CPX, a selective A₁-adenosine-receptor antagonist, regulates intracellular pH in cystic fibrosis cells

Abstract

The selective A₁-adenosine-receptor antagonist, &cyclopentyl-1,3-dipropylxanthine (CPX), has been reported to activate Clefflux from cystic fibrosis cells, such as pancreatic CFPAC-1 and lung IB3 cells bearing the cystic fibrosis transmembrane regulator(ΔF508) mutation, but has little effect on the same process in cells repaired by transfection with wild-type cystic fibrosis transmembrane regulator (O. Eidelman, C. Guay-Broder, P. J. M. van Galen, K. A. Jacobson, C. Fox, R. J. Turner, Z. I. Cabantchik, and H. B. Pollard. Proc. Natl. Acad. Sci. USA 89: 5562–5566, 1992). We report here that CPX downregulates Na⁺/H⁺ exchange activity in CFPAC-1 cells but has a much smaller effect on cells repaired with the wild-type gene. CPX also mildly decreases resting intracellular pH. In CFPAC-1 cells, this downregulation is dependent on the presence of adenosine, since pretreatment of the cells with adenosine deaminase blocks the CPX effect. We also show that, by contrast, CPX action on these cells does not lead to alterations in intracellular free Ca²⁺ concentration. We conclude that CPX affects pH regulation in CFPAC-1 cells, probably by antagonizing the tonic action of endogenous adenosine.

CYSTIC FIBROSIS (CF), the most common genetic disease in the Caucasian population, is caused by mutations in the cystic fibrosis transmembrane regulator (CFTR; Refs. ^{25, 36, 37}). The disease is manifest by defects in adenosine 3’,5’-cyclic monophosphate (cAMP)-dependent Cl⁻ secretion and reabsorption in epithelial cells (35) and by other abnormal manifestations in Na+ reabsorption (47), mucin secretion and composition (6, 30, 31), protein processing (8, 9) and trafficking (5, 34), phosphorylation (40), and intracellular pH (pH_i) regulation (2, 15, 46, 48). The most common mutation in CFTR is ΔF508, but neither the molecular basis of the resultant pathological phenotype nor the function of CFTR itself is fully understood. The current consensus is that CFTR functions as a Cl⁻ channel (1, 3, 42) and that the disease is due to impairment in Cl⁻ conductance (17). The CFTR(ΔF508) gene product is able to conduct Cl⁻ (1, 3, 8, 12) but may not traffic correctly to the cell surface (8, 34). Accordingly, therapy has focused on either replacing the defective CFTR protein in affected tissue (21, 38, 49) or resurrecting endogenous CFTR(ΔF508) function (14,26).

Conductive Cl⁻ secretion has been shown to be regulated by adenosine in a variety of epithelia (10, 24, 28, 33, 41). In a previous study, we demonstrated that nanomolar concentrations of the selective A_l-adenosine-receptor antagonist, 8-cyclopentyl-1,3-dipropylxanthine (CPX), could stimulate ³⁶Cl⁻ efflux from CF cells of diverse origin bearing the ΔF508 mutation (14), i.e., the pancreatic cell line CFPAC-1 and the lung cell line IB3. However, when these cells were corrected by gene transfer of wild-type CFTR, there no longer seemed to be a significant response to CPX (14). These data have lead us to focus further attention on the nature of the interaction between CPX and CF cells in anticipation that there might prove to be some therapeutic potential in this class of drugs. A1 receptors are linked to inhibition of CAMP production in other cell types (16,29), and examples also exist of receptor-dependent inhibition of CAMP production that is linked to activation of Na⁺/H⁺-exchange processes (22). Because these processes also occur independently of CAMP-mediated mechanisms (22), the first mechanisms we chose to investigate were the possible effects of CPX on pHi and intracellular Ca²⁺ concentration ([Ca²⁺]_i)

We report here that the selective A_l-receptor antagonists, CPX and 8-noradamantyl-1,3-dipropylxanthine (KW-3902), downregulate Na⁺/H⁺-exchange activity in pancreatic CFPAC-1 cells but have no effect on [Ca²⁺]_i. This effect is dependent on the presence of adenosine, since pretreatment of the cells with adenosine deaminase (ADA) mimics the effect of CPX. By contrast, the downregulation of Na⁺/H⁺ exchange by CPX was profoundly attenuated in CFTR-transfected CFPAC-1 cells. It is possible that the apparent parallels between CPX action on Cl⁻ efflux and Na⁺/H⁺ exchange in mutant and wild-type cells may lead to a better understanding of the basic defect in CF cells and in the mechanism of action of CPX on cells bearing the CFTR(ΔF508) mutation.

MATERIALS AND METHODS

Chemicals.

2’,7’-Bis-(Z-carboxyethyl)-5(6’)-carboxyfluorescein pentaacetoxymethyl ester (BCECF-AM) was obtained from Calbiochem (San Diego, CA); N-ethyl-N-isopropyl amiloride (EIPA) and fura 2 acetoxymethyl ester (fura 2-AM) were obtained from Molecular Probes (Eugene, OR). EIPA was added to experimental solutions from a 20 mM stock solution in dimethyl sulfoxide (DMSO). BCECF-AM and fura 2-AM were prepared as 2 mM stocks in DMSO and stored at −20°C. Bovine serum albumin (BSA) and nigercin were purchased from Sigma Chemical (St. Louis, MO). CPX was obtained from RBI (Boston, MA) and kept as a 10 mM stock in DMSO at −20°C. KW-3902 was synthesized as previously described (20).

Cell culture.

CFPAC-1 cells, as well as CFTR-transfected CFPAC-PLJ-4.6 and mock-transfected CFPAC-PLJ-6 cells, were obtained from Dr. R. Frizzell, University of Alabama at Birmingham, Birmingham, AL. IB3 (50) and IB3-S9 (13) were obtained from Dr. W. B. Guggino, Johns Hopkins University, Baltimore, MD. CFPAC-1 cells were grown in Iscove’s medium supplemented with 10% heat-inactivated fetal calf serum, 100 U/ml penicillin, 100 mglml streptomycin, 0.25 mg/ml Fungizone, and 1% wt/vol glutamine. IB3 and S9 cells were grown in LHC-8 medium supplemented with 10% fetal calf serum, 100 U/ml penicillin, 100 mg streptomycin, 0.25 mg/ml Fungizone, and 1% wt/vol glutamine. All culture media were obtained from Biofluids, Rockville, MD, and the osmolarity was 310 mosM.

Solutions.

The experiments were carried out in physiological salt solution (PSS) containing (in mM) 130 NaCl, 5.8 KCl, 1.8 CaC1₂, 0.8 MgSO₄, 0.73 NaH₂PO₄, 10 glucose, 20 N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid, and 2 glutamine, pH 7.4. In Na⁺-free PSS, NaCl was replaced with N-methyl-d-glucamine chloride and NaH₂PO₄ was replaced with KH₂PO₄.

pH_i and [Ca²⁺]_i measurements.

Cells were seeded at high density in poly-l-lysine-coated sample chambers consisting of 10 × 1o-mm cloning cylinders (Bellco Biotechnology, Vine-land, NJ) glued to no. 1 glass coverslips as previously described (45). After a 3-h incubation (5% CO₂, 37°C), the medium was exchanged to remove unattached cells. The various cell lines typically formed confluent monolayers within 3–4 days and were used for experiments l–3 days later. Confluent cell monolayers in sample chambers were loaded for 60 min at 25°C with either the pH-sensitive fluorescent dye BCECF by incubation with BCECF-AM (6 μM in PSS containing 0.01% BSA and 10 FM probenecid) or with the Ca²⁺-sensitive fluorescent dye fura 2 by incubation with fura 2-AM (2 μM in PSS containing 0.01% BSA), then washed twice and kept in PSS until use.

pH_i and [Ca²⁺]_i were monitored in single cells within a confluent monolayer at 25°C using an ARCM-MIC microfluorimeter (Spex Industries, Edison, NJ) coupled to a Nikon Diaphot microscope (Nikon, Tokyo, Japan) via a fiber-optic cable. A sample chamber prepared as described above was inserted into a Plexiglas holder designed to fit securely into the stage of the microscope. The chamber was continuously perfused at a rate of 1.5–2 ml/min so that solution changes could be effected quite rapidly. A vacuum line placed in the chamber maintained its volume at 300 μ1. The samples were viewed through a ×40 Nikon CF objective (numerical aperture 1.3). Fluorescence emission was monitored with a photomultiplier tube connected to the side port of the microscope. A pinhole aperture in front of the photomultiplier tube restricted its field of view to a single cell.

BCECF fluorescence was monitored in ratio mode by alternately exciting the specimen at 440 and 495 nm at l-s intervals. Emitted fluorescence from the area isolated by the pinhole was monitored at 530 nm. Calibration of the BCECF signal was performed using the method of Thomas et al. (43). Fura 2 fluorescence was monitored in ratio mode at excitation wavelengths of 340 and 380 nm and converted to [Ca²⁺]_i as previously described (19,45).

Acute acid loading of the cells was accomplished by the NH₄Cl prepulse technique (4). Briefly stated, cell monolayers were pretreated with 20 mM NH₄Cl in PSS, which was then exchanged for Na⁺-free PSS. On removal of NH₄Cl, the intracellular chemical equilibrium between ${\text{NH}}_4^{+}$ and NH₃ is displaced by the rapid diffusion of the permeant species NH₃ out of the cell, and the ensuing conversion of ${\text{NH}}_4^{+}$ to NH₃ results in an acute acid load. After the cells were kept 3–4 min in Na⁺-free PSS to establish the degree of acidification, the medium was changed to Na⁺-replete PSS. The initial rate of Na⁺-dependent pH_i recovery was determined by linear regression analysis of 10 points taken at 10-s intervals. To reduce cell-to-cell variability, recovery rates were normalized to those measured under control (no treatment) conditions in the same cell. In experiments where the rate of Na⁺-dependent pH_i recovery in CFPAC-1 cells was determined twice consecutively in the absence of experimental treatment, no significant difference was observed between the rates of the first and the second recoveries (0.150 ± 0.007 and 0.144 ± 0.012 pH units/min, respectively, n = 3).

Binding assays.

Adenosine-receptor binding studies were carried out on rat brain membranes as previously described (23)

Data analysis and presentation.

All values are presented as means ± SE. Significance was determined using a two-tailed t-test for paired or unpaired data. A value of P < 0.05 was assumed to indicate a statistically significant difference.

RESULTS

Effect of CPX on resting pH_i of CF-derived cells.

The effect of CPX on the resting pH_i of confluent monolayers of CF-derived cell lines bearing the ΔF508 genotype is illustrated in Fig. 1. Both tracheal (IB3) and pancreatic (CFPAC-1) cells were used for measuring pH_i after incubation with (or without) 50 nM CPX. The mean resting pH_i of confluent monolayers of CFPAC-1 and IB3 cells was 7.76 ± 0.02 (n = 4) and 7.42 ± 0.07 (n = 4), respectively, measured in nominally CO₂-free media 1 day after reaching confluence. The relatively high resting pH_i of CFPAC-1 cells is in good agreement with a previous report that used a different measuring technique (15). In our studies, however, a somewhat lower pH_i (~7.4) was observed in CFPAC-1 cells 2–3 days after confluence. In both cell lines, incubation with CPX resulted in a significant decrease (~ 0.2 pH units) in resting pH_i, suggesting that this A_l-receptor antagonist affects some mechanism involved in pH_i regulation. For wild-type CFTR-transfected CFPAC cells (CFPAC-PLJ-4.7), a resting pH_i of 7.64 ± 0.04 (n = 4) was found 1 day after attaining confluence.

Fig. 1.

Effect of 50 Nm 8-cyclopentyl-1,3-dipropylxanthine (CPX) on resting intracellular pH (pH_i) in IB3 and CFPAC-1 cells as measured in single cells. Single cells were examined within an intact, confluent monolayer grown on poly-l-lysine-coated glass coverslips. Cells were loaded with 2’,7’-bis-(2-carboxyethyl)-5(6’)-carboxyfluorescein (BCECF), and pH_i was measured by dual-excitation microfluorometry as described in MATERIALS AND METHODS in the absence (open bars) or presence (stippled bars) of 50 nM CPX. Each column represents mean ± SE of 4 experiments.

Na⁺/H⁺-exchange activity in CFPAC-1 cells.

To investigate the role of the Na⁺/H⁺ exchanger in pH_i homeostasis of the CFPAC-1 cells and in the above effect of CPX, we monitored the recovery of pH_i from an acute acid load. A typical experiment is shown in Fig. 2. The recovery of single CFPAC-1 cells from an acute acid load induced by NH₄Cl removal was monitored in the absence and presence of 50 nM CPX. When the extracellular medium was switched from PSS containing 20 mM NH₄Cl to Na⁺-free PSS, the cells rapidly acidified to a PH_i ~ 0.2 pH units below their resting value. Similar responses were seen in the presence and absence of CPX, and there was no indication of a significant component of pH_i recovery in the Na⁺-free medium in either case. However, when the extracellular medium was replaced with Na⁺-replete PSS, pH_i recovered toward resting levels, demonstrating the presence of Na⁺-dependent acid extrusion mechanism in these cells. The initial rate of pH_i recovery was 0.146 ± 0.040 pH units/min (mean ± SD, n = 8) 1 day after reaching confluence and 0.124 ± 0.033 pH units/min (n = 5) 3 days after confluence. These rates are not significantly different at the 95 and 90% levels. However, in the presence of 50 nM CPX, the recovery rate measured 1 day after reaching confluence was lowered to 0.078 ± 0.012 pH units/min (n = 6), indicating that CPX significantly (P < 0.001) blunts the Na⁺-dependent component of pH_i recovery in CFPAC-1 cells.

Fig. 2.

Representative traces depicting Na⁺ dependence of pH_i, recovery from an acid load in CFPAC-1 cells. An acid load was imposed on confluent monolayers of CFPAC-1 cells by exposure to physiological salt solution (PSS) containing 20 mM NH₄Cl for an interval of 3–4 min followed by a change to Na⁺-free medium. As shown, readdition of Nat-replete PSS was sufficient to initiate a pH, recovery. In 2nd part of experiment, 50 nM CPX was added to all solutions. Values of pH_i, converted from fluorescence ratio data using a calibration curve (see MATERIALS AND METHODS), are shown as continuous thin lines, whereas thick lines depict averages (7-point sliding window). Unsmoothed data were fit by linear regression in Na⁺-free (long dashes) and Na⁺-replete (short dashes) media as described in MATERIALS AND METHODS. Similar inhibitory effects of CPX were observed in all cells examined. Measurements were done 3 days (A) or 1 day (B) after confluence.

In HCO₃-free medium, the resting pHi of most mammalian cells is maintained by the Na⁺/H⁺ exchanger. Thus the observation that CPX depresses the resting pH_i (Fig. 1) and blunts the Na⁺-dependent pH_i recovery (Fig. 2) in CFPAC-1 cells strongly suggests that this agent is involved in the regulation of Na⁺/H⁺-exchange activity. In Fig. 3, we show that Na⁺-dependent pH_i recovery in CFPAC-1 cells is also markedly inhibited (~ 66%) by 10 μM EIPA, a relatively specific inhibitor of the Na⁺/H⁺ exchanger. These observations provide strong evidence that the Na⁺-dependent component of pH_i recovery modulated by CPX in these cells is due to Na⁺/H⁺ exchange.

Fig 3.

Effect of N-eathyl-N-isoproply amiloride(EIPA) on pH_i recovery from an acid load. CFPAC-1 monolayers were acidified by a NHdprepulse. Measurements of pHi recovery from acid load in a single cell were conducted on same cell, first in absence (open bar) and then in presence of 10 M EIPA (hatched bar). Rate of Na⁺-dependent pHi recovery in presence of EIPA was normalized to pH_i recovery rate observed in absence of EIPA. The EIPA column represents mean 2 SE of 3 experiments.

Effect of A₁ antagonists on Na⁺H⁺ exchange.

In Fig. 4, we show that the reduction of Na⁺/H⁺-exchange activity (Na⁺-dependent pH_i recovery) in CFPAC-1 cells produced by CPX is mimicked by another A₁-receptor antagonist, KW-3902 (39). The observation that these antagonists were active at concentrations in the nanomolar range suggests a direct action at an A₁ receptor. If this is the case, then we reasoned that CPX may be blocking the effect of endogenous adenosine on A₁ receptors. To test this hypothesis and to assess the possible physiological role of endogenous adenosine in pH_i regulation in CFPAC-1 cells, we preincubated cell monolayers with ADA, which converts adenosine to inosine, a compound that is inactive at extracellular adenosine receptors. Figure 5 shows that treatment of CFPAC-1 cells with ADA reduced the rate of Na⁺-dependent pH_i recovery obtained in control conditions to that observed in the presence of CPX in ADA-untreated cells. Moreover, CPX had no significant inhibitory effect on the rate of pH_i recovery in ADA-treated CFPAC-1 cells. These observations provide good evidence that endogenous adenosine has an upregulatory effect on Na⁺/H⁺-exchange activity mediated through a tonic action on A₁ receptors.

Fig. 4.

Effect of adenosine-receptor antagonists on Na⁺/H⁺ exchange in CFPAC-1 cells. Na⁺-dependent pH_i recoveries were measured on single cells in monolayers that had been preincubated for 10 min with either 50 nM CPX (n = 6 experiments) or with 50 nM 8-noradamantyl-1,3-dipropylxanthine (KW-3902; n = 5). Recovery rates are expressed as percentage of control (without drug preincubation). In both cases, observed inhibition was highly significant (P < 0.001). dpH_i/dt, change in pH_i over change in time.

Fig. 5.

Effect of adenosine deaminase (ADA) on response of CFPAC-1 cells to CPX. Recovery of pH_i was studied in control CFPAC-1 monolayers and in monolayers pretreated with 2 IU/ml ADA for 6 h and perfused during the experiment with 0.2 IU/ml ADA. Recovery rates were determined in absence (open bars) or presence (stippled bars) of 50 nM CPX as described in Fig. 2. In absence of CPX, pH_i recovery rates in ADA-pretreated cells were significantly (P < 0.02) reduced relative to rates measured in control cells.

Effect of CPX on Na⁺/H⁺ exchange in wild-type CFTR-transfected CFPAC cells.

We next examined the effect of correcting the CFTR defect on the ability of CPX to affect Na⁺/H⁺ exchange in CFPAC-1 cells. Table 1 summarizes the effects of CPX on Na⁺-dependent pHi recovery in CFPAC-1, in wild-type CFTR-transfected CFPAC-PLJ-4.7, and in mock-transfected CFPAC-PLJ-6 cells. We found that CPX had no significant effect on Na⁺-dependent pH_i recovery in wild-type CFTR-transfected CFPAC cells. By contrast, the CPX effect was evident in the mock-transfected CFPAC cells. Thus CPX action seems evident only in cells bearing the ΔF508 mutation and not the complete CFTR genotype. However, as is evident in Table 1, transfection of wild-type CFTR causes a reduction of the recovery rate in the absence of CPX, whereas little change occurs in the recovery rate in the presence of 50 nM CPX. In this regard, the effect of transfection with the normal gene is similar to pretreatment with ADA (see above).

Table 1.

Effect of wild-type CFTR transfection on intracellularpH (pH_i) recovery in CFPAC cells

	pHi Recovery Rate, pH units/min		Percent Inhibition, Matched Pairs
Cell Type	No CPX	50 nM CPX
CFPAC-1	0.144 ± 0.039 (16)	0.096 ± 0.020 (9)	38.5 ± 1.5 (9)
CFPAC-PLJ-6	0.207 ± 0.046 (8)	0.090 ± 0.025 (8)	54.4 ± 8.3 (8)
CFPAC-PLJ-4.7	0.097 ± 0.013 (9)	0.081 ± 0.013 (9)	10.9 ± 2.5 (9)

Recovery rates measured in presence or 50 nM 8-cycloen-tyl-1,3-dipropylxanthine (CPX) are given as means ± SE (n, number of experiments given in parentheses). Inhibition by 50 nM CPX of recovery rates in matched pairs (as % of control on same cell) is given in last column. Means of %inhibition values of CFPAC-1 and of CFPAC-PLJ-6 cells are each significantly different (at 0.005 level) from means of CFPAC-PLJ-4.7 cells, whereas they are not significantlv different from each other.

Effect of A₁-receptor antagonists on [Ca²⁺]_i.

In some cellular systems, the actions of A₁ receptors are known to be not only mediated by inhibition of adenylate cyclase but also linked to Ca²⁺ channels (11). Indeed, in various organs of Cftr(− / −) mice, there appears to be a strong correlation between the severity of disease and the relative level of Ca²⁺-dependent to cAMP-dependent Cl⁻ channels (7). We therefore examined the possibility that the effect of CPX could be mediated by changes in free [Ca²⁺]_i levels. [Ca²⁺]_i levels were measured in single cells using fura 2 fluorescence, and the CPX effect on [Ca²⁺]_i was analyzed in the CF cell lines CFPAC-1 and IB3 and in their respective wild-type CFTR-transfected clones CFPAC-PLJ-4.7 and IB3-S9. Addition of 50 nM CPX did not have any effect of its own on [Ca²⁺]_i levels (Fig. 6). Thus Ca²⁺ signaling is apparently not related to the effects of CPX on resting pH_i and Na⁺-dependent pH_i recovery, nor to those on Cl⁻ efflux from CFPAC-1 cells (14).

Fig. 6.

Representative Recording of CPX effect on intrecellular Ca²⁺ concentration ([Ca²⁺]_i) level in CF and in CFTR-transfected CF cells. [Ca²⁺]_i was monitored in single cells within intact, confluent monolayers of CFPAC-1 cells (A), CFTR-transfected CFPAC cells (B), IB3 cells (C), and CFTR-transfected IB3-S9 cells (D) as described in MATERIALS AND METHODS. Addition of 50 nM CPX had no effect on [Ca²⁺]_i in all cell types. Application of carbamylcholine (CCH) failed to cause a [Ca²⁺]_i increase in CFPAC-1 (A) and in CFTR-transfected CFPAC cells (B) but caused a large increase in [Ca²⁺]_i in IB3 (C) as well as in CFTR-transfected IB3-S9 cells (D). However, [Cali in both wild-type and CFTR-transfected CFPAC cells did increase in response to ionomycin (data not shown). Resting [Ca²⁺]_i were 37.1 ± 6.8 (n = 7) in CFPAC-1 cells, 47.4 ± 7.5 nM (n = 4) in CFTR-transfected CFPAC cells, 34.2 ± 1.7 (n = 5) in IB3 cells, and 32.5 ± 2.7 nM (n = 4) in CFTR-transfected IB3-S9 cells.

DISCUSSION

These data indicate that the selective A_l-receptor antagonist CPX reduces Na⁺/H⁺-exchange activity in both CFPAC-1 cells and IB3 cells and that this effect is substantially attenuated in CFTR-transfected CFPAC-1 cells. This phenomenon is remarkably similar to the selective action of CPX on ³⁶Cl efflux from the same cell types (14), leading us to conclude that a relationship may exist between the two processes. Because CPX had no significant effect on pH recovery in CFTR-transfected CF cells, we can exclude the possibility that CPX is simply acting as an inhibitor of Na ⁺/H ⁺ exchange. In addition, as shown in Fig. 6, the CPX effect is not mediated by changes in [Ca⁺²]_i. Furthermore, the CPX effect cannot merely be a consequence of elevated pH_i, since it is missing in confluent monolayers of wild-type CFTR-transfected CFPAC-PLJ-4.7 (pH_i ~ 7.6), whereas it is found in both l-day-old (pH_i 7.8) and 3-day-old (pH_i ~ 7.4) CFPAC-1 monolayers (see RESULTS). The loss of CPX action after transfection of wild-type CFTR is not related in any manner to the transfection process per se, since CPX is equally effective in mock-transfected CFPAC-PLJ-6 cells.

On the other hand, the data show that the effect of CPX on cells bearing the ΔF508 mutation is related in some way to adenosine receptors. First, the removal of endogenous adenosine by treatment of the monolayers with ADA mimics the effect of CPX and completely blunts any further lowering of the initial rate of pH_i recovery (Fig. 5). Second, KW-3902, a different specific antagonist of A_l receptors, is equally potent in inhibiting pH_i recovery (Fig. 4). Adenosine receptors are known to affect ion currents in epithelial cells, and A₁ receptors have been implicated in the regulation of Na⁺/H⁺ exchange in pig red blood cells (41). In CF airway epithelial cells, A₂ receptors were shown to be responsible for the stimulation of Cl⁻ secretion from normal but not from CF cells in the presence of 100 μM amiloride (28). Drugs in the class of amiloride, the classical inhibitor of Na⁺ channels, have been reported to express potent adenosine-receptor antagonist activity (18). For example, 5(N-methyl-N-isobutyl)-amiloride (MIBA), a close structural homologue of EIPA, has been shown to have an inhibition constant for A_l receptors of 160 nM (18). In the present work, we investigated adenosine-receptor binding of EIPA, using standard rat brain membrane binding assays (23), and found that it also had affinity for A₁ receptors, with K_i of 917 ± 261 nM. In addition, EIPA bound to rat striatal A_2A receptors with a K_i of 6.39 ± 0.97 μM. Thus the effect of EIPA may be dual, that of an A_l-receptor antagonist and that of its better known property as a Na⁺/H⁺-exchange inhibitor.

The fact that amiloride analogues do possess potent adenosine-receptor affinity may be of some relevance to the problem of CF, since amiloride itself, and in combination with ATP and UTP, has been the subject of clinical trials for the treatment of CF (27, 44). The affinity of amiloride and its analogues for A₁ receptors suggests that one might give consideration to the possibility that CPX and amiloride could share some common mechanisms of action on CF cells. However, we must temper this possibility by the fact that KW-3902 also inhibited Na⁺/H⁺-exchange activity in CFPAC-1 cells, whereas its efficacy in activating Cl⁻ efflux from CFPAC-1 cells was considerably less than that of CPX (20).

The data presented here may also have some relevance to previous suggestions that the CFTR mutation might be correlated with defective intracellular or intravesicular pH regulation (2, 15, 26, 32, 46, 48). In fact, the published data seem to be contradictory regarding the effects of mutations in CFTR on intracellular and intraorganellar pH. The observations that the change in pH across intracellular organelle membranes in CF nasal polyp cells is smaller compared with normal cells (2) is in conflict with the observation that the lysosomal acidification process is similar whether mutant or wild-type CFTR is present (46). Our data indicate that CFPAC-1 cells are uniquely susceptible to inhibition by CPX of Na⁺-dependent pH_i recovery from acute acid load. This recovery process is due to the Na⁺/H⁺ exchanger located specifically in the plasma membrane and not known to be associated with vesicular proton transport processes. These data thus do not support the concept of an intrinsic defect in vesicular pH regulation accompanying the CFTR defect in CF.

Our results do suggest that CPX may indeed function via a slight acidification of pH_i. However the results do not define the postreceptor mechanisms involved in the CPX-dependent inhibition of Na⁺/H⁺-exchange activity. Additionally, we do not know if the action of CPX on the Cl⁻ efflux may be correlated with only the variation of pH_i or whether some other intracellular mechanism activated by CPX could also be involved in this phenomenon.

In conclusion, it will be important therefore to define not only the cellular mechanism by which endogenous adenosine acts on pH_i but also the mechanism by which cellular acidification is linked to the alteration in Cl⁻ efflux that occurs following the selective A_l-antagonist action of CPX in CF cells. The fact that this effect is lost in CFPAC-1 cells repaired with wild-type CFTR would appear to indicate that the CFTR molecule may have consequences for ion distributions that go beyond Cl⁻ alone.