SEARCH

SEARCH BY CITATION

Keywords:

  • desensitization;
  • intracellular tyrosine;
  • purinergic receptors;
  • receptor modulation;
  • receptor structure;
  • recovery

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

J. Neurochem. (2012) 122, 557–567.

Abstract

ATP-activated P2X3 receptors of sensory ganglion neurons contribute to pain transduction and are involved in chronic pain signaling. Although highly homologous (97%) in rat and human species, it is unclear whether P2X3 receptors have identical function. Studying human and rat P2X3 receptors expressed in patch-clamped human embryonic kidney (HEK) cells, we investigated the role of non-conserved tyrosine residues in the C-terminal domain (rat tyrosine-393 and human tyrosine-376) as key determinants of receptor function. In comparison with rat P2X3 receptors, human P2X3 receptors were more expressed and produced larger responses with slower desensitization and faster recovery. In general, desensitization was closely related to peak current amplitude for rat and human receptors. Downsizing human receptor expression to the same level of the rat one still yielded larger responses retaining slower desensitization and faster recovery. Mutating phenylalanine-376 into tyrosine in the rat receptor did not change current amplitude; yet, it retarded desensitization onset, demonstrating how this residue was important to functionally link these two receptor states. Conversely, removing tyrosine from position 376 strongly down-regulated human receptor function. The different topology of tyrosine residues in the C-terminal domain has contrasting functional consequences and is sufficient to account for species-specific properties of this pain-transducing channel.

Abbreviations used
α,β-meATP

α,β-methylene ATP

Csk

C-terminal Src inhibitory kinase

HEK

human embryonic kidney cells

IP

immunoprecipitation

siRNA

small interfering RNA

WB

western blotting

WT

wild type

In the field of physiological and pathophysiological pain transduction, ATP-gated neuronal P2X3 receptors of trigeminal and dorsal root ganglia play a major role to transduce nociceptive stimuli to the brain and spinal cord (Khakh and North 2006; Wirkner et al. 2007; Burnstock 2009). Functional P2X3 receptor complexes at membrane level are formed as trimers, whereby each subunit is composed of a large extracellular loop, two transmembrane domains, and intracellular N- and C-termini (Jarvis and Khakh 2009; Burnstock and Kennedy 2011; Coddou et al. 2011). The cytosolic presence of these termini makes them suitable targets for interaction with certain intracellular proteins (Mo et al. 2009; Roger et al. 2010; Lalo et al. 2011) and functional coupling with other receptor types (Toulméet al. 2007; Stanchev et al. 2009; Wen and Evans 2011).

Despite a high degree of homology between rat and human P2X3 receptors (Garcia-Guzman et al. 1997), a number of studies have indicated that, when expressed in human embryonic kidney (HEK) 293T cells, human P2X3 receptors usually generate much higher currents (Wirkner et al. 2005; Stanchev et al. 2006; Bodnar et al. 2011). As rat and human receptors share conserved agonist binding sites and major consensus sites for post-translational modifications (Garcia-Guzman et al. 1997; North and Surprenant 2000; Bodnar et al. 2011), we focused on the C-terminal region where interesting differences in amino acid residues are observed which could determine the functional response efficiency. In particular, rat P2X3 receptors are inhibited through phosphorylation of their C-terminal tyrosine-393 (Y393) by the C-terminal Src inhibitory kinase (Csk); D’Arco et al. 2009). Conversely, in human P2X3 receptors, phenylalanine rather than tyrosine is expressed at the homologous 393 position, whereas tyrosine is at the position 376 (where the rat receptor contains phenylalanine). We wondered if such structural difference might result into a change in receptor activity. The aims of this study were to compare the efficiency of rat and human P2X3 expression and function in HEK 293T cells, to study the potential role of Csk on human wild-type (WT) receptors, and to probe how site-directed mutagenesis might alter species-dependent receptor properties.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Cell culture and transfection

Experiments were performed on HEK 293T cells obtained from SISSA in-house bank. HEK cells were cultured in Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum and antibiotics; the number of cell passages did not exceed 20. Transfections were performed with the calcium phosphate precipitation method (D’Arco et al. 2009). In general, cells plated in 35 mm Petri dishes were transfected with pEGFP and pcDNA3-P2X3 plasmids (0.5 μg in 1 ml each; 1 : 1 ratio). In a few experiments aimed at reducing human P2X3 receptor expression to the rat level, cells were transfected with 0.5 μg pEGFP, 0.17 μg or 0.08 μg pcDNA3-human P2X3, and 0.33 μg or 0.42 μg of pcDNA3 empty vector. After 24 h, sister dishes were used for patch clamping and protein expression experiments.

DNA constructs

WT human and rat P2X3 constructs were kindly provided by Prof. R.A. North (University of Manchester, UK). In particular, pcDNA3-P2X3 (rat sequence, NCBI accession number: CAA62594) and pcDNA3-P2X3 (human sequence, NCBI accession number: AAH74793) were used. Mutants were obtained using the QuikChange mutagenesis kit (Stratagene, La Jolla, CA, USA) with the specific primers listed in Table 1. All mutants were checked with DNA sequencing, and correct expression was confirmed by western blotting (WB) experiments. The pEGFP (Clontech, Mountain View, CA, USA) and pcDNA3-Csk (kindly provided by Dr X.Y. Huang, Cornell University; Lowry et al. 2002) were also used. Protein sequence alignment and calculation of amino acid sequence similarity and identity were performed using ClustalW2 online software (Larkin et al. 2007).

Table 1.   Primers used for site-directed mutagenesis experiments
PrimerSequenceT (°C)
  1. The list comprises the mutagenic primers used for mutant strand synthesis reaction. The temperature values (T) of primer annealing are also indicated.

Rat P2X3 F376Y  Forward  Reverse 5′-ACCAACCCAGTGTACACCAGTGACCAG-3′ 5′-CTGGTCACTGGTGTACACTGGGTTGGT-3′ 55
Rat P2X3 Y393F  Forward  Reverse 5′-GACTCGGGGGCCTTCTCTATTGGTCAC-3′ 5′-GTGACCAATAGAGAAGGCCCCCGAGTC-3′ 55
Human P2X3 Y376F  Forward  Reverse 5′-ACCAACCCAGTGTTCCCCAGCGACCAG-3′ 5′-CTGGTCACTGGTGTACACTGGGTTGGT-3′ 55
Human P2X3 F393Y  Forward  Reverse 5′-GATTCGGGGGCCTATTCCATAGGCCAC-3′ 5′-GTGACCAATAGAGAAGGCCCCCGAGTC-3′ 55

Protein analysis, immunoprecipitation, and western blotting

For western blotting (WB), cells were lysed as previously detailed (D’Arco et al. 2009) in ODG buffer (2% n-octyl-beta-D-glucopyranoside, containing 1% Nonidet P-40, 10 mM Tris pH 7.5, 150 mM NaCl plus protease inhibitors mixture; Complete, Roche Applied Science) and immunoblotted with rabbit anti-P2X3 (1 : 1000; Alomone, Jerusalem, Israel), anti-GFP (1 : 1000; Santa Cruz, Heidelberg, Germany), or anti-beta-actin (1 : 3000; Sigma, Milan, Italy). For co-immunoprecipitation studies, proteins were extracted in ODG buffer (plus 2 mM EDTA, 100 mM NaF, 20 mM orthovanadate) and immunoprecipitated as previously described (D’Arco et al. 2009) with rabbit anti-P2X3 (Santa Cruz), pulled down with magnetic beads (Millipore, Milan, Italy) for 1 h at 4°C according to the manufacturer’s instructions and immunoblotted with anti-Csk (Santa Cruz). Controls were cells mock transfected with empty vector. Signals were detected with the enhanced chemiluminescence light system ECL (Amersham Biosciences, Piscataway, NJ, USA) and recorded by the digital imaging system Alliance 4.7 (UVITEC, Cambridge, UK). Quantification of the optical density of the bands was performed with ImageJ software plug-in (Rasband, W.S., ImageJ, US National Institutes of Health, Bethesda, MD, USA, http://imagej.nih.gov/ij/, 1997-2012). Protein expression in total lysates was normalized with respect to beta-actin signal.

Biotinylation assay of surface expressed receptors

For biotinylation experiments, 24 h after transfection, HEK 293T cells were incubated with 1 mg/ml EZ-Link Sulfo-NHS-LC-biotin (Pierce, Rockford, IL, USA) as previously described (Fabbretti et al. 2006). Pull-down of biotinylated proteins was obtained with Streptavidin agarose resin (Pierce) for 2 h at 4°C according to the manufacturer’s instructions. Samples were processed for Western immunoblot using antibodies against the P2X3 receptor (1 : 1000; Alomone Laboratories). Biotinylation experiments resulted free of intracellular protein contaminants (as shown by lack of signal for beta-actin). Positive control for biotinylation assay was obtained by checking the surface expression of the transferrin receptor detected with an antibody purchased from Santa Cruz (1 : 1000). For control of correct gel loading, the expression of beta-actin in the intracellular fraction was checked. Quantification of the optical density of the bands was performed with ImageJ software plug-in.

RNA silencing

Small interfering RNA (siRNA) experiments were performed as described (D’Arco et al. 2009) with human Csk siRNA SmartPools (100 nM; Dharmacon RNAi Technology, Lafayette, CO, USA) using the DharmaFECT™-1 transfection reagent. Twenty-four hours after silencing, cells were transfected with pEGFP and pcDNA3-P2X3 plasmids and used for patch clamping and protein expression experiments 24 h later.

Electrophysiological recordings

Whole-cell voltage clamp experiments were performed at 21–24°C 24 h after transfection. Cells were continuously superfused with control solution containing (in mM): 152 NaCl, 5 KCl, 1 MgCl2, 2 CaCl2, 10 glucose, 10 HEPES (pH 7.4 adjusted with NaOH). Patch pipettes had resistance of 3–4 MΩ when filled with (in mM): 130 CsCl, 20 HEPES, 1 MgCl2, 3 ATP2-Mg3, EGTA 5, pH 7.2 adjusted with CsOH. The agonist α,β-methylene ATP (α,β-meATP; Sigma) was applied for 2 s by rapid solution changer system (Perfusion Fast-Step System SF-77B; Warmer Instruments, Hamden, CT, USA). To avoid build-up of cumulative desensitization, agonist test applications were spaced at >4-min intervals. Currents were recorded from single GFP-positive HEK cells clamped at −60 mV using the pCLAMP software Clampex 9.2 (Molecular Devices, Palo Alto, CA, USA). After establishing the whole-cell configuration, cell slow capacitance was compensated. Access resistance was never >10 MΩ and was routinely compensated by at least 70%. Analysis of the currents was performed with pCLAMP Clampfit 9.2 in terms of current peak amplitude, current density (current peak amplitude divided by cell slow capacitance), onset of desensitization (evaluated as the fast time constant τfast of the current decay during agonist application; Sokolova et al. 2006), and recovery from desensitization (assessed by a paired pulse protocol over 30-sec intervals and expressed as %; Sokolova et al. 2006). Concentration–response curves for α,β-meATP were fitted using Origin 6.0 with the logistic function:

  • image

where EC50 is the concentration of α,β-meATP [α,β] that evoked half-maximal response, Imax is the maximal current, and nH is the Hill coefficient. The plots of current amplitude obtained at different agonist concentrations (normalized with respect to the maximal response of each cell) against τfast values for current decay in the presence of α,β-meATP were displayed on a log-log scale and fitted with linear function using Origin 6.0.

Statistical analysis

Results are expressed as the mean ± SE (unless otherwise indicated) from the number of cells or experiments indicated in the figure legends. Statistical significance was evaluated with SigmaStat 3.5 (Systat Software, San Jose, CA, USA) or Origin 6.0 using the unpaired Student’s t-test (for parametric data) or the Mann–Whitney–Wilcoxon test (for non-parametric data). p values are indicated in the figure legends (p ≤ 0.05 was considered significant).

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Expression and function of rat and human P2X3 receptors

Although the amino acid sequence alignment of rat and human P2X3 sequences displays 97% similarity and 93% identity, the C-terminal region of P2X3 receptors possesses discrete amino acid differences. Previous studies have shown that human P2X3 receptors expressed in recombinant mammalian systems like HEK cells generate very large currents with 4–9 nA maximal amplitude (Wirkner et al. 2005; Stanchev et al. 2009; Bodnar et al. 2011). To compare rat and human P2X3 receptors expressed in GFP-positive HEK cells, we analyzed responses evoked by α,β-meATP (10 μM; Sokolova et al. 2006) as exemplified in Fig. 1a. At the same holding potential (−60 mV) following transfection with similar amounts of either rat (top) or human (bottom) P2X3 cDNA, human receptor responses appeared much larger, and with better recovery after 30-sec washout. Concentration–response plots (Fig. 1b) indicated larger efficacy of human P2X3 receptors in comparison with rat receptors, without difference in potency (EC50 values = 6.3 ± 1.8 and 4.7 ± 1.0 μM for rat and human receptors, respectively; n = 8–30 cells). Human receptors had lower activation threshold, as responses from rat receptors could not be observed at concentrations below 1 μM (Fig. 1b), whereas the Hill coefficients were similar (nH = 1.2 ± 0.1 for rat, and 1.2 ± 0.1 for human P2X3 receptors).

image

Figure 1.  Comparison of rat and human P2X3 receptors transfected in human embryonic kidney (HEK) cells. (a) Whole-cell current responses evoked by 10 μM α,β-methylene ATP (α,β-meATP) in HEK cells transiently transfected with rat (top) or human (bottom) P2X3 receptors (cDNA 0.5 μg/ml). Holding potential was −60 mV. (b) α,β-meATP concentration–response curves for rat (○, dashed line) and human (•, solid line) P2X3 receptors (***< 0.001 at 100 and 10 μM; *= 0.013 at 3 μM) reveal higher efficacy for human P2X3 receptors with no significant difference in potency (EC50 values are 7.6 ± 0.7 μM for rat and 6.8 ± 1.6 μM for human P2X3), or Hill coefficients (nH = 1.2 ± 0.1 for rat and nH = 1.2 ± 0.1 for human P2X3). Mean ± SE from 8 to 30 cells. (c) Left, plots of current log amplitude obtained at different agonist concentrations (normalized with respect to maximal response of each cell) against τfast log values for current decay in the presence of α,β-meATP. Note similar relation with analogous slope (−0.77 ± 0.07 and −0.79 ± 0.07 for rat and human receptors, respectively). Right, bar graph summarizes desensitization onset (fast desensitization time constant, τfast) for rat and human P2X3 currents, evoked by 2 s of 10 μM α,β-meATP (*p = 0.02; n = 30–32 cells). (d) Recovery from desensitization at 30 s expressed as ratio of initial response. Note faster recovery of the human P2X3 response (***p < 0.001; *p < 0.05; n = 30–32 cells). (e) Left, example of western immunoblotting (representative of three experiments) showing the surface (obtained after membrane biotinylation) and total expression of rat and human P2X3 cDNA (0.5 μg/ml) in HEK cells. Beta-actin was used as control loading marker of the total extract. Right, quantification of the optical density of the bands (mean ± SD) showing threefold increase (**< 0.01; n = 3) in expression of total and surface human P2X3 protein in comparison with rat one (dashed line).

Download figure to PowerPoint

As shown in Fig. 1a, inward currents rapidly and fully waned to baseline prior to the end of α,β-meATP application (Sokolova et al. 2006), indicating receptor desensitization, the onset of which was estimated with the time constant of fast current decay (τfast). We estimated the relation between current amplitude and τfast values to find out the dependence of receptor activity versus receptor desensitization. Fig. 1c shows the plot of desensitization onset values for rat and human receptors against corresponding peak current amplitudes (normalized to the maximum response for each cell). Thus, desensitization was faster for larger responses, indicating the extent of receptor activation had a major role to shift the receptor into the desensitization process (Sokolova et al. 2006; Karoly et al. 2008). After 10 μM α,β-meATP application, the τfast value was significantly larger for the human P2X3 receptor-mediated responses (Fig. 1a and c, right). To avoid very strong, long-lasting desensitization, subsequent tests were usually performed by applying 10 μM α,β-meATP in 2-sec pulses, i.e., a concentration close to the EC50 values for rat and human receptors. Recovery from desensitization measured with a double pulse protocol (30-sec interval of 10 μM α,β-meATP pulses; Fig. 1a) showed faster return of human P2X3 receptor responses (Fig. 1d). Thus, human receptors produced stronger responses with slower desensitization and faster recovery from it.

To check if such differences in receptor-mediated currents were because of differential expression of human and rat P2X3 receptors, we performed immunoblotting of transfected HEK 293T cells. As shown in Fig. 1e, for the same cDNA quantity used for transfection, human P2X3 receptors were three times more expressed than rat receptors both in the total lysate fraction and in the surface protein fraction (identified with a biotinylation method). For this reason, we next enquired if reducing the expression of human receptors by decreasing the amount of transfected cDNA (as detailed in the methods) might confer them properties analogous to rat ones. Fig. 2a shows that, in order to obtain surface human P2X3 subunit expression comparable with that of the surface rat receptor, the amount of transfected human cDNA had to be decreased from 0.5 to 0.17 μg. However, the current amplitudes exemplified in Fig. 2b (for 10 μM α,β-meATP applications) remained larger (bar graphs in Fig. 2c). Similar results were obtained when the responses were calculated in terms of current density to take into account cell size (Fig. 2c). It is noteworthy that, even when lower surface expression was observed with 0.08 μg human cDNA (see also right-hand bar graph in Fig. 2a), the peak amplitude of P2X3 currents (and their current density) was still significantly larger than the rat ones (Fig. 2b,c). Nevertheless, no difference in the desensitization onset or recovery was observed when expressing human P2X3 with either 0.17 or 0.08 μg cDNA (Fig. 2d).

image

Figure 2.  Decreasing human P2X3 expression preserves specific receptor properties. (a) Left, comparable surface (biotinylated) and total P2X3 expression were obtained when 0.5 μg/ml rat cDNA or 0.17 μg/ml human cDNA was used for transfection. Right, data quantification with mean ± SD. Values are expressed as fraction of the optical density of the rat P2X3 band (0.5 μg/ml; dashed line). *p = 0.02; n = 3. (b) Examples of currents evoked by 10 μM α,β-methylene ATP (α,β-meATP) recorded from human embryonic kidney (HEK) cells transfected with 0.5 μg/ml rat (white box) P2X3, 0.17 μg/ml or 0.08 μg/ml human (gray boxes with cDNA concentration written inside) P2X3 cDNA. (c) Summary of current amplitude (top) and current density (bottom) values indicating stronger responses by the human receptors. ***p < 0.001, n = 35–48 cells. (d) Desensitization onset (τfast, left) and desensitization recovery (right) of human P2X3 responses (10 μM α,β-meATP; 2 s) were unaffected when the cDNA amount was reduced. ***p < 0.001 with rat wild type (WT); n = 30–42 cells.

Download figure to PowerPoint

Lack of inhibition of human P2X3 receptors by Csk

Previous experiments have shown that the kinase Csk phosphorylates the C-terminal domain tyrosine-393 of the rat receptor to depress its function (D’Arco et al. 2009). We explored whether this regulatory mechanism was deficient in human WT P2X3 (lacking tyrosine-393). Co-immunoprecipitation experiments showed that co-transfected Csk was detectable in rat and human P2X3-containing fractions, but not in negative control without P2X3 transfection (Fig. 3a). The co-precipitation of Csk and P2X3 was similar for rat and human receptors.

image

Figure 3.  Lack of effect of C-terminal Src inhibitory kinase (Csk) on human P2X3 receptor function. (a) Co-immunoprecipitation experiment (representative of three) from human embryonic kidney (HEK) cells transfected with rat or human P2X3 plus Csk shows similar level of Csk in all P2X3-bound fractions. After immunoprecipitation (IP) of P2X3 protein, western blotting (WB) was performed with the Csk antibody. Lane for control consisted of mock transfection of cells with empty vector. Total lysates were analyzed with anti-P2X3, anti-Csk, and anti-beta-actin antibodies. (b) Example of western immunoblot to confirm very low expression of constitutive Csk after Csk small interfering RNA (siRNA) in HEK cells. (c) Representative currents evoked by 10 μM α,β-methylene ATP (α,β-meATP) at−60 mV holding potential from HEK cells transfected with human P2X3 cDNA and scramble RNA (control, −) or human P2X3 cDNA plus Csk siRNA (+). (d) Bar graphs show no change in the mean current amplitude and density (n = 35–41 cells). (e) Bar graphs show no change in desensitization onset (τfast, left) and recovery from it (right) of human P2X3 currents after Csk siRNA transfection (n = 32–41 cells).

Download figure to PowerPoint

As there are no commercially available selective inhibitors of Csk, we silenced endogenous Csk expressed constitutively in HEK 293T cells with siRNA and then tested any changes in α,β-meATP-mediated responses. Fig. 3b shows that siRNA-mediated silencing of Csk effectively blocked this protein expression. Although Csk siRNA potentiates rat P2X3 receptor function (D’Arco et al. 2009), this procedure had no significant effect on the 10 μM α,β-meATP-mediated current by human WT receptors or their expression (Fig. 3c–d). Likewise, Csk siRNA had no significant effect on onset of human receptor desensitization or recovery from it (Fig. 3e).

Crucial role of C-terminus tyrosine residues in shaping P2X3 receptor activity

Comparison of the amino acid sequences of the C-terminal domains between rat and human WT receptors (Fig. 4) showed that the human receptor lacked tyrosine (filled circle) at position 393 as tyrosine was instead present at position 376. To explore the role of tyrosine-393 in the rat receptor, we generated rat Y393F, rat F376Y, and rat F376Y+Y393F P2X3 (Fig. 4): in the latter two, tyrosine was moved to the same location as normally found in the human receptor. Fig. 5a (left and right) shows that in our experimental conditions 24 h after transfection, such mutations did not change P2X3 protein expression with respect to the WT receptors. When tested with 10 μM α,β-meATP (as exemplified in Fig. 5b and c), the F376Y mutant did not produce response amplitudes different from WT ones (although characterized by slower onset of desensitization), whereas the Y393F mutant generated larger currents as previously reported (D’Arco et al. 2009) with even slower onset of desensitization (Fig. 5d). Producing the F376Y+Y393F double mutant did not further change response size or desensitization properties when compared with Y393F (Fig. 5d).

image

Figure 4.  Scheme of C-termini (indicated as -COOH) of wild-type (WT) rat P2X3 (empty bar), WT human P2X3 (filled gray bar), and mutated rat receptor constructs used for this study. Numbers indicate the position of mutated amino acids, phenylalanine (F, ○), and tyrosine (Y, •). Single and double point mutations were performed at the positions 376 and 393 to make rat P2X3 sequence similar to human one, and vice versa.

Download figure to PowerPoint

image

Figure 5.  Characteristics of F376Y, Y393F, and F376Y+Y393F rat P2X3 receptors. (a) Left, western immunoblots of human embryonic kidney (HEK) cells expressing rat wild-type (WT) or mutated P2X3 receptors. Right, bar graphs show no significant difference in protein expression levels (mean ± SD, n = 3). Loading input is assessed with beta-actin. (b) Examples of currents activated by 10 μM α,β-methylene ATP (α,β-meATP) obtained from HEK cells expressing WT and mutated F376Y, Y393F, or F376Y+Y393F P2X3 receptors. Note larger current for mutants bearing Y393F substitution. (c) Bar graphs show mean peak amplitudes (top) and their current densities (bottom) induced by α,β-methylene ATP (α,β-meATP) on rat WT and mutated receptors (***< 0.005; **p < 0.01; data from 20 to 40 cells). (d) Bar graphs quantify desensitization onset τfast (left) and recovery from desensitization (right) of rat WT and mutated P2X3 receptors. Note that all mutants display slower desensitization (***< 0.005; **< 0.01; data from 14 to 30 cells) with no change in recovery.

Download figure to PowerPoint

In analogy with the rat receptor experiments, we also investigated the impact of tyrosine mutations (Y376F, F393Y, and Y376F+F393Y; see Fig. 4) on the human P2X3 receptor. Construct mutants were expressed at comparable level with WT receptors using 0.5 μg P2X3 cDNA (Fig. 6a). Unlike the results observed with the rat receptors, on human mutated receptors α,β-meATP always evoked currents significantly smaller than the WT ones. In particular, with 10 μM α,β-meATP application, there was a range of response sizes which was the lowest with Y376F and grew with F393Y to approximately 75% of the WT one (Fig. 6b and c). For all these mutants, the onset of desensitization was slowed down (Fig. 6d, left), while recovery from desensitization was significantly depressed especially for the human P2X3 Y376F mutant (Fig. 6d, right).

image

Figure 6.  Characteristics of Y376F, F393Y, Y376F+F393Y human P2X3 receptors. (a) Left, western immunoblots of human embryonic kidney (HEK) cells transfected with human wild-type (WT) or mutated P2X3 receptors confirm similar expression level (right panel, n = 3, mean ± sd). Loading input was assessed with beta-actin. (b) Examples of currents activated by 10 μM α,β-methylene ATP (α,β-meATP), obtained from HEK cells expressing WT, F393Y, Y376F+F393Y, or Y376F receptors. (c) Summary of current amplitudes (top) and current densities (bottom) of rat WT and mutated receptor responses (***< 0.001; *p < 0.05; n = 33–71 cells). Note smaller currents for all mutants with the smallest responses for Y376F. (d) Bar graphs quantify desensitization onset τfast (left) and recovery from desensitization (right) of human WT and mutated P2X3 receptors (***< 0.001; **< 0.01; *< 0.05; n = 25–69 cells). Note increase in the τfast for all mutants and strong depression of recovery from desensitization for human P2X3 Y376F.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The principal finding of the present report is the novel demonstration of the important role of C-terminus non-conserved tyrosine residues in differentially shaping the activity of rat and human P2X3 receptors.

Rat versus human P2X3 receptor characteristics

Human rather than rat receptors were more abundantly expressed by HEK cells and provided much larger current responses. Although neither the EC50 values nor the Hill coefficients were different between rat and human receptors (implying equivalent agonist receptor potency and stoichiometry), the observed maximal response amplitudes were largely dissimilar, suggesting a stronger expression of human receptors despite similar amounts of transfected cDNA. Biotinylation experiments with different amounts of cDNA to obtain human P2X3 receptor expression similar or even lower than the rat one, however, demonstrated that human receptor responses remained consistently stronger. Hence, the inter-species difference could not be simply attributed to a different amount of membrane channels, and pointed to species-dependent intrinsic properties responsible for the channel behavior. In line with this suggestion was the observation that, despite the much larger current amplitude, human P2X3 responses had slower desensitization and a much faster recovery. These results did not conform to what is expected from a cyclic operational model of receptor activity whereby desensitization is strongly related to the degree of receptor activation (Sokolova et al. 2006) as supported by the observed dependence of τfast values on response amplitude. The power-law dependence of this relation was not unexpected in view of the multiphasic process of desensitization (Sokolova et al. 2006; Karoly et al. 2008).

Our data, therefore, alluded to the possibility of intrinsic kinetic differences between rat and human receptors. Slower recovery of rat P2X3 receptors had also been reported by Pratt et al. (2005) although response amplitude and desensitization were found to be similar. The discrepancy from the present study might be attributed to methodological differences in the gene transfection procedure and the use of very large (100 μM) concentration of ATP as receptor agonist.

Intrinsic inhibitory regulation by Csk

Previous studies have indicated that phosphorylation of tyrosine-393 of the rat receptor by the intracellular kinase Csk was responsible for constraining receptor function and was a target modulated by algogenic substances like NGF (D’Arco et al. 2007). A similar role of Csk on the human receptor was unlikely because of the lack of the target residue tyrosine at position 393 (substituted in the human receptor by the structural homologous phenylalanine). In the present experiments, co-transfected Csk could be detected in immunoprecipitated rat or human P2X3 complexes, indicating its potential association with the receptor protein. To find out if constitutive Csk might indirectly regulate human receptors, we tested the effects of Csk siRNA and observed that this procedure did not significantly alter receptor responses. The lack of effect by Csk siRNA indicated that the depression of P2X3 function by this enzyme was species specific. Thus, the absence of the crucial tyrosine-393, that was necessary in the rat receptor to express Csk-mediated functional inhibition, could probably be a contributor to the generation of much larger responses by the human receptor. This hypothesis was, therefore, explored by site-directed mutagenesis of rat and human receptors.

Changing tyrosine position confers human P2X3 properties to the rat receptor

On rat receptors, replacing tyrosine-393 with phenylalanine confirmed a strong response increase (as reported by D’Arco et al. 2009) despite equivalent protein expression level. However, when tyrosine was expressed at position 376 (where it is normally found in the human receptor), the functional response of rat P2X3 receptor remained similar. This observation indicated that, at least for the rat receptor, tyrosine could determine the response amplitude only when it was at the position 393 of the C-terminus because of its critical role for expressing the inhibitory action by Csk (D’Arco et al. 2009). When tyrosine was at position 376 of rat P2X3 receptors, desensitization was also slowed down, suggesting that perhaps this residue contributed to the kinetic shift from activation to the desensitization states associated with various degree of agonist unbinding (Sokolova et al. 2006; Karoly et al. 2008). Interestingly, these data also pointed to dissociation between the mechanisms controlling the development of desensitization and recovery from it. Previous studies have shown how the latter two kinetic processes could be differentially modulated by algesic substances like CGRP (Fabbretti et al. 2006) or NGF (D’Arco et al. 2007), or discrete mutations of the extracellular receptor domain (Fabbretti et al. 2004) and the level of extracellular Ca2+ (Cook et al. 1998). In all such examples, the desensitization process was unaffected, while recovery was accelerated, implying a more prompt re-sensitization of the receptor to the agonist application. An extensive mutagenesis study of chimeras between rapidly desensitizing P2X1 and non-desensitizing P2X2 receptors (Allsopp and Evans 2011) has suggested that agonist binding and pore-forming domains of the receptor do not play a major role in the control of the time-course of the current.

Changing human P2X3 receptor responses by introducing rat-like receptor determinants

We tested three possible changes concerning C-terminus tyrosine residues of the human receptor: 1, introduction of tyrosine to position 393; 2, replacement of tyrosine-376 together with insertion of tyrosine into position 393 (making the human receptor structurally more rat-like); 3, substitution of tyrosine-376. For comparable levels of protein expression, we observed a decrease in receptor responses which was the most evident with Y376F. These data, therefore, indicated that, on the human receptor, tyrosine-376 was actually a dominant residue versus tyrosine-393 to ensure strong receptor responsiveness. In comparison, the effect of phenylalanine-393 on human P2X3 was marginal in terms of current amplitude and desensitization properties.

Y376F human receptors displayed slower (τfast value = 71 ± 3 ms) desensitization in comparison with WT human P2X3. This prolongation was only partly accounted for by the smaller current amplitude that should be accompanied by an even slower response decay (about 200 ms as predicted by the plot in Fig. 1c). Hence, tyrosine-376 of the human receptor appeared to be one, yet not the main, contributor to the process of receptor desensitization. Recovery was less intense despite the smaller response and slower desensitization confirming that, like in the case of rat receptors, the recovery process seemed distinct from the desensitization process. Although identification of the detailed molecular mechanisms responsible for the role of tyrosine-376 in enhancing receptor function requires future investigation, it seems plausible to hypothesize that this residue might allow full development of the agonist-evoked activation as well as speed up receptor transition from desensitization to rest. Thus, tyrosine-376 can be additional to the role of the conserved transmembrane tyrosine-37 (Jindrichova et al. 2011) to control the desensitization process.

In conclusion, we demonstrated that the functional difference between human and rat P2X3 receptors was at least in part determined by the position of tyrosine within the C-terminal sequence. It was possible to convert a lower-function rat receptor to a higher-function human receptor (and vice versa) by shifting the position of a non-conserved tyrosine residue. It should be emphasized that these properties were observed with recombinant receptors expressed by HEK cells; unfortunately, it was not feasible to perform similar tests with human receptors in human neurons. While it was clear that changing C-terminal tyrosine could primarily control the extent of human receptor activation, the application of this notion to in vivo sensory neurons remains conjectural in terms of ATP-mediated nociception. In fact, ion channel expression, kinetics, and recovery are highly sensitive to cell type-specific parameters, including kinase/phosphatase activity, intracellular calcium levels, and plasma membrane PIP2 levels (Zhao et al. 2007; Giniatullin et al. 2008; Köles et al. 2008). Collectively, these factors may determine the species-specific functional outcome of receptor signaling.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The financial support of Telethon Foundation, Italy (Grant no. GGP10082), and of Cariplo Foundation to A.N. is gratefully acknowledged. This work was also supported by the Italian Institute of Technology (IIT to A.N.) and by ARRS grant J3-2376-1540 (to E.F.). We are grateful to Nicol Birsa, Nežka Kavčič, and Anna Marchenkova for their help in some experiments. The authors declare that they have no competing interests.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References