INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

EVIDENCE FROM MANY CELL TYPES, including smooth muscle (7, 23), supports the essential feature of the capacitative calcium entry model (3, 26). In that model of intracellular calcium store refilling, the calcium content of intracellular calcium stores regulates or modulates calcium influx across the plasma membrane. How a decrease in the calcium content of intracellular stores stimulates the calcium influx that refills intracellular calcium stores is not known, but possibilities have included signaling by direct protein-protein interactions (15), guanyl nucleotide binding proteins (11), diffusible messengers (28), and protein tyrosine kinase (PTK) phosphorylations (13, 20, 30, 34).

Among different cell types, there may be differences in the extent to which cytosolic calcium concentration ([Ca²⁺]_i) is the determinant of intracellular store refilling. In most nonexcitable cells such as parotid acinar cells (27, 32), pancreatic acini (25), and human leukemia cells (24), it was shown that an increase in [Ca²⁺]_i was necessary for refilling of intracellular stores. However, in human fibroblasts, intracellular calcium stores refilled without increases in [Ca²⁺]_i (5). Similarly, in vascular smooth muscle, calcium influx refilled intracellular stores without stimulating contraction (7, 8, 19, 33). For canine airway smooth muscle, agonist-sensitive intracellular stores refilled without the development of tension (4), and this suggested the presence of direct or privileged refilling pathways that are separated from the inner cytosol (4, 17).

Our first goal was to assess directly whether the [Ca²⁺]_i of the inner cytosol determined the rate that caffeine-sensitive intracellular calcium stores refilled. For this, we reasoned that if the inner cytosol was the only calcium compartment determining refilling, then the effects of a calcium-channel antagonist on [Ca²⁺]_i should correlate well with the effects that the same antagonist has on refilling of intracellular stores. Therefore, we loaded isolated tracheal smooth muscle cells with fura 2 under conditions favoring detection of [Ca²⁺]_i in the inner cytosol (36) and then compared the effects that the inorganic calcium-channel antagonist nickel had on [Ca²⁺]_i versus its effects on sequential responses to caffeine. A second goal of this study was to begin to assess the role that protein kinases (PKs) played in regulating these pathways.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Cell isolation. Tracheal smooth muscle cells were dispersed from minces (1 × 1 mm) of bovine trachealis muscle cut on a tissue chopper (McIlwain). Approximately 250 mg of tissue were placed in a Coulter counter vial containing a magnetic stirring bar and 2.5 ml of a physiological salt solution (PSS; in mM: 118 NaCl, 4.7 KCl, 1.2 KH₂PO4, 1.2 MgSO₄, 25.6 NaHCO₃, 11.1 glucose, and 2.5 CaCl₂) modified to have no added CaCl₂ and containing collagenase (6 mg; Boehringer Mannheim) and elastase (3 mg; Boehringer Mannheim). The tissue minces were incubated at 37°C with constant stirring for 12 min, and then the mince was transferred to another vial and the incubation was repeated. The supernatant during the second incubation was monitored repeatedly by microscopy, and the incubation was terminated when cells began to be released from the minces. The partially digested mince was then transferred to 2 ml of PSS modified to contain 0.1 mM CaCl₂ and incubated for 3 min at 37°C with constant stirring. The cells released during this and one subsequent identical incubation were used for our studies. The dispersed cells were loaded with 0.5 µM fura 2-AM in the presence of Pluronic F-127 (0.004%) for 60 min at room temperature and then introduced into a superfusion chamber having a bottom cover glass. After adherence to the glass for 10 min at room temperature, the chamber was superfused with PSS at 37°C.

Calcium measurement. A computer-assisted fluorescent-imaging microscope system was similar to one previously described (22, 36). Fura 2-loaded cells were excited by a computer-controlled 337- and 380-nm ultraviolet light generated by a nitrogen laser and a nitrogen laser-pumped dye laser, respectively (Laser Science, Cambridge, MA). Each laser alternately fired short laser pulses (3 ns) at 30 Hz, and these alternating pulses of light were guided by a bifurcated quartz fiber to a neutral density filter at the epiport of the microscope and were then focused on the cells through a ×40 objective lens (Nikon). The fluorescent signals emitted by fura 2 were passed back through the objective to a 455-nm dichroic mirror, a 475-nm barrier filter (Omega Optics, Brattleboro, VT), and an image intensifier (Xybion Electronic Systems, San Diego, CA) and were captured by a Philips-based frame transfer charge-coupled device camera (CCTV, New York, NY). With the gain of the intensifier set at 50% maximum, slow quenching of the intensifier screen image was minimized such that a fura 2 fluorescence signal corresponding to the 380-nm laser was not detectable when only the 340-nm laser was firing. Similarly, a fluorescence signal corresponding to the 340-nm laser was not detectable when only the 380-nm laser was firing. The analog signals from the camera were digitized and stored in an imaging board, and digital outputs from this board were transferred to a personal computer (386SX, NEC) with software by Recognition Technology (Westborough, MA).

To measure [Ca²⁺]_i in cells loaded with fura 2, a background level of light from a cell-free region of the cover glass (typically <1% of the light detectable over cells) was subtracted before data acquisition, and then an 11 × 11-pixel area was selected over each cell. Areas of the cell containing the nucleus were avoided. The gray levels of fluorescence emissions stimulated by alternating pulses of 337- and 380-nm light were recorded, and their ratios were plotted. The ratio (R) was converted to [Ca²⁺]_i with the equation (12) [Ca²⁺]_i = K_D ·  · (R  R_min)/(R_max  R), where R_max and R_min are the fluorescence ratios measured in situ with permeabilized (4-bromo-A-23187) fura 2-loaded cells exposed to high (2.5 mM CaCl₂) and zero calcium, respectively;  is the ratio of fluorescence stimulated by 380-nm light in zero versus high calcium; and K_D is the equilibrium dissociation constant describing calcium binding to fura 2. Based on an in situ determination of K_D in bovine trachealis cells (18), a K_D value of 386 nM was used in converting observed fluorescence ratios to [Ca²⁺]_i.

Protocol. Cells loaded with fura 2 and attached to the glass coverslip of the perfusion chamber (0.3-ml volume) were perfused with PSS (2.5 mM CaCl₂) at 1 ml/min at 37°C for at least 30 min before the start of experiments. Then the following protocol was used. The cells were perfused with nominally calcium-free PSS for 2 min. The cells were then perfused with calcium-free PSS containing caffeine (10 mM) for 2 min, and the cell response to this first caffeine stimulation (S1) was recorded. After a 2-min wash by perfusion with calcium-free PSS, a recovery phase began. During the recovery phase, the cells were perfused with buffer containing specified reagents and calcium concentrations for defined times. At the end of the recovery phase, the cells were washed with calcium-free PSS for 2 min before being perfused again with calcium-free PSS containing caffeine (10 mM). The cell response to this second stimulation with caffeine (S2) was recorded. Whenever the effects of Ni²⁺, cyclopiazonic acid (CPA), methoxyverapamil (D-600), genistein, -cyano-(3-ethoxy-4-hydroxy-5-phenylathiomethyl)cinnamamide (ST-638), daidzein, 1-(5-isoquinolinesulfonyl)-2-methylpiperazine, dihydrochloride salt (H-7), 3-[1-[3-(amidinothio)propyl-1H-indoyl-3-yl]-3-(1-methyl-1H-indoyl-3-yl)maleimide] methane sulfonate (Ro-31-8220), the Rp diastereomer of adenosine 3',5'-cyclic monophosphothioate (Rp-cAMPS), and PD-98059 on refilling were tested, these reagents were present at specified concentrations for 2 min before the addition of caffeine and throughout the S1 recording, recovery phase, and S2 recording. In preliminary experiments, genistein and ST-638 alone had no significant effects on the intensity of fura 2 fluorescence. After 10 min of perfusion, the 340- and 380-nm fluorescent signals increased by 2 ± 3 and 4 ± 4%, respectively, in response to genistein alone (n = 5 cells). In response to ST-638 (50 µM), the 340- and 380-nm fluorescent signals decreased by 3 ± 3% [not significant (NS)] and increased by 0.2 ± 1.5% (NS), respectively (n = 5 cells).

After recordings from a single cell were made, the cell chamber was perfused with PSS (2.5 mM CaCl₂) at 37°C for 10-20 min before selection of another cell on the same coverslip. For this study, 108 tracheae were used. One to two coverslips were prepared for each trachea. The number of different cells studied per coverslip was 1-10. Each individual cell was studied only once. Multiple exposures of coverslips to caffeine did not change the baseline [Ca²⁺]_i levels from which responses to caffeine were measured; for the first and last cells recorded from on 14 separate days, [Ca²⁺]_i levels before caffeine was added differed by only 2 ± 13 nM (NS).

Reagents. Fura 2-AM and Pluronic F-127 were obtained from Molecular Probes (Eugene, OR). Genistein, daidzein, CPA, and Rp-cAMPS were obtained from RBI (Natick, MA). PD-98059 was obtained from New England Biolabs (Beverly, MA). H-7 was obtained from LC Laboratories (Woburn, MA). ST-638 and Ro-31-8220 were obtained from Calbiochem (San Diego, CA). All other reagents were obtained from Sigma (St. Louis, MO).

Data analysis. For recordings from single cells, peak calcium responses to caffeine and changes in [Ca²⁺]_i during recovery were measured from the baseline value for [Ca²⁺]_i during perfusion with calcium-free PSS. The maximum or peak change in [Ca²⁺]_i in response to S1 and S2 was expressed in nanomoles, and then the ratio of the two values (S2/S1) was calculated. For the recovery period between the S1 and S2 stimulations, the maximal levels for [Ca²⁺]_i are expressed in nanomoles. All data are expressed as means ± SE, and n is the number of cells studied. For multiple comparisons between groups of mean data, analysis of variance was followed by a Newman-Keuls test.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Effects of extracellular calcium on resting [Ca²⁺]_i and on a second response to caffeine. Perfusing cells with nominally calcium-free PSS decreased the resting [Ca²⁺]_i from 187 ± 9 to 90 ± 8 nM (n = 50 cells from 7 tracheae) within 2 min [half-time = 25 ± 3 s]. The magnitude of this decrease in [Ca²⁺]_i was not limited by an inability to detect changes in the fluorescent ratio when [Ca²⁺]_i levels were low. For example, in 13 cells, fluorescent ratios decreased from 1.50 ± 0.14 to 1.21 ± 0.09 on perfusion with calcium-free PSS. When the same cells were then perfused with calcium-free PSS containing the ionophore 4-bromo-A-23187, the ratio transiently increased and then rapidly and significantly decreased to a steady-state value of 0.72 ± 0.04 (P < 0.0005), a value agreeing closely with R_min.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Recovery of responses to caffeine. Cells were perfused with calcium-free physiological salt solution (PSS) for 2 min and then perfused with calcium-free PSS containing 10 mM caffeine. Increase in fluorescence ratio in response to stimulation with caffeine (S1) was recorded, and maximum increase in fluorescence ratio was converted to cytosolic calcium concentration ([Ca2+]i) in inner cytosol. After a 2-min exposure to caffeine, cells were washed with calcium-free PSS for 2 min. After this wash, cells were perfused with PSS with calcium (2.5 mM) for defined times (80 s for this cell), and this was designated recovery phase. After recovery phase, cells were perfused for 2 min with calcium-free PSS and then exposed to caffeine (10 mM) for a 2nd time. Increase in fluorescence ratio in response to 2nd stimulation with caffeine (S2) was recorded. Continuous recordings were not done during washes to minimize photobleaching (thick dotted line). A representative trace for a single cell is shown. Trace is ratiometric, with a fluorescence ratio of 4.0, corresponding to [Ca2+]i of 1,200 nM. [Ca2+]i increased to a maximum level of 210 nM during 80-s recovery phase, and S2-to-S1 ratio (S2/S1) was 0.52 for this particular cell.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Dependence of S2/S1 on duration of recovery phase and presence of extracellular calcium during recovery phase. Cells were stimulated with caffeine as described in Fig. 1, but duration of perfusion with extracellular calcium during recovery phase varied (0-20 min; ). In other cells, duration of recovery phase varied (0-20 min), but cells were perfused with calcium-free PSS during recovery phase (). Data are means ± SE for 3-36 cells.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Inhibition of recovery of S2/S1 by cyclopiazonic acid (CPA). Cells were initially stimulated (S1) with caffeine (10 mM) in absence of extracellular calcium. Each cell was then allowed an 80-s (A) or a 10-min (B) recovery period with and without extracellular calcium and with and without CPA. After recovery phase, a 2nd stimulation with caffeine in absence of extracellular calcium was recorded, and S2/S1 was calculated. Data are means ± SE for 4-9 cells isolated from a total of 12 tracheae. * P < 0.05 compared with recovery in calcium without CPA present.

Effects of nickel on resting [Ca²⁺]_i and S2/S1. To assess the correlation between [Ca²⁺]_i and refilling, we compared the effects that the inorganic calcium-channel antagonist nickel had on [Ca²⁺]_i versus its effects on S2/S1. In the presence of extracellular calcium, resting [Ca²⁺]_i was 186 ± 6 nM (n = 34) and this rapidly (half-time = 24 ± 5 s) decreased by 45 ± 3.5% when the cell was perfused with PSS containing nickel (2 mM). The effect of nickel on resting [Ca²⁺]_i was concentration dependent, with half-maximal decreases in resting [Ca²⁺]_i observed at 0.1-0.5 mM nickel (Fig. 4A). In contrast, half-maximal inhibition of S2/S1 (from 0.72 to 0.52) required at least 2.6 mM nickel (Fig. 4B). In these experiments, the effects of extracellular nickel depended on changes in intracellular calcium and were not due to a direct effect of nickel on fura 2 fluorescence; for example, in five cells perfused with calcium-free PSS, the addition of extracellular nickel to the perfusate for 10 min increased the 340-nm fluorescent signal by 0.2 ± 6.5% (NS) and decreased the 380-nm fluorescent signal by 0.4 ± 4.0% (NS). Also, our results were not due to an effect of nickel on S1 responses alone because, in paired experiments, the presence of nickel had no significant effects on the magnitude of the S1 responses (14 ± 16% increase with nickel; NS; n = 5). Also, for three reasons, our results could not be attributed to a problem in washing high concentrations of nickel from the chamber between recordings of different cells. First, the effects of nickel (2 mM) on the resting [Ca²⁺]_i could be reversed within 97 ± 16 s of washing (n = 5), and we always washed the cells for 10-20 min between recordings of different cells. Second, for cells never exposed to nickel versus cells exposed 5-10 times, the resting [Ca²⁺]_i differed by only 1 ± 18 nM (NS; n = 8). Third, the relative resistance of S2/S1 to nickel was evident even when coverslips were exposed to nickel (2 mM) only once (S2/S1 0.51 ± 0.09; n = 7).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effects of nickel on resting [Ca2+]i (A) and S2/S1 (B). A: resting cells perfused with PSS containing calcium (2.5 mM) were exposed to PSS containing calcium (2.5 mM) + nickel (0-5 mM), and maximum decreases in [Ca2+]i were recorded. [Ni2+], nickel concentration. Data are means ± SE for 3-50 cells. B: nickel (0-7.5 mM) was included in perfusate during an 80-s recovery phase in PSS containing calcium (2.5 mM), and S2/S1 was determined for each cell. Data are means ± SE for 6-36 cells.

Effects of nickel on [Ca²⁺]_i during the recovery phase and on S2/S1. We next compared the effects that nickel had on [Ca²⁺]_i during the 80-s recovery phase to the effects of nickel on recovery of S2/S1 (Figs. 5 and 6). When the cells were perfused with extracellular calcium (2.5 mM) during an 80-s recovery phase, [Ca²⁺]_i during the recovery phase increased maximally to 248 ± 39 nM (n = 17; Fig. 1). When no extracellular calcium was introduced during an 80-s recovery phase, [Ca²⁺]_i remained stable and the maximum [Ca²⁺]_i measured during recovery was 116 ± 19 nM (n = 13; Fig. 6A). Similarly, when the cells were perfused with extracellular calcium (2.5 mM) plus nickel (2 mM) during the 80-s recovery phase, [Ca²⁺]_i remained stable, with a maximum level of 121 ± 12 nM (n = 21).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Recovery of responses to caffeine in presence of nickel. Cells were perfused and responses were determined as described in Fig. 1 except that nickel (2 mM) was present. Continuous recordings were not done during washes to minimize photobleaching. A representative trace for a single cell is shown. Trace is ratiometric, with a fluorescence ratio of 4.0, corresponding to [Ca2+]i of 1,200 nM. In this case, [Ca2+]i remained constant during 80-s recovery phase, and S2/S1 was 0.55 for this particular cell.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Effects of nickel with different perfusates on [Ca2+]i (A) and S2/S1 (B) during recovery phase. Cells were stimulated with caffeine as described in Fig. 1, and S2/S1 was calculated. Data are means ± SE for 13-36 cells isolated from a total of 21 tracheae. * P < 0.05 compared with 2.5 mM calcium.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Effects of protein tyrosine kinase inhibitors on sequential responses to caffeine. In presence of vehicle (control) or indicated protein tyrosine kinase inhibitors, cells were perfused with calcium-free PSS for 2 min and then perfused with calcium-free PSS containing 10 mM caffeine. Increase in [Ca2+]i in response to this initial stimulation (S1) by caffeine was recorded. Cells were then washed with calcium-free PSS for 2 min. After this wash, cells were perfused with PSS containing calcium (2.5 mM) and indicated protein tyrosine kinase inhibitor for an 80-s recovery phase. After recovery phase, cells were perfused for 2 min with calcium-free PSS and then again stimulated with caffeine (S2). Indicated protein tyrosine kinase inhibitors were present continuously throughout S1, recovery phase, S2, and washes. Data are means ± SE for 6-36 cells isolated from 19 tracheae. * P < 0.05 compared with control.

View larger version (14K): [in this window] [in a new window]   Fig. 8.   Effects of genistein plus nickel on sequential responses to caffeine. Cells were initially stimulated (S1) with caffeine (10 mM) in absence of extracellular calcium. Each cell was then allowed an 80-s recovery phase in PSS containing calcium, nickel, and genistein as indicated. After recovery phase, a 2nd stimulation with caffeine (S2) in absence of extracellular calcium was recorded, and S2/S1 was calculated. When cells were exposed to nickel and genistein, they were present continuously throughout S1, recovery phase, S2, and washes. Data are means ± SE for 8-36 cells.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

This study of bovine tracheal smooth muscle cells shows that during the refilling of caffeine-sensitive SR calcium stores by extracellular calcium, the level of [Ca²⁺]_i in the inner cytosol correlates poorly with the extent of refilling. The same concentration of nickel that inhibited extracellular calcium from reaching the inner cytosol did not significantly inhibit extracellular calcium from reaching depleted caffeine-sensitive stores. Although we do not exclude the possibility that calcium ions in the inner cytosol can be taken up by caffeine-sensitive stores, we do conclude that the inner cytosol is not the only calcium-containing compartment from which the caffeine-sensitive SR calcium stores refill. The finding suggests that there are refilling pathways, poorly imaged by fura 2, that are functionally or anatomically segregated from the inner cytosol. The second major conclusion of this study is that inhibitors of PTKs antagonize refilling of caffeine-sensitive stores in bovine airway smooth muscle cells. This finding suggests that PTKs may regulate or modulate the refilling.

When cells were stimulated by caffeine in the absence of extracellular calcium, calcium was released rapidly from the intracellular stores, producing a calcium transient. Eliciting a second calcium transient in response to caffeine then depended on there being a recovery phase during which the cell was perfused by buffer containing calcium. This recovery, which depended on the presence of extracellular calcium, was also dependent on time and was inhibited by CPA, an inhibitor of SR Ca²⁺-ATPases. Given these findings, recovery of the second response to caffeine was used as an index of caffeine-sensitive calcium store refilling and is expressed as S2/S1.

When bovine airway smooth muscle cells are loaded with relatively low concentrations of fura 2-AM (0.2-0.5 µM), the resulting fluorescent signal is dominated by fluorescence from the inner, as opposed to the peripheral, cytosol (36). For cells loaded with low concentrations of fura 2, both the [Ca²⁺]_i levels of the inner cytosol and the refilling of caffeine-sensitive calcium stores were strongly dependent on the presence of extracellular calcium. Surprisingly, however, the inorganic calcium-channel antagonist nickel (2 mM) maximally decreased the resting [Ca²⁺]_i in the inner cytosolic compartment but had no significant effect on refilling. This differential effect of nickel on resting [Ca²⁺]_i versus refilling was not absolute, but the EC₅₀ for inhibiting refilling was greater than that for decreasing resting [Ca²⁺]_i. These initial findings suggested that the channels through which calcium passed to the inner cytosol to maintain resting [Ca²⁺]_i levels had a different sensitivity to nickel than the channels subserving refilling of caffeine-sensitive stores. Therefore, in other experiments, we measured [Ca²⁺]_i of the inner cytosol during the recovery phase. With an 80-s recovery phase in PSS containing calcium (2.5 mM) plus nickel (2 mM), [Ca²⁺]_i levels did not significantly increase or decrease during the recovery phase, and the maximal level of [Ca²⁺]_i during recovery was only 121 ± 12 nM, a value not significantly different from the [Ca²⁺]_i levels during the recovery phase in the absence of extracellular calcium. Nonetheless, S2/S1 values significantly increased for cells perfused with calcium plus nickel but not for cells recovered in the absence of extracellular calcium. Therefore, [Ca²⁺]_i during the recovery phase did not predict the extent of refilling. We concluded that the inner cytosol was not the only compartment through which extracellular calcium passed to refill caffeine-sensitive SR calcium stores.

Refilling of caffeine-sensitive stores has two important characteristics. First, the refilling was not dependent on calcium influx through voltage-gated channels because D-600 and high concentrations of nickel had little effect on refilling. Second, the refilling depended on SR Ca²⁺-ATPase activity because CPA effectively inhibited refilling. The first finding agrees well with a prior study (21) of cultured porcine airway smooth muscle cells where verapamil did not inhibit refilling of caffeine-sensitive stores. However, the second finding disagrees with that same prior study in which CPA had no effect on the refilling of caffeine-sensitive stores. An explanation for our different results is not certain but possibilities include differences in species, our use of acutely dispersed cells rather than cultured cells, our measurement of [Ca²⁺]_i with fura 2 versus whole cell Ca²⁺-activated chloride currents, and our assessment of the magnitude of responses to caffeine in the absence rather than the presence of extracellular calcium. Interestingly, in that prior study, thapsigargin, a different inhibitor of SR Ca²⁺-ATPase activity, had a partial inhibitory effect on the refilling of caffeine-sensitive stores, and this finding suggests that SR Ca²⁺-ATPase activity does contribute to the refilling of caffeine-sensitive stores, in agreement with our findings. It is possible that, depending on specific experimental conditions, different refilling pathways, dependent or independent of SR Ca²⁺-ATPase activity, can be detected for caffeine-sensitive stores in airway smooth muscle.

Although there is significant functional overlap between caffeine-sensitive stores and agonist-sensitive intracellular calcium stores in airway smooth muscle, refilling of these different stores cannot be assumed to be the same (21). However, similar to our findings for caffeine-sensitive stores, it is notable that many studies (4, 7, 8, 19, 33) of smooth muscle suggest that agonist-sensitive stores can be refilled from compartments that are separate from the inner cytosol. In these studies, the fact that muscle can remain quiescent during refilling suggests that some portion of calcium influx is segregated from the [Ca²⁺]_i of the inner cytosol that determines tension. Specifically, in airway smooth muscle, there have been at least two refilling pathways for agonist-sensitive stores described (4, 17, 21). One pathway is dependent on SR Ca²⁺-ATPase activity, is not dependent on voltage-gated calcium channels, and is not segregated from the inner cytosol (4, 17). A second pathway is independent of SR Ca²⁺-ATPase activity, depends on voltage-gated calcium channels, and is segregated from the cytosol (4, 17). A different study (21) of cultured porcine airway smooth muscle cells also found evidence of two pathways for the refilling of agonist-sensitive stores, but the characteristics of the pathways were different. In that study, the major refilling pathway was dependent on SR Ca²⁺-ATPase activity, was dependent on voltage-gated calcium channels, and was segregated from the cytosol. A second, minor refilling pathway was independent of both SR Ca²⁺-ATPase activity and voltage-gated calcium channels. All these findings for agonist-sensitive stores support the existence of multiple refilling pathways, at least some of which are segregated from the inner cytosol of the cell. Notably, a refilling pathway that is dependent on SR Ca²⁺-ATPase activity, independent of voltage-gated calcium channels, and yet segregated from the inner cytosol has not been described for agonist-sensitive stores. Therefore, our finding a pathway with these characteristics for the refilling of caffeine-sensitive stores constitutes additional evidence that there are differences between the pathways refilling caffeine- versus agonist-sensitive stores.

The presence of privileged or segregated refilling pathways has not been found for many nonexcitable cells. For example, in human leukemia cells, refilling depended on increases in [Ca²⁺]_i, and in that study, nickel (5 mM) effectively antagonized both increases in [Ca²⁺]_i and refilling (24). Also, in pancreatic acinar cells (25), parotid cells (32), and endothelial cells (16), refilling of intracellular stores depended on increases in [Ca²⁺]_i during refilling. Notably, however, the pathways for the refilling of intracellular stores in human fibroblasts appear to be different than in other nonexcitable cells and are similar to our findings in smooth muscle (5). In fibroblasts, repetitive calcium responses to bradykinin depended on calcium influx from the extracellular space but did not depend on increases in [Ca²⁺]_i. In that study, 5 mM nickel inhibited calcium influx to the cytosol but did not inhibit refilling of intracellular stores. Therefore, even among nonexcitable cells, there may be important cell-specific differences in SR refilling mechanisms.

Several possibilities might account for how the inner cytosol is not the only calcium compartment determining refilling of caffeine-sensitive stores in airway smooth muscle. The first possibility to explain our findings is that the SR functionally segregates calcium influx from the inner cytosol. This possibility is suggested by the superficial barrier hypothesis (8) but is not incompatible with the modified capacitative hypothesis (27). In the superficial barrier model, the SR immediately adjacent to the plasma membrane is able to take up entering calcium before it reaches the inner cytosol, and, therefore, the SR functionally segregates calcium influx from the inner cytosol. In this model, portions of the peripheral SR nearest the plasma membrane are exposed to high local calcium concentrations that lie between the SR and the plasma membrane, and it is these high local calcium concentrations that determine refilling by SR Ca²⁺-ATPases. In this model, the effects of nickel could be explained in two ways. First, the influx of calcium ions into regions lying between the plasma membrane and the SR could be mediated by channels that are relatively insensitive to nickel compared with other types of channels delivering calcium ions directly to the inner cytosol. Alternatively, and possibly more likely, high concentrations of nickel (2 mM) could nonspecifically slow the rate of calcium influx such that the fast rate of calcium uptake by the peripheral SR prevents calcium ions from reaching the inner cytosol. Common to both of these explanations for the effects of nickel, there is a functional segregation of calcium by SR uptake of calcium; that is, peripheral SR uptake of calcium prevents high concentrations of calcium at the periphery of the cell from affecting calcium concentrations in the inner cytosol. In support of this, Yamaguchi et al. (36) recently reported that in bovine airway smooth muscle cells calcium concentrations immediately beneath the plasma membrane were higher than calcium concentrations in the inner cytosol.

Anatomic segregation of the calcium influx supporting refilling could be another possible mechanism to explain how the inner cytosol is not the only calcium compartment determining refilling of caffeine-sensitive stores (4, 5, 17, 21, 26). That is, we do not exclude the possibility that there may exist anatomic pathways, poorly sensitive to nickel, that directly connect the extracellular space to at least some of the intracellular calcium stores. In support of this possibility, the peripheral SR of bovine airway smooth muscle is closely opposed to the plasma membrane, especially in regions of cavioli (6). However, direct connections between the extracellular space and the SR in any cell type have not been demonstrated morphologically.

The mechanisms regulating refilling pathways are not known. However, studies with platelets (30, 31, 34), fibroblasts (20), and colonic smooth muscle (10, 13) have suggested that depletion of intracellular calcium stores induces calcium influx across the plasma membrane via a PTK-dependent mechanism. Moreover, inhibitors of PTKs have been shown to inhibit contractions of isolated bronchioles of the rat (9). In the present study, genistein and ST-638 did partially inhibit recovery of S2/S1 during 80 s of refilling. The effect of genistein was concentration dependent, and daidzein, used as a negative control for genistein, had no effect on the recovery of S2/S1. Combined with studies (13, 20, 29) of other cells, these findings suggest that PTKs participate in the regulation of refilling caffeine-sensitive stores in airway smooth muscle, possibly by stimulating calcium influx when the stores are depleted. Because genistein inhibited recovery of S2/S1 in the presence of 2 mM nickel, our findings further suggest that tyrosine kinases regulate or modulate refilling pathways that are poorly sensitive to nickel. Because genistein and ST-638 only partially inhibited refilling, additional PTK-independent mechanisms may also contribute to the regulation of refilling.

For many cell types, PTKs have been implicated in regulating the refilling of intracellular stores primarily on the basis of the effects of putative PTK inhibitors. It remains possible that these inhibitors act by having effects on kinases other than PTK. In the present study, however, two structurally different PTK inhibitors (1, 9) had similar inhibitory effects on refilling, and daidzein, a negative control for genistein, had no effect on refilling. In other experiments, inhibitors of PKC, an inhibitor of PKA, and the putatively selective MEK inhibitor PD-98059 were all ineffective at inhibiting SR refilling. For these experiments, concentrations of these agents were chosen on the basis of previous reports (2, 14, 22) showing inhibition of PKC, PKA, or MEK. That these other kinase inhibitors were ineffective suggests that genistein and ST-638 did not inhibit rapid refilling by nonspecifically inhibiting PKC and PKA in our experiments. That PD-98059 had no inhibitory effects on refilling further suggests that MEK pathways are not the sites of the tyrosine phosphorylations important for rapid refilling. Nonetheless, nonspecific effects of genistein and ST-638 are still possible, especially because, in platelets, one study (35) showed a poor correlation between the effects that PTK inhibitors had on protein tyrosine phosphorylations versus refilling of intracellular stores.

In summary, for single airway smooth muscle cells, a concentration of nickel that effectively inhibited calcium influx to the inner cytosolic compartment poorly inhibited refilling of caffeine-sensitive calcium stores. We concluded that the inner cytosol is not the only calcium compartment from which the caffeine-sensitive stores refill. The results suggest that there is a functionally or anatomically privileged calcium influx pathway for the refilling of caffeine-sensitive stores. This pathway is relatively insensitive to inhibition by nickel and may be regulated or modulated by PTK.