Ethanol Suppresses Ureagenesis in Rat Hepatocytes

Background: Ethanol oxidation alters mitochondrial metabolism in rat hepatocytes. Results: Ethanol via acetaldehyde formation closes mitochondrial voltage-dependent anions channels (VDAC) and decreases outer membrane permeability. Conclusion: Acetaldehyde, derived from ethanol oxidation, suppresses ureagenesis by inhibiting exchange of mitochondrial substrates across the outer membrane. Significance: Acetaldehyde-mediated VDAC closure is an early adaptive event in ethanol metabolism that may contribute to alcoholic liver disease. We proposed previously that closure of voltage-dependent anion channels (VDAC) in the mitochondrial outer membrane after ethanol exposure leads to suppression of mitochondrial metabolite exchange. Because ureagenesis requires extensive mitochondrial metabolite exchange, we characterized the effect of ethanol and its metabolite, acetaldehyde (AcAld), on total and ureagenic respiration in cultured rat hepatocytes. Ureagenic substrates increased cellular respiration from 15.8 ± 0.9 nmol O2/min/106 cells (base line) to 29.4 ± 1.7 nmol O2/min/106 cells in about 30 min. Ethanol (0–200 mm) suppressed extra respiration after ureagenic substrates (ureagenic respiration) by up to 51% but not base line respiration. Urea formation also declined proportionately. Inhibition of alcohol dehydrogenase, cytochrome P450 2E1, and catalase with 4-methylpyrazole, trans-1,2-dichloroethylene, and 3-amino-1,2,3-triazole restored ethanol-suppressed ureagenic respiration by 46, 37, and 66%, respectively. By contrast, inhibition of aldehyde dehydrogenase with phenethyl isothiocyanate increased the inhibitory effect of ethanol on ureagenic respiration by an additional 60%. AcAld, an intermediate product of ethanol oxidation, suppressed ureagenic respiration with an apparent IC50 of 125 μm. AcAld also inhibited entry of 3-kDa rhodamine-conjugated dextran in the mitochondrial intermembrane space of digitonin-permeabilized hepatocytes, indicative of VDAC closure. In conclusion, AcAld, derived from ethanol metabolism, suppresses ureagenesis in hepatocytes mediated by closure of VDAC.

We proposed previously that closure of voltage-dependent anion channels (VDAC) in the mitochondrial outer membrane after ethanol exposure leads to suppression of mitochondrial metabolite exchange. Because ureagenesis requires extensive mitochondrial metabolite exchange, we characterized the effect of ethanol and its metabolite, acetaldehyde (AcAld), on total and ureagenic respiration in cultured rat hepatocytes. Ureagenic substrates increased cellular respiration from 15. 8  Balanced and coordinated functioning of hepatocytes depends on optimized exchange of metabolites between mitochondria and the cytosol, in particular the continuous exchange of hydrophilic metabolites like ATP, ADP, inorganic phosphate, and respiratory substrates. To enter the mitochondrial matrix, where most mitochondrial metabolism occurs, all metabolites must cross two separate barriers: the mitochondrial outer membrane (MOM) 4 and mitochondrial inner membrane (MIM). Although the MIM transports metabolites via a variety of selective transporters (1)(2)(3), the MOM conducts hydrophilic solutes non-selectively via a channel called mitochondrial porin or the voltage-dependent anion channel (VDAC) (4,5). The VDAC is the pathway across the MOM for entry and exit of virtually all water-soluble metabolites involved in oxidative phosphorylation and related metabolism (4 -8).
Exceptions are hydrophobic metabolites, such as molecular oxygen and short chain fatty acids, that directly permeate biological membranes. The VDAC is often considered to be constitutively open, and its potential importance in regulating mitochondrial metabolism has been largely ignored.
In hepatocytes, ureagenesis comprises a sequence of five biochemical reactions distributed between two intracellular compartments: the mitochondrial matrix and cytosol (26 -30). The compartmental separation of reactions of the urea cycle requires exchange of ornithine and citrulline, major substrates of the urea cycle, between the mitochondria and cytosol. In addition, ureagenesis is a highly energy-demanding process that requires hydrolysis of four equivalents of ATP (two each in the matrix and cytosol) per mol of urea formed (29). Overall, ureagenesis requires continuous exchange of hydrophilic metabolites between the cytosol and mitochondria, all of which must go through the VDAC. Thus, if the VDAC were to adopt a more closed state, ureagenesis would likely be one of the first metabolic processes affected.
AcAld, the toxic intermediate of ethanol metabolism, is a small neutral compound that does not require carriers or channels to cross mitochondrial membranes and enter the matrix from the cytosol. We proposed recently that the VDAC closes in response to ethanol (6,9,10,31). VDAC closure suppresses oxidation of respiratory substrates competing with AcAld to promote selective and more rapid AcAld oxidation. Here, we demonstrate that ethanol dose-dependently suppresses ureagenesis in cultured rat hepatocytes and that AcAld mediates this inhibitory effect apparently by closing the VDAC and decreasing MOM permeability.

EXPERIMENTAL PROCEDURES
Animals-All animal protocols were approved by the Institutional Animal Care and Use Committee in accordance with recommendations published in the Guide for the Care and Use of Laboratory Animals, National Academic Press, Washington, DC, 1996 (32).
Measurement of Hepatocyte Respiration-The rate of oxygen consumption (respiration) was measured in hepatocytes plated on Seahorse 24-well V7 plates (Seahorse Biosciences, Billerica, MA). Prior to plating cells, wells were coated (20 l/well) with 0.1% type-1 rat tail collagen dissolved in 0.1% acetic acid (Sigma). After drying, wells were rinsed once with 0.5 ml of warm Waymouth's medium. Aliquots of 0.1 ml of isolated hepatocyte suspension were transferred to each well (3 ϫ 10 4 cells/ well at a density of 10 5 cells/cm 2 ) and allowed to attach for 2 h in humidified 5% CO 2 /95% air at 37°C. Subsequently, the wells were supplemented with 0.4 ml of warm complete growth medium, and the hepatocytes were incubated for 16 -18 h in humidified 5% CO 2 /95% air at 37°C.
Rates of oxygen consumption (respiration) were measured in a Seahorse Bioscience XF-24 extracellular flux analyzer using 24-well plate sensor cartridges. The parameters of the measurement cycle were adjusted to avoid anoxia and were set as followings: Mix (4 min, during which the sensor plate moved up and down to equilibrate oxygen), Wait (30 s, during which oxygen sensors descended into wells to form 7-l measurement chambers above the plated hepatocytes), and Measure (1 min, 10 measurements of oxygen concentration to calculate respiratory rates) (35). Respiratory rates were calculated using Seahorse XF-24 software and the Direct ACOS fast algorithm with continuous averaging, as described (35). All incubations and measurements were done at 37°C. Prior to measurement, hepatocytes were rinsed three times with 0.45 ml of Krebs-Ringer-HEPES buffer (KRH) containing 115 mM NaCl, 5 mM KCl, 1 mM CaCl 2 , 1 mM KH 2 PO 4 , 1.2 mM MgSO 4 , 25 mM HEPES buffer (pH 7.4, with KOH) and supplemented with 27 mM NaHCO 3 (KRH-NaHCO 3 ). Hepatocytes were incubated for 30 min in a cell culture incubator in 5% CO 2 /air before transferring to a Seahorse XF-24 extracellular flux analyzer. Injection ports were filled with 50 l of KRH-NaHCO 3 supplemented with ureagenic substrates with or without ethanol or AcAld (port A) and inhibitors (port B) as specified. Respiratory rates were expressed as mol O 2 /min/10 6 cells or as percentage of base line. Oxygen consumed by hepatocytes during ureagenesis was determined by an area-under-the-curve calculation with subtraction of oxygen consumed in the absence of ureagenic substrates (base-line respiration).
Measurement of Urea Synthesis-Urea in dried aliquots of incubation medium was determined using a QuantiChrom urea assay kit according to the manufacturer's instructions (Bioassay System, Hayward CA).
Confocal images of green-fluorescing MTG and red-fluorescing RhDex were collected at each step and analyzed with ImageJ 1.43u using a customized macro algorithm. Briefly, the macro used green MTG fluorescence of mitochondria to create a "mask," which then was used to measure the intensity of red RhDex fluorescence only in areas inside mitochondria, a procedure excluding cytosolic and nuclear fluorescence outside the mask. Mean intensity of red RhDex fluorescence in MTG-labeled mitochondrial regions was taken as a measure of RhDex content in mitochondria.
Statistics-Data are presented as mean Ϯ S.E. Images shown are representative of three or more experiments. Statistical analysis was performed by Student's t test or analysis of variance using p Ͻ 0.05 as the criterion of significance.

Ethanol Suppresses Ureagenic Respiration of Cultured
Hepatocytes-Respiration by cultured rat hepatocytes was measured in a Seahorse XF24 extracellular flux analyzer and averaged 15.8 Ϯ 0.9 nmol O 2 /min/10 6 cells under basal conditions (Fig.  1A). After addition of ureagenic substrates (3 mM NH 4 Cl, 5 mM L-ornithine, and 5 mM L-lactate), respiration of cells increased to 29.4 Ϯ 1.7 nmol O 2 /min/10 6 cells within about 30 min (Fig.  1A). Ureagenic respiration of hepatocytes, defined as the difference between total respiration after addition of ureagenic substrates and basal respiration, increased to 13.6 Ϯ 0.5 nmol O 2 /min/10 6 cells (Fig. 1B). To investigate the effect of ethanol on ureagenic respiration, ethanol was added simultaneously with ureagenic substrates. Under these conditions, ethanol dose-dependently suppressed ureagenic respiration (Fig. 1A), and 200 mM ethanol decreased ureagenic respiration to 6.6 Ϯ 1.3 nmol O 2 /min/10 6 cells (B). Ethanol added to cultured hepatocytes in the absence of ureagenic substrates had no effect on basal respiration (data not shown).
Ethanol Suppresses Urea Formation in Cultured Hepatocytes-We also measured urea formation during measurements of respiration. In the absence of ureagenic substrates, urea formation was below the limits of detection (Ͻ 0.2 nmol urea/10 6 cells). After addition of ureagenic substrates, urea formation increased to 9.9 Ϯ 1.4 nmol urea/10 6 cells ( Fig. 2A). In these experiments, the total amount of ureagenic respiration averaged 12.1 Ϯ 2.5 nmol O 2 /10 6 so that the stoichiometry of urea formed to oxygen atoms consumed was 0.4, which is equivalent to an ATP/O ratio of 1.6, assuming four ATP consumed for each urea formed. Suppression of ureagenic respiration by ethanol led to a proportional inhibition of urea formation so that the ratio of urea formation to ureagenic respiration remained more or less constant (Fig. 2B). Thus, ethanol suppressed both ureagenic respiration and urea formation but did not alter the energetic efficiency of ureagenesis.
Contribution of Cytosolic Alcohol Dehydrogenase to Ethanolinduced Suppression of Ureagenic Respiration-To assess the role of ethanol metabolism via ADH in ethanol-induced suppression of ureagenic respiration, hepatocytes were pretreated before ethanol addition with 4-methylpyrazole (MP, 1 mM), an ADH inhibitor (37). In the absence of ethanol, MP did not alter ureagenic respiration (Fig. 3A, Saline). In contrast, when hepa- tocytes were subsequently treated with ethanol (100 mM), MP partially protected against ethanol-mediated suppression of ureagenic respiration and attenuated suppression by 46% after 100 mM ethanol (Fig. 3A, Ethanol).
Contribution of Catalase to Ethanol-induced Suppression of Ureagenic Respiration-To assess the role of catalase in ethanol-induced suppression of ureagenic respiration, hepatocytes were pretreated with 3-amino-1,2,4-triazole (AT, 5 mM), a catalase inhibitor (39), in the absence of ethanol. Like MP and DCE, AT had no effect on ureagenic respiration in the absence of ethanol (Fig. 3C, Saline). Also like MP and DCE, AT partially reversed the inhibitory effect of 100 mM ethanol and attenuated ethanol-induced suppression of ureagenic respiration by 66% (Fig. 3C, Ethanol).
Inhibition of Aldehyde Dehydrogenase Enhances Ethanol-induced Suppression of Ureagenic Respiration-AcAld, the product of the first step of ethanol oxidation, is further metabolized to acetate predominantly by mitochondrial ALDH2. Our finding that inhibitors of ADH, CYP450 2E1, and catalase abrogated, at least in part, ethanol-induced suppression of ureagenic respiration suggested that AcAld might be mediating ureagenic inhibition. Accordingly, we assessed the effect of phenethyl isothiocyanate (10 M), an inhibitor of ALDH (23), on ethanol-induced suppression of ureagenic respiration. In the absence of ethanol, phenethyl isothiocyanate did not inhibit ureagenic respiration (Fig. 4, Saline), but in the presence of subsequently added ethanol, phenethyl isothiocyanate enhanced inhibition of ureagenic respiration by 47% (Ethanol).
Acetaldehyde Alone Suppresses Ureagenic Respiration of Cultured Hepatocytes-Abrogation of ethanol-induced suppression of ureagenic respiration by ADH, CYP450 2E1, and catalase inhibition and enhancement by ALDH inhibition supports the conclusion that AcAld produced during ethanol oxidation mediates suppression of ureagenic respiration. Accordingly, we investigated the effect of directly added AcAld on ureagenic respiration. Because of the high volatility of AcAld, we layered mineral oil (250 l) over the incubation medium in the Seahorse plates to minimize evaporation of AcAld. Mineral oil did  not interfere with measurements of oxygen consumption (data not shown) but substantially decreased the rate of loss of AcAld from Ͼ 90% in open wells after 60 min to 40 -45% in wells covered with mineral oil (Fig. 5, compare Open with Mineral Oil). The effect of added AcAld on ureagenic respiration was then evaluated using the same protocol as for ethanol, except that the incubation medium bathing hepatocytes in each well of Seahorse plate was covered with 250 l of mineral oil. In these measurements, AcAld caused dose-dependent inhibition of ureagenic respiration (Fig. 6, A and B). The K i for inhibition of ureagenic respiration was 125 M (uncorrected for evaporative loss), as estimated from a Dixon plot (Fig. 6B, inset). In contrast to ureagenic respiration, 125 M AcAld did not inhibit uncoupled respiration stimulated by 2,4-dinitrophenol (150 M) (Fig.  6C, compare Uncoupled with Ureagenic). To determine whether AcAld increases after ethanol to a level that could account for ethanol-dependent suppression of ureagenic respiration, we measured AcAld in the medium of cultured hepatocytes exposed to ethanol for 90 min. After 60 min of incubation with 50 mM ethanol, AcAld in the incubation buffer increased to 44.6 Ϯ 0.8 M (n ϭ 2).

Acetaldehyde Suppresses Entry of Fluorescent 3-kDa Rhodamine-conjugated Dextran into the Intermembrane Space of Mitochondria of Digitonin-permeabilized Hepatocytes-Previ-
ously, we showed that ethanol inhibits entry of 3 kDa RhDex into the IMS of MTG-labeled mitochondria in permeabilized hepatocytes (9). Here, we used a similar approach to study the effect of AcAld on RhDex entry into the IMS. Intracellular mitochondria were first labeled with MTG (Fig. 7A, MTG), and then hepatocytes were treated with digitonin to permeabilize the plasma membrane. As digitonin was washed out, RhDex was added, and the cells were treated with DIDS to entrap RhDex inside the IMS. Lastly, the permeabilized cells were washed to remove excess RhDex. In fluorescence images collected at the end of this procedure, much redfluorescing RhDex colocalized with green-fluorescing MTG, indicating that 3-kDa RhDex was entrapped in the mitochondrial IMS. RhDex also accumulated in nuclei and in small vesicles of non-mitochondrial origin (Fig. 7A, RhDex), as shown previously (9). An MTG mask selectively outlining mitochondria was used to measure RhDex fluorescence only   in mitochondria and to exclude cytosolic and nuclear fluorescence outside of the mask. In comparison to untreated hepatocytes, pretreatment with AcAld (500 M) decreased RhDex uptake by 56% (Fig. 7, A and B, AcAld). Similarly, in hepatocytes pretreated with DIDS (300 M, images not shown) to inhibit the VDAC before addition of RhDex, mitochondrial RhDex uptake was decreased by 40% (Fig. 7B,  DIDS). In contrast, MTG fluorescence was not different between treatments (Fig. 7B). In these confocal imaging experiments, a layer of oil could not be used, and the medium remained open to the atmosphere. Thus, actual AcAld concentration was less due to evaporative loss.

DISCUSSION
Many factors contributing to progressive loss of liver function in alcoholic liver disease converge at the level of mitochondrial dysfunction and disturbed mitochondrial metabolism (17, 18, 40 -42). Ethanol exposure compromises mitochondrial ATP synthesis, produces steatosis, increases oxidative stress, and suppresses endogenous liver specific functions (11, 14, 16, 18 -22, 25, 37, 42-47). Although these ethanol-induced disruptions of mitochondrial metabolism are well documented, the underlying mechanism(s) of the pathogenesis of alcohol-induced liver disease remains incompletely understood.
We hypothesized previously that ethanol-dependent impairment of mitochondrial function is initiated by closure of the VDAC in the MOM, which limits exchange of water-soluble metabolites into and out of mitochondria (6,9,10). This hypothesis was supported by data showing that the MOM became limiting after ethanol treatment for permeance of substrates for respiration in the MIM and adenylate kinase in the IMS. Moreover, ethanol decreased permeance of 3-kDa RhDex into the IMS. Here, we sought to determine whether VDAC closure was directly caused by ethanol or indirectly via an ethanol metabolite, such as AcAld, using ureagenesis as an indirect measure of metabolite flux through the VDAC. Inhibitor studies implicated that AcAld was responsible for ethanol-dependent inhibition of ureagenic respiration, a conclusion that we confirmed and extended by showing that exogenously added AcAld also inhibited ureagenic respiration and, in addition, decreased MOM permeability as assessed directly by RhDex permeance into the IMS.
If ethanol causes VDAC closure, then mitochondrial processes heavily reliant on metabolite exchange with the cytosol should be most sensitive to suppression by ethanol. Urea formation is one such process, requiring exchange of ornithine for citrulline, mitochondrial release of ATP in exchange for ADP and phosphate, and uptake of respiratory substrates to support synthesis of four ATP for each urea formed. In agreement with expectations of the VDAC hypothesis, ethanol suppressed ureagenesis by cultured rat hepatocytes, as shown by decreased respiration and urea formation in the presence of ureagenic substrates. Ethanol did not inhibit basal and uncoupled respiration. Likewise, ethanol did not decrease the efficiency of ureagenesis, as determined from the ratio of urea formed to ureagenic oxygen consumed (Fig. 2B).
Inhibitor studies indicated that AcAld formed by ethanol metabolism was causing suppression of ureagenic respiration. Inhibition of ADH, CYP450 2E1, and catalase, enzymes that metabolize ethanol to AcAld, decreased ethanol-induced suppression of ureagenesis (Fig. 3). Catalase inhibition with AT abrogated ethanol-induced suppression of ureagenic respiration to a greater extent than inhibition of ADH or CYP450 (Fig.  3). However, AT also inhibits CYP450, especially at high concentrations (10 -50 mM), and, thus, the greater extent of abrogation with AT may reflect the combined effects of catalase and CYP450 inhibition (48). Additionally, ethanol metabolism switches from predominantly ADH-mediated at low concentrations of ethanol (7-12 mM) to more predominantly catalasemediated at higher doses (Ͼ 25 mM), which may also contribute to the greater effectiveness of AT in reversing ethanol-induced suppression of ureagenesis (49 -51).
In contrast to inhibition of ADH, CYP450 2E1, and catalase, inhibition of ALDH, the enzyme that oxidizes AcAld to acetate, enhanced suppression of ureagenesis (Fig. 4). Moreover, AcAld alone dose-dependently suppressed ureagenic respiration with an IC 50 of 125 M, as estimated by Dixon plot (Fig. 6B) Like ethanol, AcAld did not inhibit basal or uncoupled respiration by hepatocytes (Fig. 6C), indicating that AcAld was not directly inhibiting mitochondrial respiratory chain activity. Although the concentration of AcAld at its intracellular sites of generation is difficult to assess, measurements of AcAld accumulation in the incubation buffer showed that AcAld increased to 45 M after 60-min incubation with 50 mM ethanol. In consideration that actual intracellular AcAld concentration is likely to be even higher, this AcAld concentration generated by hepatocytes from added ethanol is comparable with concentrations of exogenous AcAld that directly produced an equivalent suppression of ureagenic respiration (Fig. 6).
AcAld also suppressed entrapment of 3-kDa RhDex into the mitochondrial IMS of permeabilized hepatocytes (Fig. 7), which was similar to the decreased RhDex permeance observed after ethanol treatment in our earlier study (9). In mitochondria from liver and other organs, the VDAC is the only pathway known for uptake of solutes like 3-kDa RhDex across the MOM. Thus, our findings, taken together, support the conclusion that the ethanol metabolism to AcAld leads to VDAC closure and suppression of mitochondrial metabolism, like ureagenesis, that requires movement of hydrophilic metabolites between mitochondria and the cytosol.
VDAC was long considered to be a component of mitochondrial permeability transition (MPT) pores (5,52,53). However, recent observations show that VDAC isoforms are not essential for MPT pore formation or induction of MPT-dependent cell death (54). However, growing evidence indicates that the open/ closed status of the VDAC regulates mitochondrial and cellular energetics through modulation of fluxes of hydrophilic metabolites across the MOM, including toxic superoxide that can form as a result of disordered mitochondrial respiration (4,5,7,8,(55)(56)(57)(58). Restriction of exchange of vital metabolites between the mitochondrial IMS and cytosol and, hence, regulation of overall mitochondrial metabolism signifies a role of VDAC as a "governator" of mitochondrial function (6).
Although AcAld appears to close mitochondrial VDAC, the molecular mechanism by which AcAld decreases VDAC permeability remains to be determined. VDAC exists as three isoforms in liver mitochondria (VDAC1, 2, and 3), and AcAldinduced VDAC closure may be isoform-specific, which could explain why ureagenic respiration and RhDex uptake into the IMS are not completely inhibited by AcAld and ethanol. The VDAC is also the target of several kinases whose inhibition or activation after AcAld might mediate VDAC closure (3,45,59). Additionally, free tubulin promotes VDAC closure, and ethanol treatment leads to tubulin acetylation and microtubular depolymerization (56,60,61). These and other mechanisms of AcAld-dependent VDAC closure will be the subject of future study.
In conclusion, our findings show that ethanol suppresses ureagenesis in cultured rat hepatocytes, an effect mediated by AcAld and associated with a decrease in nonspecific MOM permeability. Because the VDAC is the only known pathway allowing transport of hydrophilic anionic metabolites across the MOM, decreased MOM permeability in hepatocytes oxidizing ethanol is most likely due to a decrease of VDAC conductance. Such VDAC closure is adaptive in promoting selective oxidation and detoxification of AcAld, but after repeated ethanol challenge to the liver, VDAC closure may become an early hit in the multihit pathogenesis of alcoholic liver disease. Ethanol causes oxidative stresses via CYP450 2E1 metabolism, and other aldehydes formed from lipid peroxidation may also suppress VDAC function. Indeed, preliminary data indicate that malondialdehyde and 4-hydroxynonenal, products of ␤ scission in lipid peroxidation, suppress ureagenesis (62). Thus, aldehyde-induced VDAC closure and mitochondrial dysfunction may contribute to other liver diseases in which oxidative stress is an important factor.