Cystic fibrosis (CF) is caused by mutations in the gene producing the cystic fibrosis transmembrane conductance regulator (CFTR). CFTR functions as a Cl− channel. Its dysfunction limits Cl− secretion and enhances Na+ absorption, leading to viscous mucus in the airway. Ca2+-activated Cl− channels (CaCCs) are coexpressed with CFTR in the airway surface epithelia. Increases in cytosolic Ca2+ activate the epithelial CaCCs, which provides an alternative Cl− secretory pathway in CF. We developed a screening assay and screened a library for compounds that could enhance cytoplasmic Ca2+, activate the CaCC, and increase Cl− secretion. We found that spiperone, a known antipsychotic drug, is a potent intracellular Ca2+ enhancer and demonstrated that it stimulates intracellular Ca2+, not by acting in its well-known role as an antagonist of serotonin 5-HT2 or dopamine D2 receptors, but through a protein tyrosine kinase-coupled phospholipase C-dependent pathway. Spiperone activates CaCCs, which stimulates Cl− secretion in polarized human non-CF and CF airway epithelial cell monolayers in vitro and in CFTR-knockout mice in vivo. In conclusion, we have identified spiperone as a new therapeutic platform for correction of defective Cl− secretion in CF via a pathway independent of CFTR.
- cystic fibrosis therapy
- calcium-activated chloride channel
cystic fibrosis (CF) is caused by any one of a thousand mutations that affect the CF transmembrane conductance regulator (CFTR) gene. The CFTR gene produces Cl− channels that are regulated by cAMP. These channels are expressed in many cellular epithelia, including those in the lung, pancreas, intestine, hepatobiliary tract, sweat gland, and vas deferens (37). The CFTR gene was cloned in 1989 (36).
In the airway, defective CFTR impairs Cl− secretion from epithelial cells and increases their Na+ absorption. This changes the normal ion composition and dehydrates the airway surface liquid, thus decreasing its volume. Airway mucus becomes thick and is poorly cleared from the lungs, which enables bacteria to colonize the area. Eventually, respiration fails (42).
There has been considerable work to repair CFTR's Cl− transit deficiency, but rescue has not been practical. For example, the most common CFTR mutation, ΔF508-CFTR, is amenable to rescue at 27°C (7), but that temperature is too low for the human body. Some chemicals have shown that they can rescue CFTR, and the best are under phase I or phase II clinical trials.
Several non-CFTR Cl− channels, including the family of voltage-dependent Cl− channels, volume-regulated anion channels, and, of particular interest for this study, Ca2+-activated Cl− channels (CaCCs), have been identified in airway cells (34, 42). Previous studies have shown that CaCCs are present in the apical membrane of normal and CF airway epithelia (2). Even in CF, Ca2+-activated Cl− secretion is intact and provides a therapeutic target to circumvent the Cl− secretion defect in CF (34). Denufosol and Moli-1901, which stimulate the CaCC and target this therapeutic pathway, are currently being evaluated in clinical trials. Denufosol, a chemically stable P2Y2 receptor agonist, was shown to increase Cl− and fluid secretion in preclinical studies (21). It improved lung function by 2.5% in a 24-wk course of treatment in a phase III clinical trial (38). Moli-1901 (formerly known as duramycin) is a stable polypeptide. In a phase I clinical trial, Moli-1901 stimulated Cl− transport in normal and CF epithelia (45). Four weeks of treatment with aerosolized Moli-1901 caused a 2% increase in lung function (38). Although these two agents are being studied intensively, the discovery of more compounds to activate this pathway without any adverse effects would still be beneficial for the potential treatment of CF.
High-throughput screening allows assessment of a large number of compounds quickly and has been used widely in drug discovery in recent years (32). We planned to search the 2,000 compounds of the MicroSource Discovery Spectrum (MSSP) library for compounds that would enhance cytoplasmic Ca2+, which in turn would activate the CaCCs and alleviate the defective Cl− secretion in CF airways without adverse side effects.
In our previous work, we found that zinc with ATP or zinc alone reliably causes a sustained increase in cytoplasmic Ca2+ in cells bathed in a Ringer solution modified to enhance Ca2+ transfer (ECaT Ringer solution; Table 1). Zinc and ATP activate P2X purinergic receptors, which are nonselective cation channels expressed in human airway surface epithelia (29, 46, 47). The resulting increase in cytoplasmic Ca2+ translates into sustained Cl− secretion in CF and non-CF airway epithelia in vivo and in vitro (47). Zinc, however, is a biometal, so the amount of zinc taken into the body would be a concern in CF therapy. However, it was a good positive control for our screening.
Spiperone was discovered to be the most potent Ca2+ stimulator from the screening. Spiperone stimulates Ca2+ increase and subsequent Cl− secretion in normal Ringer solution; therefore, it is potentially more useful in the body. We were able to establish that spiperone uses a protein tyrosine kinase (PYK)-coupled phospholipase C (PLC) pathway to increase intracellular Ca2+, which results in increased secretion of Cl−. It is effective in cultured cell monolayers and in vivo in mice.
Although spiperone is an antipsychotic drug and a known antagonist to receptors of serotonin 2 (5HT2) and dopamine D2 (10), we also were able to establish that its action on Cl− transport is not shared by analog antipsychotics. This finding suggests that its Cl−-stimulating property is separate from its antipsychotic pathway, thus enhancing its potential use as a CF therapy.
MATERIALS AND METHODS
Cell and Monolayer Culture
Three cell lines and one primary cell culture were used for this study. IB3-1 is a CF human bronchial epithelial cell line that is heterozygous with two different CFTR mutations (ΔF508 and W1282X). Calu-3 is a non-CF human submucosal gland serous epithelial cell line. CFBE41o− (a gift from Dr. Dieter Gruenert, University of Vermont) is a CF human bronchial epithelial cell line that is homozygous for ΔF508 mutations. NHBE is a normal human bronchial/tracheal epithelial primary cell culture (Lonza, Walkersville, MD). In general, IB3-1 cells were cultured in LHC-8 medium (Invitrogen, Carlsbad, CA) supplemented with 5% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, 2 mM l-glutamine, and 1 μg/ml amphotericin B (Fungizone). Calu-3 and CFBE41o− cells were cultured in minimum essential medium (Invitrogen) supplemented with 10% FBS, 100 U/ml penicillin, 100 μg/ml streptomycin, and 2 mM l-glutamine. NHBE cells were cultured in bronchial epithelial cell growth medium (Lonza).
Culture on permeable filter supports.
For the short-circuit current measurement, Calu-3 and CFBE41o− cells were cultured in the same medium as the cells in the flask, but on the 12-mm Snapwell permeable supports (Costar, Corning, NY). Cell resistance was measured during the culture. We usually wait until Calu-3 cell resistance reaches 1,500–20,00 Ω (5–7 days) and CFBE41o− cell resistance reaches 800–12,00Ω (3–5 days) to start the experiment.
Compound Screening for Ca2+ Enhancers
The MSSP library was our source for compounds for the primary screening. The compounds are known bioactive compounds and natural products and their derivatives. The library contains 2,000 compounds on twenty-five 96-well plates, with 80 testable compounds per plate. In the master plates, each compound is suspended in DMSO at a concentration of 10 mM. A set of 96-well plates of the library in a concentration of 2 mM in DMSO was made before the screening.
Screening was performed by the Johns Hopkins University High-Throughput Biology Center Chemcore. The 96-well transparent-bottom black plates were coated with collagen I (Sigma, St. Louis, MO) overnight. IB3-1 cells were seeded onto the collagen plates at 50,000 cells per well by a Biotek multidrop system. Cells were incubated in a 37°C, 5% CO2 incubator overnight. At the time of the screening, the cells were loaded with fura-2AM for 2 h and then washed three times with PBS. The fluorescence ratio was measured in ECaT Ringer solution by a Tecan Safire fluorescence plate reader. Then small-molecule compounds (80 compounds per plate) were added by the robotic arm to the cells at a final concentration of 20 μM with zinc as the positive control and DMSO (5%) as the negative control. The fluorescence ratio was measured again immediately after the compounds were added. The final fluorescence detection was done 15 min after the second detection to find compounds that stimulate the sustained Ca2+ increase.
Fura-2 Microscope-Based Ca2+ Measurement
Cells were seeded on the collagen-coated cover glass for 48 h. Cell Ca2+ was measured with a dual-excitation-wavelength microscope system. Fura-2 fluorescence was excited at 340 and 380 nm with xenon light. Fluorescence was measured at 510-nm-emission wavelength. Data were processed and analyzed using IPLab 4.0 software. Fluorescence data of each experiment were normalized to the individual basal fluorescence value ΔF/F0 = (F − F0)/F0 (where F0 is fluorescence at baseline) to bring all the response curves to the same pretreatment starting point.
Short-Circuit Current Measurements
Calu-3 and CFBE41o− cells were cultured on Snapwell inserts until they formed a tight monolayer as monitored by measurement of transepithelial resistance. The bottom of the monolayer was removed, mounted on the Ussing chamber, and buffered by the testing solutions. Short-circuit current was measured by a multichannel voltage/current clamp (model MC6, Physiologic Instruments, San Diego, CA). Data were recorded and analyzed by Acquire and Analyze software.
Mouse Nasal Potential Difference Measurements
The studies were approved by the Johns Hopkins University Animal Care and Use Committee. The double-transgenic CF-knockout [CFKO; CFTRtm1Unc-Tg (FABPCFTR) 1Jaw/J] mice were bred at the Johns Hopkins University Animal Core Facility (JAX no. 002364). Wild-type C57Bl/6 (JAX no. 000664) strains of female mice were obtained from JAX mice (Bar Harbor, MA).
Nasal potential difference.
All mice were anesthetized with an intraperitoneal injection of a mixture of ketamine (100 μg/g body wt) and xylazine (10 μg/g body wt). Nasal potential difference (NPD) was measured using a modification of methods described by Grubb et al. (12
Values are means ± SE. All data were analyzed by paired t-test or one-way ANOVA with Tukey's post hoc multiple comparisons using GraphPad Prism software. Significance was set at P < 0.05.
Compounds That Increase Intracellular Ca2+ Were Identified by Screening a Compound Library
Our goal was to find all the compounds that could increase intracellular Ca2+ using the ECaT Ringer solution (Table 1), which was shown to be essential for obtaining a sustained Ca2+ signal induced by zinc. Zinc, used in the screening as a positive control, produced a long-lasting increase in cytoplasmic Ca2+ by targeting the P2X purinergic receptor in the ECaT Ringer solution (46, 47). Na+ is 100 times more abundant than Ca2+ in the normal Ringer solution, and Na+ transfer will predominate over Ca2+ for entry when the P2X receptor nonselective cation channels open. Thus for the ECaT Ringer solution, we eliminated Na+ from the solution and replaced it with N-methyl-d-glucamine to maintain the solution osmolality to favor Ca2+ for entry through the P2X receptor. We were searching for compounds that activate the P2X receptor channel or other Ca2+ entry channels, as well as compounds that bind to membrane or cellular proteins that are involved in intracellular Ca2+ regulation.
Compounds would be selected as potential hits if the change of their fluorescence ratio fell within 3 SD of the change of the fluorescence ratio for zinc and was >3 SD larger than the fluorescence ratio for DMSO. The average Z factor of all 25 plates is 0.6 for the 15-min detection period. Hits were solely based on the change in the fluorescence ratio. Any compound that was highly autofluorescent was not considered further.
The screening yielded 67 compounds that met the screening criteria and were selected as potential hits. Among the potential hits, 28 compounds are not commercially available, and 2 are known to create pores in the membrane. On the basis of these criteria, these 30 compounds were eliminated from further study. We further evaluated 37 hits manually using the fluorescence plate reader system. Figure 1A illustrates the change in fluorescence ratio for three of the hits compared with zinc and 5% DMSO. During the evaluations, 18 compounds induced an increase in Ca2+ that did not reach a maximum plateau value and were also not studied further. After this initial vetting process, we determined the EC50 of each of the remaining 19 molecules. For eight of those compounds, EC50 was <50 μM (Table 2). A sample of the EC50 curves of three compounds plus zinc is shown in Fig. 1B. Spiperone showed the best EC50. In addition, spiperone has a dual effect on intracellular Ca2+: activation and inhibition. Spiperone stimulates cytoplasmic Ca2+ at lower concentrations (<50 μM) but reduced stimulation at higher concentrations (>50 μM).
Spiperone Increases Intracellular Ca2+ Originating From the Endoplasmic Reticulum
Because spiperone was the most potent of the intracellular Ca2+ enhancers (EC50 = 9.3), we decided to investigate further its mechanism of action. A microscope-based fluorescence Ca2+ measurement system was utilized in this study. First, IB3-1 cells grown on a cover glass were bathed in the ECaT Ringer solution, which we used for the screen. Challenge with spiperone produced a sustained increase in Ca2+ (Fig. 2A). To ascertain the origin of this sustained increase, we repeated the experiment, except with Ca2+-free ECaT Ringer solution (Table 1). In the absence of extracellular Ca2+, the cells showed only a transient increase in cytoplasmic Ca2+ (Fig. 2B).
We also tested spiperone's effect on the Ca2+ change in cells bathed in normal Ringer solution (which contains the usual levels of Ca2+, Na+, and Mg2+) and in cells bathed in Ca2+-free normal Ringer solution (Table 1). Spiperone again induced only a transient increase in Ca2+ in both solutions (Fig. 2, C and D).
In summary, these results indicate that spiperone stimulates a sustained increase in cytoplasmic Ca2+ in the ECaT Ringer solution. Spiperone's induction of only a transient increase in cytosolic Ca2+ in normal and Ca2+-free normal Ringer solutions points to intracellular stores as the source for the transient increase.
Using Ca2+-free normal Ringer solution, we then ascertained whether the source of the increase in cytoplasmic Ca2+ was indeed the endoplasmic reticulum (ER). As part of the normal Ca2+ transit system across the ER, the Ca2+-ATPase pumps Ca2+ from the cytosol to the ER uphill against the concentration gradient. Other channels continue to cycle Ca2+ from the ER back to the cytosol. Thapsigargin blocks the Ca2+-ATPase, inhibiting the uphill uptake (43). Treatment of IB3-1 cells with thapsigargin produced a net increase of Ca2+ in the cytosol originating from the ER (Fig. 3A). In cells treated with spiperone and then with thapsigargin, there was no increase in cytosolic Ca2+ after thapsigargin treatment (Fig. 3B). The difference in these reactions to thapsigargin suggests that spiperone had emptied the ER of Ca2+.
Spiperone Appears to Act to Release Ca2+ From the ER Through a PYK-Coupled PLC
PLC is an intracellular enzyme that participates in Ca2+ signaling pathways. To establish whether spiperone acts through PLC-dependent pathways to increase cytosolic Ca2+, we preincubated the IB3-1 cells with U-73122, a known PLC inhibitor, and U-73433, a chemical analog of U-73122 that does not inhibit PLC (19). Pretreatment with U-73122 diminished the spiperone-induced Ca2+ transient, but pretreatment with U-73433 did not (Fig. 3, C and D). These data suggest that when PLC is inhibited, spiperone cannot enhance cytosolic Ca2+, which in turn suggests that spiperone releases ER Ca2+ through a PLC-dependent pathway. 2-Aminoethoxydiphenyl borate is an inhibitor of the inositol trisphosphate receptor on the ER membrane (30). Inositol trisphosphate is the downstream molecule of PLC (30). Thirty minutes of incubation with 2-aminoethoxydiphenyl borate (100 μM) totally blocked the spiperone-induced Ca2+ release from the ER (data not shown), which also suggests that the PLC pathway is involved in the spiperone-stimulated Ca2+ release mechanism.
Two major molecules activate PLC pathways to release Ca2+ from the ER: the G protein-coupled receptors and PYK (1). To establish which pathway was affected by spiperone, we incubated one set of IB3-1 cells with suramin, a general G protein-coupled receptor antagonist (9), and another set with genistein, a nonselective PYK inhibitor (1). Suramin did not reduce the spiperone-stimulated increase in Ca2+, but genistein partially blocked the increase (Fig. 3, E and F). Addition of lavendustin A (10 μM), another PYK inhibitor (31), significantly diminished the increase in Ca2+ induced by spiperone in IB3-1 cells (data not shown). All the data suggest that spiperone enhances release of Ca2+ from the ER by targeting a PYK-coupled PLC pathway.
Spiperone Released Ca2+ From the ER in Airway Cells
To assess whether spiperone induces the release of Ca2+ from other cell types, we tested its effect on Calu-3 cells, a non-CF human submucosal gland serous epithelial cell line, CFBE41o−1 cells, a CF human bronchial epithelial cell line that contains homozygous ΔF508-CFTR mutations, and NHBE, a primary normal human bronchial/tracheal epithelial cell culture. As in the IB3-1 cells, spiperone induced a transient increase in Ca2+ in Calu-3 non-CF, CFBE41o− CF, and NHBE cells (Fig. 4).
Spiperone Stimulated Cl− Secretion in Polarized Calu-3 and CFBE41o− Cell Monolayers
We needed to establish that the spiperone-induced increase of Ca2+ would produce an increase of Cl− transport relevant to CF and that it would do so in CF-relevant epithelial cells. Calu-3 and CFBE41o− cell monolayers were cultured on permeable filter supports, and “short-circuit current” was measured to evaluate the current produced by Cl− in these cells.
Na+ is a positively charged ion, and its absorption would interfere with the short-circuit current measurement for Cl− secretion. Therefore, in our normal Ringer solution, Na+ absorption across the cell membrane was blocked with amiloride (14). Cell monolayers were then exposed to the two relevant compounds: niflumic acid (NFA), which is a nonselective CaCC blocker (34), or CFTRinh-172, which is a specific CFTR blocker (41).
In Calu-3 cells, the non-CF line cell, spiperone produced a transient and then a sustained Cl− current. Figure 5, A and B, shows that transient Cl− was reduced by NFA and CFTRinh-172 and the sustained Cl− release was totally blocked. These results indicate that, at least in Calu-3 cells, spiperone not only activates the CaCCs, but it also may activate other Cl− channels, including CFTR.
In CFBE41o− cells, the CF cell line, spiperone caused only a transient Cl− secretion. When the CaCCs were blocked by NFA, spiperone's effect on Cl− secretion was reduced significantly (Fig. 5C). We further evaluated CaCC function in CFBE41o− cells by adding thapsigargin to the cells. As shown in Fig. 5D, thapsigargin blocked the reuptake of Ca2+ into the ER, causing a transient increase in Cl− secretion. Subsequent application of spiperone failed to stimulate additional Cl− transit. Taken together, our data suggest that spiperone's effect on CFBE41o− cells is to increase intracellular Ca2+ from the ER, which in turn activates CaCCs and produces a transient increase in Cl− secretion.
Antipsychotic Spiperone Analogs Did Not Have the Same Effect as Spiperone in Inducing Increases in Ca2+ and Cl− Secretion
To assess whether spiperone's antipsychotic effects were related to its function as an enhancer of Ca2+ transit and Cl− secretion, we examined the spiperone analogs haloperidol and ketanserin. Spiperone and haloperidol have a high affinity for and are antagonists to the dopamine D2 receptor (40). Spiperone and ketanserin have a high affinity for and are antagonists to the 5-HT2 receptor (10). In addition, as shown in Fig. 6A, the structure of all three drugs includes the 1-carbonyl oxygen atom and 4-flurobenzoyl moiety.
We found that ketanserin did not increase intracellular Ca2+ in IB3-1 cells, nor did it increase Cl− secretion in Calu-3 cells (Fig. 6, B and C). Haloperidol did not stimulate the Cl− current in Calu-3 cells, nor did it prevent spiperone from stimulating transepithelial Cl− currents (Fig. 6D). Even when we added both ketanserin and haloperidol to the IB3-1 cells before adding spiperone, spiperone was able to stimulate a transient increase in Ca2+ in the Ca2+-free normal Ringer solution (Fig. 6, E and F, respectively).
These data suggest that although ketanserin, haloperidol, and spiperone belong to the same family of antipsychotic drugs, their ability to induce an increase in Ca2+ and Cl− secretion in human airway epithelial cells is completely different. The data also indicate that spiperone has a different function in airway cells outside its known behavior as an antagonist of 5-HT2 and dopamine D2 receptors.
Spiperone Rescued Cl− Secretion in CFKO Mice as Assessed by Measurement of NPD
To assess the hypothesis that spiperone will stimulate Cl− secretion in vivo, we measured NPD in CF (double-transgenic CFKO) and wild-type (C57Bl/6) mice (Fig. 7). As spiperone was topically applied to nasal epithelia, we assessed Na+ and Cl− transport by measuring the voltage difference between the electrodes.
As expected for CF mice that hyperabsorb Na+ and have no CFTR, NPD was significantly hyperpolarized when we perfused the nares with baseline Ringer solution (Table 1). To allow for later hyperpolarization from Cl− secretion, we then perfused nasal epithelia with baseline Ringer solution plus amiloride to block the epithelial Na+ channel. For the next perfusion, we removed Cl− by substituting Na+ gluconate in the baseline Ringer solution (Table 1) and then removed Cl− from the mucosa by perfusing the nare with the Na+ gluconate-Ringer solution. This rinse further depolarized the epithelia, since these mice have no functioning CFTR.
We then perfused the mouse nasal epithelia with the zero-Cl− baseline Ringer solution with spiperone (1 μM), which hyperpolarized the mouse nasal potential significantly and indicated that spiperone induced an increase in Cl− conductance. Subsequent increases in the spiperone concentration (20 μM) did not boost Cl− conductance (Fig. 7, A and E). After 19 min of sustained perfusion, the NPD of the CF mouse remained significantly hyperpolarized by 1 μM spiperone, indicating that spiperone continuously stimulated nasal Cl− secretion in the CF mouse (Fig. 7B).
After we eliminated mucosal Cl− from wild-type C57Bl/6 mice, the CFTR Cl− channel was activated in the airway cells, secreted Cl−, and hyperpolarized the NPD. It was striking that addition of spiperone did not induce further hyperpolarization, which indicates that spiperone is unable to stimulate additional Cl− conductance in nasal epithelia with normally functioning CFTR (Fig. 7, C and E). The 1% DMSO control did not stimulate Cl− secretion in the wild-type mice (Fig. 7D).
By using a screening assay, we were able to rapidly assess a large number of compounds in the MSSP library to identify spiperone as the most potent Ca2+ enhancer. The discovery of spiperone also indicates a potential new CF therapy platform.
Our primary purpose was to identify possible CF therapies other than modification of the defective CFTR, an idea that has precedent. Several therapies that stimulate Cl− via increases in intracellular Ca2+ are under development. ATP and UTP operating through a P2Y2 purinergic receptor were proposed in the early 1990s to rescue Cl− secretion through the activation of CaCCs (25). Because these nucleotides are easily hydrolyzed by exonucleotidases, a nonhydrolyzed form of UTP, INS-37217 (denufosol), was developed (44). A primary end point for the phase III clinical trial of inhaled denufosol tetrasodium was recently completed. A 2.5% improvement in lung function was reported over 24 wk of treatment (21, 38). Moli-1901 interacts with the phospholipids in the membrane and activates the CaCC by elevating intracellular Ca2+ concentration (33). It stimulated Cl− secretion through the CaCC in normal and CF airway epithelia (3, 4, 45). The inhalation of nebulized Moli-1901 was well tolerated in CF patients in a phase II clinical trial. Moli-1901 showed a sustained effect in improving lung function (11).
From our data, we can infer that spiperone works via a PLC signaling pathway, because the PLC inhibitor U-73122 partially inhibited the spiperone-induced Ca2+ increase. PLC is an enzyme with a family of 13 isoforms divided into 6 different subfamilies: PLC-β, PLC-γ, PLC-δ, PLC-ε, PLC-ζ, and PLC-η (14). The PLC-β family is typically considered to be activated by a G protein-coupled receptor, whereas the PLC-γ family is activated by PYK receptors (1, 14).
In general, tyrosine kinase receptors regulate human airway goblet cell mucin secretion through a PLC-dependent pathway (1). Two major cell surface receptor tyrosine kinases, the epidermal growth factor receptor and the platelet-derived growth factor receptor families are the potential therapeutic targets in pulmonary diseases because of their important roles in chronic tissue remodeling in asthma, bronchitis, and pulmonary fibrosis (18). In our hands, the spiperone-induced increase in Ca2+ was blocked by genistein and lavendustin A, two general tyrosine kinase inhibitors (31). This result suggests that spiperone affects a PYK, activating the PLC pathway to release Ca2+ from the ER.
Spiperone acts in a similar way in several human airway epithelial cells, including the CF cell lines IB3-1 and CFBE41o−, the non-CF cell line Calu-3, which highly expresses CFTR (20), and one primary cell culture, NHBE. Importantly, in all four cell types in Ca2+-free Ringer solution, spiperone stimulated transient increases in cytosolic Ca2+.
The important question now became whether this increase in cytoplasmic Ca2+ would translate into Cl− secretion that would be of potential benefit in CF. Our results indicate, as expected, that spiperone activates CaCCs in CFBE41o− and Calu-3 cells and induces them to secrete Cl−. In a more surprising result, in Calu-3 cells, spiperone is involved in CFTR activation as well.
Looking at Cl− channels in vivo, we noted that the CaCC is downregulated in cells containing normal CFTR and upregulated in CF cells where CFTR is compromised (24). For example, in intact murine CF tracheal epithelia, application of UTP leads to a robust increase in Cl− secretion through the CaCC (17).
Our results with spiperone in wild-type and CFKO mice follow this pattern exactly. In normal mice, the CaCC is downregulated, and in the CF mice, it is upregulated. In our experiments in CF mice, spiperone stimulated a significant and sustained increase in NPD that indicated enhanced Cl− secretion in the CF mouse. By contrast, in normal mice, NPD did not change after exposure to spiperone. The observation that spiperone stimulates Cl− secretion in human CF cell lines and in intact nasal epithelium from CFKO mice enhances spiperone's potential therapeutic value.
Since spiperone is an antipsychotic, we wondered whether its antipsychotic mechanism could be tied to its ability to stimulate Cl− secretion. We chose to compare spiperone with two of its analogs, haloperidol and ketanserin. Spiperone has a high affinity for, and is an antagonist of, dopamine D2 receptors, with an IC50 of 10 nM (13, 27). Dopamine, on the other hand, is the ligand of the dopamine D2 receptor and stimulates Na+ and Cl− absorption in the rabbit ileum by interacting with D2 and α2-adrenergic receptors (8). Dopamine also has other effects on Na+ metabolism. In CF, the epithelial Na+ channel is highly active, leading to excess Na+ absorption. Helms et al. found that dopamine increases epithelial Na+ channel activity in the rat alveolar type 2 cell line L2 through the D1 receptor (15), and in alveolar type I (AT1) cells through D1, but not D2, receptors (16).
Spiperone also is a known antagonist to the 5-HT2 receptor (IC50 = 40 nM) (5, 28), which affects the release and activity of other neurotransmitters such as glutamate, GABA, and, as part of this complex net, dopamine (10). The receptor ligand 5-HT is particularly important as a well-known mediator of intestinal Cl− secretion. Exogenous 5-HT is known to induce Cl− secretion, which is inhibited by the cyclooxygenase inhibitor piroxicam in the rat distal colon (22). The 5-HT3 receptor agonists also activate Cl− secretion in the rat distal colonic mucosa (6).
Haloperidol is a potent neuroleptic drug that also binds to dopamine D2 receptors with high affinity (40). Radiolabeled [3H]spiperidone and [3H]haloperidol bind selectively and with high affinity to dopamine receptor sites in the mammalian brain (35, 40) and inhibit their function. Ketanserin is a new 5-HT2 antagonist that binds to the 5-HT2 receptor with a high affinity (26), similar to that of [3H]spiperone (28). Ketanserin is very weakly antagonistic to the dopamine receptor (26).
Of these three compounds, we found that only spiperone could stimulate Cl− secretion in epithelial cells. Cells responded to spiperone after being pretreated with haloperidol, which suggests that spiperone is not operating as an antagonist of D2 receptors to stimulate Cl− secretion in human airway epithelia in vitro. The response of the cells to ketanserin also suggested that, in this case, spiperone does not operate through a 5-HT2 receptor to activate Cl− secretion. These results may open a door to modifications that would strip away the antipsychotic properties and make spiperone-related CF therapy feasible.
We should note that these results have not been universal and that the effect of 5-HT on airway epithelium ion transport is not entirely clear (39). In another study, results were in the opposite direction: 5-HT inhibited airway epithelial Na+ absorption and stimulated Cl− secretion in canine tracheal epithelia, and spiperone (as well as ketanserin) blocked those effects (39). Our data show that, in human and mouse airway epithelial cells, spiperone stimulates Cl− secretion through a Ca2+-dependent mechanism. The most likely explanation for the difference may be drug dosing. When spiperone worked as an antagonist of the 5-HT receptor in the previous study, its IC50 for inhibition of 5-HT-stimulated Cl− secretion is 0.36 μmol (39). In our study, at least 1 μmol of the spiperone was required to induce Cl− secretion. Another reason for the discrepancies may lie in the tissue preparation. In the previous study, dog trachea was excised and mounted on two Lucite half-chambers (39), whereas we grew human airway epithelial cells in monolayers that were mounted in Ussing chambers.
Spiperone's different responses in these two studies may indicate that we have identified a new function for it beyond its known role as an antagonist of 5-HT2 and dopamine D2 receptors. Because it apparently has a number of functions in the body, it is likely that chemical modification to avoid unintended effects would be necessary before spiperone could be developed as a treatment for CF (23).
In conclusion, we have developed a new application by using the compound screening system to assess cytoplasmic Ca2+ change, and the assay has proved its value in our search for a therapeutic compound for CF. It identified a new role for spiperone as an enhancer of cytoplasmic Ca2+ that affects the CaCC and induces Cl− secretion in epithelial tissue. We established that this function of spiperone operates through the PYK-coupled PLC pathway, although its targets have yet to be identified. Spiperone activates Ca2+ release and stimulates Cl− secretion in human airway epithelia by mechanisms other than its well-known function as an antagonist of 5-HT2 and dopamine D2 receptors. Its newly found action on Cl− transport appears to be separate from its antipsychotic pathway. It rescues CF nasal epithelia in mice by inducing sustained Cl− secretion. Briefly, the discovery of spiperone represents a promising new platform on which a new therapy for CF could be developed.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2009 the American Physiological Society