|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Laboratorio de Biomembranas, Departamento de Fisiología y Biofísica, Facultad de Medicina, Universidad de Buenos Aires, Buenos Aires, Argentina
Submitted 11 September 2007 ; accepted in final form 14 December 2007
| ABSTRACT |
|---|
|
|
|---|
aquaporin 2; intracellular calcium; renal cells
Most cells respond to decrease in tonicity first by swelling and second by initiating mechanisms that allow them to recover their original volume (21, 35). This complex mechanism, called regulatory volume decrease (RVD), depends on the activation of different ion permeabilities (usually K+, Cl–, and/or HCO3–) that reverse the osmotic gradient and direction of water flow (35). Cell volume regulation mechanisms have been studied in many different cell types (18, 23, 49). However, in renal cells, the underlying mechanism that senses the change in osmolarity and/or cell volume to initiate volume regulation is not completely understood. We previously demonstrated, in a rat cortical collecting duct cell line (RCCD1) which exhibits many major functional properties of CCD (4, 10), that the presence of AQP2 in the apical membrane was crucial for a rapid activation of RVD. We also established that this fast activation of RVD is related to cystic fibrosis transmembrane conductance regulator (CFTR) and to barium-sensitive potassium channels (11). The investigation of whether aquaporins may form part of the cellular device for volume sensing and regulation is of physiological relevance. However, the exact molecular events involved in these responses are not yet known. Therefore, the aim of our present work was to investigate the signaling pathway that links AQP2 to the rapid RVD activation. Since it has been previously described that hypotonic conditions induce intracellular calcium ([Ca2+]i) increases in different cell types, we tested the hypothesis that AQP2 could have a role in activation of calcium entry by hypotonicity and its implication in cell volume regulation.
Using a fluorescent probe technique, we studied [Ca2+]i and cell volume changes in response to a hypotonic shock in two renal cell lines, one not expressing aquaporins (WT-RCCD1) (6) and another stably transfected with AQP2 (AQP2-RCCD1) that constitutively expresses AQP2 in the apical plasma membrane (11). The data presented here demonstrate that, in RCCD1 cells, the presence of AQP2 is crucial for the activation of [Ca2+]i increase by hypotonicity, which is necessary to control cell volume regulation. Furthermore, the effect was reduced by thapsigargin or by removal of extracellular Ca2+, indicating that the rise in [Ca2+]i reflects both influx from the extracellular medium and release from intracellular stores.
| MATERIALS AND METHODS |
|---|
|
|
|---|
AQP2-RCCD1 cells, stably transfected with cDNA coding for rat AQP2, were maintained in DM medium containing Geneticin (400 µg/ml, Life Technologies) as previously reported (11).
All experiments were performed on confluent cells, between passages 20 and 40, grown on coverslips during 2 or 3 days.
Cell volume changes. We measured cell volume in confluent wild-type (WT) and AQP2-RCCD1 cells grown on coverslips using a fluorescent probe technique. The use of fluorescent dyes to monitor cell volume was earlier described by other authors (8, 15, 50). We previously tested in RCCD1 cells the optimal conditions to collect fluorescence (F) (11). Briefly, we found that the best condition was to collect F from a small circular region (pinhole) of 3–1% of the total area of the cell, localized in the periphery. F from periphery was inversely proportional to the external osmolarity and showed a linear correlation with the relative external osmolarity (11). For thin optical section through the cell, a Nikon SFluor x40 numerical aperture 1.3 oil immersion objective lens was used. Coverslips were mounted on a chamber and loaded with 2 µM BCECF-AM (Molecular Probes) for 30 min at 37°C. The chamber was then placed on the stage of a Nikon TE-200 epifluorescence inverted microscope. Cells were subsequently bathed at 20°C in dye-free solution for at least 15 min before the experiment. During experiments, bathing solution was exchanged by aspirating the media and adding new media. This fluid exchange was completed within 5 s. F intensity was recorded by exciting BCECF at the isosbestic point (440 nm HBW: 10 nm), where the fluorochrome is pH insensitive.
F data were acquired every 10 s by use of a charge coupled device camera (Hamamatsu C4742–95) connected to a computer and the Metafluor acquisition program (Universal Imaging). The procedure to estimate cell water volume was previously described (11, 15). The change in cell water volume can be calculated as
![]() |
Osmotic water permeability and cell volume regulation.
The time course of V/V0 for each experiment was fitted with a single exponential function. The osmotic water permeability coefficient (Pf) was calculated from the exponential time constant (
) by the relation
![]() |
osM is the osmotic gradient, and VW is the partial molar volume of water (18 cm3/mol).
The RVD at 20 min, associated with the volumetric response of cells exposed to hypotonic medium, was calculated by the use of the following equation
![]() |
[Ca2+]i measurement. WT and AQP2-RCCD1 cells were incubated in 10 µM Fura 2-AM (Molecular Probes) for 60 min at room temperature and then washed to remove the excess dye. Pluronic F127 (0.2%; Molecular Probes) was used to dissolve the Fura 2-AM dye to prevent dye compartmentalization upon loading. The coverslips were again incubated in the experimental buffer for 15 min before the experiments. Fura-2 signal F was stimulated by dual-wavelength excitation at 340 and 380 nm. A 510 emission filter was used to collect Fura-2 signals at 10-s intervals. Ratios between the F intensity stimulated by 340/380-nm excitation were calculated.
[Ca2+]i was calibrated from maximum and minimum Fura-2 signals at the end of each experiment. Specifically, the bath solution was exchanged to 3.5 µM Ca2+ ionophore ionomycin (Sigma) with 1 mM Ca2+ in the experimental buffer to establish maximum Fura-2 signals and also to Ca2+-free experimental buffer with 100 µM EGTA (Sigma) and 3.5 µM ionomycin to establish minimum Fura-2 signals. [Ca2+]i was calculated as was previously reported (14)
![]() |
Solutions and chemicals. Isosmotic solution (osmolarity: 320 ± 4 mosM) contained (in mM) 90 NaCl, 10 NaHCO3, 5 KCl, 1 CaCl2, 0.8 MgSO4, 1 MgCl2, 100 mannitol, 20 HEPES, 5 glucose. Calcium-free solutions were made by adding EGTA (1 mM) and replacing CaCl2 by MgCl2.
Hyposmotic solutions (
osM = –100 mosM) were prepared by mannitol removal, thus maintaining the ionic strength. A more diluted solution was obtained by direct dilution of the isotonic solution with distilled water to reach a final osmolarity of 120 ± 3 mosM (
osM = –200 mosM).
The osmolalities were routinely measured by a pressure vapor osmometer (Werked). All solutions were titrated to pH 7.40 using Tris (Sigma) and bubbled with atmospheric air.
In some experiments, 100 µM gadolinium chloride (Sigma), 1 µM thapsigargin (Molecular Probes), 100 µM BAPTA-AM (Molecular Probes), and 100 µg/ml amphotericin B (Sigma) were used. BCECF-AM (2 mM), BAPTA-AM (6 mM), thapsigargin (2 mM), and amphotericin B (25 mg/ml) stock solutions were dissolved in DMSO and stored at –20°C until used.
Statistics. Values are reported as means ± SE, and n is the number of cells evaluated from three to six different experiments. For all comparisons, Student's t-test for unpaired data was applied and P < 0.05 was considered to be statistically significant.
| RESULTS |
|---|
|
|
|---|
[Ca2+]imax) was reduced to 27% in the absence of Ca2+ (Fig. 1D). In contrast, in WT-RCCD1 cells, the
[Ca2+]imax did not change and the decrease in the F intensity disappeared in a Ca2+-free medium (Fig. 1, C and D).
|
Calcium entry is required for rapid activation of RVD in AQP2-RCCD1 cells. Since we previously reported that AQP2 is crucial in the activation of rapid RVD mechanisms (11), we now evaluated whether calcium entry is required for this response. Figure 2A shows the time course of relative volume changes of AQP2-RCCD1 cells in control and Ca2+-free solutions. When a hyposmotic shock was performed, the kinetics between both conditions were different showing a decrease in RVD response in Ca2+-free experiments. The percentage of RVD at 20 min (% RVD20) was significantly lower in a Ca2+-free solution (Fig. 2C). This difference was not due to a modification of AQP2 permeability by calcium removal since Pf was not affected (Fig. 2D).
|
All these data strongly suggest that in AQP2-RCCD1 cells, Ca2+ entry triggered by hypotonicity participates in the rapid activation of RVD.
Gadolinium inhibits calcium entry and RVD in AQP2-RCCD1 cells.
One possible route for Ca2+ entry during cell swelling is via a stretch-activated channel (47). Thus, we investigated the effect of pretreatment of AQP2-RCCD1 cells with gadolinium (Gd3+), an inhibitor of stretch-activated channels (53), on the Ca2+ response to a hypotonic shock. Figure 3A shows that 100 µM Gd3+ inhibited the increase in [Ca2+]i produced by hypotonic exposure of AQP2-RCCD1 cells and reduced
[Ca2+]imax to 34% of the control.
|
These results are consistent with the hypothesis that a volume-activated and Gd3+-sensitive calcium channel mediates Ca2+ entry during swelling of AQP2-RCCD1 cells and that it has a key role in generating the RVD response.
Calcium release from intracellular stores is also involved in the [Ca2+]i increase after a hypotonic shock in AQP2-RCCD1 cells. The fact that exposure to a Ca2+-free solution or to Gd3+ did not completely abolish the [Ca2+]i response to hypotonic shock indicates that extracellular Ca2+ is not the unique source for the increased [Ca2+]i. To test whether [Ca2+]i stores are also involved in generating the volume-sensitive Ca2+ signal, we investigated the effect of cell swelling on [Ca2+]i and RVD in AQP2-RCCD1 cells after treatment with thapsigargin, an inhibitor of the Ca2+-ATPase pump which avoids the refilling of the stores (43).
Figure 4A shows the time course of the relative F in AQP2-RCCD1 cells incubated with thapsigargin (1 µM) or with the vehicle (DMSO) during 10 min before the addition of the hypotonic solution. It can be observed that there was a partial but significant attenuation of the rise following the hypotonic shock, in cells where stores were previously empty. No extra effects were observed incubating cells with thapsigargin during 20 min (Fig. 4B). In addition, there was an almost total inhibition of the rise in
[Ca2+]imax when cells were pretreated with thapsigargin in a Ca2+-free media. Similar results were obtained when [Ca2+]i was buffered incubating cells with 100 µM BAPTA/AM (Fig. 4B). Altogether, these data suggest that in AQP2-RCCD1 cells, the rise in [Ca2+]i elicited by hypotonicity is due to both intracellular release and influx from extracellular medium. Additionally, AQP2-RCCD1 cells preincubated with thapsigargin, thapsigargin in a Ca2+-free media, or buffered with BAPTA showed a significant reduction in the % RVD20 (Fig. 4C).
|
[Ca2+]imax were observed between control and thapsigargin-treated WT-RCCD1 cells (vehicle: 14 ± 4 nM; n = 60 vs. thapsigargin: 12.5 ± 2 nM; n = 60, not significant). All these results demonstrate that emptying the thapsigargin-sensitive stores before a hypotonic shock reduces the swelling-induced [Ca2+]i increase and concomitantly inhibits RVD, only in AQP2-RCCD1 cells.
To test whether an increase in the magnitude of the osmotic gradient could affect the characteristics of the Ca2+ response, cells were exposed to a more diluted solution (
osM = –200 mosM). The effects of varying osmolarity in WT and AQP2-RCCD1 cells are shown in Fig. 5. Note that the kinetics between 100 and 200 mosM are quite different in AQP2 but not in WT-RCCD1 cells (Fig. 5A). In AQP2 cells this difference disappears when cells were previously incubated with thapsigargin (Fig. 5B). Since the kinetics observed in AQP2 cells at a high osmotic gradient show a sustained rise similar to that generally described when depletion of stores induces a capacitative Ca2+ entry (CCE) (37), we tested whether the depletion of [Ca2+]i stores with thapsigargin promotes a CCE. A typical protocol used to distinguish between stored Ca2+ release and Ca2+ influx is shown in Fig. 6. Store depletion is first induced by adding thapsigargin (1 µM) in Ca2+-free medium, followed by reintroduction of Ca2+ (final [Ca2+]i = 1 mM). Figure 6A shows that preincubation of AQP2-RCCD1 cells with thapsigargin (1 µM) in an isotonic calcium-free medium induces a transient increase in [Ca2+]i (peak 1). A higher rise occurred (peak 2) when Ca2+ was restored to the medium. When the same protocol was performed with the vehicle (DMSO), no increase appeared in Ca2+-free medium; however, a small increase was observed when extracellular calcium was added. This thapsigargin-independent Ca2+ increase could be explained by partial store depletion during exposure to Ca2+-free solution or possible damage of some cells. Similar results were obtained in WT-RCCD1 cells when the same protocol was applied. Figure 6B shows the
[Ca2+]imax, calculated when Ca2+ was restored to the medium (peak 2), in AQP2-and WT-RCCD1 cells. The difference between thapsigargin and vehicle represents the CCE. The fact that these results are comparable in WT and AQP2 cells indicates that both cell lines have thapsigargin-sensitive stores and both are able to induce a CCE. However, when a hyposmotic shock is performed, the induction of a CCE appears to be dependent on the magnitude of the osmotic gradient and the expression of AQP2 in the plasma membrane.
|
|
osM = –100 mosM). There were no differences between vehicle and amphotericin-treated cells in the isotonic condition; however, when cells were exposed to the hypotonic shock, amphotericin-treated cells increased their volume more rapidly and reached a slightly higher maximal volume. This higher maximal value reach in amphotericin-treated cells could be due to the fact that in addition to water movement small ions are entering the cell driving additional water. Figure 7B shows the osmotic permeability, estimated after the exposure to the hypotonic media, of WT-RCCD1 cells treated with amphotericin or vehicle. To compare amphotericin-treated WT with AQP2-RCCD1 cells, we also performed experiments in AQP2 cells preincubated with the vehicle (Fig. 7, B–D). Preincubation of WT-RCCD1 cells with amphotericin B strongly produced an increase in Pf values; however, no changes were observed in % RVD20 (Fig. 7C) nor in
[Ca2+]imax (Fig. 7D). All these results strengthen the idea that AQP2 per se is necessary to produce an increase in [Ca2+]i critical to activate rapid RVD mechanisms.
|
| DISCUSSION |
|---|
|
|
|---|
In AQP2-RCCD1 cells, osmotic swelling produces a transient increase in the [Ca2+]i but it is insignificant in cells not expressing AQP2. The rise in [Ca2+]i may be due to the activation of Ca2+-permeable channels in the cell membrane and/or Ca2+ release from intracellular stores. Our finding, that exposure of AQP2-RCCD1 cells to a Ca2+-free solution or to 100 µM Gd3+ did not completely abolish the [Ca2+]i response, indicates that extracellular Ca2+ is not the unique source for the increased [Ca2+]i following cell swelling. Furthermore, treatment with thapsigargin reduced the increase in [Ca2+]i produced by a hypotonic shock and treatment with thapsigargin in a Ca2+-free solution almost abolished the response. All these results indicate that the [Ca2+]i increase was due to both extracellular [Ca2+] entry and [Ca2+]i release. On the other hand, the hypotonic-induced fall in [Ca2+]i may be due to Ca2+ reuptake into stores and/or to Ca2+ extrusion to extracellular medium. A calcium-binding protein, like calcyclin, could also be involved. In the kidney, calcyclin was localized all along the collecting duct and in RCCD1 cells calcyclin was evidenced in the cytoplasm (7). Our observation that in WT cells there is a decrease in [Ca2+]i, below basal levels, without an important increase, suggests that cell swelling probably induces the stimulation of calcium decreasing mechanisms independently of the activation of calcium increasing mechanisms. Moreover, this reduction in [Ca2+]i appears to be dependent on the presence of extracellular Ca2+.
The increase in [Ca2+]i produced by a hypotonic shock in AQP2-RCCD1 cells is consistent with the rise reported in many renal cells (1, 17, 26, 41, 45–47). In the rat CCD it has been described that Ca2+ entry is stimulated by hypotonic cell swelling and strongly dependent on extracellular Ca2+ but does not involve a store release of Ca2+ (17). However, it has been reported that in microperfused CCD, stretch-induced rise of [Ca2+]i includes the release from internal stores and influx from the extracellular space (51). In renal collecting duct CD8 cells and in rat inner medullary collecting duct cells, hypotonic shock is accompanied by a Ca2+ release from intracellular stores, as well as by calcium entry from the extracellular fluid (41, 45). However, none of these works evaluated whether these calcium responses were affected by AQP2 expression.
Our results also show that the increase in Ca2+ due to extracellular Ca2+ entry and to the release from intracellular stores that occurs in AQP2 cells is critical in the activation of rapid RVD mechanisms. We previously demonstrated that in AQP2-RCCD1 cells a K+ channel was implicated in this volume response. In CCD cells apical low-conductance SK and high-conductance Ca2+-activated BK channels are present (24). Thus, we propose that the increase in [Ca2+]i is necessary to turn on a Ca2+-activated K+ channel, required for rapid RVD.
Our most important finding is that the expression of AQP2 in the cell membrane is critical to produce the increase in [Ca2+]i which is necessary to activate RVD. The inhibitory effect of Gd3+ on the swelling-induced [Ca2+]i increase and on RVD suggests the involvement of a mechanosensitive calcium channel in this response. Because osmotic swelling is a kind of mechanical stretch, we suggest that the opening of stretch-activated Ca2+ channels contributes to hypotonic-induced Ca2+ influx which is necessary for the activation of volume-regulated channels in AQP2 but not in WT cells. However, the participation of stretch-activated channels in the fine-tuning of cell volume has been debatable since considerable stretch is required to activate these channels (38). We suggest that these channels may represent a last line of defense against rapid cell swelling that may occur when an AQP is present in the membrane but are not involved in the response to slow changes of cell volume as it occurs in the absence of AQPs (WT cells).
Since variation of extracellular calcium concentration has been described to modify the water permeability mediated by AQP0 (30, 31), it is important to discard the possibility that the difference observed in RVD between Ca2+-containing and Ca2+-free solutions was not due to the modification of AQP2-mediated water permeability. Several studies provided evidence for a role of intracellular and extracellular Ca2+ mobilization in vasopressin-mediated AQP2 trafficking (34). However, to our knowledge, no studies exist evaluating the direct effect of Ca2+ on the osmotic water permeability of cells that already express AQP2 in the apical membrane (independently from AQP2 traffic). In the present work, we demonstrated that the osmotic water permeability through AQP2 was not affected by removing extracellular calcium (Fig. 2D).
We hypothesize that the presence of AQP2 in the cell membrane facilitates the activation of a Ca2+ entry path, nevertheless the molecular identity of this channel remains to be defined. Members of the transient receptor potential (TRP) superfamily of ion channels such as TRPV4, TRPC6, TRPP1, and TRPP2 are attractive candidates (36). All these channels have been proposed to be implicated in the flow-induced increase in [Ca2+]i observed in CCD cells (24, 52). TRPV4 channel is highly expressed in the distal nephron and collecting duct (44), although its precise cellular localization remains controversial. The TRPV4 channel not only responds to mechanical stress but may also function as an osmosensor in CCD cells (24, 52). Our finding that AQP2 is necessary for the hypotonic induction of cytosolic Ca2+ increase is in agreement with the results recently showed for AQP5 in salivary gland epithelial cells (25). The authors demonstrated that AQP5 is required for the activation of TRPV4 by hypotonicity and that both proteins control RVD. In a recent study, Taguchi and collaborators (40) found in the epithelial layer of the human endolymphatic sac a similar distribution pattern of AQP2, V2-receptor, and TRPV4 to the one observed in the kidney and they proposed that AQP2 and TRPV4 channels are expected to be closely interconnected. TRPC6 is a nonselective cation channel that directly senses membrane stretch induced by mechanical or osmotic stimuli (39) and colocalizes with AQP2 (12). In addition, it has been recently reported that AQP2 physically associates with TRPC3 in cells of the rat renal collecting duct (13). Our findings that in WT-RCCD1, amphotericin B increases Pf, but fails to induce an increase in [Ca2+]i, strengthen the idea that a high osmotic permeability is not enough to activate calcium increase mechanisms. Although this situation is not exactly the same as that in which AQP2 is expressed in the membrane, it causes a strong increase in cell volume and membrane stretch. Therefore, we propose that AQP2 participation in the hypotonic calcium increase may be related to some specific protein interaction or coupling. To investigate whether TRPs are the proteins implicated in the observed Ca2+ entry in AQP2 cells is part of our future goal.
Our results indicate that AQP2 is also involved in the activation of Ca2+ release from intracellular stores since pretreatment of AQP2-RCCD1 cells with BAPTA-AM or with thapsigargin prevents the hypotonic-induced [Ca2+]i transient increase and abolishes rapid activation of RVD. Although the signaling pathways involved in Ca2+ release from intracellular stores were not investigated here, cumulative evidence supports the idea that some of these paths are affected by membrane stretch. For instance, the activities of phospholipase C and D have been shown to be susceptible to stretch (21) and cell swelling or membrane stretch has been proposed to induce IP3 production (9, 21). Other reports propose that mobilization of Ca2+ from intracellular stores may also result from IP3-independent processes such as prior entry of Ca2+ through stretch-sensitive calcium channel, with subsequent Ca2+-induced Ca2+ release or by a direct action of membrane stretch on the stored Ca2+ channel (28). In addition, depletion of Ca2+ stores could induce a Ca2+ influx to replenish the stores, a process called the CCE (20, 37). As we found that AQP2 and WT-RCCD1 cells are both able to produce a CCE, but the hypotonic shock only induces a CCE in AQP2 cells, becoming apparent when a higher gradient is applied, we suggest that a threshold is required to trigger this response.
Altogether our present results let us propose that when a hyposmotic shock is applied in a cell that expresses AQP2, the tension in cell membrane rapidly increases producing the activation of a calcium entry, by a stretch-activated channel, and either directly or indirectly triggers Ca2+ release from stores. These two sources of calcium may act synergistically to produce the level of [Ca2+]i increase necessary to activate a K+ channel. Cells not expressing AQP2 (WT-RCCD1) show a diminutive increase in [Ca2+]i and no rapid activation of RVD. Hence, the presence of AQP2 in the cell membrane would be crucial for the cell to sense the osmotic shock and subsequently produce the activation of signaling cascades leading to rapid RVD. This most likely indicates that AQP2, in concordance with other proteins, may form part of the cellular device for volume sensing and regulation.
| GRANTS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
U. Hasler Controlled aquaporin-2 expression in the hypertonic environment Am J Physiol Cell Physiol, April 1, 2009; 296(4): C641 - C653. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. K. Hoffmann, I. H. Lambert, and S. F. Pedersen Physiology of Cell Volume Regulation in Vertebrates Physiol Rev, January 1, 2009; 89(1): 193 - 277. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |