cells (Atkinson, 1989), osteoblasts (Xia and Ferrier, 1992), cardiac myocytes (Sigurdson et al., 1992) and mammary epithelial and tumor cells (Enomoto et al., ...
Journal of Cell Science 106, 995-1004 (1993) Printed in Great Britain © The Company of Biologists Limited 1993
995
Intercellular calcium signaling induced by extracellular adenosine 5 -triphosphate and mechanical stimulation in airway epithelial cells Michael Hansen*, Scott Boitano, Ellen R. Dirksen and Michael J. Sanderson† Department of Anatomy and Cell Biology, UCLA School of Medicine, Los Angeles, CA 90024, USA *Present address: Department of Medical Physiology, The Panum Institute, University of Copenhagen, Blegdamsvej 3, DK-2200 Copenhagen N, Denmark †Author for correspondence
SUMMARY Airway epithelial cells in culture respond to extracellular adenosine 5 -triphosphate (ATP) by increasing their intracellular Ca 2+ concentration ([Ca 2+]i). The effective concentration of ATP that elicited a Ca2+ response equal to 50% of the maximal response (EC50) was 0.5 M. Release of ATP from a pipette to form a local gradient of ATP increased [Ca2+]i of individual cells in a sequential manner. Cells closest to the pipette showed an immediate increase in [Ca2+]i while more distal cells displayed a delayed increase in [Ca2+]i. This response to the local release of ATP appeared as a wave of increasing [Ca2+]i that spread to several cells and, in this respect, was similar to the intercellularly communicated Ca2+ waves initiated by mechanical stimulation in airway epithelial cells (Sanderson et al., Cell Regul. 1, 585-596, 1990). In the presence of a unidirectional fluid flow, the Ca2+ response to a local release of ATP was biased such that virtually all the cells responding with an increase in [Ca2+]i were downstream of the release site. By contrast, an identical fluid flow did not bias the radial propagation of intercellular Ca2+ waves induced by mechanical
stimulation. Suramin, a P2-purinergic receptor antagonist, did attenuate the Ca2+ response induced by ATP but did not block the propagation of mechanically induced Ca2+ waves. Cells from young cultures (3-5 days) or those at the leading edge of an outgrowth elevated their [Ca2+]i in response to ATP. However, these cells do not respond to mechanical stimulation by the propagation of a Ca2+ wave. From these results we conclude that the intercellular Ca2+ waves elicited by mechanical stimulation are not the result of ATP or another compound released from the stimulated cell, diffusing through the extracellular fluid. This conclusion is consistent with previous experimental evidence suggesting that intercellular Ca2+ signaling in epithelial cells is mediated by the movement of inositol trisphosphate through gap junctions (Boitano et al., Science 258, 292-295, 1992).
INTRODUCTION
sequently through more distal surrounding cells (Sanderson et al., 1990). Propagating intercellular Ca2+ waves or intercellular Ca2+ signaling are also induced by mechanical stimulation in glial cells (Charles et al., 1991, 1992, 1993; Enkvist and McCarthy, 1992), endothelial cells (Demer et al., 1993; Drumheller and Hubbell, 1991; Goligorsky, 1988), kidney cells (Atkinson, 1989), osteoblasts (Xia and Ferrier, 1992), cardiac myocytes (Sigurdson et al., 1992) and mammary epithelial and tumor cells (Enomoto et al., 1992; Furuya et al., 1993). Intercellular Ca 2+ waves are also induced in astrocytes by glutamate (Cornell-Bell et al., 1990) and can occur in Obelia photocytes (Brehm et al., 1989) and hepatocytes (Sáez et al., 1989). Mechanical stimulation also increases [Ca2+]i in alveolar type II cells (Wirtz and Dobbs, 1990) and appears to have a role in the Ca2+ regulation of cell volume in a variety of cell types (McCarthy and O’Neil, 1992; Morris, 1990). In view of the diversity of the cell types that display Ca2+ waves, and the ability
Mechanical distortion of the apical membrane of a single ciliated cell in primary cultures of tracheal epithelial cells induces an increase in ciliary beat frequency (Sanderson and Dirksen, 1986). After a delay of about 1 second adjacent ciliated cells also increase their ciliary beat frequency. This communication of increased ciliary beat frequency is extended to more distal cells with the result that five or more cells, in all directions, display increases in beat frequency (Sanderson et al., 1988). With the aid of digital fluorescence microscopy and the Ca2+-specific dye fura-2, it was found that this increase in ciliary beat frequency is stimulated by a propagating increase in intracellular Ca2+ concentration ([Ca2+]i) - an intercellular Ca2+ wave (Sanderson et al., 1990). This Ca2+ wave is initiated at the point of mechanical stimulation and spreads to the borders of the stimulated cell. After a short delay at each cell border the Ca2+ wave propagates through adjacent cells and sub-
Key words: airway epithelium, ATP, intercellular communication, inositol trisphosphate, purinergic receptors, suramin
996
M. Hansen and others
of Ca2+ waves to propagate between dissimilar cell types (Sanderson et al., 1990; Charles et al., 1991), intercellular Ca 2+ signaling may be a general cellular mechanism for intercellular communication (Sanderson et al., 1994). In airway epithelial cells, halothane, an inhibitor of gap junction conductance, blocked the propagation of intercellular Ca2+ waves (Sanderson et al., 1990). Similar results were reported for osteoblasts by Xia and Ferrier (1992). In C6 glioma cells, only cells expressing the gene for the gap junction protein connexin43 propagated intercellular Ca2+ waves (Charles et al., 1992). A decrease in gap junctional conductance by the activation of protein kinase C also reduced the propagation of Ca2+ waves in astroglial cells (Enkvist and McCarthy, 1992). Collectively, these results indicate that Ca2+ waves can be communicated between cells through gap junctions. By contrast, Ca2+ waves in preparations of isolated rat basophilic leukemia (RBL) cells do not appear to require cell contact (Osipchuk and Cahalan, 1992). In these cells, a Ca 2+ wave initiated by mechanical stimulation of a single cell was biased by an extracellular flow of fluid and was attenuated by low doses of suramin, an antagonist of P2purinergic receptors (Hoyle et al., 1990; Leff et al., 1990). Similar increases in [Ca2+]i in RBL and mast cells were induced by adenosine 5′-triphosphate (ATP). From these results Osipchuk and Cahalan (1992) proposed that ATP, released by degranulation of cytoplasmic vesicles, acts as an extracellular messenger that diffuses to adjacent cells to initiate an increase in [Ca2+]i and another cycle of vesicle degranulation. Biasing of intercellular Ca2+ waves by an extracellular fluid flow has also been observed in mammary tumor cells (Enomoto et al., 1992) but not in astroglia (Finkbeiner, 1992). However, following a closer examination of the data obtained from mammary tumor cells by Enomoto et al. (1992), it appears that some communication of the Ca2+ wave occurred in a direction opposite to the fluid flow. As a result, we suggest that it may be possible that both an extra- and inter-cellular route of communication can exist concurrently. In order to clarify the role of extra- and inter-cellular signaling in airway epithelial cells it is important to determine if the release and diffusion of ATP contributes to the propagation of intercellular Ca2+ waves. In this study we have investigated the involvement of ATP by comparing and contrasting the changes in [Ca2+]i induced in young and mature primary cultures, by either local or bath applications of ATP or by mechanical stimulation in the presence or absence of unidirectional fluid flow or the P2-receptor antagonist, suramin. We have found that airway epithelial cells respond to the local application of extracellular ATP with increases in [Ca2+]i in a wave-like manner. However, these changes in [Ca2+]i induced by ATP were biased in the direction of fluid flow. By contrast, Ca 2+ waves induced by mechanical stimulation were not biased by fluid flow or attenuated in the presence of suramin. In addition, young cultures that were sensitive to ATP did not propagate Ca2+ waves. From these results we conclude that the intercellular communication of Ca2+ waves initiated by non-traumatizing mechanical stimulation is not mediated by the extracellular diffusion of ATP.
MATERIALS AND METHODS The techniques for culturing airway epithelial cells, mechanically stimulating individual cells, and the measurement of [Ca2+]i, have been described in detail elsewhere (Charles et al., 1991; Sanderson and Dirksen, 1985, 1986; Sanderson et al., 1990) and will be only briefly reviewed here.
Culture of airway epithelial cells The tracheal mucosa, obtained from New Zealand White rabbits killed by Nembutal, was dissected from the cartilaginous backing, cut into small pieces and cultured on collagen-coated coverslips. Epithelial cells derived from these explants formed monolayers of intermixed ciliated and non-ciliated cells. Cells were insensitive to mechanical stimulation in the early stages of culture (Sanderson and Dirksen, 1986) but after 5-6 days of culture cells were able to propagate extensive Ca2+ waves following mechanical stimulation. Mechanically insensitive or young cultures were 3-4 days old and mechanically sensitive or mature cultures were 7-15 days old when used for experimentation.
Measurement of [Ca2+]i Cells were loaded with fura-2 by incubation in 5 µM fura-2/AM (FURA-2-pentaacetoxymethyl ester, Calbiochem, San Diego, CA) for 1 hour at 37°C in Hanks’ balanced salt solution, without phenol-red (Gibco-BRL, Bethesda, MD) and additionally buffered with 25 mM HEPES (HBSS, pH 7.2-7.4). Cells were left for at least 30 minutes before use following loading to allow for the deesterification of fura-2-AM. Cellular fluorescence was imaged with the aid of an inverted microscope and a silicon-intensified target camera and recorded on an optical memory disc recorder (Charles et al., 1991; Sanderson et al., 1990). Images of [Ca2+]i were calculated by ratiometric imaging or by single wavelength recordings referenced to ratiometric measurements. Single wavelength recordings provide the maximal time resolution of 30 frames/second. For single wavelength recordings, an initial [Ca2+]i reference image based on 10 frames recorded at wavelengths 340 and 380 nm was obtained. Changes in [Ca2+]i were then recorded with illumination at 380 nm with additional reference images at 340 nm taken every 10-20 seconds. [Ca2+]i was calculated from the change in fluorescence intensity at 380 nm as described by Charles et al. (1991). All images were subjected to background subtraction and corrected for shading. [Ca2+]i calculated from single wavelength recordings was similar to reference [Ca2+]i calculated by ratiometric methods at selected intervals. For plots of [Ca2+]i versus time, single points encompassing an area 2.1 µm × 2.3 µm were selected at the approximate centers of each cell and the [Ca2+]i calculated for these points only. Recordings of cells stimulated mechanically or with ATP were made at real time (30 frames/second) or in time-lapse at 1 image/second (using an average of 4 sequential frames taken in real time).
Application of ATP To determine an ATP dose-response curve, ATP (Sigma Chemical Co, Missouri, A0770) was dissolved in HBSS at concentrations ranging from 0.1 to 100 µM; 3 ml of ATP solution was exchanged through the cell chamber, which had a volume of 200 µl. To apply ATP to the culture locally, 1 mM ATP in HBSS was pressure-ejected from a glass micropipette with a tip diameter of about 1 µm using a Narashige (Greenvale, NY) IM-300 pressure injection system. The tip of the pipette was placed very close to the surface of a cell without touching it. Pressure magnitude (1020 lbf/in2 (1 lbf/in2 < 6.9 kPa)) and duration (10-100 ms) were adjusted such that an ejection pulse of ATP invoked a measurable Ca2+ response from the cells. A similar ejection of solution not containing ATP did not evoke a measurable Ca2+ response. These
Ca2+ signaling in epithelial cells
997
settings were then used for subsequent experiments. The fluorescent dye 6-carboxyfluorescein was included in the ejection mixture in some cases to allow the visual detection of the ejected solution. The effects of suramin on the ATP response and mechanical stimulation were examined by bathing the cells in HBSS with 25 or 50 µM suramin (FBA Pharmaceuticals, West Haven, CT) for at least 5 minutes, and up to 1 hour before stimulation. The procedure for ATP application was slightly changed to increase the amount of data collected. In each experiment three doses of ATP were added serially to the cells. These consisted of either (a) 50 nM, 500 nM and 5 µM or (b) 100 nM, 1 µM and 10 µM. Application of each ATP solution was performed at one-minute intervals. The Ca2+ response of the cells reached a maximum level very quickly and after 20 seconds of exposure to ATP, the ATP solution was removed by washing with 2-4 ml of HBSS. This allowed sufficient time for the [Ca2+]i of the cell to return to the pre-stimulus level before the next application of ATP. Excess solution was removed before the application of solutions containing ATP. This procedure increased the exchange rate of the control solution for experimental solutions. For these experiments, control values were re-established using this modified solution exchange procedure for comparison of experimental results. In experiments where suramin was used, suramin was included in all solutions. Application of ATP to mechanically insensitive young cultures was achieved in a similar manner.
Application of a unidirectional fluid flow A steady flow of HBSS over the culture in the vicinity of the site of mechanical stimulation or local ATP release was achieved by placing a second micropipette (tip diameter 150 µm) about 300 µm from the cell of interest. The flow rate of approximately 50 µl/min, under gravity feed, was turned on immediately prior to mechanical or chemical stimulation of the cell. HBSS was filtered through a 0.22 µm filter under vacuum to remove debris.
Mechanical stimulation Mechanical stimulation of a single cell (ciliated or non-ciliated) was achieved by briefly distorting its apical surface for approximately 150 ms with a glass micropipette (tip diameter about 1 µm). The movement of the pipette was driven by a piezoelectric device under computer control (Sanderson and Dirksen, 1986).
Fig. 1. (A-C) Changes in [Ca2+]i induced by 0.5, 1.0 and 100 µM ATP. Each tracing represents [Ca2+]i for a small area (≈5 µm2) within a single tracheal epithelial cell in culture. (D) A doseresponse relationship of ATP concentration to the peak [Ca2+]i. Each point represents the mean of 50-93 cells from 20 cultures. The standard deviation is 2-16% of the change in [Ca2+]i.
RESULTS Effect of ATP on [Ca2+]i The effect of extracellular ATP on [Ca2+]i of cultured epithelial cells was investigated and typical responses of mechanically sensitive (7-15 days old) individual cells to various concentrations of ATP are shown in Figs 1 and 2AD. In general, the addition of ATP induced an increase in [Ca2+]i after a short delay. The maximal [Ca2+]i achieved was dependent on the cell and the concentration of ATP. The ciliary beat frequency of ciliated cells also increased following ATP application, but this response was not examined in detail in this study. The dose-response relationship of ATP concentration to peak [Ca 2+]i is plotted in Fig. 1D. From this dose-response curve, the effective concentration of ATP that elicits a response equal to 50% of the maximal response (EC50) is 0.5 µM. The ‘delay time’ of the response to ATP was determined as the period of time taken to reach the maximum [Ca2+]i after the addition of ATP and includes the lag before
any changes in [Ca2+]i occur and the time it takes for [Ca2+]i to increase to the maximum concentration (Table 1). At high concentrations of ATP (>1 µM) the [Ca2+]i quickly increased to a maximum. At lower ATP concentrations (
