Activation of membrane P2X7 receptors by extracellular ATP [or its analog 2′,3′-O-(4-benzoylbenzoyl)-ATP] results in the opening within several milliseconds of an integral ion channel that is permeable to small cations. If the ATP application is maintained for several seconds, two further sequelae occur: there is a gradual increase in permeability to the larger cation N-methyl-d-glucamine and the cationic propidium dye quinolinium, 4-[(3-methyl-2(3H)-benzoxazolylidene)methyl]-1-[3-(triethylammonio)propyl]diiodide (YO-PRO-1) enters the cell. The similarity in the time course of these two events has led to the widespread view that N-methyl-d-glucamine and YO-PRO-1 enter through a common permeation pathway, the “dilating” P2X7 receptor pore. Here we provide two independent lines of evidence against this view. We studied single human embryonic kidney cells expressing rat P2X7 receptors with patch-clamp recordings of membrane current and with fluorescence measurements of YO-PRO-1 uptake. First, we found that maintained application of the ATP analog did not cause any increase in N-methyl-d-glucamine permeability when the extracellular solution contained its normal sodium concentration, although YO-PRO-1 uptake was readily observed. Second, we deleted a cysteine-rich 18-amino acid segment in the intracellular juxtamembrane region of the P2X7 receptor. This mutated receptor showed normal YO-PRO-1 uptake but had no permeability to N-methyl-d-glucamine. Together, the clear differential effects of extracellular sodium ions or of mutation of the receptor strongly suggest that N-methyl-d-glucamine and YO-PRO-1 do not enter the cell by the same permeation pathway.
- cation channel
- quinolinium, 4-[(3-methyl-2(3H)-benzoxazolylidene)methyl]-1-[3-(triethylammonio)propyl]diiodide
p2z receptors were named by Gordon (10) to describe those receptors through which ATP exerted unusual actions on certain immune cells, particularly mast cells and lymphocytes. The actions were unusual in that hundredfold higher concentrations of ATP were required (>300 μM) than those effective at P2X and P2Y receptors and because one of the effects of ATP was to induce cell “permeabilization.” By permeabilization was meant the opening of a permeation pathway through which nucleotides themselves could leave cells (4) or through which large molecules could enter cells (e.g., carboxyfluorescein, ethylenediaminetetraacetic acid, ethidium; Refs. 1, 33). Another unusual feature of the P2Z receptor was that the effectiveness of ATP was much increased by removal of calcium and/or magnesium from the extracellular solution, and this has been interpreted by many to suggest that ATP4− rather than Mg-ATP was the effective agonist at this receptor (32). The ability of ATP to activate a permeation pathway to large fluorescent molecules has since been studied extensively in a range of mast cells, macrophages, lymphocytes, and related cells (reviewed in Ref. 23). Subsequent studies have also identified several further downstream cellular consequences of activating P2X7 receptors: these include release of interleukins, cytoskeletal rearrangements, L-selectin shedding, and activation of phospholipase D and p38 MAP kinase (see Ref. 23).
The P2X7 receptor was named as the seventh and last mammalian cDNA to be isolated that encodes a P2X receptor subunit (28, 31). The protein sequence of the P2X7 subunit is 35–45% identical to the other six P2X subunits, from which it differs most strikingly by its longer intracellular COOH terminus. ATP elicits within tens of milliseconds the activation of a cation-permeable channel and within several seconds the uptake of fluorescent dyes such as ethidium and quinolinium, 4-[(3-methyl-2-(3H)-benzoxazolylidene)methyl]-1-[3-(triethylammonio)propyl]diiodide (YO-PRO-1) (31). Strikingly, high concentrations of ATP (>100 μM) were required to elicit these effects, and the actions of ATP were much potentiated by removal of extracellular calcium and/or magnesium ions. For these and other reasons (such as the predominant expression of the P2X7 subunit in immune cells) it was inferred that the P2X7 receptor corresponded to the P2Z receptor, a conclusion that has since been well confirmed (see Ref. 23).
The observation that these two properties were conferred to cells by transfection with a single cDNA prompted the interpretation that they reflected the properties of a single protein. Thus ATP would initially open an ionic channel selectively permeable to small cations, and this subsequently “dilated” into a larger pore that was also permeable to larger cationic dyes (ethidium is a monovalent cation of 314 Da, and YO-PRO-1 is a divalent cation of 376 Da). This was tested directly in electrophysiological experiments by measuring the relative permeability of a large cation, N-methyl-d-glucamine (NMDG; 196 Da) (31, 35). These studies showed that either repeated (31) or sustained (35) application of agonist to cells expressing P2X7 receptors resulted in a large increase in permeability to NMDG. The finding that the rate of increase in permeability (time constant ∼7 s for NMDG) was broadly similar to the initial rate of uptake of YO-PRO-1 was consistent with the two ions entering the cell through a similar permeation pathway.
However, although P2X7 receptors respond to ATP with the rapidly developing cation current in a wide range of expression systems (see Ref. 23), the development of NMDG permeability and the uptake of fluorescent dyes are not universally observed (13, 19, 26). This introduces the possibility that the NMDG permeation/YO-PRO uptake pathway and the small cation pathway are actually formed by distinct molecules: the small cation permeation due to the P2X7 ion channel and the NMDG/YO-PRO pathway being provided by the host cell that is engaged by the activated P2X7 receptor. In the present study we made the unexpected discovery that not only can the small cationic channel be differentiated from the NMDG dilatation/YO-PRO uptake pathway but also the NMDG permeation can be differentiated from the YO-PRO uptake path.
MATERIALS AND METHODS
Human embryonic kidney (HEK)-293 cells stably expressing rat P2X7 receptors were generated as previously detailed (34). HEK-293 cells were also used to express transiently both wild-type and mutant rat P2X7 receptors by transfection of cDNA with Lipofectamine 2000 (GIBCO) as described previously (14, 28), and cells were replated on 35-mm coverslips and incubated at 37°C for 5–12 h before use.
The 54 nucleotides (1084–1137 of the P2X7 open reading frame) coding for the 18-amino acid cysteine-rich (CCRSRVYPSCKCCEPCAV) region of the receptor were deleted by overlapping PCR. Two overlapping fragments of P2X7 cDNA were amplified by PCR with Taq polymerase; primers were designed to allow deletion of the 54 nucleotides and overlapping amplification. The two amplified fragments were combined and used as DNA template in a third PCR with external primers. The resulting fragment that lacked the 54 nucleotides was substituted into the P2X7 cDNA by direct subcloning with the unique SacII and BsrGI endogenous restriction sites. The amplified region was confirmed by sequencing.
Whole cell recordings were carried out with an EPC9 patch-clamp amplifier (HEKA Elektronik) as previously detailed (7, 14, 15). The membrane potential was held at −60 mV unless otherwise indicated. The intracellular solution contained (in mM) 147 NaCl or NaF or KCl, 10 HEPES, and 10 EGTA. The standard extracellular solution contained (in mM) 147 NaCl, 2 KCl, 1 MgCl2, 2 CaCl2, 10 HEPES, and 13 glucose. Some experiments were done in extracellular solutions containing sodium only (mM: 147 NaCl, 10 HEPES, and 23 glucose), NMDG only (in mM: 154 NMDG-Cl, 10 HEPES, and 13 glucose), or a mixture of NMDG and sodium (in mM: 138 NMDG-Cl, 15 NaCl, 10 HEPES, and 13 glucose). All solutions were maintained at pH 7.3 and 300–315 mosM. The reversal potentials (Erev) were obtained by repeated applications (every 2 or 4 s) of voltage ramps of −100 mV to 40 mV with 1-s duration. The relative ion permeability, PNMDG/PNa, was derived with the following equation: PNMDG/PNa = [[Na]i exp(x)−[Na]o]/[NMDG]o, where x = ErevF/RT (F is Faraday constant, R is the gas constant, and T is the absolute temperature), [Na]i is intracellular sodium concentration, [Na]o is extracellular sodium concentration, and [NMDG]o is extracellular NMDG concentration.
Agonist concentration-current curves were fit to the Hill equation: I/Imax = 100 [[A]nH/([A]nH+ EC50nH)], where I is the peak current evoked by agonist concentration [A] expressed as the percent maximal current (Imax) evoked by 2′,3′-O-(4 benzoylbenzoyl)-ATP (BzATP), nH is the Hill coefficient and EC50 is half-maximal [A].
Immunocytochemistry was performed as previously described (17). We used the primary mouse monoclonal antibody (BabCo, Richmond, CA) against the epitope (EYMPME) tagged to the COOH terminus of the receptors at a dilution of 1:1,000 and the secondary fluorescein isothiocyanate-conjugated anti-mouse IgG antibody (Sigma) at a dilution of 1:200. We also used the monoclonal rat ectodomain anti-P2X7 antibody (16, 17) at 1 μg/ml under nonpermeabilizing conditions to compare membrane localization of receptors.
YO-PRO-1 uptake was measured with a Zeiss Axiovert 100 with a Fluar ×40 objective and the Photonics monochromator imaging system (Photonics) as previously described (15, 36). YO-PRO-1 (2 μM) was included in extracellular solutions throughout experiments, and fluorescence was measured from individual cells and averaged after subtracting background fluorescence before application of agonists. Cumulative YO-PRO-1 concentration response curves were obtained as follows. Agonists were added in increasing concentrations; each concentration was present for 90 s followed by 8- to 10-min wash. Fluorescence intensity was measured immediately before agonist washout. Because divalent cations directly reduce the fluorescence of these dyes (11), valid comparisons are not possible between responses in the presence and absence of calcium/magnesium. Therefore, concentration-response curves for YO-PRO-1 uptake were made in identical solutions for wild-type and cysteine-rich deleted mutant receptors. Data are presented where appropriate as means ± SE, and tests for statistical significance were done with Student's t-test or two-way ANOVA test as indicated.
Measurements of Membrane Current: Estimation of NMDG Permeability
Effects of extracellular sodium.
NMDG permeability can only be estimated accurately in bi-ionic conditions. Figure 1, A and B, illustrates the usual experiment, in which the intracellular solution contains predominantly sodium ions (2, 18, 35). The extracellular solution was first changed from the control solution (sodium; with 2 mM calcium and 1 mM magnesium) to NMDG (without calcium or magnesium), and some seconds later the application of BzATP was begun. The initial current through the activated P2X7 receptor was outward, indicating that the cell is more permeable to sodium than to NMDG. When the same agonist application protocol was carried out but with ramp voltages applied at 2-s intervals, Erev at the peak of the initial outward current was about −75 mV (Fig. 1B, downward arrowhead), which corresponds to PNMDG/PNa of <0.05. The current then became inward over the next 10–20 s, and Erev became less negative (Fig. 1B, upward arrowhead); the change of Erev occurred with a time constant (τ) of 6.9 ± 0.5 s (n = 7; Fig. 1D). Figure 1D, bottom, shows the computed change in PNMDG/PNa from ∼0.04 immediately after application of BzATP to 0.26 after 30 s. These experimental procedures replicate those described previously, and the numerical results are very similar (35, 36). They show that the relative permeability to NMDG increases severalfold during the first 10 s after application of the agonist BzATP. We repeated these experiments, using recording electrodes that contained 148 mM potassium, and we observed the same progressive increase in permeability when the extracellular solution was changed from sodium to NMDG (τ = 7.8 ± 2.5 s; n = 3). Similar current response and shift in Erev were observed when ATP (1 mM) was used to activate the P2X7 receptor; when ATP was the agonist, Erev shifted from −79 ± 3.2 mV to −27 ± 2 mV with a τ of 5.4 s (n = 4).
Figure 1C shows a similar experiment in which the extracellular NMDG solution contained ∼10% sodium ions (138 mM NMDG, 15 mM Na). In this case the initial Erev was about −50 mV, and this progressively became less negative during the application of BzATP (τ = 11.4 ± 1.2 s; n = 5). The relative permeability (PNMDG/PNa; see materials and methods) computed from these Erev is shown in Fig. 1D and reaches a final value of ∼0.15. This increase in permeability was significantly less than observed when the solution contained only NMDG (Fig. 1D; P < 0.01, ANOVA). This experiment indicates that the presence of even a relatively low fraction of extracellular sodium ions considerably reduces the permeability increase that is observed when NMDG is the only extracellular cation. Therefore, we next examined whether any change at all in NMDG permeability occurred in the presence of a normal extracellular sodium concentration.
Figure 2A illustrates the approach. With sodium as the only extracellular cation, BzATP elicited a large inward current (holding potential −60 mV). After 10 s in BzATP, the extracellular solution was changed from sodium (with BzATP) to NMDG (with BzATP). The current immediately became outward, and then progressively over the next 10 s or so it became inward (compare with Fig. 1A). The application of BzATP was discontinued after 30 s; it should be noted that the inward current remained even though the agonist had been removed. When the extracellular solution was changed from NMDG to the control sodium solution there was a transient increase in the inward current (presumably because the open channel is more permeable to sodium than NMDG) before it rapidly declined as the channel closed. The procedure was then repeated: after the second 10-s application of BzATP the channel remained open for 5 min after washout of the agonist and closed only when the NMDG was replaced with control solution.
These experiments show that even during 10-s application of BzATP in normal sodium-containing extracellular solution there had been no increase in permeability to NMDG even though many previous studies showed YO-PRO-1 dye uptake under these conditions and within this time period (reviewed in Ref. 23). This was clearly shown by the measurements of Erev in Fig. 2, B and C. The initial Erev after introduction of NMDG was close to −70 mV, and it then declined toward a steady-state value of about −40 mV. In other words, no increase in NMDG permeability had occurred even after 30 s of BzATP (in extracellular sodium), but this increase developed with its usual time course as soon as extracellular sodium was replaced by NMDG. We also attempted to record BzATP-evoked currents when the patch pipette contained 150 mM NMDG and zero sodium, but dialysis of the NMDG solution resulted in activation of a large Trp-like, or hemichannel, current that prevented us from determining the P2X7 receptor-mediated current in isolation (data not shown). In summary, application of agonist alone does not cause any “dilatation” of this permeation pathway; this does not occur until there is additional substitution of the extracellular sodium ions with NMDG. Figure 2D illustrates a possible scheme that might account for these experimental observations.
Effects of deleting a cysteine-rich juxtamembrane intracellular domain.
The P2X7 receptor contains an 18-amino acid segment immediately after the second membrane-spanning domain. This sequence is not found in the other P2X receptors (Fig. 3A), and this might suggest that it contributes to the marked NMDG permeability increase and/or YO-PRO-1 uptake shown by the P2X7 receptor. We therefore studied the P2X7 receptor in which this segment had been deleted (ΔCys rich). The membrane expression of this mutant form was normal, as judged by immunocytochemistry of the wild-type and ΔCys-rich receptors carrying COOH terminus epitope tags (Fig. 3A). Furthermore, maximal concentrations of BzATP gave currents similar in amplitude to those seen in wild-type receptors [amplitude to 300 μM BzATP was 3.46 ± 0.3 nA (n = 29) at wild-type and 4.1 ± 0.37 nA (n = 20) at ΔCys-rich receptor; P value = 0.18]. The concentration-response curve for the ΔCys-rich receptor was two- to threefold shifted to the left of that for the wild-type receptor (Fig. 3C). The more obvious immediate difference between the ΔCys-rich and wild-type channels was their sixfold slower closing rate [Fig. 3B; time constants of decay of the currents were 1.5 ± 0.44 s (n = 4) and 9.1 ± 0.83 s (n = 4) for wild-type and ΔCys-rich P2X7 receptors, respectively]. The finding of significant alteration in current kinetics in line with lesser alterations in agonist potency has been observed in mutagenesis studies on P2X2 receptors (14, 41). It is also consistent with work by Markwardt and colleagues (19) on human P2X7 receptors, in which they have provided evidence for two states of agonist binding, with the COOH-terminal domain influencing both activation and deactivation kinetics of one of these states.
In contrast to these minor alterations, the most striking finding, and the one that directly relates to the “large pore” properties of this receptor, was that there was no increase in NMDG permeability in the cells expressing the ΔCys-rich P2X7 receptor. Figure 4A shows a direct comparison between a cell expressing the wild-type receptor (top) and a cell expressing the ΔCys-rich receptor (bottom). Even after 30-s application of BzATP the current (at −60 mV) remained outward, and Erev remained close to −75 mV (Fig. 4B).
Measurements of YO-PRO-1 Uptake: Effects of Deleting Cysteine-Rich Juxtamembrane Intracellular Domain
Figure 5 shows that the uptake of YO-PRO-1 by cells expressing the ΔCys-rich receptor was not inhibited relative to that seen for the wild-type receptors (it was consistently slightly faster). The concentration of BzATP required to elicit YO-PRO-1 uptake in the ΔCys-rich P2X7 receptor was about fourfold less than for the wild type (Fig. 5B); this difference was similar to that observed for the membrane currents (Fig. 3C). This was observed whether the extracellular solution contained sodium or NMDG as the principal cation (Fig. 5A).
The notion has become widespread that P2X7 receptors undergo a “dilatation” of their permeation pathway during agonist applications that are maintained for several seconds (Refs. 35, 36; see Ref. 23). This has been measured in some cases as a progressive increase in PNMDG/PNa and in others as the increase in fluorescence as extracellular dyes enter the cell and bind to nucleic acids. This latter method has become a general, even automated, way of studying P2X7 receptor function. Until quite recently, it had been assumed and generally accepted that the P2X7 ion channel itself “dilated” sufficient to allow entry of large cations (NMDG) and dyes (ethidium, YO-PRO-1). This explanation is now considered unlikely in view of several studies that have indicated or shown a clear dissociation between the “small” cationic channel and the “large” (NMDG/YO-PRO-1) pore (Refs. 8, 29, 34; see Ref. 23). However, all studies to date have suggested that the mechanism(s) underlying both NMDG and dye permeation is the same. Our present experiments now make this assumption also untenable. We have demonstrated two ways in which the increase in NMDG permeability and the entry of YO-PRO-1 at the rat P2X7 receptor can be readily distinguished. First, in normal extracellular sodium concentrations there is no NMDG permeability increase, but there is YO-PRO-1 uptake. Second, deletion of a cysteine-rich juxtamembrane domain of the receptor eliminates NMDG permeability but does not reduce YO-PRO-1 uptake. In neither case is the ability of the P2X7 receptor to function as an ATP-gated channel permeable to small cations grossly impaired.
NMDG Permeability Increase
Wiley and colleagues (38–40) first demonstrated that extracellular sodium ions inhibited both calcium influx and rubidium efflux evoked by activation of endogenous P2X7 receptors in human lymphocytes and concluded that binding of sodium to this receptor inhibited further sodium influx (40). Our results support this interpretation and expand on possible mechanisms for this inhibition. We found that the progressive increase in permeability to NMDG did not occur in an extracellular solution containing the normal sodium ion concentration and was considerably inhibited even by the presence of 15 mM extracellular sodium. It is difficult to study higher sodium concentrations because of the increasing contribution that sodium ions make to the current flow. A remarkable new observation was that the NMDG-permeable channel does not close even when the agonist is removed; it closes only when sodium is reintroduced into the solution (Fig. 2A). This suggests that the receptor can bind both ligand (BzATP) and sodium ions, and that the dissociation of sodium ion from the BzATP-liganded channel allows the conformation to change into one that is NMDG permeable (Fig. 2D). The channel then becomes locked in an open, NMDG-permeable state, until sodium ions are reintroduced. The present results also appear to rule out the existence of an independent NMDG-permeable channel that is activated by P2X7 receptor activation. If this were the case, we would expect to see maximum opening of the NMDG permeation pathway on prolonged activation of the P2X7 receptor in our normal extracellular sodium solution; this did not occur.
The possible physiological significance of this behavior might be considerable. In some circumstances, ATP acts in conditions of reduced extracellular sodium. One of these occurs at the luminal surface of airway epithelial cells, where ATP induces currents in the relatively low-sodium environment of the ciliary mucus (20). Our results would indicate that even brief applications of ATP under these conditions could result in a very sustained activation of the NMDG-permeable form of the P2X7 receptor.
The transition from sodium- to NMDG-permeable state is robustly observed for P2X7 receptors expressed in HEK293 cells (this study; Refs. 35, 36). Xenopus oocytes show a similar phenomenon in some (18, 25) but not other (26) studies. In native cells in which the current induced by ATP has the features suggestive of P2X7 receptor involvement, an increase in NMDG permeability is observed in some (e.g., rat GH3 pituitary cells; Ref. 3) but not other [e.g., bovine aortic endothelial cells (27), mouse NG108-15 neuroblastoma cells (37), human Muller cells (24)] cases. Species and/or expression densities may contribute to the differences. For example, the NMDG permeability increase in the human P2X7 receptor is much less than that seen in the rat, when examined under comparable conditions (Surprenant A, unpublished observations; see also Ref. 28), and ethidium uptake by macrophages occurs only when P2X7 receptor density is high (12). The properties of the P2X2 receptor also depend on the density of expression in heterologous systems (5, 9). Importantly, in all previous studies in which both NMDG permeability changes and dye uptake have been measured, these two events have been well correlated (18, 19, 22, 35, 36), thus leading to the conclusion that NMDG permeability shifts reflect opening (dilatation) of the dye uptake pathway. Our present results with the cysteine-rich deleted receptor show this conclusion to be untenable.
The ΔCys-rich P2X7 subunit lacks an 18-amino acid segment that is likely to be just inside the cell adjoining the second transmembrane domain. The most striking difference between P2X7 ΔCys-rich and wild-type receptors is the complete absence of any increase in permeability to NMDG during prolonged agonist applications. We considered the possibility that this domain might itself constitute an intracellular ion binding site, but we observed no changes in the current when we included zinc (100 μM) or glutathione (reduced GSH or oxidized GSSG, 1 mM) in the recording electrode, or when the recordings were made with potassium rather than sodium as the main intracellular cation (Jiang L-H, unpublished observations). The sequence of the 18-amino acid domain has no obvious homologs in the databases.
Mutations that alter the propensity to become NMDG permeable have also been described for other P2X receptors. Thus it is increased relative to wild-type in rat P2X2[N333A] (36) and P2X4[G347R] (18) and prevented in P2X4[G347K] (18). We can only speculate about the possible structural explanations that might underlie this. It seems unlikely that the cysteine-rich domain is intimately involved in the ion permeation pathway, given that the human P2X5 receptor shows marked NMDG permeability but completely lacks this segment (2).
The principal focus of the present work has been a comparison between NMDG permeation and YO-PRO-1 entry. Our observations with uptake of YO-PRO-1 confirm and extend earlier experiments showing that a YO-PRO-1 entry pathway becomes activated within a few seconds of application of agonist to the P2X7 receptor. In agreement with earlier studies (12, 22, 30, 35, 39), we found that BzATP stimulates the entry of YO-PRO-1 even when all the extracellular sodium ions are replaced with NMDG. In fact, extracellular sodium appears to strongly inhibit YO-PRO-1 (or ethidium) uptake in response to activation of P2X7 receptors. Much of this inhibition can be attributed to an action of sodium at the ATP binding site, as its removal (replacement with sucrose or choline) causes a 20-fold leftward shift in ATP or BzATP EC50 for YO-PRO-1 uptake (22). YO-PRO-1 uptake was not inhibited, but rather somewhat increased, in cells expressing the P2X7 ΔCys-rich receptor. This was seen whether the extracellular solution contained sodium or NMDG (Fig. 5). Thus this mutation provides a clear separation between the property of the channel to become NMDG permeable and the property of the receptor to allow YO-PRO-1 uptake. Both mechanisms are modulated by extracellular sodium, but only the NMDG permeation pathway is strictly dependent on this cation.
Other kinds of stimuli independent of P2X receptors can elicit YO-PRO-1 entry, although most other studies involve measurements of fluorescence over a time course of minutes or hours rather than the tens of seconds reported here. Maitotoxin stimulates YO-PRO-1 uptake in human skin fibroblasts; the maitotoxin currents are not NMDG permeable (21). These findings, and the present results, are most readily interpreted to suggest that the YO-PRO-1 entry pathway is not the P2X7 receptor itself, but a distinct molecular mechanism that is switched on as a consequence of P2X receptor activation. Such a mechanism, perhaps a distinct pore or transporter, can be engaged within several seconds. A recent study using a selective blocking agent strongly implicates p38 MAP kinase in this transduction (6).
In summary, we report that the development of permeability to NMDG that occurs in rat P2X7 receptors activated by BzATP does not occur under usual physiological conditions (i.e., normal extracellular sodium, calcium, and magnesium concentrations) but would occur when extracellular sodium concentrations are greatly reduced. On the other hand, the rapid uptake of YO-PRO-1 (and, presumably, related dyes) that follows P2X7 receptor activation clearly does occur in usual physiological conditions. The YO-PRO-1 uptake is unaffected by a mutation that completely prevents the increase in NMDG permeability, suggesting again that these are distinct pathways. The simplest conclusion is that the NMDG permeability of the P2X7 receptor is controlled by an allosteric effect of sodium ions. Thus the NMDG-permeable pore is most likely resident within the ion channel itself, whereas the entry of YO-PRO-1 into cells expressing P2X7 receptors is most likely through a distinct permeation pathway.
This work was supported by The Wellcome Trust.
We thank Vicky Porteous for help with cell culture.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society