HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 567/2/621    most recent jphysiol.2005.088435v1 Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Cockayne, D. A. Articles by Ford, A. P. D. W. Search for Related Content PubMed PubMed Citation Articles by Cockayne, D. A. Articles by Ford, A. P. D. W.

¹ Roche Palo Alto, Palo Alto, CA 94304, USA
² Autonomic Neuroscience Institute, Royal Free and University College Medical School, London NW3 2PF, UK
³ Department of Biomedical Science, University of Sheffield, UK
⁴ Neurorestoration Group, CARD, Kings College London SE1 9RT, UK

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Extracellular ATP plays a role in nociceptive signalling and sensory regulation of visceral function through ionotropic receptors variably composed of P2X₂ and P2X₃ subunits. P2X₂ and P2X₃ subunits can form homomultimeric P2X₂, homomultimeric P2X₃, or heteromultimeric P2X_2/3 receptors. However, the relative contribution of these receptor subtypes to afferent functions of ATP in vivo is poorly understood. Here we describe null mutant mice lacking the P2X₂ receptor subunit (P2X₂^–/–) and double mutant mice lacking both P2X₂ and P2X₃ subunits (P2X₂/P2X₃^Dbl–/–), and compare these with previously characterized P2X₃^–/– mice. In patch-clamp studies, nodose, coeliac and superior cervical ganglia (SCG) neurones from wild-type mice responded to ATP with sustained inward currents, while dorsal root ganglia (DRG) neurones gave predominantly transient currents. Sensory neurones from P2X₂^–/– mice responded to ATP with only transient inward currents, while sympathetic neurones had barely detectable responses. Neurones from P2X₂/P2X₃^Dbl–/– mice had minimal to no response to ATP. These data indicate that P2X receptors on sensory and sympathetic ganglion neurones involve almost exclusively P2X₂ and P2X₃ subunits. P2X₂^–/– and P2X₂/P2X₃^Dbl–/– mice had reduced pain-related behaviours in response to intraplantar injection of formalin. Significantly, P2X₃^–/–, P2X₂^–/–, and P2X₂/P2X₃^Dbl–/– mice had reduced urinary bladder reflexes and decreased pelvic afferent nerve activity in response to bladder distension. No deficits in a wide variety of CNS behavioural tests were observed in P2X₂^–/– mice. Taken together, these data extend our findings for P2X₃^–/– mice, and reveal an important contribution of heteromeric P2X_2/3 receptors to nociceptive responses and mechanosensory transduction within the urinary bladder.

(Received 14 April 2005; accepted after revision 15 June 2005; first published online 16 June 2005)
Corresponding author D. A. Cockayne: Roche Palo Alto, 3431 Hillview Avenue, Palo Alto, CA 94304, USA. Email: debra.cockayne{at}roche.com

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Extracellular ATP communicates physiological signals through the activation of P2X ligand-gated cation channels and P2Y G-protein-coupled receptors. Of the seven P2X receptor subunits (P2X₁–P2X₇), evidence supports a major role for P2X₂ and P2X₃ subunits in mediating the primary sensory effects of ATP (Burnstock, 2000). P2X₃ is expressed almost exclusively on small-to-medium diameter sensory neurones, while P2X₂ is expressed on medium-to-large diameter sensory neurones, and more broadly within the peripheral and central nervous system (Burnstock, 2003). P2X₂ and P2X₃ subunits can form homomultimeric P2X₂, homomultimeric P2X₃, or heteromultimeric P2X_2/3 receptors, each with unique pharmacological properties. P2X₃ receptors are rapidly desensitizing, activated by ,ß-methylene ATP (,ß-meATP), and antagonized by 2',3'-O-trinitrophenyl-ATP (TNP-ATP); while P2X₂ receptors are slowly desensitizing, insensitive to ,ß-meATP, and less sensitive to TNP-ATP than P2X₃ receptors (IC₅₀ 2 µMversus 1 nM for P2X₃) (Virginio et al. 1998). P2X_2/3 receptors have similar pharmacology to P2X₃ receptors, but the slow desensitization kinetics of P2X₂ receptors. In dorsal root ganglia (DRG) sensory neurones, P2X₃ and P2X_2/3 receptors are the predominant P2X receptors (Rae et al. 1998; Burgard et al. 1999), whereas in nodose sensory neurones the effects of ATP appear to be mediated by P2X₂ and P2X_2/3 receptors (Virginio et al. 1998; Thomas et al. 1998). In contrast, rat and mouse autonomic ganglion neurones appear to express mainly P2X₂ receptors (Dunn et al. 2001). Whether P2X₂, P2X₃ and P2X_2/3 receptors account for all of the effects of ATP on sensory and sympathetic ganglion neurones remains unclear.

Numerous studies have shown that P2X₃ and P2X_2/3 receptors play a crucial role in nociception and mechanosensory transduction (Jarvis, 2003), but few tools have been available to define the relative contribution of these receptors to specific sensory behaviours. Studies using pharmacological agents, such as the P2X₁, P2X₃ and P2X_2/3 selective antagonist TNP-ATP (Tsuda et al. 1999a,b; Jarvis et al. 2001; Honore et al. 2002b; Ueno et al. 2003), and the P2X₃, P2X_2/3 selective antagonist A-317491 (Jarvis et al. 2002; McGaraughty et al. 2003; Wu et al. 2004), have shown that peripheral and spinal P2X₃ and P2X_2/3 receptors are involved in transmitting persistent, chronic neuropathic and inflammatory pain. P2X₃-deficient mice (Cockayne et al. 2000; Souslova et al. 2000), and animals treated with P2X₃-selective antisense (Honore et al. 2002a; Barclay et al. 2002; Inoue et al. 2003) or small interfering RNA (siRNA) (Dorn et al. 2004) revealed comparable findings. Tsuda et al. (2000) further suggested that mechanical allodynia is mediated by P2X_2/3 receptors located on capsaicin-insensitive neurones, while thermal hyperalgesia and spontaneous nocifensive behaviours are mediated by P2X₃ receptors on capsaicin-sensitive neurones. These findings suggested that P2X₃ and P2X_2/3 receptors modulate different nociceptive pathways.

P2X₃ receptors also play a role in visceral mechanosensory transduction. In the urinary bladder (Ferguson et al. 1997; Vlaskovska et al. 2001; Sun & Chai, 2002) and ureter (Knight et al. 2002) ATP is released from the urothelium upon distension. Distension leads to increased afferent nerve activity that is mimicked by ATP and ,ß-meATP, and attenuated in P2X₃-deficient mice (Vlaskovska et al. 2001; Rong et al. 2002). ATP and ,ß-meATP can directly stimulate the micturition reflex in conscious rats, and this is inhibited by TNP-ATP (Pandita & Andersson, 2004). Moreover, P2X₃-deficient mice have reduced urinary bladder reflexes (Cockayne et al. 2000). Within the gastrointestinal tract, ATP and ,ß-meATP excite extrinsic (Kirkup et al. 1999; Wynn et al. 2003) and intrinsic (Burnstock, 2001; Bertrand & Bornstein, 2002; Bian et al. 2003) afferents, and P2X₃-deficient mice have impaired peristalsis (Bian et al. 2003).

In the present study, we generated P2X₂^–/– and P2X₂/P2X₃^Dbl–/– mice to characterize further the P2X subunits mediating ATP responses in sensory and autonomic ganglia, and to further assess the role of P2X₂ and P2X₃ subunits in pain-related behaviours and urinary bladder reflexes. We have compared these findings to P2X₃^–/– mice. Our studies confirm an important role for both P2X₂ and P2X₃ subunits in sensory afferent signalling in multiple physiological systems.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Generation of P2X₂^–/– and P2X₂/P2X₃^Dbl–/– mice

The P2X₂ gene was cloned from a 12 kb EcoR1 fragment derived from a mouse embryonic stem (ES) cell 129Ola BAC genomic library (Genome Systems, St Louis, MO, USA). The P2X₂ targeting vector includes a total of 8.0 kb of genomic sequence, a LoxP-flanked neomycin-resistance gene (Neo), and the HSV-TK gene. Homologous recombination between the targeting vector and the wild-type P2X₂ allele results in deletion of a 2.6 kb region of the P2X₂ gene encompassing exons 2–11, and replacement of this sequence with the LoxP-flanked Neo gene. A NotI linearized targeting vector was electroporated into 129Ola-derived E14-1 ES cells, and clones were selected in 310 µg ml^–1 active G418 (Invitrogen, Rockville, MD, USA) and 2 µM gancyclovir (Roche Pharmaceuticals). Southern blot analysis using a 5' flanking region probe identified positive clones with the predicted 12 and 7 kb EcoRI fragments diagnostic of the wild-type and mutant P2X₂ alleles. Targeted clones were injected into C57BL/6 J (Jackson Laboratory, Bar Harbour, ME, USA) blastocysts and germline transmission of the targeted allele was established by mating the resulting male chimeras to C57BL/6 J females. P2X₂ wild-type and mutant lines were generated on a mixed 129Ola x C57BL/6 J genetic background according to the strategy described at the Banbury Conference on Genetic Background in Mice (Silva et al. 1997). Briefly, F2 P2X₂ homozygous mutant mice were generated by intercrossing F1 agouti heterozygotes. In contrast, F2 wild-type controls were produced by intercrossing F1 agouti wild-type mice and screening for F2 progeny in which the P2X₂ locus is homozygous for the 129Ola genetic background (i.e. derives from the genetic background of the ES cell). F2129Ola/129Ola P2X₂ wild-type mice were identified using a PCR assay that distinguishes a sequence polymorphism upstream of the P2X₂ gene between 129Ola and C57BL/6 J strains. The primers used for this assay were: P2X₂ 5'901S, 5'-CCT TCT GAG TGT GAG GAT GAG-3'; and P2X₂ 5'1181AS, 5'-CAT ACT TAG CAA CCT GCC TAT C-3'. The assay generated amplicons of 280 and 305 bp for the 129Ola and C57BL/6 J alleles, respectively. F2 P2X₂ wild-type mice (129Ola/129Ola at the P2X₂ locus), and F2 P2X₂ homozygous mutant mice (129Ola x C57BL/6 J) were used to establish homozygous wild-type and mutant breeding colonies. All studies were performed on F3 mice derived from these crosses. For Southern blot genotyping assays, an internal exon 1 probe was used with a BglII digest (6.7 kb wild-type and 5.0 kb mutant allele) and an EcoRI digest (12 kb wild-type and 7.0 kb mutant allele). In addition, a three-primer PCR assay was used that generated amplicons of 194 and 350 bp for the P2X₂ wild-type and mutant alleles, respectively. The primers were as follows: P2X₂ EX11288S, 5'-GTG CAG CTG CTC ATT CTG CTT-3'; P2X₂ EX2482AS, 5'-CTG CAC GAT GAA GAC GTA CCT-3'; NEO2S, 5'-ACG AGT TCT TCT GAG GGG ATC GGC-3'.

P2X₂/P2X₃^Dbl–/– mice were generated by crossing 129Ola x C57BL/6 P2X₃^–/– F2 mice (Cockayne et al. 2000) to 129Ola x C57BL/6 J P2X₂^+/– F2 mice to generate P2X₂/P2X₃ compound heterozygotes. Compound heterozygotes were bred to generate homozygous P2X₂/P2X₃ double wild-type and double knockout mice. All mice used for studies were derived from subsequently established homozygous breeding pairs, and are designated P2X₂/P2X₃^Dbl+/+ and P2X₂/P2X₃^Dbl–/– mice. The only exceptions are the wild-type controls used for formalin studies on P2X₂/P2X₃^Dbl–/–, where B6129PF2/J mice from The Jackson Laboratory were used in place of homozygous-derived P2X₂/P2X₃^Dbl+/+. All animal use procedures were overseen and approved by the Roche Palo Alto Institutional Animal Care and Use Committee, or by the UK Home Office.

Histopathology and immunohistochemistry

For histopathology, organs were fixed in 10% formalin, embedded in paraffin, and sectioned at 5 µm. Sections were stained with haematoxylin and eosin, and examined under a light microscope. For immunohistochemistry, adult mice were killed by CO₂ inhalation. DRG, nodose ganglia, trigeminal ganglia and superior cervical ganglia (SCG) were dissected and immersed in 4% paraformaldehyde and 0.2% saturated picric acid in 0.1 M phosphate-buffered saline (PBS, pH 7.2) for 4 h. Tissues were saturated with 20% sucrose in PBS, after which the ganglia were rapidly frozen, cut into 10 µm sections, and thaw-mounted on poly-L-lysine-coated slides. For immunofluorescence double labelling, sections were incubated in 10% normal horse serum in PBS for 30 min at room temperature, followed by incubation with a rabbit polyclonal antibody against the rat P2X₂ receptor subtype (Roche) (1: 800) in 10% normal horse serum and 0.1% Triton X-100 in PBS overnight. Sections were then incubated with biotinylated donkey anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA, USA) (1: 500) for 1 h, extravidin peroxidase (1: 1500) for 1 h, biotinyl tyramide for 8 min, and Streptavidin fluorescein (1: 200) for 10 min. A rabbit polyclonal antibody against the P2X₃ receptor subtype (Roche) (1: 400) was applied as a second primary antibody and detected with Cy3-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch). The sections were washed and mounted in citifluor.

Compounds used

Atropine, ATP, ß,-methylene ATP (ß,-meATP), ,ß-meATP, carbachol, dimethylphenylpiperazinium (DMPP), histamine, 5-HT, substance P, –aminobutyric acid (GABA) and tetrodotoxin (TTX) were obtained from Sigma Chemical Company (Poole, UK). Pyridoxal-5-phosphate-6-azophenyl-2',4'-disulphonic acid (PPADS) (tetrasodium salt) was supplied by Tocris Cookson Ltd (Bristol, UK). TNP-ATP was obtained from Molecular Probes (Leiden, Netherlands). Stock solutions of drugs were prepared using deionized H₂O and stored frozen. diinosine pentaphosphate (Ip₅I) was prepared by enzymatic degradation of diadenosine pentaphosphate (Ap₅A) (King et al. 1999).

Primary cell culture and electrophysiology

Single neurones from DRG, nodose, SCG and coeliac ganglia of 6- to 8-month-old mice were enzymatically isolated as previously described (Zhong et al. 1998). Briefly, mice were killed by CO₂ inhalation. Ganglia were rapidly dissected and placed in Leibovitz's L-15 medium (Life Technologies, Paisley, UK). Ganglia were desheathed, cut and incubated in 4 ml Ca²⁺/Mg²⁺-free Hanks' balanced salt solution with 10 mM Hepes buffer (pH 7.0) (HBSS) (Life Technologies) containing 1.5 mg ml^–1 collagenase (Class-II; Worthington Biochemical Corporation, Reading, UK) and 6 mg ml^–1 bovine serum albumin (Sigma) at 37°C for 40 min. Ganglia were then incubated with 4 ml HBSS containing 1 mg ml^–1 trypsin (Sigma) at 37°C for 20 min. The solution was replaced with 3 ml of growth medium comprising L-15 medium supplemented with 10% bovine serum, 50 ng ml^–1 nerve growth factor, 0.2% NaHCO₃, 5.5 mg ml^–1 glucose, 200 IU ml^–1 penicillin and 200 µg ml^–1 streptomycin. The ganglia were dissociated into single neurones by gentle trituration. The cells were then centrifuged at 160 g for 5 min, resuspended in 1 ml of growth medium, and plated onto 35 mm Petri dishes coated with 10 µg ml^–1 laminin (Sigma). Cells were maintained at 37°C in a humidified atmosphere containing 5% CO₂, and used between 2 and 48 h after plating.

Whole-cell voltage-clamp recordings were performed at room temperature using an Axopatch 200B amplifier (Axon Instruments, Union City, CA, USA). Membrane potential was held at –60 mV. The external solution contained (mM): 154 NaCl, 4.7 KCl, 1.2 MgCl₂, 2.5 CaCl₂, 10 Hepes, 5.6 glucose, and the pH was adjusted to 7.4 using NaOH. Recording electrodes (resistance 2–4 M) were filled with an internal solution which contained (mM): 56 citric acid, 3 MgCl₂, 10 CsCl, 10 NaCl, 40 Hepes, 0.1 EGTA, 10 tetraethylammonium chloride, and the pH was adjusted to 7.2 using CsOH (total Cs⁺ concentration 170 mM). Series resistance compensation of 72–75% was used in all recordings. The threshold for the minimum detectable response was set as 10 pA. Data were acquired using pCLAMP software (Axon Instruments). Signals were filtered at 2 kHz (–3 dB frequency, Bessel filter, 80 dB decade^–1).

For nodose, SCG and coeliac neurones, compounds were applied by gravity flow from independent reservoirs through a 7-barrel manifold comprising fused glass capillaries inserted into a common outlet tube (tip diameter of 200 µm) which was placed about 200 µm from the cell (Dunn et al. 1996). One barrel was used to apply drug-free solution to enable rapid termination of drug application. Solution exchange measured by changes in open tip current was complete in 200 ms; however, complete exchange of solution around an intact cell was slower (1 s). For DRG neurones, drugs were applied through a similar 4-barrel manifold controlled by computer-driven solenoid valves. The exchange of solution around the cell was complete in 100 ms. The intervals between agonist applications were 2 min for nodose, SCG and coeliac ganglia neurones, and 3.5 min for DRG neurones. These were sufficient to achieve reproducible responses. All drugs were prepared from stock solutions and diluted in extracellular bathing solution to the final concentration. Traces were acquired using Fetchex (pCLAMP software) and plotted using Origin (Microcal, Northampton, MA, USA).

Nociceptive behavioural testing

The effect of a subplantar injection of P2 receptor agonists or 5% formalin was performed as described (Cockayne et al. 2000). Mice were injected intradermally in the plantar surface of one hindpaw using a 30 g needle with varying doses of ATP or ,ß-meATP in a total volume of 30 µl, or 5% formalin in a total volume of 20 µl. For P2X agonists, the total time spent lifting the treated paw in a 4 min time bin was measured. For formalin testing, the total time spent licking, biting or flinching the treated paw was measured every 5 min between 0 and 30 min. All animals were killed at the end of the study. The investigator was blinded to the identity of all animals.

MiniPsychoscreen behavioural testing

P2X₂-deficient mice were evaluated in a MiniPsychoscreen behavioural test battery at Psychogenics, Inc., Tarrytown, NY, USA (see http://www.psychogenics.com/services.shtml for a complete description of the Psychoscreen battery and methods). Male P2X₂^+/+ and P2X₂^–/– mice were shipped to Psychogenics at 10–12 weeks of age, and were singly housed upon receipt. Mice were allowed to habituate to the colony for 2 weeks, and were handled for 3 days during this time. The test battery included home cage observations in the colony, open field, Irwin test, rotorod, grip strength, visual cliff, tail suspension, tail flick, plantar test, elevated plus maze, place recognition y-maze, prepulse inhibition of startle and pentylenetetrazole (PTZ)-induced seizure. The investigator was blinded to the identity of all animals.

Bladder in vitro pharmacology

For in vitro tissue bath studies, mice were killed and urinary bladders were dissected into a modified Krebs solution. The tissues were viewed under a dissecting microscope, stripped of adhering fat and connective tissue, and cut to give two strips of detrusor muscle, approximately 10 x 2 mm, from each bladder. Detrusor strips were weighed and prepared for isolated organ bath recording using a dissecting microscope. Silk ligatures were applied to each end of the strip; one end was attached to a rigid support and the other end to an FT03C force displacement transducer. Each strip was suspended in a 10 ml organ bath containing gassed (95% O₂ and 5% CO₂) modified Krebs solution of the following composition (mM): 133 NaCl, 4.7 KCl, 16.4 NaHCO₃, 0.6 MgSO₄, 1.4 NaH₂PO₄, 7.7 glucose and 2.5 CaCl₂, pH 7.3. Experiments were performed at 37 ± 1°C. Separate experiments were performed to study electrical field stimulation (EFS) of the parasympathetic nerves and the action of exogenously applied agonists. Mechanical activity was recorded using the software PowerLab Chart for Windows (version 4; ADInstruments, Castle Hill, NSW, Australia). An initial load of 1 g was applied to the detrusor strips, which were then allowed to equilibrate for not less than 45 min prior to the start of the experiment. The contraction due to a standard concentration of KCl (120 mM) was noted at the end of each experiment. All compounds were prepared from stock solutions and added to the organ bath (volume, 100 µl) to produce the final concentration.

Frequency–response curves to EFS were constructed for detrusor strips (100 V, 0.3 ms, 0.5–32 Hz, 15 s stimulation every 5 min) in the absence of antagonists, and then in the presence of either PPADS (30 µM, one detrusor strip from each bladder) or atropine (1 µM, one detrusor strip from each bladder). The frequency–response curves were then repeated in the presence of both PPADS (30 µM) and atropine (1 µM). All antagonists were incubated for 20 min. The curves were finally repeated in the presence of tetrodotoxin (TTX, 1 µM for 20 min). Detrusor responses to EFS are expressed as means ±S.E.M. of the percentage maximum control response for n= 5–6 determinations.

Non-cumulative concentration–response curves were constructed for various agonists including carbachol (0.01–300 µM, n= 5–11), ß,-meATP (0.1–300 µM, n= 5–7), and ATP (0.1 µM–1 mM, n= 5–6), and contractile responses are expressed as means ±S.E.M. of the percentage of the KCl (120 mM) contraction for n determinations. The time interval between application of ß,-meATP and ATP was 15–20 min to avoid desensitization. Since the concentration–response curves to these agonists did not reach a maximum, it was not possible to calculate pD2 values (–log EC₅₀ concentration). Contractile responses to single concentrations of substance P (0.3 µM, n= 6–18), 5-HT (100 µM, n= 4–17), and histamine (100 µM, n= 4–16) were also determined, and are expressed as means ±S.E.M. of the milligrams of tension developed for n determinations.

For in vitro pharmacology, statistical significance between individual groups of mice was tested by a two-way analysis of variance (ANOVA) followed by a post-hoc test. Statistical significance between single concentrations of substance P, 5–HT, and histamine for individual groups of mice was tested using a one-way ANOVA followed by Bonferroni's post-hoc analysis.

Mouse cystometry

Mice were anaesthetized with urethane, and transurethral closed cystometry was conducted as previously described (Dmitrieva et al. 1997; Cockayne et al. 2000). The bladder was cannulated transurethrally with a PE-10 polypropylene catheter. A ventral midline laparotomy was performed to confirm complete bladder emptying. Each cystometrogram consisted of slowly filling the bladder with normal saline through the transurethral catheter at a rate of 10 µl min^–1, to a total volume of 0.4 ml. The filling rate was controlled by an infusion pump (Harvard Apparatus, Holliston, MA, USA), and the pressure associated with filling was recorded via a pressure transducer using PowerLab (ADInstruments). Contractions greater than 20 cm of H₂O were taken as micturition contractions. For each cystometrogram, the volume at which active contractions occurred (micturition threshold) and the number of contractions per cystometrogram were recorded. The investigator was blinded to the identity of all animals.

Bladder afferent nerve recordings

The mouse urinary bladder/pelvic nerve preparation has been previously described in detail (Vlaskovska et al. 2001). Briefly, the whole urinary tract attached to the lower vertebrae and surrounding tissues was isolated en bloc and superfused in a recording chamber with oxygenated (5% CO₂ and 95% O₂) Krebs solution (mM: 120 NaCl, 5.9 KCl, 1.2 NaH₂PO₄, 1.2 MgSO₄, 15.4 NaHCO₃, 2.5 CaCl₂, and 11.5 glucose) at 26°C. The bladder was catheterized through the urethra and connected to a syringe-type infusion pump (sp210iw; World Precision Instruments, Sarasota, FL, USA) for intraluminal infusion with Krebs (0.1 ml min^–1). Another double-lumen catheter was inserted into the bladder through a small cut on the bladder dome for measurement of intraluminal pressure and for drainage. Both catheters were secured in place with fine sutures.

With the aid of a dissecting microscope, the pelvic nerve exiting the vertebrae was dissected. Normally, 3–4 nerve branches could be identified at this location and the finest branch was recorded with a suction glass electrode (tip diameter, 50 µm). The electrode was connected to a NeuroLog head stage (NL 100; Digitimer, UK) and an AC amplifier (NL 104; Digitimer, UK). Signals were amplified (x10 000), filtered (band-pass 200–4000 Hz), and acquired by a computer (sampling frequency 20 000 Hz) with a power 1401 A/D interface and Spike2 software (Cambridge Electronic Design, Cambridge, UK). The nerve activity was counted and plotted as rate histogram.

Following a 60 min stabilization period, repeated ramp distensions (0.1 ml min^–1, up to 50 mmHg) were performed at intervals of 10–15 min. The afferent response to distension was initially variable but usually became stabilized after three to five distensions. This stabilized afferent response was used for comparing mechanosensitivity of bladder afferents among different mouse genotypes. The volume and pressure threshold for whole nerves could be determined as the bladder afferents had no spontaneous activity with the bladder empty (e.g. Fig. 10, top panel, NA trace). In many preparations, single-unit activity could be discriminated based on the amplitude and shape of individual waveforms (e.g. Fig. 10, top panel, insert). The afferent activity at different pressure levels was determined using a script for Spike2 and the pressure–response curve was plotted in Graphpad using the Boltzmann sigmoidal equation. Values are expressed as means ±S.E.M., and the statistical significance between genotypes was compared using ANOVA.

View larger version (30K): [in this window] [in a new window]   Figure 10.  Threshold for whole-nerve response to bladder distension and single unit discrimination reveals a reduced sensitivity of pelvic afferents in P2X-deficient mice Top panel, the upper three traces are intraluminal pressure (IP), whole-nerve activity (NA) and whole-nerve discharge rate (frequency). The lower two traces are a single-unit spike (Unit) and its discharge rate (unit frequency). In the inset are superimposed spikes of that unit showing little variation in waveform. The dotted line indicates the start of bladder filling and the second solid line indicates the start of nerve firing (pressure threshold). Volume threshold is calculated from the interval between those two lines and the rate of bladder filling (0.1 ml min–1). Bottom panel, reduced sensitivity of pelvic afferents to bladder filling in P2X3–/–, P2X2–/– and P2X2/P2X3Dbl–/– mice. A, volume thresholds were significantly increased in P2X-deficient mice (n= 7–10) compared with the respective wild-type controls (n= 6–12). B, pressure thresholds were not significantly different between P2X-deficient mice (n= 6–10) and wild-type controls (n= 6–9). C, the pressure–response curves of whole nerves (C) and single units (D) in P2X-deficient mice showed shallower slopes compared with wild-type mice. *P < 0.05 and **P < 0.01 comparing the mean discharge frequency of wild-type mice with P2X-deficient mice at various intraluminal pressure levels; two way analysis of variance (ANOVA).

	Results

Top Abstract Introduction Methods Results Discussion References

Mice carrying a targeted deletion in exons 2–11 of the P2X₂ gene were generated using the gene targeting strategy shown in Fig. 1A and B). P2X₂ heterozygous (P2X₂^+/–) mice developed normally, but when intercrossed produced fewer than expected P2X₂^–/– mice (18%versus the expected 25% by Mendelian distribution, n= 794, P < 0.0001, ² test). Although P2X₂^–/– mice showed small differences in body weight compared with P2X₂^+/+ mice (Table 1), these mice were visibly and histopathologically normal for up to 1 year of age. Urinalysis, blood chemistries, and peripheral blood cell counts were also normal and similar between P2X₂^+/+ and P2X₂^–/– mice (data not shown). Overall, these findings were similar to those observed for P2X₃^–/– mice (Table 1 and Cockayne et al. 2000).

View larger version (27K): [in this window] [in a new window]   Figure 1.  Targeted disruption of the P2X2 gene and generation of P2X2–/– and P2X2/P2X3Dbl–/– mice A, gene targeting strategy showing the 12 kb genomic fragment of the mouse P2X2 gene used for deletion of exons (filled bars) 2–11. Dashed lines represent the 5' and 3' genomic arms of the targeting vector. Position of the LoxP-flanked Neo and TK selection markers are indicated. Restriction sites indicated are R, EcoRI; H, HindIII; Bg, BglII; RV, EcoRV; K, KpnI; N, NotI. A, NotI linearized targeting vector was transfected in 129Ola E14-1 ES cells to generate the targeted P2X2 locus. B, BglII Southern blot used for routine genotyping of P2X2+/+, P2X2+/– and P2X2–/– mice using the indicated exon-1-specific probe. P2X2/P2X3 double knockout mice were generated by conventional backcrossing as described in Methods. C, EcoR1 Southern blot of C57BL/6 J, P2X2/P2X3Dbl+/+ and P2X2/P2X3Dbl–/– mice probed simultaneously with a P2X2 exon-1-specific probe and a P2X3 5' flanking region probe (Cockayne et al. 2000). The P2X2 and P2X3 wild-type alleles in the P2X2/P2X3Dbl+/+ mice are represented by the expected 12 and 3 kb fragments, respectively, and the P2X2 and P2X3 mutant alleles in P2X2/P2X3Dbl–/– mice are represented by the expected 7 and 1.5 kb fragments, respectively.

As shown in Fig. 2, immunohistochemical staining of sensory (Fig. 2A–D) and sympathetic (Fig. 2E and F) ganglion neurones confirmed that P2X₂^–/– mice lack P2X₂ receptor protein (Fig 2B and F), and that P2X₂/P2X₃^Dbl–/– mice lacked protein for both the P2X₂ and P2X₃ receptor subunits (Fig. 2D). Despite the loss of P2X receptor subunits, histological staining indicated no apparent differences in the appearance, density or size distribution of neuronal cell bodies in sensory or autonomic ganglia in these mice (data not shown). There were also no apparent abnormalities in support cells or in the nerve roots around the ganglia (data not shown). These findings are consistent with our previous observations in P2X₃^–/– mice (Cockayne et al. 2000).

View larger version (145K): [in this window] [in a new window]   Figure 2.  Double immunofluorescence labelling of P2X2 and P2X3 receptors in sensory and sympathetic neurones A–D, immunofluorescence doubling labelling for P2X2 (green) and P2X3 (red) receptors in transverse sections (10 µm) of DRG (colocalization, yellow) neurones from P2X2+/+ (A), P2X2–/– (B), P2X2/P2X3Dbl+/+ (C) and P2X2/P2X3Dbl–/– (D) mice. P2X2 and P2X3 immunoreactivity was present in small-large and small-medium cells, respectively, in P2X2+/+ (A) and P2X2/P2X3Dbl+/+ (C) mice, and many DRG neurones showed colocalization of P2X2 and P2X3 receptors. P2X2 immunoreactivity was absent in P2X2–/– mice (B), while P2X3 immunoreactivity appeared unaltered. Both P2X2 and P2X3 immunoreactivity was absent in P2X2/P2X3Dbl–/– mice (D). E and F, immunofluorescence doubling labelling for P2X2 (green) and P2X3 (red) receptors in transverse sections (10 µm) of superior cervical ganglia (SCG) neurones from P2X2+/+ (E) and P2X2–/– (F) mice. Many SCG neurones in P2X2+/+ mice showed P2X2 receptor expression (E), and this P2X2 immunoreactivity was absent in P2X2–/– mice (F). Specific P2X3 immunoreactivity was not observed in SCG neurones of either P2X2+/+ or P2X2–/– mice. Scale bars, 50 µM.

Peripheral sensory and autonomic ganglion neurones differentially express P2X₂ and P2X₃ subunits, and show interspecies differences in their distribution. P2X₂, P2X₃ and P2X_2/3 receptors are dominant in rodent sensory ganglia (North, 2002; Burnstock, 2003), while P2X₂ receptors, and possibly a heteromeric receptor containing P2X₂, are found in rodent autonomic ganglia (Zhong et al. 2000; North, 2002; Calvert, 2004). Consistent with these data, P2X₂^+/+ and P2X₂/P2X₃^Dbl+/+ mice had strong P2X₃ and P2X₂ immunoreactivity in many small-to-medium and small-to-large diameter sensory neurones, respectively, of the DRG, nodosel and trigeminal ganglia, and many neurones showed colocalization of P2X₂ and P2X₃ subunits (Fig. 2A and C for DRG, and data not shown). In contrast, in sympathetic SCG neurones of P2X₂^+/+ mice, most neurones had P2X₂ but not P2X₃ immunoreactivity (Fig. 2E), consistent with reports that P2X₃ subunits do not appear to be prevalent in rat and mouse autonomic ganglion neurones (Khakh et al. 1995; Zhong et al. 2000; North, 2002).

To determine the contribution of P2X₂ and P2X₃ receptor subunits to ATP-mediated responses in distinct populations of peripheral neurones, we performed whole-cell patch-clamp analysis on dissociated sensory and sympathetic ganglion neurones from wild-type, P2X₂^–/– and P2X₂/P2X₃^Dbl–/– mice. Results were compared to whole-cell patch-clamp responses from P2X₃^–/–mice.

In DRG neurones from wild-type mice, ATP and ,ß-meATP typically evoked either transient (Fig. 3Aa) or sustained (Fig. 3Ab) responses, although in a few neurones composite responses with both transient and sustained components were observed (data not shown). For P2X₂^+/+ mice 81% (17/21) of the neurones tested responded to 100 µM ATP with either a transient (0.28 ± 0.05 nA, n= 12) or a sustained (1.22 ± 0.04 nA, n= 5) inward current (Table 2). The remaining four neurones failed to respond. In response to 30 µM,ß-meATP, 57% (12/21) of P2X₂^+/+ DRG neurones responded with a transient response of 0.23 ± 0.05 nA, while 23% (5/21) responded with a sustained response of 0.35 ± 0.1 nA, and four cells failed to respond to this agonist (data not shown). In P2X₂^–/– neurones 70% (17/24) of the neurones gave transient currents to 100 µM ATP (0.5 ± 0.09 nA), while 20% (5/24) gave small sustained responses (0.05 ± 0.02 nA) that were significantly different from those seen in the P2X₂^+/+ DRG neurones (1.22 ± 0.38 nA, P < 0.05) (Fig. 3B and Table 2). In response to 30 µM,ß-meATP, 65% (15/23) of DRG neurones responded with a transient response of 0.22 ± 0.06 nA, one cell gave a sustained response of 0.02 nA, and the remaining 7/23 cells failed to respond to this agonist (Fig. 3B, and data not shown). As previously described (Cockayne et al. 2000), all DRG neurones from P2X₃^–/– mice failed to respond to ,ß-meATP (Fig. 3Ca), and although 10% of neurones did respond to ATP, these were invariably of the sustained type (Fig. 3Cb and Table 2).

View larger version (18K): [in this window] [in a new window]   Figure 3.  Whole-cell patch-clamp recordings of DRG neurones from P2X2–/–, P2X3–/– and P2X2/P2X3Dbl–/– mice in response to P2X agonists A, wild-type dorsal root ganglia (DRG) neurones (representative traces from P2X2/P2X3Dbl+/+ mice) responded to ATP and ,ß-meATP with either rapidly desensitizing (a) or sustained (b) responses; a composite response having both rapidly and slowly desensitizing components was also observed in some neurones (data not shown). All DRG neurones examined responded to 100 µM GABA with a sustained inward current. Comparable responses were seen in DRG neurones from all wild-type lines. B, in P2X2–/– mice, DRG neurones all responded to ATP and ,ß-meATP with rapidly desensitizing transient responses. C, in P2X3–/– mice, many DRG neurones failed to respond to either ATP or ,ß-meATP, but did respond to 100 µM GABA (a). Other P2X3–/– neurones responded to ATP with a sustained inward current, but failed to respond to ,ß-meATP (b). D, in P2X2/P2X3Dbl–/– mice, most DRG neurones failed to respond to ATP or ,ß-meATP, but did respond to 100 µM GABA (a). A small percentage of neurones in double knockout mice gave small, very low amplitude responses to ATP (b), but did not respond to ,ß-meATP (data not shown).

In nodose ganglion neurones from P2X₂^+/+ mice, 100% (9/9) of the neurones tested responded to 100 µM ATP and ,ß-meATP with sustained inward currents averaging 3.8 ± 1 and 1.97 ± 0.6 nA, respectively (Fig. 4A top panel, and Table 2). In contrast, 88% (15/17) of P2X₂^–/– nodose neurones responded to ATP and ,ß-meATP, but with only a rapidly desensitizing response, such that by the end of a 1 s application the response was approximately 15% of the peak amplitude (100 µM ATP, average 0.54 ± 0.12 nA; 100 µM,ß-meATP, average 0.97 ± 0.22 nA) (Fig. 4A lower panel, and Table 2). These remaining transient responses were sensitive to the P2X antagonists Ip₅I (Fig. 4B and C) and TNP-ATP (Fig. 4C). In the presence of 1 µM Ip₅I, the response to 10 µM,ß-meATP was reduced by approximately 50%, while 10 nM TNP-ATP nearly abolished the response. In nodose neurones from P2X₂/P2X₃^Dbl+/+ mice, 100% (21/21) of the neurones tested responded to 100 µM ATP with a sustained inward current (3.8 ± 0.4 nA) (Fig. 4D and Table 2), while in P2X₂/P2X₃^Dbl–/– nodose neurones only 26% (7/27) of the neurones tested responded to 100 µM ATP with barely detectable currents (0.04 ± 0.01 nA) (Fig. 4D and Table 2). Taken together, these data indicate that P2X₂ and P2X₃ are the major P2X receptor subunits in nodose and DRG peripheral sensory neurones.

View larger version (20K): [in this window] [in a new window]   Figure 4.  Whole-cell patch-clamp recordings of nodose ganglion neurones from P2X2–/– and P2X2/P2X3Dbl–/– mice in response to P2X agonists A, comparison of responses to ATP and ,ß-meATP recorded from nodose ganglion neurones of P2X2+/+ and P2X2–/– mice. Wild-type neurones gave a sustained inward current to both agonists, while neurones from P2X2–/– mice responded with a rapidly desensitizing transient response. B and C, antagonist sensitivity of the remaining ,ß-meATP response recorded in P2X2–/– nodose neurones. B, representative responses to 10 µM,ß-meATP recorded from a single nodose neurone before, in the presence of, and after washing out the antagonist 1 µM diinosine pentaphosphate (Ip5I). C, histogram comparing the antagonist effect of 1 µM Ip5I with that of 3 nM and 10 nM TNP-ATP. Data represent means ±S.E.M. from the number of neurones shown in parenthesis. D, comparison of responses to ATP recorded in nodose ganglion neurones from P2X2/P2X3Dbl+/+ and P2X2/P2X3Dbl–/– mice.

View larger version (12K): [in this window] [in a new window]   Figure 5.  Whole-cell patch-clamp recordings of SCG neurones from P2X2–/– mice in response to ATP A, SCG neurones from P2X2+/+ mice gave a robust sustained inward current to ATP and the nicotinic agonist DMPP, but not to ,ß-meATP. B, SCG neurones from P2X2–/– mice responded to DMPP, but there was no response to ATP or ,ß-meATP.

Intradermal administration of ATP or ,ß-meATP into the hindpaw of rodents elicits a nocifensive behavioural response (Bland-Ward & Humphrey, 1997). P2X₃-containing receptors have been shown to mediate part of this response (Bland-Ward & Humphrey, 1997; Tsuda et al. 2000; Cockayne et al. 2000; Jarvis et al. 2001; Honore et al. 2002a; McGaraughty et al. 2003), but the relative contribution of homomeric P2X₃ and heteromeric P2X_2/3 receptors to this behaviour is not entirely clear. When varying doses of ATP (100 and 300 nmol) or ,ß-mATP (30, 100 and 300 nmol) were injected into the hindpaw of P2X₂^+/+ and P2X₂^–/– mice, both groups of mice produced similar responses that showed a dose-related increase in the total time spent lifting the treated hindpaw (Fig. 6A).

View larger version (16K): [in this window] [in a new window]   Figure 6.  P2X agonist and formalin-evoked nociceptive responses in P2X2–/– and P2X2/P2X3Dbl–/– mice A, hindpaw lifting response in 2- to 3-month-old female P2X2+/+ (open bars) and P2X2–/– (filled bars) mice (n= 8–10) following intraplantar injection of varying doses of ATP or ,ß-meATP in a total volume of 30 µl. Data represent the means ±S.E.M. of the total time spent lifting the treated paw in a 4 min time bin following administration of compound. B and C, nociceptive behavioural response in 3-month-old male and female P2X2+/+ (B, open bars) and P2X2–/– (B, filled bars) mice (n= 9–11), or 3- to 4-month-old male P2X2/P2X3Dbl+/+ (C, open bars) and P2X2/P2X3Dbl–/– (C, filled bars) mice (n= 12) following intraplantar injection of 5% formalin in a total volume of 20 µl. Data represent the means ±S.E.M. of the total time spent licking, biting or flinching the treated paw in the 0–5 and 5–30 min time bins following administration of formalin. *P < 0.05, **P < 0.01; Wilcoxon rank-sum exact test.

We also measured responses to noxious thermal stimuli in P2X₂^+/+ and P2X₂^–/– mice in both the plantar test and the tail-flick test, where infrared light was used as the thermal stimulus. No differences were seen between P2X₂ genotypes in these tests of acute thermal nociception (data not shown, MiniPsychoscreen test battery).

Because P2X₂ receptors are widely distributed on neurones throughout the peripheral and central nervous system, as well as on non-neuronal cells (Burnstock, 2003), they could conceivably play a role in a broad range of behavioural and neurological functions. To assess the overall integrity of the peripheral and central nervous system, P2X₂^–/– mice were profiled in a MiniPsychoscreen battery of behavioural tests. No differences were observed between P2X₂^+/+ and P2X₂^–/– mice in terms of general sensory function and motor coordination as assessed by a neurological evaluation that included the Irwin test, visual cliff, grip strength and rotorod performance (data not shown). P2X₂^–/– mice had a tendency toward hyperactivity in the open field test and in home cage behaviours. However, they showed no altered behaviour in depression or anxiety models, including the tail suspension test, open field test and the elevated plus maze (data not shown). P2X₂^–/– mice also showed no difference from wild-type controls in response to pentylenetetrazole (PTZ)-induced seizures (data not shown), indicating that there were no gross alterations in the general excitability of the CNS.

Urinary bladder reflexes in P2X₃^–/–, P2X₂^–/– and P2X₂/P2X₃^Dbl–/– mice

Cystometry studies were performed on P2X-deficient mice to assess the role of P2X receptor subtypes in sensory afferent signalling within the urinary bladder. Figure 7A illustrates representative cystometrograms from P2X₂^+/+ and P2X₂^–/– mice in a urodynamic model of acute cystometry performed under anaesthesia. Distension of the urinary bladder with a fixed infusion rate of saline (10 µl min^–1, 0.4 ml in 40 min) resulted in frequent micturition contractions in P2X₂^+/+ mice, but markedly reduced contractions in P2X₂^–/– mice. The volume threshold for micturition contractions was significantly increased in P2X₂^–/– mice (0.28 ± 0.02 ml compared with 0.16 ± 0.02 ml for P2X₂^+/+ mice, P < 0.01) (Fig. 7B), while the average number of contractions per cystometrogram was decreased (8.9 ± 3.2 compared with 23.2 ± 4.4 for P2X₂^+/+ mice, P < 0.05) (Fig. 7C). To directly compare urodynamic function in mice lacking P2X₂ and/or P2X₃ receptor subunits, micturition contractions were similarly measured in P2X₃^–/– and in P2X₂/P2X₃^Dbl–/– mice. Consistent with previous findings (Cockayne et al. 2000), an increased threshold for micturition contractions was observed for P2X₃^–/– mice (0.35 ± 0.09 ml compared with 0.12 ± 0.03 ml for P2X₃^+/+ mice, P < 0.05), while trends were observed for P2X₂/P2X₃^Dbl–/– mice (0.22 ± 0.05 ml compared with 0.14 ± 0.04 ml for P2X₂/P2X₃^Dbl+/+ mice, P= 0.164). A trend to decreases in the average number of contractions per cystometrogram was observed for P2X₃^–/– mice (20.0 ± 7.0 compared with 41.4 ± 10.7 for P2X₃^+/+ mice, P= 0.136) and P2X₂/P2X₃^Dbl–/– mice (11.7 ± 5.8 compared with 32.7 ± 8.7 for P2X₂/P2X₃^Dbl+/+ mice, P= 0.053).

View larger version (28K): [in this window] [in a new window]   Figure 7.  Urinary bladder cystometry in P2X3–/–, P2X2–/– and P2X2/P2X3Dbl–/– mice A, representative acute cystometrograms recorded from anaesthetized and transurethrally catheterized 5- to 6-month-old female P2X2+/+ and P2X2–/– mice. Traces illustrate bladder pressure recorded in response to a constant intravesical infusion of saline (10 µl min–1 for 40 min). Contractions greater than 20 cm H2O were taken as micturition contractions. B and C, quantification of the threshold for contraction (B) and the average number of contractions per cystometrogram (C) for 5- to 6-month-old female P2X3–/– (n= 7), P2X2–/– (n= 10) and P2X2/P2X3Dbl–/– (n= 7) mice tested in parallel, under similar conditions, with their respective wild-type controls (P2X3+/+ and P2X2/P2X3Dbl+/+n= 7, P2X2+/+n= 11). Data represent the means ±S.E.M.*P < 0.05, **P < 0.01; Wilcoxon rank-sum exact test.

To ensure that the differences in urinary bladder reflex contractions observed in P2X-deficient mice (Fig. 7) were not caused by changes in the properties of the detrusor smooth muscle, we measured in vitro smooth muscle contractile responses of the urinary bladder. EFS of the urinary bladder induced parasympathetic nerve-mediated responses that were frequency dependent (Fig. 8A–D, control response) and abolished by 1 µM TTX (data not shown) in P2X₂, P2X₃ and P2X₂/P2X₃ wild-type and mutant mice (Fig. 8, and data not shown). Figure 8 illustrates representative data from P2X₂/P2X₃^Dbl+/+ (Fig. 8A, male; C, female) and P2X₂/P2X₃^Dbl–/– (Fig. 8B, male; D, female) mice, where loss of P2X₂ and P2X₃ subunits had no effect on bladder smooth muscle responses to nerve-mediated stimulation (control responses). PPADS (30 µM) caused a significant inhibition of the response to EFS stimulation in all P2X wild-type and P2X-deficient mice (Fig. 8, and data not shown, P < 0.001). Interestingly, the PPADS-sensitive component was greater for female compared with male mice, and this was consistent for all groups of wild-type and null mutant mice (Fig. 8, and data not shown). The subsequent addition of atropine (1 µM) caused a further, almost complete, inhibition of responses in each group (Fig. 8 and data not shown, P < 0.001). Similarly, in the presence of atropine alone (1 µM, data not shown), there was significant inhibition of nerve-mediated responses for each group of mice, both male and female, and residual responses were almost completely inhibited by the further addition of PPADS (30 µM, data not shown). No significant differences were observed in the inhibition by PPADS (30 µM) or atropine (1 µM) between the P2X wild-type and P2X-deficient mice of either sex.

View larger version (27K): [in this window] [in a new window]   Figure 8.  Frequency–response curves of the purinergic and cholinergic components of nerve-mediated responses in bladders of P2X2/P2X3Dbl–/– mice Representative frequency–response curves in the absence (control) and presence of PPADS (30 µM), and finally in the presence of both PPADS (30 µM) and atropine (1 µM) for (A) male P2X2/P2X3Dbl+/+ mice (n= 5), (B) male P2X2/P2X3Dbl–/– mice (n= 6), (C) female P2X2/P2X3Dbl+/+ mice (n= 5), and (D) female P2X2/P2X3Dbl–/– mice (n= 5). Electrical field stimulation (EFS) was carried out at 100 V, 0.3 ms, 0.5–32 Hz, with 15 s stimulation. Data represent the means ±S.E.M. of the percentage of the maximum control response. ***P < 0.001; two-way analysis of variance (ANOVA).

Activity of pelvic afferent fibres evoked by distension of the bladders of P2X₃^–/–, P2X₂^–/– and P2X₂/P2X₃^Dbl–/– mice

To further investigate the sensory role of P2X receptor subtypes within the urinary bladder, we measured the sensitivity of bladder afferents to distension in P2X-deficient mice in a urinary bladder/pelvic nerve preparation (Vlaskovska et al. 2001). The representative traces shown in Fig. 9 illustrate the delayed afferent responsiveness to bladder distension in P2X₃^–/–, P2X₂^–/–, and P2X₂/P2X₃^Dbl–/– mice, compared to P2X wild-type controls. Volume and pressure thresholds for afferent activation were calculated as shown in the top panel of Fig. 10. The volume thresholds for whole-nerve activation were significantly elevated in P2X₂^–/– mice (0.32 ± 0.05 ml compared with 0.17 ± 0.03 ml for P2X₂^+/+ mice, P < 0.05). Similarly, the volume thresholds were increased in P2X₃^–/– mice (0.38 ± 0.06 ml compared with 0.14 ± 0.02 ml for P2X₃^+/+ mice, P < 0.01), and P2X₂/P2X₃^Dbl–/– mice (0.38 ± 0.06 ml compared with 0.13 ± 0.03 ml for P2X₂/P2X₃^Dbl+/+ mice, P < 0.01) (Fig. 10A). The pressure thresholds of bladder afferents in P2X-deficient mice also appeared to be elevated, but the differences were not significant (P > 0.05) (Fig. 10B).

View larger version (32K): [in this window] [in a new window]   Figure 9.  Pelvic nerve response to bladder filling in P2X3–/–, P2X2–/– and P2X2/P2X3Dbl–/– mice Representative traces of intraluminal pressure (IP), whole-nerve activity (NA), and whole-nerve discharge rate (frequency) are shown for P2X3 (left), P2X2 (middle) and P2X2/P2X3Dbl (right) wild-type (upper panels) and mutant (lower panels) mice. The dotted lines indicate the start of bladder distension at a constant rate of 0.1 ml min–1. Note the marked delay in bladder afferent activation in P2X-deficient mice.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
The present study supports several important conclusions about the role of P2X₂ and P2X₃ subunits in mediating certain sensory functions of ATP. First, electrophysiological studies indicate that P2X₂ and P2X₃ subunits account for virtually all ATP-mediated responses in mouse sensory and sympathetic ganglion neurones. Second, our data support a role for P2X₂ subunits in mediating pain-related behaviour. Lastly, the consistent finding of bladder hyporeflexia and reduced activity of pelvic afferent fibres in response to bladder filling, in mice lacking P2X₂ and/or P2X₃ receptor subunits, supports the involvement of both P2X₃ and P2X_2/3 receptors in mediating mechanosensory transduction within the urinary bladder.

Electrophysiological studies in P2X-deficient mice have helped to define the contribution of P2X₂, P2X₃ and P2X_2/3 receptors to ATP responses on sensory and autonomic ganglion neurones. In DRG sensory neurones, ATP and ,ß-meATP induce rapidly desensitizing, slowly desensitizing or biphasic currents, consistent with P2X₃ and P2X_2/3 receptors (Rae et al. 1998; Burgard et al. 1999). Previous studies in P2X₃^–/– mice confirmed these observations (Cockayne et al. 2000; Zhong et al. 2001), with 90% of DRG neurones having no response to ATP or ,ß-meATP, and only a small percentage (12%) responding only to ATP with slowly desensitizing currents. In contrast, nodose ganglion neurones contain predominantly P2X₂ and P2X_2/3 receptors (Virginio et al. 1998; Thomas et al. 1998), with virtually all ATP and ,ß-meATP responses being slowly desensitizing. In P2X₃^–/– mice slowly desensitizing responses to ,ß-meATP were lost, while sustained responses to ATP were maintained (Cockayne et al. 2000; Zhong et al. 2001). Taken together, these data supported the idea that ATP-induced currents in DRG neurones are mediated largely by P2X₃ receptors, while in nodose neurones P2X₂ and P2X_2/3 receptors appear more important. In the present study, we extended these findings with whole-cell patch-clamp recordings from P2X₂^–/– and P2X₂/P2X₃^Dbl–/– mice. Removal of the P2X₂ subunit left only a rapidly desensitizing response in both DRG and nodose ganglion neurones in response to ATP or ,ß-meATP, consistent with the loss of slowly desensitizing P2X₂ and P2X_2/3 receptors. The antagonist sensitivity of the remaining ,ß-meATP response in nodose neurones was consistent with a P2X₃ receptor: greater than 50% inhibition by 1 µM Ip₅I, and almost complete inhibition by 10 nM TNP-ATP. Finally, in P2X₂/P2X₃^Dbl–/– mice virtually all ATP currents were lost in DRG and nodose ganglion neurones, supporting the view that P2X₂ and P2X₃ subunits are the predominant functional P2X receptors expressed in mouse sensory neurones. Sensory ganglia in P2X₂/P2X₃^Dbl–/– mice were of normal size and histological appearance, with no apparent loss of neurones. The presence of a small residual sustained response to ATP in some DRG and nodose neurones from P2X₂/P2X₃^Dbl–/– mice indicates the presence of low levels of other P2X subunits or P2Y receptors in some neurones. P2Y₁ and P2Y₂ could play a minor role as these receptors are ATP-sensitive, and have been colocalized on sensory neurones with P2X₃ and TRPV1 receptors (Ruan & Burnstock, 2003; Gerevich & Illes, 2005). Because of the low amplitude of these responses, we have not investigated them further.

Our findings from sympathetic ganglion neurones indicate a predominant functional role for P2X₂-containing receptors (probably P2X₂ homomultimers). Thus, absence of the P2X₂ subunit leads to an almost complete loss of ATP responses in coeliac and SCG neurones. These data are consistent with the known pharmacology of P2X receptors in cultured autonomic ganglion neurones (Khakh et al. 1995; Zhong et al. 2000). As with sensory neurones, the presence of small ATP responses in a few sympathetic neurones suggests the presence of low levels of some other P2 receptors. Whether this expression is normally present or the result of compensatory upregulation is unclear. Studies using P2X₁-deficient mice have suggested that P2X₁ receptor subunits can contribute to heteromultimeric P2X₂ receptors in a small population of mouse SCG neurones (Calvert, 2004).

P2X₃ and P2X_2/3 receptors are gaining recognition for their importance in facilitating pain transmission (Jarvis, 2003), but the differential contribution of these receptors to specific pain-related behaviours is still poorly understood. An assessment of pain-related behaviours in P2X₂^–/– mice revealed deficits in nocifensive responses in the persistent, but not acute phase of the formalin model, no alterations in acute nocifensive responses to intraplantar injection of P2X agonists, and no attenuation of acute thermal nociceptive responses in the tail flick or plantar test. These data suggest that P2X_2/3 (and P2X₂) receptors are probably not critical to acute nociceptive responses. These data also suggest that homomultimeric P2