SPI-0211 activates T84 cell chloride transport and recombinant human ClC-2 chloride currents

John Cuppoletti, Danuta H. Malinowska, Kirti P. Tewari, Qiu-ju Li, Ann M. Sherry, Myra L. Patchen, Ryuji Ueno

Abstract

The purpose of this study was to determine the mechanism of action of SPI-0211 (lubiprostone), a novel bicyclic fatty acid in development for the treatment of bowel dysfunction. Adult rabbit intestine was shown to contain mRNA for ClC-2 using RT-PCR, Northern blot analysis, and in situ hybridization. T84 cells grown to confluence on permeable supports were shown to express ClC-2 channel protein in the apical membrane. SPI-0211 increased electrogenic Cl transport across the apical membrane of T84 cells, with an EC50 of ∼18 nM measured by short-circuit current (Isc) after permeabilization of the basolateral membrane with nystatin. SPI-0211 effects on Cl currents were also measured by whole cell patch clamp using the human embryonic kidney (HEK)-293 cell line stably transfected with either recombinant human ClC-2 or recombinant human cystic fibrosis transmembrane regulator (CFTR). In these studies, SPI-0211 activated ClC-2 Cl currents in a concentration-dependent manner, with an EC50 of ∼17 nM, and had no effect in nontransfected HEK-293 cells. In contrast, SPI-0211 had no effect on CFTR Cl channel currents measured in CFTR-transfected HEK-293 cells. Activation of ClC-2 by SPI-0211 was independent of PKA. Together, these studies demonstrate that SPI-0211 is a potent activator of ClC-2 Cl channels and suggest a physiologically relevant role for ClC-2 Cl channels in intestinal Cl transport after SPI-0211 administration.

  • cystic fibrosis transmembrane regulator
  • intestinal chloride channels
  • chloride channel opener

spi-0211 (lubiprostone) is a gastrointestine-targeted bicyclic fatty acid that enhances intestinal fluid secretion and has been shown to be safe and highly effective in treating constipation in clinical trials (18, 19). The gastrointestinal effects of SPI-0211 appear to be mediated by increased Cl secretion, because the Cl concentration in secreted intestinal fluid increases after SPI-0211 administration (41). The mechanism by which SPI-0211 enhances Cl secretion is not known, but an increased Cl concentration could result from enhanced Cl channel currents in gastrointestinal epithelial membranes.

In epithelial cells, it is established that Cl movement across the basolateral membrane occurs via the Na+-K+-2Cl co-transporter and that Cl movement across the apical (luminal) membrane occurs via specific Cl channels. Ca2+-activated K+ channels in the basolateral membrane help to set the driving force for Cl movement across the cell and tight junctions between cells simultaneously provide transepithelial resistance to the movement of other ions and molecules. Figure 1 schematically illustrates possible Cl transport mechanisms for epithelial cells in which both cystic fibrosis transmembrane regulator (CFTR) and ClC-2 Cl channels are present in the apical membrane.

Fig. 1.

Schematic model of Cl transport in intestinal epithelial cells. At the basolateral membrane, Cl enters the cell from blood across the Na+-K+-2Cl cotransporter. Sodium is expelled by the sodium pump, and potassium leaves via a potassium channel. Sodium is shown crossing the cell layer via a paracellular pathway, but sodium channels also exist (not shown). Both cystic fibrosis transmembrane regulator (CFTR) and ClC-2 Cl channels are present on the apical membrane and can allow Cl to exit the cell. Adapted from Pilewski and Frizzell (31).

ClC-2 is a member of the ClC Cl channel family, which consists of nine members that are widely distributed in nature (6, 20, 27, 40, 42). ClC-2 is found in the intestine (1, 5, 16, 23, 28), gastric parietal cells (26, 34), liver (32), lung (3, 6, 8, 33, 35), retina (12), parotid acinar cells (30), cardiac tissue (4), and neuronal cells (36). CFTR is also a Cl channel, and mutations in it lead to cystic fibrosis. CFTR is also widely distributed in the tissues (14, 21) and is found in the apical membrane of epithelial cells (31), including the intestine (5, 15, 29) and T84 intestinal cell lines (39). Intestinal epithelial cells are also known to contain Ca2+-activated Cl channels (2).

Several studies have demonstrated the presence of ClC-2 in intestinal tissues. However, there may be species differences in the development of ClC-2 expression, function, and localization. For example, recent studies have demonstrated that the ClC-2 protein is present in the murine intestine, albeit not in the apical membrane but at the tight junction complexes (16). In adult rabbit stomach, ClC-2 mRNA is present in parietal cells and ClC-2 protein is present in the apical membrane of the parietal cell (34). Furthermore, immunostaining indicates that ClC-2 protein is present in the apical membrane of human intestine (23). In the present study, adult rabbit was used to study the relative abundance and distribution of ClC-2 mRNA in the intestine.

The presence of ClC-2 in the apical intestinal membrane suggests a possible functional role for ClC-2 in apical Cl transport, although this remains to be fully elucidated. With the exception of Cl transport studies in guinea pig intestine (5) and murine intestine (16), studies of ClC-2 in the intact intestines of other species have not been found in the literature. ClC-2-mediated Cl transport has been studied in two human intestinal cell lines: Caco-2 (28) and T84 (1). Caco-2 cells contain ClC-2 protein, which contributes to Cl secretion, but the distribution of the protein in Caco-2 cells is at the tight junction (28). Reports of patch-clamp studies of ClC-2 in T84 cells suggest a role in Cl transport (1); however, immunostaining suggested that only a small subset of cells in the cultures contained ClC-2 when grown on glass coverslips (23). In the present study, the abundance and distribution of ClC-2 in T84 cells grown on permeable supports were investigated to determine whether T84 cells could be used to study the effects of SPI-0211 on ClC-2- and CFTR-mediated Cl transport.

CFTR is activated by PKA and inhibited by arachidonic acid (22), whereas ClC-2 activation can occur via PKA or arachidonic acid and other lipids in a PKA-independent manner (8, 38). Effects of SPI-0211 were studied in nontransfected human embryonic kidney (HEK)-293 cells and in HEK-293 cells stably transfected with human ClC-2 or human CFTR. To differentiate the effects on CFTR and ClC-2, experiments were performed in the presence and absence of myristoylated protein kinase inhibitor (mPKI), a specific PKA inhibitor.

This study shows that SPI-0211 is a potent PKA-independent activator of human ClC-2 (not CFTR), suggesting that ClC-2 may be a target for the beneficial effects of SPI-0211 (18, 19). These findings also suggest an important physiological role for ClC-2 in intestinal fluid transport.

MATERIALS AND METHODS

Materials.

SPI-0211 was obtained from Sucampo Pharmaceuticals (Sanda, Japan) as frozen aliquots of either 100 μM or 2 mM solutions in 100% DMSO. DMSO, also obtained as frozen aliquots from Sucampo Pharmaceuticals, was used to dilute the SPI-0211 and when testing for vehicle effects. AlexaFluor 546-labeled goat anti-chicken antibody and AlexaFluor 488-labeled phalloidin were obtained from Molecular Probes (Eugene, OR). MEM, Lipofectamine, hygromycin, pCEP vector, G-418, pcDNA3.1, DMEM/Ham’s F-12 medium, and the SuperScript preamplification system and random primers DNA labeling kit were purchased from Invitrogen (Carlsbad, CA). HEK-293 cells were obtained from the American Type Culture Collection (Manassas, VA). Forskolin and mPKI were obtained from Calbiochem-Novabiochem (San Diego, CA), isobutylmethyl xanthine (IBMX; a phosphodiesterase inhibitor) and nystatin were purchased from Sigma-Aldrich (St. Louis, MO), and arachidonic acid was obtained from Avanti Polar Lipids (Alabaster, AL). 1-Ethyl-2-benzimidazolinone (1-EBIO) was purchased from Tocris Cookson (Ellisville, MO). Oligotex Direct mRNA columns were obtained from Qiagen (Valencia, CA), and Takara Ex-Taq DNA polymerase was purchased from Fisher Scientific. Borosilicate glass (no. 7052) was obtained from Garner Glass (Claremont, CA). Chicken anti-ClC-2 antibody (3) was a kind gift from Dr. Carol J. Blaisdell (Department of Pediatrics, University of Maryland School of Medicine, Baltimore, MD).

Rabbit tissues.

Rabbits were anesthetized with sodium pentobarbital according to an Institutional Animal Care and Use Committee-approved animal protocol. Tissues were removed and rapidly frozen in liquid N2 for mRNA extraction or fixed in 4% paraformaldehyde and embedded in paraffin for in situ hybridization.

Detection of ClC-2 by RT-PCR.

Detection of mRNA was performed essentially as previously described (34, 35). Briefly, mRNA was prepared from rapidly frozen rabbit brain, heart, skeletal muscle, large intestine, small intestine, gastric mucosa, and lung using Oligotex Direct mRNA columns. The mRNA (2 μg) from each tissue was pretreated with DNase I, and then first-strand cDNA was synthesized using the SuperScript preamplification system. The cDNA was amplified using Ex-Taq DNA polymerase and sequence-specific primers for ClC-2 and β-actin. PCR conditions for amplification of a 468-bp cDNA fragment of the ClC-2 COOH-terminal cytoplasmic domain consisted of 35 cycles of denaturing for 45 s at 94°C, annealing for 45 s at 61°C, and elongating for 2 min at 72°C. To compare the relative levels of ClC-2 mRNA across tissues, a 669-bp cDNA fragment of human β-actin was amplified. PCR conditions for β-actin consisted of 35 cycles of denaturing for 45 s at 94°C, annealing for 45 s at 57°C, and elongating for 2 min at 72°C.

Northern blot analysis.

Northern blot analysis was performed as previously described (34, 35). The mRNA was extracted from rapidly frozen adult rabbit tissues (kidney, bladder, small intestine, and large intestine). The mRNA (2–5 μg) from various rabbit tissues was resolved on 1.2% agarose-formaldehyde gel and transferred to a nylon membrane. Rabbit ClC-2 cDNA was excised from pcDNAII, gel purified, 32P labeled using a random primers DNA labeling kit, and used to screen the blot under medium stringency. A 1.8-kb fragment of human β-actin (Clontech, Palo Alto, CA) was 32P labeled and used to screen the same blot under high stringency. Autoradiography of Northern blots was performed at −80°C.

In situ hybridization.

In situ hybridization was performed as previously described (34, 43). Sections were loaded onto salinated slides, pretreated with proteinase K, and acetylated immediately before use (43). A 261-bp cDNA fragment of the ClC-2 COOH-terminal domain was used as a template for the preparation of antisense and sense 35S-labeled cRNA fragments. Hybridization was performed as described previously (34, 43) at 55°C, followed by high-stringency washes. Autoradiography was performed for 10–42 days, and results were analyzed with a Nikon microscope and a dark-field filter. Tissue sections were then stained with hematoxylin and eosin and viewed under bright field.

Cell culture.

T84 cells obtained from Dr. Jeffrey B. Matthews (Department of Surgery, University of Cincinnati, Cincinnati, OH) (37) were grown in DMEM/Ham’s F-12 medium with 6% heat-inactivated FBS, 15 mM HEPES, 14.3 mM NaHCO3, 100 U/ml penicillin, and 100 μg/ml streptomycin sulfate. T84 cells were grown to confluence, either on 0.3-cm2 BioCoat collagen-coated permeable supports (Discovery Labware, Bedford, MA) or on 1.13-cm2 collagen-coated Snapwell permeable supports (Corning, Corning, NY). ClC-2-transfected HEK-293 cells were prepared as described by Tewari et al. (38). HEK-293 cells were transfected with His- and T7-tagged humanClC-2 cDNA in pcDNA3.1 using Lipofectamine according to the manufacturer’s instructions. Cells were selected with 300 μg/ml G418. HEK-293 cells were also stably transfected using Lipofectamine with wild-type, expressible human CFTR cDNA in pCEP vector obtained from Dr. Mitch Drumm (Department of Pediatrics, Case Western Reserve University School of Medicine, Cleveland, OH) (25). Stable transfectants were selected using hygromycin (100 μg/ml). Both ClC-2-transfected and CFTR-transfected HEK-293 cells were grown in MEM supplemented with 5% heat-inactivated horse serum, 0.1 mM nonessential amino acids, 2 mM l-glutamine, 1 mM sodium pyruvate, 100 U/ml penicillin, and 100 μg/ml streptomycin sulfate.

Immunoconfocal microscopy.

Confluent T84 cells were fixed in 4% paraformaldehyde, permeabilized with 0.5% Triton X-100 in PBS for 20 min, and washed three times in PBS. Cells were blocked with 1% BSA (RIA grade), 0.05% Tween 20, and 10% goat serum in PBS for 30 min; incubated with chicken anti-ClC-2 antibody (1:300) for 1 h (3) at 22°C; and washed three times for 5 min each with PBS containing 0.05% Tween 20. Cells were then stained for 1 h at 22°C with AlexaFluor 546-labeled goat anti-chicken antibody (1:1,000 dilution), with ClC-2 staining red and AlexaFluor 488-labeled phalloidin (20 U/ml) with F-actin staining green. After three washes as described above, the cells were mounted with Gel/Mount (Biomeda, Foster City, CA) and examined using an LSM 510 Zeiss confocal microscope with argon (488 nm excitation) and HeNe (546 nm excitation) lasers, an Axioplan upright microscope, and a 64× oil objective. X-z and x-y scans were obtained. LSM 510 software was used to process the digitized images obtained.

Short-circuit current measurements.

Two setups were used. Initially, a Plexiglas chamber was used for short-circuit current (Isc) measurements across confluent T84 cell monolayers grown on 0.3-cm2 permeable support filters (World Precision Instruments, Sarasota, FL). Electrical measurements were obtained with a 742C voltage-clamp device (Department of Biomedical Engineering, University of Iowa, Iowa City, IA). The temperature was held constant at 37°C by circulating heated water through the water jacket of the chamber. The output of the amplifier was plotted on an analog chart recorder. Isc was recorded after addition of test compounds. Background current was subtracted from Isc after the addition of compound, and Isc was then normalized to the filter area (0.3 cm2). With the use of this Isc setup, Isc measurements in T84 cells were performed as described by Loffing et al. (24). The basolateral membrane solution contained (in mM) 116 NaCl, 24 NaHCO3, 3 KCl, 2 MgCl2, 0.5 CaCl2, 3.6 sodium-HEPES, 4.4 hydrogen-HEPES (pH 7.4), and 10 glucose. The apical membrane bath solution was identical, with the exceptions that the Cl concentration was reduced by substitution of NaCl with sodium gluconate, CaCl2 was increased from 0.5 to 2 mM to account for chelation of Ca2+ by gluconate, and glucose was omitted. Both solutions were bubbled with 95% O2-5% CO2, which also served to help mix the solutions. In all cases in which this setup was used, the basolateral membrane was permeabilized with 200 μg/ml nystatin.

Subsequently, the EasyMount Ussing chamber system (6 chambers) with VCCMC6 multichannel current-voltage (I-V) clamps, purchased from Physiologic Instruments (San Diego, CA), was used for Isc measurements across confluent T84 cell monolayers. Transepithelial resistance of T84 cells was monitored with an EVOM epithelial voltohmmeter (World Precision Instruments). Cells were used when the transepithelial resistance of the monolayer was >1,200 Ω. For Isc measurements, special sliders (Physiologic Instruments) were used for Snapwell cell culture inserts (1.13 cm2). The solutions were continuously gassed with 95% O2-5% CO2, which also provided stirring, and the temperature was held constant at 37°C with a heating block. The clamps were connected to Acquire and Analyse software (Physiologic Instruments) for automatic data collection from all six of the Ussing chambers. Ag/AgCl reference electrodes were used for measuring transepithelial voltage and passing current. Isc measurements were performed as described by Devor et al. (10). For measurements in intact T84 cells, symmetrical bath solutions were used. Basolateral and apical membrane bath solutions contained (in mM) 120 NaCl, 25 NaHCO3, 3.3 KH2PO4, 0.8 K2HPO4, 1.2 MgCl2, and 1.2 CaCl2 (pH 7.4). The basolateral solution had 10 mM glucose. To ensure the absence of any Na+-glucose cotransport, the apical solution contained 10 mM mannitol (7). To examine effects on the apical membrane alone, the basolateral membrane was permeabilized with 200 μg/ml nystatin and a basolateral-to-apical Cl gradient was imposed by using an apical membrane bath solution that was identical, except that the Cl concentration was reduced by substituting sodium gluconate for NaCl and CaCl2 was increased to 4 mM to account for the chelation of Ca2+ by gluconate. In some experiments, an apical-to-basolateral Cl gradient was imposed. Under these conditions, the Cl conductance across the apical membrane is directly proportional to the Isc (24). To examine effects on the basolateral membrane, the apical membrane was permeabilized with nystatin and a basolateral-to-apical Cl gradient was imposed. In preliminary studies, Isc was measured in T84 cells treated with test drug and a final DMSO concentration of 1%. Under these conditions, it was determined that DMSO caused damage to the permeable monolayer support filters. Therefore, final DMSO concentrations were adjusted to 0.1% in subsequent experiments.

Patch-clamp measurement of whole cell Cl currents.

Patch-clamp and analytical methods were described previously (8, 13, 17, 38). Currents were elicited by voltage-clamp pulses (1,500 ms duration) between +40 and −140 mV in 20-mV increments from a beginning holding potential of −30 mV. Currents were averaged over a 50-ms time course starting at 50 ms and ending at 100 ms. The external solution was normal Tyrode solution containing (in mM) 135 NaCl, 1.8 CaCl2, 1 MgCl2, 5.4 KCl, 10 glucose, and 10 HEPES (pH 7.4 or as indicated). The pipette solution contained (in mM) 130 CsCl, 1 MgCl2, 5 EGTA, 10 HEPES (pH 7.4), and 1 ATP-Mg2+ (pH 7.4). When ATP-Mg2+ (1:1 dilution) was made, 2 mM Cl was added. These solutions would have a theoretical reversal potential of approximately −3 mV using activity coefficients of pure NaCl and pure CsCl in the bath and pipette solutions. Pipettes were prepared from borosilicate glass and pulled by a two-stage Narashige puller to produce 1- to 1.5-MΩ resistance. Data were acquired with an Axopatch CV-4 headstage, a Digidata 1200 digitizer, and an Axopatch 1D amplifier. Data were analyzed using pClamp 6.04 (Axon Instruments, Union City, CA), Lotus 1-2-3 (IBM, White Plains, NY), and Origin software (OriginLab, Northampton, MA). Initially, 100 μM SPI-0211 was diluted in DMSO to deliver 1 μM SPI-0211 and a final DMSO concentration of 1%. In later experiments, 2 mM SPI-0211 was diluted in DMSO to deliver 1 nM-10 μM SPI-0211 and a final DMSO concentration of 0.1%; the specific conditions are indicated for individual experiments. To determine maximal Cl currents in transfected HEK-293 cells, either 5 μM forskolin (a PKA activator) plus 20 μM IBMX or 1 μM arachidonic acid was added at the end of experiments. Forskolin/IBMX stimulates ClC-2 via a PKA-dependent mechanism, while stimulation of ClC-2 by arachidonic acid is PKA independent (8, 38). Each of these treatments increased the final DMSO concentration by no more than 0.1%. In some studies evaluating the role of PKA in ClC-2 activation, 0.4 μM mPKI was added to inhibit PKA before activation with test agents.

Statistical significance.

The statistical significance of the difference between two means was determined using Student’s t-test. Origin 5.0 software was used to fit curves and linear plots.

RESULTS

Identification, distribution, and localization of ClC-2 mRNA and protein in the rabbit gastrointestinal tract and in human T84 cells.

To determine whether ClC-2 was present in adult rabbit tissues, RT-PCR was used. A 459-bp fragment of ClC-2 and a 669-bp fragment of β-actin were amplified. β-Actin was used to control for the amount of cDNA used. As shown in Fig. 2A, ClC-2 mRNA was present in both the large and small intestines as well as in gastric mucosa, brain, heart, skeletal muscle, and lung. Northern blotting and in situ hybridization were performed to determine the transcript sizes, verify the RT-PCR results, and localize the mRNA. As shown in Fig. 2B, the major species detected in the small intestine and large intestine by Northern blotting is a 3.4-kb mRNA transcript. In Fig. 2C, in situ hybridization shows ClC-2 mRNA in the epithelial cells lining both the small and large intestine. No specific labeling was seen with the sense control riboprobe. Thus ClC-2 mRNA is present in the adult rabbit intestine. T84 cells, a human intestinal cell line, were then screened for the presence and localization of ClC-2 protein by immunoconfocal microscopy, and the results are shown in Fig. 3. In confluent monolayers of T84 cells grown on permeable supports, ClC-2 Cl channel protein was stained red and F-actin, staining green, outlines the cells. Fig. 3A shows the x-z scan obtained across four cells, indicated by the arrow in Fig. 3B. The ClC-2 protein was localized at or near the apical membrane in the bulk of the T84 cells. The very small amount of punctate staining in the basolateral membranes is likely to be insignificant in relation to the major staining observed at the apical membrane. As shown in Fig. 3C, when the primary antibody to ClC-2 was omitted, staining was virtually eliminated, demonstrating specificity of the antibody. The small amount of very low red staining is likely secondary antibody trapped in tissue spaces.

Fig. 2.

Identification and distribution of ClC-2 mRNA in a variety of rabbit tissues based on RT-PCR analysis (A), Northern blot analysis (B), and in situ hybridization (C). A: 468-bp cDNA fragment of the ClC-2 COOH-terminal cytoplasmic domain and 669-bp cDNA fragment of β-actin amplified by RT-PCR using the same amount of first-strand cDNA. The mRNA (2 μg) from each tissue was used. For control, first-strand cDNA synthesis was performed with sterile water instead of mRNA (No mRNA). B: tissue mRNA was hybridized with 32P-labeled rabbit ClC-2 cDNA at medium stringency and with 32P-labeled human β-actin at high stringency. The mRNA (2–5 μg) was loaded onto each lane. Positions and sizes of RNA markers are shown at left. C: in situ hybridization of small and large intestine with anti-sense and sense (control) 35S-labeled rabbit ClC-2 riboprobes. i: bright-field photomicrographs of hematoxylin and eosin (H + E)-stained tissue sections. ii and iii: dark-field photomicrographs of riboprobe labeling. a indicates location of epithelial cells, and b indicates smooth muscle cells. Magnification, ×95.

Fig. 3.

Immunoconfocal microscopy of ClC-2 Cl channels in the human T84 gastrointestinal epithelial cell line. Confluent T84 cells grown on permeable supports were fixed and permeabilized. Cells were stained with chicken anti-ClC-2 antibody and AlexaFluor 546-labeled goat anti-chicken antibody, which stains ClC-2 red, and with AlexaFluor 488-labeled phalloidin, which stains F-actin green. A: x-z scan through the cells (arrow in B) from the apical side (top) to the basolateral side (bottom, medium and support side). C: anti-ClC-2 antibody omitted. In A, B, and C, top images show ClC-2 staining, middle images show F-actin staining, and bottom images show overlays of ClC-2 and F-actin staining. ClC-2 colocalizes with actin at the apical surface (yellow) of the T84 cells.

Isc studies.

The effect of SPI-0211 on transepithelial Cl transport of T84 cells was then studied. The cells were grown on permeable supports to form a monolayer and mounted in an Ussing chamber, and Isc was measured and normalized to area. Figure 4A shows the effect on Isc of 300 μM 1-EBIO followed by 1 μM SPI-0211. Both compounds (both lipophilic) were added to the apical and basolateral sides, and because they were dissolved in DMSO, the appropriate DMSO controls are also shown. 1-EBIO caused a large, sustained increase in Isc (P < 0.001 vs. DMSO) as previously reported (9). SPI-0211 caused a further large, significant, and sustained increase in Isc (P < 0.001 vs. 1-EBIO + DMSO), while DMSO alone had no effect. The current measured (Isc) is a measure of transepithelial Cl transport (9). Thus both 1-EBIO and SPI-0211 caused increased transepithelial Cl transport. The effect of 1-EBIO on apical Cl transport is due to the opening of a basolateral Ca2+-activated K+ channel and activation of CFTR (9, 11). In further experiments, the effect of SPI-0211 added to either the apical or the basolateral membrane bath solutions was examined. Over the time course observed (15 min), SPI-0211 was slightly more effective (P < 0.05) when added to the apical side [change in IscIsc) = 72.16 ± 4.37 μA/cm2 (n = 6)] than it was when added to the basolateral side [ΔIsc = 56.15 ± 4.44 μA/cm2 (n = 6)].

Fig. 4.

Effect of 1-ethyl-2-benzimidazolinone (1-EBIO) and SPI-0211 (lubiprostone) on short-circuit current (Isc) in intact T84 cells and effect of SPI-0211 on Isc in nystatin-permeabilized T84 cells. A: Isc of intact T84 cells was allowed to equilibrate and then was measured after addition of 300 μM 1-EBIO. When the maximum Isc response was reached, 1 μM SPI-0211 was added. Isc was normalized to area. Both compounds were added to both sides. DMSO controls are also shown. Data are means ± SE. ○, DMSO (n = 4); □, 1-EBIO (n = 11), followed by DMSO (n = 3); ▴, 1-EBIO (n = 11), followed by SPI-0211 (n = 8). *P < 0.001 vs. DMSO. #P < 0.01 vs. DMSO. B: Isc was allowed to equilibrate, and then nystatin (200 μg/ml) was added to the basolateral membrane. A Cl gradient was then imposed either apically to basolaterally or vice versa. After 10-min equilibration with nystatin, 1 μM SPI-0211 was added to both sides. When the response to SPI-0211 reached its maximum, 300 μM 1-EBIO was added to both sides. DMSO (0.1%) controls are also shown. Data are means ± SE. •, SPI-0211 (n = 8); ○, DMSO (n = 6) with apical-to-basolateral Cl gradient; ▴, SPI-0211 (n = 7); ▵, DMSO (n = 5) with basolateral-to-apical Cl gradient. C: bar graph comparing the effect of SPI-0211 on Isc across apical membrane (A) (closed bar, basolateral membrane permeabilized with nystatin) and basolateral membrane (BL) (open bar, apical membrane permeabilized with nystatin) of T84 cells with a basolateral-to-apical Cl gradient imposed. DMSO controls were subtracted. Means ± SE (n = 9) are plotted. *P < 0.001 apical vs. basolateral Isc.

To investigate whether the SPI-0211 effect was on an apical Cl conductance, the basolateral membrane was permeabilized with nystatin (a polyene antibiotic that allows the movement of ions without the loss of larger molecules, such as ATP, from the cell) and a Cl gradient was imposed by reduction of the Cl concentration in the solution bathing the apical membranes (24). Under these conditions, the Cl conductance across the apical membrane is directly proportional to the Isc (24). The results are shown in Fig. 4B, together with results obtained with a reversed (apical to basolateral) Cl gradient imposed after nystatin permeabilization of the basolateral membrane. Nystatin caused small, not significantly different changes in Isc, regardless of the direction of the Cl gradient. SPI-0211 then caused a large, significant, and sustained increase (P < 0.001 vs. DMSO) in Isc with both gradients, which were not significantly different from each other. 1-EBIO was then added and had no significant effect, indicating that basolateral membrane permeabilization with nystatin was complete. These results suggest that SPI-0211 is activating a Cl conductance in the apical membrane of T84 cells. To further rule out basolateral effects, the effect of SPI-0211 on apical and basolateral membrane conductance was examined and compared. Isc was measured using a basolateral-to-apical Cl gradient, with either the basolateral membrane or the apical membrane permeabilized with nystatin. Figure 4C shows that SPI-0211 increased apical IscIsc = 80.7 ± 6.3 μA/cm2; n = 9) and had virtually no effect on basolateral IscIsc = 4.4 ± 2.5 μA/cm2; n = 9). The effect of varying SPI-0211 concentrations on Isc after nystatin permeabilization of the basolateral membrane was then examined. Figure 5A shows Isc traces after addition of 20 and 250 nM SPI-0211, and Fig. 5B shows ΔIsc normalized to area plotted as a function of SPI-0211 concentration and both data sets (measured on the 2 different Isc apparatuses) were used. The modified Michaelis-Menten hyperbolic plot was used to fit the data with the equation ΔIsc = ΔIsc max·[SPI-0211]/(EC50 + [SPI-0211]). The analysis is constricted to 0, because ΔIsc was measured. Also plotted as insets in Fig. 5B are the Lineweaver-Burk and Hill plots of the data. Table 1 shows that EC50 and ΔIsc max, maximum predicted change in current (or transport rates) obtained from both hyperbolic and Lineweaver-Burk plots, were similar: EC50 = 18.24 ± 3.72 nM and ΔIsc max = 77.82 ± 5.04 μA/cm2 (n = 11) from the hyperbolic plot, and EC50 = 18.82 ± 3.55 nM and ΔIsc max = 75.88 ± 6.91 μA/cm2 (n = 3) from the Lineweaver-Burk plot. The hyperbolic plot fit the data very well (P < 0.005): χ2 = 37.39. The Hill coefficient was 1.14 ± 0.08 (n = 8), indicating not only that the modified Michaelis-Menten hyperbolic plot is the appropriate analysis but also that the simple bimolecular first-order kinetic model is correct. These data indicate that SPI-0211 is a potent activator of apical Cl channels in T84 cells. However, because T84 cells contain both ClC-2 and CFTR Cl channels (and perhaps other channels) in their apical membranes, more definitive studies were pursued to evaluate the effects of SPI-0211 on human ClC-2 and human CFTR separately using HEK-293 cells stably transfected with either ClC-2 (8, 38) or CFTR. Nontransfected HEK-293 cells exhibit minimal endogenous Cl channel currents. For these studies, patch-clamp methodologies (17) were used to study whole cell Cl currents (8, 13, 38).

Fig. 5.

Effect of varying SPI-0211 concentration on Isc (A) or ΔIsc (B) in nystatin-permeabilized T84 cells. A: effect of 20 nM and 250 nM SPI-0211 (added to both sides) on Isc measured in T84 cells with the basolateral membrane permeabilized with nystatin (200 μg/ml) and a basolateral-to-apical Cl gradient. Data are means ± SE (n = 6). •, 20 nM SPI-0211 + 250 nM SPI-0211 + 1-EBIO; ○, 0.1% DMSO + 250 nM SPI-0211 + 1-EBIO. *P < 0.001 vs. DMSO. #P < 0.01 vs. 20 nM. **P < 0.02 vs. DMSO. B: SPI-0211 concentration is plotted against ΔIsc [means ± SE (n = 3–17)] using data obtained with both Isc setups. Data were fit with a modified Michaelis-Menten hyperbolic plot using Origin 5.0 software. Lineweaver-Burk and Hill plots are shown in the insets. All parameters calculated are shown in Table 1.

View this table:
Table 1.

Kinetic analysis of ΔIsc measured across monolayers of T84 cells in response to varying concentrations of SPI-0211

Patch-clamp studies.

Patch-clamp studies were performed in both ClC-2-transfected and nontransfected HEK-293 cells after the addition of 1 μM SPI-0211 with a final concentration of 1% DMSO. Current recordings are shown in Fig. 6, A and B; I-V relationships normalized to cell capacitance are shown in Fig. 6C; and the slope conductance is shown in Fig. 6D. Nontransfected HEK-293 cells exhibited a low Cl channel conductance of 0.016 ± 0.033 nS/pF (n = 4), which did not change significantly after addition of SPI-0211 (0.035 ± 0.014 nS/pF; n = 4) (Fig. 6D). ClC-2-transfected HEK-293 cells showed a low basal Cl channel conductance (0.196 ± 0.017 nS/pF, n = 3), which significantly increased (P < 0.001) with the addition of 1 μM SPI-0211 (Fig. 6D), to 1.820 ± 0.114 nS/pF (n = 3). Cl currents in SPI-0211-treated cells were not further increased by subsequent addition of 5 μM forskolin/20 μM IBMX or by subsequent addition of 1 μM arachidonic acid. There was no effect of 1% DMSO on basal control currents. The normalized slope conductance was 0.06 ± 0.1 nS/pF (n = 3) in the presence or absence of 1% DMSO. Arachidonic acid alone (1 μM in 1% DMSO) increased ClC-2 channel conductance to 0.694 ± 0.058 nS/pF (n = 3). These results indicate that 1 μM SPI-0211 alone fully activated ClC-2 Cl channels in the HEK-293 cells. To examine whether PKA was mediating the activating effect of SPI-0211, the effect of mPKI, the cell-permeant PKA inhibitor, on SPI-0211 activated ClC-2 Cl channel currents was investigated. As shown in Fig. 7, mPKI did not inhibit the effect of SPI-0211, which was at a maximum because no further activation with arachidonic acid was observed. However, mPKI did prevent activation by forskolin/IBMX (Fig. 7B) as previously shown (38). Therefore, SPI-0211 effects on ClC-2 Cl channels are not mediated by PKA. The effect of increasing concentrations of SPI-0211 on Cl channel currents in ClC-2-transfected HEK-293 cells was then examined. These studies were designed such that each cell would be treated with progressively increasing SPI-0211 concentrations. Preliminary studies (data not shown) indicated that accumulating DMSO increased EC50 without affecting the maximum velocity. Thus DMSO concentrations were minimized and kept constant by preparing each SPI-0211 concentration to yield a final DMSO concentration of 0.1%. This was accomplished by washing three times with 2 ml of medium before addition of the next higher SPI-0211 concentration. This procedure was used because, as shown in Fig. 8, AC, SPI-0211 could be washed out, and then the same effect was observed upon reapplying the same concentration. This result also indicates that SPI-0211 effects are reversible. In Fig. 9, the change in ClC-2 Cl channel current normalized to capacitance is plotted as a function of SPI-0211 concentration. The modified Michaelis-Menten hyperbolic plot was used to fit the data using the equation ΔI = ΔImax·[SPI-0211]/(EC50 + [SPI-0211]). The analysis is constricted to 0, because ΔI was measured. Lineweaver-Burk and Hill plots of the data are also plotted as insets in Fig. 9. Table 2 shows that the EC50 and ΔImax maximum predicted change in current (or transport rate) obtained from both hyperbolic and Lineweaver-Burk plots were not significantly different from each other: EC50 = 24.38 ± 5.80 nM and ΔImax = 56.99 ± 3.97 pA/pF (n = 7) from the hyperbolic plot, and EC50 = 17.27 ± 0.53 nM and ΔImax = 50.78 ± 2.66 pA/pF (n = 3) from the Lineweaver-Burk plot. χ2 = 15.80, indicating that the hyperbolic plot fit the data very well (P < 0.02). The Hill coefficient was 0.71 ± 0.06 (n = 5). These values are similar to those obtained in the Isc experiments shown in Fig. 6B and Table 1. Based on a Hill coefficient of 1, the modified Michaelis-Menten hyperbolic plot is the appropriate analysis to use, and the simple bimolecular first-order kinetic model is correct.

Fig. 6.

Effect of 1 μM SPI-0211 on Cl channel currents in ClC-2-transfected and nontransfected human embryonic kidney (HEK)-293 cells. A: representative whole cell Cl currents in HEK-293 cells stably transfected with recombinant human ClC-2 Cl channels without [control (C)] and with 1 μM SPI-0211 treatment and subsequently treated with 5 μM forskolin/20 μM IBMX (F/I) and then with 1 μM arachidonic acid (AA). B: representative whole cell patch-clamp Cl currents in nontransfected HEK-293 cells before and after treatment with 1 μM SPI-0211. C: current-voltage (I-V) curves for the Cl currents shown in A normalized to cell capacitance (proportional to cell membrane area) and expressed as pA/pF. Data are means ± SE (n = 3) of cells with no treatment (control; ▪) and cells subsequently treated with 1 μM SPI-0211 (•), with 5 μM F/I (▴), or with 1 μM AA (⧫). I-V curves for nontransfected cells are not shown. D: summarized data for normalized slope conductance (nS/pF) for data presented in AC. Data are means ± SE (n = 3). *P < 0.001 vs. control. #P < 0.01 vs. control.

Fig. 7.

Effect of myristoylated protein kinase inhibitor (mPKI) on SPI-0211-activated Cl channel conductance in ClC-2-transfected HEK-293 cells. The normalized slope conductance (nS/pF) is shown. Data are means ± SE (n = 3). A: effect of sequential addition of SPI-0211, F/I, and AA in the absence and presence of mPKI. B: effect of mPKI followed by F/I and then SPI-0211 compared with the effect of mPKI followed by SPI-0211 and then F/I. mPKI, 0.4 μM; SPI-0211, 1 μM; F/I, 5 μM/20 μM; AA, 1 μM. *P < 0.001 with respect to control. #P < 0.01 with respect to control. ##P < 0.05 with respect to control.

Fig. 8.

Reversibility of SPI-0211 effect on Cl currents in ClC-2-transfected HEK-293 cells. A: representative Cl currents in HEK-293 cells stably transfected with ClC-2 determined by patch clamp before (c) and after addition of 100 nM SPI-0211 in 0.1% DMSO. After washing with 3 changes of medium, I recordings were again obtained before and after 100 nM SPI-0211. B: I-V relationships for data shown in A plotted as means ± SE (n = 3). Currents are normalized to cell capacitance. ▪, control; ○, first SPI-0211; ▴, wash; ▾, second SPI-0211. C: summarized I at −140 mV plotted as means ± SE (n = 3).

Fig. 9.

Effect of varying SPI-0211 concentration on Cl currents in ClC-2-transfected HEK-293 cells. SPI-0211 concentration is plotted against the change in ClC-2 Cl channel current, ΔI (pA/pF), measured by patch clamp at −140-mV holding potential, normalized to cell capacitance, and expressed as means ± SE (n = 3). After each addition of SPI-0211 in 0.1% DMSO, 3 changes of medium were performed to wash away SPI-0211 and DMSO. The cell was then treated with the next higher concentration of SPI-0211 in 0.1% DMSO. Data were fit with a modified Michaelis-Menten hyperbolic plot using Origin 5.0 software. Lineweaver-Burk and Hill plots are shown as insets. All parameters calculated are in Table 2.

View this table:
Table 2.

Kinetic analysis of ΔI measured as Cl channel current by patch clamp of human ClC-2 expressed in HEK-293 cells in response to varying concentrations of SPI-0211

The above studies with recombinant ClC-2 not only suggested that ClC-2 is a target for SPI-0211 but also strongly suggested that the SPI-0211 effects observed with T84 cells could be mediated by effects on ClC-2. However, because T84 cells also contain CFTR, the possibility existed that CFTR might also be a target for SPI-0211. To evaluate this possibility, effects of SPI-0211 on whole cell Cl currents of HEK-293 cells stably transfected with CFTR rather than ClC-2 were examined. Figure 10 demonstrates that SPI-0211 did not increase the CFTR-mediated Cl channel current or conductance. In contrast, and as expected, forskolin/IBMX significantly increased CFTR-mediated Cl channel current and conductance. Therefore, CFTR does not appear to be a target of SPI-0211.

Fig. 10.

Effect of SPI-0211 on Cl channel currents in CFTR-transfected HEK-293 cells. Whole cell Cl currents were measured in CFTR-transfected HEK-293 cells by patch clamp. A: I-V curves for Cl currents normalized to cell capacitance: cells with no treatment (control, C, ▪), cells subsequently treated with 1 μM SPI-0211 (•), and then with F/I (▴). B: summarized data for the normalized slope conductance (nS/pF) of the data presented in A. Data are means ± SE (n = 3). *P < 0.05 vs. control.

DISCUSSION

The central hypothesis tested in this study was whether an intestinal ClC-2 Cl channel was responsible for the increased intestinal Cl observed in SPI-0211 treated animals (41) and for the beneficial effects of SPI-0211 seen in human clinical trials (18, 19). mRNA for ClC-2 was present in the adult rabbit intestine as measured by RT-PCR, Northern blot analysis, and in situ hybridization. The human intestinal T84 cell line was shown to contain ClC-2 uniformly in the apical membrane when grown to confluence on permeable supports. Isc measurements were used to determine EC50 for SPI-0211 in T84 cells. HEK-293 cells stably expressing human ClC-2 have been used in our laboratory for a number of years (8, 38), and HEK-293 cells stably expressing CFTR were developed for the present study. SPI-0211 did not activate CFTR-mediated Cl conductance but was shown, using whole cell patch clamp, to activate ClC-2 with an EC50 similar to that observed with activation of Isc (apical Cl conductance) in T84 cells. SPI-0211 activation of ClC-2 was shown to be independent of PKA because mPKI had no effect.

T84 cells have been used widely as a model system for the study of intestinal Cl transport (1, 2, 37, 39). ClC-2 has been studied by patch clamp in T84 cells grown on plastic (1). Previous localization studies suggested that ClC-2 was present in the apical membrane of T84 cells grown on glass slides, but only in a small population of such cells (23). However, in the present studies of T84 cells grown to confluence on permeable supports, the ClC-2 Cl channel protein was expressed in the apical membrane and in virtually all cells. Therefore, these T84 cells can be used to study Cl transport by the Isc technique. Isc was studied in intact T84 cells and then in cells with a basolateral-to-apical Cl gradient and the basolateral membrane permeabilized by nystatin. Under the latter conditions, Cl conductance across the apical membrane is directly proportional to Isc (24). As we have shown, transepithelial Cl transport in T84 cells is potently activated by SPI-0211. This effect is due to increased Cl transport across the apical and not the basolateral membrane. CFTR is also present in these cells (39). When correlated with the effect of SPI-0211 on recombinant ClC-2 and CFTR, the activation of Cl transport in T84 cells by SPI-0211 strongly suggests that non-CFTR Cl channels (likely ClC-2) in the apical membrane were responsible. Potent and selective inhibitors of ClC-2 do not exist, leaving open the question whether some non-CFTR Cl channel other than ClC-2 is also a target of SPI-0211 in T84 cells.

Effects of SPI-0211 on ClC-2 and CFTR were tested in HEK-293 cells transfected with recombinant human ClC-2 or CFTR Cl channels using whole cell patch clamp. HEK-293 cells have a low level of endogenous Cl currents (8, 13, 38) and have been used widely to study recombinant Cl channels (8, 13, 38) using patch-clamp techniques (17). In patch-clamp studies, it was demonstrated that SPI-0211 strongly stimulated ClC-2 channel currents but had no effect on CFTR channel currents. There was good agreement between the EC50 in T84 Isc studies and in ClC-2-expressing HEK-293 patch-clamp studies, providing evidence that ClC-2 may be the target of SPI-0211 in T84 cells. Because ClC-2 is also known to be present in the human small and large intestines (23), these results were consistent with the therapeutic benefits of SPI-0211 being the result of activation of ClC-2-mediated Cl transport in intestinal epithelia.

In most studies, the extent of ClC-2 activation by SPI-0211 in ClC-2-expressing HEK-293 cells was assessed by comparison with other well-established activators of ClC-2, such as forskolin/IBMX or arachidonic acid (8, 38). When cells were treated with forskolin/IBMX and arachidonic acid after treatment with SPI-0211, ClC-2 activation levels were similar to those with SPI-0211 alone, indicating that all of these treatments activate the same protein, namely, ClC-2, and that SPI-0211 resulted in maximum activation of ClC-2. Experiments with mPKI evaluated the role of PKA in mediating ClC-2 activation by SPI-0211. In contrast to forskolin/IBMX, SPI-0211 activation of ClC-2 did not require PKA, as previously also shown to be the case with arachidonic acid (8, 38).

These studies show that SPI-0211 is a potent activator of ClC-2 Cl channels expressed in HEK-293 cells and Cl transport in T84 cells. No effects were observed on CFTR in these systems. Therefore, SPI-0211 activation of Cl transport in the human intestine (41) may occur through activation of the ClC-2 Cl channel. These properties suggest that SPI-0211 is an excellent candidate for the clinical treatment of many gastrointestinal syndromes. These studies also suggest that ClC-2 Cl channels may play a physiological role in Cl transport in the intestine. Further studies are needed to elucidate the mechanism by which SPI-0211 activates ClC-2 Cl channels.

GRANTS

The laboratory at the University of Cincinnati acknowledges support from National Institutes of Health Grants DK-43816 and HL-58399 and United States Army Research Office Multidisciplinary University Research Initiative Grant DAAD-19-02-1-0227 (to J. Cuppoletti and D. H. Malinowska). Functional studies were supported by a grant from Sucampo Pharmaceuticals, Inc. (to J. Cuppoletti).

DISCLOSURES

J. Cuppoletti, M. L. Patchen, and R. Ueno have financial interests in Sucampo Pharmaceuticals, Inc.

J. Cuppoletti serves as a consultant to, R. Ueno is an owner of, and M. L. Patchen is a former chief executive officer of Sucampo Pharmaceuticals, Inc.

Acknowledgments

Isc measurements using the Isc chamber from World Precision Instruments were performed by Venkataraman Muthiah-Nakarajan.

Footnotes

  • 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.

REFERENCES

View Abstract