Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Leukotrienes are potent biologically active compounds known to play a pivotal role in inflammation and other cell functions. Because leukotrienes are synthesized by 5-lipoxygenase in many cells, pharmacological inhibitors of 5-lipoxegenase prove to be useful in the study of leukotriene metabolism. 3-[1-(p-Chlorobenzyl)-5-(isopropyl)-3-tert-butylthioindol-2-yl]-2, 2-dimethylpropanoic acid (MK-886) is one of the several compounds that have been used to block leukotriene biosynthesis both in vivo and in vitro. MK-886 is thought to act by inhibiting activation of 5-lipoxygenase enzyme by a protein termed 5-lipoxygenase activating protein (Rouzer et al., 1990). In vivo, it was shown that oral application of MK-886 in humans significantly inhibits leukotriene B4 biosynthesis (Depre et al., 1993) and blocks allergen-induced airway responses (Friedman et al., 1993). MK-886 inhibits leukotriene biosynthesis and antigen-induced bronchoconstriction in animal models and in asthmatic men (Young, 1991). In vitro, MK-886 exerts many effects, including preventing the translocation and activation of 5-lipoxygenase in human keratinocytes (Hegemann et al., 1995) and leukocytes (Rouzer et al., 1990), inhibiting voltage-gated K⁺ currents and activating Ca²⁺-activated K⁺ currents in rat arterial myocytes (Smirnov et al., 1998), and inhibiting DNA synthesis in leukemia cells (Khan et al., 1993). Moreover, MK-886 was found to be a potent and specific inhibitor of both leukotriene B4 and leukotriene C4 synthesis in human phagocytes (Gillard et al., 1989; Menard et al., 1990) and to induce apoptosis and exhibit antiproliferative effects in HL-60 cells (Dittmann et al., 1998).

The effect of MK-886 on Ca²⁺ homeostasis is unexplored. A transient increase in [Ca²⁺]_i is a crucial signal for numerous cell events (Clapman, 1995; Berridge, 1997). A [Ca²⁺]_i increase may occur on external stimulation as a result of Ca²⁺ entry and/or Ca²⁺ release. In nonexcitable cells that lack voltage-gated Ca²⁺ channels, one of the principle Ca²⁺ stores for the [Ca²⁺]_i increase is the inositol 1,4,5-trisphosphate (IP₃)-sensitive Ca²⁺ store (Berridge, 1993). Binding of IP₃ to its receptors on the internal stores causes active release of internal Ca²⁺. This discharge of the internal Ca²⁺ store often triggers Ca²⁺ influx, leading to a prolonged [Ca²⁺]_i increase and refilling these stores. This Ca²⁺ influx is termed "capacitative Ca²⁺ entry" (Putney and Bird, 1993).

Here we have investigated the effect of MK-886 on Ca²⁺ signaling in Madin-Darby canine kidney (MDCK) cells. We have previously shown that in this epithelial cell, IP₃-dependent agonists such as ATP (Jan et al., 1998a) and bradykinin (Jan et al., 1998b) increase [Ca²⁺]_i by releasing Ca²⁺ from the endoplasmic reticulum (ER) Ca²⁺ store, followed by capacitative Ca²⁺ entry. Additionally, thapsigargin (Jan et al., 1999a) and 2,5-di-tert-butylhydroquinone (Jan et al., 1999b) increase [Ca²⁺]_i by inhibiting the ER Ca²⁺ pump without increasing IP₃ levels, leading to Ca²⁺ release and capacitative Ca²⁺ entry. Thus, MDCK cells were chosen as a model to examine drug effects on Ca²⁺ homeostasis in nonexcitable cells.

Using fura-2 as a Ca²⁺ probe, we have found that MK-886 concentration dependently increased [Ca²⁺]_i in MDCK cells. We have established the concentration-response relationships both in the presence and absence of external Ca²⁺ and have explored the possible mechanisms underlying MK-886-induced [Ca²⁺]_i signals.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture. MDCK cells obtained from American Type Culture Collection (CRL-6253; Manassas, VA) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal bovine serum, 100 U/ml penicillin, and 100 µg/ml streptomycin at 37°C in 5% CO₂-containing humidified air.

Solutions. Ca²⁺ medium (pH 7.4) contained 140 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 10 mM HEPES, and 5 mM glucose. Ca²⁺-free medium contained no Ca²⁺ plus 1 mM EGTA (calculated [Ca²⁺] < 0.1 nM). The experimental solution contained <1% of solvent (ethanol), which did not affect [Ca²⁺]_i (n = 3).

Optical Measurements of [Ca²⁺]_i. Trypsinized cells (10⁶/ml) were loaded with 2 µM 1-[2-(5-carboxyoxazol-2-yl)-6-aminobenzofuran-5-oxy]-2-(2'-amino-5'-methylphenoxy)-ethane-N,N,N,N-tetraacetic acid pentaacetoxymethyl ester (fura-2/AM) for 30 min at 25°C in DMEM. Cells were washed and resuspended in Ca²⁺ medium before use. Fura-2 fluorescence measurements were performed in a water-jacketed cuvette (25°C) with continuous stirring; the cuvette contained 1 ml of medium and 0.5 million cells. Fluorescence was monitored with a Shimadzu RF-5301PC spectrofluorophotometer (Tokyo, Japan) by continuously recording excitation signals at 340 and 380 nm and emission signal at 510 nm at 1-s intervals. Maximum and minimum fluorescence values were obtained by adding 0.1% Triton X-100 and 20 mM EGTA sequentially at the end of an experiment. [Ca²⁺]_i was calculated as previously described (Grynkiewicz et al., 1985). Our studies showed that trypsinized cells prepared by this protocol respond to stimulation with ATP (Jan et al., 1998a), bradykinin (Jan et al., 1998b) or thapsigargin (Jan et al., 1999a) similarly to cells attached to coverslips.

Chemical Reagents. The reagents for cell culture were from Life Technologies (Grand Island, NY). Fura-2/AM was from Molecular Probes (Eugene, OR). MK-886, 1-(6-((17-3-methoxyestra-1,3,5(10)-trien-17-yl)amino)hexyl)-1H-pyrrole-2,5-dione (U73122), 1-(6-((17-3-methoxyestra-1,3,5(10)-trien-17-yl)amino)hexyl)-2,5-pyrrolidine-dione (U73343), and aristolochic acid were from BIOMOL (Plymouth Meeting, PA). The other reagents were from Sigma (St. Louis, MO).

Statistical Analysis. All values are reported as means ± S.E. of five or six experiments. Statistical comparisons were determined by using Student's paired t test, and significance was accepted when P < .05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

MK-886 Induces [Ca²⁺]_i Increases in MDCK Cells. At concentrations between 0.5 and 25 µM, MK-886 increased [Ca²⁺]_i in Ca²⁺ medium (Fig. 1A). At a concentration of 0.1 µM, MK-886 had no effect (data not shown). Over a time period of 5 min, the [Ca²⁺]_i increase consisted of an immediate initial rise and a slowly decaying phase. At a concentration of 10 µM, MK-886 induced a [Ca²⁺]_i increase that reached a maximum height 180 s later at a net value of 502 ± 12 nM (Fig. 1A, trace b; n = 6; P < .05), followed by an elevated phase that had a net height of 351 ± 10 nM at the 350-s time point. The rise of the Ca²⁺ signal was slower in response to lower concentrations of MK-886. MK-886 at concentrations 50 µM caused a persistent increase in the fura-2 ratio signal, most likely reflecting cell membrane leakage; thus, these results were not reported.

View larger version (29K): [in this window] [in a new window]   Fig. 1.   A, concentration-dependent effects of MK-886 on [Ca2+]i in fura-2-loaded MDCK cells. The concentration of MK-886 was 25 µM in trace a, 10 µM in trace b, 5 µM in trace c, 1 µM in trace d, and 0.5 µM in trace e. MK-886 was added at 30 s. The experiments were performed in Ca2+-containing medium. B, similar to A except that the experiments were performed in Ca2+-free medium (no added Ca2+ plus 1 mM EGTA). C, MK-886 (10 µM)-induced changes in the 340- and 380-nm excitation wavelength signals (emission wavelength was 510 nm). MK-886 was added at 30 s. Solid trace, 340-nm signal. Dashed trace, 380-nm signal. D, concentration-response plots of MK-886-induced [Ca2+]i increases in the presence (filled circles) and absence (open circles) of external Ca2+. The y-axis is the net area under the curve (30-350 s) of the [Ca2+]i increase. The data are means ± S. E. of five to six experiments. *P < .05 between filled circles and open circles. Traces are typical of five to six experiments.

Sources of MK-886-Induced [Ca²⁺]_i Increases. Figure 1B shows that external Ca²⁺ removal decreased the Ca²⁺ signals induced by 1 to 25 µM MK-886, both in the maximum height and the area under the curve (30-350 s). The MK-886-induced increase in the fura-2 ratio signal was not a Ca²⁺-insensitive artifact because, as shown in Fig. 1C, MK-886 (10 µM) induced an increase in the 340-nm excitation signal accompanied by a corresponding decrease in the 380-nm excitation signal. The concentration-response relationships of MK-886-induced [Ca²⁺]_i increase both in the presence and absence of external Ca²⁺ are illustrated in Fig. 1D. The y-axis represents the net area under the curve of the [Ca²⁺]_i increase.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   A, in Ca2+-free medium, CCCP (2 µM), thapsigargin (1 µM), cyclopiazonic acid (CPA; 100 µM), and MK-886 (10 µM) were added at 30, 400, 800, and 900 s, respectively. B, in Ca2+-free medium, MK-886 (10 µM), thapsigargin (1 µM), and CCCP (2 µM) were added at 30, 650, and 790 s, respectively. Traces are typical of five to six experiments.

Effects of MK-886 on Capacitative Ca²⁺ Entry. Because it was reported that depletion of internal Ca²⁺ stores often triggers capacitative Ca²⁺ entry in MDCK cells (Jan et al., 1998a,b,c, 1999a,b,c,d), experiments were performed to examine whether MK-886-induced Ca²⁺ influx was via capacitative Ca²⁺ entry. Capacitative Ca²⁺ entry was measured by adding 3 mM Ca²⁺ to cells pretreated with MK-886 in Ca²⁺-free medium. Figure 3A shows that after depleting the internal Ca²⁺ store for 880 s with 10 µM MK-886, Ca²⁺ induced an [Ca²⁺]_i increase with a net maximum height of 487 ± 15 nM (trace a), which was higher than control (39 ± 5 nM; trace b) by 12-fold (n = 6; P < .05).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Effects of MK-886 on capacitative Ca2+ entry. Capacitative Ca2+ entry was induced by depleting internal Ca2+ stores in Ca2+-free medium by MK-886 preincubation, followed by addition of 3 mM CaCl2. A, trace a, MK-886 (10 µM) was added at 30 s, followed by CaCl2 at 900 s. Traces b and c, after MK-886 preincubation for 700 s, SKF96365 (50 µM; trace b) and econazole (25 µM; trace c) were added, respectively, before CaCl2. Trace d, CaCl2 was added without preincubation with MK-886, SKF96365, or econazole. B (similar to A), after MK-886 pretreatment for 700 s propranolol (0.1 mM; trace a) or aristolochic acid (40 µM; trace c) was added, followed by CaCl2. Traces are typical of five to six experiments.

Mechanism of MK-886-Induced Internal Ca²⁺ Release. The pathway by which MK-886 releases Ca²⁺ was investigated by examining the effect of inhibiting phospholipase C-dependent IP₃ formation. We have previously shown that ATP (10 µM) induces significant Ca²⁺ release in an IP₃-dependent manner (Jan et al., 1998c). Shown in Fig. 4A, trace a, is a typical [Ca²⁺]_i increase induced by 10 µM ATP. Incubation with U73122 (1 µM), a phospholipase C inhibitor (Thompson et al., 1991), for 220 s abolished the [Ca²⁺]_i increase induced by ATP (10 µM) (Fig. 4A, trace c; n = 6; P < .05). This implies that U73122 pretreatment effectively blocked phospholipase C-dependent IP₃ production. After U73122 pretreatment for 330 s, application of MK-886 (10 µM) induced a [Ca²⁺]_i increase with a net maximum height of 140 ± 7 nM, which is 82 ± 5% (n = 6; P < .05) of control (Fig. 4, trace b). U73343, an inactive U73122 analog, neither altered the resting [Ca²⁺]_i nor the [Ca²⁺]_i increases induced by ATP and MK-886 (data not shown). The effect of aristolochic acid and propranolol on MK-886-induced internal Ca²⁺ release was examined. Figure 4B shows that pretreatment with aristolochic acid (40 µM) for 300 s inhibited 10 µM MK-886-induced [Ca²⁺]_i increase by 50 ± 6% in net peak height (trace c versus trace a; n = 6; P < .05). Aristolochic acid did not alter the resting [Ca²⁺]_i. Likewise, pretreatment with propranolol (0.1 mM) for 300 s markedly reduced 10 µM MK-886-induced [Ca²⁺]_i peak by 51 ± 7% (Fig. 4B, trace b versus trace a; n = 6; P < .05) without significantly increasing the resting [Ca²⁺]_i.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   A, trace a, ATP (10 µM) was added at 30 s. Trace b, control effect of MK-886 (10 µM; added at 350 s). Trace c, U73122 (2 µM) was added at 30 s, followed by ATP (10 µM) at 260 s and MK-886 (10 µM) at 350 s, respectively. B, trace a, control effect of MK-886 (10 µM) added at 330 s. Trace b, propranolol (0.1 mM) was added at 30 s, followed by MK-886 (10 µM) at 330 s. Trace c, aristolochic acid (40 µM) was added at 30 s, followed by MK-886 (10 µM) at 330 s. The experiments in A and B were both performed in Ca2+-free medium. Traces are typical of five to six experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This report is the first to demonstrate that MK-886, widely used as a 5-lipoxygenase inhibitor, induced a significant [Ca²⁺]_i increase in a nonexcitable epithelial cell at concentrations commonly used to inhibit lipoxygenases. It is rather unlikely that the MK-886-induced increase in [Ca²⁺]_i resulted from its inhibition of arachidonic acid metabolism because the other lipoxygenase inhibitors tested, such as baicalein (50 µM), 5,8,11,14-eicosatetraynoic acid (ETYA; 100-200 µM), caffeic acid (5-50 µM), esculetin (5-50 µM), and L-655238 (80-100 µM) did not alter the resting [Ca²⁺]_i (data not shown). It is also unlikely that MK-886 increases [Ca²⁺]_i by causing plasma membrane leakage because first, trypan blue assay performed 10 min after cells were exposed to 25 µM MK-886 revealed no increased cell death than that in control; and second, Fig. 1A shows that the [Ca²⁺]_i signals induced by 25 µM MK-886 reached a peak 3 min after drug application and started to decline afterward. If the increase in fura-2 ratio signal was due to Ca²⁺ influx through damaged plasma membrane, the fluorescence signal would increase persistently.

MK-886 triggers both Ca²⁺ influx and Ca²⁺ release at concentrations of 0.5 to 25 µM because the Ca²⁺ signals were partly decreased by Ca²⁺ removal. The rise and decay phases were both reduced by Ca²⁺ removal, suggesting that the [Ca²⁺]_i increase involves Ca²⁺ influx throughout the whole course of measurement. Another line of evidence that MK-886 induces Ca²⁺ influx comes from Fig. 3 that illustrates that MK-886 (10 µM) induced capacitative Ca²⁺ entry.

The internal Ca²⁺ sources for MK-886-induced [Ca²⁺]_i increase consist of thapsigargin-sensitive ER stores, CCCP-sensitive mitochondrial stores, and other unidentified stores. This is because in Ca²⁺-free medium, pretreatment with 10 µM MK-886 prevented 1 µM thapsigargin and 2 µM CCCP from releasing more Ca²⁺; and conversely, after pretreating with CCCP and thapsigargin, MK-886 still released a significant amount of Ca²⁺. This is interesting because all the other Ca²⁺-mobilizing substances we have tested so far in MDCK cells, such as ATP, bradykinin, U73122, cyclopiazonic acid, 2,5-di-tert-butylhydroquinone, econazole, and SKF96365, release Ca²⁺ solely from thapsigargin-sensitive stores (Jan et al., 1998a,b,c, 1999a,b,c,d). Another possible candidate of internal Ca²⁺ stores is the ryanodine-sensitive store. However, it was previously shown that MDCK cells probably do not possess functional ryanodine receptors because neither ryanodine (1-50 µM) nor caffeine (10-20 mM) increases the resting [Ca²⁺]_i (Jan et al., 1998b). The unidentified Ca²⁺ stores were not further investigated due to the lack of selective pharmacological tools.

We have examined whether the Ca²⁺ release induced by MK-886 was mediated by a rise in cytosolic IP₃ levels by using U73122, a phospholipase C inhibitor, to suppress IP₃ formation. U73122 pretreatment resulted in a slight depletion of Ca²⁺ stores but completely blocked IP₃ formation because ATP (10 µM) added subsequently did not increase [Ca²⁺]_i. The lack of effect of ATP on Ca²⁺ release could not be due to U73122-induced partial depletion of Ca²⁺ stores because MK-886 added afterward still induced a [Ca²⁺]_i increase with a peak height only 15% smaller than that of control. The smaller [Ca²⁺]_i increase that MK-886 induced after U73122 pretreatment was most likely due to U73122-induced partial depletion of Ca²⁺ stores. Thus, it seems unlikely that IP₃ has a dominant role in mediating MK-886-induced Ca²⁺ release.

It was shown in MDCK cells (Kennedy et al., 1997) that Ca²⁺-mobilizing agents, such as bradykinin, can initiate a complex signaling cascade that includes early activation of upstream enzymes, including phospholipase C and phospholipase D, and phospholipase A₂-dependent release of arachidonic acid. Phospholipase A₂ is thought to move from cytosol to plasma membranes after an increase in [Ca²⁺]_i. Interestingly, we found that MK-886-induced Ca²⁺ entry was inhibited by 50% by inhibiting phospholipase A₂ with aristolochic acid but was not affected by suppressing phospholipase D with propranolol. This suggests that phospholipase A₂-coupled events, such as arachidonic acid synthesis, may have a positive feedback action both on MK-886-induced Ca²⁺ release and Ca²⁺ influx. In contrast, phospholipase D-associated events might be significantly involved in the modulation of MK-886-induced Ca²⁺ release but not Ca²⁺ influx. However, phospholipase D and phospholipase A₂ may play a significant role in regulating MK-886-induced Ca²⁺ release because inhibition of phospholipase D and phospholipase A₂ with propranolol (0.1 mM) and aristolochic acid (40 µM) for 300 s reduced the MK-886 response by as much as 50% in peak height without significantly depleting Ca²⁺ stores.

We found that 10 µM MK-886 activates capacitative Ca²⁺ entry. This is consistent with the data shown in Fig. 1 that MK-886 activated significant Ca²⁺ influx. SKF96365 and econazole partly suppressed this Ca²⁺ entry, consistent with our previous results that these two drugs exerted partial inhibition of the capacitative Ca²⁺ entry induced by thapsigargin, cyclopiazonic acid, and UTP (Jan et al., 1999c,d).

Because another lipoxygenase inhibitor, nordihydroguaiaretic acid, has been shown to activate Ca²⁺-dependent K⁺ channels in other cells (Hatton and Peers, 1996; Nagano et al., 1996), it is possible that the Ca²⁺ entry triggered by MK-886 in MDCK cells was caused by an increased driving force for Ca²⁺ influx that resulted from MK-886-induced membrane hyperpolarization by activating Ca²⁺-dependent K⁺ channels. We examined this possibility by investigating the effect of valinomycin, a K⁺ ionophore, on [Ca²⁺]_i. Valinomycin was expected to hyperpolarize the cell by increasing K⁺ efflux. Our data suggest that during the 5 min of incubation with 10 to 100 µM valinomycin, the resting [Ca²⁺]_i did not significantly increase (data not shown). Likewise, pretreatment with 10 to 20 mM tetraethylammonium and 10 µM charybdotoxin to inhibit K⁺ currents did not alter MK-886-induced [Ca²⁺]_i increase. Thus, it appears that MK-886-induced [Ca²⁺]_i increase is dissociated from its effects on membrane potential.

Figure 3A shows that in Ca²⁺-free medium, the [Ca²⁺]_i increase induced by 10 µM MK-886 remained elevated above prestimulatory baseline by approximately 50 nM, 700 s after drug addition. In contrast, it was shown that the [Ca²⁺]_i increases induced by other ligands, such as ATP, bradykinin, thapsigargin, and 2,5-di-tert-butylhydroquinone (Jan et al., 1998a,b, 1999a,b), returned to baseline in less than 400 s after drug addition. One possible explanation is that MK-886 inhibits the mechanism underlying the efflux or sequestration of the mobilized Ca²⁺. We have found similar phenomena in econazole- and SKF96365-induced Ca²⁺ release (Jan et al., 1999c,d).

Collectively, we have characterized the [Ca²⁺]_i increase induced by MK-886 in MDCK cells and have attempted to examine the possible underlying mechanisms. We have found several important effects of MK-886: 1) concentration dependently increasing [Ca²⁺]_i at ranges commonly used to inhibit 5-lipoxygenase. [For example, in a study performed in myocytes, MK-886 was used at a concentration of 10 µM to demonstrate that this drug inhibited voltage-gated K⁺ currents while it activated Ca²⁺-activated K⁺ currents (Smirnov et al., 1998)]; 2) activating Ca²⁺ influx and Ca²⁺ release; 3) releasing Ca²⁺ from thapsigargin-sensitive ER stores, CCCP-sensitive mitochondrial stores, and other stores; 4) triggering capacitative Ca²⁺ entry, which was inhibited by SKF96365, econazole, and aristolochic acid; and 5) mobilizing internal Ca²⁺ in an IP₃-independent, phospholipase D-, and phospholipase A₂-dependent manner. Given the fact that MK-886 increases [Ca²⁺]_i in MDCK cells at concentrations commonly used by investigators to inhibit lipoxygenases in most cell types, we caution the use of this drug as a specific lipoxygenase inhibitor, especially in situations that increases in [Ca²⁺]_i caused by Ca²⁺ influx and/or Ca²⁺ release may affect the results.