PV-Cre::GCaMP6 n=mice n=DRGs n=proprioceptors alive1 n=proprioceptors responding to CNO Normality Raw Data Normalized to CNO condition (fold change) Statistic Statistic Kruskal- Wallis P=0.4233 Kruskal-Wallis P=0.4300 AMPLITUDE OF PROPRIOCEPTORS RESPONDING TO CNO GFAP-hM3Dq::PV-Cre::GCaMP6 n=mice n=DRGs n=proprioceptors responding to CNO Normality Raw Data Normalized to CNO condition (fold change) Statistic Statistic Kruskal-Wallis P=0.0072 Kruskal-Wallis P=0.0072 Cumulative frequency Kolmogorov- Smirnov Cumulative frequency Kolmogorov- Smirnov RISE TIME OF PROPRIOCEPTORS RESPONDING TO CNO GFAP-hM3Dq::PV-Cre::GCaMP6 n=mice n=DRGs n=proprioceptors responding to CNO Normality Raw Data Normalized to CNO condition (fold change) Statistic Statistic Kruskal-Wallis P=0.0011 Kruskal-Wallis P=0.0011 Cumulative frequency Kolmogorov- Smirnov Cumulative frequency Kolmogorov- Smirnov DURATION OF PROPRIOCEPTORS RESPONDING TO CNO GFAP-hM3Dq::PV-Cre::GCaMP6 n=mice n=DRGs n=proprioceptors responding to CNO Normality Raw Data Normalized to CNO condition (fold change) Statistic Statistic Kruskal-Wallis P<0.0001 Kruskal-Wallis P<0.0001 Cumulative frequency Kolmogorov- Smirnov Cumulative frequency Kolmogorov- Smirnov Supplementary table 5preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint n=cells responding to ATP-gS Amplitude of cells responding to ATP-gS(DF/F0) Rise time (s) Duration (s) Onset (s) Proprioceptors (PV-Cre::GCaMP6) 6 18 262 11 2.7 ± 1.04 21.5 ± 4.2 72.2 ± 11 .2 53.7 ± 11.6 SGCs (Cx43-CreERT2::GCaMP6) 2 2 52 8 28.8 ± 5.04 29.7 ± 2.3 113.0 ± 11 .3 57.5 ± 8.5 n=proprioceptors responding to ADP-bS Amplitude of proprioceptors responding to ADP-bS(DF/F0) Rise time (s) Duration (s) Onset (s) 2 6 56 3 7.1 ± 3.7 7.7 ± 3.3 36 ± 3.5 26.7 ± 14.2 Supplementary table 6. CALCIUM IMAGING - MECHANISM UNERLYING CNO/SGC-MEDIATED Ca 2+ RESPONSES IN PROPRIOCEPTORS (related to Fig. 3 and Suplementary Fig. 10) n=mice n=DRGs n=cells alive Mean ± SEM ATP-gS application ADP-bS application Proprioceptors (PV-Cre::GCaMP6) n=mice n=DRGs n=proprioceptors alive Mean ± SEM Supplementary table 6preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint M ouse line (single trangenics) E xpressed transgene and cell type Re ferences GF AP-hM3Dq h M3Dq in SGCs (13) C x43-CreERT2 C reERT2 recombinase expressed in SGCs (22) P V-Cre C re recombinase expressed in proprioceptors (21) GC aMP6 l ox-STOP-lox GCaMP6 under ubiquitous CAG promoter (23) N ewly genereated mouse lines (multiple transgenics) E xpressed transgene and cell type C x43-CreERT2::GCaMP6 G CaMP6 in SGCs GF AP-hM3Dq::Cx43-CreERT2::GCaMP6 h M3Dq and GCaMP6 in SGCs P V-Cre::GCaMP6 G CaMP6 in proprioceptors GF AP-hM3Dq::PV-Cre::GCaMP6 h M3Dq in SGCs and GCaMP6 in proprioceptors GF AP-hM3Dq::Cx43-CreERT2::PV-Cre::GCaMP6 h M3Dq in SGCs and GCaMP6 in both SGCs and proprioceptors Supplementary table 7 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint A ntibodies N ame C ompany/Cat. # S pecies D ilution Pr imary antibodies GAPDH Mil ipore/MAB360 Mou se 1:2 0,000 (W B) GFP Invitro gen/A10262 Ch icken 1:1 ,000 (IHC) GLAST Frontie r Institute/Af660 Ra bbit 1:5 ,000 (IHC) HA tag Ro che/11867423001 Rat 1:1 ,000 (IHC) Na +/K+-ATPase D-AF48 8 Sa nta Cruz/sc-48 345 AF488 Mou se 1:1 ,000 (IHC) Neu rofilament Mil lipore/AB9568 Ra bbit 1:1 ,000 (IHC) pE RK1/2 Ce ll Signaling/4370 Ra bbit 1:1 ,000 (W B & I HC) Parva lbumin Swa nt/PV27 Ra bbit 1:1 ,000 (IHC) S econdary antibodies Ale xa Fluor® 546 anti-r abbit Invitro gen/A-1103 5 Goa t 1:1 ,000 (IHC) Ale xa Fluor® 488 anti-ch icken Invitro gen/A-1103 9 Goa t 1:1 ,000 (IHC) HR P anti-rab bit Jackso n Immunoresearch/111-035-003 Goa t 1:1 0,000 (W B) HR P anti-mou se Jackso n Immunoresearch/115-035-003 Goa t 1:1 0,000 (W B) Dr ugs N ame C ompany/Cat.# Tota l bath applic ation d uration Fi nal concentration A gonists – ex viv o Ad enosine Ab cam/ab120498 2mi n 1µM (Ca2+ ima ging) ADP-ES Sig ma/A8016 2mi n 30 µM (Ca2+ i maging) ATP-JS Tocris/40 80 2mi n 50µM (Ca2+ i maging) Ca rbachol Ab cam/ab141354 2mi n 10µM (Ca2+ i maging) CNO NIH (Bryan Roth Lab) 2mi n 10µM (Ca2+ i maging) DH PG Ab cam/ab120020 2mi n 50µM (Ca2+ i maging) Gluta mate Ab cam/ab120049 2mi n 20 0µm (Ca2+ i maging) Hi stamine Sig ma/H7125 2mi n 10µM (Ca2+ i maging) A ntagonists – e x vivo AIDA En zo/ALX550101M025 3min 10 0µM (Ca2+ i maging) AP5 Ab cam/ab120003 3min 50µM (Ca2+ i maging) Bicu culine Sig ma/14340 3min 50µM (Ca2+ i maging) CGP3 5348 Ab cam/ab120167 10 min 10 0µM (Ca2+ i maging) LY 341495 Ab cam/ab120818 10 min 1µM (Ca2+ ima ging) MRS 22 79 Tocris/21 58 3min 50 0nM (Ca2+ i maging) MSOP Ab cam/ab141381 10 min 10 0µM (Ca2+ i maging) NBQX Ab cam/ab120046 3min 10µM (Ca2+ i maging) PPAD S Ab cam/ab120009 3min 10 0µM (Ca2+ i maging) R0-3 Tocris/30 52 3min 10µM (Ca2+ i maging) Dr ugs – In vi vo N ame C ompany/Cat.# Fi nal concentration CNO NIH (Bryan Roth Lab) 1mg /kg Tamoxi fen Sig ma/T5648 1mg /day (1 0mg/mL) Supplementary table 8preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 5, 2024. ; https://doi.org/10.1101/2024.09.04.611156doi: bioRxiv preprint