Caspofungin induces the release of Ca2+ ions from internal stores by activating ryanodine receptor-dependent pathways in human tracheal epithelial cells

The antimycotic drug caspofungin is known to alter the cell function of cardiomyocytes and the cilia-bearing cells of the tracheal epithelium. The objective of this study was to investigate the homeostasis of intracellular Ca2+ concentration ([Ca2+]i) after exposure to caspofungin in isolated human tracheal epithelial cells. The [Ca2+]i was measured using the ratiometric fluoroprobe FURA-2 AM. We recorded two groups of epithelial cells with distinct responses to caspofungin exposure, which demonstrated either a rapid transient rise in [Ca2+]i or a sustained elevation of [Ca2+]i. Both patterns of Ca2+ kinetics were still observed when an influx of transmembraneous Ca2+ ions was pharmacologically inhibited. Furthermore, in extracellular buffer solutions without Ca2+ ions, caspofungin exposure still evoked this characteristic rise in [Ca2+]i. To shed light on the origin of the Ca2+ ions responsible for the elevation in [Ca2+]i we investigated the possible intracellular storage of Ca2+ ions. The depletion of mitochondrial Ca2+ stores using 25 µM 2,4-dinitrophenol (DNP) did not prevent the caspofungin-induced rise in [Ca2+]i, which was rapid and transient. However, the application of caffeine (30 mM) to discharge Ca2+ ions that were presumably stored in the endoplasmic reticulum (ER) prior to caspofungin exposure completely inhibited the caspofungin-induced changes in [Ca2+]i levels. When the ER-bound IP3 receptors were blocked by 2-APB (40 µM), we observed a delayed transient rise in [Ca2+]i as a response to the caspofungin. Inhibition of the ryanodine receptors (RyR) using 40 µM ryanodine completely prevented the caspofungin-induced elevation of [Ca2+]i. In summary, caspofungin has been shown to trigger an increase in [Ca2+]i independent from extracellular Ca2+ ions by liberating the Ca2+ ions stored in the ER, mainly via a RyR pathway.

www.nature.com/scientificreports/ endoplasmic reticulum (ER) and the mitochondria. The ER can reduce [Ca 2+ ] i via activation of the sarcoplasmic ATPase (SERCA)-carrying Ca 2+ ions into the ER. The release of Ca 2+ ions from the ER is mainly orchestrated by activating the ryanodine receptors (RyR) or IP 3 receptors to enhance [Ca 2+ ] i . There may also be alternative pathways that lead to Ca 2+ leakage from the ER; however, these are still unknown. Mitochondria can buffer Ca 2+ ions when [Ca 2+ ] i exceeds a threshold of 500 nM and they slowly release Ca 2+ into the cytosol when [Ca 2+ ] i falls below the aforementioned threshold 9 . It is still not known whether further intracellular Ca 2+ stores such as lysosomes contribute to the regulation of [Ca 2+ ] i . Different plasma membrane-bound Ca 2+ channels are known as Ca 2+ influx pathways, which include transient receptor potential (TRP) channels, store-operated calcium (SOC) channels or voltage-dependent calcium channels that can all enhance [Ca 2+ ] i . Consequently, any disturbance or rise in [Ca 2+ ] i in epithelial airway cells can be achieved by the activation of different Ca 2+ pathways or by liberating them from internal stores. Elevated [Ca 2+ ] i directly contribute to altered cell function, e.g., mucociliary clearance or mucus secretion, which can either lead to colonization of the lower airways by pathogens or can lead to improved clearance rates. When treating mycotic infections in critically ill patients, echinocandines represent the established first-line therapy of antimycotic substances including caspofungin. Caspofungin is used under clinical conditions, with a high capability to treat systemic or local Candida spp. infections. In healthy men, a mean maximum serum concentration Cmax = 9.1-11 µM was achieved after a loading dose of 70 mg caspofungin 10 .
In order to successfully treat mycotic infections in different organs, caspofungin has to reach therapeutic concentrations in many tissues or regions including the liver and the lower airways of the lungs that exceed plasma concentrations 11,12 . This distribution pattern depends on the physiology of specific organs 13 . In critically ill patients and in perfused isolated rat hearts, we have previously reported that caspofungin has severe effects on cardio-circulatory function [14][15][16] . Caused by impaired contractility of cardiomyocytes as a result of changes in [Ca 2+ ] i levels 17 . It is therefore of interest to determine whether caspofungin may also interact with the homeostasis of [Ca 2+ ] i in airway epithelial cells. [Ca 2+ ] i is part of many cellular signal cascades that are a precondition to the altering of cell functions such as the beat rate of cilia. In order to find out, we investigated whether caspofungin also changes the [Ca 2+ ] i in isolated human tracheal epithelial cells (HTEpC) and tried to elucidate the underlying regulatory mechanism.
We measured [Ca 2+ ] i using the fluoroprobe FURA-2 AM. To determine the amount of Ca 2+ ions in the ER, we used Mag-Fluo-4 AM. Pharmacological approaches were used to identify signal pathways that are involved in the regulation of [Ca 2+ ] i . The data gathered by this study should expand our knowledge of the interactions of this antifungal drug in terms of the regulation of [Ca 2+ ] i in airway epithelial cells, which can be considered to be part of the immune system.

Materials and methods
Calcium imaging in isolated epithelial cells. For calcium imaging, we used the human tracheal epithelial cell line (HTEpC, C12644, PromoCell, Heidelberg, Germany). Cells were cultivated in an Airway Epithelial Cell Growth Medium Kit (C-21160) containing the Airway Epithelial Cell Growth Medium Supplement Pack (C-39160, PromoCell, Heidelberg, Germany). The airway epithelial cells were kept in a humidified chamber at 37 °C with air containing 5% CO 2 . For the Ca 2+ imaging, cells were seeded onto laminin-coated coverslips. Dye loading with 2.5 µM FURA-2 AM (Biotium, Fremont, CA, USA) in darkness was performed in a 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer for 30 min at 37 °C. For the contents of the HEPES buffer, see the section on drugs and buffer solutions. After the loading period, the cells were rinsed in fresh HEPES buffer and were transferred to the recording chamber of an upright fluorescence microscope equipped with 20 × immersion lenses (BX50 WI, Olympus, Hamburg, Germany), which contains 2 ml HEPES. The excitation light was provided by a 50 W xenon lamp. The microscope was equipped with a dichromatic excitation longpass mirror (400 nm).
The ratiometric dye, FURA-2 AM, was excited at 340 nm and 380 nm (± 8 nm) when equipped with bandpass excitation filters. The emitted fluorescence was directed through a dichromatic shortpass filter of 560 nm to a bandpass filter of 510 nm.
imaging of luminal ca 2+ ions in the ER of tracheal epithelial cells. For calcium imaging in the ER, HTE cells were seeded on laminin-coated coverslips. To assess the intraluminal Ca 2+ concentration ([Ca 2+ ] ER ) within the ER, the cells were loaded with 5 µM of the low-affinity calcium indicator Mag-Fluo-4 AM (Kd = 22 µM, Invitrogen, Paisley, UK) in darkness at 37 °C for 60 min. Afterward, the cells were rinsed in HEPES buffer for 20 min and were then transferred into the recording chamber of the fluorescence microscope equipped with 20 × immersion lenses (BX50 WI, Olympus, Hamburg, Germany), which also contains 2 ml HEPES. Experiments were performed in both Ca 2+ -containing and Ca 2+ -free HEPES buffer (for contents see below). As single wavelength indicators are difficult to calibrate, we used normalized Mag-Fluo-4 signals (F/F o ) to express changes in the luminal [Ca 2+ ] ER . Pharmacological agents were applied with a blunt pipette into the recording chamber. The Mag-Fluo-4 was then excited at 490 nm (± 10 nm) and the emitted fluorescence was directed through a dichromatic shortpass filter of 560 nm to a bandpass filter of 510 nm.
fluorometric RoS measurements in tracheal epithelial cells under exposure to caspofungin. As  Statistical analysis. A Mann-Whitney U test was used to compare equivalent measuring points from different experiments, and a Wilcoxon rank-sum test was used to compare dependent variables. Statistical data evaluation and testing were performed using the GraphPad PRISM (Version 5.04) software (GraphPad Software, La Jolla, CA, USA). 2+ ] i was expressed as changes in the FURA-2-fluorescence ratio. For the controls, isolated tracheal epithelial cells were rested in Ca 2+ -containing or in Ca 2+ -free buffer solutions for more than 800 s. In the cells resting in Ca 2+ -containing buffer, the FURA-2-fluorescence ratio did not change significantly during this period (1.09 ± 0.08, p = 0.12, compared to initial FURA-2-fluorescence ratio). At the end of all the experiments, we applied a brief pulse of KCl (200 mM), and only the cells that responded to this vitality test by showing a significant increase in the FURA-2-fluorescence ratio were included for further statistical evaluation (Fig. 1). Cells resting in the Ca 2+ -free buffer medium also showed no changes in the FURA-2-fluorescence ratio throughout the entire exposure period (0.89 ± 0.03, p = 0.09, Wilcoxon ranksum test, Fig. 1C, D). However, these cells also responded to a brief KCl pulse with an increase in the FURA-2-fluorescence ratio.

Results changes in [ca 2+ ] i in tracheal epithelial cells. The [Ca
When the tracheal epithelial cells were exposed to caspofungin (60 µM) in Ca 2+ -containing buffer, almost all of the investigated cells immediately showed an elevation in [Ca 2+ ] i . We observed a majority of cells responding with two subsequent Ca 2+ -transients with a maximum of 2.46 ± 0.29 (FURA-2-fluorescence ratio, n = 30). The first rapid increasing Ca 2+ transient returned to the baseline, while the second one showed a rapid increase in [Ca 2+ ] i which did not return to the baseline during the remaining exposure period ( Fig. 2A). Increasing the caspofungin concentration to 120 µM resulted in prolonged elevation of [Ca 2+ ] i that remained above the baseline at the end of the exposure period. There was a caspofungin dose-dependency of maximum amplitudes of initial Ca 2+ transients (Fig. 2B). The application of NiCl 2 was used to inhibit Ca 2+ entry via the plasma-membrane-bound Ca 2+ channels. However, it did not prevent the peak rise in [Ca 2+ ] i induced by caspofungin (60 µM, 2.47 ± 0.18, p = 0.093, n = 30, Fig. 2B 2+ ] i in all of the investigated cells (n = 30) which was biphasic, and which remained above the baseline level until the end of the experimental observation period (Fig. 3A). The peak rise in [Ca 2+ ] i was significantly higher in the Ca 2+ -free buffer solutions (2.95 ± 0.72, n = 30) than the peak transients observed in the Ca 2+ -containing buffer solutions (2.27 ± 0.17, n = 30, p < 0.01; Fig. 3B). caspofungin liberates ca 2+ from internal stores. In order to determine the source of the Ca 2+ ions that increase the levels of [Ca 2+ ] i during caspofungin exposure, we pharmacologically depleted the known intracellular Ca 2+ stores, e.g., DNP (25 µM) was used in Ca 2+ -free buffer solutions to deplete the mitochondrial Ca 2+ stores. This uncoupling of the respiratory chain induces a complete depolarization of the mitochondrial membrane potential eliciting an efflux of Ca 2+ ions from these organelles. www.nature.com/scientificreports/ The Ca 2+ efflux from the mitochondria was visible by brief Ca 2+ transients (increase in FURA-2 ratio) that almost immediately returned to baseline levels (Fig. 4A). The following exposure to caspofungin (60 µM), still in the presence of DNP, evoked transient elevation in [Ca 2+ ] i that always returned to the baseline. The transient increases in Ca 2+ ions, when exposed to caspofungin in the presence of DNP, were significantly less than the transient increases in Ca 2+ ions when the mitochondrial stores were not depleted (2.41 ± 0.07, n = 11, p < 0.01, Fig. 4B).
Ca 2+ stores in the ER were identified and depleted by cyclopiazonic acid (CPA, 10 µM), a reversible inhibitor of SERCA. In a Ca 2+ -free buffer solution, CPA induced a prolonged increase in [Ca 2+ ] i that slowly returned to baseline levels (Fig. 5A). Cell vitality was verified by exposure to a depolarizing concentration of KCl (200 mM, Fig. 5A). The depletion of ER Ca 2+ -stores by caffeine (30 mM) led to a rapid but transient increase in [Ca 2+ ] i . The subsequent application of CPA (10 µM) had no further effect, demonstrating that caffeine had already depleted the ER Ca 2+ stores, which are identical to CPA-sensitive stores (Fig. 5B).
In further experiments, exposing the cells to caffeine led to a brief Ca 2+ transients of the FURA-2 fluorescence ratio that almost immediately returned to the baseline. The subsequent application of caspofungin (60 µM or 120 µM) had no effect on any tracheal epithelial cells investigated.   Fig. 6A, B). The  www.nature.com/scientificreports/ subsequent application of caspofungin still induced single Ca 2+ transients with a significantly reduced peak to 2.04 ± 0.08 (n = 30, p < 0.001) compared to Ca 2+ transients induced by caspofungin alone (2.95 ± 0.72, Fig. 6B). However, we observed no prolonged elevation in the FURA-2 fluorescence ratio, which we observed when the Ca 2+ -free buffer solution was exposed to caspofungin (Fig. 3A). In a further series of experiments, we inhibited the ryanodine receptors using 40 µM ryanodine. This concentration of ryanodine did not induce an efflux of Ca 2+ from ER stores, as demonstrated by the unaltered FURA-2 fluorescence ratio (Fig. 6C, D). The application of caspofungin in the presence of ryanodine did not evoke any increase in the FURA-2 fluorescence ratio, and no Ca 2+ transients in any of the investigated cells were observed (Fig. 6C, D). The [Ca 2+ ] i was no different to the values obtained prior to caspofungin application (1.1 ± 0.01, n = 30, p = 0.52). At the end of each experiment, Ca 2+ transients were still elicited by applying KCl, demonstrating the vitality of the cells (Fig. 6C).  (Fig. 7A), Ca 2+ free HEPES buffer (Fig. 7B), and Ca 2+ -free HEPES buffer after prior loading with BAPTA-AM (Fig. 7C). None of these conditions reduced the Mag-Fluo-4 fluorescence signal displaying [Ca 2+ ] ER , demonstrating that the ER Ca 2+ stores were stable under resting conditions without significant leakage of Ca 2+ ions into the cytosol (Fig. 7A). However, exposure to caspofungin rapidly altered the Mag-Fluo-4 fluorescence. In all the cells that were investigated, we observed a rapid decrease in fluorescence within a few seconds of exposure (Fig. 7D, E) that immediately recovered, before a slower and more prolonged decline in the fluorescence showed the [Ca 2+ ] ER . The amplitude of the initial downward spike of Mag-Fluo-4 fluorescence and the degree of the prolonged reduction was concentration-dependent (Fig. 8A, B). A final exposure to the Ca 2+ -containing buffer solution with a depolarizing KCl concentration www.nature.com/scientificreports/ demonstrated a recovery of the Mag-Fluo-4 signal equivalent to the replenishment of Ca 2+ stores within the ER (Fig. 7D). The initial sharp decrease in Mag-Fluo-4 fluorescence completely disappeared when the cells were loaded with BAPTA prior to caspofungin exposure; only the slow, prolonged phase of the declining Mag-Fluo-4 was still visible. In addition, replenishing the ER Ca 2+ stores by Ca 2+ -containing buffer and depolarizing the cells with KCl only had a limited effect, assuming that the cytosolic Ca 2+ ions were instantly buffered by BAPTA (Fig. 7F).
The total depletion of ER Ca 2+ stores was achieved by the application of caffeine (30 mM). The stores were replenished after subsequent exposure to a buffer solution containing depolarizing KCl (200 mM) and Ca 2+ (2.5 mM), inducing a maximum rise in Mag-Fluo-4 fluorescence (Fig. 9A). In the presence of caffeine, exposure to caspofungin had no further effect. Neither was the initial downward Ca 2+ spike visible nor was the prolonged reduction of Mag-Fluo-4 fluorescence further diminished (Fig. 9B, C).

ROS generation under caspofungin exposure in the tracheal epithelium.
Here, we were able to determine whether the observed effects of caspofungin on intracellular Ca 2+ stores were caused by cellular ROS generation. In freshly isolated mice tracheae, we measured ROS generation under exposure to caspofungin (120 µM) using a fluorescence dye. The results showed no significant ROS generation in the control experiment or when the tracheae were exposed to caspofungin for 30 min (Fig. 9D).  ] i that is followed by a prolonged increase in the beating rate 8 . We found no evidence that caspofungin induces ROS generation since ROS are known to interfere with calcium-signaling pathways. The stimulation of cilia-bearing cells in order to enhance their beating rate is the main task to rapidly convey heavy loads of particles and pathogens. This mechanism depends on the liberation of Ca 2+ ions from internal stores 18 . It is also in line with reports that evidenced the activation of the ciliary beat frequency following the kinetics of [Ca 2+ ] i 19 . We were also able to demonstrate that caspofungin activates intracellular signal transduction cascades that lead to the release of Ca 2+ ions from ER stores via RyR. We also demonstrated that caspofungin depleted the Ca 2+ stores in the ER.
However, although we still cannot conclude that caspofungin directly affects RyR, we have excluded any interaction with membrane-bound receptors or with their downstream reaction cascades. Therefore, with the present data, we can assume that caspofungin can penetrate tracheal epithelial cells and to activate specifically RyR. This is similar to cardiomyocytes, in which caspofungin also triggers the liberation of Ca 2+ ions from the ER via RyR 17 .

The permeability of caspofungin and its diffusion into mammalian tissues and cells. Caspo-
fungin inhibits the synthesis of 1,3-β-D-glucan, which is an essential component of the fungal cell wall 20,21 . While caspofungin underlies very low excretion kinetics that result in high sustained plasma levels, it attains therapeutic concentrations in various tissues 22 . It is also characterized by its ability to diffuse into many organs and tissues in order to reach the concentrations necessary to treat invasive candidiasis; thus, its penetration into the central nervous system (CNS) is poor 23 . Even when the meninges are inflamed, the concentrations of caspofungin in the CNS are far below serum concentrations 24 . In contrast, caspofungin concentrations are highest in the liver, lungs, kidney, and spleen, where caspofungin can deploy its antifungal properties 13 . In the lungs, the concentration of caspofungin leads to favorable response rates of up to 62% in neutropenic patients suffering from pulmonary invasive fungal disease 25 . www.nature.com/scientificreports/ In mammals, caspofungin mainly acts extracellularly against most Candida species. However, caspofungin, a membrane-anchored cyclic lipopeptide, is also able to penetrate mammalian cells and can reach high concentrations within the tracheal epithelium allowing it to act upon cilia-bearing cells. We can also conclude that caspofungin concentrations are also high in the epithelial lining fluid, and it may act from the luminal side onto epithelial cells.
Since caspofungin is used to treat pulmonary infections caused by Candida spp., it is able to diffuse via the epithelium into the airway lumen so that it can reach these therapeutic concentrations. We found evidence that caspofungin can reach intracellular concentrations in isolated lower airway cells that are high enough to liberate Ca 2+ ions from cell organelles. So far, the effects of caspofungin on intracellular organelles have only been described for a few cell types. In stimulated mammalian macrophages, caspofungin has been shown to reach concentrations that will have an effect on trapped Candida glabrata 26 . Recently, we described a change in the contractility of isolated cardiomyocytes via the liberation of Ca 2+ ions from internal Ca 2+ stores 15,17 . These data Figure 9. Caspofungin depletes caffeine-sensitive ER Ca 2+ stores. (A) In the Ca 2+ -free buffer solution, exposure to caffeine (30 mM) led to a rapid and prolonged reduction in Mag-Fluo-4-fluorescence. The ER stores were replenished by a voltage-gated Ca 2+ influx (exposure to 200 mM K + ions). (B) Caffeine exposure induces the discharge of ER Ca 2+ stores prior to caspofungin application. In addition, caspofungin did not induce a further reduction in Mag-Fluo-4-fluorescence, demonstrating that ER stores were already depleted by caffeine. (C) The addition of caffeine led to a significant reduction in the Mag-Fluo-4-fluorescence indicating that the Ca 2+ stores in the ER were empty. Subsequent application of caspofungin in the presence of caffeine did not alter the Mag-Fluo-4-fluorescence any further. (D) Isolated mice tracheae were either exposed to caspofungin (120 µM) or kept as controls in buffer solutions. Under both conditions, we observed only minor alterations in ROS generation displayed by emission fluorescence at 590 nm. (the horizontal bars in the experimental recordings depict the exposure periods of defined pharmacological agents. n = number of individual investigated cells, ns = not significant, ***p < 0.001, two-way ANOVA). www.nature.com/scientificreports/ support our present findings and lead us to postulate that the diffusion of caspofungin into mammalian cells is not restricted to particular cell types; it seems rather a general characteristic of this antimycotic substance.
The effects of caspofungin on intracellular Ca 2+ stores. In our present study, we identified that caspofungin could trigger the release of Ca 2+ ions from Ca 2+ stores in the ER, mainly via a caffeine/ryanodinesensitive pathway. We found no evidence for the liberation of Ca 2+ ions stored in mitochondria, although these organelles do store Ca 2+ ions under resting conditions (see Fig. 4). However, they primarily buffer Ca 2+ ions when [Ca 2+ ] i exceeds a threshold of approximately 500 nM 9,27 . In liver cells and cardiomyocytes, caspofungin is known to disturb the electron transport within the mitochondrial respiratory chain by inhibiting complex I and III. The effect on complex III is probably caused by its interference with cytochrome C 28 . However, it is not known whether this mechanism also depolarizes mitochondrial membrane potential, which is essential for Ca 2+ buffering in these organelles 9 . Here we measured no effect of caspofungin onto mitochondrial Ca 2+ stores. When the mitochondrial Ca 2+ stores were depleted by DNP, exposure to caspofungin had little effect (see Fig. 4). This small decrease in caspofungin response can be explained by the inhibition of ATP synthesis due to the application of DNP application 29 . ATP usually powers the sarco-endoplasmic calcium ATPase (SERCA) that refills the ER's Ca 2+ stores. Therefore, the interrupted mitochondrial ATP synthesis after exposure to DNP reduces SERCA activity, which then leads to a slow Ca 2+ leakage from the ER stores via different pathways 9,30 . Eventually, this results in a slightly reduced Ca 2+ response during subsequent caspofungin application, which was what was observed here. Our observations provide evidence that the mitochondria in tracheal epithelial cells store Ca 2+ ions under resting conditions but are not affected by caspofungin.
Under steady-state conditions, the influx of Ca 2+ ions into the cytosol from the ER stores is balanced on the one hand by RyR, IP 3 , or alternative leaks that drive the efflux from the ER, and on the other hand by SERCA, which regulates Ca 2+ influx into the ER resulting in a zero net flux of Ca 2+ ions into the cytosol 31 33 . The refilling of the ER stores cannot be achieved under these conditions, which is the reason why we observed no further Ca 2+ transients when caspofungin was applied after the ER stores were depleted by the addition of caffeine. When caspofungin is applied to cells that have not had their ER stores depleted by caffeine, we observed Ca 2+ transients in most of the cells.
In our experiments using Mag-Fluo-4 as an indicator for ER luminal Ca 2+ contents, we observed a rapid decline in Mag-Fluo-4-fluorescence followed by a partial recovery, which eventually led to a prolonged decline in Mag-Fluo-4-fluorescence. The partial recovery is most probably a response to the SERCA balancing the efflux and influx of Ca 2+ ions from the ER, as is generally assumed for many cell types 34 . A similar mechanism of Ca 2+ depletion from the ER has been described previously for smooth muscle cells derived from different organs in rats 35,36 . Here, the prolonged decline of Mag-Fluo-4-fluorescence was caused by the continuous presence of caspofungin activating RyR.
Under these conditions, SERCA was unable to balance the influx and efflux of Ca 2+ ions from the ER stores. Most probably, this net efflux is provoked by caspofungin triggering the increase in open state probability of RyR, since the kinetics of Ca 2+ transients are similar. The clearance of cytosolic Ca 2+ transients is caused by alternative clearance pathways, e.g., the transmembrane efflux, buffering, or by an uptake in alternative Ca 2+ stores like mitochondria. We interpret these findings such that caspofungin specifically activates ER-bound RyR and depletes these Ca 2+ stores. Caspofungin now provokes an imbalance in favor of a net Ca 2+ efflux that results in an immediate increase in [Ca 2+ ] i , which is either transient or sustained. The rapid amplitude in [Ca 2+ ] i under caspofungin exposure is caused by the steep function of Ca 2+ content in the ER as has previously been described in cardiomyocytes 37,38 . This Ca 2+ content contributes to the regulation of ciliary beat activity by the activation of different signal cascades 8 .
Alternatively, caspofungin may promote the dissociation of Ca 2+ ions from cytosolic buffers as an alternative source for increasing [Ca 2+ ] i . In many cell types, cytosolic Ca 2+ is strongly buffered, and for every free cytosolic Ca 2+ ion approximately 100 to 200 Ca 2+ ions are bound to buffers 39,40 . In cardiomyocytes, the major buffers are troponin and SERCA, whereas in airway epithelial cells the Ca 2+ is supposed to be buffered by calmodulin 8,34 . However, in this case, the kinetics of increased [Ca 2+ ] i would be different.
We have not observed biphasic Ca 2+ transients or sustained elevations of [Ca 2+ ] i . This is because Ca 2+ pumps on the plasma membrane, the ER and mitochondrial Ca 2+ uptake would still be fully active and would be sufficient to immediately reduce elevated [Ca 2+ ] i content.
Since the initial binding of Ca 2+ ions to SERCA contributes significantly to Ca 2+ buffering, it may be that caspofungin contributes to the dissocitaion of Ca 2+ ions from these binding sites 41 . Nevertheless, in this case, we would observe Ca 2+ transients while RyR was inhibited by ryanodine, or when the ER stores were depleted by caffeine. However, we did not observe these Ca 2+ kinetics. These data argue against the assumption that caspofungin drives Ca 2+ ions out of the buffering sites of the SERCA. Furthermore, the inhibition of SERCA by caspofungin would lead to a slow, sustained increase in [Ca 2+ ] i which then returns to the baseline, as we observed for CPA (Fig. 5) 9,30 . In contrast, we did not observe a slow increase in [Ca 2+ ] i under caspofungin exposure; the kinetic was always steep and rapid. In contrast, in many of the airway cells observed, the decay in the elevated [Ca 2+ ] i levels was slow. This prolonged kinetic may be caused either by reduced ATP content and subsequently reduced activity of ATP-driven ion pumps or by the inhibition of ion transporters. However, it is questionable whether caspofungin reduces ATP generation or inhibits ion transport and this needs to be investigated further. www.nature.com/scientificreports/ In our experiments, we noted a reduced response to caspofungin while using 2-APB, which binds to the IP 3 receptor. Since phospholipase-C activation and liberation of Ca 2+ ions from Ca 2+ stores via the IP 3 receptor contribute to Ca 2+ wave propagation in airway epithelia, we investigated whether caspofungin interferes with this signal cascade 42 . The reduced response to caspofungin exposure when 2-APB was applied may be caused by the activation of these receptors by caspofungin in competition with 2-APB. This assumption is supported by the fact that the response to caspofungin was delayed in the presence of 2-APB. Therefore, 2-APB might only be effective when caspofungin concentration is high enough to prevent 2-APB from binding. Further findings revealed that ryanodine abolished the calcium signals elicited by Caspofungin. Controversially, IP3-mediated calcium release seems not to occur in parallel to RyR-mediated calcium release. Nevertheless, we observed a reduced response to caspofungin in the presence of 2-APB. 2-APB is also known to modulate TRP channel, gap junctions, STIM ORAI channel conductance and I CRAC in higher concentrations 43,44 . Owing to this lack of specificity for the IP 3 receptor, 2-APB could also impair other calcium pathways and not particularly the activity of the IP 3 receptor. On the other hand, our experiments were performed in calcium free buffer solution, so store-operated calcium entry (SOCE) could not occur. Further experiments are necessary to explore whether the IP3 receptor pathway is affected by using a more specific IP 3 receptor blocker such as Xestospongin C or to examine SOCE pathways, which is more complicated due to diverse interactions and a finetuned orchestration of multiple targets. Regardless, we conclude that RyR is the principal target of caspofungin since we measured the main effects of caspofungin via this Ca 2+ efflux pathway.
The observed sustained elevation of Ca 2+ ions under caspofungin exposure may be due to several reasons, as a similar effect has been described when these cells are exposed to histamine or acetylcholine 8,45 . Most probably, this prolonged kinetic is caused by an imbalance between the influx and efflux of Ca 2+ ions from the cytosol due to a slow clearance rate when reaching baseline levels of [Ca 2+ ] i during the observation period. These findings imply that caspofungin has effects on other cell organelles. It may also interfere with mitochondrial ATP synthesis, as has been described for mammalian mitochondria 28 . Under these circumstances, the remaining ATP supply from mitochondria, glycolysis or phosphocreatine stores is too low to fully activate Ca 2+ ATPase in the plasma membrane. This mechanism prevents a rapid return of [Ca 2+ ] i to baseline levels. Beyond plasma membrane-bound, sodium-calcium exchangers are also activated when [Ca 2+ ] i is high and may blunt the peak of observed the Ca 2+ transients 46 .
We also found no evidence that caspofungin depolarizes mitochondrial membrane potential, which is the driving force behind Ca 2+ storage in these organelles. This is supported by our findings that the number of Ca 2+ transients was only slightly altered when mitochondria were depolarized using DNP. Therefore, in many tracheal epithelial cells, caspofungin may have additional effects on other cellular structures that cause a prolonged elevation of [Ca 2+ ] i , with the slow return to the baseline level due to an energetic restriction.
HTE cells, which are isolated cells from the human tracheal epithelium, are suitable for investigating the different effects and signal pathways under stable conditions, whereas freshly isolated human tracheal cells should only be used in further studies to underline these observations. Additionally, a contribution of alternative Ca 2+ influx pathways, e.g. store operated calcium entry (SOCE) or IP 3 receptors in the intact tracheal epithelium, should also be considered.

conclusion
Caspofungin liberates Ca 2+ ions from internal ER stores and enhances [Ca 2+ ] i either transiently or in a sustained kinetic. This mechanism likely occurs due to the activation of a RyR pathway. A sustained elevation in [Ca 2+ ] i levels may be supported via energetic imbalances during caspofungin application. Further studies should focus on RyR activation by caspofungin, detecting its binding site to these receptors, and determining how cellular ATP synthesis is compromised by caspofungin.