HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Salathe, M. Articles by Bookman, R. J. Search for Related Content PubMed PubMed Citation Articles by Salathe, M. Articles by Bookman, R. J.

Agonist-stimulated calcium decreases in ovine ciliated airway epithelial cells: role of mitochondria

Matthias Salathe*†, Pedro I. Ivonnet†, Thomas Lieb* and Richard J. Bookman†

MS 1445 Received 24 July 2000; accepted after revision 6 October 2000.

	ABSTRACT

Top Abstract Introduction Methods Results Discussion References

1. In ovine ciliated tracheal epithelial cells, acetylcholine (ACh) activates signal transduction pathways that not only transiently increase cytoplasmic Ca²⁺ ([Ca²⁺]_i) but also actively lower [Ca²⁺]_i. The pathway for decreasing [Ca²⁺]_i is clearly revealed after depletion of intracellular Ca²⁺ stores by thapsigargin (Tg), 2,5-di-(tert-butyl)-1,4-benzohydroquinone or NiCl₂. Measurements with microinjected fura-2 excluded a [Ca²⁺] measurement artefact.

2. A four-compartment model to simulate calcium transients in non-excitable cells (consisting of a plasma membrane Ca²⁺ pump and channel; Ca²⁺ store with pump and channel; and cytosolic Ca²⁺ buffer) could not account for the observed [Ca²⁺]_i decrease. We therefore explored, by simulation and experimentation, several different mechanisms that could account for it.

3. The ACh-stimulated [Ca²⁺]_i decrease was not due to an inhibition of Ca²⁺ influx (Ca²⁺ channel blockers or absence of extracellular calcium had no effect), activation of a plasma membrane Ca²⁺-ATPase (two inhibitors, vanadate (30 mM) and lanthanum (10 mM), had no effect) or inhibition of the Na⁺-Ca²⁺ exchanger (replacing extracellular Na⁺ with N-methylglucamine had no effect).

4. The application of mitochondrial uncouplers (5 µM CCCP or 5 µM FCCP), eliminated the ACh-induced [Ca²⁺]_i decrease. Addition of CCCP at the nadir of the decrease restored intracellular calcium levels of Tg-treated cells to baseline faster than controls not exposed to mitochondrial uncouplers. CCCP application to naïve cells did not block the ACh-induced transient increase in [Ca²⁺]_i.

5. These data suggest that ACh-induced [Ca²⁺]_i decreases in ciliated cells are caused by stimulated Ca²⁺ uptake into mitochondria.

	INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

It has been well established that a single G-protein coupled receptor can simultaneously activate multiple effectors by signalling through multiple G-proteins (Mohuczy-Dominiak & Garg, 1992a,b; Felder, 1995; Ecelbarger et al. 1996), by signalling through distinct actions of the monomer and the dimer G-protein subunits (e.g. Kofuji et al. 1995), and finally by signalling through direct interaction with a variety of effectors through the C-terminal domains (e.g. Hall et al. 1998a,b; Luttrell et al. 1999).

We have shown that acetylcholine (ACh) not only transiently increases both the cytosolic Ca²⁺ concentration ([Ca²⁺]_i) and ciliary beat frequency (CBF) of single ciliated ovine epithelial cells but also consistently decreases [Ca²⁺]_i and CBF after depletion of intracellular, inositol 1,4,5-trisphosphate (IP₃)-sensitive Ca²⁺ stores by thapsigargin (Tg) (Salathe & Bookman, 1995; Salathe et al. 1997). Both increases and decreases of [Ca²⁺]_i were mediated by M3 receptors (Salathe et al. 1997). Data revealing simultaneous activation of signalling pathways to increase and decrease [Ca²⁺]_i have been established in the literature for more than 10 years. An early report from platelets indicated that thrombin can act in such a way and implicated a Ca²⁺-ATPase in lowering [Ca²⁺]_i (Rink & Sage, 1987). Reports from other cell types followed, which included pancreatic acinar cells (Muallem et al. 1988; Tepikin et al. 1992), macrophages (Randriamampita & Trautmann, 1990), and hepatocytes (Duddy et al. 1989). In pancreatic acinar cells, the stimulated Ca²⁺ efflux was directly proportional to the elevated [Ca²⁺]_i levels (Muallem et al. 1988; Tepikin et al. 1992), suggesting a [Ca²⁺]_i-activated extrusion mechanism. In murine peritoneal macrophages, platelet-activating factor (PAF) increased Ca²⁺ extrusion presumably via arachidonic acid stimulation of a plasma membrane Ca²⁺-ATPase (Randriamampita & Trautmann, 1990). The extrusion pathway could only be activated when [Ca²⁺]_i was elevated above its original baseline by either thapsigargin (Tg) or PAF and this activation never resulted in a [Ca²⁺]_i decrease below the original baseline (Randriamampita & Trautmann, 1990). Similar findings were obtained in hepatocytes where vasopressin, angiotensin II and ATP all produced a rapid return of [Ca²⁺]_i to baseline after [Ca²⁺]_i elevation by 2,5-di-(tert-butyl)-1,4-benzohydroquinone (BHQ), which acts to empty IP₃-sensitive intracellular Ca²⁺ stores similar to Tg (Duddy et al. 1989). The hormones apparently stimulated an undefined efflux pathway to remove BHQ-mobilized Ca²⁺ from the cytosol.

The results in ovine ciliated epithelial cells (Salathe et al. 1997), however, differed somewhat from these earlier reports. In many cells, the [Ca²⁺]_i decreases went below baseline, indicating that an increase in [Ca²⁺]_i was not necessary for activation of the lowering pathway. In addition, in some ciliated cells (with apparently replete internal Ca²⁺ stores and no exposure to Tg), the [Ca²⁺]_i lowering pathway seemed to dominate. These cells responded to ACh only with a small and rapidly transient Ca²⁺ and CBF increase followed by a longer lasting decrease of both signals below the original baseline. In addition, a few cells showed no initial [Ca²⁺]_i rise, but only undershot the baseline in response to ACh.

With this report, we continue to analyse the [Ca²⁺]_i lowering pathway in ciliated cells. We ruled out a measurement artefact by showing that the [Ca²⁺]_i decreases occurred whether cells were loaded with fura-2 AM or injected with fura-2. A four-compartment model (Korngreen et al. 1997) to simulate calcium transients in non-excitable cells (consisting of a plasma membrane Ca²⁺ pump and channel, Ca²⁺ store with pump and channel, and cytosolic Ca²⁺ buffer) could not account for the observed [Ca²⁺]_i decrease. We therefore explored, by simulation and experimentation, several different mechanisms to account for it; these potential mechanisms included: (1) inhibition of Ca²⁺ influx; (2) activation of Ca²⁺ extrusion via a plasma membrane Ca²⁺-ATPase; (3) activation of Ca²⁺ extrusion via the Na⁺-Ca²⁺ exchanger; (4) activation of a Tg-insensitive uptake mechanism into internal stores; and (5) activation of Ca²⁺ uptake into mitochondria. As it turns out, stimulated uptake of Ca into mitochondria seems to be responsible for the ACh-induced [Ca²⁺]_i decrease in ovine ciliated cells.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Preparation of tracheal epithelial cultures

Primary cultures of sheep tracheal epithelial cells were prepared as previously described (Salathe & Bookman, 1995). Briefly, tracheas were removed immediately from ewes killed by an overdose of pentobarbital (30 mg kg^-1 intravenously, according to NIH and local animal care use committee approved protocols). The mucosa, was dissected from the underlying cartilage under sterile conditions and incubated in 0·05 % protease (Sigma, type 14) in Dulbecco's modified Eagle's medium overnight at 4°C. Epithelial cells, released by vigorous shaking, were harvested by centrifugation and plated on collagen-coated glass coverslips (human placental collagen, type VI, Sigma) at a density of 0·5 × 10⁶ cells cm^-2 in a minimal volume of 100 µl cm^-2 (equal to 1 ml per 35 mm dish). The culture medium consisted of 50 % DMEM, 50 % Ham's F-12 nutrient supplemented with insulin (10 µg ml^-1), transferrin (5 µg ml^-1), hydrocortisone (0·36 µg ml^-1), tri-iodothyronine (20 ng ml^-1), endothelial cell growth supplement (7·5 µg ml^-1), penicillin (100 U ml^-1), and streptomycin (100 µg ml^-1). The medium was exchanged every other day. Cultures were used 2-14 days after plating.

Measurement of [Ca²⁺]_i

Incubation protocol for dye-loading. After rinsing coverslips several times with Hanks' balanced salt solution buffered with 10 mM Hepes and adjusted to pH 7·4 (referred to as HBSS), the cells were loaded with 4 µM fura-2 AM, 2·5 % fetal calf serum (Hyclone, UT, USA), in HBSS for 45 min at room temperature on a rocking table.

Microinjection of ciliated cells. Micropipettes were pulled on a Sutter P-97 Flaming-Brown microelectrode puller and filled with 50 µM fura-2 (K⁺ salt) in 100 mM potassium glutamate buffered with 10 mM Hepes and adjusted to pH 7·2. Injection of ciliated cells was performed with continuous flow through the pipette. The cells were impaled for a short period of time (1-2 s) during which they were bathed in a Ca²⁺-free solution. Using this technique, 3 of 10 attempted injections resulted in ciliated cells that were filled with dye and which continued to beat.

Imaging hardware and software. With fura-2 as the [Ca²⁺]_i indicator, 10 nm wide filters centred on 340 and 380 nm were used (Chroma Technology Corp., Brattleboro, VT, USA). Ratiometric Ca²⁺ imaging was performed with an UV excitation system (Horrigan & Bookman, 1995) that permits the output of a 100 W mercury lamp to be rapidly switched to different interference filters in less than 1 ms. The emitted fluorescence (light between 510 and 600 nm) was visualised through a ×40, 1·3 NA objective lens (CF Fluor DL series, Nikon Inc.) and directed to a KS-1381 microchannel plate image intensifier (Opelco Inc., Washington, DC, USA), coupled to a Hamamatsu 2400 CCD camera. On-line processing of the signal was accomplished with an IC300 Imaging Workstation (Inovision Corp., Durham, NC, USA) using Inovision's 'ratio-tool' software running on a SunSparc 10. Individual cells were identified as regions of interest (ROIs). The fluorescence ratio within each ROI (340/380 excitation) was computed on a pixel-by-pixel basis, pixels which failed to reach a threshold value were rejected, averaging 16 video frames at each of the two excitation wavelengths. Ratios were computed every 10 s, or more frequently as needed (every 4 s during the time of exposure to different stimuli).

Extracellular calibration and computation of free Ca²⁺. An intracellular calibration of the fura-2 signal could not be obtained in ciliated cells, because the effects of the ionophore ionomycin (10 µM; Calbiochem, CA, USA) were overcome by these cells. A simpler calibration procedure was thus adopted. The fluorescence intensity at each wavelength was measured with a calcium-free and a saturating calcium aliquot of buffered 10 µM fura-2 (K⁺ salt) in solution containing 150 mM KCl and 10 mM Hepes and adjusted to pH 7·4. With these values we calculated the fluorescence ratios in the absence (R_min) and presence of saturating Ca²⁺ (R_max) and the buffering capacity () which were used to transform the data into Ca²⁺ concentrations using the equation of Grynkiewicz et al. (1985) assuming a dissociation constant (K_d) of 250 nM. The reported Ca²⁺ values are thus only an approximation since the true K_d of the dye in the cytoplasm of these cells is unknown and no corrections for cytoplasmic viscosity have been made (Poenie, 1990).

Modelling of Ca²⁺ transients

A four-compartment model to simulate Ca²⁺ transients was adapted from a published paper (Korngreen et al. 1997). This model has been developed for non-excitable cells and consists of extracellular space, cytosol, Ca²⁺ buffer and Ca²⁺ store. Ca²⁺ exchange between these compartments is provided by several pumps and channels and the initial simulation parameters have been directly adopted from this paper. These parameters are: plasma membrane channel (maximal rate of calcium flux, 0·000158 s^-1), plasma membrane pump (maximum rate, 1·5 µM s^-1; K_d, 0·6 µM), store channel (rate of calcium flux, 7·5 s^-1), store pump (maximum rate, 500 µM s^-1; K_d, 0·1 µM), association with buffer (rate, 601 µM^-1 s^-1), dissociation from buffer (97 s^-1), and a stable extracellular Ca²⁺ concentration of 1500 µM. The initial [Ca²⁺] in the cytoplasm was 0·108 µM, and in the store 180 µM. The free buffer concentration was assumed to be 180 µM. In contrast to the originally published model with on/off activation parameters, ACh stimulation was modelled by increasing the basal flux (towards the cytosol) through the plasma membrane and store channels in such a way as to simulate the actual occurrence in vivo. The stimulation was initially rapidly increased to a preset maximum (for details see Korngreen et al. 1997) and then relaxed back to its basal activity within 75 s using an exponential decay function.

Thapsigargin exposure was simulated by lowering the maximum rate of the store pump to 5 µM s^-1. In addition, the ensuing mandatory Ca²⁺ influx through the plasma membrane channels (depletion activated Ca²⁺ influx) was modelled by increasing the basal Ca²⁺ flux through this channel exponentially from the baseline modifier of 0·2 to 0·8 following the example of Korngreen et al. (1997).

Statistics

For statistical analysis, one-way analysis of variance was used to compare the means of more than two groups using JMP software from SAS Institute Inc. (Cary, NC, USA). If a significant difference was found, a group-by-group comparison was performed using the Tukey-Kramer honestly significant difference test. Two groups were compared using Student's unpaired t test. P < 0·05 was considered significant.

Experimental protocols

As mentioned in the introduction, we tested five hypotheses. Using the multicompartment model, we first carried out simulations to evaluate the hypotheses theoretically. We then carried out direct measurements.

For the experiments, a [Ca²⁺]_i baseline was recorded for several minutes from single fura-2 loaded cells bathed in HBSS. To avoid mechanically triggered [Ca²⁺]_i transients in ciliated cells (Sanderson & Dirksen, 1989; Sanderson et al. 1990), solution exchanges were carried out gently, typically taking 20-30 s for a 10 ml wash (to fully exchange the 0·5 ml working volume of the chamber). The internal Ca²⁺ stores were depleted by exposure to 1 µM Tg at least 15 min before any further manipulation; this exposure period depletes internal Ca²⁺ stores by > 95 % (Mathes & Thompson, 1995). Unless stated otherwise, ACh was added to the bath at a final concentration of 10 µM for a period of 2 min.

Hypothesis 1 (inhibition of Ca²⁺ influx leads to a decrease in [Ca²⁺]_i) was tested using two different strategies: (a) by eliminating extracellular Ca²⁺ before ACh application and (b) by assessing the influence of different Ca²⁺ channel blockers on the drop response. For the first approach, the bathing fluid of cells with depleted internal Ca²⁺ stores was replaced with 0 Ca²⁺-HBSS at least 2 min before exposure to 10 µM ACh in 0 Ca²⁺-HBSS. The '0 Ca²⁺'-HBSS solution was prepared by adding 1 mM EGTA to divalent cation-free HBSS, restoring the free Mg²⁺ concentration (0·9 mM) according to Fabiato & Fabiato (1978) resulting in a calculated free Ca²⁺ concentration of < 70 nM. For the second approach, the calcium channel blockers cadmium (100 µM) and nickel (100 µM) were added to the bathing fluid at least 2 min before exposure to ACh.

Hypothesis 2 (activation of Ca²⁺ extrusion via a plasma membrane Ca²⁺-ATPase) was tested by two different means. Since plasma membrane Ca²⁺-ATPase activity can be stimulated by calmodulin (Carafoli & Stauffer, 1994), we tested the ability of two different calmodulin antagonists (phenoxybenzamine (10 µM) and trifluoperazine (10 µM)) to prevent the ACh-induced drop in [Ca²⁺]_i in Tg-pretreated cells. The antagonist was added to the bathing fluid at least 2 min before exposure to ACh. Since plasma membrane Ca²⁺-ATPases are also activated through pathways that do not involve calmodulin, we inhibited them directly by high concentrations of orthovanadate (30 mM) or lanthanum (10 mM). Vanadate is an unspecific inhibitor of many ATPases, including the plasma membrane Ca²⁺-ATPase, the endoplasmic reticulum Ca²⁺-ATPase and dynein. Lanthanum, on the other hand, although unspecific in the test tube, is a relatively selective blocker in intact cells since it does not enter the cell and thus only gains access to ATPases in the plasma membrane (Carafoli & Stauffer, 1994).

Hypothesis 3 (activation of Ca²⁺ extrusion via the Na⁺-Ca²⁺ exchanger) was tested by eliminating Na⁺ from the bathing fluid before application of ACh. This was achieved by replacing Na⁺ with N-methylglucamine (KCO₃ and K₂HPO₄ replaced NaCO₃ and Na₂HPO₄, respectively; the osmolarity was restored to 297 mosmol l^-1 with glucose). Na⁺-free HBSS replaced the regular bath solution before or after Tg treatment and 5-30 min before application of 10 µM ACh.

Hypothesis 4 (activation of Ca²⁺ uptake into a Tg-insensitive store) was tested during the experiments for the plasma membrane Ca²⁺-ATPase. As previously mentioned, orthovanadate is an unspecific inhibitor of many ATPases. Thus, these experiments were also used to assess whether an intracellular Ca²⁺-ATPase was stimulated by ACh by comparing the results obtained using orthovanadate with those using La³⁺ (see above).

Hypothesis 5 (activation of Ca²⁺ uptake into mitochondria) was tested using the mitochondrial uncouplers carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP) and carbonyl cyanide m-chlorophenylhydrazone (CCCP). These uncoupling agents (5 µM each) were added to the bathing solution 5 min before ACh exposure or at the nadir of the [Ca²⁺]_i decrease. To show that the effect of CCCP on the muscarinic signalling pathway was not unspecific, naïve cells (not treated with Tg) were exposed for 1 min to 5 µM CCCP before the application of 10 µM ACh plus 5 µM CCCP.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

General characteristics of the [Ca²⁺]_i decrease

As mentioned above, some ciliated cells (with apparently replete internal Ca²⁺ stores) responded to 10 µM ACh with a short, transient [Ca²⁺]_i increase followed by a longer lasting decrease of [Ca²⁺]_i below the original baseline (Fig. 1A and B). A more typical response of a single ciliated cell to 10 µM ACh before and after treatment with 1 µM Tg, however, is illustrated in Fig. 1C. Exposure to 1 µM Tg caused a transient increase in the ratio signal. Application of ACh, at least 15 min after the initial exposure to Tg, caused a decrease in [Ca²⁺]_i of 33 ± 2 nM from a post-Tg baseline of 125 ± 5 nM in all of the cells (mean ± S.E.M., n = 91 cells from 10 coverslips). In 60 % of the cases, the response to ACh decreased [Ca²⁺]_i to a level below the pre-Tg level.

		View larger version [in this window] [in a new window]
Figure 1. ACh-induced Ca2+ decreases in ciliated cells A, the [Ca2+]i response to 10 µM ACh seen in a few ciliated cells where a short, transient [Ca2+]i increase was followed by a transient decrease below the original baseline. B, very few cells showed no initial [Ca2+]i increase with 10 µM ACh application, but only undershot the baseline. C, a typical [Ca2+]i transient of a ciliated cell in response to 10 µM ACh. Thapsigargin (Tg, 1 µM) also caused a transient increase in [Ca2+]i which relaxed back to a plateau level above baseline. A second application of 10 µM ACh caused a decrease in [Ca2+]i below the pre-Tg baseline. D, a cell injected with the free acid of fura-2, which assures cytoplasmic measurement of [Ca2+]i, revealed responses to 10 µM ACh before and after 1 µM Tg similar to those of the cell loaded with fura-2 AM (C).

To make sure that the [Ca²⁺]_i decrease was not due to loading of fura-2 AM into cellular compartments (e.g. measuring the [Ca²⁺] of the endoplasmic reticulum stores, Golovina & Blaustein, 1997), we injected ciliated cells with the free acid fura-2. Upon ACh exposure, these cells (n = 5) still revealed a [Ca²⁺]_i decrease of 22 ± 3 nM from a baseline of 94 ± 3 nM (Fig. 1D). We therefore continued to use fura-2 AM loaded cells for the remainder of the experiments.

The ACh-stimulated decrease in [Ca²⁺]_i could be repetitively provoked. Using 25 cells, the [Ca²⁺]_i decrease was observed three times with repeated 2 min exposures to 10 µM ACh, 15 min apart from each other (Fig. 2A). In order to ensure that the observed [Ca²⁺]_i decrease is not dependent on the pretreatment of the cells with Tg, internal Ca²⁺ stores of ciliated cells were depleted by another inhibitor of the endoplasmic reticulum Ca²⁺-ATPase, 2,5-di-(tert-butyl)-1,4-benzohydroquinone (BHQ) (Fig. 2B). Upon exposure to 20 µM BHQ, [Ca²⁺]_i increased transiently then relaxed back to 116 ± 6 nM. Exposure to 10 µM ACh then decreased [Ca²⁺]_i to 86 ± 5 nM. The [Ca²⁺]_i decrease was not significantly different from the decrease seen with Tg pretreatment (P > 0·05).

Both of the methods designed to reveal the [Ca²⁺]_i lowering pathway presented thus far relied on pharmacological inhibition of endoplasmic reticulum Ca²⁺-ATPase. To empty internal Ca²⁺ stores by different means, ciliated epithelial cells were exposed multiple times to 10 µM ACh (to produce Ca²⁺ release from stores) in the presence of 100 µM NiCl₂ to prevent Ca²⁺ influx (which is necessary to replenish the Ca²⁺ stores). As seen in Fig. 2C, 10 µM ACh initially induced a transient [Ca²⁺]_i increase in the absence of NiCl₂; however, after depletion of the stores in the presence of NiCl₂, 10 µM ACh induced a transient [Ca²⁺]_i decrease (after a brief initial increase) which did not recover fully in the continued presence of NiCl₂. After removing NiCl₂, the baseline [Ca²⁺]_i recovered partially, however, ACh still provoked only short transient increases followed by significant decreases in [Ca²⁺]_i, which is most probably due to incomplete refilling of the internal Ca²⁺ stores. It should be noted here that of the 65 cells tested, 50 (77 %) responded with the pattern shown in Fig. 2C; the rest did not reveal any [Ca²⁺]_i drops but continued to show only transient [Ca²⁺]_i increases in the presence of NiCl₂ although with lower peak amplitude compared with the responses in the absence of NiCl₂.

		View larger version [in this window] [in a new window]
Figure 2. Features of ACh-induced [Ca2+]i decreases in ciliated cells A, [Ca2+]i responses to 10 µM ACh could be repetitively provoked in a ciliated cell. B, to deplete internal Ca2+ stores other than by Tg, ciliated cells were treated with 20 µM BHQ which also lead to a transient [Ca2+]i increase. ACh (10 µM) then lead to a [Ca2+]i decrease, similar to that seen after Tg exposure. C, to deplete internal Ca2+ stores in a third way, cells were treated with 100 µM NiCl2, a broad Ca2+ channel blocker, thereby preventing the refilling of Tg-sensitive stores with Ca2+. NiCl2 itself had no effect on [Ca2+]i, but after preventing Ca2+ influx, ACh exposure now lead to only a small [Ca2+]i increase followed by a decrease below the pre-ACh baseline which never recovered in the continued presence of NiCl2. The [Ca2+]i response recovered to baseline after wash out of NiCl2 (right), however, the ACh transients never fully recovered to pre-NiCl2 conditions.

Modelling of Ca²⁺ transients

If an agonist causes a decrease in [Ca²⁺]_i, Ca²⁺ must either be buffered in the cytoplasm or move to a different compartment. Extrusion, reuptake and/or buffering should be able to account for the decrease. In order to distinguish possible mechanisms, we used a previously published model to simulate Ca²⁺ transients in non-excitable cells (Korngreen et al. 1997). This model assumes four compartments that participate in Ca²⁺ handling: the cytosol, a Ca²⁺ buffer in the cytosol, a single intracellular Ca²⁺ store, and the extracellular space (Fig. 3A). Ca²⁺ is exchanged between these compartments by a plasma membrane Ca²⁺ pump and a plasma membrane Ca²⁺ channel, by binding to and being released from a Ca²⁺ buffer, as well as by a store channel and store pump. Because the original model did not reveal any Ca²⁺ decreases upon ACh stimulation in cells with depleted intracellular Ca²⁺ stores, we modified the activation patterns of the model as described in Methods. Even with this change, simulated Ca²⁺ transients failed to account for the observed post-Tg [Ca²⁺]_i decrease (Fig. 3B).

		View larger version [in this window] [in a new window]
Figure 3. Simulation of ACh-induced [Ca2+]i transients using a published model A, a model to simulate Ca2+ transients was adapted from a recently published paper (Korngreen et al. 1997). The details of the model are given in Methods. ACh stimulation was modelled by increasing the basal flux (towards the cytosol) through the plasma membrane and store channels using an exponential function. Tg exposure was simulated by lowering the maximum rate of the store pump. In addition, the ensuing mandatory Ca2+ influx through the plasma membrane channels (depletion activated Ca2+ influx) was modelled by increasing the basal Ca2+ flux through this channel. B, these simulations revealed a reasonable [Ca2+]i transient upon exposure to ACh and a steady [Ca2+]i increase in response to Tg. However, these simulations failed to account for the [Ca2+]i decrease seen upon ACh exposure after Ca2+ store depletion.

Testing the hypotheses

Hypothesis 1

ACh-stimulated decreases in [Ca²⁺]_i are accounted for by inhibiting steady Ca²⁺ influx through plasma membrane Ca²⁺ channels (see Fig. 4).

		View larger version [in this window] [in a new window]
Figure 4. Hypotheses for ACh-induced [Ca2+]i decreases The figure shows five proposed mechanisms by which ACh could decrease [Ca2+]i in cells treated with Tg. 1, inhibition of Ca2+ influx; 2, activation of Ca2+ extrusion via a plasma membrane Ca2+-ATPase; 3, activation of Ca2+ extrusion via the Na+-Ca2+ exchanger; 4, activation of a Tg-insensitive uptake mechanism into internal stores and 5, activation of Ca2+ uptake into mitochondria.

Simulation. The simulations were intended to give us a qualitative picture of the expected changes in [Ca²⁺]_i for each hypothesis. They were not intended to be used for fitting the experimental data or representing the data quantitatively.

To simulate ACh-stimulated decreases in [Ca²⁺]_i in Tg-pretreated cells for hypothesis 1, the steady post-Tg Ca²⁺ influx was temporarily decreased by reducing the activating factor for the plasma membrane Ca²⁺ channel from 0·8 to 0·05 (Fig. 5A). This led to a decrease in the post-Tg [Ca²⁺]_i similar to the responses observed under experimental conditions. Blocking this influx by reducing the external [Ca²⁺] to 0 mM eliminated any further decrease in [Ca²⁺]_i as expected (Fig. 5B).

		View larger version [in this window] [in a new window]
Figure 5. Simulation and experimental testing of hypothesis 1: inhibition of Ca2+ influx A, to simulate ACh-stimulated decreases in [Ca2+]i in Tg-pretreated cells according to hypothesis 1 (see Fig. 4), the steady post-Tg Ca2+ influx was decreased in the presence of ACh by reducing the activating factor for the plasma membrane Ca2+ channel from 0·8 to 0·05. This led to a temporary decrease in the post-Tg [Ca2+]i. B, blocking any Ca2+ influx in these simulations by removing extracellular calcium, lead to a decrease in the post-Tg [Ca2+]i and eliminated the ACh-induced decrease in [Ca2+]i. C, exposing ciliated cells to 0 mM extracellular Ca2+ revealed that Ca2+ influx occurred after exposure to 1 µM Tg as predicted from the simulations shown in B, the [Ca2+]i signal decreased to a new plateau after 0 Ca2+-HBSS replaced regular HBSS. Despite the removal of extracellular Ca2+, 10 µM ACh still provoked a decrease in [Ca2+]i. D, application of 100 µM Cd2+ again showed that Ca2+ influx occurred after depletion of intracellular Ca2+ stores by Tg. Despite influx inhibition with this fairly broad Ca2+ channel blocker, 10 µM ACh still provoked a decrease in [Ca2+]i.

Ciliated cells. Ciliated cells treated with Tg were exposed to 0 Ca²⁺-HBSS before ACh application in order to eliminate Ca²⁺ influx (Fig. 5C). The exchange of regular HBSS with 0 Ca²⁺-HBSS resulted in a decrease in the post-Tg baseline from 119 ± 3·3 nM (mean ± S.E.M. from 151 cells on 11 coverslips) to 78 ± 2·4 nM. This decrease supports a steady Ca²⁺ influx after Ca²⁺ depletion of internal stores. Despite the inhibition of this influx, application of 10 µM ACh resulted in a further [Ca²⁺]_i decrease (Fig. 3C) of 36·8 ± 1·5 nM (mean ± S.E.M.) and this decrease was not significantly different from date-matched control cells where ACh induced a decrease of 34·3 ± 2·1 nM from a post-Tg baseline of 79·4 ± 4·8 nM (42 cells on 9 coverslips).

To confirm the findings with 0 Ca²⁺-HBSS, we used the Ca²⁺ channel blockers cadmium and nickel. The application of 100 µM Cd²⁺ to the bath of cells pretreated with Tg again revealed the presence of Ca²⁺ influx after intracellular Ca²⁺ store depletion. [Ca²⁺]_i decreased by 22·3 ± 1·6 nM from a post-Tg baseline of 157·4 ± 5·8 nM (44 cells on 3 coverslips). Despite Ca²⁺ influx inhibition with this fairly broad Ca²⁺ channel blocker, 10 µM ACh still provoked a [Ca²⁺]_i decrease of 55·5 ± 4·2 nM (Fig. 5D).

Upon exposure to 100 µM Ni²⁺, [Ca²⁺]_i decreased by 21·9 ± 1·4 nM from a post-Tg baseline of 98·7 ± 4·5 nM (57 cells on 3 coverslips) again supporting the presence of Ca²⁺ influx. Ni²⁺ was unable to inhibit the ACh-induced [Ca²⁺]_i decrease which measured 37·5 ± 2·2 nM. This decrease was not significantly different from the decreases obtained in control cells (P > 0·05).

The decrease in the post-Tg baseline upon Ca²⁺ influx inhibition was significantly greater with 0 Ca²⁺-HBSS compared with Cd²⁺ and Ni²⁺ (P < 0·05). The reason for this difference is not clear, but the difference could be due to incomplete inhibition of the Ca²⁺ influx by the chosen Cd²⁺ and Ni²⁺ concentrations. In the presence of the Ca²⁺ channel blocker Cd²⁺, the ACh-induced [Ca²⁺]_i decrease was significantly greater than in control cells, cells in 0 Ca²⁺-HBSS, and cells treated with Ni²⁺ (ANOVA, P < 0·05). One has to consider, however, that in some experiments of the Cd²⁺ series, the post-Tg baseline was higher than in the control and 0 Ca²⁺-HBSS experiments, making the relevance of such a significance uncertain. In fact, when we compared the effects of Ni²⁺ and Cd²⁺ on cells with similar heights of post-Tg Ca²⁺ plateau this apparent difference was eliminated.

In summary, Ca²⁺ influx is present in ciliated cells after treatment with Tg. Inhibiting this influx, however, does not eliminate the ACh-induced [Ca²⁺]_i decrease. While we cannot be certain that we have successfully blocked all possible plasma membrane Ca²⁺ channels, we consider these results, taken together with the 0 extracellular Ca²⁺ experiments, as sufficient to refute hypothesis 1 in ciliated cells.

Hypothesis 2

Activation of Ca²⁺ extrusion via a plasma membrane Ca²⁺-ATPase accounts for the [Ca²⁺]_i decrease.

Simulation. To simulate an ACh-stimulated decrease in [Ca²⁺]_i for hypothesis 2, the plasma membrane Ca²⁺-ATPase was stimulated using an exponential function by transiently increasing its activating factor (from 0·2 to 0·85). This led to a decrease in the post-Tg [Ca²⁺]_i (Fig. 6A), demonstrating the plausibility of the hypothesis. Inhibiting this influx by reducing the agonist-induced activation of the plasma membrane Ca²⁺-ATPase to 20 % of the simulation shown in Fig. 6A, did not abolish, but clearly diminished the resulting decrease of [Ca²⁺]_i upon ACh exposure (Fig. 6B).

		View larger version [in this window] [in a new window]
Figure 6. Simulation and experimental testing of hypothesis 2: activation of plasma membrane Ca2+-ATPase A, to simulate activation of the plasma membrane Ca2+-ATPase during ACh exposure, the activating factor was transiently increased from 0·2 to 0·85 following exponential kinetics both for the increase and the decay back to baseline. This led to a reduction in the pre-Tg ACh-induced [Ca2+]i peak and to an ACh-induced decrease in the post-Tg [Ca2+]i. B, the same simulation was run while the activation of the plasma membrane Ca2+-ATPase was reduced by 80 % throughout the experiment. This procedure reduced the ACh-induced [Ca2+]i decrease significantly. C, exposing ciliated cells to '0' extracellular Ca2+ to inhibit Ca2+ influx and 30 mM orthovanadate (an inhibitor of ATPases in general) did not eliminate the post-Tg, ACh-induced decrease in [Ca2+]i. D, application of 10 mM La3+ inhibited Ca2+ influx in ciliated cells pretreated with Tg, but did not inhibit the ACh-induced decrease in [Ca2+]i. Evidence that this concentration of La3+ in fact blocked the plasma membrane Ca2+-ATPase is shown in Fig. 9B.

Ciliated cells. To test this hypothesis, we looked at the ability of trifluoperazine (10 µM; n = 58 from 3 coverslips) and phenoxybenzamine (10 µM; n = 17 from 3 coverslips), both calmodulin inhibitors, to inhibit the ACh-induced [Ca²⁺]_i decrease. Trifluoperazine was invariably cytotoxic in the presence of UV light leading to ciliostasis and cell death. The results from these experiments were therefore not further analysed. Phenoxybenzamine, on the other hand, completely inhibited the [Ca²⁺]_i decrease in 50 % of the cells, but left [Ca²⁺]_i unchanged in the other 50 %. Phenoxybenzamine has been reported to alkylate muscarinic receptors, rendering them inactive (Eglen & Harris, 1993). This possibility, together with the lack of a positive control for phenoxybenzamine's ability to inhibit calmodulin's action in this system, and the possibility of ATPase activation through pathways not involving calmodulin (Carafoli & Stauffer, 1994) encouraged us to test hypothesis 2 by different, additional means.

ATPases, including the plasma membrane Ca²⁺-ATPase and dynein, can be inhibited by high concentrations of vanadate or lanthanum (Carafoli & Stauffer, 1994). The application of 30 mM orthovanadate to Tg-treated cells led (1) to ciliostasis, providing evidence that vanadate crossed the plasma membrane and inhibited dynein, and (2) to an increase in [Ca²⁺]_i supporting the hypothesis that, at this concentration, vanadate blocked the plasma membrane Ca²⁺-ATPase (unopposed Ca²⁺ influx; not shown). Exposing these cells to 0 Ca²⁺-HBSS resulted in a [Ca²⁺]_i decrease, perhaps due to Ca²⁺ efflux buffered by EGTA. Most importantly, however, exposing ciliated cells to 0 Ca²⁺-HBSS and 30 mM vanadate did not eliminate the ACh-induced [Ca²⁺]_i decrease. From a post-Tg baseline of 148 ± 10 nM, the [Ca²⁺]_i decreased to 103 ± 10 nM in the presence of vanadate and 0 Ca²⁺-HBSS and, upon ACh application, to 79 ± 8 nM (12 cells on 2 coverslips) (Fig. 6C). This ACh-induced decrease in [Ca²⁺]_i was neither significantly different from a historic control nor from culture- and date-matched control cells exposed solely to 0 Ca²⁺-HBSS (P > 0·05). It is expected that, if present, vanadate also blocked Tg-insensitive Ca²⁺-ATPases on organellar membranes. Such Ca²⁺-ATPases (see hypothesis 5 in Fig. 1) are therefore unlikely to account for the [Ca²⁺]_i decrease.

To confirm the results obtained with vanadate, we used lanthanum, another Ca²⁺-ATPase inhibitor. La³⁺ does not cross the plasma membrane in intact cells and thus more selectively inhibits the plasma membrane Ca²⁺-ATPase. In addition, La³⁺ blocks Ca²⁺ entry into cells as well as the Na⁺-Ca²⁺ exchanger (Kimura et al. 1986). When 10 mM La³⁺ was added to ciliated cells pretreated with Tg, [Ca²⁺]_i decreased from a post-Tg baseline of 171 ± 6 nM to 133 ± 6 nM (29 cells from 7 coverslips), attesting to a post-Tg Ca²⁺ influx. Despite the presence of La³⁺, the ACh-induced [Ca²⁺]_i decrease was not inhibited. [Ca²⁺]_i decreased further to 101 ± 4 nM (Fig. 6D). These decreases were neither significantly different from a historic control nor from culture- and date-matched control cells. Evidence that this concentration of La³⁺ in fact blocked the plasma membrane Ca²⁺-ATPase will be given below when we discuss the CCCP results (see Fig. 9B).

In summary, these experiments suggested that stimulation of the plasma membrane Ca²⁺-ATPase was not required for the ACh-induced [Ca²⁺]_i decrease in ciliated cells and also suggested that the Na⁺-Ca²⁺ exchanger was not involved in this response as confirmed in later experiments discussed below.

Hypothesis 3

ACh-induced activation of Ca²⁺ extrusion via the Na⁺-Ca²⁺ exchanger accounts for the [Ca²⁺]_i decrease.

Simulation. Since simulation of this scenario would be similar to that performed for hypothesis 2, we did not pursue a specific simulation.

Ciliated cells. To inhibit the Na⁺-Ca²⁺ exchanger, known to be present in airway epithelia (Murphy et al. 1988), we examined the effects of Na⁺-free HBSS on the ACh-induced [Ca²⁺]_i decrease. Neither short (3-5 min, Fig. 7A) nor long (> 30 min, Fig. 7B) incubations in Na⁺-free HBSS before or after Tg treatment inhibited the ACh-induced [Ca²⁺]_i decrease. [Ca²⁺]_i decreased by 52 ± 3 nM (68 cells from 7 coverslips) upon ACh application in Na⁺-free HBSS, a decrease not different from culture- and date-matched controls solely exposed to ACh. These experiments suggested therefore that Ca²⁺ extrusion by the Na⁺-Ca²⁺ exchanger does not play a role in the ACh-induced decrease response. The La³⁺ experiments outlined above add further confirmation that the Na⁺-Ca²⁺ exchanger does not contribute to the ACh-induced [Ca²⁺]_i decrease.

		View larger version [in this window] [in a new window]
Figure 7. Effects of Na+-Ca2+ exchanger inhibition on the post-Tg [Ca2+]i decrease in response to ACh To examine the role of the Na+-Ca2+ exchanger known to be present in airway epithelial cells, ciliated cells were exposed to ACh in Na+-free HBSS (see Methods). Neither short (A) nor long (B) incubations in Na+-free HBSS inhibited the ACh-induced decrease in [Ca2+]i.

Hypothesis 4

ACh-induced activation of Ca²⁺ uptake into a non-mitochondrial Tg-insensitive store accounts for the decrease in [Ca²⁺]_i.

Simulation. The simulation of this hypothesis is identical to hypothesis 5 (see below and Fig. 8).

Experiments. The experiments for this hypothesis have already been described above. The failure of orthovanadate (30 mM) to block the decrease in [Ca²⁺]_i make it unlikely that a non-mitochondrial, Tg-insensitive, intracellular ATPase was playing a role in the [Ca²⁺]_i decrease (Fig. 6C).

Hypothesis 5

ACh-induced activation of Ca²⁺ uptake into mitochondria accounts for the decrease in [Ca²⁺]_i.

Simulation. To simulate ACh-stimulated decreases in [Ca²⁺]_i for hypothesis 4 and 5, a new, Tg-insensitive store was created inside the model cell with the potential to increase Ca²⁺ uptake upon ACh stimulation. The parameters of the store Ca²⁺-ATPase, as well as the Ca²⁺ leak channel, were chosen to be similar to those for the Tg-sensitive store but with a slower rate of calcium flux and a slower maximum rate of the pump (new store channel with calcium flux rate, 0·9 s^-1; store pump with a maximum rate, 100 µM s^-1; K_d, 0·2 µM and [Ca²⁺]_i, 25 µM). The fine tuning of the parameters was aimed at obtaining a stable steady-state [Ca²⁺]_i. Using this model, and stimulating the Tg-insensitive store Ca²⁺ pump using an exponential function, a [Ca²⁺]_i decrease similar to the one observed under experimental conditions could be observed (Fig. 8A). Blocking this Ca²⁺ store by reducing the V_max of the Tg-insensitive store Ca²⁺ pump to one-tenth of its original value (e.g. using CCCP for uncoupling mitochondria) induced a transient increase in [Ca²⁺]_i and also eliminated the ACh-stimulated [Ca²⁺]_i decrease as expected (Fig. 8B).

		View larger version [in this window] [in a new window]
Figure 8. Simulation of hypotheses 4 and 5: activation of Ca2+ uptake into a Tg-insensitive store A, a new, Tg-insensitive store was created inside the cells with the potential to increase Ca2+ uptake upon ACh stimulation. The parameters of the store are given in Methods. The usual ACh exposure simulation was expanded by transiently increasing the store pump's activation factor. With this, a [Ca2+]i decrease could be observed when Tg-sensitive stores were depleted. B, blocking this Ca2+ store by reducing the Vmax of the Tg-insensitive store Ca2+ pump to one-tenth of its original value (e.g. using CCCP for uncoupling mitochondria) induced a transient increase in [Ca2+]i and eliminated the ACh-stimulated [Ca2+]i decrease.

Ciliated cells. To test the possibility that the uptake of Ca²⁺ into mitochondria explains the ACh-induced [Ca²⁺]_i decrease in ciliated airway epithelial cells, we used the mitochondrial uncouplers CCCP and FCCP. These compounds uncouple oxidative phosphorylation thereby disrupting the Ca²⁺ uniporter's activity (Jouaville et al. 1995; Herrington et al. 1996). Ethanol (1:2000), the vehicle for CCCP and FCCP, did not inhibit the ACh-induced [Ca²⁺]_i decrease (Fig. 9A). Addition of 5 µM FCCP or CCCP to cells treated with 1 µM Tg lead to a temporary [Ca²⁺]_i increase as predicted by the simulation (Fig. 9A and C). More importantly, cells treated either with 5 µM CCCP (n = 87 cells from 10 coverslips) or FCCP (n = 6 cells from 2 coverslips) and 1 µM Tg no longer showed an ACh-induced [Ca²⁺]_i decrease. In addition, when CCCP was added at the nadir of the [Ca²⁺]_i decrease, it reversed the [Ca²⁺]_i decrease back to baseline at a much faster rate than in the absence of CCCP, reverting the [Ca²⁺]_i towards baseline within less than 1 min (Fig. 9D). These results strongly suggest that the Ca²⁺ taken up into mitochondria leaks out of these organelles to restore [Ca²⁺]_i. Further, they suggest that no other mechanism is working to decrease [Ca²⁺]_i in ciliated cells.

		View larger version [in this window] [in a new window]
Figure 9. Effects of mitochondrial uncouplers on the [Ca2+]i decrease in response to ACh in ciliated cells A, application of 5 µM CCCP to cells treated with 1 µM Tg led, as predicted in Fig. 8B, to a temporary increase in [Ca2+]i. The ACh-induced decrease in [Ca2+]i was lost. B, adding 10 mM La3+ and 5 µM CCCP to a ciliated cell led to a steady increase in [Ca2+]i, suggesting that plasma membrane Ca2+-ATPases are active when cells are exposed to CCCP alone and that La3+ blocked the plasma membrane Ca2+-ATPase at this concentration. After CCCP exposure, the [Ca2+]i decrease in response to ACh was lost. C, application of 5 µM FCCP to cells treated with 1 µM Tg also lead to a temporary increase in [Ca2+]i. Again, the ACh-induced [Ca2+]i drop was absent. D, adding 5 µM CCCP to a ciliated cell at the nadir of the drop response led to faster recovery of [Ca2+]i back to baseline compared with a control cell (dashed grey trace).

A concern with these results is the loss of cellular ATP due to mitochondrial uncoupling, and the consequent deprivation of fuel to ATPases. Two arguments suggest that the mitochondrial disruption did not lead to general ATP depletion and a breakdown of cellular metabolism. First, the cilia of all cells exposed to CCCP and FCCP were still beating at the end of the experiments, albeit at a lower frequency (visual evaluation through the oculars). This indicates the presence of sufficient ATP to fuel ciliary motility. Second, even if the residual ciliary beating could be attributed to local ATP production, adding 10 mM La³⁺ plus 5 µM CCCP to ciliated cells lead to a steady [Ca²⁺]_i increase (from Ca²⁺ leaking out of mitochondria, Fig. 9B). This suggests that La³⁺ blocked a plasma membrane Ca²⁺-ATPase that was still active when the cells were exposed to CCCP alone (Fig. 9B). Simulations of this scenario supported this line of reasoning (not shown). This result also confirms that La³⁺ at this concentration blocks a plasma membrane Ca²⁺-ATPase (see above). Finally, an unspecific effect of CCCP on the ACh signalling pathway was ruled out by exposing cells (n = 20 from 4 coverslips) to ACh and subsequently 5 µM CCCP 1 min before CCCP plus 10 µM ACh. CCCP was effective within 1 min as evidenced by ciliary slowing (see also Fig. 9D: increase in [Ca²⁺]_i from nadir occurs within less than 1 min after CCCP application). CCCP pre-exposure did not eliminate the transient increase in [Ca²⁺]_i upon exposure to ACh, in fact the transient had a higher peak level and slower decrease back towards baseline compared with the first ACh-induced [Ca²⁺]_i transient (consistent with the known role of mitochondria to decrease [Ca²⁺]_i after agonist stimulation).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

We have explored different hypotheses in an effort to find the mechanism by which ciliated cells actively decrease [Ca²⁺]_i upon muscarinic receptor activation both using simulations of [Ca²⁺]_i behaviour in non-excitable cells as well as carrying out direct measurements on ciliated airway epithelial cells. Korngreen et al. (1997) developed a realistic model of Ca²⁺ transients in response to receptor stimulation in non-excitable cells. We have evaluated this model to see whether it could account for the Ca²⁺ responses seen in ciliated cells. The model, as proposed, did not predict any [Ca²⁺]_i decreases upon receptor stimulation when cells were pretreated with Tg. Further, changing the activation parameters of the model to a more physiological pattern (rapid rise and exponential decay of the stimulation pattern) also failed to account for the observed Ca²⁺ decrease. The hypotheses outlined in Fig. 4 were then incorporated into the model. The simulations helped to establish the plausibility of each hypothesis and were intended to provide qualitative data of [Ca²⁺]_i changes. We did not fit the parameters of this model to experimental data in a quantitative manner; therefore, the magnitude of the [Ca²⁺]_i changes of the model is not in exact agreement with the experimental results. Nevertheless, the qualitative predictions were useful in supporting the [Ca²⁺]_i behaviour of different antagonists in all the tested hypotheses.

Most of the experiments were carried out by loading the cells with fura-2 AM, the acetoxymethyl ester and membrane-permeable form of the dye. Therefore, fura-2 could have accumulated in several subcellular compartments, especially the endoplasmic reticulum (Golovina & Blaustein, 1997). This raises an obvious concern, namely that fura-2 fluorescence reported a calcium signal from this compartment. To ensure that our data were not flawed, we have chosen conditions that discourage fura-2 AM loading into compartments by incubating cells at room temperature and for less than an 1 h (Golovina & Blaustein, 1997). To confirm that this approach reports calcium changes in the cytoplasm, we injected cells with the free acid form of fura-2 assuring only cytosolic distribution of the dye. Data from injected cells were similar to those obtained with fura-2 AM loaded cells (Fig. 1D). Finally, we evaluated our technique for AM dye loading with confocal microscopy and failed to obtain signals from mitochondria or other stores. We have also shown that ciliary beat frequency decreases with the calcium concentration in ciliated cells pretreated with Tg (Salathe et al. 1997). Although two different signalling pathways might account for this similarity, it is more likely to support the notion that our measurements from fura-2 AM loaded cells mainly reflect the cytosolic calcium concentration. We therefore used the easier approach of loading with the membrane-permeable form of fura-2 for all subsequent experiments.

The experiments were inspired by previous reports of agonist-induced [Ca²⁺]_i decreases (Rink & Sage, 1987; Muallem et al. 1988; Duddy et al. 1989; Randriamampita & Trautmann, 1990; Tepikin et al. 1992). Most of these reports concluded that activation of a plasma membrane Ca²⁺-ATPase was responsible for the observed [Ca²⁺]_i decrease. As evident from our results, however, the activation of a plasma membrane Ca²⁺-ATPase is not critical for the [Ca²⁺]_i decrease in ciliated cells. Here, experiments with mitochondrial uncouplers (CCCP and FCCP) suggested that stimulated mitochondrial Ca²⁺ uptake is responsible for the observed decrease in [Ca²⁺]_i.

This finding was somewhat surprising, because, historically, stimulated Ca²⁺ uptake into mitochondria has not been described. More and more evidence accumulates, however, that mitochondria are important participants in the Ca²⁺ homeostasis of a cell: Jouaville et al. (1995) showed that mitochondria are essential for the synchronisation of Ca²⁺ waves in Xenopus oocytes and several papers reported on the role of mitochondria in clearing Ca²⁺ from the cytosol after IP₃-mediated Ca²⁺ release from internal stores or during Ca²⁺ oscillations (Rutter et al. 1993; Hajnoczky et al. 1995; Montero et al. 1995; Herrington et al. 1996). All the authors, however, were puzzled by their findings in light of the reports that the Ca²⁺ uptake system into mitochondria, namely the Ca²⁺ uniporter, is active only at a [Ca²⁺]_i above 1 µM. This assumption was based on the measurement of a high K_d of the uniporter to Ca²⁺ in vitro. Rizzuto et al. (1993) thus proposed a close spatial association of the endoplasmic reticulum (from where the Ca²⁺ is released) and mitochondria. This association might allow mitochondria to 'see' locally high Ca²⁺ concentrations, sufficient for Ca²⁺ uptake into mitochondria via the Ca²⁺ uniporter. This model, however, would still not explain our results of stimulated Ca²⁺ uptake into mitochondria after depletion of internal Ca²⁺ stores. However, Sparagna et al. (1995) described a new uptake mechanism into mitochondria which is active at or below 200 nM [Ca²⁺]_i. Therefore, some evidence supports the idea that mitochondrial Ca²⁺ uptake is possible at the [Ca²⁺]_i levels we measured. But the identity of the second messengers that might mediate G-protein coupled receptor activation to such Ca²⁺ uptake into mitochondria remains unknown.

There are still concerns regarding the results obtained in ciliated cells and it is worth reviewing them. An ATP depletion of the cell during mitochondrial uncoupling could have rendered an ATPase non-functional. Arguments against this explanation are: (1) cilia were still beating, albeit slowly, and (2) the plasma membrane Ca²⁺-ATPase was still active (see Fig. 9B). Although local ATP production in cilia has not been ruled out (as proposed via a creatine shuttle), the still active plasma membrane ATPase shows that the cells must have had considerable ATP reserves. In addition, the vanadate experiments make it unlikely that a different, intracellular ATPase was paralysed by ATP depletion because such an ATPase did not play a role in the [Ca²⁺]_i decrease.

So how can we reconcile the difference between most published reports (implicating an ATPase) and our finding (implicating mitochondria)? A major function of ciliated cells is ciliary beating, which requires a lot of energy from ATP. Thus ciliated cells are loaded with mitochondria. It is possible that a single second messenger activates both mitochondria and ATPases and that the magnitude of the effect depends on the availability of the target. If that is true, ciliated and other cells with lots of mitochondria will show a larger mitochondria-dependent decrease in [Ca²⁺]_i and simultaneously transduce the signal to increase ATP production. Alternatively, non-ciliated cells with relatively low mitochondrial content would primarily signal through plasma membrane ATPases to actively decrease [Ca²⁺]_i. Both cell types would still use mitochondria for passive uptake of Ca²⁺ at high levels of cytosolic Ca²⁺ (Herrington et al. 1996).

In summary, we have presented evidence that the activation of muscarinic receptors, in addition to releasing Ca²⁺ from internal stores, stimulates at least one additional pathway to actively decrease [Ca²⁺]_i. In ciliated cells, the data suggest that this pathway stimulates Ca²⁺ uptake into mitochondria. It is interesting to note that the same stimulus that increases ciliary beating and thereby increases demand for ATP also simultaneously signals to the mitochondria, via Ca²⁺, to increase ATP production (Hubbard & McHugh, 1996). Further studies delineating the exact signalling pathways are needed.

	REFERENCES

Top Abstract Introduction Methods Results Discussion References

Carafoli, E. & Stauffer, T. (1994). The plasma membrane calcium pump: functional domains, regulation of the activity, and tissue specificity of isoform expression. Journal of Neurobiology 25, 312-324	[Medline]
Duddy, S. K., Kass, G. E. & Orrenius, S. (1989). Ca2+-mobilizing hormones stimulate Ca2+ efflux from hepatocytes. Journal of Biological Chemistry 264, 20863-20866	[Abstract]
Ecelbarger, C. A., Chou, C. L., Lolait, S. J., Knepper, M. A. & DiGiovanni, S. R. (1996). Evidence for dual signaling pathways for V2 vasopressin receptor in rat inner medullary collecting duct. American Journal of Physiology 270, F623-633	[Medline]
Eglen, R. M. & Harris, G. C. (1993). Selective inactivation of muscarinic M2 and M3 receptors in guinea-pig ileum and atria in vitro. British Journal of Pharmacology 109, 946-952	[Medline]
Fabiato, A. & Fabiato, F. (1978). Effects of pH on the myofilaments and the sarcoplasmic reticulum of skinned cells from cardiac and skeletal muscles. The Journal of Physiology 276, 233-255	[Abstract]
Felder, C. C. (1995). Muscarinic acetylcholine receptors: signal transduction through multiple effectors. FASEB Journal 9, 619-625	[Abstract]
Golovina, V. A. & Blaustein, M. P. (1997). Spatially and functionally distinct Ca2+ stores in sarcoplasmic and endoplasmic reticulum. Science 275, 1643-1648	[Abstract/Full Text]
Grynkiewicz, G., Poenie, M. & Tsien, R. Y. (1985). A new generation of Ca2+ indicators with greatly improved fluorescence properties. Journal of Biological Chemistry 260, 3440-3450	[Abstract]
Hajnoczky, G., Robb-Gaspers, L. D., Seitz, M. B. & Thomas, A. P. (1995). Decoding of cytosolic calcium oscillations in the mitochondria. Cell 82, 415-424	[Medline]
Hall, R. A., Ostedgaard, L. S., Premont, R. T., Blitzer, J. T., Rahman, N., Welsh, M. J. & Lefkowitz, R. J. (1998a). A C-terminal motif found in the beta2-adrenergic receptor, P2Y1 receptor and cystic fibrosis transmembrane conductance regulator determines binding to the Na+/H+ exchanger regulatory factor family of PDZ proteins. Proceedings of the National Academy of Sciences of the USA 95, 8496-8501	[Abstract/Full Text]
Hall, R. A., Premont, R. T., Chow, C. W., Blitzer, J. T., Pitcher, J. A., Claing, A., Stoffel, R. H., Barak, L. S., Shenolikar, S., Weinman, E. J., Grinstein, S. & Lefkowitz, R. J. (1998b). The beta2-adrenergic receptor interacts with the Na+/H+-exchanger regulatory factor to control Na+/H+ exchange. Nature 392, 626-630	[Medline]
Herrington, J., Park, Y. B., Babcock, D. F. & Hille, B. (1996). Dominant role of mitochondria in clearance of large Ca2+ loads from rat adrenal chromaffin cells. Neuron 16, 219-228	[Medline]
Horrigan, F. & Bookman, R. J. (1995). Construction and performance of a rapidly switching, filter-based fluorescence excitation system. Biophysical Journal 68, A289.
Hubbard, M. J. & McHugh, N. J. (1996). Mitochondrial ATP synthase F1-beta-subunit is a calcium-binding protein. FEBS Letters 391, 323-329	[Medline]
Jouaville, L. S., Ichas, F., Holmuhamedov, E. L., Camacho, P. & Lechleiter, J. D. (1995). Synchronization of calcium waves by mitochondrial substrates in Xenopus laevis oocytes. Nature 377, 438-441	[Medline]
Kimura, J., Noma, A. & Irisawa, H. (1986). Na-Ca exchange current in mammalian heart cells. Nature 319, 596-597	[Medline]
Kofuji, P., Davidson, N. & Lester, H. A. (1995). Evidence that neuronal G-protein-gated inwardly rectifying K+ channels are activated by G beta gamma subunits and function as heteromultimers. Proceedings of the National Academy of Sciences of the USA 92, 6542-6546	[Abstract]
Korngreen, A., Gold'shtein, V. & Priel, Z. (1997). A realistic model of biphasic calcium transients in electrically nonexcitable cells. Biophysical Journal 73, 659-673	[Abstract]
Luttrell, L. M., Ferguson, S. S., Daaka, Y., Miller, W. E., Maudsley, S., Della Rocca, G. J., Lin, F., Kawakatsu, H., Owada, K., Luttrell, D. K., Caron, M. G. & Lefkowitz, R. J. (1999). Beta-arrestin-dependent formation of beta2 adrenergic receptor-Src protein kinase complexes. Science 283, 655-661	[Abstract/Full Text]
Mathes, C. & Thompson, S. H. (1995). The relationship between depletion of intracellular Ca2+ stores and activation of Ca2+ current by muscarinic receptors in neuroblastoma cells. Journal of General Physiology 106, 975-993	[Abstract]
Mohuczy-Dominiak, D. & Garg, L. C. (1992a). Characterization of cholinergic receptors in Madin-Darby canine kidney cells. Journal of the American Society of Nephrology 3, 170-181	[Abstract]
Mohuczy-Dominiak, D. & Garg, L. C. (1992b). Muscarinic receptors in MDCK cells are coupled to multiple messenger systems. American Journal of Physiology 263, C1289-1294	[Medline]
Montero, M., Brini, M., Marsault, R., Alvarez, J., Sitia, R., Pozzan, T. & Rizzuto, R. (1995). Monitoring dynamic changes in free Ca2+ concentration in the endoplasmic reticulum of intact cells. EMBO Journal 14, 5467-5475	[Abstract]
Muallem, S., Pandol, S. J. & Beeker, T. G. (1988). Calcium mobilizing hormones activate the plasma membrane Ca2+ pump of pancreatic acinar cells. Journal of Membrane Biology 106, 57-69	[Medline]
Murphy, E., Cheng, E., Yankaskas, J., Stutts, M. J. & Boucher, R. C. (1988). Cell calcium levels of normal and cystic fibrosis nasal epithelium. Pediatric Research 24, 79-84	[Abstract]
Poenie, M. (1990). Alteration of intracellular Fura-2 fluorescence by viscosity: a simple correction. Cell Calcium 11, 85-91	[Medline]
Randriamampita, C. & Trautmann, A. (1990). Arachidonic acid activates Ca2+ extrusion in macrophages. Journal of Biological Chemistry 265, 18059-18062	[Abstract]
Rink, T. J. & Sage, S. O. (1987). Stimulated calcium efflux from fura-2-loaded human platelets. The Journal of Physiology 393, 513-524	[Abstract]
Rizzuto, R., Brini, M., Murgia, M. & Pozzan, T. (1993). Microdomains with high Ca2+ close to IP3-sensitive channels that are sensed by neighboring mitochondria. Science 262, 744-747	[Medline]
Rutter, G. A., Theler, J. M., Murgia, M., Wollheim, C. B., Pozzan, T. & Rizzuto, R. (1993). Stimulated Ca2+ influx raises mitochondrial free Ca2+ to supramicromolar levels in a pancreatic beta-cell line. Possible role in glucose and agonist-induced insulin secretion. Journal of Biological Chemistry 268, 22385-22390	[Abstract]
Salathe, M. & Bookman, R. J. (1995). Coupling of [Ca2+]i and ciliary beating in cultured tracheal epithelial cells. Journal of Cell Science 108, 431-440	[Abstract]
Salathe, M., Lipson, E., Ivonnet, P. I. & Bookman, R. J. (1997). Muscarinic signal transduction in tracheal epithelial cells: effects of acetylcholine on intracellular Ca2+ and ciliary beating. American Journal of Physiology 272, L301-310	[Medline]
Sanderson, M. J., Charles, A. C. & Dirksen, E. R. (1990). Mechanical stimulation and intercellular communication increase intracellular calcium in epithelial cells. Cell Regulation 1, 585-596	[Medline]
Sanderson, M. J. & Dirksen, E. R. (1989). Mechanosensitive and beta-adrenergic control of the ciliary beat frequency of mammalian respiratory tract cells in culture. American Review of Respiratory Disease 139, 432-440	[Medline]
Sparagna, G. C., Gunter, K. K., Shen, S. S. & Gunter, T. E. (1995). Mitochondrial calcium uptake from physiological-type pulses of calcium - A description of the rapid uptake mode. Journal of Biological Chemistry 270, 27510-27515	[Abstract/Full Text]
Tepikin, A. V., Voronina, S. G., Gallacher, D. V. & Petersen, O. H. (1992). Acetylcholine-evoked increase in the cytoplasmic Ca2+ concentration and Ca2+ extrusion measured simultaneously in single mouse pancreatic acinar cells. Journal of Biological Chemistry 267, 3569-3572	[Abstract]

Acknowledgements

We wish to thank our colleagues Drs Adam Wanner and Gregory E. Conner for helpful discussions, Dr William M. Abraham for his invaluable support, and Sara Donoghue for her technical support. Matthias Salathe was a Howard Hughes Medical Institute Postdoctoral Fellow. Supported by NIH awards HL-55341 and HL-60644, American Lung Association of Florida, and the Howard Hughes Medical Institute.

Corresponding author

M. Salathe: Division of Pulmonary and Critical Care Medicine (R-47), University of Miami School of Medicine, 1600 N.W. 10th Avenue, Miami, FL 33136, USA.

Email: msalathe{at}miami.edu

This article has been cited by other articles:

				B.-X. Zhang, X. Ma, W. Zhang, C.-K. Yeh, A. Lin, J. Luo, E. A. Sprague, R. H. Swerdlow, and M. S. Katz Polyunsaturated fatty acids mobilize intracellular Ca2+ in NT2 human teratocarcinoma cells by causing release of Ca2+ from mitochondria Am J Physiol Cell Physiol, May 1, 2006; 290(5): C1321 - C1333. [Abstract] [Full Text] [PDF]

				M. Kawakami, T. Nagira, T. Hayashi, C. Shimamoto, T. Kubota, H. Mori, H. Yoshida, and T. Nakahari Hypo-osmotic potentiation of acetylcholine-stimulated ciliary beat frequency through ATP release in rat tracheal ciliary cells Exp Physiol, November 1, 2004; 89(6): 739 - 751. [Abstract] [Full Text] [PDF]

				T. Lieb, C. W. Frei, J. I. Frohock, R. J. Bookman, and M. Salathe Prolonged increase in ciliary beat frequency after short-term purinergic stimulation in human airway epithelial cells J. Physiol., January 9, 2002; 538(2): 633 - 646. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Salathe, M. Articles by Bookman, R. J. Search for Related Content PubMed PubMed Citation Articles by Salathe, M. Articles by Bookman, R. J.