Activation of Intra-nodose Ganglion P2X7 Receptors Elicit Increases in Neuronal Activity

Vagus nerve innervates several organs including the heart, stomach, and pancreas among others. Somas of sensory neurons that project through the vagal nerve are located in the nodose ganglion. The presence of purinergic receptors has been reported in neurons and satellite glial cells in several sensory ganglia. In the nodose ganglion, calcium depletion-induced increases in neuron activity can be partly reversed by P2X7 blockers applied directly into the ganglion. The later suggest a possible role of P2X7 receptors in the modulation of neuronal activity within this sensory ganglion. We aimed to characterize the response to P2X7 activation in nodose ganglion neurons under physiological conditions. Using an ex vivo preparation for electrophysiological recordings of the neural discharges of nodose ganglion neurons, we found that treatments with ATP induce transient neuronal activity increases. Also, we found a concentration-dependent increase in neural activity in response to Bz-ATP (ED50 = 0.62 mM, a selective P2X7 receptor agonist), with a clear desensitization pattern when applied every ~ 30 s. Electrophysiological recordings from isolated nodose ganglion neurons reveal no differences in the responses to Bz-ATP and ATP. Finally, we showed that the P2X7 receptor was expressed in the rat nodose ganglion, both in neurons and satellite glial cells. Additionally, a P2X7 receptor negative allosteric modulator decreased the duration of Bz-ATP-induced maximal responses without affecting their amplitude. Our results show the presence of functional P2X7 receptors under physiological conditions within the nodose ganglion of the rat, and suggest that ATP modulation of nodose ganglion activity may be in part mediated by the activation of P2X7 receptors.


Introduction
The vagus nerve is the X cranial nerve and innervates several organs including the heart, stomach, and pancreas among others. Somas of sensory neurons that project through the vagus nerve are located in the nodose and jugular ganglia. Nodose ganglion (NG) neurons express receptors for several neurotransmitters, including acetylcholine (Cooper 1984;Kichko et al. 2013), glutamate (Hoang and Hay 2001;Czaja et al. 2006), serotonin (Wallis et al. 1982;Higashi and Nishi 1982), histamine (Thompson et al. 2000), GABA (Wallis et al. 1982;Ashworth-Preece et al. 1997), and ATP (Krishtal et al. 1983;Khakh et al. 1995;Wang et al. 2009). All these transmitters have been proposed to be involved in the central communication of NG afferent activity, or the generation and/or modulation of efferent activity to the periphery (Deuchars et al. 2001;Kichko et al. 2013;Yokoyama et al. 2015). On the other hand, there is a growing amount of evidence that reveals an intra-ganglia capacity of "cross-talking" in sensory ganglia, based on the property of satellite glial cells and their neighboring neurons to have a crosstalk through an unknown chemical neuro-glio transmitter(s) (Suadicani et al. 2010;Christie et al. 2015;Retamal et al. 2017). This intercellular communication can potentially modulate the excitatory status of sensory neurons and, thus, the afferent discharge (Hanani 2010;Retamal et al. 2014b;Hossain et al. 2017;Yamakita et al. 2018). Recently, it has been shown that an increased neuronal activity induced by low extracellular calcium in the NG can be partly reversed by a P2X7 receptor blocker applied to the ganglion (Retamal et al. 2014a), suggesting a possible involvement of P2X7 receptors in the generation of neuronal activity. These results suggest that ATP can be a good candidate for this glial-to-neuron communication. Accordingly, ATP has been involved in the generation and/or modulation of vagal visceral afferents related to nociceptor modality (Cockayne et al. 2005;Wang et al. 2009;Wan et al. 2010), although other sensory modalities could also be modulated through ATP receptors (Cockayne et al. 2005;Song et al. 2012;Adriaensen et al. 2015). The presence of ATP receptors, both ionotropic (P2X; Lewis et al. 1995;Vulchanova et al. 1997;Xiang et al. 1998;Hubscher et al. 2001;Song et al. 2012;Wang et al. 2014)) and metabotropic (P2Y; Fong et al. 2002;Ruan and Burnstock 2003)) have been reported in NG neurons. P2X receptors are homo-and/or hetero-trimers composed of subunits (P2X1-P2X7), most of which have been found to be expressed in the NG (Lewis et al. 1995;Deuchars et al. 2001;Hubscher et al. 2001;Wang et al. 2014;Kupari et al. 2019), although most of the physiological responses appear to be mediated by P2X2/3 heterotrimeric receptors (Khakh et al. 1995;Thomas et al. 1998;Virginio et al. 1998;Dunn et al. 2000;Cockayne et al. 2005;Tan et al. 2009;Wang et al. 2014). Nevertheless, P2X7 receptors have been reported to be localized in the central terminals of myelinated fibers of NG neurons (Deuchars et al. 2001). The activation of these receptors modulates the release of glutamate from the afferent nerve endings through a presynaptic mechanism (Deuchars et al. 2001). Whether the activation of P2X7 receptors at the neuronal soma could modulate their activity under physiological conditions is still not known. Accordingly, the main aim of the present study was to characterize the responses induced by ATP and a P2X7 receptor agonist applied to the neuronal soma of NG. We found that the P2X7 receptor was constitutively expressed in the NG in both neurons and satellite glial cells, and that selective activation of P2X7 receptors increased NG neuronal activity in a concentration-dependent manner. Also, we found that repeated activation of P2X7 receptors leads to desensitization of NG neural responses.

Animals
Recordings were obtained from fourteen preparations obtained from thirteen male Sprague-Dawley rats weighing 297 ± 25 g (mean ± SE). The animals were anaesthetized with sodium pentobarbitone (60 mg/kg, i.p.) or ketamine/xylazine (75/7.5 mg/kg. i.p.), and supplemented with additional doses when necessary to maintain a light level of surgical anesthesia (Stage 3, plane 2). At the end of the experiments, animals were euthanized with a pentobarbitone (180 mg/kg, i.p.) or ketamine (240 mg/kg, i.p.) overdose. The Commission of Bioethics and Biosafety from our respective Universities approved all the experimental protocols, which were performed according to guidelines of the National Fund for Scientific and Technological Development (FONDECYT, Chile) and the "Guide for the Care and Use of Laboratory Animals" from the Institute of Laboratory Animal Research Commission on Life Sciences, National Research Council (National Academy Press, Washington, D.C. 1996).

Ganglion Extraction
Vagus nerve and the nodose-petrosal-jugular complex (NPJc) were exposed through a midline incision in the neck. The vagus nerve was dissected from surrounding tissue and cut 2-4 cm distal to the NPJc peripheral limit. Then, the central process of the NPJc was exposed and cut about 1 mm from apparent central limit of the ganglion, and the obtained tissues were placed in cold Hanks' balanced salt solution (HBSS). This procedure was repeated contralaterally.

Vagus Nerve Recording
The connective tissue over the NPJc and the vagus nerve was carefully removed using tweezers and Vannas microscissors. The preparations were transferred into a two-compartment chamber kept at 38.0 ± 0.5 °C, and superfused with HBSS supplemented with 5 mM HEPES pH 7.43 buffer, equilibrated with air, and a flowing of 1.2 mL/min. The NPJc samples were placed in the 0.3 mL capacity lower compartment, over a pair of platinum electrodes, and pinned to the bottom of the chamber (Alcayaga et al. 1998). The electrodes were connected to a stimulator, and a thermistor was located in the superfusion channel near the NPJc surface. The vagus nerve (VN) was placed on paired Pt recording electrodes, and lifted into the upper compartment of the chamber filled with mineral oil. The recording electrodes were connected to an AC-preamplifier (Model 1800; A-M Systems, USA), and the resulting electroneurogram was amplified, displayed on an oscilloscope, and recorded on video-cassette tapes. The electroneurogram was also fed to a spike amplitude discriminator, of which standardized pulses were counted at 1 s intervals to assess the vagus nerve frequency of discharge (ƒ VN ), which was also digitized on line through an analog to digital conversion board, displayed on a computer using a custommade program, and saved as ASCII encoded text files for later analysis. Drugs were applied in 10-50 µL boluses by means of micropipettes, whose fine tips were placed about 1 mm distance from the exposed surface of the NPJc.

Intracellular Recordings
Isolated NPJc were placed in a recording chamber, their connective tissue was carefully removed and the resulting ganglia were pinned to the bottom of the chamber and continuously superfused with HBSS supplemented with 5 mM HEPES pH 7.43 buffer, equilibrated with air, at 37 °C. Neurons were impaled, under microscopic guidance, with glass pulled microelectrodes filled with KCl 3 M (20-50 MΩ), connected in turn with an electrometer (Axon 900) that allows the recording of resting membrane and action potentials evoked by current injection or application of ATP and its agonists.

Cell Culture
SH-SY5Y cells were seeded on 12-well plates with 12 mm glass coverslips in order to obtain a 60-80% confluence after a 24 h incubation. These cells were fixed for 10 min in a 4% paraformaldehyde (PFA) in PBS solution, then washed three times for 5 min with Ca/Mg DPBS (Corning, Corning, NY, USA), and permeabilized for 10 min with a 0.02% Triton X-100 in PBS solution. After blocking with a 10% BSA solution, an anti-P2X7 (1:100 dilution, APR-004, Alomone Labs, Jerusalem, Israel) primary antibody was incubated with or without its corresponding blocking peptide (1:25 dilution, BLP-PR004, Alomone Labs, Jerusalem, Israel) overnight at 4 °C. Then, coverslips were washed three times for 5 min in Ca/Mg DPBS. Anti-rabbit IgG Alexa-Fluor 488 (1:500 dilution, A11008, Thermo Fisher Scientific, Waltham, MA, USA) was used as secondary antibody, performing a 2 h incubation at room temperature. After washing with PBS, cells were mounted in glass slides using ProLong® Diamond Antifade Mountant with DAPI fluorescence mounting medium (P36971, Thermo Fisher Scientific, Waltham, MA, USA). Single focal images were taken using a Zeiss LSM 800 confocal microscope (Carl Zeiss, Heidelberg, Germany) with a Plan-Apochromat 63 × /1.46 oil objective. Images were acquired as 16-bit, 1024 × 1024 pixels, avoiding signal saturation, pinhole adjusted to 1 Airy unit, gain between 630 and 770 V, and laser power ranging from 1.48 to 21.32% for the 488 nm laser. Acquired images were processed using ZEN Imaging Software 3.4 (Carl Zeiss, Heidelberg, Germany).

Nodose Ganglia
Anesthetized rats were perfused intracardially with phosphate buffered saline (PBS; pH 7.4) for 10 min, followed by buffered PFA 4% (Sigma-Aldrich) for 10 min. Nodose ganglia were visually identified and harvested from rats, and fixed by immersion in buffered PFA 4% for 12 h at 4 °C followed by three 5 min washes in PBS, a sucrose gradient (5%, 10%, 20%, 30% in PBS) treatment, and then embedded in optimum cutting media. Cryostat sections (20 μm) of the nodose ganglia were obtained and mounted on Superfrost Plus slides (Thermo Fisher Scientific). Sections were blocked/permeabilized in 0.5% Triton X-100, 2% fish skin gelatin (Sigma-Aldrich), 1% bovine serum albumin in PBS for 1 h at room temperature. Sections were incubated overnight at 4 °C with a mouse anti-P2X7 monoclonal antibody (1:100 in the same blocking media, Millipore). Negative controls were performed by omitting the incubation with the primary antibody. After being washed with PBS, tissue sections were incubated for 1 h at room temperature with an Alexa-Fluor 488 rabbit anti mouse IgG (1:200, Molecular Probes) antibody. Finally, sections were mounted in DAPIcontaining media (Vectashield, Vector Laboratory) and visualized using a confocal laser microscope (Leica).

Data and Statistical Analyses
Changes in the frequency of discharge (Δƒ VN ) induced by drugs were calculated as difference between the maximal frequency (max ƒ VN ) achieved during an evoked response and the mean basal activity (bas ƒ VN ), computed along the 30 s period prior to drug administration. The mean Δƒ VN s, directly or standardized to the maximal response induced by ATP (Δƒ VN /Δƒ max ATP) on each preparation, were related to the Bz-ATP dose by a sigmoid curve (Y = Y max / [1 + (ED 50 /D) S ]), where D represents the applied dose, ED 50 the dose that evoked half-maximal response, S the Hill slope factor determining the steepness of each curve, and Y max , was the mean maximal Δƒ VN induced by Bz-ATP or the mean maximal Δƒ VN /Δƒ max ATP, depending on the adjusted curve. Results are presented as mean ± standard error (SE). Two populations were compared using the Student´s t test. Difference between groups was assessed using parametric (one-way ANOVA) or non-parametric analysis of variance (Kruskal-Wallis test) and post-hoc test (Holm-Sidak's or Dunn's multiple comparisons tests), according to the vari-ables´ distribution. Analyses were carried out using Graph-Pad Prism version 7.05 for Windows (GraphPad Software, La Jolla California USA, www. graph pad. com). Obtained p values < 0.05 were considered as statistically significant.

Vagus Nerve Recording
Application of a single bole (10 μl) of ATP (90 mM) on top of the superfused ganglion produced a brisk increase in ƒ VN (Fig. 1A, B, continuous line) that returned to basal levels within 15 to 40 s after its application. Similarly, 1 mM Bz-ATP produced a fast increase of ƒ VN (Fig. 1A, B, segmented line), with a mean duration of 19.7 ± 3.8 s, which was not significantly different to the mean duration of the responses induced by ATP (21.4 ± 3.7 s) (p > 0.05; paired Student's t test; n = 7; Fig. 2). The maximal amplitude of the responses induced by Bz-ATP presented a large variation between different preparations (two examples in Fig. 1A, B), but the mean maximal Δƒ VN induced by Bz-ATP (202.2 ± 40.6 Hz) was significantly lower than that evoked by the largest ATP dose (312.7 ± 44.8 Hz) (p < 0.05; paired Student's t test; n = 7; Fig. 2). Responses induced by Bz-ATP present a high degree of temporal desensitization.
Consecutive applications of Bz-ATP (1 mM) every 30 s produced a time-dependent reduction of the amplitude and duration of the ƒ VN increases, with the fourth application being devoid of effects (Fig. 3A, B). Increasing the period between stimuli reduced the magnitude of the desensitization, with an interval longer than about 1 min to evoke responses of similar amplitude (Fig. 3C).
The amplitude of the responses induced by Bz-ATP was dose dependent (Fig. 4A), with a threshold of about 0.1 mM, further increasing up to a plateau of maximal activity near 1-2 mM (Fig. 4B). The mean duration of Δƒ VN induced by Bz-ATP were correlated with the applied doses (r 2 = 0.99; p < 0.05) (n = 7), with a mean maximal discharge of 240.0 ± 17.0 Hz, a half-maximal response dose (ED 50 ) of 0.62 ± 0.08 mM, and a slope factor of 1.46 ± 0.16 (Fig. 5A). Similarly, the mean responses induced by Bz-ATP, standardized to the maximal response evoked by ATP on each preparation, were correlated with the applied doses (r 2 = 0.99; p < 0.05), presented a maximal response of 0.70 ± 0.05, an ED 50 of 0.58 ± 0.08 mM, and a slope factor of 1.37 ± 0.15 (Fig. 5B).

Nodose Neuron Recordings
The recordings obtained from 19 neurons, stimulated with ATP (n = 8), α,β-methylene ATP (α,β-meATP;  (1 mM). Mean Δƒ VN max induced by ATP was significantly larger (p < 0.05; Student paired t test) than the ones evoked by Bz-ATP in the same animal model. Conversely, the mean duration was not significantly different (p < 0.05; Student paired t test) between responses induced by both drugs. #p < 0.05; Student paired t test (n = 7) 1 3 n = 6) or Bz-ATP (n = 5) (Fig. 7A), showed that in control conditions, there were no statistically significant differences in the resting membrane potential between neurons on each group ( Fig. 7B; p > 0.05, Kruskal-Wallis test). After a short delay, all three agonists evoked a significant (p < 0.01; Kruskal-Wallis test; Fig. 7A, black arrow) depolarization with mean magnitude significantly larger for α,β-meATP (23.55 ± 3.76 mV) than for ATP (5.68 ± 1.45 mV; p = 0.013; Dunn's multiple comparisons test) or Bz-ATP (4.58 ± 1.29 mV; p = 0.012; Dunn's multiple comparisons test); no significant differences (p > 0.99; Dunn's multiple comparisons test) were observed between ATP-and Bz-ATP-induced depolarizations (Fig. 7C). The time necessary to attain the maximal depolarization level induced by the three different agonists was significantly different (p < 0.01), with α,β-meATP-induced responses (10.65 ± 2.16 s) being significantly different from ATP-(26.74 ± 2.08 s; p = 0.004; Dunn's multiple comparisons test) but not from Bz-ATP-induced responses (16.37 ± 2.49 s; p > 0.99; Dunn's multiple comparisons test) (Fig. 7D). Non-significant differences (p = 0.11; Dunn's multiple comparisons test) were observed between ATP-and Bz-ATP-induced time to maximal depolarization (Fig. 7D). Thus, the mean slope of the depolarization was significantly different between the agonists (Fig. 7E), with the mean slope of the α,β-meATP-induced responses (10.71 ± 2.78 mV/s) being significantly different from the one of ATP-induced responses (0.29 ± 0.07 mV/s; p = 0.0009; Dunn's multiple comparisons test) but not from the Bz-ATP-induced one (0.73 ± 0.22 mV/s; p = 0.17; Dunn's multiple comparisons test). Altogether, Fig. 3 Responses induced by consecutive Bz-ATP applications present temporal desensitization. A Frequency discharge (ƒ VN ) in a single preparation was maximally increased by the first application of Bz-ATP (Ap 1), with following applications every 30 s presenting a decreasing amplitude and duration. The fourth application (Ap 4) produced no significant modification of ƒ VN . B Mean increases (n = 3) in frequency discharge (Δƒ VN ; empty bars) and duration (stripped bars) of the responses to Bz-ATP applications were reduced in consecutive (30 s apart) applications. C Mean standardized increases (n = 4) in frequency discharge induced after a certain delay (Δƒ t ) with respect to the initial response (Δƒ 0 ). The reduction of the response was maximal after 24 s, recovering linearly after approximately 55 s. Bars: SEM   Fig. 4 Responses to Bz-ATP increased with increasing doses in a single preparation. A Frequency discharge (ƒ VN ) increased in a dose-dependent manner in response to the application (arrowhead) of Bz-ATP. B The increases in frequency discharge (Δƒ VN ) were related to the dose, with a threshold around 0.1 mM, with an ED 50 near 0.3 mM, and reaching a plateau for doses above 0.5 mM. Error bar = SEM. n = 3 1 3 these results show that the responses evoked by ATP and Bz-ATP are very similar, and this could indicate that the ATP-mediated in vivo response could be through P2X7 receptors.

Localization of P2X7 Receptors in the Rat Nodose Ganglia
First, we confirmed the primary antibody selectivity by Western blot and immunofluorescence. As shown in Fig. 8A, there is a correlation between the P2X7 immunosignal and its protein expression levels in cultured SH-SY5Y cells. These results confirm and extent previous reports on antibody specificity (Larsson et al. 2002;Hevia et al. 2019). It is worth noting that P2X7 immunoreactivity in SH-SY5Y cells was absent in the negative control (non-primary antibody incubation) and blocking peptide conditions (Fig. 8B). Similar immunofluorescent images of sensory neurons have been previously reported (Vit et al. 2008;Jack et al. 2011;Liu et al. 2014;Dhandapani et al. 2018). This allows us to suggest the presence of positive P2X7 receptor-stained neurons within the ganglion with variable intensity (Fig. 8C, arrows). Moreover, it seems that P2X7 receptor is present in both large and small diameter neurons. Nuclei of immune-stained neurons appear to be centrally located (Fig. 8C, zoom). It is well known that satellite glial cells are surrounding the neuronal bodies. Because some staining was observed in places where satellite cells must be located, we suggest that this cell type also expressed (in some degree) P2X7 receptors. These results suggest that P2X7 receptors could be expressed in nodose ganglion neurons, but also in satellite glial cells, supporting the idea that this receptor participates in neuronto-glial crosstalk. Future experiments will be focused in a detailed description of the exact localization of these receptors within the ganglion.

Discussion
In this work, we found that the application of ATP and Bz-ATP to the NG increases neuronal activity measured as an increase of discharge in the vagus nerve, but the evoked response induced by Bz-ATP was smaller compared to that induced by ATP (with an ED 50 for Bz-ATP close to 600 µM). However, single NG neurons recordings indicated similar electrophysiological modifications induced by ATP and BZ-ATP, suggesting that in vivo the response to ATP is more complex and may involve other players, such as satellite glial cells. Thus, it is well known that NG neurons express various types of P2X (Lewis et al. 1995;Vulchanova et al. 1997;Xiang et al. 1998;Hubscher et al. 2001;Song et al. 2012;Wang et al. 2014) and P2Y receptors (Fong et al. 2002;Ruan and Burnstock 2003). However, there are no reports showing the exact entity of these receptors presents in NG satellite glial cells, but pharmacological (Feldman-Goriachnik et al. 2015) and RT-PCR (Yokoyama et al. 2015) studies demonstrate that satellite glial cells cultured from NG already have functional receptors of both types: P2X and P2Y. Therefore, satellite glial cells perfectly could have participated in the final response to ATP or Bz-ATP observed in this work.
In our knowledge, this is the first time that the presence of P2X7 receptor in NG neurons has been reported, although P2X7 mRNA is present in most NG neurons (Kupari et al.2019). Additionally, the presence of P2X7 in satellite glial cells of other sensory ganglia, such as trigeminal ganglion (Nowodworska et al. 2017), dorsal root ganglion (DRG) (Chen et al. 2012;Liu et al. 2015) has been previously reported, suggesting that NG satellite cells could also express this receptor, moreover when these cells present its mRNA (Kupari et al. 2019). Our immunostaining studies suggested that satellite glial cells also express P2X7 (in a Fig. 5 Relationship between the mean increases in frequency discharge (Δƒ VN ) and the dose of Bz-ATP applied. A Mean duration of Δƒ VN were significantly related (r 2 = 0.99; p < 0.05) to the Bz-ATP dose in a logistic manner, with a threshold below 0.1 mM, a maximal response of about 240 Hz, a half-maximal response for a dose of 0.62 mM, and a slope factor of 1.46. B The relationship between the increases induced by Bz-ATP (Δƒ VN ), standardized by the maximal response induced by ATP on each preparation (Δƒ VN /Δƒ max ATP) was highly significant (r 2 = 0.99; p < 0.05), with a maximal value of about 0.7, a half-maximal response for a dose of 0.58 mM, and a slope factor of 1.37. Error bars = SEM. n = 5 less degree compared to NG neurons). This result is congruent with previous publications in the DRG, where activation of P2X7 in satellite glial cells induce the ATP release from them, which in turn activate P2Y receptors in neurons, downregulating the expression of P2X3 receptors (Chen et al. 2012). Thus, the activation of P2X7 receptors in satellite glial cells could directly modulate the electrical activity and/or protein expression of sensory neurons. Therefore, satellite glial cells into the nodose ganglion could modulate sensory afferences/efferences of neurons that project from the vagal nerve, we suggest to explore this hypothesis in future experiments.
In the NG, the activation of sensory neurons by a decrease of extracellular divalent cations is reduced in part by Pannexin 1 (Panx1) and P2X7 inhibitors (Retamal et al. 2014a). It is well known that P2X7 receptor closely interacts with Panx1 channels, where the activation of P2X7 by ATP induces the opening of channels formed by Panx1 (Pelegrin and Surprenant 2006;Kim and Kang 2011;Gulbransen et al. 2012;Bravo et al. 2015). Accordingly, it has been proposed that activation of connexin hemichannels and pannexons induced the release of ATP to the extracellular space, which in turn can active P2X7 or other P2X in the soma of sensory neurons, inducing an increase of the intracellular calcium concentration (Suadicani et al. 2010) and electrophysiological activity (Retamal et al. 2017). In support to this idea, both P2X7 (Chen et al. 2016;Wu et al. 2017) and Panx1 (Zhang et al. 2015;Hanstein et al. 2016) when activated/ overexpressed increase sensory neuron activity, generally associated to pain pathogenesis, suggesting that the P2X7/ Panx1 axis play an important role in the control of sensory neuronal activity during pain conditions.
In ex vivo experiments, the response to ATP was higher compared to that induced by Bz-ATP. However, single NG neuron responses to ATP and Bz-ATP were not statistically different. This could be explained by the fact that vagus nerve recordings include all neuronal populations within the ganglia while single neuronal recordings allowed the recording of a small proportion of the whole number of NG neurons. Indeed, Kupari et al. (2019) reported 18 Then, our results showing the constitutive expression of the P2X7 receptor in both neurons and satellite glial cells within the rat NG confirm previous results showing mRNA expression of the P2X7 receptor in the NG but also provided novel data showing that these receptors are fully functional under physiological conditions. We would like to note that despite the ATP concentration in the bolus can be considered high (90 mM), the final concentration within the ganglion can be much lower. Because the ganglion in under constant superfusion (1.2 ml/min), the 10 μl bolus is rapidly washed out, thus, decreasing ATP concentration very fast. On the other hand, the ATP in the bath solution enters the ganglion by diffusion, through first the connective tissue capsule and then extracellular matrix that is between cell (Haberberger et al. 2019). On the other Fig. 7 Intracellular responses of nodose ganglion neurons to ATP and its agonists. A Response induced by application (arrow) to the superfusion medium of ATP (black trace), α,β-methylene ATP (red trace) and benzoyl-ATP (blue trace) on a single rat nodose ganglion (NG) neuron. After a delay, the neuron depolarized, and in the case of ATP application, the neuron fired a train of action potentials (inset). Dotted line indicates resting membrane potential (− 48 mV) in the figure and spiking threshold (− 41 mV) in the inset. The action potentials appeared cropped in the figure at approximately + 10 mV. B The control mean resting membrane potential was similar in all three groups of neurons. C Depolarizations induced by local application of α,β-meATP (n = 6) were significantly larger (p > 0.05; Kruskal-Wallis test) than those induced by ATP (n = 8) or Bz-ATP (n = 5) that were of similar amplitude. D The mean time to maximal depolarization was significantly larger (p < 0.05; Kruskal-Wallis test) for ATP-than for α,β-meATPinduced responses, with no significant differences between Bz-ATP-induced responses and the other agonists. E Mean slope of the agonist-induced depolarization was significantly larger (p < 0.05; Kruskal-Wallis test) for α,β-meATP-than for ATP-induced responses, with no significant differences between Bz-ATP-induced responses and the other agonists 1 3 Fig. 8 P2X7 is expressed in nodose ganglion cells. A Western blot analysis of P2X7 expression in cell lines with different P2X7 expression. B P2X7 immunofluorescence in SH-SY5Y cells without (left panel) or with (center panel) primary antibody. A blocking peptide condition was added to confirm signal specificity (right panel). C Images of immunofluorescence staining showing P2X7 receptor (green) and nuclei morphology (blue) in rat nodose ganglia. Arrows indicate neurons that display strong positive reaction for P2X7 receptor. Double arrowheads indicate P2X7 immunostaining in cells that surround neuronal bodies within the ganglia. As negative control, a condition without primary antibody was included. Right picture corresponds to a high magnification of the white box in the main picture; Scale bar = 100 µm. Right, high magnification; Scale bar = 20 µm hand, the maximal frequency discharge in an ex vivo petrosal ganglion model was attained by an extracellular ATP concentration of around 500 mM (Alcayaga et al. 2006), Moreover, the dose-response curve obtained from isolated petrosal ganglion neurons indicates that the maximal response is reached by ATP concentrations around 50 µM Iturriaga et al. 2007). These results indicate that to obtain similar responses, the concertation of ATP should be at least 200 time higher for ex vivo than for single-cell in vitro experiments.
Finally, we observed that the inhibitory effect of AZ 10606120 on Bz-ATP response was only partial. It may be because, despite that Bz-ATP is the most used agonist for P2X7 receptors, it can activate several different P2X receptors as well. Thus, for example, Bz-ATP is able to activate P2X1-P2X5 (Syed and Kennedy 2012). Therefore, the final response to Bz-ATP could be a mixture between P2X7 and other P2X receptors. Interestingly, the effect of AZ 10606120 on the Bz-ATP was to decreased the duration of its response (Fig. 6A), suggesting that (1) the initial Bz-ATP response is mediated by other P2X receptors and the slowest is mediated mostly by the activation of P2X 7 receptors and/ or (2) the slowest response is mediated (at least in part) by the activation/crosstalk between satellite glial cells and NG neurons, which will take more time if the response is directly mediated only by activation of NG neurons.

Conclusions
We and others have proposed a sensory model based on glial-to-neuron cell communication, which is mediated, at least in part, by purinergic signaling (Zhang et al. 2007;Gu et al. 2010;Rozanski et al. 2013). The present results provide new evidence showing that the P2X7 receptors may play a physiological role in the regulation of NG sensory function. Therefore, our findings provide a new molecular pathway involved in sensory processing in the rat NG. Data availability All data generated and analyzed during this study are included in this published article.

Declarations
Competing Interests The authors declare that they have no competing interests.
Ethical Approval All experiments were conducted in accordance to the guidelines of the National Fund for Scientific and Technological Research (FONDECYT, Chile) and the Guide for the Care and Use of Laboratory Animals (National Research Council of the National Academies, USA). Bio-Ethics Committee from the Facultad de Ciencias of the Universidad de Chile approved the experimental protocols.

Consent for Publication
Not applicable.