The Role of P2Y1 Purinergic Receptors and Cytosolic Ca2+ in Hypotonically Activated Osmolyte Efflux from a Rat Hepatoma Cell Line*

Exposure of HTC rat hepatoma cells to a 33% decrease in extracellular osmolality caused the cytosolic Ca2+ concentration ([Ca2+] i ) to increase transiently by ∼90 nm. This rise in [Ca2+] i was inhibited strongly by apyrase, grade VII (which has a low ATP/ADPase ratio) but not by apyrase grade VI (which has a high ATP/ADPase ratio) or hexokinase, indicating that extracellular ADP and/or ATP play a role in the [Ca2+] i increase. The hypotonically induced rise in [Ca2+] i was prevented by the prior discharge of the intracellular Ca2+ store of the cells by thapsigargin. Removal of extracellular Ca2+ or inhibition of Ca2+ influx by 1–10 μm Gd3+depleted the thapsigargin-sensitive Ca2+ stores and thereby diminished the rise in [Ca2+] i . The hypotonically induced rise in [Ca2+] i was prevented by adenosine 2′-phosphate-5′-phosphate (A2P5P) and pyridoxyl-5′-phosphate-6-azophenyl-2′,4′-disulfonate, inhibitors of purinergic P2Y1 receptors for which ADP is a major agonist. Both inhibitors also blocked the rise in [Ca2+] i elicited by addition of ADP to cells in isotonic medium, whereas A2P5P had no effect on the rise in [Ca2+] i elicited by the addition of the P2Y2 and P2Y4 receptor agonist, UTP. HTC cells were shown to express mRNA encoding for rat P2Y1, P2Y2, and P2Y6 receptors. Inhibition of the hypotonically induced rise in [Ca2+] i blocked hypotonically induced K+ (86Rb+) efflux, modulated the hypotonically induced efflux of taurine, but had no significant effect on Cl− (125I−) efflux. The interaction of extracellular ATP and/or ADP with P2Y1purinergic receptors therefore plays a role in the response of HTC cells to osmotic swelling but does not account for activation of all the efflux pathways involved in the volume-regulatory response.

Cells swollen by a decrease in the osmolality of the external medium return to their original volume by a process known as regulatory volume decrease (RVD). 1 The signaling pathways activated during RVD have not yet been elucidated fully but result in the efflux of the inorganic ions, K ϩ and Cl Ϫ , and organic molecules, with a concomitant efflux of water and a decrease in cell volume. Transient increases in cytoplasmic Ca 2ϩ concentration ([Ca 2ϩ ] i ) have been observed in response to osmotic swelling in many, but not all, cell types (1)(2)(3) and may contribute to RVD by direct activation or modulation of K ϩ , Cl Ϫ , or organic osmolyte channels and/or by activating signaling pathways involved in RVD.
Recent reports (4 -16) have demonstrated that cells respond to mechanical stimuli, including osmotic swelling, by releasing ATP. ATP is known to bind to two main classes of purinergic receptors present on the cell plasma membrane; P2Y receptors are G-protein-linked receptors, and P2X receptors are intrinsic ion channels (17). Activation of either class of receptor can result in a transient increase in [Ca 2ϩ ] i . In a number of cell types apyrase (an ATP/ADP phosphatase) and P2 receptor antagonists such as suramin and PPADS have been shown to inhibit, at least partially, swelling and mechanically stimulated cytoplasmic Ca 2ϩ signals (5,13,15,16), thus providing evidence for a link between extracellular ATP release and transient increases in [Ca 2ϩ ] i .
In a study of HTC rat hepatoma cells, Fitz and co-workers (4,6) proposed that extracellular ATP, released in response to a decrease in the extracellular osmolality, could act in an autocrine manner by binding to purinergic receptors and thereby stimulating signaling pathways responsible for activating the Cl Ϫ efflux required for RVD. A recent report (18) demonstrated a rise in [Ca 2ϩ ] i in HTC cells in response to osmotic swelling and showed that depletion of thapsigargin-sensitive intracellular stores of Ca 2ϩ inhibited both swelling-activated K ϩ and Cl Ϫ currents. In contrast to the findings in other cell types (13,15,16), however, the rise in [Ca 2ϩ ] i was not inhibited by either apyrase or inhibitors of purinergic receptors, from which it was concluded that the swelling-induced release of ATP was not responsible for the rise in [Ca 2ϩ ] i (18).
In this study we have investigated the relationship between a decrease in extracellular osmolality, extracellular nucleotides, [Ca 2ϩ ] i , and ion and organic osmolyte efflux in HTC cells. In experiments comparing the actions of three enzyme preparations that hydrolyze ATP, we found that the hypotonically induced rise in [Ca 2ϩ ] i was inhibited significantly only under conditions in which extracellular ADP as well as ATP was removed efficiently. The hypotonically induced rise in [Ca 2ϩ ] i in HTC cells was compared with the increase in [Ca 2ϩ ] i elicited by the addition of ATP, UTP, or ADP to cells in isotonic medium, and the effect of submicromolar concentrations of ATP on the efflux of K ϩ ( 86 Rb ϩ ), Cl Ϫ ( 125 I Ϫ ), and the organic osmolyte taurine from cells in isotonic medium was compared with the effect of a decrease in extracellular osmolality. Our results confirm the earlier finding by Roe and colleagues (18) that the hypotonically induced rise in [Ca 2ϩ ] i seen in HTC cells is a consequence of the release of Ca 2ϩ from intracellular stores. However, our results also show that, in contrast to the conclusion drawn by Roe and colleagues (18), extracellular adenine nucleotides do play a role in this rise in [Ca 2ϩ ] i and in activating/modulating the consequent volume-regulatory efflux of K ϩ ( 86 Rb ϩ ) and taurine, but not Cl Ϫ ( 125 I Ϫ ), from the cells.

EXPERIMENTAL PROCEDURES
Materials-Fura-2 AM and pluronic F127 were obtained from Molecular Probes (Eugene, OR). Ionomycin was from Calbiochem-Novabiochem or Sigma. All other biochemicals were from Sigma. Radionuclides 86 RbCl, Na 125 I, and [ 3 H]taurine were obtained from Amersham Biosciences.
Solutions-Isotonic HEPES-buffered saline (HBS) was composed of the following (in mM): HEPES (10); NaCl (140); KCl (5); CaCl 2 (1.3); MgCl 2 (0.5); glucose (5), adjusted to pH 7.4 with NaOH. The osmolality was measured by freezing point depression in an Advanced Osmometer, model 3D3 (Advanced Instruments, Norwood, MA) and adjusted to 300 mOsm kg Ϫ1 (range of 299 -301 mOsm kg Ϫ1 ) with NaCl. The hypotonic (200 mOsm kg Ϫ1 ) solutions used in efflux experiments were prepared as above, except that the NaCl was reduced to 90 mM. For the measurement of [Ca 2ϩ ] i cells were subjected to a decrease in the extracellular osmolality by dilution of the isotonic HBS with a low-osmolality solution (35 mOsm kg Ϫ1 ) of the same composition as HBS but from which the NaCl was omitted. The final (measured) osmolality of the hypotonic medium was, in all cases, within the range 198 -204 mOsm kg Ϫ1 .
In experiments in which CaCl 2 was omitted from the buffers, the osmolality was adjusted to the required value with NaCl. In experiments in which extracellular Ca 2ϩ was chelated with EGTA at the time of dilution, the low-osmolality solution used to reduce the osmolality to 200 mOsm kg Ϫ1 contained 5 mM EGTA, and the concentration of the other components was reduced by 50% so that the osmolality of the diluting solution remained in the range 33-35 mOsm kg Ϫ1 . The final concentration of EGTA in the extracellular solution was 1.84 mM and that of free Ca 2ϩ (estimated using a program based on the equations described in Ref. 20) was 50 nM.
Apyrase (grade VII and VI) and hexokinase (type F) were each dissolved in water at a concentration of 2000 units/ml and stored as aliquots at Ϫ20°C. Solutions of 100ϫ the final required concentration were prepared in HBS and in the appropriate NaCl-reduced HBS solutions on the day of the experiment. ATP, ADP, 2-MeSADP, PPADS, and A2P5P were prepared as stock solutions in water at 5-50 mM. Frozen aliquots were thawed on the day of use and kept on ice. ADP was preincubated at 37°C for 1 h with hexokinase (20 units/ml) and glucose (20 mM) in order to remove contaminating ATP. Ionomycin and thapsigargin were prepared as 10 and 2 mM stocks, respectively, in dimethyl sulfoxide (Me 2 SO) and stored as aliquots at Ϫ20°C. Fura-2 AM was prepared on the day of the experiment as a 10 and assuming a K d of 224 nM (21). The parameters S f,2 , S b,2 , R min , and R max were all estimated in cells treated with 10 M ionomycin, using the internal calibration procedure described by Kao (22). S f,2 and S b,2 are the fluorescence emission (at 510 nm) obtained on excitation at 380 nm, after chelation of Ca 2ϩ with EGTA and in the presence of excess Ca 2ϩ , respectively. R min and R max are the ratios of the fluorescence emission (at 510 nm) obtained on excitation at 340 nm to that obtained on excitation at 380 nm, following chelation of Ca 2ϩ with EGTA, and in the presence of excess Ca 2ϩ , respectively.
The maximum increase in [Ca 2ϩ ] i (⌬[Ca 2ϩ ] i ), occurring in response to different stimuli was calculated by subtracting an average of 15-20 data points acquired immediately prior to the stimulus from an average of 15 data points acquired during the period when the rise in [Ca 2ϩ ] i was maximal.
Reverse Transcriptase-PCR-Total RNA was prepared from HTC cells (ϳ5 ϫ 10 6 ) using the NucleoSpin RNA II kit (Macherey-Nagel, Duren, Germany) according to the manufacturer's instructions. 2 g of RNA was incubated with oligo(dT) 15 primer (0.5 g, Invitrogen) for 10 min at 65°C in a total volume of 12 l and then chilled on ice. Dithiothreitol (10 mM, final), dNTPs (500 M, final), 5ϫ first-strand buffer (Invitrogen) and RNaseOUT ribonuclease inhibitor (20 units, Invitrogen) were added (total volume, 20 l) and incubated for 2 min at 42°C. In order to prepare first strand cDNA, Superscript II-RT (200 units, Invitrogen) was added, and the mixture was incubated for 2 h at 42°C, before terminating the reaction by heating to 75°C for 15 min. In order to detect any contamination of the RNA by genomic DNA, control incubations were performed in samples in which the reverse transcriptase was omitted.
PCRs were carried out for four rat P2Y receptor subtypes, rP2Y 1 , rP2Y 2 , rP2Y 4 and rP2Y 6 , as well as for the housekeeping gene actin (using the 5Ј-primer set, Gene Checker Kit, Invitrogen). The P2Y primers pairs, designed to amplify a 400-bp fragment from each receptor subtype, are shown in Table I. The PCR mixtures contained 2% (v/v) first strand cDNA solution, 100 pmol of each of the sense and antisense primers, 2 units of Taq polymerase (Qiagen), 20% (v/v) Q solution (Qiagen), and 200 M dNTPs. The amplification conditions were 2 min at 94°C, then 30 cycles (30 s at 94°C, 1 min at 50°C, 1 min at 72°C), and then 10 min at 72°C. Amplification products were separated on a 1% agarose gel by electrophoresis.
Efflux Measurements-HTC cells (ϳ2 ϫ 10 5 cells) were plated in 35-mm dishes and grown for ϳ24 h under normal culture conditions before being loaded with radioisotopes for 1-2 h at 37°C. In some cases 86 Rb ϩ and 125 I Ϫ efflux measurements were carried out using cells dual-labeled with both isotopes (0.3 Ci/ml for 86 Rb ϩ and 5 Ci/ml for 125 I Ϫ ). Taurine efflux measurements were carried out using separate dishes, prepared on the same day, and labeled with [ 3 H]taurine (0.1 Ci/ml). At the end of the loading period extracellular radiolabel was removed by repeated addition then removal of 1 ml of HBS (300 mOsm kg Ϫ1 ; 8-10 times). The flux experiment was then commenced by the careful addition to the cells of 1 ml of HBS (300 mOsm kg Ϫ1 ). After 4 -5 min the extracellular solution was transferred to a vial for scintillation counting, and the extracellular solution was replaced by HBS (300 mOsm kg Ϫ1 ) with or without grade VII apyrase or inhibitors. After a further 4 -5 min, the procedure was repeated using hypotonic HBS (200 mOsm kg Ϫ1 ) that also contained the appropriate agent. The radioactivity remaining in the cells at the end of the 4 -5-min period in hypotonic medium was estimated by lysing the cells with 0.5 M NaOH. Efflux measurements were carried out at room temperature (23°C) except where specified otherwise. Unidirectional efflux rate constants (k) were calculated using the expression: k ϭ ln(X t /X tϪ1 )/⌬t, where X t is the fraction of isotope remaining in the cells at time point (t), X tϪ1 is the fraction of isotope remaining at the previous time point, and ⌬t is the interval between the two time points.
Data  Fig.  1A. Smaller decreases in the osmolality gave rise to smaller rises in [Ca 2ϩ ] i , with significant increases still being observed in response to reducing the osmolality from 300 to 250 mOsm kg Ϫ1 (not shown).
Grade VII apyrase is an ATP/ADP phosphatase with a low ATP/ADPase ratio. Inclusion of 3 units/ml of this enzyme preparation in the extracellular solution 5 min prior to and during a decrease in the osmolality caused a marked reduction in the hypotonically induced ⌬[Ca 2ϩ ] i to 12 Ϯ 2 nM (n ϭ 10), 13% of the corresponding same-day controls (91 Ϯ 12 nM, Fig. 1B). In experiments in which the initial resting [Ca 2ϩ ] i was found to be above 100 nM (in isotonic medium), the addition of apyrase (grade VII) also caused a significant decrease in the resting [Ca 2ϩ ] i (results not shown), implicating extracellular nucleotides in the elevated resting [Ca 2ϩ ] i .
Grade VI apyrase is an ATP/ADP phosphatase with a high ATP/ADPase ratio. In contrast to the results obtained with grade VII apyrase, inclusion of grade VI apyrase (3 units/ml) did not decrease the hypotonically induced ⌬[Ca 2ϩ ] i . The ⌬[Ca 2ϩ ] i in the presence of grade VI apyrase was 100 Ϯ 15 nM (n ϭ 5), not significantly different from the value of 79 Ϯ 6 nM observed for the same-day controls (Fig. 1B).
Hexokinase, which reacts with ATP to phosphorylate glucose and form ADP, was also without significant effect on the hypotonically induced ⌬[Ca 2ϩ ] i observed in response to a reduction in the extracellular osmolality (Fig. 1B). The ⌬[Ca 2ϩ ] i in the presence of hexokinase was 125 Ϯ 20 nM, not significantly different from the 115 Ϯ 20 nM increase seen for the same-day controls (n ϭ 3).
The transient increase in [Ca 2ϩ ] i observed on reduction of the osmolality of the extracellular solution was not due to mechanical disturbance or to the decrease in ionic strength. Addition of an identical volume of isotonic buffer did not alter the measured resting [Ca 2ϩ ] i nor did the addition of an isotonic mannitol buffer, which maintained osmolality but reduced ionic strength of the extracellular solution (results not shown).
Hypotonic Swelling Releases Ca 2ϩ from Internal Stores-The role of extracellular Ca 2ϩ in the hypotonically induced rise in [Ca 2ϩ ] i was investigated by the following: 1) omitting Ca 2ϩ from the extracellular solution; 2) chelating extracellular Ca 2ϩ simultaneously with the reduction in osmolality; and 3) pretreating cells with thapsigargin in order to deplete the endo- plasmic reticulum (ER) Ca 2ϩ stores (23) prior to the reduction in osmolality.
When HTC cells were placed for 15-20 min, in solutions from which Ca 2ϩ had been omitted, the resting [Ca 2ϩ ] i decreased to 21 Ϯ 4 nM (n ϭ 10). Under these conditions the ⌬[Ca 2ϩ ] i observed in response to a decrease in the extracellular osmolality from 300 to 200 mOsm kg Ϫ1 was reduced to 12 Ϯ 2 nM (n ϭ 6), compared with 94 Ϯ 16 nM (n ϭ 8) for same-day controls in solutions containing 1.3 mM extracellular Ca 2ϩ . One possible explanation for this reduction in ⌬[Ca 2ϩ ] i is that extracellular Ca 2ϩ is the major source of Ca 2ϩ required for the ⌬[Ca 2ϩ ] i observed on osmotic swelling. However, when cells that had been subjected to a decrease in osmolality were treated subsequently with thapsigargin in order to release any Ca 2ϩ remaining in the ER Ca 2ϩ stores (23), the thapsigargin-induced ⌬[Ca 2ϩ ] i in cells bathed in the Ca 2ϩ -free solution (57 Ϯ 6 nM, n ϭ 6) was ϳ10-fold lower than that in cells bathed in solution containing 1.3 mM extracellular Ca 2ϩ (680 Ϯ 90 nM, n ϭ 8). Exposure of HTC cells to a Ca 2ϩ -free solution therefore caused a marked decrease in the size of the intracellular Ca 2ϩ pool. Representative [Ca 2ϩ ] i traces are shown in Fig. 2, A and B.
In Fig. 2C the ⌬[Ca 2ϩ ] i observed in response to a decrease in the osmolality (in cells in both Ca 2ϩ -free and Ca 2ϩ -containing media) is expressed relative to the ⌬[Ca 2ϩ ] i observed in response to the subsequent addition of thapsigargin. For cells in both solutions the ratio is the same. These data are consistent with the hypothesis that the Ca 2ϩ giving rise to the hypotonically induced increase in [Ca 2ϩ ] i was derived from the ER Ca 2ϩ store. The amount of Ca 2ϩ in this store was reduced substantially when the cells were transferred from a Ca 2ϩ -containing to a Ca 2ϩ -free medium. Nevertheless, the proportion of the total amount of ER store Ca 2ϩ store that was released in response to a hypotonic shock was the same in each case.
In order to confirm that entry of extracellular Ca 2ϩ is not required for the hypotonically induced ⌬[Ca 2ϩ ] i , the normal (1.3 mM extracellular Ca 2ϩ ) isotonic solution was diluted with a low osmolality solution containing 5 mM EGTA, so that the extracellular concentration of free Ca 2ϩ was reduced from 1.3 mM to ϳ50 nM simultaneously with the reduction in osmolality from 300 to 200 mOsm kg Ϫ1 . A typical trace is shown in Fig. 3A. Under these conditions the decrease in osmolality caused [Ca 2ϩ ] i to undergo an initial rise of 104 Ϯ 16 nM (n ϭ 6), not significantly different from same-day controls in cells exposed to 1.3 mM free Ca 2ϩ (Fig. 3B). The initial rise was followed by a decrease in [Ca 2ϩ ] i , to ϳ20 nM.
When an isotonic solution containing EGTA was added (such that the final free Ca 2ϩ concentration was ϳ50 nM) there was no increase in [Ca 2ϩ ] i ; [Ca 2ϩ ] i simply decreased to ϳ20 nM (Fig.  3C), close to the value measured in cells exposed to solutions containing no added Ca 2ϩ ( Fig. 2A). For cells in the EGTAcontaining solutions (containing ϳ50 nM free Ca 2ϩ ) the size of the thapsigargin-sensitive Ca 2ϩ store was reduced ϳ10-fold relative to that seen in cells in media containing 1.3 mM free Ca 2ϩ (compare Fig. 3, A and C, with Fig. 2B), as was the case when Ca 2ϩ was omitted from the solutions ( Fig. 2A).
When thapsigargin (200 nM) was added to HTC cells bathed in isotonic medium containing 1.3 mM Ca 2ϩ , [Ca 2ϩ ] i increased to a maximum of 1000 Ϯ 40 nM (n ϭ 5) above resting levels and then gradually declined to a value of ϳ400 nM after 30 min. A representative Ca 2ϩ trace is shown in Fig. 3D. When the cells were subjected to a decrease in the extracellular osmolality after 30 min or more in the presence of thapsigargin, [Ca 2ϩ ] i showed no sign of increasing, again consistent with the thapsigargin-sensitive ER Ca 2ϩ store providing the source of Ca 2ϩ for the hypotonically induced increase in [Ca 2ϩ ] i . Instead the decrease in osmolality caused [Ca 2ϩ ] i to decline rapidly to ϳ200 nM, as would be expected from a rapid increase in cell volume and consequent dilution of the cytoplasm. The initial decline was followed by a slower increase, probably a consequence of the cells undergoing RVD (Fig. 3D).
Gd 3ϩ Inhibits Hypotonically Induced Increases in [Ca 2ϩ ] i -Gd 3ϩ , a nonspecific inhibitor of stretch-activated non-selective cation channels, as well as of various Ca 2ϩ channels (24) has been found to inhibit RVD in a number of hypotonically swollen cells including HTC cells (10,25,26). It has also been proposed that Gd 3ϩ inhibits swelling-activated ATP release with an IC 50 of between 10 and 100 M (8, 10,11,27), although the extent to which this is true has recently been questioned. Gd 3ϩ has been found to interfere with the luciferase assay for ATP (28), and after accounting for this inhibitory action, Maroto and Hamill (29) found that 10 M Gd 3ϩ reduced mechanosensitive release of ATP from Xenopus oocytes by only 7%. consistent with Gd 3ϩ blocking an entry pathway for divalent cations. Gd 3ϩ was also found to reduce the amount of Ca 2ϩ that could be released from thapsigargin-sensitive Ca 2ϩ stores. The ⌬[Ca 2ϩ ] i observed on addition of thapsigargin, 4 min after Gd 3ϩ treatment, was 128 Ϯ 7 nM (n ϭ 3) compared with 1100 Ϯ 80 nM for the untreated control cells (results not shown). These data are consistent with the effect of Gd 3ϩ on the hypotonically induced increase in [Ca 2ϩ ] i being a consequence of the depletion of the (thapsigargin-sensitive) intracellular Ca 2ϩ stores.
P2Y 1 Antagonists Inhibit the Hypotonically Induced Rise in Cytoplasmic [Ca 2ϩ ]-The failure of grade VI apyrase (which converts ATP to ADP more efficiently than the subsequent hydrolysis of ADP to AMP) or of hexokinase (which converts ATP to ADP) to inhibit the hypotonically induced increase in [Ca 2ϩ ] i (Fig. 1B) is consistent with the hypothesis that ADP is able to stimulate purinergic receptor(s) involved in the hypotonically induced increase in [Ca 2ϩ ] i . The finding that the Ca 2ϩ giving rise to the hypotonically induced increase in [Ca 2ϩ ] is derived from intracellular Ca 2ϩ stores, rather than directly from the extracellular solution, implicates P2Y (rather than P2X) receptors in the process. ADP is a strong agonist for P2Y 1 , but not for P2Y 2 , P2Y 4 , or P2Y 6 purinergic receptors (17,30). In this study two purinergic receptor antagonists, A2P5P and PPADS, were tested for their ability to inhibit the rise in [Ca 2ϩ ] i caused by a decrease in osmolality from 300 to 200 mOsm kg Ϫ1 . A2P5P has been shown to inhibit P2Y 1 but not other P2Y receptors (31,32). PPADS is a nonspecific inhibitor of purinergic receptors showing a significant inhibition of P2Y 1 but weaker or no inhibition of the other P2Y receptors (17,30).
The hypotonically induced increase in [Ca 2ϩ ] i was inhibited by more than 90% both by 100 M A2P5P and by 10 M PPADS. Representative traces are shown in Fig. 5A with averaged data shown in the fourth panel. Addition of PPADS interfered to a small degree with the fluorescence signal of Fura-2, causing an apparent jump in the ratio of 340/380 nm (not shown); for these experiments, calibration of the fluorescence signals was therefore carried out in the presence of PPADS. The purinergic receptor inhibitors suramin and reactive blue 2 could not be used in our system as they interfered strongly with the Fura-2 fluorescence signals.
To assess whether the effect of A2P5P and PPADS on the hypotonically activated ⌬[Ca 2ϩ ] i in HTC cells could be accounted for by inhibition of purinergic receptors that are selective for ADP rather than those (such as P2Y 2 , P2Y 4 , or P2Y 6 receptors) for which ATP and/or UTP or UDP are major agonists (30,31), the effects of ADP and UTP on [Ca 2ϩ ] i in cells in isotonic medium were compared. The concentration of nucleotides were chosen to give a maximum ⌬[Ca 2ϩ ] i of the same order of magnitude and no more than twice that observed in response to a reduction in the extracellular osmolality from 300 to 200 mOsm kg Ϫ1 . Samples of ADP were pre-treated with hexokinase to eliminate contamination by ATP, and the experiments involving ADP were performed in the presence of 2 units/ml hexokinase with 5 mM glucose in the extracellular solution.
HTC cells in an isotonic medium responded to 2 M ADP with a rapid rise in [Ca 2ϩ ] i , before returning to resting levels within 5-6 min. The average maximum response was 180 Ϯ 20 nM (n ϭ 3). A representative trace is shown in Fig. 5B. A very similar response was observed when 2-MeSADP, another strong P2Y 1 agonist (30), was applied to the cells (results not shown). Both A2P5P and PPADS inhibited the response to ADP (Fig. 5B), reducing the ⌬[Ca 2ϩ ] i to 10 -15% of the control response. Thus the action of both inhibitors on the ADP-activated increase in [Ca 2ϩ ] i closely resembled their effects on the hypotonically induced increase in [Ca 2ϩ ] i (Fig. 5A).
On addition of UTP (Fig. 5C) or ATP (Fig. 5D) to cells in isotonic medium, [Ca 2ϩ ] i rose to a maximum and then decreased to levels above that of the resting cell [Ca 2ϩ ] i for the duration of the experiment. The averaged ⌬[Ca 2ϩ ] i values observed in response to the addition of 500 nM UTP and 500 nM ATP were 130 Ϯ 8 (n ϭ 4) and 140 Ϯ 15 nM (n ϭ 9), respectively. As expected for an inhibitor specific for the P2Y 1 receptor, A2P5P (100 M) did not inhibit the response of HTC cells to UTP (Fig. 5C, 2nd panel). PPADS (10 M) did cause a significant reduction in the UTP-induced ⌬[Ca 2ϩ ] i (Fig. 5C, 3rd panel), consistent with this antagonist having a mild inhibitory action on P2Y 2 and/or another of the purinergic receptors in HTC cells. Inhibition of the ATP-induced ⌬[Ca 2ϩ ] i by PPADS was ϳ50%, significantly greater than that observed for the UTP-induced ⌬[Ca 2ϩ ] i (p ϭ 0.01).
P2Y 1 Receptors Are Expressed in HTC Cells-The physiological and pharmacological data implicate P2Y 1 receptors in the response of HTC cells to a decrease in the extracellular osmolality. To test which of the P2Y receptors are actually expressed in HTC cells, 400-bp fragments for each of the four rat P2Y receptors so far cloned (rP2Y 1 , rP2Y 2 , rP2Y 4 , and rP2Y 6 ) were amplified by RT-PCR. The rP2Y 2 400-bp fragment gave the FIG. 6. HTC cells express mRNA for rat P2Y 1 , P2Y 2 , and P2Y 6 receptors. Total RNA from HTC cells was reverse-transcribed into cDNA, and PCR was performed using specific primers designed to synthesize fragments of 400 bp for each of the rat receptor subtypes, rP2Y 1 , rP2Y 2 , rP2Y 4 and rP2Y 6 , and for a 1000-bp fragment of the housekeeping gene for actin. Lanes 1 show controls for genomic DNA contamination in which the reverse transcriptase was omitted from the reaction mixture. Lanes 2 show the products of the RT-PCR mixtures. The volume of the PCR reaction product applied to the gel for actin (10 l) was half that of the P2Y products (20 l). Lane L is a 100-bp DNA ladder. strongest signal, although significant amounts of rP2Y 1 and rP2Y 6 mRNA were also detected (Fig. 6, lanes 2). No product corresponding to 400-bp fragments was detected with the rP2Y 4 primers. No fragments were detected when the reverse transcriptase was omitted from the RT-PCR mixture (Fig. 6, lanes 1), indicating the absence of contamination by genomic DNA.
Hypotonically Activated Ion and Organic Osmolyte Efflux-Reduction of the extracellular osmolality from 300 to 200 mOsm kg Ϫ1 resulted in an increase in the efflux of 86 Rb ϩ (a marker for K ϩ ), 125 I Ϫ (a marker for Cl Ϫ ), and [ 3 H]taurine (the radiolabeled form of a common organic osmolyte) from HTC cells (Fig. 7). In the presence of grade VII apyrase (3 units/ml, closed bars), which was shown in Fig. 1 to inhibit the hypotonically induced ⌬[Ca 2ϩ ] i strongly, the hypotonically induced increase in 86 Rb ϩ efflux rate constant was inhibited completely (Fig. 7A). Apyrase had no significant effect on hypotonically induced 125 I Ϫ efflux but caused a partial inhibition of the hypotonically induced taurine efflux (Fig. 7A).
Gd 3ϩ also inhibited the hypotonically induced ⌬[Ca 2ϩ ] i (Fig.  4B). Fig. 7B shows the effect of this ion on the efflux of solutes from HTC cells in isotonic and hypotonic (200 mOsm kg Ϫ1 ) media. 1 M Gd 3ϩ caused a significant (p ϭ 0.02) inhibition of 86 Rb ϩ efflux in hypotonic (but not isotonic) medium, whereas 10 M Gd 3ϩ eliminated the hypotonically induced 86 Rb ϩ efflux completely. In the case of taurine, there was a progressive decrease in hypotonically induced efflux with increasing [Gd 3ϩ ]. By contrast, hypotonically induced 125 I Ϫ efflux was not inhibited significantly by either 1 or 10 M Gd 3ϩ (p ϭ 0.5). PPADS (10 M) strongly inhibited 86 Rb ϩ efflux in a hypotonic solution as well as causing significant inhibition of hypotonically activated 125 I Ϫ and [ 3 H]taurine efflux (Fig. 7C).
Comparison of ATP-activated Efflux with Hypotonically Induced Efflux-The effect of 500 nM ATP on solute efflux was compared with that of a 33% decrease in osmolality in paired experiments (Fig. 8). Measurements of [Ca 2ϩ ] i on the same days as the efflux experiments gave a mean value for ⌬[Ca 2ϩ ] i of 146 Ϯ 16 nM (n ϭ 8) on the addition of 500 nM ATP in isotonic solutions, not significantly different (p ϭ 0.15) from 110 Ϯ 16 nM (n ϭ 7) obtained in response to a decrease in osmolality. Exposure of HTC cells in isotonic solution to 500 nM ATP caused the 86 Rb ϩ efflux rate constant to increase by more than 2-fold, from a resting level of 0.30 Ϯ 0.02 to 0.65 Ϯ 0.07 h Ϫ1 . This increase was significantly smaller (p ϭ 0.006) than that observed to occur in response to a 33% reduction in the osmolality, i.e. from 0.29 Ϯ 0.005 to 1.12 Ϯ 0.09 h Ϫ1 . The ATPactivated 86 Rb ϩ efflux was inhibited strongly by 10 M PPADS (Fig. 9) and 10 M Gd 3ϩ (results not shown), as was the hypotonically induced 86 Rb ϩ efflux (Fig. 7, B and C).
A small, but significant, increase in the 125 I Ϫ efflux rate constant from 1.1 Ϯ 0.1 to 1.7 Ϯ 0.08 h Ϫ1 was observed on the addition of 500 nM ATP (Fig. 8). This small rise was dwarfed by the 5-fold increase in 125 I Ϫ efflux that occurred in response to a 33% reduction in osmolality (from 1.4 Ϯ 0.09 to 7.2 Ϯ 0.5 h Ϫ1 ). 500 nM ATP had no significant effect on taurine efflux (Fig. 8). Extracellular ATP at 100 M was found to cause much larger increases in the rate of efflux of both 86 Rb ϩ and 125 I Ϫ , whereas taurine efflux remained unaffected (results not shown).

HTC Cells Undergo a Hypotonically Induced Rise in [Ca 2ϩ ] i
Due to the Release of Ca 2ϩ from Intracellular Stores-Exposure of HTC cells to a reduction of the extracellular osmolality resulted in a transient increase in [Ca 2ϩ ] i . The same observation has been made by Roe et al. (18) who also presented evidence that this is due to the release of Ca 2ϩ from intracellular stores. Our own data provide further support for this view. A number of maneuvers shown here to decrease the size of the intracellular Ca 2ϩ pool reduced or eliminated the hypotonically induced increase in [Ca 2ϩ ] i . These included preincubation of cells in a Ca 2ϩ -free medium ( Fig. 2A), pretreatment of the cells with thapsigargin (Fig. 3D), or blockade of Ca 2ϩ influx by Gd 3ϩ (Fig. 4). Conversely, chelation of extracellular Ca 2ϩ simultaneously with reduction of the extracellular osmolality (thereby eliminating the extracellular Ca 2ϩ pool as a source of Ca 2ϩ ) had no significant effect on the magnitude of the ⌬[Ca 2ϩ ] i observed (Fig. 3A).
The findings that omission of extracellular Ca 2ϩ ( Fig. 2A) and the presence of 1 M Gd 3ϩ (Fig. 4A) both decreased the resting [Ca 2ϩ ] i value (as well as the amount of Ca 2ϩ that was available from the thapsigargin-sensitive store) are indicative of a dynamic equilibrium between intra-and extracellular Ca 2ϩ in HTC cells, with both the resting [Ca 2ϩ ] i and the amount of Ca 2ϩ in the stores reflecting a balance between the influx and efflux of Ca 2ϩ . Omission of Ca 2ϩ from the external medium, or addition of Gd 3ϩ both decrease the influx of Ca 2ϩ , thereby decreasing both cytoplasmic [Ca 2ϩ ] i and the amount of Ca 2ϩ in the stores. It is as a result of the depletion of Ca 2ϩ stores that both maneuvers cause a marked decrease in the The Hypotonically Induced Increase in [Ca 2ϩ ] i in HTC Cells Is Mediated by Activation of P2Y Purinergic Receptors-It is now well established that cells exposed to a range of physical stresses, including hypo-osmotic shock, release ATP. The mechanism by which this occurs has yet to be elucidated but involves a process that can be inhibited by anion channel blockers and is independent of cell lysis (4,7,13,36,37). In their recent study, Roe and colleagues (18) showed that grade VI apyrase did not prevent the hypotonically induced increase in [Ca 2ϩ ] i in HTC cells, and they concluded from this that the increase was not mediated by extracellular ATP/purinergic receptors. In this study we have confirmed the findings by Roe et al. (18) with grade VI apyrase, but we have also obtained independent evidence for the involvement of P2Y 1 purinergic receptors in the hypotonically induced increase in [Ca 2ϩ ] i and the ensuing volume-regulatory response.
Two isoforms of apyrase (EC 3.6.1.5) have been isolated from potatoes: one isoform has an ATPase/ADPase ratio of ϳ10:1, and the other has an ATPase/ADPase ratio of ϳ1:1 (38,39). Grade VI apyrase is predominantly the former, the high ratio isoform, and grade VII apyrase is predominantly the latter, the low ratio isoform (Sigma catalogue). Thus grade VI enzyme should form ADP faster than it can hydrolyze it to AMP (leading to an increase in [ADP]), whereas the grade VII enzyme should be able to catalyze the hydrolysis of ADP at a rate that keeps pace with its formation. Grade VII apyrase (and not hexokinase or apyrase VI) was shown here to inhibit the hypotonically induced increase in [Ca 2ϩ ] i (Fig. 1B). These results are consistent with the hypothesis that ADP, either as well as or instead of ATP, is active in eliciting the hypotonically induced increase in [Ca 2ϩ ] i in HTC cells through the stimulation of purinergic receptors. The finding that the Ca 2ϩ is intracellular in origin implicates the G-protein-linked P2Y receptors (rather than the ionotropic P2X receptors) in this rise in [Ca 2ϩ ] i .

HTC Cells Express a Number of Functional P2Y
Receptors-In this study it was shown by using RT-PCR that mRNA transcripts for P2Y 1 , P2Y 2 , and P2Y 6 , but not P2Y 4 , receptors are present in HTC cells (note that the rat homologues of the P2Y 11 , P2Y 12 , and P2Y 13 receptors (30,40) have not been cloned as yet). mRNA transcripts for P2Y 2 receptors have been detected previously (41) in HTC cells and those for P2Y 1 , P2Y 2 , P2Y 4 , and P2Y 6 receptors have been found in rat hepatocytes (42), human hepatocytes, and the human hepatoma cell lines, Hep G2 and HuH-7 (43).
Previous studies by Fitz and co-workers (44) have provided evidence for the presence of functional P2Y 2 receptors in HTC cells. The results of the present study provide physiological and pharmacological evidence for the presence of functional P2Y 1 receptors. ADP, which interacts with P2Y 1 receptors with an affinity ϳ100 -300 times greater than that does ATP (45), caused an increase in [Ca 2ϩ ] i when added to cells in isotonic medium (Fig. 5B). The increase was blocked by A2P5P, an antagonist believed to be specific for P2Y 1 receptors (31), as well as by PPADS, a less specific purinergic receptor antagonist (17,30).
The addition of UTP (an agonist for P2Y 2 and P2Y 4 receptors but not P2Y 1 receptors) also caused an increase in [Ca 2ϩ ] i when added to cells in isotonic medium. The UTP-induced ⌬[Ca 2ϩ ] i was unaffected by A2P5P (consistent with A2P5P acting as a specific inhibitor of P2Y 1 receptors in our system) but underwent an ϳ30% reduction in the presence of 10 M PPADS (Fig.  5C). ATP (an agonist for rat P2Y 1 , P2Y 2 , and P2Y 4 receptors) again caused an increase in [Ca 2ϩ ] i when added to cells in isotonic medium, and this was reduced by ϳ50% in the presence of PPADS (Fig. 5D). The inhibition of the ATP-induced ⌬[Ca 2ϩ ] i by PPADS was significantly greater than the inhibition of the UTP-induced ⌬[Ca 2ϩ ] i (p ϭ 0.01). This might be accounted for if the response of HTC cells to extracellular ATP was due to the combination of a PPADS-sensitive P2Y 1 component and a less PPADS-sensitive P2Y 2 component, whereas the response of HTC cells to extracellular UTP was due to P2Y 2 receptors. The possible contribution of P2Y 6 receptors to nucleotide-induced increases in [Ca 2ϩ ] i in HTC cells was not tested. The UTP used in the experiments giving rise to Fig. 5C may have been contaminated with UDP, and therefore the observed increase in [Ca 2ϩ ] i could be due, in part, to a P2Y 6 response.
Rat hepatocytes, which express both P2Y 1 and P2Y 2 receptors, have been shown to behave in a similar fashion to that seen here for HTC cells. In rat hepatocytes the antagonist A3P5P, a close structural analogue of A2P5P, inhibits the [Ca 2ϩ ] i rise in response to 2-MeSADP (a strong agonist of P2Y 1 receptors) but does not inhibit the rise in response to UTP (46).
The Hypotonically Induced Increase in [Ca 2ϩ ] i in HTC Cells Is Mediated by P2Y 1 Receptors-A2P5P and PPADS both inhibited the rise in [Ca 2ϩ ] i seen in cells subjected to a decrease in the extracellular osmolality. The pattern of inhibition of the hypotonically activated increase in [Ca 2ϩ ] i resembled very closely that seen for the inhibition of the increase in [Ca 2ϩ ] i induced by the addition of the P2Y 1 agonist ADP to cells in isotonic medium (Fig. 5, A and B). These data are consistent with the hypothesis that in HTC cells P2Y 1 , and not P2Y 2 , receptors are activated by adenine nucleotides released in response to a hypotonic shock.
The previous observation by Roe and colleagues (18) that suramin, a nonspecific antagonist of P2 receptors which should inhibit both P2Y 1 and P2Y 2 receptors, does not inhibit the rise in [Ca 2ϩ ] i in response to a hypotonic shock is difficult to explain. Both suramin and PPADS have been shown to result in at least partial inhibition of the swelling-induced increase in [Ca 2ϩ ] i in other cell types (13,15,16). We were unable to use suramin in our system due to its interference in the detection of the Fura-2 fluorescence.
Why Do Released Nucleotides Activate P2Y 1 and Not P2Y 2 Receptors?-The hypothesis that adenine nucleotides, released in response to moderate (33%) hypo-osmotic shock, interact with P2Y 1 receptors and thereby activate the release of Ca 2ϩ from intracellular stores raises the question of why the ATP that is released, and has been detected previously (10) using the luciferin-luciferase reaction, fails to induce significant activation of P2Y 2 receptors.
One possible explanation for the preferential activation of P2Y 1 over P2Y 2 receptors may lie in the relative affinities of the two receptors for ATP. The EC 50 for activation of P2Y 1 recep- tors by ATP has been reported to be ϳ40 nM (45), whereas that for activation of P2Y 2 receptors by ATP is 200 -300 nM (41,47). Although the concentration of ATP in the immediate vicinity of the membrane of hypotonically swollen cells is not known, it may be insufficient to induce significant activation of the comparatively low affinity P2Y 2 receptors yet be sufficiently high to cause activation of the higher affinity P2Y 1 receptors.
Another possible explanation for the preferential activation of P2Y 1 receptors by hypotonically released ATP might lie in a co-localization of P2Y 1 receptors with the ATP release pathway(s) and/or with ecto-apyrases that are known to be active in liver cell plasma membranes (48 -50). The EC 50 for activation by ADP of P2Y 1 receptors has been estimated as 0.13 nM, 300-fold lower than the corresponding value for ATP (45). If ADP is formed from hypotonically released ATP close to P2Y 1 receptors, it may be sufficient to saturate these receptors. This would explain why the addition of hexokinase and apyrase VI (both of which should have converted all released ATP to ADP) did not cause a significant increase in the hypotonically induced ⌬[Ca 2ϩ ] i (Fig. 1B).
The P2Y 1 -mediated Rise in [Ca 2ϩ ] i Activates a K ϩ Channel-The increased efflux of K ϩ ( 86 Rb ϩ ) from HTC cells following a reduction in the extracellular osmolality was inhibited by grade VII apyrase, 10 M Gd 3ϩ , and 10 M PPADS (Fig. 7), all of which prevented the hypotonically induced increase in [Ca 2ϩ ] i . This is in accord with the results of Roe et al. (18) who showed that swelling-activated K ϩ currents in HTC cells were inhibited by chelation of intracellular Ca 2ϩ . These data, together with our finding in another study 2 that the hypotonically activated efflux of 86 Rb ϩ was inhibited strongly by charybdotoxin but not apamin, are consistent with the hypothesis that a charybdotoxin-sensitive, Ca 2ϩ -activated K ϩ channel is involved in the response of HTC cells to hypotonic swelling.
The ⌬[Ca 2ϩ ] i induced by a 33% reduction in the extracellular osmolality was not significantly different from that observed on application of 500 nM ATP to HTC cells, whereas the increase in the 86 Rb ϩ efflux rate constant elicited in response to the hypotonic shock was significantly higher than that activated by 500 nM ATP (Fig. 8). If the hypotonically activated ⌬[Ca 2ϩ ] i is a result of the activation of P2Y 1 receptors and if the addition of exogenous ATP to cells in isotonic medium results in activation of both P2Y 1 and P2Y 2 receptors, this might infer that the [Ca 2ϩ ] i rise elicited by the P2Y 1 receptor is more effective in activating K ϩ efflux than the ⌬[Ca 2ϩ ] i caused by a combination of P2Y 1 and P2Y 2 receptors. The observation that the 86 Rb ϩ efflux activated by 500 nM ATP was completely blocked by PPADS (Fig. 9), even though the ⌬[Ca 2ϩ ] i was reduced by only 50% (Fig. 5), is consistent with the hypothesis that the Ca 2ϩ for activating this K ϩ channel is provided primarily by P2Y 1 activation even when ATP is added to the external solution and is able to activate both P2Y 1 and P2Y 2 receptors. This would argue for co-localization of P2Y 1 receptors and Ca 2ϩ -activated K ϩ channels in HTC cells.
Hypotonically Activated Cl Ϫ ( 125 I Ϫ ) Efflux Is Not Dependent on the P2Y 1 -mediated Rise in [Ca 2ϩ ] i -The increase in the 125 I Ϫ efflux rate constant seen in response to application of 500 nM ATP to HTC cells in isotonic medium was Ͻ10% that observed in cells subjected to a 33% decrease in the extracellular osmolality. Higher concentrations of ATP (e.g. 100 M) did cause a much larger increase in 125 I Ϫ efflux (results not shown) and have been reported previously to activate Cl Ϫ currents in HTC cells (51). The hypotonically activated 125 I Ϫ efflux did not require a rise in [Ca 2ϩ ] i as neither extracellular grade VII apyrase nor Gd 3ϩ , both of which inhibited the rise in [Ca 2ϩ ] i , caused a significant inhibition of 125 I Ϫ efflux (Fig. 7, A and B). The inhibition of 125 I Ϫ efflux seen in the presence of PPADS (Fig. 7C) is likely to be due to direct blockade of the efflux pathway because this organic anion and other purinergic antagonists have been found to inhibit hypotonically activated anion currents (12,52).
These results contrast with those of Fitz and colleagues who have shown that apyrase inhibits hypotonically activated anion currents in HTC cells (4,6), hepatocytes (14), and in a biliary cell line (9). Swelling-activated Cl Ϫ currents in HTC cells were also partially inhibited on chelation of intracellular Ca 2ϩ (18). We cannot explain the reason for this discrepancy. Our results are in agreement with reports in human epithelial intestine 407 cells and murine C127 cells in which the swelling-activated release of ATP was not required for activation of a Cl Ϫ current (11,12,27) nor for hypotonically induced efflux of 125 I Ϫ (12).
Taurine efflux was unaffected by the addition of exogenous ATP to HTC cells in isotonic medium. However, prevention of the hypotonically induced ⌬[Ca 2ϩ ] i by apyrase and Gd 3ϩ led to a partial inhibition of taurine efflux. This may be attributed to the signaling pathways involved in the hypo-osmotically induced activation of taurine efflux functioning more effectively when [Ca 2ϩ ] i undergoes a transient increase in response to hypo-osmotic swelling. In cultured astrocytes taurine efflux was partially inhibited by chelation of intracellular Ca 2ϩ (53).
The finding that hypotonically activated taurine efflux was more sensitive to [Ca 2ϩ ] i than was 125 I Ϫ efflux is relevant to the ongoing debate about the relationship between swelling-activated anion/osmolyte efflux pathways (54). It has been proposed that swelling-activated anion and taurine efflux occur through a common pathway in HTC (55) and other cell types (56,57). Although such a common pathway may be present, the FIG. 10. Model for the autocrine action of ATP released in response to a hypotonic shock in HTC cells. Released ATP acts on P2Y 1 receptors and/or is hydrolyzed via ecto-apyrases (apyrase) to ADP which is a much more potent agonist of P2Y 1 receptors. Binding of ADP and/or ATP to the P2Y 1 receptor activates phospholipase C (PLC) which liberates inositol trisphosphate (IP 3 ). Inositol trisphosphate activates inositol trisphosphate receptors (IP 3 R) located in the endoplasmic reticulum, and the resulting increase in [Ca 2ϩ ] i activates the efflux of K ϩ and modulates that of taurine. finding here that hypotonically activated taurine efflux showed a dependence on [Ca 2ϩ ] i , whereas hypotonically activated 125 I Ϫ efflux did not, is consistent with the operation of an organic osmolyte pathway that is more sensitive to [Ca 2ϩ ] i than are anion efflux pathways.
Based on the results obtained in this study we present a model indicating how hypotonically released ATP may act on extracellular purinergic receptors in order to influence the activation of ion and osmolyte efflux during RVD in HTC cells (Fig. 10). Osmotic swelling induces the release of ATP which acts on P2Y 1 receptors and/or is converted (via ecto-apyrases at the cell surface) to ADP which is a much more potent agonist of P2Y 1 receptors. P2Y 1 receptor activation leads to the release of Ca 2ϩ from intracellular stores, inducing the activation of Ca 2ϩactivated K ϩ channels as well as modulating the activity of the organic osmolyte efflux pathway(s). This model differs from those put forward in previous studies (4,18) in two key aspects. Activation of purinergic receptors does underlie the hypotonically induced rise in [Ca 2ϩ ] i , and this activation (and rise in [Ca 2ϩ ] i ) is not required for the hypotonically activated efflux of Cl Ϫ ( 125 I Ϫ ). However, our results are consistent with the previous proposal by Roe and colleagues (18) that Ca 2ϩ , released from intracellular stores, plays a key role in hypotonically induced K ϩ channel activation.
In the in vivo situation liver cell volume increases in response to physiological stimuli such as nutrient uptake and the hormone insulin, and the hydration state of the cell has been shown to be a controlling factor in the subsequent changes in cell function (58). Hä ussinger and colleagues (58,59) have proposed that G-protein activation is a primary event in the signaling cascades (e.g. the mitogen-activated protein kinase (MAPK) pathway) that link cell swelling to subsequent metabolic and functional changes. In this study we have shown that the G-protein-linked P2Y 1 receptor is activated by extracellular nucleotides in response to osmotic swelling in HTC cells, a rat hepatoma cell line, causing a transient increase in [Ca 2ϩ ] i . Although the osmotic shock used for this study was outside the normal physiological range, a decrease in osmolality to 250 mOsm kg Ϫ1 (which would lead to swelling comparable with physiological stimuli in the liver (58)) did result in a significant rise in [Ca 2ϩ ] i (results not shown). ATP, released in response to swelling, has been shown activate the MAPK pathway in intestine 407 cells (12), and the binding of adenine nucleotides to P2Y 1 receptors has been shown to activate the MAPK pathway (60). Thus in addition to activating K ϩ efflux in response to an osmotic shock, the binding of extracellular adenine nucleotides to purinergic receptors could also be involved in activating the signaling cascades responsible for the consequent metabolic changes observed in swollen cells.