SEARCH

SEARCH BY CITATION

Keywords:

  • P2X7 receptor;
  • rat cortex;
  • radioligand binding;
  • occupancy

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflict of interest
  9. References
  10. Supporting Information

BACKGROUND AND PURPOSE The P2X7 receptor is implicated in inflammation and pain and is therefore a potential target for therapeutic intervention. Here, the development of a native tissue radioligand binding, localization and ex vivo occupancy assay for centrally penetrant P2X7 receptor antagonists is described.

EXPERIMENTAL APPROACH Autoradiography studies using the P2X7 antagonist radioligand [3H]-A-804598 were carried out in rat brain and spinal cord. Subsequent in vitro binding and ex vivo occupancy assays were performed using rat cortex homogenate.

KEY RESULTS P2X7 expression was shown to be widespread throughout the rat brain, and in the grey matter of the spinal cord. In binding assays in rat cortex homogenate, ∼60% specific binding was achieved at equilibrium. In kinetic binding assays, kon and koff values of 0.0021·min−1·nM−1 and 0.0070·min−1 were determined, and the Kd derived from kinetic measurements was consistent with that derived from saturation analysis. Novel P2X7 antagonists inhibited the binding of [3H]-A-804598 to rat cortex P2X7 receptors with Ki values of <40 nM. In an ex vivo occupancy assay, a P2X7 antagonist dosed orally to rats caused a concentration-dependent inhibition of the specific binding of [3H]-A-804598 to rat cortex.

CONCLUSIONS AND IMPLICATIONS The present study describes the development of an assay that allows localization of P2X7 receptors, the measurement of the binding affinity of P2X7 receptor antagonists in native tissue, and provides a means of determining central P2X7 receptor occupancy. These assays could form an important part of a P2X7 drug discovery programme.


Abbreviations
BSA

bovine serum albumin

BZ-ATP

2′3′-O-(4-benzoylbenzoyl)adenosine 5-triphosphate

DMSO

dimethyl sulphoxide

FLIPR

fluorometric imaging plate reader

L4–L6

lumbar vertebrae in the rat spinal cord

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflict of interest
  9. References
  10. Supporting Information

The P2X7 receptor is a member of the P2X family of ATP-gated ion channels. It differs from other P2X family members in that it is functional only in homomeric forms, and it is activated only by high concentrations of ATP (North, 2002). Brief exposure of the receptor to agonists results in elevated intracellular calcium due to the opening of cation channels, whereas prolonged exposure results in the formation of a non-selective pore, allowing the influx of larger cations as well as large hydrophilic molecules (Ferrari et al., 1996; Rassendren et al., 1997). The P2X7 receptor was cloned initially from rat (Surprenant et al., 1996), and the human and mouse receptors were subsequently cloned and shown to have around 80% homology with the rat receptor (Rassendren et al., 1997; Chessell et al., 1998).

Northern blotting studies showed a widespread tissue distribution of P2X7 mRNA throughout the rat immune system and also in other organs such as brain, spinal cord, skeletal muscle, lung and placenta (Collo et al., 1997; Rassendren et al., 1997). In the central nervous system (CNS), expression of the receptor on astrocytes, microglial cells and neurones has been demonstrated using immunohistochemical methods (Deuchars et al., 2001; Kukley et al., 2001; Xiang and Burnstock, 2005). However, the specificity of some commercially available P2X7 antibodies has been called into question by researchers who demonstrated staining in tissues from P2X7 knock-out mice (Sim et al., 2004; Anderson and Nedergaard, 2006). This field has now been complicated further by the discovery that knock-out animals may still express truncated splice variants of P2X7 due to incomplete inactivation of the gene (Nicke et al., 2009). This confusion highlights the need to develop specific, selective and high affinity radioligands to demonstrate a clearer representation of receptor protein localization.

Activation of P2X7 receptors on cells of the immune system triggers the release of a number of cytokines and inflammatory mediators (Ferrari et al., 1997; Sperlágh et al., 1998; Bulanova et al., 2005). In the nervous system, P2X7 receptors have been shown to play a role in the modulation of neurotransmitter release (Deuchars et al., 2001; Sperlágh et al., 2002), and in microglial and astroglial activation (Bianco et al., 2006; Suadicani et al., 2006). Hence there is therapeutic rationale for the use of P2X7 antagonists in the treatment of a variety of disease states, from inflammatory diseases such as rheumatoid arthritis, to the amelioration of neuropathic pain.

The synthesis of a radiolabelled P2X7 receptor antagonist, [3H]-A-804598, was recently reported in the literature, and the binding characteristics of this ligand at rat recombinant P2X7 receptors were described (Donnelly-Roberts et al., 2009). The aim of the current study was to use this ligand to investigate P2X7 receptor localization in the rat brain and spinal cord, and to develop a rat native tissue P2X7 homogenate binding assay. In addition, the suitability of the homogenate binding assay methodology to measure ex vivo occupancy of dosed P2X7 receptor antagonists was investigated, in order to provide a means of demonstrating target engagement of centrally penetrant P2X7 receptor antagonists.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflict of interest
  9. References
  10. Supporting Information

Animals and tissue collection

Male Sprague-Dawley rats (200–300 g), purchased from Charles River Laboratories (UK) were kept in a fixed 12 h light–dark cycle with food and water provided ad libitum. Rats were culled by decapitation and tissues were removed and frozen on isopentane cooled to −40°C on dry ice. All experiments were conducted in compliance with the Home Office Guidance on the operation of the Animals (Scientific Procedures) Act (1986) under the authority granted in personal and project licenses, and were reviewed and approved by the ethical review panel.

Fluorometric imaging plate reader (FLIPR) assay

A FLIPR assay was used to confirm the functional potency of A-804598 and two other novel P2X7 antagonists; Compound A (Nelson et al., 2006) and A-740003 (Honore et al., 2006). rP2X7-1321N1 cells (cells expressing the rat recombinant P2X7 receptor) were maintained in a humidified 37°C incubator under 5% CO2. The culture media was MEM Alpha containing 10% dialysed foetal bovine serum, 100 U·mL−1 penicillin G sodium, 100 µg·mL−1 streptomycin sulphate and 350 µg·mL−1 geneticin. The cells were plated at 2 × 105 cells·mL−1 into 384-well poly-D-lysine coated black/clear bottom plates. The plates were incubated overnight in a humidified 37°C incubator under 5% CO2.

The FLIPR Calcium 4 assay kit (Molecular Devices, Wokingham, UK) was used for these experiments. The dye supplied with the kit was resuspended in assay buffer (20 mM HEPES buffered Hank's balanced salt solution at pH 7.4). The cell plate was quickly inverted over a sink to remove the culture media and it was replaced with 25 µL per well of the buffered dye. The cells were then placed back in the incubator for 60 min and the plate then transferred to a FLIPR Tetra. The cells were pre-incubated with P2X7 antagonists (25 µL) for 10 min, and BZ-ATP (100 µM; an approximate EC80 concentration) was then added and the calcium influx recorded as relative fluorescence units.

In vitro autoradiography

Cryostat sections (20 µm) were cut coronally through a rat brain and the L4–L6 region of rat spinal cord, and thaw-mounted onto poly-L-lysine coated slides. Slides were stored at −80°C until use. Sections were incubated for 2 h at 4°C with 5 nM [3H]-A-804598 in assay buffer (50 mM Trizma/0.1% BSA, pH 7.4 at 4°C) either in the absence (total binding) or presence (non-specific binding) of 10 µM A-740003. Sections were rinsed (2 × 5 min) in ice-cold 50 mM Trizma, followed by a brief dip in distilled water, and dried under a stream of cold air. Slides were imaged using a Beta-imager (BioSpace, Paris, France) for 12 h under high resolution.

In vitro homogenate binding

[3H]-A-804598 binding was performed in assay buffer (50 mM Trizma/0.1% BSA, pH 7.4 at 4°C) at 4°C for 2 h in a total volume of 250 µL, using 300 µg of cortex homogenate per well. Non-specific binding was defined using 10 µM A-740003. Concentration-dependent (saturation) binding was performed using 12 concentrations of [3H]-A-804598 (0.5–50 nM). For association experiments, cortex homogenate was incubated with 5 nM [3H]-A-804598 for intervals from 2 min to 5 h in the presence and absence of 10 µM A-740003. For dissociation experiments, cortex homogenate was incubated with 5 nM [3H]-A-804598 for 2 h to allow association of [3H]-A-804598. After this time an excess (10 µM) of A-740003 was added to each well at various times, and dissociation followed from 2 min to 4 h. Competition binding experiments were performed using 5 nM [3H]-A-804598, an approximate Kd concentration, in the presence and absence of P2X7 antagonists. IC50 values were determined using 10 half-log serial dilutions in duplicate.

Binding assays were terminated by separation of bound and free radioligand by filtration through GF/B filters pre-soaked in 0.5% polyethyleneimine using a Brandel harvester. Filter circles were punched out using a Brandel punch-out machine, and counted on a Tricarb liquid scintillation counter (Perkin Elmer, Seer Green, UK).

Ex vivo receptor occupancy studies

Rats (four per treatment group) were orally dosed with Compound A at 3, 10, 30 and 100 mg·kg−1 or vehicle (0.5% methylcellulose) and subsequently killed by decapitation at 1 h post dose. Trunk blood samples were collected immediately after decapitation and the plasma was separated by centrifugation for determination of drug concentration. The cortex was removed and frozen immediately in isopentane (cooled to −40°C with dry ice) and stored at −80°C until use.

For the ex vivo homogenate binding assay, cortex samples were weighed and homogenized in 12 mL assay buffer·g−1 weight tissue (giving a protein concentration of approximately 1.75 mg per well). The binding assay was performed as detailed in the in vitro method, except that the assay incubation time was 15 min and the radioligand concentration was 10 nM. A small volume of each sample homogenate was frozen on dry ice and retained for determination of drug concentration.

Measurement of drug concentrations in plasma/homogenate

Control blood and brains were obtained from rats killed by administration of CO2. Plasma was separated from blood by centrifugation. Brain was homogenized in 50 mM Trizma buffer/0.1% BSA at pH 7.4, 4°C, using a dilution of 1 g tissue in 12 mL buffer.

Calibration lines were constructed in 2 mL 96-well polypropylene plates. Calibration lines were made using 20 µL matrix (plasma or brain homogenate) spiked with compound ranging 5–20 000 ng·mL−1. Study samples were incorporated and extracted via liquid–liquid extraction (Brignol et al., 2001) using pH 10 borate buffer and tert-butyl-methyl-ether. For samples that exceeded the calibration range, further dilutions in control matrix were carried out for the analysis. Extracted plates were dried under nitrogen. Samples in the extracted plates were reconstituted in 800 µL MF5 (Romil solvent; 90% water, containing 2 mM ammonium acetate and 0.027% formic acid, and 10% methanol); 10 µL was injected via a CTC-PAL autosampler on to an ultra high performance liquid chromatography-tandem mass spectrometry system. Using an Agilent 1200RR pump, a 2.1 min ballistic gradient (from 10–90% methanol/water containing 2 mM ammonium acetate and 0.027% formic acid at a flow of 1 mL·min−1) was directed through a Zorbax Eclipse XDB C18 column (4.6 × 50 mm, 1.8 µm). Compound levels were detected by positive ion multiple reaction monitoring using a Turbo Ion Spray Sciex API 4000 mass spectrometer (AB SCIEX, Warrington, UK).

The free fraction (Fu) of Compound A (molecular weight 339.71) was measured via protein binding experiments. An automated equilibrium dialysis method was carried out utilizing the rapid equilibrium dialysis device (Thermo Scientific, Loughborough, UK). The prepared rapid equilibrium dialysis device was incubated on an orbital shaker for 4 h at 37°C within a humidified CO2 chamber. Plasma and brain homogenate binding were measured individually using dialysed matrix/buffer calibration lines to quantify concentrations using ultra high performance liquid chromatography-tandem mass spectrometry. These quantified values were used for calculating protein binding. The free fraction (Fu) of Compound A in plasma and brain was calculated to be 0.425 and 0.29 respectively.

Data analysis

FLIPR assay.  Data are reported as mean ± SEM of n experiments. Data were analysed using the non-linear regression functions within PRISM (GraphPad Software, San Diego, CA, USA).

Autoradiography.  Digital images were analysed using BetaVision image analysis software (BioSpace) and binding density in a defined brain region (according to the rat brain atlas by Paxinos and Watson, 1998) was measured as digital light units per mm2.

Homogenate binding.  Data are reported as mean ± SEM of n experiments. Data were analysed using the non-linear regression functions within PRISM. The equilibrium dissociation constant (Kd) and the binding site density (Bmax) were derived from radioligand saturation curves using the Langmuir equation RL = RtL/(Kd+ L) where L is the free ligand concentration, RL is the concentration of receptor-bound ligand at equilibrium and Rt is the total receptor concentration. IC50 (the concentration of compound producing 50% inhibition of specific binding) and Hill coefficients were derived from fitting to a four-parameter logistic equation. Apparent Ki values were derived using the Cheng and Prusoff equation; Ki= IC50/(1 +[L]/Kd) where [L] is the free concentration of radioligand and Kd is the appropriate equilibrium dissociation constant derived from the saturation analysis (Cheng and Prusoff, 1973). Apparent Ki values were converted to apparent pKi values for illustration purposes. In kinetic experiments, the observed association rate constant (kobs) and the association and dissociation half-lives were derived directly from the one-phase non-linear exponential models within PRISM. The dissociation rate constant (koff in units of·time−1) was calculated as 0.693/t1/2 off, The association rate constant (kon in units of·time−1·nM−1) was derived by the pseudo first-order method, which takes into account the concentration of radioligand used, the observed association rate constant (kobs in units of·time−1) and the dissociation rate constant (koff) such that (kon=kobskoff)/[L]. The ratio of koff/kon was also used to estimate Kd.

For ex vivo receptor occupancy studies, the specific binding in each compound-treated sample was calculated as a percentage of the mean vehicle specific binding according to the following equation: (compound-treated specific binding/mean vehicle specific binding) × 100. The resulting value was then subtracted from 100% to give the percentage occupancy in the compound-treated sample.

Materials

[3H]-A-804598 (2-cyano-1-[(1S)-1-phenylethyl]-3-quinolin-5-ylguanidine) was custom synthesized by GE Healthcare (UK). An extra tritium atom was incorporated to increase the specific activity from 8 Ci·mmol−1 (Donnelly-Roberts et al., 2009) to 53 Ci·mmol−1. A-804598, Compound A (3-[1-(2-chloro-3-trifluoromethylphenyl)-1H-tetrazol-5-ylmethyl] pyridine hydrochloride; Nelson et al., 2006) and A-740003 (N-(1-{[(cyanoimino)(5-quinolinylamino) methyl]amino}-2,2-dimethylpropyl)-2-(3,4-dimethoxyphenyl)acetamide; Honore et al., 2006) were synthesized at Pfizer, Sandwich, UK (chemical structures shown in Figure 1). Cell culture reagents were purchased from Invitrogen. The FLIPR Calcium 4 kit was purchased from Molecular Devices. All other supplements and chemicals were purchased from Sigma (UK).

image

Figure 1. Chemical structure of A-804598 (A), Compound A (B) and A-740003 (C).

Download figure to PowerPoint

For FLIPR assays, P2X7 antagonists were dissolved and serial dilutions prepared in 100% dimethyl sulphoxide (DMSO). The DMSO stocks were then diluted into assay buffer such that the final DMSO concentration did not exceed 1%. The equivalent concentration of DMSO was present in control wells. BZ-ATP was dissolved and diluted in assay buffer.

For in vitro homogenate binding, P2X7 antagonists were dissolved at 10 mM in 100% DMSO. Dilutions were made in assay buffer containing DMSO such that the % DMSO was maintained at 10% (1% in the assay). The equivalent concentration of DMSO was present in total binding wells. BZ-ATP was dissolved and diluted in assay buffer.

The drug/molecular target nomenclature in this paper conforms to the British Journal of Pharmacology Guide to Receptors and Channels (Alexander et al., 2009).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflict of interest
  9. References
  10. Supporting Information

Functional assay

A-804598, Compound A and A-740003 produced a concentration-dependent inhibition of BZ-ATP-stimulated calcium influx with IC50 values (±SEM) of 28.71 ± 5.32 nM, 89.71 ± 11.39 nM and 72.27 ± 11.74 nM respectively (n= 3, Figure 2).

image

Figure 2. Effect of P2X7 antagonists on BZ-ATP (100 µM)-stimulated calcium influx in cells expressing the rat recombinant P2X7 receptor. The data are expressed as mean ± SEM (vertical lines) of three individual experiments.

Download figure to PowerPoint

In vitro autoradiography

Autoradiographic analysis of coronal rat brain sections (Figure 3, Table 1) revealed that there was specific binding of [3H]-A-804598 throughout the rat brain. The relative intensity of binding signal was colliculus >> hypothalamic nuclei (Figure 3D) = olfactory nucleus (Figure 3A) > thalamic nuclei (Figure 3D) = piriform cortex (Figure 3B) > hippocampus (Figure 3D,E) = spinal trigeminal nucleus and tract (Figure 3G) > pontine nuclei (Figure 3F) = amygdala (Figure 3D) = other cortical regions (Figure 3A–G) > cerebellum (Figure 3G) > caudate putamen (Figure 3C). High levels of [3H]-A-804598 specific binding were also found in the grey matter of the L4–L6 region of the rat spinal cord.

image

Figure 3. Localization of P2X7 receptors in coronal sections of rat brain (A–G) and spinal cord (L4–L6 region; H). For each figure, total binding sections are shown on the left, and non-specific binding sections (defined with 10 µM A-740003) are shown on the right. Amg, amygdala; Cer, cerebellum; Coll, colliculus (superior and inferior); CP, caudate putamen; CTX, cortex; Hip, hippocampus; Ht, hypothalamus; ON, olfactory nucleus; PAG, periaqueductal grey; Pir CTX, piriform cortex; PN, pontine nuclei; Th, thalamic nuclei, STN/STT, spinal trigeminal nucleus and spinal trigeminal tract.

Download figure to PowerPoint

Table 1.  Quantification of [3H]-A-804598 binding signal in rat brain and spinal cord regions
Brain regionSpecific binding (cpm·mm-2)
Colliculus6.96
Grey matter spinal cord4.03
Hypothalamic nuclei4.02
Olfactory nucleus3.95
Thalamic nuclei3.69
Piriform cortex3.54
Spinal trigeminal nucleus and tract3.29
Hippocampus3.23
Pontine nuclei2.96
Other cortical regions2.93
Amygdala2.83
Cerebellum2.21
Caudate putamen1.67

In vitro homogenate binding

Specific binding of 5 nM [3H]-A-804598 represented approximately 60% of total binding to rat cortex homogenate (data not shown). Specific binding of [3H]-A-804598 was time-dependent, reaching a plateau after approximately 2 h, and fully reversible after the addition of an excess of A-740003 (10 µM) (Figure 4). Saturation binding confirmed that [3H]-A-804598 binds with high affinity to rat cortex homogenate (Figure 5). Statistical analysis confirmed that a 1-site model was the most appropriate fit for the data obtained. Table 2 shows a summary of the binding properties of [3H]-A-804598 to the rat cortex P2X7 receptor. A two sample equal variance t-test confirmed that there was no significant difference between the Kd value estimated from saturation analysis and the Kd obtained from the ratios of koff/kon

image

Figure 4. Binding kinetics of 5 nM [3H]-A-804598 to rat cortex homogenate (300 µg protein per well). Association time-course (A) and dissociation time-course initiated by 10 µM A-740003 (B). The data are expressed as mean ± SEM (vertical lines) of three individual experiments.

Download figure to PowerPoint

image

Figure 5. Saturation binding of [3H]-A-804598 (0.5–50 nM) to rat cortex homogenate (300 µg protein per well). The data are expressed as mean ± SEM (vertical lines) of three individual experiments.

Download figure to PowerPoint

Table 2.  Summary of binding parameters of [3H]-A-804598 at the rat cortex P2X7 receptor
ParameterMeasured value
  1. Data shown are mean ± SEM (n= 3).

t1/2 (on) (min)45.12 ± 6.81
t1/2 (off) (min)88.24 ± 10.85
koff (min−1)0.0070 ± 0.0007
kon (min−1·nM−1)0.0021 ± 0.0003
koff/kon (nM)3.38 ± 0.53
Kd (nM) from saturation plot3.08 ± 0.56
Bmax (fmol·mg−1)112.90 ± 5.57

Competition experiments (Figure 6, Table 3) showed that novel P2X7 antagonists and the unlabelled form of the radioligand caused full inhibition of the specific binding of [3H]-A-804598 to the rat cortex P2X7 receptor. Although Hill slopes did differ somewhat from unity, apparent Ki values were calculated to control for differences in radioactivity concentration between experiments. A-804598 was the most potent compound tested, with a Ki value of 7.51 nM. A-740003 and Compound A had Ki values of 26.51 and 37.53 nM respectively. In contrast, within the concentration range tested, BZ-ATP did not reach full inhibition and was significantly less potent, with a Ki value of 11.30 µM.

image

Figure 6. Competition binding of P2X7 antagonists against 5 nM [3H]-A-804598 binding to rat cortex homogenate (300 µg protein per well). The data are expressed as the mean ± SEM (vertical lines) of three individual experiments.

Download figure to PowerPoint

Table 3.  Apparent pKi values for P2X7 antagonists in rat cortex binding assay
AntagonistpKi
  1. Data shown are means ± SEM; n= 3.

A-8045988.12 ± 0.07 (Hill slope 0.85 ± 0.09)
A-7400037.57 ± 0.04 (Hill slope 1.49 ± 0.19)
Compound A7.43 ± 0.03 (Hill slope 1.28 ± 0.13)
BZ-ATP4.95 ± 0.15 (Hill slope 0.84 ± 0.13)

Ex vivo receptor occupancy

A series of optimization experiments were carried out in order to adapt the binding assay for use in ex vivo occupancy determination. In order to minimize the potential for drug dissociation from treated tissues, the radioligand incubation period was shortened from 2 h to 15 min. To retain an adequate specific binding signal using the reduced incubation time, it was necessary to increase the radioligand concentration to 10 nM, and tissue protein concentration to approximately 1.75 mg per well. Under these conditions, specific binding in vehicle-treated cortex samples was maintained at 50–60%. Compound A showed dose–dependent occupancy in the rat cortex (Figure 7) with an ED50 value of 39.31 mg·kg−1. Corresponding plasma and cortex exposure EC50 levels were 312.34 ng·mL−1 and 472.46 ng·g−1 respectively.

image

Figure 7. Ex vivo occupancy of Compound A at the P2X7 receptor in rat cortex. Data are expressed as mean ± SEM (vertical lines) of four rats per treatment group.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflict of interest
  9. References
  10. Supporting Information

The present study describes for the first time the autoradiographic localization of P2X7 receptors in the rat brain and spinal cord, and characterization of the binding properties of the P2X7 antagonist radioligand [3H]-A-804598 in rat native tissue. Additionally, the homogenate binding methodology was successfully adapted to provide a robust ex vivo occupancy assay to measure target engagement of centrally penetrant P2X7 antagonists.

Before conducting radioligand binding and autoradiography experiments with [3H]-A-804598, the functional potency of the unlabelled form of the radioligand, as well as that of two other novel P2X7 antagonists, was assessed using a FLIPR assay in cells expressing the rat recombinant P2X7 receptor. All three compounds potently inhibited BzATP-stimulated calcium influx, with IC50 values of less than 100 nM. However, A-804598 was the most potent compound tested with an IC50 of 28.7 nM. This is consistent with data reported by Donnelly-Roberts et al. (2009), who used similar methodology and generated an IC50 value of 9.9 nM for this compound.

Autoradiography experiments showed that binding of [3H]-A-804598 was widespread throughout the rat CNS. In the literature, P2X7 has been reported to be expressed on microglia and astrocytes (Deuchars et al., 2001; Kukley et al., 2001; Xiang and Burnstock, 2005). Conflicting data exist in the literature as to the expression of the P2X7 receptor on neurones. Collo et al. (1997) used the techniques of in situ hybridization and immunohistochemistry to demonstrate the expression of P2X7 receptors on rat microglia but not neurones. Similarly, Sim et al. (2004) used immunohistochemistry, immunoblot and immunoprecipitation and found no evidence for the expression of the P2X7 receptor on mouse hippocampal neurones. In contrast, multiple research groups using a variety of techniques (in situ hybridization, immunohistochemistry, immunocytochemistry, RT-PCR) have reported data showing that P2X7 receptors are expressed on neurones (Deuchars et al., 2001; Armstrong et al., 2002; Atkinson et al., 2002; 2004; Lundy et al., 2002; Sperlágh et al., 2002; Ishii et al., 2003; Wang et al., 2004; Díaz-Hernandez et al., 2008; Marcoli et al., 2008).

Our autoradiography data could represent expression of the receptor on microglia, astrocytes or neurones. All cell types would be expected to be present throughout the CNS, and autoradiography does not offer sufficient resolution to determine the cellular localization of the binding of [3H]-A-804598. However, it was evident that there were discrete regions where moderate-high binding intensity was observed. These include the colliculus, cerebral cortex; in particular the piriform cortex; hippocampus, hypothalamic and thalamic nuclei, amygdala, pontine nuclei, spinal trigeminal nucleus and tract, and the grey matter of the spinal cord. This is in agreement with results obtained by Yu et al. (2008); they recently published a study describing the use of in situ hybridization to demonstrate that P2X7 receptor mRNA was widely distributed throughout the rat brain. In the Yu study, the piriform cortex, hippocampus, pontine nuclei and the anterior horn of the spinal cord were among the areas of high P2X7 mRNA expression. Marcoli et al. (2008) demonstrated expression of P2X7 receptors on the presynaptic glutaminergic nerve terminals of rat cortical neurones. Calcium entry through these receptors was shown to evoke glutamate release, suggesting a role for the P2X7 receptor in cortical glutaminergic transmission. Wang et al. (2004) demonstrated the expression of P2X7 receptors on spinal cord neurones and found that inhibition of the receptor ameliorated acute spinal cord injury. Díaz-Hernandez et al. (2008) identified functional P2X7 receptors on the growth cones of hippocampal neurones in culture, where they appeared to play a role in neuronal growth and branching. Our data provide evidence for moderate-high expression of P2X7 in the cortex, hippocampus and spinal cord, and therefore add further support to the above findings.

The equilibrium and kinetic binding properties of [3H]-A-804598 at the rat native P2X7 receptor were determined. The association half-time and Kd were consistent with those reported by Donnelly-Roberts et al. (2009) who investigated the binding properties of this radioligand at the rat recombinant P2X7 receptor. One discrepancy was in the dissociation half-time of the radioligand. Donnelly-Roberts et al. reported that the radioligand had fully dissociated from the rat recombinant P2X7 within 30 min, whereas in the present study, a dissociation half-time of 88 min was calculated. However, in the Donnelly-Roberts paper, dissociation was reported as having been conducted at room temperature as opposed to 4°C, which could account for the quicker off-rate of the radioligand.

The novel P2X7 antagonists caused full and potent inhibition of the specific binding of [3H]-A-804598 to rat cortex. The standard P2X7 receptor agonist BZ-ATP had µM potency in the binding assay. This is consistent with its potency in our FLIPR assay (EC50 60 µM) as well as its reported P2X7 receptor potency in diverse assay systems; for example, patch clamping experiments in rat recombinant P2X7 cells (Hibell et al., 2000) and ethidium bromide uptake in rat peritoneal cells (Chen et al., 2005).

In the ex vivo occupancy assay, Compound A showed dose–dependent occupancy at the cortex P2X7 receptor, reaching maximal occupancy at 100 mg·kg−1. It is anticipated that the current ex vivo occupancy data may provide confidence for demonstrating target engagement using positron emission tomography imaging in clinical studies.

In summary, a P2X7 receptor radioligand binding assay in rat native tissue has been reported for the first time. Specific binding of [3H]-A-804598 was shown to be of high affinity, saturable and reversible in rat cortex homogenate. Novel P2X7 antagonists inhibited specific binding of the radioligand to rat cortex with affinities consistent with their functional potency at the rat recombinant P2X7 receptor. The radioligand can also be used to measure ex vivo occupancy of P2X7 antagonists in rat cortex, which could provide an important biomarker in pre-clinical studies. Given the therapeutic potential for compounds that bind to the P2X7 receptor, a native tissue P2X7 receptor binding and occupancy assay could form an important component of a drug discovery programme.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflict of interest
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. Conflict of interest
  9. References
  10. Supporting Information

Supporting Information: Teaching Materials; Figs 1–7 as PowerPoint slide.

FilenameFormatSizeDescription
BPH_1025_sm_Figs1-7.pptx452KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.