Abstract
P2X receptors are highly expressed throughout the nervous system, where ATP has been shown to be a neurotransmitter. The aim of this study was to characterize P2X receptor expression within sympathetic postganglionic neurons from the superior cervical ganglia. Reverse transcription-polymerase chain reaction showed the presence of mRNA for all P2X receptors, raising the possibility of multiple subunit expression within these ganglia. Whole-cell patch-clamp and calcium imaging studies revealed a heterogeneous population of P2X receptors in ∼70% of neurons. We propose that the heterogeneity in properties could be caused by differential expression and/or subunit composition of the P2X receptor. The dominant phenotype was P2X2-like; neurons showed slow desensitization, sensitivity to antagonists, and a profile of ionic modulation that is characteristic of P2X2 receptors: potentiation by acidification and extracellular Zn2+ and attenuation by high extracellular Ca2+ and pH. A subpopulation of neurons (10-15%) were α,β-methylene ATP (α,β-meATP) sensitive, and in neurons from P2X1 receptor-deficient mice the α,β-meATP response was reduced to 2% of all neurons, demonstrating a direct role for P2X1 subunits. Control α,β-meATP responses were eliminated by high extracellular Ca2+ and pH, indicating the presence of heteromeric channels incorporating the properties of P2X1 and P2X2 receptors. This study demonstrates that in neurons, the P2X1 receptor can contribute to the properties of heteromeric P2X receptors. Given the expression of P2X1 receptors in a range of neurons, it seems likely that regulation of the properties of P2X receptors by this subunit is more widespread.
ATP is copackaged and coreleased with classic neurotransmitters throughout the nervous system (Bodin and Burnstock, 2001). Purinergic receptor-mediated currents were first described in sensory (Krishtal et al., 1983) and spinal cord neurons (Jahr and Jessell, 1983) and have subsequently been shown to mediate synaptic events and regulation of transmission in a range of both peripheral and central neurons (Dunn et al., 2001; Khakh, 2001; Robertson et al., 2001). In the peripheral nervous system, ATP is coreleased with noradrenaline from sympathetic nerves (Sneddon and Burnstock, 1984) and can stimulate noradrenaline release (von Kugelgen et al., 1997; Boehm, 1999) and mediate fast synaptic transmission (Evans et al., 1992; Silinsky et al., 1992). ATP is an agonist at P2X receptor ion channels (North, 2002) and some G-protein-coupled P2Y receptors (Abbracchio et al., 2003). Seven P2X receptor genes (P2X1-7) have been identified; the functional receptors are thought to form as either homo- or heterotrimers, and a range of phenotypes has been described for recombinant receptors based on their subunit composition (North, 2002). In addition splice variants of P2X receptor subunits have also been described resulting in additional phenotypic variation (Brandle et al., 1997; Simon et al., 1997).
The superior cervical ganglia (SCG) are peripheral sympathetic ganglia involved in a range of activities, including innervation of the lacrimal gland, the salivary glands (submaxillary, sublingual, and parotid), and the dilator pupillae muscles in the eye. Previous studies suggested that rat peripheral sympathetic SCG neurons have a predominantly P2X2-like receptor phenotype [α,β-methylene ATP (α,β-meATP)-insensitive and potentiated by zinc](Cloues et al., 1993; Khakh et al., 1995; Boehm, 1999; Dunn et al., 2001). In this study, RT-PCR analysis showed the presence of multiple P2X receptor subunits within the SCG, and patch-clamp and calcium imaging demonstrated a significant population (∼10-15%) of α,β-meATP-sensitive neurons. α,β-meATP sensitivity is associated with the expression of P2X1, P2X3 (North, 2002), or P2X6 (R. J. Evans, unpublished observations) receptor subunits. We have 1) used patch clamp electrophysiology and calcium imaging of acutely dissociated adult SCG neurons to determine the properties of SCG neuron P2X receptors and 2) compared the properties of SCG neurons from wild-type (WT) and P2X1 receptor-deficient “knockout” (KO) mice (Mulryan et al., 2000). We have shown that the SCG has a heterogeneous population of P2X receptors and that the predominant isoform is most likely to be the P2X2 receptor; we demonstrate a direct functional contribution of the P2X1 receptor to receptor heteromers in ∼10% of neurons.
Materials and Methods
Primary Cell Culture. Adult MF1 mice of either sex were given a lethal intraperitoneal injection of pentobarbitone, and death was confirmed by femoral exsanguination. For each culture, SCG from two animals were removed immediately, cut in half, and placed in ice-cold Hanks' buffered salt solution (HBSS). The tissue was placed in 0.1% l-cysteine and 0.05% papain (enzyme activity, 39.2 units/ml) in HBSS and digested for 15 min at 37°C. It was then transferred to 0.05% collagenase (enzyme activity, 215.5 units/ml) and 4% dispase (enzyme activity, 4.48 units/ml) in HBSS for 40 min at 37°C. The pieces of ganglia were washed with Leibovitz's 15 (L-15) media before being placed in 2.5 ml of feeding media (L-15 media supplemented with 1% fetal calf serum, 0.22% sodium bicarbonate, 0.54% glucose; 0.01 ng/ml nerve growth factor, 30 μg/ml penicillin, and 30 μg/ml streptomycin). Mechanical trituration was used to separate the cells, which were seeded onto poly-d-lysine-coated coverslips and incubated for 1.5 h before a further 1 ml of feeding media was added.
Electrophysiology. Electrophysiological recordings were made for 1 to 4 days after dissociation, using the whole-cell patch-clamp configuration, at room temperature. Microelectrodes were pulled from thin-walled, filamented borosilicate glass and had resistances of 3 to 7 MΩ when filled with internal solution: 140 mM potassium gluconate, 10 mM EGTA, 10 mM HEPES, and 5 mM NaCl, pH adjusted to 7.3 with KOH. The bath was continuously perfused with extracellular solution containing: 150 mM NaCl, 10 mM HEPES, 11 mM glucose, 2.5 mM KCl, 2.5 mM CaCl2, and 1 mM MgCl2, pH adjusted to 7.3 with NaOH. Neurons were voltage-clamped at -70 mV (after correction for the tip potential) using an Axopatch 200B amplifier (Axon Instruments Inc., Union City, CA). Data were collected with pClamp 8.1 software and analyzed using Origin 6 (OriginLab Corp, Northampton, MA).
Agonists were applied for 1 s via a fast-flow U-tube (Evans and Kennedy, 1994). The off rate of agonist was dependent both on the U-tube used and the positioning relative to the cell, and changes in these parameters probably account for the variation in decay of P2X receptor current after removal of agonist. Antagonists and extracellular test solutions were perfused into the bath for 5 min before agonist application. Reproducible responses were obtainable with 5-min intervals between agonist applications. All experiments were performed on at least three separate neuronal cell cultures (2 mice per culture), except where stated in the text.
Fluorescent Calcium Imaging. SCG neurons were plated onto 16-mm coverslips and kept in culture for up to 3 days. Cells were loaded with Fluo-3 acetoxymethyl ester (final concentration, 1 μM) for 30 min at 37°C. The coverslips were washed in extracellular perfusion solution (as for electrophysiological experiments) for 30 min at 37°C. Argon laser excitation of the dye was at a wavelength of 488 nm with the filters set to capture emission at wavelengths greater than 510 nm. The neurons were imaged using a Fluoview FV300 confocal microscope (Olympus, Tokyo, Japan) at room temperature. Real-time fluorescence was captured by Olympus Fluoview ver. 4.2 software at a frequency of 0.5 Hz, for a continuous time period of ∼360 s. Agonist application was via U-tube for 40 s after 30 s of baseline fluorescence was acquired. Agonist applications occurred at an interval of 15 min. Antagonists were applied for 8 min before test solution application. Extracellular solution changes were performed 2 min before agonist application to allow complete exchange of the bath solution.
Neurons were identified visually. When extracellular solution was applied to the neurons via the U-tube, 34.4% (142 of 413) of neurons gave an increase in fluorescence (puff response). In the presence of 100 μM suramin, 84.2% (16 of 19) of puff responses were abolished. The remaining 15.8% (3 of 19) of neurons had a response that was 85.7 ± 14.3% of the initial artifact. This indicates that ATP may be released from some cells (either neurons or glia) because of mechanical stimulation of the neurons, causing local activation. None of the neurons with puff artifacts was included in data analysis.
Neuronal viability/responsiveness was confirmed by Ca2+ influx in response to 60 mM K+ perfusion solution; only cells that responded were included in the analysis (for both WT and KO mice). Average background fluorescence was measured from a cell-free area of the field of view. Baseline measurements were calculated as a mean of the first 10 s of recording. Neuronal traces were corrected for background and baseline and then expressed as a ratio of the baseline (referred to as self-ratio) to compensate for differential loading of the neurons. Comparisons were made between the normalized peak calcium changes under different agonist conditions. All experiments were performed on at least three separate neuronal cell cultures (2 mice per culture).
RT-PCR. Total RNA was isolated from six intact SCGs, and SCG cultures using the RNeasy Mini Kit (QIAGEN, Valencia, CA). The RNA was treated with DNase I (Sigma) to remove any DNA contamination. Half of the purified RNA was reverse-transcribed to cDNA using SuperScript II-reverse transcriptase (Invitrogen, Carlsbad, CA), and the remainder of the RNA was used as a negative control to ensure that there was no genomic contamination. Details of all the primers and the predicted size of the amplified products are given in Table 1. PCR reactions were carried out with 2 μl of cDNA and a final concentration of 500 nM of each primer, except for the P2X7 primers, where 3 μl of cDNA was required. Amplification took place in a Techne Genius PCR machine (Techne, Cambridge, UK) using the following protocol: denaturation at 94°C for 5 min; repeated cycles of denaturation at 94°C, followed by annealing at 57°C and extension at 72°C, each for 30 s; final extension at 72°C for 10 min. Thirty-five cycles were used in most cases, except for tyrosine hydroxylase (TH) and β-actin amplification, for which 30 cycles were used. The amplification products were examined by electrophoresis on a 1% agarose gel and visualized using ethidium bromide under UV light. Positive controls used were TH for confirmation of the sympathetic nature of the tissue, and β-actin was used to confirm that there was no genomic DNA contamination. The primers for β-actin were designed such that they would amplify an intronic sequence of the gene, whereas mRNA would have spliced out this intron. Based on the size of the amplified product, any genomic contamination could therefore be detected. To confirm the identity of the two bands amplified for the P2X4 receptor, the bands were sequenced using BigDye Terminator Version 1.0 sequencing kits (Applied Biosystems, Foster City, CA). Whole SCG were taken for RNA extraction; therefore, the sample contains RNA from non-neuronal cells (e.g., glia and blood cells).
Data Analysis. There was no significant difference between the amplitude of response (analysis of variance) or proportion of neurons responding (t test) over the 4-day test period for the electrophysiological data; therefore, the data have been pooled for all days. Statistical tests were performed using an unpaired t test, with a P value of <0.05 considered significant.
ATP was tested on each neuron. In responsive neurons, ATP was also applied after the test solution, to determine any run-down of responses. All electrophysiological data and fluorescent calcium imaging data are expressed as a percentage of the mean ATP responses measured before and after the test solution, except in the case of the antagonists, in which case it is the percentage of the initial ATP response. Data are expressed as the mean ± S.E.M., unless otherwise stated.
Concentration-response data were collected from individual neurons that had three or more concentrations of ATP tested on them, and an average data set was also obtained. The response to 1 mM ATP was taken as 100%, and the rest of the data points were normalized. Curves were generated in Origin 6 using the Hill equation: y = (Vmax × xnH)/(EC50nH + xnH) where Vmax is the maximum response (100%), nH is Hill coefficient, and EC50 is the concentration required to evoke a half-maximal response. pEC50 corresponds to -log10 of the EC50 concentration. The effects of the antagonists suramin and PPADS were calculated as the percentage inhibition of the peak response to ATP.
Results
Expression of P2X Receptor Subunit mRNA. To determine which P2X isoforms could be expressed by the mouse SCG, RT-PCR was performed on isolated total RNA (the samples corresponded to whole SCG and comprised a mixed population of cell types, including neurons, glia, and blood cells). No genomic DNA contamination of the RNA samples was found using the β-actin primers (data not shown), and primers for TH amplified a band of the expected size (∼460 bp), confirming the sympathetic origin of the tissue. Using subtype-selective primers, bands of the appropriate size were amplified corresponding to P2X1, P2X2, P2X3, P2X4, and P2X5 receptors (n = 3 separate RNA extractions; Fig. 1) from reactions containing reverse transcriptase but not when it was absent, further confirming the lack of DNA contamination of the samples. Two bands were seen for the P2X4 receptor, one at the expected ∼560 bp and the other at ∼630 bp. On sequencing, these were identified as two splice variants of the P2X4 receptor, consistent with previous reports (Townsend-Nicholson et al., 1999). The product amplified using the P2X7 primers was at the limit of detection and required 3 μl of cDNA template for a band to be evident (two of three separate RNA extractions). A band corresponding to P2XM, which has been identified as the murine P2X6 receptor (Robertson et al., 2001), was also just detectable (three of three). Dissociation of parasympathetic neurons from rat submandibular ganglia changes the profile of P2X receptor subunits as measured immunohistochemically. In intact ganglia, only P2X5 receptors are detected; however after dissociation, high levels of P2X2 and P2X4 receptors were reported (Smith et al., 2001). In the present study, the same complement of P2X receptors was amplified from SCG neurons that had been in culture for 24 h (data not shown), indicating that there was no major change in the properties of P2X receptors in these neurons after short-term culture. These results indicate that multiple P2X receptor subunits may be expressed in the SCG and that individual neurons may express multiple P2X receptor subunits, findings that are consistent with those of previous studies (Collo et al., 1996; Dunn et al., 2001).
P2X Receptor-Mediated Currents. In 54.6% of neurons (318 of 582; Fig. 2A) ATP (30 μM-3 mM) evoked fast inward currents that were slowly desensitizing (peak current decayed to 83.6 ± 2.6% after 1-s application, n = 20). In 2.2% (13 of 582) of SCG neurons, the response to ATP had fast and slowly desensitizing components (Fig. 2A) and showed significantly greater decay in current amplitude during ATP application (the peak current decayed to 69.4 ± 7.2% of the peak after 1 s, n = 9). ATP was inactive at the remaining 43.1% (251 of 582) of SCG neurons. However, all of the mouse SCG neurons tested gave inward currents in response to ACh (30 μM-1 mM; n = 113; range for 100 μM ACh, 73-5880 pA). This demonstrates that there is some species variation in the proportion of SCG neurons expressing P2X receptors, because ATP-evoked currents were recorded from 100% of rat (Khakh et al., 1995) and guinea pig (Zhong et al., 2000) SCG neurons.
For the ATP-sensitive neurons, the mean amplitude of slowly desensitizing responses to ATP (100 μM) was 292.4 ± 36.5 pA (n = 137) with a range of 21 to 2060 pA. There was no correlation between amplitude of the current and the membrane capacitance (a measure of cell size; correlation coefficient, 0.1; membrane capacitance, 22.1 ± 1.1 pF; n = 134). The majority of neurons with slowly desensitizing P2X currents had small amplitudes; however, for 8% (11 of 137) of neurons, the ATP current was greater than 1 nA. The neurons that gave combined fast and slow desensitizing P2X currents had ATP (100 μM) evoked amplitudes that were significantly greater (2305.1 ± 457.9 pA; n = 7; p < 0.001) than slowly desensitizing responses and also had higher mean cell capacitance (32.3 ± 2.8; n = 7; p < 0.05).
For the SCG neurons that had a slowly desensitizing P2X receptor phenotype, ATP had a mean EC50 of ∼35 μM (pEC50 of 4.7 ± 0.2, n = 12 neurons; Fig. 3, A and D). However, there was a considerable range in sensitivity, with EC50 values of 2.66 to 120.8 μM, with a mean Hill coefficient of 1.1 ± 0.1. This range of sensitivity is unlikely to result from variation in expression of the receptors (Clyne et al., 2003) because there was no correlation (correlation coefficient -0.03) between the amplitude of the response and the EC50 value.
The dominant mouse SCG P2X receptor is most likely to correspond to the expression of P2X2 receptor subunits (Cockayne et al., 2002). To corroborate this, other characteristics that are specific to the P2X2 receptor were examined (Table 2). Suramin (100 μM) was an antagonist (n = 20 cells) and reduced the response to ATP (100 μM) by 78.9 ± 3.6% (Fig. 3B, n = 5 cells, from one culture, two animals), with recovery to 66.1 ± 7.7% of the initial ATP response after wash-off (10 min). The IC50 for suramin at these neurons was 18.4 μM (n = 5 cells). Iso-PPADS (30 μM) almost completely abolished (98.5 ± 1% inhibition, n = 4 cells from one culture, two animals; Fig. 3C) the response to 100 μM ATP. This effect was only slowly reversible, with a 46 ± 3.7% recovery after a 40-min wash period (n = 3). The P2X2 receptor can also be modulated by extracellular ions including H+, Zn2+ (King et al., 1996), and Ca2+ (Evans et al., 1996). Acidification of the extracellular solution to pH 6.3 potentiated the response to 30 μM ATP by 56 ± 45.5% (n = 7 cells from two cultures, four animals; Fig. 4A). In contrast an increase in extracellular pH to 8.3 attenuated the response to ATP to 24 ± 3.5% of the initial ATP response (30 μM; n = 5 cells from two cultures, four animals). Addition of zinc (10 μM) to the extracellular solution potentiated the P2X current to 233.3 ± 42.9% of the control ATP response (30 μM; n = 6 cells from two cultures, four animals; Fig. 4B). Increasing the extracellular Ca2+ concentration from 2.5 to 30 mM abolished the ATP evoked P2X current in SCG neurons (n = 6 cells, from three cultures, six animals; Fig. 4C; similar results were seen in calcium imaging studies—see below). Taken together, these results (31 cells from 7 cultures from 14 animals) support the conclusions of Cockayne et al. (2002) that the P2X2 receptor phenotype dominates the properties of mouse SCG neurons.
The results from the RT-PCR suggest that receptor subunits other than P2X2 may be expressed in the SCG and that they contribute to the P2X receptor phenotype. α,β-meATP is an agonist at receptors containing P2X1, P2X3, or P2X6 receptor subunits. For neurons that showed slow desensitization to ATP, α,β-meATP (100 μM) evoked small currents (15.0 ± 0.5% of the response to ATP, 100 μM; Fig. 2D) in 25% of cells (two of eight) and was ineffective in the remainder of cells tested (six of eight; Fig. 2E). In contrast, all neurons that had a fast desensitizing component responded to α,β-meATP (three of three), with an inward current of 57.2 ± 10.4% of the response to ATP (100 μM; Fig. 2C). These results show that the P2X receptors in the SCG can be subdivided into at least three pharmacologically distinct P2X receptor phenotypes: 1) slowly desensitizing α,β-meATP insensitive (∼41% of all neurons), 2) slowly desensitizing α,β-meATP-sensitive (∼14% of all neurons), and 3) mixed fast/slow desensitizing α,β-meATP-sensitive neurons (∼2% of all neurons). In the guinea pig SCG α,β-meATP sensitivity is thought to result from the coexpression of P2X2/3 heteromeric channels (Zhong et al., 2000). However there is marked species variation in the properties of SCG P2X receptors; for example, in the rat SCG, α,β-meATP is ineffective as an agonist but produces a concentration-dependent attenuation of the response to ATP (Khakh et al., 1995). The α,β-meATP sensitivity in this study most probably suggests that P2X1, P2X3, and/or P2X6 receptor subunits contribute to the P2X phenotype of ∼10 to 15% of mouse SCG neurons.
P2X Receptor-Mediated Changes in Intracellular Calcium. The functional activity of P2X receptors in mouse SCG neurons was also examined using fluorescent calcium imaging. This technique has the advantage that several neurons can be imaged at once and allows higher throughput for characterization of neuronal P2X receptors. Of the neurons that responded to 60 mM K+ perfusion solution, 80.4% (701 of 872) gave an increase in fluorescence to ATP application (67.3% of all neurons, including those that failed to respond to 60 mM potassium, responded to ATP application). The mean self-ratio for the fluorescence increase to 100 μM ATP was 1.54 ± 0.03 (n = 379), which equates to a rise of 54% above the baseline levels of Ca2+ in the neurons. Removal of Ca2+ from the extracellular perfusion solution resulted in no increase in [Ca2+]i when ATP was applied (n = 26; Fig. 5A), indicating that calcium influx is essential for the response and ruling out a role of ATP sensitive P2Y receptors (calcium responses to the P2Y receptor agonists UTP and UDP were still recorded in the absence of extracellular calcium; data not shown).
ATP evoked concentration-dependent changes in intracellular calcium with an EC50 of ∼26 μM (pEC50 of 4.3 ± 0.1, with a range of 25.7-95 μM and a Hill coefficient of 2 ± 0.3, n = 6; Fig. 6, A and B). In common with P2X receptor-mediated currents, increasing extracellular calcium to 10 mM reduced the rise in intracellular calcium in response to ATP (100 μM) by 89 ± 6% (n = 12; Fig. 5A), indicating the contribution of P2X2 receptor subunits. These results are essentially identical to those determined in patch-clamp studies (Fig. 6C) and indicate that calcium imaging is a robust and reliable method for characterization of P2X receptors in SCG neurons.
The data from both patch clamp and calcium imaging studies on ATP and ionic sensitivity, combined with the slowly desensitizing nature of the P2X response, suggests that the P2X2 receptor subunit dominates the properties of SCG P2X receptors. This is consistent with an abstract on P2X2 receptor knockout mice (Cockayne et al., 2002). However, Cockayne et al. (2002) indicated there was a small residual response to ATP application, suggesting that there were additional subunits contributing to the native P2X receptor phenotype in these neurons. We have confirmed the contribution of other P2X receptor subunits in this study by showing that the P2X1, P2X3, and P2X6 receptor subunit agonist α,β-meATP (100 μM) evoked inward currents in patch-clamp studies and rises in intracellular calcium from ∼10% of neurons (48 of 510 neurons, mean amplitude 28.3 ± 4.4% of the 100 μM ATP response; Fig. 7A). To determine whether P2X1 receptor subunits contribute to the α,β-meATP sensitivity of SCG neurons, we compared normal and P2X1 receptor-deficient mice (Mulryan et al., 2000).
P2X1 Receptor-Deficient Mice Reveal a Role for the P2X1 Receptor in SCG Neurons. In calcium imaging studies, responses to ATP (100 μM) were seen in 44.3% (116 of 262) of neurons from P2X1 receptor-deficient (KO) mice, which is significantly reduced compared with responses seen in WT mice. The magnitude of increase in [Ca2+]i in response to ATP in KO neurons was not significantly different from that of WT, with a mean self-ratio of 1.6 ± 0.06 (n = 134). α,β-meATP (100 μM) caused a rise in [Ca2+]i in only 1.7% (2 of 116) of the neurons that were responsive to ATP (the amplitude of the response to α,β-meATP had self-ratios of 1.16 and 1.47), compared with 9.4% of neurons from WT mice (Fig. 7B). The expression pattern of P2X receptor transcripts between WT and KO mice (data not shown) was the same except for the absence of the P2X1 receptor transcript from SCG from P2X1 receptor-deficient mice and suggests that there is no major compensatory change in P2X receptor expression associated with the P2X1 receptor deficiency. These results demonstrate that the P2X1 receptor is the predominant isoform responsible for ∼80% of the α,β-meATP sensitivity seen in SCG neurons. The reduction in numbers of ATP-sensitive neurons in P2X1 receptor-deficient mice may also indicate a role of the P2X1 receptor in assembly and trafficking of functional P2X receptors in the SCG. The residual α,β-meATP-sensitive response in ∼2% of SCG neurons from P2X1 KO mice suggests that P2X3 or P2X6 receptor subunits may contribute to agonist sensitivity in these neurons.
P2X1 Receptors Form Functional Heteromeric P2X1/2 Channels in the SCG. The data with the P2X1 receptor-deficient mice demonstrate that this receptor subunit gives the α,β-meATP sensitivity to ∼8% of SCG neurons. The α,β-meATP-sensitive currents from the SCG in patch-clamp studies showed relatively slow desensitization (50% decay, >2 s). Slowly desensitizing recombinant P2X1 receptor-mediated responses have been described previously (Parker, 1998). However, the predominant P2X1 receptor phenotype reported for recombinant P2X1 receptors, and for native mouse homomeric P2X1 receptors in the bladder, vas deferens, and arteries, is of rapid desensitization [50% decay, <300 ms (Lewis and Evans, 2000; Mulryan et al., 2000; Vial and Evans, 2000, 2002; North, 2002)]. In addition, the ionic sensitivity of α,β-meATP evoked responses described in this study (see below) are inconsistent with the properties of homomeric P2X1 receptors. Studies on recombinant receptors have indicated that P2X1 receptors can heteropolymerize with other P2X receptor subunits, including the P2X2 isoform (Torres et al., 1999; Brown et al., 2002). Homomeric P2X1 receptors are unaffected by high extracellular calcium (Evans et al., 1996) and increasing pH (Stoop et al., 1997). In contrast, P2X2 receptor subunits are inhibited by high extra-cellular calcium (Evans et al., 1996) and increase in pH (Stoop et al., 1997). When P2X receptor subunits come together to form heteromeric channels, the resultant properties are a composite of those of the constituent subunits; e.g., the P2X2/3 heteromer shows the agonist sensitivity of the P2X3 subunit and the time course of the P2X2 subunit. The properties of heteromeric channels that have been characterized are detailed in Table 2. Knockout studies indicate that the P2X2 receptor subunit predominates in the SCG (Cockayne et al., 2002). To determine whether heteromeric channels comprising the agonist sensitivity of P2X1 receptors and the ionic sensitivity of P2X2 receptors are functionally expressed in the SCG, we have tested the effects of high extracellular calcium and pH on α,β-meATP (100 μM)-evoked calcium responses. Ca2+ (10 mM) in the extracellular solution abolished the response to α,β-meATP in eight of nine cells; in the remaining cell, the response was unaffected and it is possible that this corresponds to the residual α,β-meATP-sensitive response recorded from P2X1 receptor-deficient mouse cultures. Raising the pH to 8.3 reduced the response to α,β-meATP (100 μM) by 93.7 ± 6.3% (n = 9; Fig. 7C). These results, taken together with the ∼80% loss of α,β-meATP-sensitive responses from P2X1 receptor-deficient mouse SCG neurons, suggest that heteromeric channels incorporating the properties of P2X1 and P2X2 receptor subunits are expressed in the SCG.
P2X Receptors and Calcium Signaling in Neurons. Changes in intracellular calcium can have profound effects on neuronal signaling. We have investigated the mechanisms associated with changes in intracellular calcium levels after P2X receptor activation. Calcium influx could be caused by conductance through the P2X receptor, and/or from activation of voltage-gated calcium channels (VGCCs). Extracellular Cd2+ (100 μM) blocks VGCCs and abolished the K+ induced increases in [Ca2+]i in SCG neurons (n = 10). Application of 100 μMCd2+ did not affect the ATP-induced [Ca2+]i increase (100.3 ± 9.1% of the response to 100 μM ATP; n = 8; Fig. 5B), indicating that the calcium rise is caused by the calcium permeability of the P2X receptor. The calcium influx through the P2X receptor could be amplified by calcium-induced calcium release from intracellular stores. Cyclopiazonic acid (CPA) depletes intracellular Ca2+ stores by blocking the endoplasmic reticulum Ca2+ pump. CPA (30 μM with a 15-min preincubation) caused a small increase in the background fluorescence (10.3 ± 2.8%, n = 37) and a change in the time course of the fluorescence signal (n = 37; Fig. 5C). There was a slowing of the initial phase of calcium rise and a reduction in the plateau (60.9 ± 9.6% of control response, n = 37), and this level was maintained until the CPA was washed off. The continued elevation of intracellular Ca2+ was probably caused by block of the neuron's normal calcium sequestration mechanisms. The block was reversible, and ATP responses returned to normal after a 20-min wash period. These results indicate that calcium-induced calcium release contributes to P2X receptor signaling in neurons. This amplification of the response (which would be expected to improve the signal-to-noise ratio) may help to explain the slightly larger proportion of P2X receptor mediated responses recorded from SCG neurons in calcium imaging studies. These studies demonstrate that P2X receptors can act as a direct source for calcium influx into SCG neurons, independent of the activation of voltage-dependent calcium channels. Similar direct influx has been reported in other neurons [e.g., spinal horn interneurons (Rhee et al., 2000)] and is associated with regulation of transmitter release. However, this study also reveals that calcium influx is not the sole source of the rise in intracellular calcium, that calcium-induced calcium release from intracellular stores contributes to the response, and that the ER sequestration of calcium plays a significant role in the termination of the calcium signal after P2X receptor activation.
Discussion
Pharmacological profiles and knockout mice have contributed significantly to our understanding of the physiological roles of some P2X receptor isoforms in the nervous system. For example, a role for P2X3 receptors in sensory perception was initially proposed because of pharmacological studies (Chen et al., 1995; Lewis et al., 1995). This work was confirmed using P2X3 knockout mice, which had reduced nociceptive responses to tissue damage and urinary bladder hyporeflexia (Cockayne et al., 2000; Souslova et al., 2000). In this study, we have characterized P2X receptor-mediated responses in mouse SCG and used P2X1 receptor-deficient mice to demonstrate a contribution of these receptors to neuronal function.
We have shown using a combination of electrophysiology and calcium imaging that functional P2X receptors are expressed on ∼60 to 70% of mouse SCG neurons. In contrast, nicotinic acetylcholine receptor-mediated responses were recorded from all neurons tested. This raises the possibility that P2X receptor expression is coded and corresponds to the target of particular SCG neurons. This has been demonstrated for neuropeptide expression in the rat SCG (Bergner et al., 2000) and for 5-HT3 receptors in superior cervical, dorsal root, and nodose ganglia (Morales and Wang, 2002). The marked heterogeneity in properties of P2X receptors (e.g., the α,β-meATP sensitivity of ∼10 to 15% of neurons in the SCG) also indicates that there is further specialization in receptor expression and function.
The predominant phenotype of P2X receptors in the mouse SCG, based on the time course and antagonist/ionic sensitivity, indicates a substantial contribution of P2X2 receptor subunits; however, our studies did reveal some heterogeneity in the responses. These results are consistent with an abstract on P2X2 receptor knockout mice that showed the P2X2 receptor isoform dominates the properties of SCG P2X receptors (Cockayne et al., 2002). However, the fact that small residual P2X receptor-mediated responses were recorded in SCG neurons from P2X2/P2X3 double-knockout mice (Cockayne et al., 2002) indicates that additional P2X receptor subunits contribute to receptor expression in sympathetic neurons. This is supported by three pieces of evidence: 1) RT-PCR, in situ hybridization, and immunohistochemical studies (Collo et al., 1996; Dunn et al., 2001) indicate that multiple P2X receptor subunits are present in the SCG, and coimmunoprecipitation studies have demonstrated that P2X2 receptor subunits may associate with P2X1,3,5&6 receptor subunits (Torres et al., 1999), 2) patch clamp and calcium imaging studies showed that there was marked variation in ATP sensitivity of P2X receptor-mediated responses (EC50 values ranged from 2.5 to 120 μM); and 3) P2X1 receptor subunit-dependent α,β-meATP sensitivity was demonstrated in a subpopulation of neurons. It is likely therefore that the variation in agonist sensitivity of mouse SCG neuron P2X receptors results, at least in part, from the expression of a range of different homomeric and/or heteromeric subunit combinations within a single neuron, as has been demonstrated for guinea pig SCG neurons (Zhong et al., 2000) and sensory neurons (Grubb and Evans, 1998; Thomas et al., 1998; Burgard et al., 1999). In this study, we have focused on the contribution of P2X1 receptors to the phenotype of SCG neurons, and the contribution of other P2X receptor subunits to P2X receptor-mediated responses in SCG remains to be determined. It is also possible that the properties of SCG P2X receptors may be subject to additional modification by phosphorylation (Boue-Grabot et al., 2000), splice variants (Brandle et al., 1997; Simon et al., 1997), or additional regulatory factors.
The functional characterization of heteromeric P2X receptor channels and their contribution to native phenotypes has relied on distinct properties of heteromeric channels [e.g., α,β-meATP-sensitive, nondesensitizing P2X2/3 receptors (Lewis et al., 1995)]. α,β-meATP sensitivity is associated with the expression of P2X1, P2X3, and P2X6 receptor subunits (Table 2). Previous studies have demonstrated a functional role of α,β-meATP-sensitive P2X1 receptors in smooth muscle and blood cells (Mulryan et al., 2000; Vial et al., 2002; Oury et al., 2003). P2X1 receptor expression in neurons has been shown by in situ hybridization (Collo et al., 1996) and immunohistochemical studies (Vulchanova et al., 1996), and it has been suggested, based on electrophysiological and immunohistochemical studies, that P2X1 receptors contribute to the rapidly desensitizing response in a subset of rat dorsal root ganglion neurons in which P2X3 receptor immunoreactivity was below the limit of detection (Petruska et al., 2000a,b). Once again, this suggests that there may be species variation in P2X receptor-mediated responses in neurons, because α,β-meATP-sensitive P2X receptor currents were abolished in sensory neurons from P2X3 receptor-deficient mice (Cockayne et al., 2000; Souslova et al., 2000; Zhong et al., 2001). In this study, the 80% reduction of α,β-meATP-sensitive responses in neurons from P2X1 receptor-deficient mice shows directly that P2X1 receptor subunits can be functionally expressed in neurons and results in α,β-meATP sensitivity in ∼10% of SCG neurons.
This study demonstrates that P2X1 receptor subunits contribute to the heteromeric assembly of P2X receptors in neurons, because the time course and ionic modulation of P2X1 receptor subunit-dependent α,β-meATP-evoked responses were different from homomeric P2X1 receptor channels (Evans et al., 1996; Stoop et al., 1997; Mulryan et al., 2000). The reason for this preferential incorporation of P2X1 receptor subunits into heteromeric rather than homomeric channels remains to be determined but could reflect the relative abundance of different P2X receptor subunits being processed in the neuron. In calcium imaging studies, we demonstrated that the α,β-meATP-evoked responses show sensitivity to ionic modulation (inhibition by high extracellular calcium and increasing pH) that bear the hallmark of the contribution of P2X2 receptors and not that of homomeric P2X1 receptors (Evans et al., 1996; Stoop et al., 1997). Given the preponderance of functional of P2X2 receptors in the SCG (Cockayne et al., 2002), it seems likely that the α,β-meATP-sensitive receptors described in this study incorporate both P2X1 and P2X2 receptor subunits. Recombinant heteromeric rat P2X1/2 receptors have been described previously in which the phenotype has the rapidly inactivating component of the P2X1 receptor and responses are increased by raised pH (Brown et al., 2002). However, in our study, the abolition of the response to high pH and the relatively slow time course of the α,β-meATP-sensitive response contrast with those of Brown et al. (2002) and suggest that a heteromeric channel incorporating only P2X1 and P2X2 receptor subunits is unlikely to account for the responses observed (unless the responses possibly reflect different subunit assembly stoichiometry or species differences). Because P2X receptors are thought to form as trimers, this suggests that, in addition to P2X1 and P2X2 receptor subunits, an additional P2X receptor subunit contributes to α,β-meATP-sensitive heterotrimeric channels in the SCG.
In summary we have shown there are at least three broad categories of P2X receptor expression in SCG neurons: 1) neurons with a phenotype dominated by the properties of P2X2 receptors, 2) α,β-meATP-sensitive neurons that reflect the expression of P2X1 receptor subunits, most probably as part of heteromeric channels, and 3) neurons that have no detectable functional P2X receptor expression. This study shows a role of P2X1 receptors in neurons. However, in contrast to homomeric P2X1 channel expression in smooth muscle and platelets, we propose that it forms heteromeric P2X receptors, containing P2X1 and P2X2 receptor subunits, in SCG neurons. P2X1 receptors are thought to be expressed widely in the nervous system, and a contribution of this receptor subunit to functional P2X receptors in other neuronal preparations seems likely.
Acknowledgments
We thank Dr. Catherine Vial for her contributions to the RT-PCR studies.
Footnotes
-
This work was supported by a Medical Research Council PhD studentship and the Wellcome Trust.
-
ABBREVIATIONS:SCG, superior cervical ganglia; α,β-meATP, α,β-methylene ATP; RT-PCR, reverse transcription-polymerase chain reaction; WT, wild type; KO, knockout; TH, tyrosine hydroxylase; PPADS, pyridoxal-phosphate-6-azophenyl-2′,4′-disulfonate; Bp, base pair(s); VGCC, voltage-gated calcium channel; CPA, cyclopiazonic acid.
- Received June 17, 2003.
- Accepted September 25, 2003.
- The American Society for Pharmacology and Experimental Therapeutics