Liver cells from rats chronically fed a Lieber-De Carli diet for 3 wk presented a marked decreased in tissue Mg2+ content and an inability to extrude Mg2+ into the extracellular compartment upon stimulation with catecholamine, isoproterenol, or cell-permeant cAMP analogs. This defect in Mg2+ extrusion was observed in both intact cells and purified liver plasma membrane vesicles. Inhibition of adrenergic or cAMP-mediated Mg2+ extrusion was also observed in freshly isolated hepatocytes from control rats incubated acutely in vitro with varying doses of ethanol (EtOH) for 8 min. In this model, however, the defect in Mg2+ extrusion was observed in intact cells but not in plasma membrane vesicles. In the chronic model, upon removal of EtOH from the diet hepatic Mg2+ content and extrusion required ∼10 days to return to normal level both in isolated cells and plasma membrane vesicles. In hepatocytes acutely treated with EtOH for 8 min, more than 60 min were necessary for Mg2+ content and extrusion to recover and return to the level observed in EtOH-untreated cells. Taken together, these data suggest that in the acute model the defect in Mg2+ extrusion is the result of a limited refilling of the cellular compartment(s) from which Mg2+ is mobilized upon adrenergic stimulation rather than a mere defect in adrenergic cellular signaling. The chronic EtOH model, instead, presents a transient but selective defect of the Mg2+ extrusion mechanisms in addition to the limited refilling of the cellular compartments.
- Mg2+ homeostasis
- plasma membrane
magnesium (Mg2+) is the second most abundant cation in mammalian cells after potassium (15). Within the cell, magnesium is mainly localized inside mitochondria, endoplasmic reticulum, nucleus, and cytoplasm (18) and regulates numerous biological functions within each of these compartments (37). The cellular abundance of Mg2+ and its physiological relevance for various cell functions contrast with the limited knowledge of the mechanisms regulating its homeostasis and transport. In the absence of hormonal or metabolic stimuli, mammalian cells retain their physiological Mg2+ content unaltered even when large outward-oriented Mg2+ gradients are imposed (45). Following hormone administration, however, large and fast fluxes of Mg2+ in and out of mammalian cells have been observed (35, 36, 47). Although no mammalian Mg2+ extrusion mechanism has been cloned to date, experimental evidence supports the notion that the majority of mammalian cells utilize a Na+-dependent Mg2+ extrusion pathway, tentatively identified as a Na+/Mg2+ exchanger (12), as the main route for Mg2+ efflux. Under conditions in which no extracellular Na+ is available to sustain the exchanger operation (34) or nonspecific Na+ transport inhibitors [i.e., amiloride, imipramine, or quinidine (11, 17, 43)] are present in the extracellular milieu, the operation of an alternative Na+-independent Mg2+ extrusion pathway is uncovered (19). This pathway can use extracellular Ca2+, Mn2+, or even anions to mobilize Mg2+ (reviewed in Ref. 19). The operation of these two distinct Mg2+ extrusion mechanisms has also been observed in purified liver plasma membrane vesicles (3). The putative Na+/Mg2+ exchanger is specifically located in the basolateral domain of the hepatocyte whereas the Na+-independent Mg2+ extrusion pathway is uniquely located in the apical domain of the cell (4). In addition, these two pathways can be differentiated on the basis of their distinct sensitivity to the inhibitors imipramine and amiloride, respectively (4), or their activation by different adrenergic agonists (9, 10). In liver cells, the administration of glucagon or β-adrenoceptor agonist (e.g., isoproterenol) selectively activates the Na+-dependent Mg2+ extrusion pathways (9) whereas the stimulation of α1-adrenoceptor by phenylephrine specifically activates the Na+-independent Mg2+ extrusion pathway (9, 10). Consistent with this differential activation, the administration of epinephrine activates both Na+-dependent and Na+-independent Mg2+ extrusion pathways (9, 22) and results in an extrusion of Mg2+ equivalent to the sum of the amounts of Mg2+ separately mobilized by α1- and β-adrenoceptor stimulation (9, 22). The Na+-dependent Mg2+ extrusion elicited by glucagon or isoproterenol (9) can be mimicked by the administration of forskolin or cell-permeant cAMP analogs (9, 35) and is prevented by the administration of the Rp-cAMP isomer to the cell (44). This observation has therefore fostered the notion that the increase in cellular cAMP resulting from the activation of glucagon or β-adrenergic receptor or from the use of cell-permeant cAMP analogs phosphorylates the putative Na+/Mg2+ exchanger or a nearby regulatory protein and determines Mg2+ extrusion from the cell (16, 27). A similar dependence of the Na+-dependent but not the Na+-independent Mg2+ extrusion pathway on cAMP-mediated phosphorylation has been observed in purified plasma membrane vesicles (5, 7).
Hypomagnesemia and a decrease in tissue Mg2+ content have been reported in various pathological conditions including alcoholism (25). A marked hepatic Mg2+ loss has been observed in acute and chronic animal models of alcohol consumption (40, 47), as well as in chronic alcoholic patients (33). Furthermore, clinical and experimental evidence suggests that Mg2+ supplementation ameliorates several hepatic metabolic functions compromised by alcohol abuse (48). Yet the mechanisms responsible for tissue Mg2+ loss, the relevance of a decrease in tissue Mg2+ content for the onset of alcohol-related complications, and the modality by which Mg2+ supplementation exerts its beneficial effects have not been investigated in detail. Our group has reported that the acute perfusion of liver with increasing doses of ethanol (EtOH) results in a marked, time- and dose-dependent extrusion of Mg2+ from the organ into the perfusate via a Na+-dependent mechanism. The Mg2+ extrusion is associated to a transient decrease in cellular ATP content consequent to EtOH metabolism since both Mg2+ extrusion and ATP loss are prevented by the alcohol dehydrogenase inhibitor 4-methyl-pyrazole (4-MP) (40). We have also reported that, in hepatocytes acutely treated with EtOH, adrenergic stimulation is impaired as a result of an altered Mg2+ homeostasis and distribution within cellular compartments (46). Following administration of EtOH in the diet for 3 wk according to the Lieber-De Carli model, hepatic Mg2+ and ATP content decrease by ∼25 and ∼17%, respectively, and the hepatocytes lose the ability to reaccumulate Mg2+ from the extracellular space to restore cellular Mg2+ homeostasis (47).
In the present study, hepatocytes and plasma membrane vesicles were used as experimental models to investigate the time frame necessary for Mg2+ extrusion to recover and return to a normal level upon acute or chronic EtOH administration and alcohol removal. The obtained results indicate that ∼70 min in the acute EtOH model and 10 days in the chronic EtOH model are necessary to fully restore Mg2+ extrusion in liver cells. Similar results are obtained in both models following stimulation by cell-permeant cAMP analog. As total Mg2+ content returns to normal level within a similar time frame for both EtOH models, it appears that the defect in Mg2+ extrusion predominantly depends on a defective restoration of Mg2+ content within cellular compartments rather than a mere defect in adrenergic signaling. In addition, the chronic EtOH model evidences a defective operation of the Mg2+ extrusion mechanisms operating in the cell membrane.
MATERIALS AND METHODS
Collagenase (CLS2) was from Worthington. Luciferin-luciferase and lactate dehydrogenase (LDH) enzymatic kits and all the other analytical grade chemicals and reagents were from Sigma (Sigma, St. Louis, MO). The diets for ethanol-treated and control animals were from Bio-Serve (Frenchtown, NJ).
Male Sprague-Dawley rats (180–200 g body wt) were randomly divided into control- and EtOH-treated groups and housed individually in metabolic cages. EtOH-treated rats were maintained for 3 wk with a liquid Lieber-De Carli diet containing 67 ml EtOH/l (6% EtOH, final concentration, vol/vol). Control rats received a daily amount of control diet isocaloric to the Lieber-De Carli (pair-fed protocol) (47). For further comparison, pellet-fed Sprague-Dawley rats of paired age or paired weight were used. Weight gain was recorded weekly for all experimental groups. For the withdrawal studies, after 3 wk of EtOH administration, the alcohol diet was suspended and the rats were fed the control diet for a period of time varying from 2 to 15 days.
Determination of tissue cation content.
Animals were anesthetized by an intraperitoneal injection of saturated pentabarbital sodium solution (50 mg/ml). Following attainment of deep anesthesia, the abdomen of the animal was opened and the liver exposed. Liver was collected, rinsed in 250 mM sucrose, blotted on absorbing paper, weighed, minced, and homogenized in 10% HNO3. Following overnight extraction, the acid mixture was centrifuged at 2,000 g × 5 min to sediment denaturated protein. The Na+, K+, Ca2+, and Mg2+ content of the acid extracts were measured by atomic absorbance spectrophotometry (AAS) in a Perkin-Elmer 3100, calibrated with proper standards (40).
Collagenase dispersed cells.
Collagenase-dispersed rat hepatocytes were isolated from EtOH-fed rats, control diet-fed rats, and rats receiving pellet chow diet according to the procedure of Seglen (38). After the isolation, hepatocytes were resuspended, at the final concentration of ∼1 × 106 cells/ml, in a medium composed of (in mM) 120 NaCl, 3 KCl, 1.2 KH2PO4, 12 NaHCO3, 1.2 CaCl2, 1.2 MgCl2, 10 HEPES, and glucose 10, pH 7.2 at 37°C, under O2-CO2 (95:5 vol/vol) flow and kept at room temperature until used. Cell viability was 87 ± 3, 86 ± 2, and 86 ± 4% for EtOH-fed rats, control diet-fed rats, and chow diet rats, respectively (n = 10 for each experimental group), as assessed by Trypan blue exclusion test, and did not significantly change over 3–4 h (85 ± 4, 83 ± 2, 82 ± 5%, respectively, n = 10 for each group). For the determination of Mg2+ transport, 1 ml of cell suspension was transferred into a microfuge tube and the cells were rapidly sedimented at 600 g × 30 s. The supernatants were removed and the cells washed with 1 ml of a medium having the same composition as the one reported above but devoid of Mg2+ (incubation medium). The cells were then transferred in 10 ml of incubation medium, prewarmed at 37°C, and incubated therein under continuous stirring and O2-CO2 flow. After a few minutes of equilibration, the reported concentrations of adrenergic agonist or dibutyryl-cAMP (abbreviated as cAMP in the figures) as a cell-permeant cAMP analog were added to the incubation system. At the time points reported in the figures, 700 μl aliquots of the incubation mixture were withdrawn in duplicate, and the cells were sedimented in microfuge tubes. The Mg2+ content in the supernatant was measured by AAS.
Acute EtOH administration.
Aliquots of collagenase-dispersed hepatocytes isolated from chow-fed rats or control diet-fed rats were preincubated with 0.01, 0.1, or 1% EtOH for 8 min at 37°C. At the end of the preincubation, the cells were sedimented at 600 g × 1 min. The supernatant was removed and assessed for Mg2+ content by AAS. The cell pellet was transferred to a prewarmed volume of the incubation medium indicated previously. After a few minutes of equilibration, the cells were stimulated for Mg2+ extrusion by addition of adrenergic agonist or dibutyryl-cAMP at the reported doses.
Cellular Mg2+ content.
To estimate total cellular Mg2+ content, hepatocytes from the different experimental groups were sedimented at 600 g × 1 min. The cell pellet was then digested overnight in 10% HNO3 (40, 47). The Mg2+ content of the acid extract was measured by AAS following sedimentation of the denaturated protein at 2,000 g × 5 min in a refrigerated Beckman J-6B centrifuge.
Chow diet-fed male Sprague-Dawley rats (250–350 g body wt) were anesthetized by intraperitoneal injection of pentobarbital sodium (50 mg/ml). The abdomen was opened, and the liver was perfused via the portal vein with a medium containing 120 mM NaCl, 3 mM KCl, 1 mM CaCl2, 1.2 mM KH2PO4, 0.6 mM MgCl2, 12 mM NaHCO3, 15 mM glucose, and 10 mM HEPES, pH 7.2 at 37°C, preequilibrated with O2-CO2 (95:5 vol/vol), at the flow of 3.5–4 ml·g−1·min−1. Following the cannulation, the liver was rapidly removed and placed on a platform. After 5 min of equilibration, the perfusion medium was replaced with another having similar composition but devoid of Mg2+ (Mg2+-free medium). Contaminant Mg2+ content of the medium was measured by AAS and found to be ∼10–12 μM. Following 8 min of washout with Mg2+-free medium (time 0), samples of the effluent were collected at 30-s intervals, and Mg2+ content of the perfusate was measured by AAS. The samples collected during the first 10 min provided a baseline for the subsequent infusion of EtOH or adrenergic agonist. EtOH was infused in a range of concentrations between 0.01% (1.5 mM) and 1% (150 mM) for 8 min (40). After 8 min from the time of EtOH introduction into the perfusion system the livers were removed from the perfusion apparatus, gently blotted on adsorbing paper, and weighed. Two aliquots of ∼1 g each were removed. One aliquot was homogenized in 5 volumes of 10% HNO3 and digested overnight to measure total tissue Mg2+ content. The protein content of the acid extract was sedimented in a refrigerated Beckman J-6B centrifuge (2,000 g × 10 min), and the Mg2+ content in the supernatant was measured by AAS. The second aliquot was homogenized (20%, wt/vol) in 5% perchloric acid and digested for 5–10 min in ice. The acid mixture was then neutralized by addition of KHCO3, and the denaturated protein was sedimented in a refrigerated Beckman J-6B (1,500 g × 10 min). The supernatants were removed and stored at −20°C until used for ATP determination.
Sedimentation of aliquots of the perfusate in microfuge tubes (14,000 g × 10 min) and protein assay excluded the persistence of nonresident cellular and plasma proteins in the collected medium (40). The absence of cell damage was assessed by enzymatically measuring LDH activity in aliquots of the perfusate at 1-min intervals throughout the procedure. The release of K+ from damaged cells into the perfusate was measured by AAS in aliquots of the perfusate (40).
Plasma membrane isolation.
Total liver plasma membrane vesicles (tLPM) were purified from rats fed a Lieber-De Carli diet, control diet, or Purina chow diet as well as from livers acutely perfused with varying doses of EtOH or sham perfused. Plasma membrane vesicles were purified and enzymatically assessed for purity using 5′-nucleotidase, cytochrome-c oxidase, and glucose-6-phosphatase activities as markers for plasma membrane, mitochondria, and endoplasmic reticulum, respectively, as described in detail by Cefaratti et al. (3, 4). Negligible levels of cytochrome-c oxidase and glucose-6-phosphatase activities were detected in tLPM preparations (data not shown), whereas 5′-nucleotidase and Na+-K+-ATPase were found to be enriched 11- and 14-fold, respectively, in both EtOH-treated and control diet tLPM as well as in plasma membranes from chow-fed rats. The 5′-nucleotidase activity was 0.054 ± 0.009 and 0.058 ± 0.002 in the liver homogenate of control diet and EtOH diet animals, respectively, and 0.667 ± 0.092 and 0.648 ± 0.107 in the corresponding tLPM preparations (n = 8 preparations for all experimental conditions, each tested in duplicate). The activity of the Na+-K+-ATPase was 0.028 ± 0.016 and 0.029 ± 0.015 in the homogenate of control diet and EtOH diet animals, respectively, and 0.315 ± 0.081 and 0.338 ± 0.068 in the corresponding tLPM preparations (n = 8 preparations for all experimental conditions, each tested in duplicate). Similar values of enrichment were observed in tLPM purified from livers obtained from animals fed a pellet chow diet, refed animals at various time points after removal of EtOH from the diet, and livers perfused in vitro with EtOH (data not shown). The activities of 5′-nucleotidase and Na+-K+-ATPase were also used to evaluate vesicle orientation, as previously reported (3). The comparison of these activities to those of detergent-disrupted vesicles (considered as 100%) confirmed our early report (3) that ≥90% of EtOH-treated and control tLPM were in the “inside-in” configuration following loading with Mg2+ irrespective of EtOH administration. To exclude contamination, the tLPM endogenous carryover of cations and adenine phosphonucleotides were measured by AAS and HPLC, respectively, and found to be negligible (not shown).
Loading of LPM and determination of Mg2+ fluxes.
Five-milliliter aliquots of liver plasma membrane vesicles (LPM) were resuspended in 25 ml of 250 mM sucrose-25 mM HEPES (pH 7.4 with Tris) in the presence of 20 mM NaCl, and loaded by four passes in a Thomas C Potter with a tight-fitting pestle at 4°C (1). Following sedimentation of the mixture at 34,500 g × 10 min in a Sorvall SS-34 rotor to remove excess of extravesicular cation, Na+-loaded vesicles were resuspended in 5 ml of 250 mM sucrose-25 mM HEPES (pH 7.4 with Tris) and stored in ice until used. Loading efficiency was assessed by treating the vesicles with detergent (Triton X-100) and measuring the amount of Na+ extruded in the extravesicular space or retained into the vesicle pellet by AAS, as reported previously (3).
Mg2+ fluxes were measured by AAS.
An aliquot of Mg2+-loaded LPM was incubated in the medium mentioned above, at 37°C, under continuous stirring, at the final concentration of ∼300 μg protein/ml. After 2 min of equilibration, aliquots of the incubation mixture were withdrawn in duplicate at 2-min intervals, and the vesicles were sedimented in microfuge tubes at 7,000 g × 45 s (3). Total Mg2+ content in the supernatants was measured by AAS. The pellet was digested overnight in 500 μl of 10% HNO3. The following day the denaturated protein was sedimented in microfuge tubes, and the Mg2+ and Na+ content of the acid extract measured by AAS. The first two time points after the equilibration period (t = 0 min) were used to establish a baseline. Following the withdrawal of the second sample, 50 mM Na + or 500 μM Ca2+ was added to the incubation mixture, and the incubation continued for 6 additional min. These Na+ or Ca2+ concentrations were selected because they elicit maximal Mg2+ extrusion from LPM suspensions (3). As Mg2+ content in the supernatant varied considerably among preparations as a result of the loading procedure and carryover, the data are reported as the net variation in extravesicular Mg2+ or intravesicular Na+ content, normalized per milligram of protein, for simplicity. To calculate the net Mg2+ extrusion, the Mg2+ content in the supernatant at the first two time points was calculated, averaged, and subtracted from the values of the subsequent time points of incubation.
Dephosphorylation and rephosphorylation of liver plasma membrane vesicles.
For these experiments, liver plasma membrane vesicles were purified in the presence of 20 units alkaline phosphatase per gram tissue throughout the purification procedure (dephosphorylation procedure), as described in details elsewhere (5). For the in vitro rephosphorylation studies, the alkaline phosphatase-treated vesicles were loaded with 2 mg of catalytic subunit of protein kinase A and 20 mM ATP per 5 mg of vesicle protein at the time of Mg2+ loading as previously described (5). The vesicles were then sedimented at 20,000 rpm for 20 min to remove excess cation and nontrapped protein kinase A catalytic subunit and resuspended in the incubation medium (250 mM sucrose, 25 mM HEPES, pH 7.4 with Tris) described previously at the final concentration of 5 mg protein/ml and stored in ice until used. Mg2+ transport was assessed as described above.
HNE-induced modifications of liver proteins.
Aliquots of liver and liver plasma membrane vesicles from EtOH-fed and refed animals were homogenized at 10% wt/vol in 200 mM sucrose, 20 mM HEPES, pH 7.4, in the presence of a cocktail of protease inhibitors (Sigma). One volume of the homogenate was incubated for 2 min at 80°C in Laemmli buffer containing 1% β-mercaptoethanol and 1% protease inhibitor cocktail (Sigma, St. Louis). Liver and plasma membrane proteins (15 μg/lane) were resolved on a 12% SDS-PAGE gel, and electrotransferred onto a nitrocellulose immobilization membrane (Schleicher & Schuell). Western blot analysis was performed incubating the nitrocellulose membrane overnight at 4°C with antibodies recognizing 2:1 amino acids-4-hydroxynonenal (HNE) adducts. These antibodies are highly specific in recognizing only the cross-linked adduct between two amino acids and one HNE molecule and not the hemiacetalic Michael adduct between protein and HNE, as we have used them successfully in other studies (24). The nitrocellulose membranes were extensively washed with PBS, and the primary antibody binding was visualized via peroxidase-conjugated secondary antibody and ECL Western blotting detection reagents (Amersham Biosciences). Primary antibody binding was visualized utilizing peroxidase-conjugated secondary antibody and ECL detection reagents.
Assessment of PEt formation.
Formation of phosphatidylethanol (PEt) was assessed as described in detail by Exton and collaborators (1) in aliquots of liver tissue and collagenase-dispersed hepatocytes. Liver tissue and hepatocytes were extracted with 1.5 ml CHCl3-MeOH (1:2 vol/vol) followed by CHCl3 (0.5 ml) and H2O (0.5 ml). PEt was separated by TLC on silica gel F-254 plates. The plates were developed with ethyl acetate-isooctane-acetic acid (9:5:2), dried and developed by dipping in CuSO4 (5% wt/vol) and phosphoric acid (4% wt/vol), followed by heating at 190°C for 20 min (1). Lipid content was quantitated by densitometry.
Determination of ATP levels and protein content.
Cellular ATP level was measured in isolated hepatocytes by using a combination of luciferin-luciferase assay and HPLC detection system as reported previously (40).
Protein content was measured according to the procedure of Lowry (26), using BSA as standard.
The data are reported as means ± SE. Data were first analyzed by one-way ANOVA. Multiple means were then compared by Tukey's multiple comparison test performed with a q value established for statistical significance of P < 0.05.
Male Sprague-Dawley rats fed for 3 wk with a 6% EtOH-enriched liquid diet weighed 9% less than pair-fed rats receiving an isocaloric control liquid diet (379.8 ± 6.5 vs. 417.6 ± 5.2 g body wt, respectively, n = 11 for each experimental group), and 15% less than rats maintained with a regular Purina chow pellet diet (447.1 ± 8.6 g body wt, n = 11). Following withdrawal of EtOH from the diet and feeding with liquid control diet for 15 days, these animals remained ∼5 and ∼8% underweight compared with animals fed exclusively the control diet (401.3 ± 5.6 vs. 421.3 ± 7.8 g body wt, respectively, n = 11 for each experimental group) and pellet-fed rats (433.5 ± 6.9, n = 10), respectively.
Consistent with a previous observation of ours (47), hepatic cation determination indicated a marked decrease in total Mg2+ content in EtOH-fed animals (Table 1). Despite a >18% increase in liver mass compared with liquid diet controls (14.27 ± 0.61 vs. 12.09 ± 0.37 g of tissue, respectively, n = 10, P < 0.05), and pellet-fed rats (12.54 ± 0.39 g, n = 10, P < 0.05), the liver of EtOH-fed rats presented a 20% decrease in total Mg2+ content compared with livers of control liquid or pellet diet-fed rats (53.5 ± 2.9 vs. 66.9 ± 5.8 vs. 67.3 ± 4.9 nmol Mg2+/mg protein, respectively, n = 10, P < 0.01). The liver of EtOH-fed animals also presented a decrease in total K+ and an increase in total Na+ and total Ca2+ content (Table 1). In contrast, Mg2+ and K+ content increased in the urine (Table 1).
The decrease in total cellular Mg2+ content was confirmed in collagenase-dispersed hepatocytes (28.78 ± 0.58 vs. 35.18 ± 0.79 vs. 34.97 ± 0.85 nmol/mg protein from EtOH-fed, control liquid, and control pellet diet, respectively, n = 9, P < 0.01). Consistent with a decrease in cellular Mg2+ content and a depletion of Mg2+ mobilizable pools, the hepatocytes from EtOH-fed rats had also lost the ability to extrude Mg2+ upon hormonal stimulation. As Fig. 1 shows, hepatocytes from rats fed a liquid control diet responded to the administration of catecholamine by extruding between 2 and 3 nmol/mg protein over 6 min from the administration of phenylephrine or isoproterenol (Fig. 1, A and B). A quantitatively similar Mg2+ extrusion has been observed in hepatocytes from pellet-fed rats (not shown, see Refs. 6 and 7). Prolonging the incubation for a longer period of time did not result in an increased extrusion of Mg2+ (not shown), consistent with our previous studies (9, 10, 35). since Mg2+ extrusion is maximal at time = 8 min (i.e., time = 6 min from agonist addition), the subsequent results are reported as net extrusion at this time point for simplicity. A quantitatively similar Mg2+ extrusion was observed in hepatocytes stimulated by the cell-permeant cAMP analog dibutyryl-cAMP (100 μM, labeled as cAMP in this and the following figures for simplicity) (Fig. 1B), or thapsigargin (Fig. 1B), an agent that mimics α1-adrenoceptor-mediated Mg2+ extrusion (9). In contrast, Mg2+ extrusion was significantly hampered in hepatocytes from EtOH-fed animals (Figs. 1, A and B) irrespective of the agonist or dose (not shown) used. Because hepatocytes from control rats on pellet diet and liquid diet presented similar amounts of cellular Mg2+ and ATP and responded in a manner qualitatively and quantitatively similar to the administration of Mg2+-extruding agents (not shown; see Refs. 9, 10, and 40 for comparison), for the remainder of our study we used hepatocytes from rats fed a liquid control diet, since they better mimic the experimental conditions of rats receiving the Lieber-De Carli diet.
The operation of distinct Mg2+ extrusion mechanisms can be evidenced in purified liver plasma membrane vesicles (3, 4). The advantage of this experimental model is that Mg2+ transport can be assessed independent of cellular compartments and buffering systems. Hence we used liver plasma membrane vesicles to ascertain whether the defect in Mg2+ extrusion occurred at the level of the transport mechanisms rather than the regulatory signaling components. Figure 2A shows that vesicles from control diet-fed animals extruded ∼145 and ∼158 nmol Mg2+/mg protein within 1 min upon addition of 50 mM Na+ or 500 μM Ca2+, respectively (Fig. 2B). In contrast, Mg2+ extrusion from plasma membrane vesicles purified from livers of EtOH-fed rats was almost completely inhibited (Fig. 2, A and B). Concentrations of Na+ and Ca2+ larger than those reported in the figure were not utilized because these concentrations already elicit maximal Mg2+ extrusion in plasma membrane vesicles (3, 4).
Similar Mg2+ extrusion experiments were performed in hepatocytes from control diet animals acutely treated for 8 min with concentrations of EtOH varying from 0.01 to 1% (40) or in plasma membrane isolated from livers perfused for 8 min with similar doses of EtOH (6). Upon administration of 0.01%, 0.1, or 1% EtOH, hepatocytes released 1.63 ± 0.39, 3.07 ± 0.53, and 4.12 ± 0.63 nmol Mg2+·mg protein−1·8 min−1, respectively, in the extracellular medium (n = 5, P < 0.01 for all three EtOH concentrations vs. the corresponding time point in untreated cells). A dose-dependent mobilization of Mg2+ was also observed in the perfused livers (1.28, 1.95, and 3.52 μmol Mg2+ following 8 min infusion of 0.01, 0.1, and 1% EtOH, respectively, n = 5, P < 0.01 for all three EtOH concentrations vs. control perfused livers, see also Ref. 40 for comparison). Isolated hepatocytes pretreated with EtOH exhibited a marked decrease in phenylephrine, isoproterenol, or cell-permeant cAMP-induced Mg2+ extrusion compared with the extrusion observed in EtOH-untreated cells (Fig. 3A). This decrease directly depended on the dose of EtOH administered. A marked decrease in adrenoceptor-mediated Mg2+ mobilization persisted also in hepatocytes pretreated with 4-MP (50 μM) as an inhibitor of alcohol metabolism (40). In the presence of the inhibitor, however, the Mg2+ extrusion elicited by the cell-permeant cAMP or thapsigargin was largely retained (Fig. 3A). In contrast, the amplitude of Mg2+ extrusion was unaffected in plasma membrane vesicles purified from livers perfused with EtOH for 8 min irrespective of the dose of alcohol administered. Plasma membrane vesicles purified from livers previously perfused with 0.01, 0.1, or 1% EtOH (Fig. 3B), the highest dose of EtOH that can reliably be infused without release of LDH or cell damage (40), extruded between 150 to 155 and between 180 to 190 nmol Mg2+·mg protein−1·1 min−1 when stimulated by Na+ or Ca2+, respectively, compared with ∼145 ± 26 and 176 ± 32 nmol Mg2+·mg protein−1·1 min−1 extruded upon addition of extravesicular Na+ or Ca2+, respectively, to vesicles purified from sham-perfused livers (Fig. 3B).
We next investigated the recovery time of cellular Mg2+ content following acute and chronic exposure to EtOH. For the chronic model, following 3 wk of administration, EtOH was removed from the diet and the animals were fed for varying periods of time with the liquid control diet. As Table 2 indicates, at least 10 days were necessary for hepatic Mg2+ content to return to normal level. This restoration was accompanied by the normalization of all other cation content within the liver tissue as well as in the urine. The extrusion of Mg2+ in collagenase-dispersed hepatocytes isolated from animals at different time points after EtOH withdrawal from the diet also returned to the level observed in EtOH-untreated hepatocytes within 10 days (Fig. 4A). Prolonging the feeding with control diet for more than 10 days (i.e., 15 days, Fig. 4A) did not further increase the amplitude of Mg2+ extrusion irrespective of the agonist utilized. A similar stepwise restoration was observed in purified liver plasma membrane loaded with 20 mM Mg2+ and stimulated by extravesicular addition of Na+ or Ca2+ (Fig. 4B). When similar experiments were performed on liver cells following acute treatment with 1% EtOH for 8 min (Fig. 5), Mg2+ extrusion returned to normal level within 75 min from EtOH removal. Prolonging the incubation up to 90 min did not increase Mg2+ extrusion. For hepatocytes pretreated with 0.01 or 0.1% EtOH, the recovery was also fully attained at time = 75 min from EtOH removal (∼95 ± 3%, n = 5, not shown), suggesting that the time course of recovery was independent of the dose of EtOH administered. In all these conditions, incubating the cells in supraphysiological extracellular Mg2+ concentrations (e.g., 2 or 5 mM Mg2+, not shown) did not shorten the recovery time. Concentrations of Mg2+ larger than 5 mM in the incubation system or the effect of administering a diet enriched in Mg2+ following EtOH withdrawal to shorten the recovery time were not tested at this time. The administration of doses of agonists larger than those reported in Figs. 4A and 5 at any time point prior to the return to maximum Mg2+ extrusion was also ineffective at accelerating the restoration of Mg2+ extrusion (not shown). Because the amplitude of Mg2+ extrusion in both the chronic and acute model was not enhanced or restored by the stimulation with 100 μM dibutyryl-cAMP (Figs. 4A and 5) or 200 μM 8-Cl-cAMP (not shown), it can reasonably be excluded that the defective extrusion depended merely on a defective or not optimal production of cAMP within the hepatocytes following administration of β-adrenergic agonist. As indicated previously (Fig. 3B), plasma membrane vesicles from livers perfused acutely with EtOH presented a Mg2+ extrusion quantitatively and qualitatively similar to that observed in vesicles purified from sham-perfused livers. Plasma membrane vesicles isolated at later time points (e.g., 15, 30, or 45 min) following perfusion with EtOH did not show significant differences compared with vesicles isolated at similar time points from sham-perfused livers or nonperfused livers. Hence this model was not used for the recovery study upon acute EtOH administration.
Chronic EtOH administration is associated with desensitization of receptors coupled to adenylyl cyclase activity (30). We have already reported that acute EtOH treatment is also associated with a decreased production of cAMP following adrenergic stimulation (46). Hence we investigated whether Mg2+ extrusion in tLPM from EtOH-fed and refed animals could be restored by in vitro phosphorylation according to our published protocols (5). This approach, however, did not restore the amplitude of Mg2+ extrusion in the vesicles. Plasma membrane vesicles isolated after 3 wk of EtOH diet extruded 27.42 ± 0.51 nmol Mg2+·mg protein−1·2 min−1 (see also Figs. 3B and 4B) compared with 26.94 ± 0.57 nmol Mg2+·mg protein−1·2 min−1 extruded from in vitro rephosphorylated vesicles, n = 3 preparations for each experimental conditions, each tested in duplicate. Plasma membrane vesicles purified from rats refed 5 days without EtOH extruded 67.95 ± 1.01 vs. 68.29 ± 0.98 nmol Mg2+·mg protein−1·2 min−1 extruded from in vitro rephosphorylated vesicles, n = 3 preparations for each experimental conditions, each tested in duplicate (see Fig. 4B). These results were consistent with the reported observation that the addition of cell-permeant cAMP analogs to hepatocytes isolated from similar experimental animals was ineffective at restoring Mg2+ extrusion (Figs. 1 and 3).
Because of the impossibility to test for a direct effect of EtOH on the Mg2+ extrusion mechanisms (neither of the two mentioned mechanisms has been cloned), we tested our preparations for the formation of HNE and PEt. Generated from reactive oxygen species formation and lipid peroxidation, HNE is a highly reactive aldehyde forming adducts with phospholipids and proteins (31). On the other hand, PEt is a bioproduct of ethanol consumption, which has received a high degree of attention as a potential biomarker in alcoholics. This product, formed through the activity of phospholipase D (1) can be detected within biological membrane as well as in the plasma (20).
Figure 6A shows formation of adducts between HNE and plasma membrane proteins following 3 wk of EtOH administration. The formation of these adducts tends to decrease following removal of EtOH from the diet (Fig. 6, A and B) with a time course comparable to that of Mg2+ extrusion restoration. Also in the acute EtOH model, a trend toward an increased formation of HNE/protein adducts was observed in hepatocytes treated with 1% EtOH, but the process did not achieve statistical significance (not shown), most likely due to the short time of exposure to EtOH. Hepatocytes incubated with lower concentrations of EtOH (0.01 or 0.1%) presented negligible amounts of HNE adducts (not shown).
The formation and disappearance of PEt within the liver upon 3 wk exposure to EtOH also exhibits a similar time course (Fig. 7A). At variance of HNE adducts, however, the amount of PEt detected in the tissue declined at a slower rate over the 2 wk period of refeeding (Fig. 7A). A marked production of PEt was also observed in hepatocytes exposed acutely to 1% EtOH for 8 min (Fig. 7B). Formation of PEt albeit to a smaller extent was also observed in hepatocytes treated in vitro with 0.1% EtOH (∼12 to 15 pmol wet wt) or 0.01% EtOH (∼7 to 8 pmol/g wet wt).
During the last decade, our laboratory and other groups as well have reported that the increase in cellular cAMP level that follows the administration of either hormones (isoproterenol, catecholamine, glucagons, or PGE2), forskolin, or cell-permeant cAMP analogs results in an extrusion of Mg2+ from the cell (see Table 1 in Ref. 40 for a summary of tested cells and tissues). Although the physiological significance of Mg2+ extrusion is not fully appreciated, it appears that in liver cells it is closely associated with the output of glucose that accompanies the infusion of hormones like glucagon or catecholamine (9, 41). In addition, experimental (40, 47, 48) and clinical (13) data indicate that chronic EtOH consumption alters physiological functions and ion pattern within several tissues. Tissue Na+ and Ca2+ content have been reported to increase as a result of an altered operation of transport mechanisms (14, 39) and cell signaling (21). A decrease in tissue Mg2+ content has also been reported to occur in brain, heart, and other tissues, but the causes for the decrease have not been fully elucidated. Reports from our laboratory indicate that both the acute (40) and chronic (47) administration of EtOH result in a marked decrease in total hepatic Mg2+ content, which is associated with a >17% decrease in total cellular ATP level (40). Ethanol administration also results in an inhibition of Mg2+ extrusion and accumulation mechanisms (46). The present study was undertaken to elucidate the time frame necessary for EtOH-mediated inhibition of Mg2+ extrusion to be restored to its full amplitude and hepatic Mg2+ homeostasis to return to basal level.
Magnesium Homeostasis and Transport Upon Chronic EtOH Feeding
Hepatocytes from rats receiving 6% EtOH in the diet for 3 wk showed a ∼20% decrease in total Mg2+ content compared with hepatocytes from pair-fed control rats, in good agreement with previous reports (47). The hepatocytes also presented a prolonged inability to rapidly restore hepatic Mg2+ homeostasis as determined by the cation determination at different time points following EtOH withdrawal from the diet (Tables 1 and 2). As a result of Mg2+ loss and delayed renormalization in Mg2+ content, both isolated cells and purified plasma membrane vesicles presented a marked inability to extrude Mg2+ following hormonal stimulation, administration of second messengers, or challenge by extravesicular counterions such as Na+ or Ca2+. The results with cell-permeant cAMP analog exclude that the defect merely lays in a defective production of cAMP and related phosphorylation within the cells. Although it could be argued that the lack of Mg2+ extrusion is secondary to the depletion of cellular pools from which Mg2+ is mobilized following hormonal stimulation, the results in plasma membrane vesicles suggest that both the Na+-dependent and the Na+-independent (Ca2+-dependent) Mg2+ extrusion mechanisms operating in the cell membrane are affected by EtOH administration. Because neither of these Mg2+ extrusion mechanisms has presently been cloned, we are not in the position to determine whether EtOH impairs their operation directly by interacting with the transporters in a manner similar to what has already been observed for neuronal and cellular proteins (23) or indirectly by affecting either the phospholipids environment in which the transporters operate or the proteins therein located via formation of elevated levels of PEt (32) or 1-hydroxyl-ethyl radical (29). An additional modality by which EtOH can impair the transporters operation is by generating lipid peroxidation products such as HNE (42), which in turn can modify membrane proteins or phospholipids by forming irreversible adducts on specific amino acids or carbonyls (31). As Mg2+ extrusion returns to normal level within ∼10 days in both cells and plasma membrane vesicles, it is tempting to associate the return to a normal extrusion with the progressive refilling in total cellular Mg2+ content over time. This explanation, however, is not completely consistent with the observation that Mg2+ extrusion remains defective in plasma membranes up to day 7 upon EtOH removal despite the fact that the vesicles were filled with 20 mM Mg2+ to resemble the Mg2+ concentration present within the hepatocyte under physiological conditions (15, 18). As vesicles purified from refed animals seem to retain a Km similar to that observed in vesicles from control diet-fed animals (Fig. 4B), it does not appear that EtOH administration has affected the affinity of the transporters for the intravesicular Mg2+ concentration. Similarly, it does not appear that lack of cAMP-mediated phosphorylation is a potential cause because in vitro phosphorylation of tLPM does not restore Mg2+ extrusion. Hence indication may be there, albeit indirect, that a proper number of functional transporters have to be expressed in the hepatocyte cell membrane to restore Mg2+ extrusion. The time frame of ∼2 wk would indeed be consistent with the time necessary for the transporters to be newly synthesized and expressed in the cell membrane. The lack of cloning information and detecting antibodies, however, prevents us from validating this possibility. We did observe, however, formation of adducts between HNE and cell membrane proteins after 3 wk of EtOH administration. The amount of adducts tended to decrease upon removal of EtOH from the diet. Because these proteins have not yet been identified, their involvement in the (defective) Mg2+ extrusion remains questionable. It has to be noted that the use of antibodies to detect the formation of stable adducts between HNE and proteins is a highly specific approach (24), which indicates that reactive oxygen species were generated and lipid peroxidation occurred. This approach, however, does not quantitate the extent of lipid peroxidation taking place or the amount of other lipid peroxidation products (e.g., malonyldialdehyde) formed within the hepatocyte. In addition, we detected formation of PEt in the cell membrane of liver cells upon 3 wk of EtOH administration in the diet. The amount of PEt present in the cell membrane tended to decrease during refeeding time. Additional studies are required to determine whether PEt formation plays a specific role in altering Mg2+ homeostasis and transport upon ethanol administration.
In terms of cation composition, the decrease in Mg2+ content was paralleled by a decrease in K+ content and an increase in Na+ and Ca2+ content. The modalities by which the cellular concentrations of these cations are affected by EtOH are not completely elucidated. It has been reported that EtOH administration inhibits K+ transport at the cell membrane level (28). As cellular K+ content returns to normal with a time course similar to Mg2+, it is possible that the restoration of K+ homeostasis plays a permissible role on Mg2+ entry, most likely via modulation of cell membrane potential. The increase in Na+ content has also been observed following EtOH administration (2, 28), and it has been attributed to changes in the activity rate of the Na+-K+-ATPase or the Na+/H+ exchanger. A third possibility is that the Na+ increase depends in part on the extrusion of Mg2+ via the putative Na+/Mg2+ exchanger. Inhibition of this exchanger by amiloride, in fact, largely prevents the loss of Mg2+ by EtOH (40). Ethanol also alters hepatic Ca2+ homeostasis by impairing the activity of sarco(endo)plasmic reticulum Ca2+-ATPase pumps as well as the release of Ca2+ from the endoplasmic reticulum (21). It is not fully investigated whether the open probability of Ca2+ channels in the hepatocyte cell membrane is also affected following exposure to EtOH. It is similarly unclear whether under the latter condition Ca2+ enters the hepatocyte in exchange for Mg2+ through the Na+-independent mechanism. The absence of a specific inhibitor for this transporter prevents at this time to confirm or disprove this possibility. It is well documented, however, that the perturbation of Ca2+ homeostasis observed following EtOH exposure leads to activation of endoplasmic reticulum stress and nonmitochondrial apoptotic pathway (8).
Magnesium Homeostasis and Transport Upon Acute EtOH Administration
Also the acute administration of EtOH impairs the ability of liver cells to extrude Mg2+ effectively upon addition of adrenergic agonists or second messengers. At variance of what observed in the chronic EtOH model, however, the defect in this model appears to lay predominantly in the EtOH-mediated depletion of the cellular Mg2+ pools from which the cation is mobilized upon adrenergic stimulation and on the associated signaling mechanisms at the cell membrane, as previously reported (46). This notion is corroborated by the observation that plasma membrane vesicles purified from livers perfused with varying doses of EtOH for only 8 min and artificially loaded with Mg2+ do not show any defect in Mg2+ extrusion even upon infusion of the organ with 1% EtOH (i.e., ∼150 mM) (Fig. 3B and also Ref. 6). Work in progress in our laboratory would suggest that the administration of 0.01% EtOH for 8 min is sufficient to inhibit the accumulation of Mg2+ from the extracellular compartment for more than 60 min, and larger doses of EtOH (i.e., 0.1% or 1%) do not change significantly the time necessary for Mg2+ accumulation to be fully restored (L. M. Torres and A. Romani, personal observation). Hence it appears that in this model of acute EtOH administration the restoration of Mg2+ extrusion merely depends on the refilling of the cellular pools, a process that occurs within a time frame similar to that observed for the Mg2+ extrusion for all the concentrations of EtOH tested so far. Irrespective of the dose of EtOH administered, in fact, 60 min was the minimum time necessary for the extrusion process to return to a detectable level (at least 75% of total). In this model, a trend toward an increased formation of HNE-protein adducts was also detected under these experimental conditions, but the phenomenon did not achieve statistical significance (not shown). In contrast, a marked increase in PEt formation was observed. Although not tested under our experimental conditions, it is interesting to note that the PEt formation increases significantly in hepatocytes pretreated with EtOH and exposed to various Ca2+-mobilizing hormones (e.g., vasopressin or catecholamine) (1). In view of this observation and our results, additional experiments need to be carried out assess the exact role of PEt in the defective Mg2+ mobilization especially under conditions of hormonal stimulation.
To our knowledge, this is the first study to investigate the recovery time of Mg2+ homeostasis and extrusion following acute and chronic EtOH administration. The results reported here indicate that a single dose of EtOH is sufficient to impair Mg2+ homeostasis and extrusion for more than 60 min. Although this impairment is relatively limited in time, it can be envisioned that repeated doses of EtOH staggered over time can induce a persistent decrease in cellular Mg2+ homeostasis and distribution, which can ultimately affect the operation of specific enzymes within select organelles [e.g., mitochondria dehydrogenases (33)]. As expected, the alteration in Mg2+ homeostasis and transport persists for a prolonged period of time following a protracted EtOH administration. A prolonged administration of EtOH, in fact, results in a major loss of cellular Mg2+ within the hepatocytes and also affects the ability of the cells to redistribute and transport Mg2+ effectively as a result of a direct or indirect effect of alcohol on the transporters and/or the surrounding membrane environment. The short- and long-term implications of an impaired Mg2+ content on liver metabolism are presently under investigation.
This research was supported by National Institute on Alcohol Abuse and Alcoholism Grant AA-11593 to A. Romani.
- Copyright © 2009 the American Physiological Society