Materials and Methods

DNA Constructs.

The wild-type P2X₁receptor cDNA was cloned from a rat heart cDNA library provided by Dr. M. Tamkun (Vanderbilt University, Nashville, TN). P2X₅ receptor cDNA was a gift of Dr. G. Buell (Ares-Serono, Geneva, Switzerland).

Cell Culture and Transfection Protocol.

HEK-293, COS-7, and Chinese hamster embryonic fibroblast cells were maintained in exponential growth in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal bovine serum, 2 mM glutamine, 50 U/ml penicillin G, and 50 μg/ml streptomycin (Life Technologies, Inc., Rockville, MD). CHO-K1 cells were maintained in 50:50 Dulbecco’s modified Eagle’s medium/F12 medium (Cellgro, Herndon, VA) with the same supplements plus nonessential amino acid solution (Cellgro). All cells were incubated at 37°C in a humidified atmosphere with 5% CO₂. On reaching 70 to 80% confluence in 75cm² tissue culture flasks, the cells were trypsinized and then plated at a density of 3 × 10⁵ cells/35-mm culture dish. The next day they were washed twice with serum free Opti-MEM (Life Technologies, Inc.) and transiently transfected with wild-type P2X₁and/or P2X₅ receptor cDNAs. A mixture of 1 μg total cDNA and 6 μl Lipofectamine (Life Technologies, Inc.) in 1 ml of Opti-MEM was added to each plate for 5 h after which the bathing solution was replaced with the supplemented culture media. The cells were returned to the incubator and used for electrophysiological experiments 36 to 48 h later.

Electrophysiological Recordings.

A suspension of transiently transfected cells were obtained by mechanical dissociation of the contents of a single 35-mm culture dish using a fire-polished Pasteur pipette. An aliquot of the cell suspension was transferred to a recording chamber mounted on the stage of an inverted microscope (Nikon Diaphot 200). The cells were continuously perfused with an extracellular bath solution containing: 150 mM NaCl, 1 mM CaCl₂, 1 mM MgCl₂, 10 mM glucose, 10 mM HEPES, pH 7.4 with NaOH, and whole-cell voltage clamp was performed using an AxoPatch 200A amplifier (Axon Instruments, Foster City, CA) with low resistance (1–2 MΩ) 7052 borosilicate glass electrodes (Garner Glass, Claremont, CA). The intracellular recording solution consisted of: 150 mM CsCl, 10 mM HEPES, 10 mM EGTA, and 10 mM tetraethylammonium-Cl, pH 7.4 with CsOH. Whole cell currents were measured at room temperature from single cells held at −40 mV. Drugs were applied by manually moving the electrode and attached cell into the line of flow of solutions exiting one of a group of inlet tubes feeding the recording chamber. Complete exchange of solution was obtained in less than 100 ms (τ ≈ 24 ms; T.M.E., unpublished observation). Unless otherwise stated, agonists were applied for about 2 s, separated by 1.5- to 2-min intervals; the exception was the P2X₁ receptor, where 3-min intervals were required to ensure recovery from desensitization. Antagonists were applied after a stable control ATP-evoked current amplitude was measured. Continuous application of antagonist began 1 min before the start of the test dose of agonist. Current-voltage relationships were determined for peak, plateau, and tail currents by measuring the response to 1 μM ATP in cells expressing P2X_1/5and voltage-clamped at a range of potentials (−100 to +60 mV). Averaged data were fit with a third order polynomial to estimate the reversal potential. Liquid junction potentials were estimated to be 4.9 mV using Patcher’s Power Tools (Dr. F. Mendez, Göttingen, Germany); no offset correction was used in the present study.

Drugs.

Pyridoxal phosphate-6-azophenyl-2′,4′-disulfonic acid (PPADS) was purchased from Research Biochemicals, Inc. (Natick, MA), suramin was obtained from the Centers for Disease Control (Atlanta, GA), and 2′,3′-O-(2, 4, 6-trinitrophenyl) ATP (TNP-ATP) was purchased from Molecular Probes (Eugene, OR). All other drugs were purchased from Sigma (St. Louis, MO).

Data Analysis.

Agonist concentration-response curves were generated by measuring peak currents evoked by a range of agonist concentrations. Currents from a single cell were normalized to those evoked by a test dose of 100 μM ATP in the same cell. The appropriate data from all cells were pooled and then fit using a Levenburg-Marquardt least-squares algorithm (IgorPro; Wavemetrics, Lake Oswego, OR) to the following equation:I=Imax(1−1/(1+([agonist]]/EC50)nH)) Equation 1where I_max is the maximum current at saturating concentrations of agonist, EC₅₀ is the agonist concentration at which the response is half-maximal, andn_H is the Hill coefficient. Concentration-inhibition curves for antagonists were determined in a similar manner. The ability of a range of antagonist concentrations to inhibit peak currents evoked by 3 μM ATP was measured, the data were pooled, and the IC₅₀ was determined from the best fit toI=Imax(1/(1+([antagonist]/IC50)nH)) Equation 2where I_max is the normalized response to ATP in the absence of antagonist and IC₅₀ is the antagonist concentration at which a 50% inhibition is apparent. Estimates of EC₅₀and IC₅₀ are reported as mean ± S.D. as determined from the curve fits of the pooled data. All other data are mean ± S.E. Significance was determined using Student’st test.

Results

Homomeric P2X₁ and P2X₅ receptors differ in the size and shape of their ATP-gated currents and in their rank order of agonist and antagonist potencies. Cotransfection of HEK-293 cells with the cDNAs encoding the P2X₁ and P2X₅isoforms results in expression of a heteromultimeric P2X_1/5 channel with a novel phenotype (Torres et al., 1998). We find that the phenotype of this heteromeric receptor incorporates features of both individual isoforms. Furthermore, the P2X_1/5 receptor has unique qualities that are not present in either homomeric P2X₁ or P2X₅ receptors.

The Effect of ATP.

Figure 1A shows representative currents evoked by 30 μM ATP in HEK-293 cells expressing P2X₁, P2X₅, and P2X_1/5 receptors. ATP-gated current through the P2X₁ receptor differs from current through the P2X₅ receptor in two ways: current through P2X₁ is larger and is quicker to desensitize. By contrast, current through the heteromeric P2X_1/5receptor is biphasic and has a fast inward current that quickly diminishes to a sustained plateau. That the biphasic current is not due to the simple superposition of P2X₁ and P2X₅ homo-oligomeric currents is supported by three findings. First, repeated and closely spaced applications of ATP produce near complete desensitization of homomeric P2X₁ receptors but causes little or no desensitization of peak current through the P2X_1/5 receptor (Torres et al., 1998). Second, the agonist profile of the sustained phase does not match that of the nondesensitizing P2X₅ receptor (see below). Third, modest concentrations of ATP elicits rebound tail currents in cells expressing the P2X_1/5 receptor, and this current is never recorded from cells expressing either P2X₁ or P2X₅ receptors alone. A typical response to an application of 30 μM ATP is shown in Fig. 1A (right). Current during the drug application was biphasic and had measurable peak and plateau currents. ATP-gated current began to fade immediately on washout of agonist. This acute decline of membrane current reflects the rapid rate of solution exchange that we expect using the drug delivery system described in Materials and Methods. The initial decline was cut short, however, by the development of an additional phase of inward current that gradually reached a peak and then declined to the predrug-application holding current level. In most cells, the rebound tail current was as large or larger than the plateau current seen during the application of ATP. Tail currents were also measured in other mammalian expression systems including CHO-K1 (Fig. 1B), COS-7 (Fig. 1C), and Chinese hamster embryonic fibroblast (data not shown) cells, and are not unique to HEK-293 cells. The rebound tail current, however, is unique to the P2X_1/5 receptor, and we never recorded a similar current on removal of ATP or any other agonist from cells expressing any homomeric or other known heteromeric P2X receptor (data not shown). ATP-induced rebound currents have been demonstrated at high agonist concentrations in PC12 cells expressing P2X receptors as a result of acidification of the extracellular solution by addition of millimolar concentrations of ATP (Stoop and Quayle, 1998). The rebound tail current of P2X_1/5 receptors was not caused by acidification, because it was apparent at much lower concentrations (1–100 μM) of ATP that did not change the pH of the bathing solution and because tail current through the P2X_1/5receptor is relatively insensitive to changes in pH (W.R.H. and T.M.E., unpublished observation).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Unique ATP-activated currents from cells coexpressing P2X₁ and/or P2X₅ receptor isoforms. Typical responses to 30 μM ATP are shown in cells voltage-clamped at −40 mV. A, HEK-293 cell transfected with cDNA encoding homomeric P2X₁ receptors showed the characteristic rapid desensitizing whole-cell current in response to ATP (left). By contrast, a different HEK-293 cell transfected with the cDNA encoding the P2X₅ receptor gave a small nondesensitizing current in response to ATP (middle). A third HEK-293 cell transfected with both cDNAs gave a biphasic current; note the pronounced tail of rebound inward current on washout of ATP (right). B and C, current phenotype observed from the hetero-oligomeric receptor in response to ATP was not unique to HEK-293 cells as similar currents were also observed in CHO-K1 (B) and COS-7 (C) cells coexpressing the two isoforms. Solid lines above all traces in this and the following figures indicate the approximate time of agonist application.

Current-voltage curves were generated by measuring 1 μM ATP-gated currents of individual HEK-293 cells expressing the P2X_1/5 receptor. Individual currents evoked at a range of voltages (−100 to +60 mV) were measured at three different times during the ATP response, and the averaged currents at each potential were plotted against voltage (Fig.2). The current-voltage relationships were not different for the peak, plateau, and rebound tail currents, suggesting that they all evolve from a common population of channels. Some inward rectification was seen at positive potentials, and all three currents had reversal potentials near 5 mV before correction for liquid junction potentials. Thus, P2X_1/5receptors resemble other homomeric and heteromeric P2X receptors in being nonselective cation channels (Bean and Friel, 1993).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Current-voltage relationships for P2X_1/5receptor. Current-voltage relationships were examined by measuring the response to 1 μM ATP of individual HEK-293 cells expressing P2X_1/5 receptors and held at membrane potentials ranging from −100 to +60 mV. For each response the amplitude of the current was measured at three different phases (see inset). For all three phases, current amplitudes were normalized to those elicited at −40 mV and the averaged currents ± S.E.M. for all cells (n ≥ 3) from a given potential were plotted against voltage. I-V curves for each measured phases were then generated by fitting the averaged data with a third order polynomial. Data shown is not corrected for liquid junction potentials estimated to be about 4.9 mV. Inset, typical current elicited by 1 μM ATP from a cell held at −40 mV. Peak, plateau, and rebound phases are labeled at points where current amplitudes were measured for generation of I-V curves.

The Effects of Other Agonists.

Raw currents elicited by applications of common purinoceptor agonists to cells expressing P2X_1/5 are shown in Fig.3 where each row shows the response of a single cell to a different agonist applied at concentrations close to their EC₂₀, EC₅₀, and EC₉₅ values. All agonists evoked monophasic, nondesensitizing currents when applied at the lowest effective concentrations. At the highest concentrations, the currents were biphasic and resembled those evoked by 100 μM ATP with two important differences. First, the ratio of size of the plateau current to the size of the rapidly developing peak current varied among agonists. The difference was greatest for ATP and 2-meSATP, which evoked large peak currents and relatively small plateau currents. By contrast, CTP evoked peak currents that were only slightly larger than the sustained plateaus. Second, applications of some (ATP, 2-meSATP, ADP) but not all agonists were followed by generation of rebound tail currents. The size of the tail current was inversely related to the amplitude of the plateau current in that agonists (ATP, 2-meSATP) that generated small plateaus when applied at low concentrations also evoked large and long-lasting tail currents.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Raw currents elicited by select P2X receptor agonists on P2X_1/5 receptors. Whole-cell currents recorded from HEK-293 cells held at −40 mV and expressing the hetero-oligomeric P2X receptor composed of P2X₁ and P2X₅ subunits. Each family of three traces was recorded from an individual cell and represents typical responses to three different concentrations of the agonist listed. Calibration: 500 pA, 2 s.

Agonist Sensitivity.

Agonist potency was determined from the concentration-response curves of peak currents shown in Fig.4; the results are summarized in Table1. ATP and 2-meSATP were full and potent agonists with EC₅₀ values of 0.7 ± 0.1 and 1.3 ± 0.3 μM, respectively. All other drugs evoked smaller peak currents and were less potent. The rank order of agonist potency measured from EC₅₀ values for each drug was ATP ≥ 2-meSATP > adenosine 5′-O-(3-thiotriphosphate) (ATPγS) > α,β-methylene ATP (α,β-meATP) > ADP > CTP. α,β-methylene ADP (α,β-meADP); UTP; GTP; and AMP were poor agonists that evoked less than 5% of the maximal ATP current even when applied at the highest (100 μM) concentration. Most agonists (ATPγS; α,β-meATP; ADP; and CTP) evoked smaller peak currents than did ATP, suggesting that these drugs may be partial agonists. If so, then one would predict that concentration-response curves of these partial agonists would have lower slopes than those of the full agonist like ATP. However, we measured no significant differences in the Hill coefficients of these drugs and ATP, and other factors may explain the apparent change in efficacy. One possibility is that some agonists cause a rapid receptor desensitization that distorts the shape of their concentration-response curves. The most likely candidates are ATP and 2-meSATP, which apparently cause the greatest degree of rapid desensitization, and these agonists may be even more potent than our estimates reported here. A similar change in efficacy without a change in slope is also apparent in agonist concentration-response curves of the P2X₁receptor (Valera et al., 1994).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Concentration-response curves for the actions of P2X receptor agonists on P2X_1/5 receptor. Peak currents evoked by a range of concentrations of each agonist were normalized to that obtained from 100 μM ATP. Only one other agonist was applied to each individual cell. Each point represents averaged normalized data ± S.E.M. Solid lines are the best fit to the pooled data as described inMaterials and Methods.

Antagonist Sensitivity.

Determination of the subunit composition of native P2X receptors is hampered by a lack of selective purinergic antagonists (Ralevic and Burnstock, 1998). However, a number of agents have been used to discriminate between the P2X and P2Y receptor families, and some newer drugs have shown limited selectivity for the different P2X isoforms. We tested three antagonists against ATP-evoked currents of the recombinant P2X_1/5receptor. Two of these, suramin and PPADS, have been widely tested against purinergic responses of native tissues. The third, TNP-ATP, is a more selective antagonist of P2X₁ and P2X₃ receptors than of any other isoform (Virginio et al., 1998b). We measured the ability of increasing concentrations of each antagonist to block the currents evoked by applications of 3 μM ATP. Figure 5shows raw data for each antagonist. The first trace in each family of currents represents the baseline ATP current before application of antagonist; as expected, ATP-generated inward current had the three distinct phases (rapid peak, sustained plateau, and rebound tail current) described above. The middle traces show the effect of different concentrations of antagonist, and the last trace shows the effect of the first postantagonist application of ATP. For suramin and TNP-ATP, a progressively larger block of ATP-gated current was apparent as the antagonist concentration increased (Fig. 5, top and middle). Peak current was affected first, although all three current components were antagonized. Near complete block of the ATP-induced current was attained with maximal concentrations of these antagonists (≥100 μM suramin and ≥30 μM TNP-ATP), and complete recovery of the response to ATP was seen within minutes of their washout (data not shown). By contrast, the effect of PPADS was complex and varied with concentration. When applied in low concentrations (for example, see 0.1 μM PPADS, Fig. 5, bottom), PPADS had little effect on the peak ATP-gated current but markedly potentiated the plateau currents. These potentiations were irreversible (we tested up to 22 min after washout of PPADS). At higher concentrations, PPADS progressively inhibited all three current components until a near complete block of the ATP-response was achieved. The inhibition by PPADS reversed when the antagonist was removed. The molecular mechanism of the potentiation by PPADS is unknown. One possibility, however, is that some (ATP, 2-meSATP, ADP), but not all, agonists (for example, α,β-meATP) produce a pronounced receptor desensitization that can be modulated by PPADS; this hypothesis is supported by the data in Fig.6. In this experiment, the ability of two concentrations of PPADS to alter currents evoked by α,β-meATP and ATP were measured in a single HEK-293 cell expressing P2X_1/5 receptors. As expected, α,β-meATP evoked large peak and plateau currents, whereas the plateau current evoked by ATP was smaller. If the smaller plateau current measured in ATP reflects an agonist-specific receptor desensitization sensitive to PPADS, then a low concentration of this antagonist would be expected to alter the shape of the ATP-gated current so that it more closely resembles that caused by α,β-meATP. This is exactly what happens. The shapes of the currents evoked by both ATP and α,β-meATP were similar when applied in the presence of 0.1 μM PPADS, a concentration of antagonist that caused only a small effect on the peak current amplitudes. The data also suggest that a separate population of lower-affinity PPADS binding sites is responsible for inhibition of both α,β-meATP and ATP currents, as higher concentrations of PPADS inhibited peak currents caused by either antagonist to an equal degree (see Fig. 6).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Effects of P2X antagonists on raw current elicited by ATP through P2X_1/5 receptors. Whole-cell currents recorded from HEK-293 cells expressing P2X_1/5 receptors. Each family of five traces was recorded from an individual cell voltage-clamped at −40 mV and demonstrates the inhibitory effects of increasing concentrations of the listed antagonist on responses to 3 μM ATP. The first and last trace in each family represents the control ATP response and first ATP response after washout of the antagonist, respectively. Open horizontal bars denote currents recorded in the presence of the given antagonist. All antagonists were applied 1 min before, as well as during, ATP application. Calibration: 400 pA, 5 s.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Effects of PPADS on raw P2X_1/5 currents elicited by ATP and α,β-meATP. Whole-cell currents recorded from HEK-293 cells expressing P2X_1/5 receptors. All currents were recorded from a single cell voltage clamped at −40 mV and represent the effects of two different doses of PPADS on both 10 μM α,β-meATP- and 1 μM ATP-evoked currents. Open horizontal bars denote currents recorded in the presence of PPADS. Calibration: 200 pA, 5 s.

Antagonist IC₅₀ values were determined from the inhibition of peak currents. Concentration-response curves are shown in Fig. 7, and the data are summarized in Table 2. TNP-ATP is a potent antagonist of homomeric P2X₁ receptors with an IC₅₀ of 1 to 3 nM when tested against ATP applied at concentrations close to its EC₅₀ (Virginio et al., 1998b; W.R.H. and T.M.E., unpublished observation). We found that this drug was far less potent in inhibiting peak current through P2X_1/5 receptor, where it had an IC₅₀ of 0.36 ± 0.13 μM. Homomeric P2X₁ and P2X₅ receptors do not differ dramatically in their sensitivity to PPADS and suramin. Not surprising then is the finding that the sensitivity of the heteromeric P2X_1/5 receptor to PPADS (IC₅₀ equals 0.63 ± 0.05 μM) and suramin (IC₅₀ equals 1.6 ± 0.7 μM) resembled those of the homomeric receptors.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7

Concentration-inhibition curves for P2X receptor antagonists on P2X_1/5 receptors. The curves were generated by measuring peak currents evoked by 3 μM ATP in the presence of varying concentrations of each antagonist in single cells. Only one antagonist was applied to each individual cell. The currents were normalized to those evoked by 3 μM ATP in the same cell. Each point represents averaged normalized data ± S.E.M. Solid lines are the best fit to the pooled data as described in Materials and Methods.

The Effect of Extracellular Calcium.

ATP-gated current through the P2X₁ receptor is relatively insensitive to changes in extracellular [Ca²⁺] (Evans et al., 1996; Virginio et al., 1998a; Fig. 8, top). The sensitivities of P2X₅ and P2X_1/5 receptors are unknown. We measured the ability of increasing concentrations of [Ca²⁺]_o to alter ATP-gated current in cells expressing either P2X₁, P2X₅, or P2X_1/5 receptors. Current evoked by 15 μM ATP through P2X₅ receptors was strongly inhibited by raising [Ca²⁺]_o with an apparent IC₅₀ of 6.7 ± 1.9 mM (Fig. 8, middle). By contrast, peak currents through P2X_1/5 receptors were unaffected by changes in [Ca²⁺]_o (Fig. 8, bottom). However, increasing [Ca²⁺]_o potentiated plateau currents evoked by 3 μM ATP, perhaps by the same mechanism that underlies the potentiation of this phase of current by PPADS. No enhancement was seen when Ba²⁺ was substituted for extracellular Ca²⁺, and chelating [Ca²⁺]_i by addition of 20 mM BAPTA to the pipette solution had no effect on the magnitude of the rebound current (data not shown). These results suggest that the effect of Ca²⁺ does not involve an intracellular pathway, but may involve a Ca²⁺-specific effect on a part of the channel protein accessible from the bath solution. A summary of the effects of [Ca²⁺]_o on currents through P2X₁, P2X₅, and P2X_1/5 is presented in Fig.9.

• Download figure
• Open in new tab
• Download powerpoint

Figure 8

Sensitivity of P2X_1/5 receptors to changes in extracellular calcium concentration. The effects of increasing concentrations of extracellular calcium on ATP-activated currents recorded from cells expressing P2X₁ and/or P2X₅ receptors were investigated. Typical responses are shown. Changing the extracellular calcium concentration from 1 to 10 or 30 μM had no effect on the amplitude of the current elicited by 3 μM ATP in cells expressing only P2X₁ receptors (top). However, a change in the extracellular calcium concentration from 1 to 10 μM greatly inhibited currents evoked by 15 μM ATP in a cell expressing P2X₅ receptors alone; further increasing the extracellular calcium to 30 mM completely and reversibly abolished the ATP-activated current (middle). In a cell expressing the hetero-oligomeric P2X_1/5 receptor, increasing calcium concentrations had no effect on peak current amplitude elicited by 3 μM ATP, although a reversible dose-dependent increase in the plateau phase was observed (bottom). Current through P2X₁ receptor was filtered at 300 Hz; other traces were filtered at 5 KHz.

• Download figure
• Open in new tab
• Download powerpoint

Figure 9

Pooled data from experiments described in Fig. 8. Data are mean ± S.E. Data significantly different (p < .01) from control currents recorded in 1 mM [Ca²⁺]_o are marked with asterisks.

Discussion

The unique properties of some native ATP-induced currents might result from formation of novel P2X receptors by hetero-oligomeric assembly of different receptor isoforms. Evidence is now mounting to support this hypothesis. Multiple P2X receptor isoforms are expressed in many of the same tissues (Ralevic and Burnstock, 1998). Furthermore, most P2X receptor isoforms (the exception is P2X₇) can form hetero-oligomeric assemblies (Torres et al., 1999). Finally, novel receptors formed by heteropolymerization of P2X₂ and P2X₃ (Lewis et al., 1995), P2X₁ and P2X₅ (Torres et al., 1998), and P2X₄ and P2X₆ (Le et al., 1998) have been shown to functionally assemble in heterologous cell systems. A more complete survey of ATP-gated currents of native tissues therefore critically depends on a complete functional characterization of the possible combinations of heteromultimeric assemblies. We present here an extensive characterization of the novel P2X_1/5receptor expressed in HEK-293 cells. Results from in situ hybridization, PCR/cloning analysis, and immunocytochemistry experiments have implicated the coexistence of P2X₁ and P2X₅ receptors in heart (Bogdanov et al., 1998) and spinal cord tissue (Collo et al., 1996), and our results provide a basis for subsequent investigation of the effects of extracellular ATP in those tissues.

The pharmacological profile of the P2X_1/5receptor is different from that of either the P2X₁ and P2X₅homo-oligomeric receptors and borrows unique traits from each isoform. The agonist potency ranked by EC₅₀ was ATP ≥ 2-meSATP > ATPγS > α,β-meATP > ADP > CTP. Thus, like the P2X₅ receptor, ATP is slightly more potent than 2 meSATP at the P2X_1/5receptors. However, unlike the P2X₅ receptor (but like P2X₁ receptors), the P2X_1/5 receptor is appreciably sensitive to α,β-meATP. Indeed, the ability of α,β-meATP to activate biphasic currents with large plateaus is probably the most distinguishing characteristic of the P2X_1/5phenotype. P2X_1/5 receptors also show large sustained currents in response to the pyrimidine nucleotide, CTP, a property not shared by the P2X₅ receptor (Garcia-Guzman et al., 1996). P2X_1/5 receptors are as sensitive to the antagonists suramin and PPADS as are P2X₁ and P2X₅ receptors, but they are markedly less sensitive to TNP-ATP than the P2X₁ receptor. Furthermore, the novel heteromultimeric receptor resembles the P2X₁receptor in the inability of a rise in extracellular calcium to block peak ATP-gated current.

The dual effects (e.g., potentiation and block) of PPADS on the shape of ATP-gated currents are unique to P2X_1/5receptors. One explanation is that cotransfection of HEK-293 cells with cDNAs encoding the P2X₁ and P2X₅ isoforms leads to at least two separate populations of heteromultimeric receptors that differ in their response to PPADS. If so, then the agonist and antagonist profiles of these two populations must be very similar because all of the concentration-response curves were well fit using models that assume either a homogenous receptor population or multiple populations with identical parameters. Another possibility is that PPADS (and [Ca²⁺]_o) affect receptor desensitization in an agonist-specific manner. Thus, full agonists like ATP and 2-meSATP cause a rapid decline in current as the receptor desensitizes, whereas partial agonists like α,β-meATP cause less desensitization and have larger plateaus. ADP is an exception to this rule; it is not a full agonist but it does have a pronounced rebound tail current when applied in high (300 μM) concentrations. This prominent tail current may result from contamination of ADP with ATP. Regardless, if the rate of desensitization is lengthened by application by PPADS, then we would expect this drug to have its greatest effect on those responses that show the greatest desensitization; this is exactly the case. Application of 0.1 μM PPADS greatly increases the size of the plateau current evoked by ATP, changing the shape of the response so that it closely resembles the current evoked by α,β-meATP. This model assumes that there are at least two populations of PPADS binding sites: a high-affinity site that results in change in receptor desensitization and a lower-affinity site that results in noncompetitive inhibition of current. This is because higher concentrations of PPADS inhibited all current components equally well. The presence of multiple populations of PPADS binding sites is also supported by the Hill slope of this antagonist’s concentration-response curves (Fig. 7 and Table 2). Furthermore, raising [Ca²⁺]_o would have an effect like PPADS if the rate of desensitization can also be modulated by Ca²⁺. Indeed, the time course of desensitization of homomeric P2X₃ receptors is modulated by changes in [Ca²⁺]_o (Cook and McCleskey, 1997), and a similar mechanism may be present in P2X_1/5 receptors. It will be interesting to see if site-directed mutagenesis identifies common or different sites underlying the similar effects of Ca²⁺_o and PPADS on desensitization.

The agonist-dependent change in receptor desensitization may also explain the genesis of the rebound tail current. Full agonists that have pronounced desensitization also have the largest rebound tail currents. If recovery from desensitization can only occur from the open state, then a channel would need to cycle from the desensitized, to the open, to the closed state on removing agonist; rebound current could occur if the cycle is slow. Of course, other explanations are also possible. For example, the tail current may result from relief of open channel block perhaps by ATP itself. However, the size of the tail current is an approximately linear function of voltage, and this lack of voltage dependence argues against such a mechanism. Additional experiments are needed to more fully understand the mechanisms that underlie generation of this current component.

Several native fast-acting ATP responses have been recorded although the subtype of P2X receptor mediating these responses is still unknown. In pyramidal neurons of the CA1 area of the rat hippocampus, both excitatory postsynaptic currents from hippocampal slices and ATP-activated whole cell currents sensitive to PPADS have been identified (Pankratov et al., 1998). ATP and similar analogs also elicit fast rising inward currents attributable to activation of P2X receptors in trigeminal mesencephalic nucleus neurons of the rat (Khakh et al., 1997). Finally, P2X purinoceptors were shown to mediate excitatory effects in rat locus ceruleus neurons (Sansum et al., 1998). The specific P2X receptor isoform(s) which mediate these different purinergic responses have yet to be conclusively identified, and it is possible that a hetero-oligomeric P2X receptor might account for the different responses. mRNAs encoding the P2X₁ and P2X₅ isoforms are both found in large neurons of the ventral horn of the rat spinal cord (Collo et al., 1996), raising the possibility that native heteromeric P2X_1/5receptors may exist. Here, we have presented data further characterizing the pharmacological profile of the novel recombinant P2X_1/5 ionotropic receptor. These data should provide the foundation for functional identification of purinergic responses of native systems.