GASTROENTEROLOGY 2005;128:695–707
Deoxycholic Acid Activates Protein Kinase C and Phospholipase C Via Increased Ca2ⴙ Entry at Plasma Membrane BONNIE W. LAU,* MATILDE COLELLA,*,‡ WARREN C. RUDER,* MARIANNA RANIERI,‡ SILVANA CURCI,* and ALDEBARAN M. HOFER* *Boston VA Healthcare System and the Department of Surgery, Harvard Medical School, Brigham and Women’s Hospital, West Roxbury, Massachusetts; and ‡Dipartimento di Fisiologia Generale ed Ambientale, Universita’ di Bari, Bari, Italy
Background & Aims: Secondary bile acids like deoxycholic acid (DCA) are well-established tumor promoters that may exert their pathologic actions by interfering with intracellular signaling cascades. Methods: We evaluated the effects of DCA on Ca2ⴙ signaling in BHK-21 fibroblasts using fura-2 and mag-fura-2 to measure cytoplasmic and intraluminal internal stores [Ca2ⴙ], respectively. Furthermore, green fluorescent protein (GFP)-based probes were used to monitor time courses of phospholipase C (PLC) activation (pleckstrin-homology [PH]-PLC␦GFP), and translocation of protein kinase C (PKC) and a major PKC substrate, myristolated alanine–rich C-kinase substrate (MARCKS). Results: DCA (50 –250 mol/L) caused profound Ca2ⴙ release from intracellular stores of intact or permeabilized cells. Correspondingly, DCA increased cytoplasmic Ca2ⴙ to levels that were ⬃120% of those stimulated by Ca2ⴙ-mobilizing agonists in the presence of external Ca2ⴙ, and ⬃60% of control in Ca2ⴙ-free solutions. DCA also caused dramatic translocation of PH-PLC␦-GFP, and conventional, Ca2ⴙ/diacylglycerol (DAG)-dependent isoforms of PKC (PKC-I and PKC-␣), and MARCKS-GFP, but only in Ca2ⴙ-containing solutions. DCA had no effect on localization of a novel (PKC␦) or an atypical (PKC) PKC isoform. Conclusions: Data are consistent with a model in which DCA directly induces both Ca2ⴙ release from internal stores and persistent Ca2ⴙ entry at the plasma membrane. The resulting microdomains of high Ca2ⴙ levels beneath the plasma membrane appear to directly activate PLC, resulting in modest InsP3 and DAG production. Furthermore, the increased Ca2ⴙ entry stimulates vigorous recruitment of conventional PKC isoforms to the plasma membrane.
econdary bile acids such as deoxychlolic acid (DCA) are generated via bacterial deconjugation of secreted endogenous bile in the lower intestine. These natural steroid compounds are implicated in the pathogenesis of many disorders of the gastrointestinal tract. For example, it is well documented that DCA can act as a tumor
S
promoter in the colon and esophagus, but precisely how this occurs has not been established.1 In early studies it was assumed that the irritant and cytotoxic effects of bile acids derived solely from their indiscriminate solubilizing actions on biological membranes, when used at micellar concentrations. Recently, it has become apparent that physiologic concentrations of bile acids can affect cellular function in a very complex way, by modifying gene expression and by intersecting with intracellular signaling pathways that control growth, proliferation, and apoptosis.2,3 Several specific molecular targets for bile recently have been identified. For example, certain bile constituents (particularly lithocholic acid and derivatives, but also DCA) are effective ligands for nuclear steroid receptors such as the pregnane X receptor, the farnesoid X receptor, and the nuclear vitamin D receptor, permitting bile acids to control distinct patterns of gene expression.4 – 6 DCA also causes ligand-independent transactivation of epidermal growth factor (EGF) receptors, leading to signaling through mitogen-activated protein kinase pathways.7,8 In addition, there is compelling functional evidence indicating that secondary bile compounds are linked to stimulation of cyclic adenosine 3=,5=-cyclic monophosphate signaling,9,10 which may explain in part how bile can provoke inappropriate fluid secretion in gastrointestinal epithelia. Recently, Maruyama et al11 and Kawamata et al12 independently identified 2 novel mammalian G-protein– coupled receptors activated by DCA and other bile acids at low micromolar concentraAbbreviations used in this paper: 2-APB, 2-aminoethoxydiphenyl borate; BK, bradykinin; DAG, diacylglycerol; DCA, deoxycholic acid; GFP, green fluorescent protein; InsP3, inositol 1,4,5-trisphosphate; MARCKS, myristolated alanine–rich C-kinase substrate; PH, pleckstrinhomology; PIP2, phosphatidylinositol 4,5-bisphosphate; PKC, protein kinase C. © 2005 by the American Gastroenterological Association 0016-5085/05/$30.00 doi:10.1053/j.gastro.2004.12.046
696
LAU ET AL
tions. These G-protein– coupled receptors were coupled to cyclic adenosine 3=,5=-cyclic monophosphate generation via Gs, but apparently were not linked to intracellular Ca2⫹ signaling cascades. It has long been recognized that treatment of cells with DCA and other bile constituents can cause increases in intracellular [Ca2⫹],13–16 although details of the molecular mechanism(s) underlying this phenomenon still are not entirely clear. Ca2⫹-mobilizing hormones and neurotransmitters typically work by binding to G-protein– coupled receptors that couple to phospholipase C (PLC) (eg, via Gq). Alternatively, other types of agonists, such as growth factors, activate tyrosine receptor kinases linked to PLC␥, such as the EGF receptor.17 Receptor occupancy permits hydrolysis of the membrane lipid PIP2 through PLC, resulting in the generation of InsP3 and dicylglycerol (DAG). DAG is an important cofactor for the activation of many isoforms of protein kinase C (PKC), whereas InsP3 is a potent releaser of internal Ca2⫹ stores. Emptying of internal stores, in turn, results in the opening of capacitative Ca2⫹ entry pathways in the plasma membrane, also known as storeoperated channels.18 Early observations established that specific bile acids increase cytoplasmic [Ca2⫹] ([Ca2⫹]cyt) by releasing intracellular [Ca2⫹] stores.19,20 In pancreatic acinar cells, taurolithocholic acid 3-sulfate has been shown to elicit Ca2⫹ signals with many features typical of those stimulated by native Ca2⫹ mobilizing hormones, such as oscillations and polarized signaling.21 DCA and other bile acids also are known to activate PKC (as measured by biochemical assays) in a variety of cell systems, providing an attractive explanation for the tumor-promoting properties of certain bile acids.22–24 DCA-stimulated Ca2⫹ and PKC signaling can produce other significant shortterm alterations in cellular physiology. For example, in colonic fibroblasts, the increase in [Ca2⫹]cyt and PKC activation stimulated by DCA were shown to elicit the production of the inflammatory mediator prostaglandin E2 via inducible cyclooxygenase-2 enzymes.25 In light of the potential pathophysiologic implications of these downstream effects, it is of immediate interest to identify precisely how DCA activates PKC and Ca2⫹ signaling pathways. In the present study, we used imaging techniques and a panel of fluorescent biosensors to explore how DCA impacts Ca2⫹ signaling cascades and PKC activation in single living BHK-21 fibroblasts. Although carcinoma arises from epithelial cells, the neighboring stromal cells, including fibroblasts, now are firmly established to be active participants in the development of cancer and in the invasive growth of tumors,26 including gastric tu-
GASTROENTEROLOGY Vol. 128, No. 3
mors.27 It therefore is also important to examine the actions of bile acids in stromal cell types if a complete picture of the process of tumor development in the gastrointestinal tract is to be obtained. Our measurements of intracellular [Ca2⫹] and intraluminal endoplasmic reticulum (ER) Ca2⫹ using compartmentalized low-affinity indicators confirmed that DCA caused release of internal stores and Ca2⫹ entry in this cell type. Real-time measurements of GFP-tagged PKCs showed that DCA treatment led to the redistribution of 2 conventional isoforms of PKC (PKC1 and PKC␣) with a pattern similar to that stimulated by bradykinin (BK) and adenosine triphosphate (ATP). These isoforms require both Ca2⫹ and DAG for their translocation (which in turn reflects activation of the kinase). However, DCA did not cause appreciable translocation of the novel isoform, PKC␦ (which depends only on DAG for activation), or the atypical isoform, PKC-, which is DAGand Ca2⫹-independent. In addition we show here that myristolated alanine–rich C-kinase substrate (MARCKS), a major substrate of PKC that undergoes translocation only on phosphorylation (eg, by PKC),28,29 also redistributed after DCA treatment in a similar manner to that triggered by agonist. Given the similarities between the actions of DCA on intracellular Ca2⫹ signaling and those of authentic Ca2⫹ mobilizing agonists, we considered the possibility that DCA might cause activation of PLC, perhaps by binding to a cell surface receptor. By using a genetically encoded GFP-based reporter of PLC activation, PH-PLC␦EGFP,30 we found that DCA, similar to agonist, stimulated PIP2 hydrolysis. The subsequent production of InsP3 and DAG would be expected to cause release of InsP3-sensitive internal Ca2⫹ stores, and explain the redistribution of conventional PKCs. However, further investigation revealed that the action of DCA on PKC and PLC differed from those of agonist in several ways, and were dependent strictly on Ca2⫹ entry from the extracellular space. Our data are consistent with a model in which Ca2⫹ entry initiated by DCA (possibly through an ionophore-like mechanism) can by itself drive the activation of downstream signaling elements (PLC, PKC, and MARCKS) normally associated with the stimulation of G-protein– coupled receptors or receptor tyrosine kinase– coupled Ca2⫹ signaling events.
Materials and Methods Cell Culture BHK-21 and Caco-2 cells were grown in Dulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum and .5% antibiotics (penicillin/streptomycin), and were
March 2005
maintained in a humidified incubator at 37°C in the presence of 5% CO2/95% air. Subcultures were prepared by trypsinization, and the cells were not used for more than 10 passages after receipt from the distributor. Cells were seeded at low density on glass coverslips and used the following day for imaging measurements of cytoplasmic or intraorganellar free [Ca2⫹] with fura-2-AM, or mag-fura-2-AM, respectively. Alternatively, cells were transfected with GFP reporters 24 hours after plating and used the next day for fluorescence imaging studies.
Fluorescence Imaging Experiments Calcium imaging. Intracellular [Ca2⫹] were measured using the ratiometric indicator, fura-2.31 Measurements of [Ca2⫹] within the agonist- and InsP3-sensitive internal store in intact or digitonin-permeabilized cells were obtained by using compartmentalized mag-fura-2, as described previously.32–34 Cells were loaded in tissue culture medium at 37°C for 20 minutes with fura-2-AM (7 mol/L) or for 35– 45 minutes with mag-fura-2-AM (2–5 mol/L). When cells were preloaded with BAPTA-AM, the chelator was added during the last 20 minutes of dye loading to ensure cytosolic retention (final concentration, 40 mol/L). Fura-2 and mag-fura-2 ratios were acquired using a ratio imaging set-up running Metafluor software (Universal Imaging, West Chester, PA). Coverslips with dye-loaded cells were mounted in an open-topped perfusion chamber (Series 20; Warner Instrument Corporation, Hamden, CT) and placed on the heated stage of a Nikon TE200 inverted microscope (Melville, NY). Cells were excited alternately at 340 nm and 380 nm for 80 ms through a 40⫻ (numerical apertune [NA], 1.4) oil immersion objective. The excitation wavelengths were generated using a microprocessor controlled filter wheel (Sutter Instruments, Novato, CA) placed in the path of a 100-W mercury light source. Pairs of fluorescence images (emission collected above 510 nm) were captured by a Hamamatsu ORCA ER CCD camera (Bridgewater, NJ) every 4 seconds, and converted to a ratio image by the Metafluor software. Measurement of intraluminal ER [Ca2ⴙ] in permeabilized cells. As described previously,35 mag-fura-2–loaded cells were rinsed briefly in a high K⫹ solution (in mmol/L: 125 KCl, 25 NaCl, .1 MgCl2, 10 Hepes, pH 7.20), and then exposed for 2–3 minutes to an intracellular buffer (the same solution with free Ca2⫹ clamped to 170 nmol/L using Ca2⫹/ ethylene glycol-bis(-aminoethyl ether)-N,N,N=,N=-tetraacetic acid buffers and supplemented with 1 mmol/L Na2ATP) also containing 5 g/mL digitonin at 37°C. After plasma membrane permeabilization, cells were superfused continuously with intracellular buffer (without digitonin). Measurements of mag-fura-2 fluorescence were performed as described earlier for intact cells.
Fluorescence Imaging of GFP-Based Reporters Cells were transfected transiently with GFP-based indicator(s) using Effectene transfection reagent (Qiagen, Valencia,
DCA ACTIVATES PLC THROUGH CA2ⴙ ENTRY
697
CA). Fluorescence images of cells expressing the various probes were acquired every 4 seconds using the Metafluor imaging set-up and the perfusion apparatus described earlier. GFPs and venus (a variant of GFP) were excited at 485 nm, and emission was collected above 530 nm. PKC translocation was measured using enhanced EGFPconjugated PKC chimeras of ␣, 1, ␦, and isoforms.36,37 Plasmids were generous gifts from our colleagues Dr. Sarino Rizzuto and Dr. Paolo Pinton, University of Ferrara, Italy, and Dr. Paulo J. Magalhães (PKC␣-venus), University of Padova, Italy. Synthesis of intracellular InsP3/PLC activation was monitored with a GFP-tagged pleckstrin homology domain of PLC␦1 (PH-PLC␦1-EGFP),30,38,39 a kind gift from Dr. Stephen S. G. Ferguson, London, Ontario, Canada. Activation of MARCKS was monitored by following translocation of an EGFP-tagged MARCKS, a key target of PKC phosphorylation.28 This probe was graciously supplied by Professor Naoki Saito, Kobe University, Japan. Nuclear [Ca2⫹] was measured using the 480/410 nm excitation ratio (emission ⬎510 nm) of nuclear-targeted ratiometric pericam,40 a kind gift from Dr. Atsushi Miyawaki, RIKEN, Japan. The nuclear localization of this construct permitted simultaneous measurement of PLC activation (with PHPLC␦1-EGFP, which is excluded from the nucleus) and nuclear Ca2⫹ changes in the same cell. In some experiments, PH-PLC␦1-EGFP– or PKC-GFP– expressing cells were loaded with fura-2 for simultaneous measurement of intracellular [Ca2⫹] and PLC/PKC activation. Cells were excited alternately at 340, 380, and 485 nm through the GFP filter set (dichroic 505 nm DRLP, emission 530 nm DF30), permitting capture of a portion of the fura-2 signal. This allowed for qualitative assessment of the time course of intracellular [Ca2⫹] changes in the same cell without interfering with the acquisition of the GFP channel.
Solutions and Materials Unless otherwise stated, all chemicals were purchased from Sigma (St. Louis, MO). Experiments were performed with a Ringer’s solution containing (in mmol/L): 121 NaCl, 2.4 K2HPO4, .4 KH2PO4, 1.2 CaCl2, 1.2 MgCl2, 5.5 glucose, 10 HEPES/NaOH, pH 7.40. Bradykinin and ionomycin were from Calbiochem-Novabiochem Corp. (La Jolla, CA); InsP3, fura-2-AM, mag-fura-2-AM, and BAPTA-AM were obtained from Molecular Probes (Eugene, OR). When dimethyl sulfoxide or ethanol was used as a solvent, the final solvent concentration never exceeded .01% or .1%, respectively. Because in our experience the response to agonist stimulation in BHK-21 cells can be inconsistent, ie, some cells respond only to bradykinin whereas another subset of cells may respond only to ATP, we used these agonists together to ensure that a response always was obtained for our control agonist stimulation.
Data Analysis Responses to DCA and Ca2⫹ mobilizing agonists were compared in the same cell, wherever possible. This approach
698
LAU ET AL
Figure 1. DCA-stimulated intracellular [Ca2⫹] changes as measured by fura-2 in single BHK-21 cells. Response of fura-2–loaded BHK-21 cells to the Ca2⫹-mobilizing agonists, BK (100 nmol/L) and ATP (100 mol/L) compared with [Ca2⫹] increase elicited by DCA (250 mol/L). Changes in the fura-2 ratio in (A) Ca2⫹-containing solutions and in (B) Ca2⫹-free solutions.
eliminated concerns about the variability of the starting fura-2 or mag-fura-2 ratio, or the baseline fluorescence of the GFPbased probes. Where appropriate, paired data were assessed for statistical significance using the Student t test. Data are expressed as means ⫾ SEM with n equal to the number of experimental runs.
Results The flat morphology of BHK-21 fibroblasts facilitates imaging of the translocation of GFP-based probes using conventional (nonconfocal or non-total internal reflection fluorescence) fluorescence imaging methods. In addition, we and others have used this cell model extensively for measurements of [Ca2⫹] changes in the internal Ca2⫹ store using compartmentalized mag-fura-2, both in intact and permeabilized cells.32,34,35,41 DCA is the predominant secondary bile acid, and can be present at relatively high concentrations (5–500 mol/L) in certain compartments of the alimentary tract. The results shown in Figure 1 confirm that DCA causes increases of intracellular [Ca2⫹] in BHK-21 cells, as has been described previously for a number of other cell models.13,14,42 In the presence of 1 mmol/L extracellular Ca2⫹, 250 mol/L DCA elicited cytosolic Ca2⫹ transients that were slightly but significantly larger (116.88% ⫾ 6.59% of control peak; P ⬍ .03) than those produced by the Ca2⫹-mobilizing agonists BK (100 nmol/ L) and ATP (100 mol/L) in the same cells (Figure 1A; n ⫽ 9 experiments, 72 cells). The response to 250 mol/L DCA was decreased significantly in the absence
GASTROENTEROLOGY Vol. 128, No. 3
of external Ca2⫹ (Figure 1B; 51.2% ⫾ 9.00% of control response to BK and ATP; n ⫽ 9, 72 cells, P ⬍ .001). Similar results were obtained with 350 mol/L DCA, and increases of intracellular [Ca2⫹] could be detected with DCA concentrations as low as 50 mol/L (not shown). These results suggest that DCA was able to release Ca2⫹ from internal stores, but that a significant component of the Ca2⫹ signal was caused by Ca2⫹ entry across the plasma membrane. In separate experiments, the colon carcinoma cell line Caco-2 was used as a model of a cell type directly exposed to secondary bile acids under pathophysiologic conditions. The results (not shown) were qualitatively similar to those obtained with BHK-21 cells. In the presence of 1 mmol/L external Ca2⫹, Caco-2 cells responded to 250 mol/L DCA with an increase of intracellular [Ca2⫹] that was smaller (38.8% ⫾ 2.2% of control peak n ⫽ 17, 107 cells, P ⬍ .001) than that recorded in response to the Ca2⫹-mobilizing agonists ATP (100 mol/L) and carbachol (100 mol/L). As was the case in BHK-21 cells, in Caco-2 cells the response to DCA was decreased significantly in the absence of external Ca2⫹ (25.3% ⫾ 4.5% P ⬍ .05 n ⫽ 8, 51 cells, compared with the decrease observed in 1 mmol/L Ca2⫹). We next examined the Ca2⫹ release process in more detail in intact and permeabilized BHK-21 cells loaded with the low-affinity Ca2⫹ indicator mag-fura-2. We have shown previously that compartmentalized dye monitors [Ca2⫹] changes solely in a thapsigargin- and InsP3sensitive internal Ca2⫹ store in this cell type.35 Figure 2A shows that DCA at concentrations between 50 and 500 mol/L caused a decrease in the intraluminal free [Ca2⫹] as measured by organelle-trapped mag-fura-2 in intact cells (n ⫽ 7). Significantly, the Ca2⫹-releasing action of DCA did not require Ca2⫹ in the extracellular bath. This phenomenon also was observed in digitonin-permeabilized cells (Figure 2B), where DCA concentrations as low as 50 mol/L were able to effectively decrease [Ca2⫹] in the ER lumen (n ⫽ 4). Figure 2C further shows the direct effect of 150 mol/L DCA on intraluminal [Ca2⫹] compared with Ca2⫹ release elicited by 6 mol/L InsP3 (similar results were obtained in n ⫽ 3 independent experiments). The collective results of Figures 2A–C suggest that DCA can cross the plasma membrane and exert direct actions on internal Ca2⫹ stores. We attempted to determine whether DCA enters BHK-21 cells by passive permeation, or through a specific bile acid transporter. Parallel experiments using HPTLC (high performance thin layer chromatography) showed that DCA readily partitions into the membrane fraction of BHK-21 cells (not shown, F. Lopez and M. Colella, unpublished re-
March 2005
DCA ACTIVATES PLC THROUGH CA2ⴙ ENTRY
699
Figure 2. DCA causes the release of intracellular Ca2⫹ stores, as measured by compartmentalized mag-fura-2 in single BHK-21 cells. BHK-21 cells were loaded with mag-fura-2 as described in the Materials and Methods section. (A) Intact cells: BK and ATP cause the expected rapid decrease in the 340/380-nm excitation ratio of the low-affinity Ca2⫹ indicator, mag-fura-2, caused by release of Ca2⫹ from InsP3-sensitive internal stores. Complete recovery occurred on simultaneous washout of agonists and readmission of Ca2⫹ to the bath. Stepwise increases in the concentration of DCA (50 –500 mol/L), also administered in a Ca2⫹-free solution, caused a profound decrease in the ratio. (B, C) Digitonin-permeabilized cells. Examples showing the direct Ca2⫹-releasing action of DCA as measured by ER-trapped mag-fura-2 compared with a maximal dose of InsP3 (6 mol/L) in the same cells. (B) Effect of varying doses of DCA (25, 50, 100, 200 mol/L) on stored Ca2⫹. (C) Effect of 150 mol/L of DCA compared with 6 mol/L of InsP3.
sults). Organic anion transporters represent the major permeation pathway for bile acids in many cell types. Imaging experiments with fura-2 showed that probenecid (100 mol/L), a general organic anion transport inhibitor, did not inhibit the DCA-induced intracellular Ca2⫹ increase, excluding the possibility that DCA can be transported via a probenecid-sensitive carrier (not shown). Furthermore, because some of these carriers are known to be Na⫹-dependent,43 we also performed experiments in Na⫹-free conditions. The results, however, were not conclusive because in the absence of Na⫹, the response to DCA was decreased only partially (77% ⫾ 8.0% of the agonist response, n ⫽ 4, 26 cells). This was, however, significantly smaller (P ⬍ .0001) than the response obtained in control Na⫹-containing Ringer’s solutions (116.88% ⫾ 6.59%; see Results section), meaning that a partial involvement of an Na⫹-coupled DCA transport cannot be excluded. In addition to its direct action on internal store membranes, it appears that DCA also affects Ca2⫹ entry across the plasma membrane. Fura-2–loaded BHK-21 cells were preincubated for 30 minutes with the irreversible Sarco-endoplasmic reticulum calcium ATPase (SERCA) inhibitor, thapsigargin (100 nmol/L), to deplete internal Ca2⫹ stores. Cells then were transferred to Ca2⫹-free
medium, and after ⬃10 minutes, the measurement was begun. Addition of DCA in Ca2⫹-free media under these conditions had no effect on the fura-2 ratio (not shown), indicating that the Ca2⫹ stores released by DCA (eg, as shown in Figure 1B) were thapsigargin sensitive. As shown in Figure 3A, switching from Ca2⫹-free solutions to a solution containing 1 mmol/L Ca2⫹ produced the expected increase in the ratio owing to Ca2⫹ entry via capacitative pathways. Ca2⫹ entry was augmented significantly in the presence of increasing concentrations of DCA (50, 100, 200, 300 mol/L). Because thapsigargin treatment completely eliminated stored Ca2⫹ (no Ca2⫹ release was observed in these cells after ionomycin treatment; not shown), these results indicate that DCA either enhanced capacitative entry, or acted as a carrier for Ca2⫹ across the plasma membrane. We attempted to explore the nature of the DCAinduced Ca2⫹ release and Ca2⫹ entry process using the compound 2-aminoethoxydiphenyl borate (2-APB). 2-APB originally was described as a selective blocker of Ca2⫹ release through InsP3 receptors, but is now recognized to be a relatively nonspecific inhibitor of several different cellular Ca2⫹ transport pathways; in particular, 2-APB is effective in blocking store-operated Ca2⫹ entry in many cell models.44,45 In our experiments in fura-2–loaded
700
LAU ET AL
Figure 3. (A) DCA enhances Ca2⫹ entry in thapsigargin-pretreated cells, as measured by fura-2. Fura-2–loaded BHK-21 cells were preincubated with the irreversible microsomal Ca2⫹-ATPase inhibitor, thapsigargin (100 nmol/L), in Ca2⫹-free solutions before the start of the experiment. At the time shown, 1 mmol/L Ca2⫹ was readmitted to the bath, producing the expected peak and plateau of Ca2⫹ entry through store-operated channels. Stepwise addition of increasing DCA concentrations (50, 100, 200, 300 mol/L) produced further increases in the fura-2 ratio, indicating enhanced Ca2⫹ entry. (B) 2-APB has differential effects on agonist- and DCA-induced Ca2⫹ increases in BHK-21 cells, as measured by fura-2. In 1 set of experiments, the peak response of cells to agonist stimulation (100 nmol/ L BK and 100 mol/L ATP) was compared with the peak in the same cells in the presence of the nonspecific blocker, 2-APB (50 mol/L) (left panel). The inhibitory action of 2-APB on agonist-induced Ca2⫹ increases was dramatic and highly significant (P ⬍ .0001) in the absence of extracellular Ca2⫹ (bars, 0 Ca2⫹), but the effect was barely significant in the presence of Ca2⫹ (bar, 1 Ca2⫹). In contrast, 2-APB produced only a modest inhibition of the DCA response in Ca2⫹-free solutions (right panel; 0 Ca2⫹), as also was the case in Ca2⫹-containing solutions (1 Ca2⫹). Note that the absolute magnitude of the fura-2 ratio changes cannot be compared among the different groups (except for the paired data shown, which were obtained in the same cell) because they are derived from different experimental groups on separate days.
GASTROENTEROLOGY Vol. 128, No. 3
BHK-21 cells, the response to 250 mol/L DCA in Ca2⫹-free solutions was slightly but significantly attenuated in the presence of 50 mol/L 2-APB, as compared with the control response in the same cells (67.8% ⫾ 3.6% of control; data from 94 cells in 4 independent experiments, P ⬍ .01; not shown). At the same time, 2-APB almost completely blocked the response to BK/ ATP in Ca2⫹-free conditions (5.9% ⫾ 2.0% of control; data from 36 cells, n ⫽ 2 experiments, P ⬍ .0001), indicating that this agent is efficacious in preventing the release of InsP3-sensitive intracellular stores. Taken together, these results would suggest that DCA is capable of inducing mobilization of Ca2⫹ stores through pathways in addition to the 2-APB–sensitive intracellular release channels normally activated by agonists. These data are summarized in Figure 3B. We next examined the action of 2-APB in the presence of extracellular Ca2⫹ (Figure 3B). The response to DCA again was decreased partially but significantly in the presence of 2-APB (63.7% ⫾ 3.5% of control Ca2⫹ increase in the same cell; data from 66 cells, n ⫽ 4 experiments; P ⬍ .0001). Interpretation of this result was confounded, however, by the fact that 2-APB only partially inhibited the response to BK and ATP in BHK-21 cells in the presence of Ca2⫹ (75.2% ⫾ 7.06% of control, data from 2 cells, n ⫽ 2 experiments; P ⬍ .002). Therefore, we can only conclude that the plasma membrane Ca2⫹ entry pathways stimulated by either agonist or DCA were not entirely sensitive to 2-APB. Separate experiments using the potent InsP3 receptor inhibitor, heparin, confirmed that DCA directly releases Ca2⫹ stores independent of InsP3 production or InsP3 receptor activation. Heparin (500 g/mL) completely blocked the response to InsP3 (5 mol/L), but did not prevent store release owing to DCA (125 or 250 mol/L) in permeabilized mag-fura-2–loaded BHK-21 cells (not shown; n ⫽ 4 experiments, 4 cells). DCA has been reported to stimulate PKC activity. Three distinct classes of PKC isoforms have been described: conventional, novel, and atypical. Conventional isoforms are distinguished by their dependence on Ca2⫹ and DAG for their activation, whereas the novel isoforms require only DAG. Atypical PKCs are not activated by either Ca2⫹ or DAG. The activation of many PKCs can be assayed indirectly based on their cellular localization.36,37 For example, on stimulation, many conventional and novel isoforms rapidly traffic from cytosol to plasma membrane where they reversibly associate with DAG and lipid constituents. This sometimes is followed by phase of accumulation of the kinase in the nucleus. We used PKC chimeras tagged with fluorescent proteins (EGFP or the variant, venus) transiently expressed in BHK-21 cells
March 2005
DCA ACTIVATES PLC THROUGH CA2ⴙ ENTRY
701
to examine the actions of DCA on a panel of PKC isoforms. This allowed us to make detailed comparisons of the actions of DCA with those of native agonists in living cells in real time. Figure 4A shows the fluorescence of PKC1-EGFP expressed in a BHK-21 cell. At rest, the probe was distributed in the cytoplasm, largely excluded from the nucleus (upper panel, a). On stimulation with either agonists (100 nmol/L BK and 100 mol/L ATP) or with 250 mol/L DCA, PKC1-GFP underwent a rapid translocation to the plasma membrane and partially trafficked to the nucleus, causing depletion in cytoplasmic fluorescence and a slight increase in nuclear fluorescence (lower panel, b). The time courses of the GFP fluorescence changes from the indicated cellular regions (nuclear and cytoplasmic) are shown in the trace below. The magnitude of the translocation (as measured by the change in cytoplasmic fluorescence) in DCA was 140.6% ⫾ 33% of the control response to agonist, but this difference was not statistically significant (22 cells, n ⫽ 16 experiments). Similar results were obtained in experiments with 350 mol/L DCA. DCA caused a very similar pattern of translocation in trials with another conventional PKC isoform, PKC␣venus. In this case, the magnitude of the cytoplasmic fluorescence change was significantly greater than that induced by BK/ATP in the same cells (188% ⫾ 27.4%; data from 40 cells, n ⫽ 15; P ⬍ .002). In contrast, DCA did not appreciably affect the translocation of a novel isoform, PKC␦-EGFP, as compared with the brisk activation induced by the phorbol ester, phorbol 12-myristate 13-acetate (PMA) (100 nmol/L; data from 16 cells, n ⫽ 8 experiments; not shown). In addition, there was no effect of DCA whatsoever on the atypical PKC (17 cells, n ⫽ 13; not shown). These last results provide confir-
4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ Figure 4. DCA causes Ca2⫹-dependent activation of conventional isoforms of PKC, as measured by translocation of GFP-labeled PKCs in living BHK-21 cells. (A) Upper panels show fluorescence images (480 nm excitation) of a representative BHK-21 cell expressing PKC1-EGFP (a) before and (b) during treatment with 250 mol/L DCA. Fluorescence intensities were recorded from the selected regions in the cytoplasm (blue circle) and nucleus (green circle) and plotted vs. time in the lower panel. The time points from the corresponding images are indicated by circles marked a and b on the trace recording. The pattern of translocation of PKC1-EGFP and another conventional PKC isoform (PKC␣-venus; venus is a GFP variant) was similar for agonist and DCA. (B) Fluorescence intensity changes recorded from cytoplasmic and nuclear regions of a PKC␣-venus– expressing during DCA treatment. Regions were selected as shown in the previous panel. This record shows that PKC␣-venus failed to translocate upon DCA treatment in Ca2⫹-free solutions until Ca2⫹ was readmitted to the bath. Similar results were obtained using PKCEGFP (not shown).
702
LAU ET AL
GASTROENTEROLOGY Vol. 128, No. 3
Figure 5. DCA causes Ca2⫹-dependent translocation of a major PKC substrate, MARCKS. Right panel: Fluorescence images of typical BHK-21 cells expressing EGFP-tagged MARCKS (a) before and (b) after treatment with 250 mol/L DCA. Reciprocal changes in fluorescence intensities were recorded from the selected regions indicated in the cytoplasm (blue circle) and plasma membrane (red circle) and plotted vs. time in the lower panel. Left panel: The time points from the corresponding images are indicated by a and b on the trace recording. DCA caused redistribution of EGFP fluorescence similar to that elicited by agonist, but only in the presence of 1 mmol/L Ca2⫹.
mation that DCA was not interfering with GFP fluorescence (eg, because of changes in intracellular pH) or causing other nonspecific cellular actions that would cause PKC to redistribute (ie, gross membrane disruption). In addition, DCA had no effect on BHK cells expressing EGFP alone (14 cells, n ⫽ 6 experiments, not shown), providing additional control for potential artifacts. The effect of DCA on the redistribution of conventional PKCs was strongly Ca2⫹-dependent. After preincubation with the high-affinity Ca2⫹ chelator BAPTAAM, DCA was not able to induce translocation of PKC1-EGFP (typical of 7 cells, n ⫽ 3 experiments). In addition, in the absence of external Ca2⫹, DCA was unable to cause translocation of PKC␣, but was able to do so after subsequent re-addition of 1 mmol/L Ca2⫹ and DCA. A representative example of this result in a PKC␣venus– expressing BHK cell is shown in Figure 4B (typical of responses to DCA observed in 6 of 10 PKC␣venus– expressing cells in 6 experiments). Identical findings were made in cells expressing PKC1-EGFP (7 of 7 cells in 3 experiments).
Translocation of conventional and novel PKCs generally is regarded as a marker of PKC activation. We wanted to confirm whether the DCA-induced translocation of conventional PKCs truly reflected physiologic activation of PKC by looking at the phosphorylation of a downstream target of PKC, MARCKS.28,29,46 At rest, MARCKS is associated with the plasma membrane, bound via its hydrophobic myristate chain and a cluster of basic amino acids containing multiple phosphorylation sites. After phosphorylation, the basic cluster becomes more negatively charged, facilitating its release from the plasma membrane to the cytosol. As shown in Figure 5, DCA caused robust relocation of EGFP-labeled MARCKS with a time course that generally was similar to that induced by agonist treatment with BK and ATP (representative of 6 of 6 cells in 5 experiments). The redistribution caused by DCA required Ca2⫹ in the extracellular bath, consistent with our earlier-described finding that DCA-induced PKC translocation did not occur without Ca2⫹ entry. We took advantage of the recent development of a GFP-based indicator that allowed us to examine the
March 2005
DCA ACTIVATES PLC THROUGH CA2ⴙ ENTRY
703
action of DCA on PLC turnover in real-time in single BHK-21 cells. PH-PLC-EGFP incorporates a GFPtagged PH domain of a particular PLC isoform, PLC␦. The PH domain is the portion of the PLC enzyme that interacts with the phospholipid head group of its substrate, PIP2. At rest, most of the probe is associated with the plasma membrane, revealing a clearly demarcated rim of peripheral fluorescence. On PIP2 hydrolysis, the PH domain relocates to the cytosol because it has substantially higher affinity for InsP3 than for its parent molecule, PIP2.30,47 Another important factor contributing to the loss of association with the membrane is that the substrate becomes depleted at the same time.38 Okubo et al39 showed that InsP3 was both necessary and sufficient for translocation of the PH domain of PLC in living Purkinje neurons. This would suggest that the movement of PH-PLC␦-EGFP from plasma membrane to cytoplasm on stimulation reflects PIP2 hydrolysis, and provides an indirect measure of intracellular InsP3 formation via PLC activation. The response of a typical PH-PLC␦-EGFP– expressing BHK-21 fibroblast to BK and ATP is shown in Figure 6A. Fluorescence intensity changes (measured at 480 nm) in the indicated membrane and cytoplasmic regions of the cell are shown in the corresponding trace. In these experiments, cells also were cotransfected with an indicator for nuclear [Ca2⫹], nuclear ratiometric pericam. The 480/410 nm excitation ratio of this sensor provides a measure of nuclear-free [Ca2⫹].40 These 2 probes exhibit discrete cellular localizations, and the PH-PLC␦-EGFP
4™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™™ Figure 6. DCA treatment is associated with activation of PLC. (A) Simultaneous measurement of PLC activation, as measured by translocation of PH-PLC␦-EGFP, and nuclear [Ca2⫹] using nuclear-targeted ratiometric pericam, as described in the Materials and Methods section. Left images (a–e): Fluorescence of a representative cell excited at 480 nm, showing primarily PH-PLC␦-EGFP in the periphery of the cell. (a) At rest, EGFP fluorescence was confined to the outer margins of the cell because the probe was associated mostly with the plasma membrane. (b) After treatment with BK/ATP or (d) 250 mol/L DCA, there was dramatic redistribution of the sensor from membrane to cytoplasm. The reciprocal changes in fluorescence intensities were recorded from the selected regions indicated in the cytoplasm (blue circle) and plasma membrane (red circle) and plotted vs. time in the right panel. The time points from the corresponding images are indicated by a through e on the trace. The 480/410 excitation ratio of the Ca2⫹ indicator nuclear ratiometric pericam (green trace) was acquired from a region comprising the nucleus, the fluorescence of which was demarcated clearly when observed at 410 nm (inset in lower panel, 410 nm; see text for details). The trace shows that translocation of PH-PLC␦-EGFP only occurred when Ca2⫹ was present in the external bathing medium. (B) Expanded view of a similar experiment as the one shown in A, showing the strict dependence of the translocation process on Ca2⫹ entry. Although there was a transient increase in [Ca2⫹] after DCA treatment as measured by the nuclear ratiometric pericam, no translocation occurred until Ca2⫹ was present outside the cell.
704
LAU ET AL
has extremely low fluorescence when excited at 410 nm. Meanwhile, 410 nm is the optimal excitation wavelength for nuclear ratiometric pericam (see Figure 6A, inset, 410 nm), so it was possible to monitor intranuclear [Ca2⫹] (which closely approximates the Ca2⫹ signals recorded from the cytoplasm) and PH-PLC␦-EGFP translocation at the same time. As shown in Figure 6A, stimulation with DCA caused robust redistribution of PH-PLC␦-EGFP that was qualitatively and quantitatively very similar to that produced by agonist. In paired experiments in which cells were stimulated with BK/ ATP followed by 250 –350 mol/L DCA (or vice versa), the DCA response (as measured by the increase in cytoplasmic fluorescence intensity) was 152.5% ⫾ 20.6% of agonist response (not significant, data from 21 cells; 17 experiments). However, as the experiment in Figure 6A shows, movements of the probe occurred only when Ca2⫹ was present outside the cell (typical of 12 of 15 cells, 12 experiments). It is noteworthy that in separate experiments it was possible to observe the transient translocation of the biosensor when cells were stimulated with BK and ATP in Ca2⫹-free solutions (not shown; observed in 17 of 23 cells, 7 experiments). This latter result is consistent with the well-known ability of agonists to release intracellular InsP3-sensitive stores in Ca2⫹-free solutions.34 An expanded recording that clearly shows the strict dependence of this translocation phenomenon on Ca2⫹ entry, and not release of Ca2⫹ from internal stores, is shown in Figure 6B. Addition of DCA in Ca2⫹-free solutions gave the expected slow transient of intracellular Ca2⫹, as measured by nuclear ratiometric pericam in the same cell (see Figure 1). However, measurable translocation of PH-PLC␦-EGFP did not occur until Ca2⫹ was readmitted to the external solutions, on which a dramatic relocation of the probe was observed. Similar responses were obtained in cells that were coloaded with fura-2 (to measure intracellular Ca2⫹; n ⫽ 4 experiments, 13 of 15 cells), and also in cells that expressed only the PH-PLC␦EGFP biosensor (n ⫽ 2 experiments, 2 cells). The translocation of the probe in DCA or agonist-treated cells was not apparent when cells were preincubated with BAPTAAM, to prevent cytoplasmic Ca2⫹ increases (not shown; data from 6 experiments, 25 cells). We attempted to explore further the actions of DCA on PIP2 turnover as measured by the translocation of the PH-PLC␦-EGFP probe using the PLC inhibitor U73122. We found, however, that this drug by itself (used at 25–50 mol/L) caused morphologic changes to the cell, and a pattern of redistribution of PH-PLC␦EGFP that was distinct from that caused by agonist or DCA (data not shown). Furthermore, use of this drug is
GASTROENTEROLOGY Vol. 128, No. 3
confounded by the fact that it appears to potently release internal stores in BHK-21 cells, as measured by compartmentalized mag-fura-2 (Colella and Ranieri, unpublished data). This could explain fully the inhibitory actions of U73122 on agonist- and DCA-induced Ca2⫹ increases that we observed in our studies (data not shown).
Discussion The disruptive effects of bile acids on structure and function of cells in the gastrointestinal tract (hepatocytes, pancreatic, gastric, and intestinal cells of the underlying stroma) have been documented extensively.1 Although healthy organisms are able to protect themselves from the potential cytotoxicity of bile acids, for example, by conjugating them to glycine or taurine, and then by transporting them out of cells, there are a number of pathologic conditions (cholestasis, metabolic disorders that increase the levels of bile acids, bile acid malabsorption, chronic reflux of gastric fluids, and so forth) in which unconjugated, therefore hydrophobic, bile acids can exert cytotoxic actions. Under these conditions, extracellular and/or intracellular concentrations of bile acids can reach abnormally high levels. These compounds then become relevant causative agents in a number of diseases such as hepatitis, chronic liver diseases, pancreatitis, gastroesophageal reflux disease, and so forth.1 In addition, a vast literature documents that the hydrophobic and highly cytotoxic DCA and chenodeoxycolic acid play a key role in colorectal carcinogenesis,48 as well as apoptosis in colon cancer cell lines.49 Bile acids have been implicated in the development of many other types of cancer, including esophageal adenocarcinoma associated with chronic gastric reflux.50 Despite the clear clinical significance of these agents, relatively little is known about the basic mechanisms that underlie bile acid injury. In the present study we took advantage of the current availability of sensitive GFP-based probes to examine how a particular bile acid, DCA, intersects with Ca2⫹ signaling pathways and the signaling cascades downstream of this second messenger in single living cells. This permitted us to compare the time courses and dynamic properties of DCA-induced PKC translocation and PLC activation with those elicited by native Ca2⫹ mobilizing agonists in the same cell. The most remarkable finding of this study was that DCA was able to activate PLC effectively, with an efficacy similar to that of BK and ATP, as visualized by the cellular redistribution of PH-PLC␦-EGFP. On first inspection, this finding was highly suggestive of DCA working through a G-
March 2005
protein– or tyrosine kinase– coupled cell surface receptor, or some other nonreceptor protein kinase system that is able to activate PLC (see Rhee and Bae for review51). However, the actions of DCA also can be explained readily by a receptor-independent mechanism whereby entering Ca2⫹ directly activates PLC. Our data are consistent with DCA causing the following sequence of events. Initially, DCA precipitates the release of intracellular Ca2⫹ stores through a direct action of this bile compound on the store membrane, as shown by our experiments in permeabilized mag-fura-2–loaded cells (Figure 2B and C). These experiments also show that DCA caused depletion of intraluminal Ca2⫹ in intact cells in the absence of extracellular Ca2⫹ (Figure 2A), presumably by crossing the plasma membrane, and acting on the ER in a fashion similar to that observed in the permeabilized cells. Concomitantly, DCA directly facilitates the admittance of Ca2⫹ across the plasma membrane, as shown by our fura-2 experiments in thapsigargin-treated cells (Figure 3A). Meanwhile, the release of internal stores activates store-operated Ca2⫹ channels, which together with the Ca2⫹ entry elicited directly by the DCA acting at the cell surface, cause local increase of Ca2⫹ under the plasma membrane. Our data further show that this Ca2⫹ is sufficient to stimulate PIP2 hydrolysis directly, as evidenced by the translocation of PH-PLC␦1-EGFP, with the attendant generation of InsP3 and DAG. Preventing the intracellular Ca2⫹ increase with BAPTA abolished this response. Although our experiments using the inhibitor 2-APB (Figure 3) would suggest that a portion of the store release stimulated by DCA is independent of InsP3, the additional InsP3 produced by PLC after Ca2⫹ entry nevertheless likely contributes to further release of stores. This would result in an even greater activation of capacitative Ca2⫹ entry, further stimulation of PIP2 hydrolysis, and so forth. Thus, the initial slow admission of Ca2⫹ into the cell triggered by DCA establishes a cycle that amplifies the Ca2⫹ release and entry processes. The question that remains unanswered is precisely how DCA is able to cause the initial Ca2⫹ release and Ca2⫹ entry. Kim et al52 provided evidence that a mixture of bile acids applied to pancreatic acinar cells could inhibit SERCAs. This action required cellular uptake of the bile compounds by specific bile acid transporters. This potential mechanism and an alternative one in which DCA interacts with cell surface receptors coupled to some highly Ca2⫹-dependent PLC isoforms different from the enzyme linked to BK/ATP stimulation cannot be excluded completely by the present data. It seems, however, that the most straightforward interpretation of our findings is that DCA enters the cell by some means
DCA ACTIVATES PLC THROUGH CA2ⴙ ENTRY
705
and exerts its actions on the ER membrane through an ionophore-like mechanism, as well as having a direct ionophore effect on the plasma membrane. Indeed, older studies using reconstituted systems, such as artificial black-lipid bilayers,53 membrane vesicles,54 or red cell preparations,55 attest to the Ca2⫹-selective ionophoric characteristics of DCA. Moreover, treatment of cells with the Ca2⫹ ionophore, ionomycin, also was able to cause translocation of PH-PLC␦1-EGFP in a manner strictly dependent on extracellular Ca2⫹ (Lau and Hofer, unpublished results from 16 cells, 7 experiments), showing that the biological effects of DCA can be reproduced simply by increasing intracellular Ca2⫹ levels through Ca2⫹ entry at the plasma membrane. The efficacy of bile acids in releasing stores will be enhanced by the presence of specific bile uptake pathways,49 such as organic anion transporters. It also is known, however, that the highly hydrophobic bile compound, DCA, can passively cross plasma membranes.56,57 We were unable to determine whether the entry of DCA in BHK cells was through passive permeation pathways or via a specific transport process, but this will be a very important consideration when gauging the potential of DCA for exerting pathologic actions in different cell types of the alimentary tract. An important consequence of the combined PLC stimulation and Ca2⫹ entry provoked by DCA is that conventional isoforms of PKC become activated. DCA has been described previously as a non–phorbol-ester–type tumor promoter capable of activating PKC directly.22,23 Our data would suggest, however, that DCA-induced Ca2⫹ entry causes PLC turnover, and produces enough DAG to be permissive for translocation of the conventional (Ca2⫹- and DAG-dependent) isoforms of PKC, but not sufficient to cause palpable translocation of the novel PKC␦, which is dependent only on DAG. Evidence that activated conventional PKCs were competent to phosphorylate targets after DCA treatment is provided by the data shown in Figure 5, in which a major substrate of PKC, MARCKS, also was shown by translocation assay to have been activated. Again, the redistribution of MARCKS was dependent strictly on the entry of Ca2⫹ from the external medium. The notion that PIP2 turnover can be stimulated directly by Ca2⫹ entering across the plasma membrane has profound implications, and might have been anticipated from longstanding biochemical evidence regarding the Ca2⫹-dependence of all PLC isoforms.51 That physiologic modes of Ca2⫹ entry can elicit PIP2 hydrolysis directly in living cells is now supported by several recent reports.39,46,58,59 In one very elegant and technically sophisticated imaging study, Mogami et al46 showed that oscillatory influx of Ca2⫹
706
LAU ET AL
through voltage-dependent Ca2⫹ channels in Ins-1 cells was sufficient to cause PLC activation and DAG generation, and this was translated into persistent phosphorylation of MARCKS by conventional (PKC␣) and novel (PKC) isoforms of PKC. The results of this study differed slightly from those of the present work in that we did not detect robust activation of the novel PKC␦ after DCA-induced Ca2⫹ influx. In another study in the insulin-secreting line MIN6 and primary pancreatic islet cells, Pinton et al59 likewise provided evidence that localized domains of increased Ca2⫹ under the plasma membrane had a powerful effect in recruiting conventional PKCII, but not the novel isoform PKC␦, and additionally showed that different types of stimuli caused different patterns of PKC redistribution. Several years ago Okubo et al39 made the important observation that Ca2⫹ fluxes through AMPA receptors in neurons can activate PLC directly (as measured by PH-PLC␦ redistribution). Convincing evidence also recently has been presented indicating that the stimulatory actions of Ca2⫹ on PLC can account for the regenerative properties of Ca2⫹ waves in very large cells such as oocytes.60 These previous results, along with our own data, highlight the fact that Ca2⫹ entry is by itself sufficient to drive the activation of several downstream pathways normally associated with Ca2⫹ signaling through receptor stimulation. Our results provide one piece of the puzzle regarding the complex actions of DCA on cellular signal transduction, and potentially may contribute to the understanding of how this natural compound exerts its injurious effects on cells and organ systems of the alimentary tract.
References 1. Hofmann AF. The continuing importance of bile acids in liver and intestinal disease. Arch Intern Med 1999;159:2647–2658. 2. Bouscarel B, Kroll SD, Fromm H. Signal transduction and hepatocellular bile acid transport: cross talk between bile acids and second messengers. Gastroenterology 1999;117:433– 452. 3. Beuers U. Effects of bile acids on hepatocellular signaling and secretion. Yale J Biol Med 1997;70:341–346. 4. Makishima M, Okamoto AY, Repa JJ, Tu H, Learned RM, Luk A, Hull MV, Lustig KD, Mangelsdorf DJ, Shan B. Identification of a nuclear receptor for bile acids. Science 1999;284:1362–1365. 5. Makishima M, Lu TT, Xie W, Whitfield GK, Domoto H, Evans RM, Haussler MR, Mangelsdorf DJ. Vitamin D receptor as an intestinal bile acid sensor. Science 2002;296:1313–1316. 6. Goodwin B, Kliewer SA. Nuclear receptors. I. Nuclear receptors and bile acid homeostasis. Am J Physiol 2002;282:G926 –G931. 7. Werneburg NW, Yoon JH, Higuchi H, Gores GJ. Bile acids activate EGF receptor via a TGF-alpha-dependent mechanism in human cholangiocyte cell lines. Am J Physiol 2003;285:G31–G36. 8. Qiao L, Studer E, Leach K, McKinstry R, Gupta S, Decker R, Kukreja R, Valerie K, Nagarkatti P, El Deiry W, Molkentin J, Schmidt-Ullrich R, Fisher PB, Grant S, Hylemon PB, Dent P. Deoxycholic acid (DCA) causes ligand-independent activation of epidermal growth factor receptor (EGFR) and FAS receptor in primary hepatocytes: inhibition of EGFR/mitogen-activated protein kinase-signaling module enhances DCA-induced apoptosis. Mol Biol Cell 2001;12:2629 –2645.
GASTROENTEROLOGY Vol. 128, No. 3
9. Potter GD. Bile acid diarrhea. Dig Dis 1998;16:118 –124. 10. Mauricio AC, Slawik M, Heitzmann D, von Hahn T, Warth R, Bleich M, Greger R. Deoxycholic acid (DOC) affects the transport properties of distal colon. Pflugers Arch 2000;439:532–540. 11. Maruyama T, Miyamoto Y, Nakamura T, Tamai Y, Okada H, Sugiyama E, Itadani H, Tanaka K. Identification of membrane-type receptor for bile acids (M-BAR). Biochem Biophys Res Commun 2002;298:714 –719. 12. Kawamata Y, Fujii R, Hosoya M, Harada M, Yoshida H, Miwa M, Fukusumi S, Habata Y, Itoh T, Shintani Y, Hinuma S, Fujisawa Y, Fujino M. A G protein-coupled receptor responsive to bile acids. J Biol Chem 2003;278:9435–9440. 13. Montrose MH, Lester R, Zimniak P, Anwer MS, Murer H. Bile acids increase cellular free calcium in cultured kidney cells (LLCPK). Pflugers Arch 1988;412:164 –171. 14. Anwer MS, Engelking LR, Nolan K, Sullivan D, Zimniak P, Lester R. Hepatotoxic bile acids increase cytosolic Ca⫹⫹ activity of isolated rat hepatocytes. Hepatology 1988;8:887– 891. 15. Li M, Vemulapalli R, Ullah A, Izu L, Duffey ME, Lance P. Downregulation of a human colonic sialyltransferase by a secondary bile acid and a phorbol ester. Am J Physiol 1998;274:G599 –G606. 16. Dziki AJ, Batzri S, Harmon JW, Molloy M. Cellular hypercalcemia is an early event in deoxycholate injury of rabbit gastric mucosal cells. Am J Physiol 1995;269:G287–G296. 17. Berridge MJ, Bootman MD, Roderick HL. Calcium signalling: dynamics, homeostasis and remodelling. Nat Rev Mol Cell Biol 2003;4:517–529. 18. Berridge MJ, Lipp P, Bootman MD. The versatility and universality of calcium signalling. Nat Rev Mol Cell Biol 2000;1:11–21. 19. Combettes L, Berthon B, Doucet E, Erlinger S, Claret M. Characteristics of bile acid-mediated Ca2⫹ release from permeabilized liver cells and liver microsomes. J Biol Chem 1989;264:157–167. 20. Combettes L, Dumont M, Berthon B, Erlinger S, Claret M. Release of calcium from the endoplasmic reticulum by bile acids in rat liver cells. J Biol Chem 1988;263:2299 –2303. 21. Voronina S, Longbottom R, Sutton R, Petersen OH, Tepikin A. Bile acids induce calcium signals in mouse pancreatic acinar cells: implications for bile-induced pancreatic pathology. J Physiol 2002;540:49 –55. 22. Huang XP, Fan XT, Desjeux JF, Castagna M. Bile acids, nonphorbol-ester-type tumor promoters, stimulate the phosphorylation of protein kinase C substrates in human platelets and colon cell line HT29. Int J Cancer 1992;52:444 – 450. 23. Milovic V, Teller IC, Faust D, Caspary WF, Stein J. Effects of deoxycholate on human colon cancer cells: apoptosis or proliferation. Eur J Clin Invest 2002;32:29 –34. 24. Craven PA, Pfanstiel J, DeRubertis FR. Role of activation of protein kinase C in the stimulation of colonic epithelial proliferation and reactive oxygen formation by bile acids. J Clin Invest 1987;79:532–541. 25. Zhu Y, Hua P, Rafiq S, Waffner EJ, Duffey ME, Lance P. Ca2⫹- and PKC-dependent stimulation of PGE2 synthesis by deoxycholic acid in human colonic fibroblasts. Am J Physiol 2002;283:G503–G510. 26. Kataoka H, Tanaka H, Nagaike K, Uchiyama S, Itoh H. Role of cancer cell-stroma interaction in invasive growth of cancer cells. Hum Cell 2003;16:1–14. 27. Nejjari M, Anderson W, Pourreyron C, Jacquier MF, Scoazec JY, Remy L. The role of fibroblasts in the modulation of integrindependent interactions between the gastric cell line HGT-1 and fibronectin. Int J Cancer 2004;112:560. 28. Ohmori S, Sakai N, Shirai Y, Yamamoto H, Miyamoto E, Shimizu N, Saito N. Importance of protein kinase C targeting for the phosphorylation of its substrate, myristoylated alanine-rich Ckinase substrate. J Biol Chem 2000;275:26449 –26457. 29. Sawano A, Hama H, Saito N, Miyawaki A. Multicolor imaging of Ca(2⫹) and protein kinase C signals using novel epifluorescence microscopy. Biophys J 2002;82:1076 –1085.
March 2005
30. Dale LB, Babwah AV, Bhattacharya M, Kelvin DJ, Ferguson SS. Spatial-temporal patterning of metabotropic glutamate receptormediated inositol 1,4,5-triphosphate, calcium, and protein kinase C oscillations: protein kinase C-dependent receptor phosphorylation is not required. J Biol Chem 2001;276:35900 –35908. 31. Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca2⫹ indicators with greatly improved fluorescence properties. J Biol Chem 1985;260:3440 –3450. 32. Caroppo R, Colella M, Colasuonno A, DeLuisi A, Debellis L, Curci S, Hofer AM. A reassessment of the effects of luminal [Ca2⫹] on inositol 1,4,5-trisphosphate-induced Ca2⫹ release from internal stores. J Biol Chem 2003;278:39503–39508. 33. Hofer AM, Machen TE. Technique for in situ measurement of calcium in intracellular inositol 1,4,5-trisphosphate-sensitive stores using the fluorescent indicator mag-fura-2. Proc Natl Acad Sci U S A 1993;90:2598 –2602. 34. Hofer AM, Landolfi B, Debellis L, Pozzan T, Curci S. Free [Ca2⫹] dynamics measured in agonist-sensitive stores of single living intact cells: a new look at the refilling process. EMBO J 1998; 17:1986 –1995. 35. Hofer AM. Measurement of free [Ca2⫹] changes in agonistsensitive internal stores using compartmentalized fluorescent indicators. Methods Mol Biol 1999;114:249 –265. 36. Chiesa A, Rapizzi E, Tosello V, Pinton P, de Virgilio M, Fogarty KE, Rizzuto R. Recombinant aequorin and green fluorescent protein as valuable tools in the study of cell signalling. Biochem J 2001; 355:1–12. 37. Sakai N, Sasaki K, Ikegaki N, Shirai Y, Ono Y, Saito N. Direct visualization of the translocation of the gamma-subspecies of protein kinase C in living cells using fusion proteins with green fluorescent protein. J Cell Biol 1997;139:1465–1476. 38. Stauffer TP, Ahn S, Meyer T. Receptor-induced transient reduction in plasma membrane PtdIns(4,5)P2 concentration monitored in living cells. Curr Biol 1998;8:343–346. 39. Okubo Y, Kakizawa S, Hirose K, Iino M. Visualization of IP(3) dynamics reveals a novel AMPA receptor-triggered IP(3) production pathway mediated by voltage-dependent Ca(2⫹) influx in Purkinje cells. Neuron 2001;32:113–122. 40. Nagai T, Sawano A, Park ES, Miyawaki A. Circularly permuted green fluorescent proteins engineered to sense Ca2⫹. Proc Natl Acad Sci U S A 2001;98:3197–3202. 41. Landolfi B, Curci S, Debellis L, Pozzan T, Hofer AM. Ca2⫹ homeostasis in the agonist-sensitive internal store: functional interactions between mitochondria and the ER measured in situ in intact cells. J Cell Biol 1998;142:1235–1243. 42. Molloy M, Batzri S, Dziki AJ, Goldberg WJ, Hale DA, Harmon JW. Reversibility of deoxycholate-induced cellular hypercalcemia in rabbit gastric mucosal cells. Surgery 1996;119:89 –97. 43. Erlinger S. Review article: new insights into the mechanisms of hepatic transport and bile secretion. J Gastroenterol Hepatol 1996;11:575–579. 44. Peppiatt CM, Collins TJ, Mackenzie L, Conway SJ, Holmes AB, Bootman MD, Berridge MJ, Seo JT, Roderick HL. 2-Aminoethoxydiphenyl borate (2-APB) antagonises inositol 1,4,5-trisphosphate-induced calcium release, inhibits calcium pumps and has a use-dependent and slowly reversible action on store-operated calcium entry channels. Cell Calcium 2003;34:97–108. 45. Bootman MD, Collins TJ, Mackenzie L, Roderick HL, Berridge MJ, Peppiatt CM. 2-aminoethoxydiphenyl borate (2-APB) is a reliable blocker of store-operated Ca2⫹ entry but an inconsistent inhibitor of InsP3-induced Ca2⫹ release. FASEB J 2002;16:1145–1150. 46. Mogami H, Zhang H, Suzuki Y, Urano T, Saito N, Kojima I, Petersen OH. Decoding of short-lived Ca2⫹ influx signals into long term substrate phosphorylation through activation of two distinct classes of protein kinase C. J Biol Chem 2003;278:9896 –9904. 47. Xu C, Watras J, Loew LM. Kinetic analysis of receptor-activated phosphoinositide turnover. J Cell Biol 2003;161:779 –791.
DCA ACTIVATES PLC THROUGH CA2ⴙ ENTRY
707
48. Debruyne PR, Bruyneel EA, Li X, Zimber A, Gespach C, Mareel MM. The role of bile acids in carcinogenesis. Mutat Res 2001; 480 – 481:359 –369. 49. Powell AA, LaRue JM, Batta AK, Martinez JD. Bile acid hydrophobicity is correlated with induction of apoptosis and/or growth arrest in HCT116 cells. Biochem J 2001;356:481– 486. 50. Kaur BS, Triadafilopoulos G. Acid- and bile-induced PGE(2) release and hyperproliferation in Barrett’s esophagus are COX-2 and PKCepsilon dependent. Am J Physiol 2002;283:G327–G334. 51. Rhee SG, Bae YS. Regulation of phosphoinositide-specific phospholipase C isozymes. J Biol Chem 1997;272:15045–15048. 52. Kim JY, Kim KH, Lee JA, Namkung W, Sun AQ, Ananthanarayanan M, Suchy FJ, Shin DM, Muallem S, Lee MG. Transporter-mediated bile acid uptake causes Ca2⫹-dependent cell death in rat pancreatic acinar cells. Gastroenterology 2002;122:1941–1953. 53. Abramson JJ, Shamoo AE. Anionic detergents as divalent cation ionophores across black lipid membranes. J Membr Biol 1979; 50:241–255. 54. Maenz DD, Forsyth GW. Ricinoleate and deoxycholate are calcium ionophores in jejunal brush border vesicles. J Membr Biol 1982;70:125–133. 55. Oelberg DG, Dubinsky WP, Sackman JW, Wang LB, Adcock EW, Lester R. Bile salts induce calcium uptake in vitro by human erythrocytes. Hepatology 1987;7:245–252. 56. Aldini R, Montagnani M, Roda A, Hrelia S, Biagi PL, Roda E. Intestinal absorption of bile acids in the rabbit: different transport rates in jejunum and ileum. Gastroenterology 1996;110:459–468. 57. Cabral DJ, Small DM, Lilly HS, Hamilton JA. Transbilayer movement of bile acids in model membranes. Biochemistry 1987;26: 1801–1804. 58. Thore S, Dyachok O, Tengholm A. Oscillations of phospholipase C activity triggered by depolarization and Ca2⫹ influx in insulinsecreting cells. J Biol Chem 2004;279:19396 –19400. 59. Pinton P, Tsuboi T, Ainscow EK, Pozzan T, Rizzuto R, Rutter GA. Dynamics of glucose-induced membrane recruitment of protein kinase C beta II in living pancreatic islet beta-cells. J Biol Chem 2002;277:37702–37710. 60. Wagner J, Fall CP, Hong F, Sims CE, Allbritton NL, Fontanilla RA, Moraru II, Loew LM, Nuccitelli R. Wave of IP3 production accompanies the fertilization Ca2⫹ wave in the egg of the frog, Xenopus laevis: theoretical and experimental support. Cell Calcium 2004;35:433–447.
Received August 12, 2004. Accepted December 2, 2004. Address requests for reprints to: Aldebaran M. Hofer, PhD, Boston VA Healthcare System and the Department of Surgery, Harvard Medical School, Brigham and Women’s Hospital, 1400 VFW Parkway, West Roxbury, Massachusetts 02132. e-mail:
[email protected]; fax: (617) 363-5592. B.W.L. was the recipient of an AGA Student Research Fellowship from the Foundation for Digestive Health and Nutrition. This work was supported by grants from the Medical Research Service of the Veteran’s Administration, a Harvard Digestive Diseases Center grant (National Institutes of Health DK-3485), and support from the Brigham Surgical Group (to A.M.H.). M.R. was supported by a doctoral fellowship awarded from the University of Bari (I). The authors thank their colleagues around the world who kindly supplied the various GFP-based indicators used in this study; Dr. Atsushi Miyawaki for generously providing the nuclear ratiometric pericam complementary DNA; Professor Naoki Saito for the kind gift of the MARCKS-EGFP; Dr. Stephen S. G. Ferguson, London Ontario, Canada, and colleagues for their invaluable help in sending their PH-PLC␦1EGFP construct; Dr. Sarino Rizzuto and Dr. Paolo Pinton at the University Ferrara, Italy, and Dr. Paulo J. Magalhães at the University of Padova, Italy, for graciously supplying the PKC biosensors; and Dr. Francesco Lopez for help with the experiments using HTPLC.
