Ethanol Disrupts Hormone-Induced Calcium Signaling in Liver.

Receptor-coupled phospholipase C (PLC) is an important target for the actions of ethanol. In the ex vivo perfused rat liver, concentrations of ethanol >100 mM were required to induce a rise in cytosolic calcium (Ca2+) suggesting that these responses may only occur after binge ethanol consumption. Conversely, pharmacologically achievable concentrations of ethanol (≤30 mM) decreased the frequency and magnitude of hormone-stimulated cytosolic and nuclear Ca2+ oscillations and the parallel translocation of protein kinase C-β to the membrane. Ethanol also inhibited gap junction communication resulting in the loss of coordinated and spatially organized intercellular Ca2+ waves in hepatic lobules. Increasing the hormone concentration overcame the effects of ethanol on the frequency of Ca2+ oscillations and amplitude of the individual Ca2+ transients; however, the Ca2+ responses in the intact liver remained disorganized at the intercellular level, suggesting that gap junctions were still inhibited. Pretreating hepatocytes with an alcohol dehydrogenase inhibitor suppressed the effects of ethanol on hormone-induced Ca2+ increases, whereas inhibiting aldehyde dehydrogenase potentiated the inhibitory actions of ethanol, suggesting that acetaldehyde is the underlying mediator. Acute ethanol intoxication inhibited the rate of rise and the magnitude of hormone-stimulated production of inositol 1,4,5-trisphosphate (IP3), but had no effect on the size of Ca2+ spikes induced by photolysis of caged IP3. These findings suggest that ethanol inhibits PLC activity, but does not affect IP3 receptor function. We propose that by suppressing hormone-stimulated PLC activity, ethanol interferes with the dynamic modulation of [IP3] that is required to generate large, amplitude Ca2+ oscillations.

Introduction

Calcium mobilization from intracellular stores and Ca2+-influx from the extracellular space are ubiquitous signals used to control a wide range of cellular processes. Calcium-mobilizing hormones, such as α1-adrenergic receptor agonists and vasopressin, regulate liver function by activating the phosphoinositide-specific phospholipase C (PI-PLC) signaling cascade to produce transient increases in cytosolic free calcium concentration ([Ca2+]i). Increases in [Ca2+]i are achieved by the activation of receptor-associated trimeric G proteins (Gq/G11), stimulation of PLCβ activity, hydrolysis of the lipid precursor phosphatidylinositol 4,5-bisphosphate, and the corelease of the second messenger’s inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). DAG stimulates the activity of protein kinase C (PKC), while IP3 binds to receptors in the endoplasmic reticulum to release internal Ca2+ stores., Depletion of the internal Ca2+ stores, in turn, stimulates store-operated calcium entry to sustain the hormone signal and refill the internal Ca2+ stores. In nonexcitable cells, such as pancreatic acinar cells and hepatocytes, these Ca2+ signals are manifest as periodic Ca2+ increases referred to as Ca2+ oscillations, and the individual Ca2+ transients propagate across the cell from a discrete origin as Ca2+ waves. The frequency of Ca2+ oscillations is determined by the concentration of the hormone, while the amplitude and kinetics of individual [Ca2+]i spikes as well as the rates at which the Ca2+ increase propagates across the cell are largely independent of agonist dose. These oscillatory calcium signals are rapidly transferred into the mitochondrial matrix thereby stimulating Ca2+-sensitive dehydrogenases leading to enhanced production of NAD(P)H, and synthesis of mitochondrial ATP. PLC-linked [Ca2+]i spikes also propagate into the hepatocyte nucleus stimulating the activity of Ca2+-sensitive transcription factors and kinases that, in turn, regulate cell proliferation and responses to acute liver injury.,

Previous studies have identified multiple sites of interaction between ethanol intoxication and the phosphoinositide signaling system. In pancreatic acinar cells, nonoxidative metabolites of ethanol can mobilize internal Ca2+ stores through an IP3 receptor mechanism, whereas ethanol targets the activity of PLC in hepatocytes., The addition of ethanol to isolated hepatocyte suspensions, in the absence of receptor agonists, has been shown to stimulate receptor-coupled PLC, evident by the accumulation of [3H]inositol, a rise in [Ca2+]i, and the Ca2+-dependent activation of phosphorylase a. However, ethanol-stimulated PLC activity requires high concentrations of ethanol (≥100 mM), is short-lived and followed by desensitization of the pathway. The direct ethanol-induced Ca2+ increases are also sensitive to experimental conditions, and are either absent or markedly diminished in hepatocytes maintained in primary culture. Importantly, it is unclear if blood alcohol concentrations can rise to the levels that activate PLC in vivo except after excessive binge ethanol intake, or if these Ca2+ signals play a role in the development of liver injury.

It has also been documented that acute ethanol treatment suppresses the cellular responses to multiple hormones coupled to the activation of PLCβ.,,, Here, the ethanol concentrations required to inhibit receptor-coupled PLC activity are lower compared to those required to activate the enzyme directly and are unaffected by culture conditions. Moreover, the mechanism by which ethanol suppresses PLC is thought to be distinct from the actions of ethanol that result in the activation of PLC. The sites of acute ethanol action have not been fully defined. Ethanol may interfere with receptor-G protein coupling and/or potentiate a PKC-dependent negative feedback loop onto PLC activity.,,, Either mechanism could explain the decrease in hormone-stimulated production of IP3. Ethanol has also been reported to inhibit IP3 binding and to reduce the efficacy of IP3 to activate the intracellular IP3-gated Ca2+ release channels in digitonin-permeabilized hepatocytes. Perturbing IP3 receptor function would also contribute to the suppression of PLC-linked Ca2+ signals.

We have previously shown that hepatocytes are desensitized to the inhibitory actions of acute ethanol treatment on hormone-induced Ca2+ signals after chronic ethanol feeding. Hepatocytes isolated from ethanol-fed rats or mice exhibit higher levels of receptor-mediated PLC activity and a leftward shift in the dose-response curves for Ca2+-mobilizing stimuli, resulting in more sustained increases in [Ca2+]i at low hormone concentrations rather than the typical oscillatory Ca2+ signals. These compensatory mechanisms would allow the liver to respond to Ca2+-mobilizing hormones even in the presence of intoxicating levels of ethanol, but may also predispose the tissue to further ethanol-induced injury by causing Ca2+ overload or over stimulation of PKC during periods of ethanol withdrawal.

There is increasing evidence that dysregulation in Ca2+ handling and Ca2+ signaling is a potential driver in the development of numerous liver diseases including nutrient-induced fatty liver disease, hepatic insulin resistance,, cholestasis,, liver regeneration,, hepatocellular carcinoma, and metastases.,. In this study, we first demonstrate that ethanol concentrations >100 mM are required to induce [Ca2+]i increases in hepatocytes within the intact perfused rat liver. These ethanol-induced Ca2+ oscillations occurred in small numbers of cells and may only manifest after a high dose of acute ethanol feeding. By contrast, ethanol concentrations that are readily achievable in vivo potently inhibit hormone-induced Ca2+ oscillations and decrease the magnitude of both cytosolic and nuclear Ca2+ oscillations in the ex vivo perfused liver. Moreover, these levels of ethanol also inhibited gap junction communication in intact liver, causing the loss of coordinated and spatially organized intercellular Ca2+ waves, which are thought to synchronize the metabolic output of the liver. The actions of ethanol on hormone-stimulated Ca2+ oscillations required metabolism and were largely prevented by pretreatment with 4-methylpyrazole, an alcohol dehydrogenase inhibitor, indicating a role for acetaldehyde in inhibiting IP3/PLC-linked Ca2+ oscillations. Ethanol treatment decreased both the rate of rise and magnitude of hormone-stimulated IP3 formation. We predict that by slowing down the dynamic changes in [IP3], ethanol interferes with Ca2+-IP3 cross-coupling causing the early termination of [Ca2+]i spikes. Indeed, increasing the concentration of the hormone and presumably enhancing the rates of IP3 production overcame the effects of ethanol on the frequency of Ca2+ oscillations and amplitude of the Ca2+ spike. Ethanol-dependent inhibition of receptor-operated Ca2+ oscillations was paralleled by the loss of conventional PKCβ translocation to the membrane. Taken together, this study demonstrates that acute ethanol intoxication suppresses receptor-mediated activation of PLC, which translates into the bilateral inhibition of both Ca2+ and PKC signaling pathways in hepatocytes.

Materials and Methods

Reagents

Hydroxyethyl piperazineethanesulfonic acid (HEPES) and all other chemicals were obtained from Sigma-Aldrich (St Louis, MO, USA) or VWR BDH chemicals (Radnor, PA, USA). Fura-2/AM, fluo-4/AM, calcein/AM, pluronic acid F-127, and fluorescein conjugated-BSA were obtained from Thermo Fisher Scientific (Waltham, MA, USA), bromosulfophthalein (BSP) was from Fluka (Fluka, Buchs, Switzerland), and 190 proof ethyl alcohol from Decon, Labs Inc. Heat shock, fraction V BSA was from Roche (Basel, Switzerland). BSA was extensively dialyzed against normal saline plus 10 mM HEPES using Spectra/Por 12–14 kDa cut-off dialysis tubing (Thermo Fisher Scientific).

Dr. K. Mikoshiba (RIKEN) generously provided the plasmid pIRIS-1 and pRatiometric-pericam-nu was a gift from A. Miyawaki (RIKEN). pGP-CMV-GCaMP6f was a gift from Douglas Kim & GENIE Project (Addgene plasmid # 40755; http://n2t.net/addgene:40755; RRID:Addgene_40755). Adenovirus particles expressing PKCβII-enhanced green fluorescent protein (Ad-PKCβII-EGFP) were a gift from Dr. R. Rizzuto (University of Padua).

Ethical Approval and Animals

Animal studies were approved by the Institutional Animal Care and Use Committee at Rutgers, New Jersey Medical School. Male Sprague Dawley rats (weighing 200–250 g; Taconic Biosciences, Rensselaer, NY, USA) were housed in ventilated cages under a 12:12 h dark:light cycle in a fully Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) accredited animal facility. Rats were given ad libitum access to rodent chow and water until the day of experimentation. Rats were anesthetized with an I.P. injection of Nembutal (60 mg/kg, Diamondback Drugs, Scottsdale, AZ, USA) diluted 1:1 with PBS. The depth of anesthesia was assessed by relaxation of muscle tone and a loss of reflex responses to external stimuli. Liver tissue was harvested under a surgical plane of anesthesia and used for isolation of hepatocytes or establishing ex vivo perfused liver preparations. The organ donor did not recover from surgery.

Adenovirus Production

Ad-PKCβII-EGFP virus particles were produced in HEK 293 cells. EGFP expression was used to monitor production of viruses. Virus particles were extracted, purified by ultracentrifugation on a CsCl gradient then dialyzed extensively against 20 mM Tris, pH 8.0, 25 mM NaCl, and 2.5% glycerol. Viral stocks (1013 particles/mL) were stored at −80° C until used.

Ex vivo Perfused Liver

The livers of anesthetized male Sprague Dawley rats (200–250 g) were perfused in situ via the hepatic portal vein with a HEPES-buffered balanced salt solution (HBSS) as described previously. The perfusion buffer was composed of (in mM): 121 NaCl, 25 Hepes (pH 7.4 at @30°C), 5 NaHCO3, 4.7 KCl, 1.2 KH2PO4, 1.2 MgSO4, 1.3 CaCl2, 5.5 glucose, 0.5 glutamine, 3 lactate, 0.3 pyruvate, 0.2 BSP, and 0.1% (w/v) dialyzed BSA, equilibrated with 100% O2 at 30°C. BSP, an organic anion transport inhibitor, was included to inhibit efflux of Ca2+ indicator dyes. The median lobe of the liver was used for these studies and other lobes were tied off with silk thread and excised to increase the loading efficiency of indicator dyes for confocal imaging. Ca2+-sensitive indicator dyes or calcein/AM were loaded by recirculating perfusion buffer supplemented with 5 µM of the acetoxymethyl ester, 0.02% (v/v) Pluronic F-127, and 2% (w/v) dialyzed BSA for 40–50 min. Following dye loading, the liver was transferred to a custom-built imaging chamber fitted onto the microscope stage as described previously.

Hepatocyte Isolation and Culture

Hepatocytes were isolated by a two-step collagenase perfusion of livers from fed male Sprague Dawley rats essentially as described previously. Hepatocytes (500 000–750 000 cells) were plated onto glass coverslips coated with type 1 collagen (rat tail, Corning; 10 μg/cm2). Cells were maintained for 30–60 min in a humidified atmosphere of 5% CO2 and 95% air at 37°C in complete Williams’ Medium E (WEM, Thermo Fisher Scientific) supplemented with 5% (v/v) fetal bovine serum (Atlanta Biologicals), penicillin, (10 units/mL), streptomycin (10 μg/mL), gentamycin sulfate (50 μg/mL), and glutamine (2 mM). In some experiments, hepatocytes were transduced with Ad-PKCβII-EGFP (100 particles/cell) or transfected by electroporation using an Amaxa Rat/Mouse Hepatocyte Nucleofector Kit according to the manufacturer’s instructions (Lonza, Basel, Switzerland). Hepatocytes used in electroporation or viral transduction protocols were maintained in complete WEM containing 140 nM insulin for 3–4 h and then insulin was reduced to 14 nM and the cells cultured for an additional 16–20 h to allow expression of the transgene.

Imaging Measurements of [Ca2+]i in the ex vivo Perfused Liver

Confocal images were acquired with a Radiance 2002-MP laser scanning multiphoton confocal microscope (Bio-Rad, Hercules, CA, USA) using either 10 × 0.5 NA or 20 × 0.75 NA PlanFluor objectives (Nikon, Tokyo, Japan) using the manufacturer’s software as described previously. A Mira/Verdi Ti:sapphire laser (Coherent Inc. Santa Clara, CA, USA) was used for multiphoton excitation of the fura-2 and the 488 nm argon laser line was utilized for single-photon excitation of fluo-4. Fura2 images (810 nm excitation, 460–600 nm emission) and fluo-4 images (488 nm excitation, 520–600 nm emission) were captured every 4 s. At the end of the experiment, fluorescein-labeled BSA (F-BSA) was injected into the perfusion line to identify the different lobular zones as described previously. In some experiments, fura-2 fluorescence (excitation 340 and 380 nm, emission 420–600 nm) was also detected using a cooled charge-coupled device (CCD) camera (Photometrics) as described previously. Values for [Ca2+] were calculated from the 340/380 nm ratio after correction for autofluorescence and out of focus information as described in. Autofluorescence images were acquired by perfusing the liver for 10–15 min with 100 µM MnCl2 in Ca2+-free HBSS buffer to quench cytosolic fura-2, and images were deblurred with deconvolution algorithms to remove out of focus fluorescence.

When applicable, minor movements of the liver tissue in the x,y direction were corrected using the StackReg plugin for ImageJ. Hormone-induced increase in fura-2 or fluo-4 fluorescence intensities were analyzed using in-house software or algorithms written for ImageJ (NIH, Bethesda, MD, USA). The lobular organization of hormone-induced increases in Ca2+ is shown as the actual fluorescence intensity values depicted on a linear gray scale with the magnitude of increases in [Ca2+] at each time point shown with a red overlay as described previously., Changes in [Ca2+]i were calculated by subtracting sequential confocal images to yield a differential image stack and the fluorescence intensity at each time point is proportional to the change in Ca2+. Difference images were processed with a median filter and a minimum threshold intensity to reduce noise.

Fluorescence Recovery after Photobleaching

Fluorescence recovery after photobleaching (FRAP) experiments were carried out in calcein/AM-loaded livers. Routines written in the manufacturer’s macro language were used to control the bleaching laser pulse and image acquisition. Photobleaching was carried out by briefly focusing the laser onto a single hepatocyte and then acquiring full-field images. Confocal images of calcein fluorescence were acquired every 5–10 s using a Bio-Rad MRC-600 laser scanning confocal microscope (488 nm excitation, 515–600 nm emission) and a 60 × 1.3 NA SPlan objective. A linear fit was used to calculate the initial rate of fluorescence recovery.

Single-Cell Imaging Measurements and Data Analysis

Single-cell imaging experiments were performed in HBSS containing 100–200 μM sulfobromophthalein and 0.25% (w/v) BSA, pH 7.4 at 37°C. Cell cultures were loaded with fura-2 by incubation with 2–5 μM fura-2/AM plus Pluronic F-127 (0.02% v/v) for 20–40 min in HBSS. The cells were washed twice with HBSS and then transferred to a thermostatically regulated microscope chamber (37°C). Fura-2 fluorescence images (excitation, 340 ± 10 and 380 ± 10 nm, emission 420–600 nm) were acquired at 3–4 s intervals with a cooled CCD camera as previously described., Calibration of fura-2 in terms of [Ca2+]i was calculated from the 340/380 nm ratio after correcting for cellular autofluorescence. Autofluorescence values were obtained at the end of each experiment by permeabilizing the cells with digitonin in an intracellular-like media containing a Ca2+/EGTA buffer and Mg-ATP to release cytosolic fura-2. The fura-2 calibration parameters were determined in vitro using a Kd value of 224 nM. For simultaneous measurement of hormone-induced [Ca2+]i and PKCβII-EGFP responses, fura-2-loaded hepatocytes were excited with 340, 380, and 485 nm, using a computer-controlled filter wheel and shutter. The emitted fluorescence was collected using a fura-2/BCECF dichroic beam splitter and a 515 nm long bandpass emission filter (Chroma Technology) as described previously. For these experiments, the fura-2 ratio was not converted into [Ca2+]i due to a small amount of EGFP contamination into the fura-2 fluorescence intensity signal. Changes in PKCβII-EGFP distribution were monitored by taking the ratio of the GFP fluorescence intensity changes in the membrane versus cytosol. The resulting membrane:cytosol ratio was normalized to the baseline values prior to data analysis. The amplitude, rate of rise, rate of decline, and area under the curve (AUC) for [Ca2+]i spikes and the translocation of PKCβII-EGFP to the membrane were determined using algorithms (Brumer R & Thomas A, unpublished work) written in MATLAB (MathWorks, Natick, MA, USA).

Flash Photolysis of Caged IP3

Cytosolic calcium changes induced by photolysis of caged IP3 were measured using the Ca2+ indicator fluo4-AM. Cells were loaded with 2 μM caged IP3 (Sichem Gmbh) in HBSS for 45 min; followed by 30 min with 5 μM fluo4/AM. Cells were washed and mounted on the stage of an Axiovert200 (Zeiss) spinning disc confocal microscope. Fluo4 fluorescence images (Argon laser excitation 488 nm, emission >510 nm) were acquired at 2 Hz with a CCD camera using the data acquisition software Piper ControlTM (Stanford photonics). Photolysis of caged IP3 was achieved by light pulses (1 ns duration with a wavelength of 337 nm and 1.45 mJ of energy) from a nitrogen-charged UV flash lamp (Photon Technology International) guided through the objective (C-Achromatx20/1.2). Data analysis was performed using ImageJ (NIH).

Simultaneous Measurement of [Ca2+]i and IRIS-1 in Single Cells

Hepatocytes were transfected with 2–5 µg of IRIS-1 via electroporation (Nucleofector™; Lonza), plated onto collagen (10 µg/cm3)-coated glass coverslips then cultured overnight in insulin-containing WEM. Transfected cultures were incubated with 1 μM indo-1/AM and Pluronic acid F-127 (0.02% v/v) for 20–30 min at 37°C in HBSS. Simultaneous indo-1 and IRIS-1 fluorescence images were acquired by sequential illumination with 436 ± 20 nm then 360 ± 40 nm excitation using a 455 nm long bandpass dichroic mirror. IRIS-1 donor and acceptor fluorescence images were separated with a 505 nm long bandpass dichroic mirror and directed to 480 ± 30 nm (CFP) or 535 ± 40 nm (Venus) emission filters using an image beam splitter (Optical Insights™). The CFP donor emission image recorded at 480 nm was divided by the acceptor emission image recorded at 535 nm to obtain the FRET emission ratio. Indo-1 fluorescence intensities decrease with increasing [Ca2+]i in both emission channels. These images are summed to increase the signal output and reported as the ratio of (F/F0)−1.

Statistical Analysis

Data are presented as means ± SD for the number of livers, separate hepatocyte preparations, or cells as indicated in the figure legend. Statistical differences from the relevant controls were calculated by Student’s t-test or one-way ANOVA for more than two groups (Prism; GraphPad Software Inc., San Diego, CA, USA). A value of P < 0.05 was considered statistically significant.

Results

High Doses of Ethanol Induce [Ca2+]i Spikes in the ex vivo Perfused Liver

Relatively high concentrations of ethanol have previously been shown to induce [Ca2+]i increases in hepatocyte suspensions by activating receptor-coupled PLC activity.,,, In this study, we tested whether lower concentrations of ethanol can also induce [Ca2+]i spikes in the ex vivo perfused liver; a tissue preparation that maintains the normal architecture of the liver and cell-to-cell communication. Perfused rat livers were loaded with Ca2+-sensitive indicator dyes (see “Materials and Methods” section) and challenged with increasing concentrations of ethanol. Live tissue confocal imaging studies revealed that concentrations of ethanol up to 30 mM did not induce Ca2+ responses after 10–20 min of perfusion (not shown), whereas ethanol concentrations ≥100 mM induced [Ca2+]i oscillations in a few hepatocytes scattered throughout the liver lobule. Increases in [Ca2+]i were detected in about 1% of the cells within a field of view after perfusion with 100 mM ethanol (1.2 ± 1.3%; mean ± SD, n = 3 livers) and increased to about 5–10% in the presence of 300 mM (Figure 1A; 6 ± 8%; mean ± SD, n = 3). In Figure 1A, the red overlay was created from a maximal projection through time of a differential image series, showing the peak [Ca2+]i increase in all the responsive hepatocytes during the 10 min perfusion period. The red color shows the lobular distribution of ethanol-responsive cells, confined predominately to the periportal regions, and the color intensity reflects the magnitude of ethanol-induced Ca2+ activity. The traces in Figure 1B show the [Ca2+]i responses in two adjacent cells along the same hepatic plate after perfusing the liver with 300 mM ethanol. The data indicate that ethanol-induced [Ca2+]i increases are asynchronous and do not propagate as intercellular Ca2+ waves, as shown for Ca2+-mobilizing hormones, eg, see Figure 2A. The biological significance of these high ethanol-induced Ca2+ responses is not clear as these levels are not typically achieved in animal models of ethanol feeding and may only occur in extremely intoxicated individuals (see “Discussion” section). Therefore, we decided to focus on ethanol actions on hormone-induced Ca2+ signaling, as these effects occur at lower pharmacologically relevant ethanol concentrations.

Figure 1.

High Doses of Ethanol Stimulate Asynchronous [Ca2+]i Increases in the Perfused Liver. A rat liver loaded with fura-2/AM was perfused with 300 mM ethanol for 10 min. (A) A differential image time series was calculated to localize positive [Ca2+]i changes, and this was used to generate a maximum intensity projection (red) superimposed on the original fluorescence image (gray). Thus, the red overlay shows the location of ethanol responsive cells and the color intensity reflects the peak Ca2+. The white box shows the area displayed as an insert in panel b at higher resolution. (B) Colored traces are the [Ca2+]i responses in two adjacent hepatocytes during the perfusion with ethanol (bar). The lobular location of the cells is depicted by the similarly colored arrows shown on the insert.

Figure 2.

Effect of Acute Ethanol Treatment on Vasopressin-Induced Intercellular Ca2+ Waves in the Intact Liver. (A and B) A fluo-3/AM loaded rat liver was infused with 100 pM vasopressin (A) followed by perfusion with 20 mM ethanol in the continuing presence of vasopressin (B). Confocal images (563 µm × 686 µm) of fluo-3 fluorescence are displayed in a linear grayscale with increases in fluorescence intensity at each time point displayed as a red overlay. Time (s) after vasopressin (A, 126–336 s) or ethanol (B, 15–105 s) reach the liver are shown in the top left of each image. In (B), the bottom right two panels show the Ca2+ response at 70 s and 156 s (70 s_w, 156 s_w) after beginning the washout of ethanol. Periportal (PP) zone and the boundary (white line) of a lobule are illustrated in the first image of Panel A. White scale bar denotes 100 µm. (C) Traces are the [Ca2+]i responses from three hepatocytes positioned sequentially along a hepatic plate. The Ca2+ traces are displaced along the y-axis for clarity. The insert on the right shows the cell-to-cell propagation of one Ca2+ spike induced by vasopressin. The colors and numbers of the cells correspond to the similar colored and numbered Ca2+ traces. Data were acquired at higher magnification (×20) in a separate liver preparation.

Acute Ethanol Treatment Inhibits Hormone-Induced Intercellular Ca2+ Waves in the Intact Liver

As reported previously,, perfusion with submaximal concentrations of vasopressin (30–100 pM) initially induce periodic [Ca2+]i spikes (oscillations) in a small number of hepatocytes surrounding the periportal tract, which then spread progressively along the hepatic plates generating a wave of [Ca2+]i increase throughout the liver lobule (Figure 2A). The average frequency of vasopressin-induced Ca2+ oscillations was 0.71 ± 0.2 and 0.66 ± 0.2 spikes/min (mean ± SD, n = 17 livers) in the periportal and pericentral zones, respectively. In the continuous presence of hormone, each Ca2+ spike generated in the initiating periportal cells propagated throughout the lobule as an intercellular Ca2+ wave. The images in Figure 2B show the effect of perfusing the same liver with 20 mM ethanol in the continuing presence of the vasopressin. The arrival of ethanol into the lobule often induced a [Ca2+]i increase in a few cells scattered throughout field of view (Figure 2B, 15 s), and this was followed by a strong inhibition of vasopressin-induced Ca2+ oscillations and the loss of coordinated intercellular Ca2+ waves (Figure 2B, 45–105 s). Experiments carried out at higher magnification revealed that Ca2+ spikes generated within pacemaker-like hepatocytes did not propagate from cell-to-cell along the hepatic plate in the presence of ethanol, suggesting that gap junction communication was impaired (Figure 2C).

The frequency of the Ca2+ oscillations elicited by vasopressin in the intact liver was calculated in the absence and presence of ethanol in the perfusate, and the data expressed as a fold-change. Ethanol reduced the frequency of hormone-induced Ca2+ oscillations in a dose-dependent fashion in hepatocytes located in both the periportal and pericentral zones, with IC50 values of 1.4 and 2 mM, respectively (Figure 3A and B). In addition to its effects on the rates of Ca2+ spiking, ethanol also decreased the magnitude of vasopressin-induced [Ca2+]i increases. This effect is best seen in fura-2/AM loaded livers where the ratiometric method can be used to control for tissue movement and dye leakage (Figure 3C). The effect of ethanol on the amplitude of [Ca2+]i spikes was also observed in isolated hepatocytes (see below). The action of ethanol on hormone-induced Ca2+ oscillations was fully reversible (Figure 2C), and spatially organized intralobular Ca2+ waves reformed 1–2 min after commencing the washout of ethanol (Figure 2B, 70 s_w, 156 s_w). Increasing the concentration of vasopressin in the perfusate also overcame the inhibitory effects of ethanol on the frequency of Ca2+ oscillations (Figure 3D and E). However, even with increased vasopressin, the Ca2+ responses in the presence of ethanol were still not organized on a lobular scale, and intercellular Ca2+ waves propagated through only a small number of cells before terminating. These findings suggest that actions of ethanol on the Ca2+ oscillation machinery can be overcome by increasing the rate of IP3 formation, but the higher levels of IP3 are not capable of reversing the effects of ethanol on gap junction communication.

Figure 3.

Effect of Acute Ethanol Administration on Vasopressin-Induced Ca2+ Signaling in the Perfused Liver. (A) Single-cell [Ca2+]i trace showing the effects of increasing concentrations of ethanol on vasopressin-induced Ca2+ oscillations in the intact perfused liver. The concentrations and duration of vasopressin and ethanol perfusion are shown by the horizontal bars. (B) The frequency of [Ca2+]i spiking induced by vasopressin (30–100 pM) was determined in single hepatocytes within the periportal (PP, red) or pericentral (PC, green) zones (30–125 cells/zone/liver) in the absence and presence of ethanol in the perfusion buffer. The data are expressed as the fold-change for the indicated ethanol concentrations. Data were acquired from fura-2/AM or fluo-4/AM loaded livers (n = 17 livers). Symbols are the mean ± SD, or mean data for experiments with less than three livers (see also Table 1). (C–E) Fura-2/AM-loaded rat livers were stimulated with vasopressin (VP) followed by perfusion with ethanol plus hormone at the indicated concentrations. Fura-2 fluorescence images were deconvolved prior to calculating the ratio as described previously. (E) Summary data from the experiment shown in Panel D, depicting the frequency as Ca2+ spikes per minute for the indicated conditions. Rates of Ca2+ spiking were determined in individual hepatocytes from two different lobules. *Significantly different from other conditions; one-way ANOVA, Tukey post hoc test P < 0.01.

Table 1.

Effect of Acute Ethanol Treatment on the Frequency of Vasopressin-Induced Ca2+ Oscillations in the ex vivo Perfused Rat Liver

Ethanol (mM)	Periportal	Pericentral	n
0.3	0.97	0.88	1
1	0.83 ± 0.07	0.78 ± 0.06	5
10	0.57 ± 0.03	0.52 ± .03	3
20	0.54 ± 0.06	0.51 ± 0.03	8
50	0.47	0.41	2
300	0.50	0.44	1

The frequency of Ca2+ spiking was calculated before and after addition of the indicated concentrations of ethanol and expressed as a fold-change. Data are the means ± SD or the means if less than three livers were used in the experiment, n = number of liver preparations.

Ethanol Inhibits Gap Junction Communication in the ex vivo Perfused Liver

Gap junctions are one target for the inhibitory actions of ethanol and loss of intercellular communication could help to explain the disruption of intercellular Ca2+ waves (Figure 2B). Previous studies have demonstrated the ability to assess gap junction communication by monitoring the rates of FRAP in both isolated hepatocyte couplets and hepatocytes within the perfused liver. Here, ex vivo perfused rat livers were loaded with a Ca2+-insensitive indicator dye, calcein/AM, followed by a brief laser-induced photobleaching of a single hepatocyte or cluster of hepatocytes and fluorescence recovery monitored by confocal microscopy (Figure 4A). The FRAP assay was carried out before and then after perfusing the liver for 5–10 min with 20 mM ethanol in the same hepatic lobule, but using different hepatocytes for each assay to avoid laser-induced injury (Figure 4B). The initial rates of fluorescent recovery were determined, and summary data are presented in Figure 4C. The data indicate that ethanol perfusion markedly inhibits the rate of fluorescence recovery. These results are consistent with acute ethanol intoxication suppressing gap junction communication, which could explain the loss of coordinated propagating intercellular Ca2+ waves.

Figure 4.

Acute Ethanol Perfusion Suppresses Gap Junction Communication. (A) Confocal images (211 µm × 141 µm) of a calcein/AM loaded perfused rat liver. Images show the laser-induced photobleaching and recovery of the cell highlighted by the arrowhead. White scale bar denotes 20 µm. (B and C) Fluorescence recovery after photobleaching (FRAP) assays were carried out before and after perfusion with 20 mM ethanol. (B) Traces are the normalized fluorescence intensity changes from two different cells within the same hepatic lobule. (C) Summary data showing the initial rates of fluorescence recovery in the absence or presence of 20 mM ethanol. Data points are recovery rates from individual cells. Mean ± SD are also shown, n = 3 livers, *Paired t-test P < 0.05.

The Effects of Ethanol on Hormone-Induced Ca2+ Oscillations in Isolated Hepatocytes

The remaining experiments in this study were carried out in primary cultured hepatocytes, which maintain the hormone-stimulated Ca2+ oscillation properties described above for the intact liver. Consistent with the intact liver data, the addition of ethanol inhibited Ca2+ oscillations induced by either phenylephrine, an α1-adrengeric agonist, or vasopressin in fura-2/AM loaded hepatocytes (Figure 5A and B). In some experiments, hepatocytes were transfected with genetically encoded calcium indicators (GECIs) targeted to the cytosol or nucleus, and then maintained in primary culture overnight. We used GECIs for these experiments because these biosensors can be selectively targeted within the cell, and they are also relatively immobile Ca2+ binding proteins, thus less likely to perturb local calcium dynamics. The traces in Figure 5C and D show examples of vasopressin-induced Ca2+ oscillations measured with GCaMP6f fluorescence before and after addition of 30 mM ethanol. Ethanol addition caused a modest but persistent decrease in the peak height of GCaMP6f Ca2+ oscillations (Figure 5C) or a progressive loss in peak height until Ca2+ spiking ceased (Figure 5D). The amplitude of hormone-induced GCaMP6f spikes decreased an average of 13 ± 8% after the addition of ethanol (mean ± SD, n = 10 GCaMPf6 expressing hepatocytes from two cell preparations and four independent transfections, P < 0.01 paired t-test). Addition of a higher concentration of vasopressin reversed the effects of ethanol on both magnitude and frequency of cytosolic Ca2+ oscillations measured with GCaMP6f (Figure 5D). When added in the absence of hormone, ethanol (1–50 mM) had no effect on [Ca2+]i (not shown).

Figure 5.

Acute Ethanol Treatment Suppresses Hormone-Induced Ca2+ Oscillations in Isolated Hepatocytes. (A and B) Freshly-plated rat hepatocytes were loaded with fura-2/AM then stimulated with submaximal concentrations of phenylephrine (PE) or vasopressin (VP) where indicated by the arrowheads. After 5–10 min, the cultures were acutely treated with 30 mM ethanol. In separate studies, hepatocytes were transfected with GCaMPf6 (C and D) or ratiometric-pericam-nu (E and F) and then maintained in primary culture overnight. Traces are single-cell Ca2+ responses illustrating the effects of ethanol treatment on the magnitude of agonist-induced cytosolic (C and D) or nuclear (E and F) Ca2+ spikes.

We have previously demonstrated that intracellular Ca2+ waves propagate into the hepatic nucleus initiating a rise in nucleoplasmic [Ca2+]. Hepatocyte cultures were transfected with ratiometric-pericam-nu to investigate the effects of ethanol treatment on nuclear Ca2+ signals. Notably, the acute addition of ethanol decreased both the frequency and amplitude of vasopressin (Figure 5E) or phenylephrine (Figure 5F) stimulated Ca2+ oscillations in the nucleoplasm. The magnitude of vasopressin-induced nuclear Ca2+ spikes decreased an average of 11 ± 7% after the addition of 30 mM ethanol (mean ± SD, n = 6 hepatocytes from three different cell preparations, P < 0.05 paired t-test). Similarly, the nuclear Ca2+ spikes generated by α1-adrenergic stimulation were also suppressed by ethanol treatment, with a decrease in amplitude of 10 ± 4% (mean ± SD, n = 6 hepatocytes from two different cell preparations, P < 0.01 paired t-test).

Acute Ethanol Intoxication Impairs the Translocation of Conventional PKC

The activation of receptor-coupled PLC and subsequent spikes in [Ca2+]i are expected to drive the translocation of conventional isoforms of PKC from the cytosol to the membrane where the kinase binds and is further activated by DAG. In the next series of experiments, we investigated the actions of ethanol on the translocation of conventional PKCβII to the membrane. Hepatocyte cultures were transduced overnight with adenoviruses encoding enhanced green fluorescent protein (EGFP) tagged PKCβII (gift from R. Rizzuto). Live-cell confocal imaging was used to measure the distribution of GFP fluorescence in hepatocytes expressing PKCβII-EGFP before (Figure 6A, left) and after stimulating with 1 µM ATP (Figure 6A, right). Image analysis revealed that purinergic stimulation induced a decrease in GFP fluorescence in the perinuclear region and accumulation of the probe in a plasma membrane domain. We next measured hormone-induced [Ca2+]i responses and PKCβII translocation simultaneously in fura-2-loaded and PKCβII-EGFP expressing hepatocytes. A ratio was calculated between the reciprocal changes GFP fluorescence intensity in membrane versus the cytoplasm to evaluate the translocation of PKC. The traces in Figure 6B show the measurements of vasopressin-induced oscillations in [Ca2+]i and parallel translocation of PKCβII-EGFP to and from the membrane. The insert on the right depicts Ca2+ and PKC responses on an expanded time scale and normalized to baseline values for clarity. The data show that onset of the [Ca2+]i spike occurs prior to detectable changes in PKCβII-EGFP distribution, whereas the PKCβII-EGFP is still translocating to the membrane after the [Ca2+]i spike has peaked (Figure 6B). PKCβII-EGFP translocated back to the cytosol during the declining phase of the Ca2+ spike, but clearly lagged behind the Ca2+ signal (Figure 6B).

Figure 6.

Effect of Ethanol on Hormone-Induced Translocation of PKCβII. Hepatocyte primary cultures were transduced overnight with adenoviruses encoding EGFP-tagged PKCβII then used in confocal microscopy (A and D) or coloaded with fura-2/AM and used in wide-field fluorescence studies (B and C). (A) Representative confocal images showing translocation of PKCβII-EGFP induced by 1 µM ATP. White scale bar denotes 15 µm. (B) Simultaneous measurement of vasopressin-induced Ca2+ oscillations and PKCβII-EGFP translocation. Translocation of PKCβII was monitored by taking the ratio of the GFP fluorescence intensity changes in the membrane versus cytosol. Insert a single Ca2+ spike and translocation of PKCβII on an expanded time scale and normalized to the initial ratio and peak increase for clarity. (C) Effect of 30 mM ethanol on vasopressin-induced Ca2+ oscillations and translocation of PKCβII or (D) the translocation of PKCβII alone. The PKCβII membrane:cytosol ratio was normalized to the initial ratio. In Panel C, Ca2+ increases were induced by the addition of 1 nM vasopressin. (E and F) Summary data showing the AUC for hormone-induced [Ca2+]i spikes and the normalized PKCβ translocation ratio in the presence of vasopressin (VP) and after the acute addition of 30 mM ethanol (VP + ethanol). Data were calculated from 2 to 4 baseline-separated [Ca2+]i spikes or translocation of PKCβII before and after the addition of 30 mM ethanol. *Significantly different from hormone alone; P < 0.01, paired t-test. n = 17–22 cells from three independent hepatocyte preparations.

The addition of ethanol suppressed the magnitude and frequency of vasopressin-induced Ca2+ oscillations, and these effects translated into decreases in the extent and frequency of PKCβII-EGFP translocation to the membrane (Figure 6C). The actions of ethanol on hormone-induced PKCßII translocation were similar in both fura-2/AM-loaded and unloaded hepatocytes (Figure 6D). We calculated the peak vasopressin-induced increases in the fura-2 ratio and extent of PKC translocation, as well as the rates of rise, rates of fall, and AUC for both parameters, before and after ethanol treatment in the same cell. Summary results are shown in Figure 6E and F and Table 2. The acute ethanol intoxication significantly decreased the magnitude of both hormone-induced [Ca2+]i spikes and translocation of PKC to membrane (Figure 6E and F). Moreover, ethanol slowed down the kinetics of the Ca2+ spikes and the translocation of PKC (Table 2).

Table 2.

Effect of Ethanol Treatment on Vasopressin-Induced Calcium Increases and PKC Translocation

	Vasopressin	Vasopressin + 30 mM ethanol
ΔRatio (Ca2+)	0.89 ± 0.5	0.59 ± 0.3*
Ratio rise/s	0.13 ± 0.01	0.08 ± 0.06*
Ratio fall/s	−0.05 ± 0.03	−0.04 ± 0.02*
Ratio AUC	19.1 ± 12.4	10.4 ± 5.8*
ΔPKC	0.5 ± 0.4	0.23 ± 0.2*
PKC rise/s	0.04 ± 0.03	0.03 ± 0.04ns
PKC fall/s	−0.02 ± 0.02	−0.01 ± 0.01*
PKC AUC	17.7 ± 18.3	5.2 ± 3.4*

Primary cultured hepatocytes expressing PKCβII-EGFP were loaded with fura-2/AM and then stimulated with vasopressin to evoke baseline-separated [Ca2+]i oscillations. Ethanol (30 mM) was added 10–20 min after hormone stimulation. The data are the means ± SD for the peak increase in the fura-2 excitation ratio or PKC translocation, the rates of rise and fall in fura-2 ratio or PKC translocation, and the AUC for fura-2 ratio or PKC translocation. Data were calculated from 2 to 4 spikes before and after addition of ethanol in the same cell, n = 22 cells from three separate cell preparations.

Statistically different from vasopressin alone, paired t-test, P < 0.05. ns, not significant.

The Actions of Ethanol on Hormone-Induced Ca2+ Signaling Require Ethanol Metabolism

The main pathway for ethanol clearance in hepatocytes is the sequential oxidation by alcohol dehydrogenase (ADH) to produce acetaldehyde followed by the conversion of acetaldehyde into acetate in the mitochondrial matrix by aldehyde dehydrogenase (ALDH). The catabolism of ethanol causes a marked reduction of NAD+ to NADH, leading to a rise in the NADH/NAD+ redox ratio. This stimulates the production of reactive oxygen species (ROS),, which can potentiate the opening of IP3 receptors., Hepatocyte cultures were incubated for 30 min in the presence of 4-methylpyrazole, an alcohol dehydrogenase inhibitor, or cyanamide, an aldehyde dehydrogenase inhibitor, prior to the addition of submaximal concentrations of phenylephrine. Cyanamide is a prodrug that requires bioconversion into the active compound necessitating the preincubation step.

The frequency of phenylephrine-induced Ca2+ oscillations was reduced in a dose-dependent manner with increasing concentrations of ethanol in primary cultured hepatocytes (Figure 7B, red symbols). Pretreating hepatocytes with 4-methylpyrazole to block ADH prevented the ethanol-induced increase in NADH fluorescence, consistent with the inhibition of ethanol metabolism (Figure 7C). This inhibition of the key first step of ethanol metabolism in liver largely ablated the effects of low ethanol concentrations on hormone-induced Ca2+ oscillations (Figure 7A). The dose-response for ethanol inhibition of Ca2+ oscillations in the presence of 4-methylpyrazole (Figure 7B, green symbols) shows a substantial right-shift compared to ethanol alone. At higher concentrations of ethanol (≥30 mM) there was a residual effect to reduce the frequency of agonist-induced Ca2+ oscillations, which may reflect a direct effect of ethanol or the possibility that saturating concentrations of ethanol can partially overcome the competitive block of 4-methylpyrazole. The finding that the acute effects of ethanol on hormone-stimulated Ca2+ signaling depend, at least in part, on ethanol metabolism via ADH is an important finding that can point to potential mechanisms.

Figure 7.

Ethanol Metabolism Is Required for the Inhibitory Actions of Ethanol on Ca2+ Oscillations. (A–C) Hepatocytes were pretreated for 30 min with 500 µM 4-methylprazole (4-MP), 100 µM cyanamide, or buffer prior to data acquisition. (A) Representative single-cell Ca2+ trace showing the lack of effect of ethanol on Ca2+ spiking in the presence of 4-MP. (B) Summary data showing the effects of increasing concentrations of ethanol (0.1–100 mM) on the frequency of phenylephrine (PE)-induced Ca2+ oscillations. The changes in frequency are expressed as a fold-change for each cell analyzed. Data points are the means ± SEM, n = 3–12 separate experiments carried out in 20 different hepatocytes preparations with 25–75 cells per measurement. (C) The increases in hepatic NAD(P)H fluorescence in response to ethanol (100 mM) were determined in the absence (left) or presence (right) of 4-MP. NAD(P)H fluorescence was normalized to the initial intensity and to the maximal fluorescence intensities recorded after the addition of rotenone plus ß-hydroxybutyrate; n = 50–100 cells/assay. (D and E) Hepatocytes incubated with N,N′-dimethylthiourea (1 mM, DMTU) or N,N′-diphenyl 1,4-phenylenediamine (5 µM, DPPD) for 60–90 min prior to data acquisition. Traces are single-cell Ca2+ responses depicting the effects of ethanol on Ca2+ spiking in the presence of the indicated antioxidants.

We further investigated the role of ethanol metabolism by inhibiting the clearance of the ethanol metabolite acetaldehyde with cyanamide. In contrast to the inhibition of ADH with 4-methylpyrazole, ALDH inhibition potentiated the effects of ethanol on agonist-induced Ca2+ oscillations (Figure 7B blue symbols). The IC50 values for ethanol-induced inhibition of Ca2+ oscillations were 5 mM, 79 mM or 0.8 mM in the presence of ethanol alone, 4-methylpyrazole plus ethanol or cyanamide plus ethanol, respectively. In a separate set of experiments, hepatocyte cultures were pretreated with the antioxidants N,N'-dimethylthiourea (1 mM, DMTU) or N,N′-diphenyl 1,4-phenylenediamine (5 µM, DPPD). DMTU has been shown previously to suppress ROS production and apoptosis in hepatocytes during acute ethanol intoxication,, while the addition of DPPD inhibits oxidant-induced increases in lipid peroxidation., The data show that neither DMTU nor DPPD pretreatment blocked the inhibitory effects of ethanol on agonist-induced Ca2+ oscillations (Figure 7D and E). Moreover, treatment with Trolox (100 µM), a water-soluble vitamin E analog, or DMSO (100 mM) at concentrations that scavenge hydroxyl radicals were also ineffective at blocking the effects of ethanol on Ca2+ signaling. The fold-change in the frequency of phenylephrine-induced Ca2+ oscillations was 0.51 ± 0.2 in the presence of ethanol alone, while in cells treated with DMTU, DPPD, Trolox, or DMSO prior to the addition of ethanol, the fold-change was 0.38 ± 0.1, 0.37 ± 0.2, 0.43 ± 0.1, or 0.56 ± 0.1 respectively (mean ± SD, n = 2–4 separate hepatocyte preparations). These data indicate that ethanol catabolism is required to suppress IP3-dependent Ca2+ signals and implicate acetaldehyde, but not the formation of ROS, as the underlying effector.

Ethanol Does Not Perturb IP3 Receptor Function, but Inhibits the Hormone-Stimulated Formation of IP3

To assess the effects of acute ethanol treatment on IP3 receptor function, cultured hepatocytes were coloaded with fluo-4/AM and caged-IP3, and then exposed to a series of UV flashes to photorelease IP3. Photolysis of caged IP3 produces a variety of [Ca2+]i increases in hepatocytes ranging from no response, through to single Ca2+ spikes, repetitive Ca2+ oscillations, to maximal peak and plateau types of Ca2+ signals., Importantly, we have previously shown that the Ca2+ spikes produced by uncaging IP3 are predominately mediated by the intrinsic properties of the IP3 receptors with little to no contribution from receptor-coupled PLC activity., The addition of 30 mM ethanol did not qualitatively alter the types of Ca2+ responses induced by uncaging IP3 compared to control (Figure 8A), nor did ethanol treatment affect the amplitude or width of IP3-dependent Ca2+ spikes (Figure 8B and C). Moreover, the addition of 100 µM acetaldehyde to digitonin-permeabilized hepatocyte suspensions did not affect Ca2+-uptake into internal stores (not shown) or the amount of Ca2+ mobilized by IP3. The addition of 500 nM IP3 released 2.3 ± 0.2 nmoles Ca2+/mg protein compared to 3.3 ± 1.0 nmoles Ca2+/mg protein in the absence and presence of acetaldehyde, respectively (mean ± SD, n = 3 hepatocyte preparations, ns). These data suggest that these pharmacologically relevant concentrations of ethanol or its metabolite, acetaldehyde, do not alter the properties of IP3 receptors.

Figure 8.

Ethanol Treatment Does Not Affect IP3 Receptor Sensitivity but Suppresses Hormone-Induced Production of IP3. Rat hepatocytes were coloaded with cell-permeant caged-IP3 and fluo-4/AM (A–C). IP3 was uncaged with 4 UV pulses from a nitrogen-charged UV laser in the absence or presence of 30 mM ethanol (5 min pretreatment). (A) Summary data showing the categories of Ca2+ responses observed after uncaging IP3. The effects of ethanol treatment on the amplitude (B) and width (C) of Ca2+ spikes induced by uncaging IP3. Data points are the means of 2–3 replicate determinations carried out for each hepatocyte preparation (n = 3). (D–G) Hepatocytes transfected overnight with IRIS-1 were loaded with Indo-1/AM and then stimulated with vasopressin (1–3 nM). (D) Representative traces showing the simultaneously measurement of vasopressin-induced increases in [Ca2+]i (red trace) and IRIS-1 emission ratio (blue trace) (E) Vasopressin-induced increases in the IRIS-1 emission ratio were aligned to the rising phase of the spike. Traces are the averaged responses calculated from three IRIS-1 spikes before and after the addition of ethanol (30 mM). (F and G) Summary data showing the rates of rise and AUC for vasopressin-induced spikes in the IRIS-1 emission ratio before (VP) and after the addition of 30 mM ethanol (VP + ethanol). Data were calculated from 2 to 4 IRIS-1 spikes before and after addition of ethanol in the same cell. *Significantly different from vasopressin alone; P < 0.01, paired t-test. n = 10 cells from two different hepatocyte preparations and three independent transfections.

We initially utilized biochemical assays to investigate the actions of ethanol on hormone-stimulated PLC activity in cultured hepatocytes. These assays were unable to detect any differences between vasopressin-stimulated increases in IP3 mass or in the accumulation of [H3]-inositol phosphates in the absence or presence of 30 mM ethanol, which may reflect the relatively low sensitivity of these assays at physiological hormone concentrations that give rise to Ca2+ oscillations. As an alternative approach, hepatocytes were transfected overnight with a recombinant IP3-sensitive biosensor IRIS-1 and then loaded with the Ca2+ indicator dye Indol-1/AM. We have previously used this protocol to demonstrate that there are cross-coupled oscillations of IP3 and Ca2+ during hormone stimulation of hepatocytes, and that these reflect a requirement for positive feedback of Ca2+ on PLC activity to generate the hormone-induced Ca2+ oscillations. As reported previously, the onset of each Ca2+ and IP3 spike occurs in parallel, whereas the declining phase of each IP3 spike lags slightly behind the concomitant Ca2+ spike (Figure 8D). In order to increase the signal to noise of the IRIS-1 FRET ratio, IP3 spike traces were averaged from 2 to 4 sequential oscillations aligned to the rising phase of the individual spikes. This allowed quantitation of the IP3 spike kinetics following hormone stimulation before (Figure 8E, left) and after (Figure 8E, right) the addition of 30 mM ethanol. Ethanol treatment decreased the amplitude of the vasopressin-stimulated IP3 oscillations, slowed the rates of rise and fall of the individual IP3 spikes and decreased the AUC (Figure 8F and G; Table 3). Taken together, our findings suggest that ethanol inhibits receptor-coupled PLC activity which, in turn, suppresses the cross-coupling between IP3 and Ca2+ dynamics that initiates and drives the rapid rising phase of the oscillatory Ca2+ spikes.

Table 3.

Effect of Ethanol Treatment on Vasopressin-Induced Increases in IRIS-1 Emission Ratio

	Vasopressin	Vasopressin + 30 mM Ethanol
ΔIRIS-1	8.5e-3 ± 4.7e-3	5.9e-3 ± 4e-3*
Rise/s	5.0e-4 ± 3e-4	3.7e-4 ± 3e-4*
Fall/s	−3.3e-4 ± 2e-4	−2.7e-4 ± 2e-4ns
AUC	0.33 ± 0.24	0.18 ± 0.14*

Hepatocytes expressing IRIS-1 were loaded with Indo-1/AM and then stimulated with vasopressin (VP, 1–3 nM) to evoke baseline separated Ca2+ oscillations. Ethanol (30 mM) was added 10–20 min after vasopressin stimulation. The data are the means ± SD for the peak VP-induced increases the IRIS-1 emission ratio, the rates of rise and fall of the IRIS-1 increase, and the AUC for the IRIS-1 spike. Data were calculated from 2 to 4 IRIS-1 spikes before and after addition of ethanol in the same cell, n = 10 cells from two different cell preparations and three independent transfections. *Statistically different from vasopressin alone, paired t-test, P < 0.05. ns, not significant.

Discussion

The natural progression of alcohol-related liver disease (ALD) follows a well-characterized pattern. Ethanol consumption initially induces steatosis or fatty liver over the course of a few days to several weeks of heavy drinking and this phenotype can manifest in virtually all alcohol drinkers. Alcohol-induced fatty liver can progress onto liver inflammation and fibrosis in a subset of alcoholics, but this usually requires decades of excessive alcohol consumption. Despite the overt toxicity of ethanol and its metabolite, acetaldehyde, the factors driving the progression from the early stages of alcoholic fatty liver, which are reversible, to more advanced stages of liver injury remain debatable and areas of active research., Ethanol is a relatively nonspecific drug with pleiotropic effects on most organs in the body and, thus ethanol consumption can initiate multiple mechanisms, in both the liver and extrahepatic tissues, that can potentially contribute to the development of liver injury., These mechanisms include ethanol-dependent changes in the gut microbiome and disruption of gut barrier function, adaptive responses in hepatic ethanol metabolism, hyperactivation of lipolysis in adipose tissue, acute pancreatitis, stimulation of the innate immune system, and alterations in the composition and secretion of exosomes. These changes can combine with environmental factors such as the consumption of dietary unsaturated fats to produce a multifaceted and complex pathology. We have proposed that enhanced receptor-coupled PI-PLC activity induced by long-term and repeated exposure to ethanol may be another mechanism contributing to hepatic injury,; the inappropriate activation of PLC-linked signaling cascades could underlie some of the pleiotropic actions of ethanol on liver function.

In this study, we investigated both the acute effects of ethanol on [Ca2+]i and its actions on hormone-induced Ca2+ signaling. Our studies using an ex vivo perfused liver model indicate that concentrations of ethanol up to and above the legal intoxication limit (20–30 mM) do not have a direct effect on [Ca2+]i. Ethanol in the perfusate had to be increased to 100 mM or more before inducing Ca2+ spikes in a small percentage of cells in the intact liver. The biological significance of these ethanol-induced Ca2+ responses is not readily clear. Blood alcohol concentrations in excess of 100 mM are not observed in healthy volunteers given ethanol,, however, there are reports of alcoholics presenting at emergency departments with plasma ethanol concentrations between 100 and 170 mM,, as well as a single report of an alcoholic who survived a blood alcohol level of 300 mM. Thus, although high concentrations of ethanol can directly activate the G-protein-coupled PLC to generate IP3 and elicit Ca2+ signals in the liver,,,, this probably only occurs after bouts of excessive binge drinking in chronic alcoholics. In this context, we have previously demonstrated that chronic ethanol consumption enhances the efficacy of hormones to stimulate Ca2+ signaling by sensitizing receptor-coupled PLC activity in liver, which may be a compensatory mechanism for the acute inhibitory effects of ethanol described here. If such a compensatory mechanism also increases the efficacy of ethanol to mobilize internal Ca2+ stores, then an acute binge on top of chronic ethanol consumption could produce a prolonged and sustained rise in [Ca2+]i that could cause liver injury. It is currently unknown if chronic ethanol consumption or feeding can modify the Ca2+ responses induced by acute ethanol treatment.

Bearing in mind the range of blood alcohol levels normally encountered, the more relevant effects of ethanol described here are to suppress hormonal signaling. Acute ethanol intoxication has been reported to affect multiple targets that could contribute to the inhibition of phosphoinositide-dependent Ca2+ signals in rat hepatocytes., However, these previous experiments were carried out predominately with high concentrations of ethanol (>100 mM), raising questions of whether these targets are also affected at pharmacologically achievable concentrations of ethanol. Our data show that low doses of ethanol potently suppress hormone-induced Ca2+ oscillations and the parallel translocation of conventional PKCß to the membrane. The half-maximal inhibitory concentration of ethanol on Ca2+ oscillations was in the low millimolar range, which is similar to the reported Kms of the two main enzymes that metabolize ethanol to acetaldehyde in liver: ADH (Km ≈2 mM) and CYP2E1 (Km ≈10 mM). By using 4-methylpyrazole, which is a potent inhibitor of both ADH and CYP2E1, we showed that the inhibitory effects of ethanol on agonist-induced Ca2+ oscillations are largely dependent on ethanol metabolism. The effects of 10 mM ethanol and below were completely blocked by 4-methylpyrazole, and this compound also significantly reduced the inhibitory effects of ethanol up to 100 mM (Figure 7B). The residual inhibitory effect of high ethanol concentrations could reflect a direct action of ethanol on IP3-dependent Ca2+ signaling, or it could be because 4-methylpyrazole does not completely block ethanol oxidation by CYP2E1.

Another important effect of ethanol observed in our ex vivo perfused rat liver studies was the marked reduction in cell-to-cell communication in the presence of ethanol. This clearly underlies the suppression of hepatocyte-to-hepatocyte propagation of Ca2+ signals in the intact perfused liver, and hence to the disruption of the organization and coordination of intercellular Ca2+ waves during agonist perfusion in the presence of ethanol. Moreover, the reduction in cell-cell communication will reduce integrated lobular signaling, which we recently reported plays a key role in generating the global response of the liver to glycogenic hormones. We have shown previously that the coordination of lobular Ca2+ waves during hormonal stimulation of the liver is dependent on coupling of cell signaling through gap junctions, and this study shows that acute ethanol interferes with this coupling. This is consistent with previous work showing the inhibitory effects of ethanol on gap junctions, and significantly, this effect was also reported to require ethanol metabolism.

Taken together, the findings discussed above indicate that the inhibitory actions of pharmacologically relevant doses of ethanol on hepatic Ca2+ signaling are mediated by metabolically derived acetaldehyde, and are most likely not a direct effect of ethanol. It should be noted that ethanol also reacts with long-chain fatty acids to produce fatty acid ethyl esters (FAEEs) through a nonoxidative pathway catalyzed by fatty acid ethyl ester synthase. The nonoxidative production of FAEEs is thought to be the primary mechanism whereby ethanol induces alcoholic pancreatitis., These studies indicate that pancreatic acinar cell injury is caused by a sustained rise in [Ca2+]i triggered by FAEE dependent Ca2+ release from the ER and store-operated Ca2+ entry.,, The role of FAEEs in alcoholic liver disease is less clear. The activity of fatty acid ethyl ester synthase in rat tissues is ≈60-fold higher in the pancreas compared to liver. Moreover, in human hepatoma cell lines, nonoxidative production of FAEEs accounts for <1% of total ethanol metabolism, and the detection of cellular FAEEs above baseline values requires high concentrations of ethanol (80 mM) and long incubation periods (hours)., These studies suggest that FAEEs may not be involved in, at least, the acute effects of ethanol on Ca2+ signaling in hepatocytes. Nevertheless, it is possible that this mechanism could play a role in liver damage in chronic alcoholics. Our studies show acetaldehyde is responsible for impairing Ca2+ signaling, but do not address how acetaldehyde can affect receptor-coupled PLC activity. Acetaldehyde can form both stable and unstable adducts with hepatic proteins, particularly on lysine residues,, and such acetaldehyde-adducts could modify or impair the function of the receptor-stimulated Ca2+ signaling machinery.

We also considered ethanol-induced increases in cellular NADH or the production of ROS as potential mediators of ethanol action on Ca2+ signaling. Several lines of evidence that argue against a role for NADH: 1) ethanol-induced increases in NADH are maximal at ≈2 mM in the perfused liver (not shown, but see Kashiwagi et al.), whereas the effects of ethanol on agonist-induced Ca2+ oscillations are not (Figures 3B and 7B); 2) an increase in cytosolic NADH levels does not appear to affect the activity of PLC; and 3) inhibiting ALDH enhances the effects of ethanol on Ca2+ oscillations while decreasing NADH formation (Figure 7B). Acute ethanol intoxication has been reported to induce the formation of ROS in mitochondria. However, the effects of ethanol on agonist-induced Ca2+ signaling were not prevented by the ROS scavengers DPPD and DMTU. Notably, production of mitochondrial ROS and cell death induced by the reoxygenation of anoxic hepatocytes has been shown to be markedly suppressed by DPPD. Our data show that similar concentrations of DPPD cannot block or reverse the actions of ethanol on agonist-induced Ca2+ oscillations (Figure 7E). These data indicate that ethanol metabolism to increase NADH and stimulate ROS formation does not play a significant role in the short-term effects of ethanol on agonist-induced Ca2+ oscillations.

In conclusion, our data show that pharmacologically relevant concentrations of ethanol suppress receptor-stimulated PLC activity and slow down the associated oscillatory increases in both [IP3] and [Ca2+]i. We have previously demonstrated that rapid dynamic changes in IP3 levels are required to produce the repetitive, large-amplitude Ca2+ oscillations induced by PLC-linked hormones. By decreasing the rate of rise and the peak increase in [IP3], ethanol partially uncouples the positive feedback between IP3 and Ca2+ during the rising phase of the Ca2+ spike, leading to early termination of the [Ca2+]i increase or complete cessation of Ca2+ spiking. This idea is consistent with the observation that increasing the concentration of hormone, and presumably also the rates of IP3 production, reverses the effects of ethanol on the frequency of Ca2+ oscillations and amplitude of the Ca2+ spike (Figures 3 and 5). In the short term, the acute inhibitory effects of ethanol on receptor-coupled PLC will perturb the normal hormonal regulation of hepatic metabolism. However, after long-term ethanol exposure, the liver overcomes the acute inhibitory actions of ethanol on IP3 production by increasing the efficacy of hormones to stimulate PLC activity. This adaptive response would enable hepatocytes to respond to Ca2+-mobilizing hormones and increase metabolic output even in the presence of high levels of circulating ethanol. However, when blood alcohol levels fall during times of sobriety this compensatory mechanism could lead to the overstimulation of receptor-coupled PLC and hyperactivation of downstream Ca2+ and PKC-sensitive targets.