Anal. Chem. 2000, 72, 2001-2007
Direct Observation of Calcium-Independent Intercellular ATP Signaling in Astrocytes Ziqiang Wang,† Philip G. Haydon,‡ and Edward S. Yeung*,†
Ames LaboratorysUSDOE and Department of Chemistry, Iowa State University, Ames, Iowa 50011, and Neuroscience Program, Department of Zoology and Genetics, Iowa State University, Ames, Iowa 50011
Adenosine triphosphate (ATP) is assumed to be involved in the regulation of many extracellular signaling systems including calcium wave propagation. So far all supportive evidence is indirect, such as monitoring changes in intracellular calcium on application of extracellular ATP or off-site measurement of ATP from superfusates. Furthermore, the causal relationships among the various signaling agents are still unclear. A novel chemiluminescence dynamic imaging method was developed to monitor ATP release from living biological cells. The assay has linear response over 3 orders of magnitude for fixed concentrations of enzyme and cofactors, with a correlation coefficient of 0.999. The detectability of ATP is down to 10-8 M at millisecond exposure times with an intensified charge-coupled device camera. The direct imaging of ATP waves in astrocyte cultures was performed together with Fluo-3-Ca imaging at millisecond temporal resolution and micrometer-scale spatial resolution. We discovered that extracellular ATP mediates intercellular calcium wave propagation, but surprisingly, release and propagation of ATP are not calcium dependent. Therefore, ATP rather than Ca or IP3 is the primary intercellular signaling messenger. ATP is widely distributed in almost every type of biological cell. It is best known as the energy substrate for various cellular metabolic functions and the regulator of biological activities, including hydrogen production for electron transfer, biosynthesis, photosynthesis, mitosis, respiration, DNA replication, RNA synthesis, muscle contraction, membrane ion-channel pump, and hormonal effects.1,2 In recent years, ATP in extracellular fluids in the central nervous system (CNS) was found to have modulation effects on many neuronal activities, such as neurotransmission, epithelial secretion, endocrine/exocrine secretion, cardiovascular performance, and immune or inflammatory reactions.3,4 In several types of neuronal systems such as auditory organ and visionary †
Ames LaboratorysUSDOE and Department of Chemistry. Neuroscience Program, Department of Zoology and Genetics. (1) Bridger, W. A.; Henderson, J. F. Cell ATP; John Wiley & Sons: New York, 1983. (2) De Robertis, E. D. P.; Saez, F. A.; De Robertis, E. M. F. Cell and Molecular Biology; Saunders: Philadelphia, 1980. (3) Dubyak, G. R.; el Moatassim, C. Am. J. Physiol. 1993, 265, C577-C606. (4) Bruns, R. F. Ann. N. Y. Acad. Sci. 1990, 603, 211-226. ‡
10.1021/ac9912146 CCC: $19.00 Published on Web 03/31/2000
© 2000 American Chemical Society
retina, lines of supporting evidence5,6 have been established regarding the role of ATP as a neurotransmitter or neuromodulator. Several kinds of techniques were developed to measure the ATP contents in biological samples, including UV absorption,7-9 electrochemical detection,10,11 and luminescence assay.12-19 UV absorption detection is a common technique for measuring components in biological fluids, but it is neither very sensitive nor very specific. Electrochemical detection including chronoampemetry, cyclic voltammetry, and sinusoidal voltammetry has been used to quantify purine base nucleic acids including ATP with improved sensitivity. While the detection limit for ATP is on the order of 70-200 nM at the electrode surface, many biogenic species interfere with the measurement. The most widely used technique to measure ATP in biological fluids and tissues is bio/chemiluminescence involving the enzyme firefly luciferase. The reaction has fast response (approximately milliseconds) and a broad linear range for ATP. This reaction is specific since only ATP will react with the enzyme and not other adenosine-containing nucleotides such as AMP or ADP. The sensitivity is extremely high because the background signal is theoretically zero. The detection limit easily reaches concentrations of 10-10 M or amounts of 10-18 mol at normal conditions. Because of these advantages, various applications have been successfully based on ATP measurements in (5) Newman, E. A.; Zahs, K. R. Science 1997, 275, 844-847. (6) Wangemann, P. Aud. Neurosci. 1996, 2, 187-192. (7) Fujitaki, J. M.; Nord, L. D.; Willis, R. C.; Robins, R. K. J. Biochem. Biophys. Methods 1992, 25 (1), 1-10. (8) Au, J. L.; Su, M. H.; Wientjes, M. G. Clin. Chem. 1989, 35 (1), 48-51. (9) Karon, B. S.; Nissen, E. R.; Voss, J.; Thomas, D. D. Anal. Biochem. 1995, 227 (2), 328-333. (10) Singhal, P.; Kuhr, W. G. Anal. Chem. 1997, 69, 3552-3557. (11) Bianchi, A.; Domenech, A.; Garcia-Espana, E. Anal. Chem. 1993, 65, 31373142. (12) Stanley, P.; McCarthy, B.; Smither, R. ATP Luminescence; Blackwell Scientific Publications: Oxford, U.K., 1989. (13) Abraham, E.; Okunieff, P.; Scala, S.; Vos, P.; Oosterveld, M.; Chen, A.; Shrivastav, B. Science 1997, 275, 1324-1326. (14) Bitler, B.; McElroy, W. Arch. Biochem. Biophys. 1957, 72, 358-368. (15) Denburg, J.; Reiko, L.; McElroy, W. Arch. Biochem. Biophys. 1969, 134, 381-394. (16) Rhodes, W.; McElroy, W. J. Biol. Chem. 1958, 233, 1528-3157. (17) Schifman, R.; Wieden, M.; Broker, J.; Chery, M.; Delduca, M.; Norgard, K.; Palen, C.; Reis, N.; Swanson, J.; White, J. J. Clin. Microbiol. 1984, 20, 644648. (18) Molin, O.; Nilsson, L. J. Clin. Microbiol. 1983, 18, 521-525. (19) Deluca, M., McElroy, W., Eds. Bioluminescence and Chemiluminescence; Academic Press: New York, 1981.
Analytical Chemistry, Vol. 72, No. 9, May 1, 2000 2001
many types of biological tissues and samples under different environments.13, 18, 20-23 Intercellular signaling in glial5,24,25 and other cells26-28 is a subject of intense current interest. Direct monitoring of the temporal and spatial distributions of the various messengers has been elusive except for Ca.29,30 Intracellular cAMP levels can be followed based on the loss of fluorescence resonance energy transfer on dissociation of cAMP-dependent protein kinase.31 Inositol 1,4,5-trisphosphate (IP3) has been detected as the translocation of green fluorescent protein-tagged pleckstrin homology domain from the plasma membrane to the cytoplasm.26 However, imaging of ATP in microscopy has not been reported using any reporter system. With the advent of intensified charge-coupled device (ICCD) cameras with single-photon detection sensitivity, chemiluminescence associated with the ATP-dependent reaction between luciferase and luciferin32 can be used to image ATP at levels down to 10-8 M in the millisecond time scale. Similar imaging experiments should be feasible in a broad spectrum of biological systems.33,34 EXPERIMENTAL SECTION Reagent and Chemicals. Sodium chloride, potassium chloride, magnesium chloride, calcium chloride, D-glucose, HEPES free acid, luciferase from firefly Photinus pyralis (EC 1.13.12.7), D-luciferin, thapsigargin, suramin, and ATP were purchased from Sigma Chemical Co. (St. Louis, MO). Fluo-3, Fluo-3 AM, and BAPTA-AM were obtained from Molecular Probes (Eugene, OR). U-73122 and U-73343 were from Research Biochemicals (Natick, MA). Dimethyl sulfoxide (DMSO) was purchased from Fisher Scientific (Pittsburgh, PA). All reagents were used without further purification. Buffers. All experiments were carried out in fresh saline prepared with purified and deionized water. The buffered saline contained NaCl (135 mM), KCl (5 mM) MgCl2 (7 mM), CaCl2 (2 mM), D-glucose (6 mM), and HEPES (10 mM). The final pH was adjusted to 7.35. All experiments were done at room temperature. (20) Dadoo, R.; Seto, A.; Colon, L.; Zare, R. Anal. Chem. 1994, 66, 303-306. (21) Leach, F. R. J. Appl. Biochem. 1981, 3, 473-517. (22) Lin, S.; Cohen, H. P. Anal. Biochem. 1968, 24, 531-540. (23) Imbeault, N.; Paquet, M.; Cote, R. J. Water Qual. Res. Can. 1998, 33, 403415. (24) Cotrina, M. L.; Lin, J. H. C.; Alves-Rodrigues, A.; Liu, S.; Li, J.; Azmi-Ghadimi, H.; Kang, J.; Naus, C. C. G.; Nedergaard, M. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 15735-15740. (25) Guthrie, P. B.; Knappenberger, J.; Segal, M.; Bennett, M. V. L.; Charles, A. C.; Kater, S. B. J. Neurosci. 1999, 19 (2), 520-528. (26) Hirose, K.; Kadowaki, S.; Tanabe, M.; Takeshima, H.; Iino, M. Science 1999, 284, 1527-1530. (27) York, J. D.; Odom, A. R.; Murphy, R.; Ives, E. B.; Wente, S. R. Science 1999, 285, 96-100. (28) Osipchuk, Y.; Cahalan, M. Nature 1992, 359, 241-244. (29) Poenie, M.; Alderton, J.; Tsien, R. Y.; Steinhardt, R. A. Nature 1985, 315, 147-149. (30) Williams, D. A.; Fogarty, K. E.; Tsien, R. Y.; Fay, F. S. Nature 1985, 318, 558-561. (31) Adams, S. R.; Harootunian, A. T.; Buechler, Y. J.; Taylor, S. S.; Tsien, R. Y. Nature 1991, 349, 694-697. (32) McElroy, W. D.; Seliger, H. H. In Light and Life; McElroy, W. D., Glass, B., Eds.; Johns Hopkins Press: Baltimore, MD, 1961. (33) Redman, R. S.; Silinsky, E. M. J. Physiol. 1994, 477, 117-127. (34) Von Kugelgen, I.; Allgaier, C.; Schobert, A.; Starke, K. Neuroscience 1994, 61 (2), 199-202.
2002
Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
Cell Culture. The enriched astrocyte cultures were prepared as follows:35 Cortices of 1-4-day postnatal rat pups were dissected, dissociated, and placed into culture flasks containing phenol-free modified minimal essential medium (MMEM) (Eagle’s minimum essential medium, 2 mM glutamine, 1 mM sodium pyruvate, 100 units/mL penicillin and 100 mg/mL streptomycin) supplemented with 10% fetal bovine serum. These flasks were placed in a humidified 5% CO2/95% air atmosphere for 8-10 days. The flasks were then rinsed twice with ice-cold MMEM, tapped abruptly to dislodge any loose cells (neurons), and then placed on an orbital shaker for 1.5 h at 260 rpm. The flasks were again rinsed twice with ice-cold MMEM, tapped, and returned to the shaker for 18 h. The adherent cells were detached by trypsinization (0.1%). The detached cells were spun at 750 rpm for 10 min, and the supernatant was discarded. The cells were resuspended in MMEM and plated onto 12-mm round microscope cover slips. They were placed at 37 °C in a humidified 5% CO2/95% air atmosphere and allowed to grow to confluency (∼4 days). Before the experiment began each day, the culture dish was washed with fresh saline described above to eliminate potential bacterial growth and other interference factors. Instrumentation. The imaging system consisted of an inverted light microscope (Nikon Diaphot 300, Fryer, Edina, MN) and a microplate-coupled ICCD (EEV 576 × 384, Princeton Instruments, Trenton, NJ) attached to the camera mount of the microscope to record images. The signal was collected by a 20× objective (N/A 0.75, Zeiss, Germany). The experiments were carried out in a flow chamber on the microscope stage. The flow chamber was made from a piece of Plexiglas slide by drilling a hole with a diameter of 8 mm and a depth of 1.8 mm. The volume was thus ∼90 µL. For standard ATP calibration experiments, a microsyringe was inserted into the chamber through a side channel for injection of ATP solution. For in vitro cell experiments, a mechanical tip made from a tungsten rod was used to stimulate the cells grown on the cover slip in the flow chamber. The mechanical tip was etched to 6-8-µm diameter and controlled by a micromanipulator, as depicted in Figure 1. In both the standard calibration test and in vitro cell experiment, the aliquot of luciferase enzyme mixture was first placed in chamber before the experiment began. The solution was kept still while data were collected. For simultaneous monitoring of the calcium-Fluo-3 fluorescence signal, a 488-nm argon ion laser (Cyonics, San Jose, CA) was coupled to the system. A 488-nm notch filter (Oriel) was put in front of the ICCD to exclude stray light. A customized mechanical shutter was placed in front of the laser. It was controlled by the programmed digital I/O signal sent by the ICCD controller so that the fluorescence signal alternated with ATP signal. When the ICCD was imaging ATP, the shutter was closed to block the laser beam. So, only chemiluminescence was detected by the ICCD. At the next frame, the shutter was open and both chemiluminescence and laser-induced fluorescence signals were acquired. However, the fluorescence intensity was much higher (over 100×) than chemiluminescence so that the latter could be neglected in the image. Then, this cycle of shutter programming was repeated. In every experiment, the ICCD was always allowed to run for a few frames to record the (35) Banker, G.; Goslin, K. Culturing Nerve Cells; MIT Press: Cambridge, MA, 1991.
Figure 1. Instrumentation for simultaneous in vitro imaging of ATP and Ca.
background signal before ATP injection or cell stimulation. Standard ATP Calibration. Aliquots of standard ATP solution with concentrations ranging from 10-8 to 10-6 M were injected into the flow chamber to react with the luciferase-luciferin mixture. The light signal generated was recorded sequentially by the ICCD at 1-s exposure time. The concentrations of the components in the chamber were 40 µg/mL and 0.5 mM for luciferase and D-luciferin, respectively. Each ATP data point was the average of triplicate measurements. In Vitro Astrocyte Monitoring. The astrocyte culture cover slip was carefully attached to the lower surface of the flow chamber with the cell side facing up. The flow chamber was then placed on the microscope stage and filled with reaction reagents prepared in saline. Stimulation to the cells to cause ATP release was via gentle tapping to the cell membrane with the mechanical tip, as described elsewhere.36,37 The alternating ATP signal and calcium signal were recorded sequentially with ICCD at a 2-s frame rate. The exposure times were 500 and 50-100 ms for the ATP and calcium signals, respectively. The final concentrations were 100200 µg/mL and 100-200 µM for luciferase and luciferin, respectively. The laser power at the focus was controlled to less than 1 mW to avoid possible cell damage. Various Pretreatments of Astrocytes. Intracellular calcium concentration changes were monitored by using the fluorescent indicator Fluo-3.38 The loading solution was prepared by dissolving solid Fluo-3 AM into dry DMSO and then diluting with saline to the final concentration. The astrocyte cultures were loaded with Fluo-3 AM (∼10 µM) for 30-45 min at 37 °C. The cover slip was then rinsed and kept in fresh saline for 30 min to let Fluo-3 deesterify before use.39 The intracellular calcium chelator reagent 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetate (BAPTA) AM (25 µM) was
coloaded with Fluo-3 AM into astrocyte cultures for 45 min, and the cover slip was rinsed before the experiment. The preparation of BAPTA AM solution is similar to that of Fluo-3 AM. The calcium blocker thapsigargin was dissolved in saline (1 µM), and the cover slip containing cells was immersed into this solution for 30 min and rinsed before the experiment.37,40 Suramin was dissolved in saline and was add to chamber with a final concentration of 100 µM for 5 min before cell stimulation. The phospholipase-C (PL-C) inhibitor aminosteroid 1-[6-[[17β3-methoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-1H-pyrrole-2,5dione (U-73122) was dissolved in saline (10 µM) and was placed in flow chamber for at least 5 min before the experiment. The solution was kept in the chamber throughout the experiment, i.e., with no rinse step. Similar experiments were also carried out in parallel with its inactive form, U-73343 (10 µM).41-43 Cell-Free Gaps. The cell-free lanes (50-200 µm) were made between regions of confluent astrocytes by physically drawing a
(36) Charles, A. C. Dev. Neurosci. 1994, 16, 196-206. (37) Araque, A.; Parpura, V.; Sanzgiri, R. P.; Haydon, P. G. Eur. J. Neurosci. 1998, 10, 2129-2142. (38) Kao, J. P. Y.; Hartootunian, A. T.; Tsien, R. Y. J. Biol. Chem. 1989, 14, 8179-8184. (39) Hassinger, T. D.; Guthrie, P. B.; Atkinson, P. B.; Bennett, M. V. L.; Kater, S. B. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 13268-13273.
(40) Araque, A.; Sanzgiri, R. P.; Parpura, V.; Haydon, P. G. J. Neurosci. 1998, 18 (17), 6822-6829. (41) Smallridge, R. C.; Kiang, J. C.; Gist, I. D.; Fein, H. G.; Galloway, R. J. Endocrinology 1992, 131 (4), 1883-1888. (42) Yule, D. I.; Williams, J. A. J. Biol. Chem. 1992, 267, 13830-13835. (43) Yamada, M.; Yamada, M.; Richelson, E. Eur. J. Pharmacol. 1992, 226, 187188.
Figure 2. Time course of ATP reaction and (inset) standard calibration curve.
Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
2003
Figure 3. Temporally and spatially resolved ATP release patterns from astrocytes taken by an ICCD camera at 2-s intervals with exposure times of 500 ms. The image sequence starts from upper left to right and then down. The first corresponds to the bright-field image of the astrocyte culture, showing the microtip used for mechanical stimulation positioned over a glial cell, and the second corresponds to the frame at which the stimulation is applied. Scale bar at lower right, 50 µm. See also movie file Figure 3-ATP.AVI in Supporting Information.
A
Figure 4. Simultaneous monitoring of ATP and [Ca2+]i from astrocyte cultures showing that the calcium wave is synchronized with the released ATP upon stimulation. The exposure time for ATP is 500 ms and that for [Ca2+]i is 50 ms. (A) (a) Bright-field image showing the position of the stimulation tip followed by ATP images. Times are relative to the time of stimulation. See also movie file Figure 4-ATP.AVI in Supporting Information. (b) Corresponding calcium images taken at 0.5-s delay relative to each ATP image shown above it. See also movie file Figure 4-Ca.AVI in Supporting Information. Cell 1 is at the position of stimulation. Initially, only the calcium level in cell 1 increased. As the experiment progressed, cells 2 and 3, then cells 4 and 5, and finally cell 6 showed increases in calcium level synchronized with the arrival of the ATP wave shown in (a). The last image showed calcium levels falling back to normal coinciding with the fading of the ATP wave. (B) Time course of ATP releases from four individual cells. (C) Correspondent calcium changes in the same four cells showing start times synchronized with ATP release.
fine mechanical tip across the cover slip and allowing 30 min for cell recovery. At the time of experiment, the cell-free lanes were bordered by healthy cells loaded with Fluo-3. These displayed normal baseline fluorescence compared with controls. 2004 Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
RESULTS AND DISCUSSION Dynamic Imaging of ATP by Chemiluminescence. The chemiluminescence assay for ATP is known. D-Luciferin (LH2) combines with luciferase (E) to form the enzyme-coenzyme
Figure 5. Repetitive stimulation of astrocytes. (A) Integrated ATP release intensities: 1, first stimulation; 2, second stimulation to the same cells after a few minutes; 3, third stimulation. Cells are numbered according to their distances from the stimulation point. Cell 1 is the closest and cell 3 is the farthest. (B) Simultaneous calcium changes in cell 1.
E-LH2 intermediate. When ATP exists in solution, this intermediate will quickly react with it to form the E-L-AMP complex on the enzyme surface and release pyrophosphate. Under normal aerated conditions, the E-L-AMP complex will dissociate to E-L and AMP, accompanied by light generation around 560 nm. This reaction is catalyzed by Mg2+ in solution,44 which was supplied from the saline in our experiment. Without Mg2+, the reaction efficiency is greatly reduced (over 90%). The optimal pH is 7.8 for maximum enzyme activity.45 At pH 7.35, which was our experimental condition, the enzyme activity was at ∼80% of maximum. It is reported44 that a few anions, including SCN-, I-, NO3-, and Br-, have inhibitory effects on the reaction. Because these anions are generally not present in cells and are not in our experimental buffer solutions, they are not of concern here. Figure 2 shows the light intensity as a function of time from the injection of standard ATP solution observed by using an ICCD camera. After the reactants are mixed, there is at first a rapid rise in light intensity. Then the signal falls down to the second phase where there is a relatively slow decrease over time. This is mainly due to reaction kinetics hindered by the generation and accumulation of pyrophosphate from the reaction.32 The height of the first peak in Figure 2 is found to be linear with ATP concentration when ATP is the limiting reagent. The inset in Figure 2 shows the standard calibration curve from our imaging setup. The linearity of this enzyme reaction is retained, with a (44) Denburg, J.; McElroy, W. Arch. Biochem. Biophys. 1970, 141, 668-675. (45) Seliger, H. H.; McElroy, W. D. Arch. Biochem. Biophys. 1960, 88, 136141.
correlation coefficient of 0.999 over at least 3 orders of magnitude. The detection limit is down to 10-8 M ATP with subsecond ICCD exposure. The ATP concentration in extracellular fluids in the CNS is reported to range from a few hundred nanomolar to millimolar,25,46 which makes this a broadly applicable imaging method in neurochemical analysis. Monitoring ATP Releases from Rat Astrocyte Cultures. Confluent astrocytes47 were stimulated mechanically via microtips (Figure 3). While not a natural stimulus, mechanical stimuli provide a high degree of spatiotemporal control that mimics the action of physiological ligands.48 Consistent with the extracellular signaling hypothesis, physical contact, which is known to cause a calcium wave, immediately generated an extracellular ATP signal. As soon as an ATP signal appeared, the tip was withdrawn vertically. Extracellular diffusion during exposure limits the spatial resolution of these ATP images to 10 µm. Consistent with previous observations of calcium wave propagation,39 the average effective ATP wave excursion distance is 258 ( 50 µm (n ) 25). This argues against the single point release model, in which the effective signal travel distance is 110 ( 30 µm,49,50 and implies distinct contributions from cells along the path. The existence of a finite propagation range for the ATP wave, plus the fact that the intensities decrease gradually along the path, support the postulation51,52 that subpopulations of astrocytes express different specific purinergic receptor subtypes. This dictates that not all astrocytes have the same sensitivity to ATP and they do not release identical amounts of ATP when stimulated. From the injection of standard ATP solution, the concentration of ATP released into the extracellular fluid can be estimated to be at the micromolar range, which agrees with previous reports.25,46 The basal release is not detectable. Since intracellular ATP concentrations are in the millimolar range, the present reporter scheme cannot detect those changes that result from stimulation. Synchronization of ATP Wave and Evoked Intercellular Calcium Wave. ATP is known to evoke intercellular calcium wave propagation from glial5,53 and other28 cells. The astrocyte cultures were loaded with Fluo-338 and the fluorescence on binding with intracellular Ca, [Ca2+]i, was monitored simultaneously with the chemiluminescence signal from the released ATP. On stimulation, the intercellular calcium wave is found to be synchronized spatially and temporally with the extracellular ATP wave (Figure 4). An important feature is that [Ca2+]i persists at an elevated level for some time after stimulation while the ATP level decreases soon after stimulation. This is an inherent difference between intracellular components that are trapped and extracellular components that can diffuse away. Also, there is a dramatic decrease in the ATP signal but only a modest decrease in the [Ca2+]i signal as a function of distance from the point of stimulation. Yet, the calcium wave does not propagate beyond the excursion range of the ATP wave. This suggests that the two waves are (46) Silinsky, E. M.; Redman, R. S. J. Physiol. 1996, 492, 815-822. (47) Levison, S. W.; McCarthy, K. D. In Culturing Nerve Cells; Banker, G., Goslin, K., Eds.; MIT Press: Cambridge, MA, 1991; pp 309-336. (48) Sanzgiri, R. P.; Araque, A.; Haydon, P. G. J. Neurobiol. 1999, 41, 221-229. (49) Crank, J. The Mathematics of Diffusion; Clarendon: Oxford, U.K., 1975. (50) Hazel, J. R.; Sidell, B. D. Anal. Biochem. 1987, 166, 335-341. (51) Pearce, B.; Langley, D. Brain Res. 1994, 660, 329-332. (52) Ho, C.; Hicks, J.; Salter, M. W. Br. J. Pharmacol. 1995, 116, 2909-2918. (53) Van Den Pol, A. N.; Finkbeiner, S. M.; Cornell-Bell, A. H. J. Neurosci. 1992, 12, 2648-2664.
Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
2005
Figure 6. Diffusion of ATP across a cell-free zone. ATP evokes calcium elevations in astrocytes physically isolated from the releasing cells. (a) Bright-field image showing the stimulation position and a physically induced cell-free zone followed by images showing the ATP wave propagating to different distances at different times. See also movie file Figure 6-ATP.AVI in Supporting Information. (b) Corresponding simultaneous calcium-Fluo-3 fluorescence images. See also movie file Figure 6-Ca.AVI in Supporting Information. Cell 1 is at the stimulation position. Cells on the same side of the gap (2 and 3) showed evoked [Ca2+]i elevation first according to the order of their distances to the stimulation position. After a delay, [Ca2+]i elevation was observed from cells on the other side of the gap (4), synchronized to ATP crossing the gap.
not related to each other in a straightforward manner. Repetitive Stimulation of Astrocytes. To verify that the mechanical stimulation process did not create any artifacts, the same astrocyte cell can be stimulated again after 2 min. A second ATP wave and the associated calcium wave can be observed at a reduced intensity (Figure 5), confirming the integrity of the cells. However, no additional signals were detected at the third and subsequent stimulations, indicating a state of overstimulation. Afterward, a more drastic mechanical stimulation was applied. A small and localized ATP signal of short duration was recorded with the absence of any calcium signal. This last event is consistent with leakage after physically breaking open the cell and diffusion away from a single point rather than release followed by a physiological ATP wave. Signaling among Physically Isolated Cells. There are conflicting reports as to whether signaling in astrocytes is extracellular or requires gap junctions.39,54 We tested signaling on physically isolated astrocytes.39 We made cell-free gaps with 50-200-µm widths in confluent cultures by depleting astrocytes locally with a mechanical tip (Figure 6a, bright-field image). After 30-min cell recovery, the stimulated ATP wave was able to travel across the gap (
