INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

ATP AND ITS METABOLITES ARE potent autocrine and paracrine agonists that modulate cellular responses through activation of purinergic receptors (3, 4, 10, 24, 31, 32). ATP and its metabolites are released by macrophages and mast cells in inflammatory responses, by platelets for self-aggregation, by dorsal root ganglia in neural networks, and by purinergic and autonomic presynaptic nerve terminals alone or together with other neurotransmitters (3, 4, 10, 24, 31, 32). As such, autocrine and paracrine nucleotide and nucleoside agonists act within a tissue for platelet aggregation, neurotransmission, pain perception, modulation of vascular tone, modulation of skeletal muscle and heart contractility, mast cell and immune cell activity, and cell volume regulation (3, 4, 10, 24, 31, 32). It is important to emphasize, however, that nucleotides are not blood-borne hormones because they are subject to profound degradation in the general circulation (10). Concerning modulation of cell volume, ATP, released by a conductive transport mechanism, was implicated recently as an essential autocrine regulator of cell volume in rat hepatoma cells (32). ATP release mechanisms have not been studied in epithelia grown as polarized monolayers.

ATP or purinergic receptors are expressed by cells to receive the extracellular ATP signal in these autocrine and paracrine regulatory events. At least six members of a family of G protein-coupled P2Y purinergic receptors and seven isoforms of a family of ATP-gated P2X Ca²⁺-permeable cation channel receptors have been identified (3, 4, 31). P2Y receptors couple to phospholipases through the pertussis toxin-sensitive G_i/G_o subclass or the pertussis toxin-insensitive G_q/G₁₁ subclass of heterotrimeric G proteins and trigger increases in intracellular Ca²⁺ and phospholipid signaling. P2X receptors depolarize the membrane voltage and increase intracellular Ca²⁺ by acting as an ATP receptor and a Ca²⁺-permeable, nonselective cation channel. Adenosine is a potent metabolite of ATP that has its own family of G protein-coupled adenosine receptors with numerous subtypes such as A₁, A₂, and A₃ (20). Once ATP is released by an epithelium, P2Y and P2X receptors expressed by the same epithelium bind that ATP and transduce the extracellular ATP signal.

Despite this knowledge, the cellular and molecular mechanisms whereby epithelial cells release ATP are not understood. Moreover, the physiological role of ATP release mechanisms in epithelial cells or extracellular nucleotide and nucleoside signaling in tissues lined by epithelial cells is understood poorly. As such, we tested the hypothesis in the form of a question: Do epithelial cells release ATP under basal and stimulated conditions and, if so, can we study ATP release from polarized epithelial monolayers? Moreover, we wanted to determine whether cystic fibrosis (CF) epithelial ATP release was compromised due to a lack of wild-type CF transmembrane conductance regulator (CFTR). To test these hypotheses, a highly sensitive bioluminescence detection assay for ATP released by polarized epithelial monolayers was developed. This assay is performed in a reduced serum medium to maximize cell viability. This assay was designed to study epithelial monolayers within a sealed chamber of a luminometer in real time, to minimize cell perturbation and maximize sensitivity. With this assay, we optimized the study of ATP release under basal and stimulated conditions in non-CF and CF epithelial cells grown as monolayers. A variety of epithelial cell primary and immortalized cultures derived from lung, gastrointestinal tissues, and kidney have also been studied with this assay. Results herein show that epithelial cells release ATP under basal conditions, that non-CF epithelial monolayers release ATP preferentially into the apical medium under basal conditions, and that hypotonicity triggers the release of ATP from epithelial cell monolayers across both apical and basolateral membranes in a reversible manner. Importantly, CF epithelial monolayers fail to release ATP across the apical membrane and fail to respond significantly to hypotonic challenge. Loss of extracellular ATP signaling in CF epithelia may contribute to the pathophysiology of CF.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Cell culture. All epithelial cell lines were grown on diluted Vitrogen (collagen types I and IV diluted 1:15 in Dulbecco's PBS; CelTrix, Santa Clara, CA)-coated 35-mm culture dishes and 25- or 75-cm² culture flasks (Falcon/Fisher, Suwanee, GA) or on filter supports (Millicell HA 0.45-µm, 12-mm-diameter insert, catalog no. PIHA01250, Millipore/Fisher, Bedford, MA) in MEM with Earle's salts and L-glutamine (catalog no. 10-010-CM, Cellgro-Mediatech, Herndon, VA) supplemented with 10% fetal bovine serum (FBS; certified, heat-inactivated FBS, GIBCO BRL, Grand Island, NY), 1× of 100 U/ml penicillin-100 µg/ml streptomycin solution (100× stock; GIBCO BRL), 1× or 2 mM L-glutamine (100× or 200 mM stock; GIBCO BRL), and 1-2 ml fungizone solution (GIBCO BRL). Airway epithelial primary cultures were purchased from Clonetics and were grown in a defined serum-free bronchial epithelial basal medium supplemented with bovine pituitary extract (2 ml), insulin (5 mg/ml), hydrocortisone (0.5 mg/ml), 1,000× gentamicin (0.5 ml), retinoic acid (0.1 µg/ml), transferrin (10 mg/ml), triiodothyronine (6.5 µg/ml), epinephrine (0.5 mg/ml), and human epidermal growth factor (0.5 µg/ml) diluted into 500 ml basal medium. Primary cultures were also grown on diluted Vitrogen. These cell lines and primary cultures form monolayers in air-fluid interface culture. When seeded, culture medium bathed both sides of the monolayers for 2 days to allow the cells to attach and grow. Thereafter, the monolayers were fed only on the basolateral side of the permeable filter support. When no leak was detectable from the basolateral to the apical side of the monolayer, the monolayers were studied. Routinely, monolayers were studied between days 8 and 12 in air-fluid interface culture.

Data analysis and statistics. Bioluminescence, in arbitrary light units (ALU), was recorded continuously with 15-s photon collection intervals in a laboratory notebook. Data were compiled into Microsoft Excel spreadsheets in which the mean ± SE was calculated for each time point in each set of experimental time courses (unless otherwise indicated). Data were then plotted in SigmaPlot for Windows using the same ALU values. Statistics were performed using SigmaStat for Windows. Paired Student's t-test and ANOVA were performed when appropriate; a P value of <0.05 was considered significant. Distributions of the ATP bioluminescence data from confluent cultures on dishes vs. apically directed release from polarized monolayers are also shown and were Gaussian in character.

Materials. Standard curves of ATP (Mg salt; Sigma) at known concentrations were performed with 2 mg/ml luciferase-luciferin reagent in Opti-MEM I medium by serial dilution from a 0.5 M ATP stock (made fresh at the time of performance of standard curves) to approximate the concentrations of ATP released from cells (see Fig. 3). Once a month, standard curves were performed to authenticate the detection reagent. The Sigma detection reagents were consistent from vial to vial. The luciferase-luciferin reagents lyophilized from Tris buffers (Sigma) were stable as stocks for several days and as dilutions in the detection medium for up to 12 h during 1 day of experimentation. Luciferase-luciferin reagent lyophilized from a glycine buffer had to be made up fresh each hour and was used as a stock for each day of experimentation, as it was not as stable over time. Measurements at each dose of ATP were performed in triplicate; luminescence values were stable among those three measurements (i.e., no "bleaching" or other instability in the signal was observed). Differences in efficacy of luciferase-luciferin reagents were observed among different commercial sources. The Promega reagents were tried initially, because the Turner luminometer was offered by Promega. However, their luciferase-luciferin reagent was already resuspended in a prelysis buffer for luc reporter gene assays and was not useful for these studies. The Sigma products were the most useful and were two- to threefold more efficacious in the ATP-catalyzed bioluminescence signal than a similar 2 mg/ml concentration of Calbiochem luciferase-luciferin reagent at a standard 1 µM MgATP concentration. Gadolinium chloride (GdCl₃) and apyrase were obtained from Sigma. GdCl₃ and the series of osmoles used in these studies do not affect significantly the ability of luciferase to detect ATP at this luminometer sensitivity (see Fig. 3). Other substances, like FBS and glibenclamide, a CFTR Cl channel inhibitor (Sigma), do have inhibitory effects on luciferase (see Fig. 3).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Bioluminescence detection assay of released ATP from cells: rationale for assay design and development of the assay. Figure 1 provides an illustration of how these measurements of released ATP were performed on epithelial cells grown to confluence in a culture dish and, more importantly, adapted for epithelial cells grown as a polarized monolayer. This assay was essential to test the hypothesis proposed above. Previous studies have used luminometers that require injection of the cells into an injection port, addition of cell suspensions in a cuvette held in a cuvette holder, or cells grown on a coverslip mounted in a cuvette with the aid of forcep manipulation (1, 2, 11, 12). There is potential for cell damage in each of these assays. As such, each of these studies differ from this assay in one important respect: the cells were studied on their substrates with no disruption or damage and without perturbation in our newly designed assay. This luminometer has a chamber that accommodates a platform that holds epithelial monolayers grown on 12-mm filters and epithelial or other cell cultures grown to confluence in 35-mm dishes. In other studies, cell coverslips were handled with forceps during removal from culture and mounting in a cuvette (11, 12). Other studies used trypsinized cells (1, 2, 11, 12). As such, the design of this assay represents a significant advance in the study of ATP release mechanisms and extracellular ATP signaling.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Schematic of ATP release assays performed in a Turner TD20/20 luminometer illustrates how ATP bioluminescence release and detection assays are performed. This is a simplified model of Turner TD20/20 luminometer without injection ports or other attachments. This luminometer was designed originally to perform photon collection from lysates of cells transfected transiently with mammalian expression vectors containing luc reporter genes. Its design was adapted by our laboratory to perform detection assays of cellular ATP release from epithelial cell cultures as depicted here. See text for details.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Cell model illustrates relative contributions of ATP release mechanisms, ATP consumption by luciferase enzyme, and degradation of released ATP by ecto-ATPases and related enzymes. Although depicted as only apical release in this model, ATP release is also observed across basolateral membrane, albeit at lesser magnitudes (see Figs. 4-7). See text for details. Ppi, phosphorylated Pi; H, amount of light produced.

Bioluminescence detection of ATP released from epithelial cells grown to confluence on culture dishes. To test the hypothesis proposed above, an assay was developed to detect ATP released from epithelial cells grown to confluence in 35-mm culture dishes. This assay utilizes a lyophilized luciferase-luciferin reagent resuspended in an Opti-MEM I reduced serum medium (GIBCO BRL) to detect ATP released from the cells and to keep cells viable. This is the same medium recommended for transfection of mammalian cells with cationic lipids; cells are incubated in this medium for as long as 24 h with no loss of viability. Gentle washes with PBS remove serum-containing medium. Serum inhibits the luminescence reaction; thus serum-containing medium cannot be used as a vehicle for this assay. Figure 3A shows that some inhibition of the luciferase enzyme is apparent with 5 and 10% serum added to the Opti-MEM I reduced serum medium, whereas the luminescence signal at different ATP concentrations was not different in PBS vs. Opti-MEM I medium. Thus, because there was no difference in ATP detection in the Opti-MEM I medium but an enhancement in cell viability in a medium vs. a Ringer solution, Opti-MEM I was used as the medium or vehicle in these experiments. The contents of the Life Technologies Opti-MEM I medium are proprietary; however, a personal communication with a Life Technologies representative revealed that no serum is present in the medium but that some specialized factors and salts are supplemented into the Opti-MEM I medium to allow the serum percentage to be reduced for a given cell without affecting growth rate or viability. Luciferase detection of ATP is not affected by 100 µM GdCl₃ (used in subsequent experiments) but is inhibited by glibenclamide, a CFTR inhibitor (Fig. 3B). Figure 3C shows that luciferase consumption of ATP is not affected by a reduction or an increase in NaCl concentration. Other salts or inert osmoles also had no significant effect on luciferase detection of ATP (data not shown). Taken together, these results validate the use of the Opti-MEM I medium as a medium for this assay to maximize cell viability without compromising ATP detection. Moreover, glibenclamide may be troublesome to utilize for the study of CFTR regulation of ATP release and was avoided in this study.

View larger version (24K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   View larger version (22K): [in this window] [in a new window]   Fig. 3.   Detection of ATP by luciferase is optimal in Opti-MEM I medium. A: standard curves of ATP detection by luciferase in Opti-MEM I, PBS, and Opti-MEM I supplemented with 5 and 10% fetal bovine serum (FBS). These results show that serum inhibits ability of luciferase to detect ATP (representative of 3 different standard curves). B: similar curves showing that glibenclamide inhibits luciferase, whereas GdCl3 does not (representative of 3 different standard curves). C: reduction or increase in NaCl concentrations does not affect luciferase detection of ATP (representative of 3 different standard curves). Values for all standard curves (luminescence in arbitrary light units) and for all ATP concentrations ([ATP]) tested (only a subset are plotted) are as follows (with n given in parentheses): no ATP, 0.075 ± 0.008 (30); 1015 M, 0.513 ± 0.268 (5); 1014 M, 0.782 ± 0.352 (5); 1013 M, 1.002 ± 0.342 (5); 1012 M, 1.746 ± 0.410 (15); 1011 M, 4.135 ± 0.999 (15); 1010 M, 8.150 ± 1.855 (15); 109 M, 20.29 ± 4.169 (15); 108 M, 54.72 ± 11.41 (15); 107 M, 302.9 ± 67.25 (15); 106 M, 2,184 ± 417.1 (15); 105 M, 7,259 ± 805.4 (15); 104 M, 9,423 ± 318.7 (15); 103 M, >9,999 (15).

Bioluminescence detection of ATP released from epithelial cells grown as polarized monolayers: development of the assay. Epithelial cells were seeded at high density (at least 10⁵ cells per 12-mm filter) onto collagen-coated permeable supports (Millicell 12-mm-diameter, permeable filter cups). The epithelial monolayers were grown with medium bathing the apical and basolateral sides of the filter support for 2 days. After that initial period, the apical side was devoid of medium, whereas the basolateral side was fed daily ("air-fluid interface" culture). Monolayers were grown in this manner until no fluid leaked from the basolateral space into the apical side or filter cup. This reflected a transepithelial resistance of 200  · cm² (after subtracting the resistance of the filter itself), as measured with an EVOM meter (World Precision Instruments, Sarasota, FL) and a monolayer tight to fluid for at least 24 h. Cells were then fed with fresh medium on both sides of the monolayer and incubated overnight for subsequent experimentation the following day. This also served to wash away any cells that had been shed by the monolayer. Monolayers were washed two times in PBS on both sides, 200 µl of Opti-MEM I medium containing 2 mg/ml luciferase-luciferin reagent (Sigma) were added to the cells on one side of the monolayer (i.e., the side on which ATP is detected), and 200 µl of Opti-MEM I medium without detection reagent were added on the contralateral side. [The filter was placed on the lid of a 35-mm culture dish in a 200-µl drop of medium lacking detection reagent to bathe the basolateral side, and an equal volume of medium containing detection reagent was added into the filter cup on the apical side for measurement of apically directed ATP release, and vice versa for measurement of basolaterally directed ATP release (see Fig. 1)]. The filter was placed on a platform, lowered into a chamber in complete darkness within a simplified model Turner TD20/20 luminometer, and studied immediately within the luminometer in real time (Fig. 1). Time courses measuring the ATP in the medium over time were done over the course of 10-15 min in continuous 15-s photon collection intervals.

View larger version (32K): [in this window] [in a new window]   View larger version (26K): [in this window] [in a new window]   View larger version (30K): [in this window] [in a new window]   View larger version (29K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   View larger version (21K): [in this window] [in a new window]   Fig. 4.   ATP is released across apical and basolateral membranes of epithelial monolayers under basal conditions. Distributions of peak ATP bioluminescence values under basal conditions for all experiments performed on non-cystic fibrosis (CF) epithelial monolayers grown on collagen-coated filter supports measuring apical release (A), on CF epithelial monolayers grown on collagen-coated filter supports measuring apical release (B), on non-CF epithelial monolayers measuring basolateral release (C), and on CF epithelial monolayers measuring basolateral release (D). NHAE, non-CF human airway epithelia. Mean ± SE are shown by filled circle and error bars to left of data and are also given in text with number of experiments. Lowest and highest luminescence values are also given to indicate range of data. E: summarized raw luminescence data in arbitrary light units (ALU) are plotted on y-axis vs. time (in minutes) on x-axis. Sustained apically directed ATP release (n = 10) was markedly greater than basolaterally directed ATP release (n = 8). GdCl3 (100 µM) inhibited ~50% of apically directed ATP release and 100% of basolaterally directed ATP release. Apyrase abolished remaining luminescence signal not inhibited by GdCl3. Apically directed ATP release was significantly greater than basolateral release at all time points (P < 0.05 by ANOVA with a Bonferroni ad hoc test). APM, apical membrane; BLM, basolateral membrane. F: data are plotted as %luminescence (100% being luminescence value of medium spiked with 10 µM ATP before exposure to cells) compared with aliquots of medium removed after exposure to cells. Cells degrade ATP fully at 8 h following exposure (n = 4; luminescence of each extracted medium sample measured in triplicate).

Development of an ATP degradation assay. Because specific inhibitors of ecto-ATPases and ecto-apyrases that do not also affect the luciferase enzyme are not available, ATP degradation assays were performed in parallel to ATP release assays as follows. Cells were grown to confluence in T25 flasks (4-8 flasks for each cell type). ATP (10 µM) was supplemented in the culture medium for the 16HBE14o cells, CFBE41o cells, and CFPAC-1 cells. Medium not exposed to cells was analyzed by bioluminescence to record a standard bioluminescence value for 10 µM MgATP in the medium in the absence of cells. The ATP-containing medium was then added to the confluent cultures, and aliquots of the medium were removed from the flask and analyzed by bioluminescence over time at 15 min (or the average time required to perform an ATP bioluminescence detection assay of released ATP), 30 min, and 1, 2, 4, 8, and 24 h. Luminometry measurements of the medium aliquots were performed in triplicate by mixing them with an equal volume of medium containing luciferase-luciferin reagent in the absence of cells. Data were normalized to bioluminescence of the "spiked" medium containing 10 µM MgATP that was never exposed to cells (100% value as time 0).

Degradation of ATP by ecto-ATPases dampens the bioluminescence detection of released ATP to a small but significant extent. Not only can epithelial cells release ATP but they also may have the capacity to degrade ATP by expression or secretion of ecto-ATPases or ecto-apyrases. In studies that paralleled ATP release assays, a known quantity of ATP (10 µM) was added to the culture medium, and the disappearance of ATP was measured over time in an ATP degradation assay. The results of a degradation assay performed on 16HBE14o cells, CFBE41o cells, and CFPAC-1 cells is shown in Fig. 4F. Approximately 10-20% of the ATP added to the culture medium is degraded over a 15-min incubation period by epithelial cell cultures, whereas 40-60% of the ATP is gone at 1 h (38.3 ± 3.0%; n = 4). Virtually all of the ATP, however, is gone at 8 h (3.8 ± 4.1%; n = 4) in both non-CF and CF cultures. These results show that, although ATP is released by epithelial cells, it is also subject to degradation, albeit at a much slower rate. The rates of ATP consumption by the luciferase enzyme (milliseconds; due to molar excess of the enzyme) and ATP release (seconds), however, are much greater than the rate of degradation (minutes to hours). The molar excess of the detection reagent ensures that most of the released ATP will be consumed immediately and detected as photons or bioluminescence; however, this assay design does not rule out that degradation of ATP by ecto-ATPases may dampen the signal, especially in cells that express high amounts of ecto-ATPases and ecto-apyrases. Of interest, the magnitude of ATP degradation by CF epithelia was much less than that of non-CF epithelia over the 1- to 4-h period of the assay.

Hypotonicity stimulates apically and basolaterally directed ATP release from 16HBE14o monolayers. Studies in rat hepatoma cells have shown recently that conductive release of ATP under hypotonic conditions is essential for control of regulatory volume decrease (RVD) during cell volume regulation (32). ATP release was measured as a conductive transport event in this study by patch-clamp recording (32). We tested the hypothesis that hypotonicity may be an important stimulus for ATP release in epithelial cells and that this newly developed bioluminescence detection assay may detect hypotonicity-induced ATP release by a new and different method. Figure 5A shows the effect of an addition of isotonic medium (as a control) followed by a 33% dilution of the medium volume with distilled water. Opti-MEM I reduced serum medium osmolality was 298 ± 4 mosmol/l. Dilutions with water of the volumes indicated was also analyzed by osmometry to determine the actual percent dilution of the medium osmolality (data not shown). Both the medium and the water had a similar concentration of luciferaseluciferin reagent (2 mg/ml) in this and all subsequent experiments. In either medium or water additions, similar manipulation of the preparation was performed. The luminometer chamber was opened, the platform supporting the filter was raised, the aliquot of medium or water was added by pipettor, the platform was lowered into the chamber, and the luminometer chamber was closed. Thus mechanical control of the experiments is also performed in these studies. Addition of medium had no significant effect on the luminescence of the monolayer preparation when added to either the apical or the basolateral medium (Fig. 5A). Addition of distilled water to dilute the osmotic strength of the medium augmented ATP release into the apical medium fourfold and ATP release into the basolateral medium threefold (Fig. 5A). Addition of apyrase abolished the signal completely, illustrating that hypotonicity-induced release of ATP by the epithelium increased the bioluminescence signal. These results suggest that hypotonicity is a profound stimulus for ATP release.

View larger version (20K): [in this window] [in a new window]   View larger version (28K): [in this window] [in a new window]   View larger version (31K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   View larger version (27K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Fig. 5.   ATP release by non-CF 16HBE14o epithelial monolayers into both apical and basolateral media is potentiated markedly by hypotonicity. A: apically and basolaterally directed ATP release measured under basal conditions (as in Fig. 3) for 2 min, followed by addition of isotonic medium (100 µl; 33% dilution of 200 µl of isotonic medium already present), and then followed by addition of distilled water (150 µl; 33% dilution of 300 µl medium volume) that stimulated ATP release across either apical (n = 6) or basolateral (n = 4) membrane. Hypotonicity-induced ATP release across apical and basolateral membranes was significant at all time points (P < 0.05 by paired Student's t-test vs. medium controls and unstimulated magnitudes). B: similar experiments showing increasingly severe dilutions of medium osmolality and their effects on apically (top) and basolaterally (bottom) directed ATP release (n = 6 and 5, respectively). Corresponding isotonic medium controls are also shown. C and D: effects of GdCl3 (100 µM) are shown on apically (C) and basolaterally (D) directed ATP release (n = 4 each; significant inhibition, P < 0.05 for each data set). HYPO, hypotonicity. E and F: representative raw luminescence data vs. time showing that only dilution of osmolality on side of epithelium where ATP release is being detected stimulates ATP release profoundly. Luciferase-luciferin reagent was present only in apical (E) or only in basolateral (F) medium.

View larger version (21K): [in this window] [in a new window]   View larger version (22K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   View larger version (30K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   Fig. 6.   Mild dilutions of medium osmolality trigger ATP release into apical medium in a reversible manner in non-CF epithelial monolayers. Apically directed ATP release was stimulated by "mild," "normal," and "severe" dilutions of medium osmolality. A: mild dilutions of medium osmolality (4, 13, and 22% dilutions; n = 4). B: normal dilutions of medium osmolality (24, 33, and 41% dilutions; n = 4). C: severe dilutions of medium osmolality (37 and 50% dilutions; n = 4). Similar additions of isotonic media at same volumes are also shown. All dilutions caused a significant increase in bioluminescence or ATP release (P < 0.05 by paired Student's t-test) vs. medium controls and unstimulated magnitudes. Addition of 50 mM NaCl boluses reversed responses completely to unstimulated levels in a similarly significant manner (P < 0.05 by paired Student's t-test vs. peak stimulated value). D: summary of NHAE monolayer data testing reversal of hypotonicity-induced ATP release into apical medium with a panel of different osmoles. Results of inhibition with 100 µM GdCl3 are compared with 100 mM readdition of salts or 200 mM readdition of sucrose or mannitol following 50% dilution of medium osmolality (~300 mosmol/l before addition of distilled water); n = 4-8. Open bars show significantly less or no ability to reverse response. * P < 0.05 for reversal of ATP release response by paired Student's t-test. E and F: representative raw luminescence data showing effect or lack of effect in reversing hypotonicity-induced ATP release of different osmoles, including NaCl vs. sucrose and mannitol (E, top), NaCl vs. N-methyl-D-glucamine chloride (NMDGCl) and sodium gluconate (NaGlu) (E, bottom), and NaCl vs. NaI (F).

Sensitivity to hypotonicity is lost and the magnitude of hypotonicity-induced ATP release is attenuated markedly in CF epithelia. To follow the studies of isotonic or basal ATP release showing a loss of apically directed ATP release in CF epithelia, we examined the relative response to hypotonic challenge in non-CF vs. CF epithelia. These results are shown in Fig. 7. Whereas non-CF epithelia respond to a hypotonic challenge robustly, CF epithelia fail to respond to hypotonic challenge. Figure 7, A and B, compares hypotonicity-induced ATP release across both the apical and basolateral membranes of non-CF and CF monolayers on the same luminescence scale. The difference is more dramatic for apically directed ATP release; however, basolaterally directed ATP release in response to hypotonic challenge is also attenuated somewhat. The loss of sensitivity to a mild hypotonic challenge in CF epithelia vs. non-CF epithelia is shown with the summarized data in Fig. 7C. With 13% dilution of the medium osmolality, no significant increase in ATP release is evident into the apical medium of CF monolayers. In contrast, non-CF 16HBE14o monolayers release ATP robustly. When a more severe 41% dilution is performed, only small increases in ATP release across the apical membrane are observed in CF epithelia (<5 ALU), whereas an increase of ~45 ALU is seen in non-CF epithelia. Taken together, these results show that, in CF epithelia lacking functional CFTR, extracellular ATP signaling under isotonic and hypotonic conditions is attenuated markedly. These results suggest that CFTR itself may play a role in sensing/transducing changes in external osmolality or reduction in external anion concentration, an ability that may be lost in CF cells. Moreover, because CF epithelia do release ATP (albeit when exposed to more severe hypotonic challenge), these results suggest that the ATP release mechanism is expressed by CF epithelia and may be an entity separate from but regulated by CFTR.

View larger version (16K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Fig. 7.   CF epithelia fail to release ATP in response to a hypotonic challenge. Apically and basolaterally directed ATP release in non-CF 16HBE14o (A) and CF CFBE41o (B) epithelial monolayers measured under basal conditions for 2 min and after addition of distilled water (successive 67-µl additions for 24, 33, and 41% dilution of medium osmolality). C and D: summarized data showing significant response to dilution of medium osmolality (* P < 0.005 by paired Student's t-test) in non-CF 16HBE14o monolayers [wild-type CF transmembrane conductance regulator (CFTR)] vs. CF monolayers (F508 CFTR) for both a mild (13%; C) and a severe (41%; D) dilution.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Strengths and weaknesses of the bioluminescence detection assays of released ATP from epithelial cells. To test the hypothesis that epithelial cells release ATP to support essential physiological processes that occur in tissues lined by epithelial cells, a highly sensitive assay was developed to measure ATP released from epithelial monolayers adherent to a filter support inside a luminometer. An illustration of the design and use of the preparations examining ATP release from nonpolarized confluent cell cultures and polarized epithelial monolayers are shown in Fig. 1. Establishment of this assay was the primary goal of this study and this paper, and this is the first description of such an assay. The strengths of this assay are that no endogenous ATP is present in the medium containing the luciferase-luciferin reagent, nanomolar quantities of ATP can be measured in the system with no background, and adherent and viable cells growing on collagen-coated substrates can be studied in this luminometer (fitted with a platform for dishes or dish lids holding filters) in a medium rather than a Ringer or saline solution. This assay was adapted into a microassay with small filter supports (see Fig. 1) to enhance productivity, to reduce variability, and to study the "sidedness" of ATP release. This assay can also be done on a single Xenopus oocyte injected with wild-type and mutant forms of CFTR (Q. Jiang, E. M. Schwiebert, W. B. Guggino, J. K. Foskett, and J. F. Engelhardt, unpublished observations), provided that the sensitivity of the Turner luminometer is maximized to 100%. The sensitivity of the Turner luminometer for our studies was set at 40%. The disadvantages are that this assay is limited to cells in culture and is a noncirculating system. However, it is possible to apply the technology of this assay to other preparations. Adaptations of the luminometer would permit an interface with an Ussing chamber system or a perfused tissue preparation to perform luminometry in a circulating environment and measure transepithelial ion and fluid transport simultaneously. Applications for this assay concerning the study of epithelial cells derived from specific tissues are outlined below (see Physiological and pathophysiological applications and epithelial paradigms for the study of extracellular ATP signaling).

Comparison with other bioluminescence measurements of ATP release. Other ATP bioluminometry studies have utilized cells in suspension in a cuvette, cells in suspension injected through a syringe port in the luminometer (not unlike that of an HPLC), cells grown on a coverslip but manipulated and mounted with forceps in a cuvette vertically and at an angle within a luminometer, and samples derived from medium conditioned by cells and injected into the luminometer (1, 2, 11, 12, 23). In all cases, cells were manipulated or potentially disturbed or damaged in such assays. Some loss of the ATP may occur during the transfer and processing of medium samples conditioned with ATP. Other types of luminometers were also used for these studies that were limited to injection ports or cuvette holders.

Factors that influence detection of released ATP in this assay. The detection of ATP released by epithelial cell cultures and monolayers is governed by at least three major processes: cellular release mechanisms in epithelia, consumption by the luciferin-luciferase reaction, and ecto-ATPases and ecto-apyrases expressed by or secreted by epithelia. ATP release occurs under basal conditions and, in some cases, at substantial magnitudes, as measured by the bioluminescence detection assays. Bioluminescence is created by immediate consumption of the luciferase enzyme, present in the medium in molar excess. This rate or magnitude of ATP release (measured as luminescence) is increased rapidly by hypotonicity and reversed rapidly by GdCl₃ or readdition of osmoles to reestablish isotonic conditions. The explanation for rapid increases and decreases in bioluminescence is that the ATP released is consumed immediately by the luciferase enzyme on a millisecond time scale. Bioluminescence, however, is being measured in continuous, 15-s collection intervals by the luminometer. Thus bioluminescence detection of ATP is equivalent to the consumption of ATP by luciferase: one photon collected for every molecule of ATP. ATP is also consumed in this system by ecto-ATPases and ecto-apyrases that compete with the luciferase enzyme for the ATP substrate. Over the 15 min required for the ATP release assays, ~10-20% of the ATP present in the medium is consumed by these competing enzymes expressed or secreted by the epithelial cells themselves, as measured in parallel but separate ATP degradation assays. Thus the amounts of ATP being released and being detected, through consumption by the luciferase enzyme, are underestimated slightly in this assay due to the activity of these degradative enzymes. Specific inhibitors of these enzymes that do not also affect the luciferase enzyme itself are lacking; thus simultaneous assessment of ATP release and degradation is not possible at this time.

Physiological roles of extracellular ATP signaling. Why would a cell want to release its ATP? In nonepithelial cell systems, ATP release is essential for platelet self-aggregation and pain perception via neurotransmission in dorsal root ganglia (10). ATP is released by nerves innervating urinary bladder to cause P2X receptor-mediated stimulation and contraction of urinary bladder smooth muscle (30). ATP is concentrated to millimolar or higher amounts in chromaffin granules with epinephrine, secretory granules in mast cells with histamine, and in presynaptic vesicles of the autonomic nervous system with norepinephrine and acetylcholine (3, 4, 10, 31). Interestingly, ATP and ADP is released by platelets, and ATP is released by dorsal root ganglia and presynaptic nerve terminals by exocytic mechanisms. Preliminary data suggest that Ca²⁺ agonists promote apically directed but not basolaterally directed ATP release from epithelial monolayers derived from multiple tissues (E. M. Schwiebert, unpublished observations). In epithelial cell systems, exogenous ATP has been shown to modulate vasopressin regulation of water channels in renal collecting duct, stimulate Cl and fluid secretion in airway and gastrointestinal epithelia, and inhibit Na⁺ absorption in nasal and renal epithelial cells (7, 14, 15, 17, 19, 26, 28). Until now, however, the sources of ATP important for releasing this agonist to exert these effects has not been studied. We propose that the epithelium itself elaborates this ATP to modulate transepithelial ion and water transport by multiple mechanisms.

Extracellular ATP signaling has importance in autocrine control of epithelial cell volume. Recently, Fitz, Roman, and colleagues (32) showed that conductive release of ATP was essential for cell volume regulation in rat hepatoma cells. Hypotonic cell swelling induced a significant ATP whole cell conductance that was shown to be vital for subsequent regulatory volume decrease (RVD) and the opening of swelling-activated Cl channels involved in RVD (32). ATP scavengers and purinergic receptor antagonists attenuated RVD (32). Therefore, cells release ATP for highly specialized physiological processes that occur within tissues to self-regulate their own function or the function of neighboring cells and to respond to changes in their environment such as a decrease in osmolality at the external surface of their plasma membranes. Once released, ATP, as an extracellular agonist, exerts its effects on cell function through its purinergic receptors. Data provided by this study agree with those described by Fitz and co-workers (32) and show that hypotonicity promotes ATP release across both the apical and basolateral membrane domains of epithelial cells. This study extends that work, detecting ATP with a different assay, and provides a physiological role for ATP release. A biological process such as RVD during cell volume regulation is an ideal setting for ATP release cell biology.

Mechanisms of ATP release in epithelia. At least three putative mechanisms may underlie isotonic (basal or unstimulated) ATP release as well as hypotonicity-induced ATP release in 16HBE14o monolayers. An ATP channel regulated or gated by the ATP-binding cassette transporter, CFTR, has been observed by many laboratories (2, 21, 23, 26, 29). CFTR mRNA and protein is expressed abundantly by the 16HBE14o cells, Calu-3 cells, and NHAE primary cultures (L. M. Schwiebert and E. M. Schwiebert, unpublished observations), and CFTR Cl channels have been observed routinely in the cell lines (Ref. 13 and K. L. Marrs and E. M. Schwiebert, unpublished observations). Conductive transport of ATP was also measured in response to hypotonicity in rat hepatoma cells (32). Thus conductive transport of ATP, down a highly favorable concentration gradient (>100,000-fold; 3-5 mM ATP is present in the cytosol, whereas nanomolar or lower amounts are present outside of cells), out of the cell may account for these responses. In support of a conductive transport mechanism, the broad-specificity mechanosensitive ion channel blocker GdCl₃ inhibits at least a portion of isotonic and hypotonicity-induced ATP release.

Physiological and pathophysiological applications and epithelial paradigms for the study of extracellular ATP signaling. Extracellular ATP signaling may be defective or detrimental in diseases that affect epithelial cell function. Loss of extracellular nucleotide signaling due to a lack of the ATP channel regulator CFTR may cause defective regulation of multiple ionic conductances and an abnormal ionic and osmotic composition of the airway surface fluid that covers the apical surface of CF epithelia. Our data show definitively that apically directed ATP release under basal conditions is lost in CF epithelia compared with non-CF epithelia. Moreover, hypotonicity-induced ATP release is also attenuated greatly in CF epithelia. Because extracellular ATP stimulates Cl and fluid secretion and inhibits Na⁺ absorption in epithelia (7, 14, 15, 17, 19, 26, 28), a loss of CFTR and a subsequent loss of ATP release and extracellular ATP signaling may underlie the lack of Cl transport and fluid secretion coupled with the heightened Na⁺ absorption that are hallmarks of CF transport pathophysiology.