INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

PULMONARY SURFACTANT, a mixture of phospholipids and lipoproteins, is secreted by alveolar type II (ATII) cells via the exocytotic release of lamellar body (LB) contents into the alveolar space. Consequently, fusion of LBs with the plasma membrane is the key step linking cellular synthesis of surfactant with its dispersion over the air-liquid interface where it lowers surface tension and increases pulmonary compliance. The regulation of secretion in ATII cells is complex, involving hormones as well as local chemical and physical factors (for reviews, see Refs. 4, 29, 32). How this multitude of different stimuli converge into a single physiological response is still incompletely understood.

Haller et al. (12) and Mair et al. (20) have recently developed a method to resolve single LB fusion events and to analyze the onset, time course, and extent of the cellular responses. This method is based on FM 1-43, a fluorescent marker of surfactant, that can be used in combination with fura 2, allowing simultaneous recordings of the intracellular Ca²⁺ concentration ([Ca²⁺]_i) and exocytosis of LBs. An earlier study by Haller et al. (12) indicated that in the majority of cells stimulated with ATP plus isoproterenol, the onset of exocytosis coincides with intracellular Ca²⁺ release. Furthermore, Mair et al. (20) have recently shown that micromolar Ca²⁺ concentrations promote exocytosis when applied via the patch pipette in the whole cell patch-clamp configuration. This concurs with most investigators' opinions that elevations in Ca²⁺ are associated with increased secretion and support the findings of Dobbs et al. (7) and Pian et al. (26) that Ca²⁺ may trigger secretion, even though a decisive Ca²⁺ dependence of exocytosis could not be inferred from these studies. In addition, quantitative aspects, in particular, relationships between the level of [Ca²⁺]_i and exocytosis, have not yet been investigated in these cells.

To specifically address these issues, we used three approaches to modulate [Ca²⁺]_i and to study the early effects (within minutes after stimulation) on LB exocytosis in single cells, 1) [Ca²⁺]_i was selectively elevated by flash photolysis of a cell-permeant caged Ca²⁺ compound [o-nitrophenyl EGTA-AM (NP-EGTA-AM)], 2) [Ca²⁺]_i was gradually increased in continuously permeabilized cells, and 3) the amplitude of the ATP-induced Ca²⁺ signal was attenuated by the Ca²⁺-chelating effect of NP-EGTA. The present study is the first demonstration of simultaneous measurements of [Ca²⁺]_i and exocytosis in single ATII cells combined with selective [Ca²⁺]_i elevation by flash photolysis. Our findings suggest a prominent role of Ca²⁺ in stimulus-secretion coupling and an exocytotic mechanism with a high affinity for this messenger.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell preparation. ATII cells were isolated from Sprague-Dawley rats according to the procedure of Dobbs et al. (6) with the slight modifications summarized in Ref. 12. After isolation, the cells were seeded on glass coverslips at low density (40 cells/mm²), cultured in DMEM, and used 1 day thereafter.

Optical setup and ultraviolet flash. Excitation light was generated by a polychromatic illumination system (TILL Photonics) as previously described (21). Excitation bands centered at 340 or 380 (for fura 2) and 475 nm (for FM 1-43), each of 10 ms duration, were consecutively applied after dark intervals of 1 s (sample interval of 1 Hz). Excitation light was directed through a quartz glass fiber, attenuated by a gray filter to 10% transmission, and fed laterally into a two-port condenser unit (TILL Photonics) attached to the epifluorescence port of an inverted microscope (Axiovert 35, Zeiss). The source of the ultraviolet (UV)-flash light was a pulsed xenon arc lamp (pulse length 0.5 ms, wavelength ~320-390 nm) coupled by a quartz-light guide into the rear port of the same condenser unit. Both beams were centered at a sapphire window by which they were made parallel to each other and to the epifluorescence path of the microscope. Alignment of the sapphire allowed transmission of 92% of the UV flash and reflection of 8% of the excitation light intensity into an achromatic UV condenser. Incident light was further deflected by a 500-nm dichroic mirror into a Zeiss Fluar ×40 1.3-numerical aperture oil objective. With this optical configuration, the discharged flash energy of 80 J yields ~2 × 10²³ photons/m² in the specimen plane. The intensity of the excitation light, however, was below the level at which Ca²⁺ release due to photolysis of NP-EGTA could be observed. The emitted fluorescence was directed through a 520-nm long-pass filter and a Viewfinder (TILL Photonics) to a photomultiplier tube (Hamamatsu). The Viewfinder contained adjustable field stops by which the collected fluorescence could be restricted to the area of one single cell. Correspondingly, excitation and flash light were aligned by separate field stops to the same area in the specimen plane (~20 µm). Cellular fluorescence was simultaneously recorded with a charge-coupled device video camera (Variocam, PCO Computer Optics). Control of excitation light and data acquisition were performed with the fura extension of the Pulse software (Heka). Data were analyzed with DatGraf (Cyclobios) and SigmaPlot (Jandel Scientific). Data are reported as arithmetic means ± SE.

Fluorescence measurements. Cells were preincubated at 37°C in DMEM with fura 2-AM (1.2 µmol/l, 15 min) and NP-EGTA-AM (1-10 µmol/l, 5-30 min; both from Molecular Probes). Coverslips with stained cells were mounted into a perfusion chamber placed on the stage of the microscope and rinsed at 25°C with bath solution A (in mmol/l: 140 NaCl, 5 KCl, 1 MgCl₂, 2 CaCl₂, 5 glucose, and 10 HEPES, pH 7.4). Exocytosis was recorded in the continuous presence of 0.5 µmol/l of FM 1-43 (Molecular Probes) in the bath (12). Although interference between the FM 1-43 fluorescence (F_FM1-43) and fura 2 fluorescence (F_fura2) intensity ratios is generally small, the fura 2 signal was corrected off-line for the progressive increase in F_FM1-43 (see [Ca²⁺]_i calculation).

As an alternative approach to flash photolysis, we aimed at gradually raising [Ca²⁺]_i from resting levels up to micromolar concentrations. We found that this is most effectively achieved by exposing ATII cells to a linear perfusion gradient ranging from 100% of control conditions (solution A; see above) to 100% of a "maximum Ca²⁺" solution B (65-125 mmol/l of NaCl, 5-25 mmol/l of CaCl₂, and 10-50 nmol/l of ionomycin; other constituents were the same as in solution A). FM 1-43 was permanently present at 0.5 µmol/l in solutions A and B. Both solutions were mixed by means of a self-made gradient pump and supplied to the cells. Gradient perfusion typically lasted 10 min.

Dose-response relationships between ATP and LB exocytosis were determined with a video-imaging system as previously described (12). In each experiment, the mean cellular F_FM1-43 of up to 50 cells was measured before and 10 min after stimulation with ATP and corrected for the respective baseline value.

[Ca²⁺]_i calculation. [Ca²⁺]_i was calculated as previously described (11, 13). At 380-nm excitation, FM 1-43 contributes, to a small extent, to the fura 2 emission, which leads to a slight underestimation of the actual [Ca²⁺]_i. To minimize this effect, FM 1-43 was applied at the lowest possible concentration but was still sufficient for detection (0.5 µmol/l). Furthermore, before background subtraction and ratio (R) calculation (340-to-380 nm), the F_fura2 at 380 nm was corrected by subtracting (F_FM1-43 × 0.0127). Because we found that overestimation of F_fura2 at 380 nm is proportional to F_FM1-43, we corrected the former by subtraction of this empirically determined value. A dissociation constant (K_d) of 224 nmol/l was used for the fura 2-Ca²⁺ complex (11). All other calibration factors [background fluorescence (F_b), minimum ratio (R_min), maximum ratio (R_max), and proportionality coefficients for fluorescence of free to Ca²⁺-bound dye (S_f2/S_b2)] were determined at the end of each individual experiment as previously described (13). Where calibration was not feasible or indefinite, the calibration values from other experiments on the same day were pooled and the resulting average values were applied. Determination of R_min is particularly difficult in ATII cells because even in the absence of Ca²⁺ and the presence of high concentrations of both EGTA (5 mmol/l) and ionomycin (20 µmol/l) in the bath, Ca²⁺ is only slowly extruded from the cells. The decline in the 340- to 380-nm ratio, however, could be fitted by a single exponential in most cases, and R_min was determined from extrapolation of the fitted ratio calculation. In experiments applying flash photolysis, the fura 2 calibration factors had to be corrected for the flash artifacts. In a separate set of experiments, these flash-induced changes in calibration factors were analyzed in detail and were as follows (expressed as mean percent change from the preflash values; n = 15 measurements): F_b at 340 nm, 3.6 ± 1.1%; F_b at 380 nm, 2.1 ± 1.4%; R_min, 3.0 ± 1.0%; R_max, 5.2 ± 1.3%; and S_f2/S_b2, 4.4 ± 1.1%. The flash-induced change in K_d was estimated with the following protocol: fura 2- and Ca²⁺-preloaded cells (10 µmol/l of ionomycin plus 5 mmol/l of Ca²⁺) were exposed to a solution with a measured Ca²⁺ activity of 447 nmol/l (in mmol/l: 7 NaCl, 130 KCl, 1 MgCl₂, 7.906 CaCl₂, 10 HEPES, 10 EGTA, 2 × 10⁵ ionomycin, and 0.01 nigericin, pH 7.25). As a result, [Ca²⁺]_i slowly decreased within 30-60 min. When stable [Ca²⁺]_i readings in the range of the expected K_d (224 nmol/l) were reached, the cells were exposed to the UV flash, and the change in the fura 2 ratio [R to postflash R (R_post)] was analyzed. With the assumption of identical [Ca²⁺]_i values before and after flash and with the respective pre- and postflash calibration factors, the postflash K_d-to-K_d ratio can be calculated according to Grynkiewicz et al. (11) as {1/(postflash S_f2/postflash S_b2) × [(postflash R_max  R_post)/(R_post  postflash R_min)]}/{1/(S_f2/S_b2) × [(R_max  R)/(R  R_min)]}. In our experiments, the postflash K_d increased by 10.6 ± 2.1% of the preflash value (n = 15 measurements). Based on the reported K_d of 224 nmol/l, this yields 248 nmol/l, which was subsequently used for calculating postflash [Ca²⁺]_i. The significance of applying corrected calibration values for [Ca²⁺]_i is shown in Fig. 1.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Stimulation of exocytosis by photolysis of caged Ca2+. Ultraviolet (UV) flash (arrow)-induced transient rise in intracellular Ca2+ concentration ([Ca2+]i) and simultaneous recording of FM 1-43 fluorescence intensity (FFM1-43) in a single type II cell are shown. During flash, connection to photomultiplier tube was switched off (dotted line). Importance of correcting fura 2 calibration factors for flash artifacts is shown by [Ca2+]i calculated without such corrections at 3 conditions () during measurement. arb, Arbitrary. Left inset: enlargement of [Ca2+]i peak showing single data points sampled at 1 Hz. Right inset: background-corrected fura 2 ratio image (340-to-380 nm) of a single cell acquired 850 ms after flash, demonstrating homogeneous Ca2+ elevations throughout cell.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Stimulation of LB exocytosis by photolysis of caged Ca²⁺. To selectively and rapidly increase [Ca²⁺]_i in intact ATII cells, we used the photoactivable (caged) Ca²⁺ compound NP-EGTA. Photolytic cleavage of the molecule (uncaging) is accomplished by a short UV illumination, resulting in an immediate and uniform rise in [Ca²⁺]_i (Fig. 1). Cells were loaded with the cell-permeant NP-EGTA-AM. Within the cell, the ester groups are removed by esterases and free cytosolic Ca²⁺ is bound by complexation. On flash, Ca²⁺ is rapidly released from NP-EGTA, leading to [Ca²⁺]_i spikes of up to 5 µmol/l. After the peak (which was, in fact, a short plateau in most cases), [Ca²⁺]_i declined, presumably due to sequestration into intracellular organelles and concomitant Ca²⁺ extrusion into the extracellular space (Fig. 1).

Defining the Ca²⁺ threshold for exocytosis. The amplitude of flash-induced [Ca²⁺]_i transients depends primarily on the uncaging efficiency of the actinic light and the amount of Ca²⁺ bound to the intracellular chelator. The latter can be modulated by varying the extracellular concentration of NP-EGTA-AM. Using 1-10 µmol/l, we were able to generate [Ca²⁺]_i peaks from several nanomoles per liter up to a few micromoles per liter. As evident from Fig. 2A, most cells responded when [Ca²⁺]_i peaked above ~300 nmol/l. A considerable portion of the cells could not be activated at all. Responding cells could be clearly discriminated from such nonresponders by observation of FM 1-43-stained spots after termination of the measurements. Although we presently have no explanation for these nonresponders, which are always present in our primary culture of ATII cells irrespective of the method of stimulation (12), we can rule out that they are simply dead cells: they are identical to responding cells with respect to cell morphology, calcium homeostasis, plasma membrane integrity (Haller, unpublished observations), and acidification of LBs (12).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Determination of [Ca2+]i threshold for lamellar body (LB) exocytosis. A: peak values of UV flash-induced [Ca2+]i transients vs. FFM1-43 increase (FFM1-43) determined before and 10 min after UV flash in single cells as in Fig. 1 (n = 9 preparations). Responding cells () were discriminated from nonresponding cells () by observation of FM 1-43-stained spots after termination of measurements (10 min). Inset: peak values of UV flash-induced [Ca2+]i (binning interval = exponent 0.2) from responding cells in A vs. delay between UV flash and onset of FFM1-43. * Delay is significantly higher at high compared with low [Ca2+]i, P < 0.05. B: peak values of UV flash-induced [Ca2+]i transients (data in A) were binned as in A, inset, and plotted against percentage of responders. Data were fitted according to y = a  b/(1 + cx)(1/d), where a, b, c, and d are constants. Threshold [Ca2+]i was defined at 90% of extrapolated asymptote (arrow), yielding 320 nmol/l.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   A: stimulation of exocytosis by a gradual rise in [Ca2+]i in a single cell. Cellular Ca2+ influx was slowly enhanced by cell perfusion with a continuous rise in ionomycin (0-20 nmol/l) and Ca2+ (2-20 mmol/l; gradient was started at time 0; FM 1-43 was continuously present at 0.5 µmol/l; see MATERIALS AND METHODS). [Ca2+]i (*) at initiation of exocytosis (right dotted line) was noted for histogram in B. [Ca2+]i threshold values were determined 30 s before onset of FFM1-43 (left dotted line and arrow), corresponding to peak of delay histogram in Fig. 4. B: distribution of [Ca2+]i values determined in A. Data were binned in intervals of 50 nmol/l [Ca2+]i and plotted against no. of observations (n = 10 preparations) C: data corrected for a constant exocytotic delay of 30 s in A, arrow. Both histograms were fitted by a 6th-order polynomial. Frequency peak in C yielded a [Ca2+]i threshold of 330 nmol/l.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Comparison of delay histograms in response to UV flash and ATP (10 µmol/l). Delay denotes time between first detectable onset of [Ca2+]i increase (measured by photometry) and onset of exocytosis (determined by observation with a charge-coupled device video camera). Cells were binned in intervals of 20 s and are expressed as percentage of total measurements (n = 36 for UV flash and 70 for ATP). Inset: dose-response curve for ATP (n = 15 measurements, with each n representing mean of up to 50 cells from 11 preparations). Data were fitted by a logistic dose-response function.

[Ca²⁺]_i threshold and agonist-induced exocytosis. ATP is a potent stimulus in ATII cells, acting via cAMP formation and P₂ receptor-dependent phosphoinositide hydrolysis, with subsequent Ca²⁺ mobilization (8, 27, 28). The importance of the transient rise in [Ca²⁺]_i on ATP-induced surfactant secretion, however, is still a matter of debate. Based on the above findings, we speculated that the effects of ATP on early exocytosis might be sufficiently interpreted by its Ca²⁺-mobilizing action. We tested this hypothesis by modulating ATP-induced [Ca²⁺]_i rises in NP-EGTA-loaded cells at "low" (nominally Ca²⁺-free) or "high" (1 mmol/l) extracellular Ca²⁺. By acting as a high-affinity Ca²⁺ buffer, cytosolic NP-EGTA binds Ca²⁺ and diminishes the ATP-induced [Ca²⁺]_i increase. In the low-Ca²⁺ bath, the intracellular concentration of NP-EGTA was sufficient to completely block the ATP-induced [Ca²⁺]_i signal (Fig. 5B), whereas ATP at 1 mmol/l of Ca²⁺ exceeded the buffer capacity of this chelator, resulting in a delayed and small [Ca²⁺]_i increase (Fig. 5A). Although these [Ca²⁺]_i peaks were much smaller than ATP-evoked Ca²⁺ signals without NP-EGTA (usually >500 nmol/l), they were close to the threshold concentration in every single case and were sufficient to stimulate exocytosis. Subsequent photolysis of NP-EGTA increased [Ca²⁺]_i severalfold but had no additional stimulatory effect on exocytosis. In the nominally Ca²⁺-free solution (Fig. 5B), ATP failed to create a detectable increase in [Ca²⁺]_i, and exocytosis was significantly, although not completely, inhibited. The remaining exocytotic activity could be a result of constitutive exocytosis, which always contributes ~10% to the total observed exocytosis on stimulation (12). The subsequent Ca²⁺ release by flash photolysis of NP-EGTA significantly enhanced or initiated exocytosis in all responding cells under study. The results from all three protocols (flash photolysis, gradient perfusion, and ATP stimulation) strongly support the above hypothesis and suggest that in ATII cells the Ca²⁺ threshold is the major determinant for early exocytotic activity, with other factors playing a negligible role during that phase.

View larger version (24K): [in this window] [in a new window]   Fig. 5.   Effects of ATP on [Ca2+]i and exocytosis (FFM1-43) in single alveolar type II cells in presence of cytosolic Ca2+ chelator o-nitrophenyl EGTA (NP-EGTA). Cells were preincubated for 15 min at 37°C with 2 µmol/l of fura 2-AM and 5 µmol/l of NP-EGTA-AM. Before measurement, cells were exposed for 5 min to bath solution A (see MATERIALS AND METHODS) containing 0.5 µmol/l of FM 1-43, 10 mmol/l of NP-EGTA-AM, and no added CaCl2. Flash photolysis caused a bleaching artifact of FFM1-43 (down deflection). A: 10 µmol/l of ATP and 1 mmol/l of external Ca2+ (+Ca2+) were added (straight arrow). By this treatment, [Ca2+]i increased to levels (411 ± 63 nmol/l; note that by averaging individual measurements, resulting [Ca2+]i peak is lower than individual values) still sufficient to stimulate exocytosis. Subsequent photolysis with NP-EGTA (flash arrow) increased [Ca2+]i severalfold but had no additional effect on exocytosis. B: in absence of external Ca2+ (Ca2+), stimulation by 10 µmol/l of ATP did not lead to a detectable [Ca2+]i increase, and exocytosis was negligible. However, flash induced Ca2+ release beyond threshold-initiated exocytosis. max, Maximum. Before individual FFM1-43 tracings were averaged, data were normalized between 0 (before addition of ATP) and 100% (maximum increase). Tracings are arithmetic means (thick lines) ± SE (thin lines); n = 8 measurements from 4 preparations. Nonresponding cells were omitted. During breaks, photomultiplier was switched off.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Evidence for the involvement of Ca²⁺ in LB exocytosis. Ca²⁺ has long been recognized as a modulator of surfactant secretion. This notion emerged from observations that various kinds of stimulation (e.g., ATP) increased [Ca²⁺]_i and phospholipid secretion similarly (27, 33) and that some of them require a certain basal [Ca²⁺]_i (26, 30). Furthermore, mechanical stretch promotes secretion by an apparently Ca²⁺-dependent process (33). The interpretation of these studies with regard to the role of Ca²⁺, however, is limited because the use of agonists does not allow an accurate dissection of the relative contribution of branching signaling pathways, and physical treatments such as stretch also modulate Ca²⁺ by processes not clearly defined. So far, the most compelling support for [Ca²⁺]_i in surfactant secretion comes from measurement with Ca²⁺ ionophores (7, 26, 30). In addition, Mair et al. (20) have recently shown that micromolar [Ca²⁺]_i values promote exocytosis when applied via the patch pipette in the whole cell patch-clamp configuration.

Methods to modulate [Ca²⁺]_i. The present study is a further step forward in the analysis of the physiological role of Ca²⁺ in LB exocytosis with experimental protocols that modulate [Ca²⁺]_i in a most selective way in single cells. This was achieved by flash photolysis of caged Ca²⁺, which has been used for studies of stimulus-secretion coupling in a variety of other secretory cells (14, 15, 31). With this technique, activation of Ca²⁺-independent signaling pathways are bypassed. Furthermore, it can be assumed that Ca²⁺ is uniformly elevated within the cytosol, excluding the possibility of spatially restricted Ca²⁺ gradients (31). This is a crucial point because agonist stimulation could lead to high local [Ca²⁺]_i in the vicinity of LBs despite a moderate or even low overall [Ca²⁺]_i signal in the cytosol, which would lead to a severe underestimation of the actual Ca²⁺ requirement for exocytosis. The same problem may also arise during pulsed addition of ionophores and in stretch-induced Ca²⁺ mobilizations, creating high local Ca²⁺ gradients beneath the plasma membrane or close to intracellular compartments. The issue of localized versus global Ca²⁺ effects in secretion and the problem to relate exocytosis with [Ca²⁺]_i in a quantitative manner are discussed elsewhere in great detail (23, 24). The experiments with gradient perfusions were similarly designed to yield a uniform but slow increase in [Ca²⁺]_i.

[Ca²⁺]_i threshold for LB exocytosis. Both methods effectively triggered LB exocytosis irrespective of whether [Ca²⁺]_i was elevated instantaneously (UV flash) or slowly (gradient experiments). This is in contrast to some nonexcitable cells (e.g., mast cells) where Ca²⁺ was suggested merely as a cofactor of exocytosis, probably not sufficient to trigger this process by itself (1, 9, 22). Because secretion in our cells was stimulated by photolytic Ca²⁺ increase, which is a transient event, we conclude that a sustained [Ca²⁺]_i is not a prerequisite for stimulation. Furthermore, both methods of calculating [Ca²⁺]_i thresholds revealed a consistently and unexpectedly low value. This finding is corroborated, although indirectly, by attenuating the rise in [Ca²⁺]_i below that value during cellular stimulation with ATP, which blocked exocytosis (Fig. 5B). From all these observations, strong evidence emerges that in ATII cells the Ca²⁺ threshold is a major determinant for the early exocytotic activity during the first minutes of stimulation.

Comparative aspects of Ca²⁺ in stimulus-secretion coupling. Ca²⁺ acts as a second messenger-coupling stimulus to secretion in a variety of cells (15, 25, 31). In ATII cells, LB exocytosis is triggered by a mechanism with a high affinity for Ca²⁺. Compared with synaptic vesicle fusion, where Ca²⁺ has to reach levels of hundreds of micromoles per liter, LB exocytosis is effectively initiated near 320 nmol/l. In fact, this threshold is among the lowest known for mammalian cells, comparable with human neutrophils or pancreatic -cells (3, 19). The differences in Ca²⁺ requirements to initiate exocytosis in various cell types could imply different secretory mechanisms or different regulatory controls and seem to be inversely related to vesicle size and the time course of secretion (2, 25). LBs have diameters in the micrometer range (20), and the exocytotic activity of ATII cells is among the slowest known, corresponding to the above hypothesis (1, 12). From a physiological point of view, the time needed to release the secretory products corresponds to different cellular functions. In ATII cells, the time-limiting step is probably not exocytosis itself but the subsequent slow release of the hydrophobic surfactant material and its transformation into molecular forms that are able to adsorb to the air-liquid interface in the alveoli. The high Ca²⁺ sensitivity of LB exocytosis could reflect the physiological function of ATII cells to release surfactant on a variety of different signals. In particular, this sensitivity toward Ca²⁺ could be explained by its role in responding to alveolar distension with effective secretion. As shown by Wirtz and Dobbs (33), a single stretch of ATII cells causes a transient increase in [Ca²⁺]_i followed by a sustained stimulation of surfactant secretion. Both Ca²⁺ mobilization and exocytosis exhibited a dose dependence to the magnitude of the stretch, whereas cAMP was not enhanced, supporting the concept of a sensitive, Ca²⁺-mediated mechanism. Because Lee et al. (18) found gap junction complexes between type I and type II cells and because intercellular propagation of Ca²⁺ waves with declining amplitude are described for monolayers comprising both cell types (17), the high Ca²⁺ sensitivity could be an effective mechanism to sense alveolar distension remote of the stimulus itself. Proteins that have been suggested to mediate secretion in ATII cells in a Ca²⁺-dependent fashion are annexin II and annexin VII (29). A likely candidate for a high-affinity Ca²⁺ sensor is also calmodulin, which has been reported to be reversibly associated with LBs (16). Nevertheless, the Ca²⁺ sensor(s) in ATII cells, which participates in the regulation of the final steps of LB exocytosis, and the role of soluble N-ethylmaleimide-sensitive factor attachment protein receptor proteins in this highly Ca²⁺-sensitive process remain to be determined.