Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Almost one third of chronic alcoholics exhibit cardiac dysfunction (Urbano-Marquez et al., 1989), and a large proportion of these (Preedy and Richardson, 1994; Urbano-Marquez et al., 1995) develop alcoholic heart muscle disease (AHMD). Individuals with AHMD do not usually have vitamin deficiencies (Alexander et al., 1977a) or impaired nutritional status (Alexander et al., 1977b). Therefore, the development of AHMD is a function of alcohol intake rather than poor nutrition. The diagnosis of AHMD is made on the basis of deteriorating cardiac function, increased heart size, and a long history of alcohol abuse (Urbano-Marquez et al., 1989). Clinical characteristics of AHMD are reduced left ventricular ejection fraction, dilated cardiac chambers, and enlarged left ventricular mass. Evaluations of biopsy specimens at the light microscopic level demonstrate diffuse fibrosis, increased endocardial thickness, and cardiac myocyte hypertrophy (Preedy and Richardson, 1994; Urbano-Marquez et al., 1995). Electron microscopy reveals a loss or disruption of myofibrils and dilated sarcoplasmic reticulum. Swollen mitochondria with disorganized cristae are reported in most morphological studies. Apart from the history of alcoholism, these features are consistent with other dilated cardiomyopathies.

Neither the mechanism nor the specific toxic agent that produces AHMD has been established (Redetzki et al., 1983; Capasso et al., 1991, 1992; Zou et al., 1991; Mikami et al., 1993; Preedy and Richardson, 1994; Thomas et al., 1994). One potential toxin is the major metabolite of ethanol, acetaldehyde. Acetaldehyde has been found to concentrate in the heart (Espinet and Argiles, 1984) and has significant actions on cardiac function (Savage et al., 1995; Patel et al., 1997; Ren et al., 1997). Acute exposure to acetaldehyde produces a biphasic dose-dependent effect on cardiac contractility (Savage et al., 1995). At low doses, contractility is enhanced by a sympathetically mediated action, whereas at higher doses, contractility is decreased. The latter effect appears to be a direct action of acetaldehyde on the cardiac myocyte because inhibition is seen even in isolated myocytes (Ren et al., 1997). Acetaldehyde may contribute to the production of AHMD because it is highly reactive, producing measurable acetaldehyde-protein adducts even at very low concentrations (Hoerner et al., 1988). In addition, many patients with AHMD have circulating antibodies to cardiac acetaldehyde-protein adducts (Hoerner et al., 1988; Harcombe et al., 1995). However, a lack of suitable methods for chronically altering acetaldehyde concentration in vivo has prevented a definitive test of the acetaldehyde hypothesis. Confirmation or disapproval of this hypothesis is essential to understanding the basis of the disease. Unfortunately, the current evidence was largely indirect and correlative. Until now, more direct tests have had serious technical problems. Earlier experiments used metabolic inhibitors to alter the concentration of acetaldehyde (Weishaar et al., 1978; Hillbom et al., 1983; Preedy and Richardson, 1994). However, these inhibitors are only partially effective and nonspecific, greatly inhibit alcohol intake, and are potentially toxic and difficult to maintain chronically in experimental animals. The direct administration of acetaldehyde is also unsuitable for chronic studies. In this study, we increased cardiac exposure to acetaldehyde by genetically increasing the activity of enzymes that produce it, specifically in the heart. Our results indicate that a 4-fold increase in cardiac acetaldehyde levels greatly increases the rate of onset of AHMD. These results are consistent with the hypothesis that acetaldehyde plays an important role in alcohol-induced cardiomyopathy.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Development of Transgenic Mice. The MyADH transgene was produced by replacement of the catalase coding sequences in the transgene MyCAT (Kang et al., 1996) with the coding sequences for rat ADH I (Crabb and Edenberg, 1987). FVB mice obtained from the University of North Dakota Biomedical Research Center were used to produce transgenic lines containing the MyADH transgene. Standard procedures were used for producing transgenic animals. A second transgene containing a cDNA for the enzyme tyrosinase was coinjected with MyADH. The enzyme tyrosinase produces coat color pigmentation in albino mice (Overbeek et al., 1991) and was used to conveniently identify transgenic animals. All animal procedures were approved by the U.S. Department of Agriculture-certified institutional animal care committee.

Analysis of Transgene Expression. The activity and tissue specificity of the transgene were determined by measuring RNA on Northern blots and enzyme activity in tissue homogenates. RNA for Northern blots was prepared with RNazol as we previously described (Voss-McCowan et al., 1994; Kang et al., 1997) from heart, liver, skeletal muscle, brain, kidney, and lung of transgenic and normal mice. A 150-nucleotide fragment of the 3' untranslated sequence of the transgene was labeled with ³²P and used as probe. This sequence had little detectable cross-reactivity with any other mouse mRNA. The strength of the hybridization signal was determined on a PhosphorImager (model 445 SI; Molecular Dynamics, Sunnyvale, CA), and quantification was facilitated with image analysis software.

Chronic Treatment of Mice with Ethanol. Age- and sex-matched transgenic and normal animals between 9 and 12 weeks of age were placed on a nutritionally complete, all-liquid ethanol-containing diet (Lieber and DeCarli, 1982). Diets were purchased from Dyets Inc. (Bethlehem,PA). The standard diet contains 6% alcohol, which we found to produce essentially 100% mortality rates in FVB mice. Therefore, we used 4% alcohol diets containing maltose dextrin to compensate for the reduced calories. Normal FVB mice were found to have a higher intake than control mice and lost weight when exactly pair fed with transgenic animals. To compensate for this, control animals were given access to 25% higher intake than that consumed by transgenic mice. Over the course of the study, control animals consumed 17% more diet per day than transgenic animals. Despite this, transgenic animals gained 4 g b.wt. over the course of study, whereas control animals gained less than 1 g b.wt.

Alcohol-Induced Modifications in Cardiac Gene Expression. To assess changes in gene expression in AHMD and to determine how they were affected by overexpression of ADH, cardiac RNA was isolated and probed on Northern blots. Mice were sacrificed by cervical dislocation. Hearts were weighed, and RNA was isolated as described above. Northern blots were probed with labeled oligonucleotides for atrial natriuretic factor (ANF), -skeletal actin (SkActin), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA. Probes were labeled using polynucleotide kinase and [-³²P]ATP. Sequences of the probes were obtained from Jones et al. (1996). Hybridization for each probe was be quantified on Molecular Dynamics PhosphorImager and normalized to the signal for GAPDH mRNA.

Measurement of Contractility. Mice were anesthetized by i.p. injection of 100 mg/kg ketamine and 32 mg/kg xylazine coadministered with 100 IU of heparin. When deep anesthesia was achieved, the left side of the chest was opened. A small incision was made in the descending aorta to allow insertion of a cannula. Immediately on insertion of the cannula into the aorta, the inferior vena cava was cut, and flow was initiated at a rate of 1.5 ml/min. The innominate artery, common carotid artery, and subclavian artery were then ligated and cut. The heart was removed from the chest and transferred to the perfusion apparatus. From the point of cannulation and throughout the entire procedure, the heart was retrogradely perfused with Krebs-Henseleit buffer (KH) consisting of 120 mM NaCl, 20 mM NaHCO₃, 4.6 mM KCl, 1.2 mM KH₂PO₄, 1.2 mM MgCl₂, 1.25 mM CaCl₂, and 11 mM glucose. KH buffer was prefiltered through a 0.22-µm filter. Throughout the perfusion, KH buffer was continuously equilibrated with 95% O₂/5% CO₂, which maintained pH 7.4.

Electron Microscopy. Anesthesia of the mice was achieved through i.p. injection of sodium pentobarbital (70 mg/kg). The abdomen and thorax were opened, and the right atrium was cut to allow the release of blood. Perfusion fixation was immediately initiated at a flow rate of 5 ml/min for 1 to 3 min with washout solution made up as 6.6 g of paraformaldehyde, 1.15 ml of 70% glutaraldehyde, 0.5 g of CaCl₂, 0.25 g of AlCl₃, and 1 g of procaine hydrochloride in 360 ml of 0.2 M sodium cacodylate and distilled water to make 1000 ml. A small amount (1-3 drops) of 1 N NaOH was required to get the paraformaldehyde into solution. The entire solution is finally brought to pH 7.4. Washout solution is followed by perfusion with chilled (4°C) Karnovsky's fixative (pH 7.4) at a rate of approximately 3 ml/min for 10 min. The heart is removed with surgical scissors and immersed immediately in chilled Karnovsky's fixative (Karnovsky, 1965). The heart is cut to form small tissue blocks (1 mm³) while immersed in fixative. Tissue fixation continues in the cold (4°C) for 2 to 4 h in a 10:1 fluid/tissue ratio. This is followed by rinsing in 0.2 M sodium cacodylate buffer (pH 7.4), postfixation in 1% OsO₄ at 4°C (90 min), rinsing in distilled water, and en bloc staining in 1% tannic acid (60 min in dark) and 0.5% aqueous uranyl acetate (4°C, 90 min). Tissue blocks are dehydrated through graded ethanols and propylene oxide, embedded in Epon/Araldite, and cured for 48 h at 60°C. Thin sections (silver-gray interference color) are cut on an MT-2B DuPont-Sorvall ultramicrotome equipped with a Diatome diamond knife. Sections are collected on naked copper (300-mesh) grids and stained with lead citrate and uranyl acetate (4% in absolute ethanol). These are observed with either a JEOL 100-S or Hitachi 7500 transmission electron microscope at initial magnifications of ×1000 to ×35,000.

Measurement of Acetaldehyde. Acetaldehyde was measured according to a modification of the procedure of Eriksson et al. (1984). In brief, hearts were removed from anesthetized mice, blotted to remove excess blood, weighed, and immediately homogenized using a Brinkmann Tissuemizor in ice-cold 0.6 M perchloric acid. The sample was then centrifuged at 10,000g at 4°C. Triplicate aliquots of each sample were assayed within 16 h of isolation. Then, 2-ml sealed vials containing 0.4 ml of the extracts are placed in an oven at 60°C, and approximately 0.5 ml of the headspace gas was removed through the septum on the cap with a gas-tight syringe. This sample was transferred to a 200-µl loop injection system on a Varian 3700 gas chromatograph equipped with a flame ionization detector. Acetaldehyde and other components were separated on a 15-m VOCOL capillary column (Supelco, Inc.) with a 1-µm film thickness and an inner diameter of 0.25 mm. The oven temperature was ramped from an initial temperature of 35-80°C at a rate of 30°C/min. The flow rate through the column is 2.3 ml/min of hydrogen, and a flame ionization detector was used. Under these conditions, acetaldehyde was separated from methanol and ethanol with baseline resolution with a total separation time of less than 1 min.

Measurement of Cardiac NAD⁺ and NADH. These cofactors were measured according to spectrophotometric procedures (Klingenberg, 1985). Briefly, the heart was excised from anesthetized mice and pulverized in liquid nitrogen. NADH was extracted with alcoholic KOH and NAD⁺ extracted in 0.6 N perchloric acid. Extracts were neutralized and centrifuged. NAD⁺ was measured spectrophotometrically by the production of NADH during the oxidation of alcohol by yeast ADH. NADH was measured by the drop in absorbance at 340 nm during the NADH-dependent reduction of pyruvate by lactate dehydrogenase.

Statistical Analysis. The difference between the mean values of two groups was determined using two-tailed Student's t test. Comparisons of multiple mean values were made with one-way ANOVA and Scheffé's post hoc test. Statistical analysis was performed with the program SPSS 8.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Three lines of mice constructed with the MyADH transgene were assayed for enzyme activity. As shown in Fig. 1, ADH activity in the hearts of transgenic mice was increased by up to 40-fold over levels measured in nontransgenic, FVB mice. MyADH line 239, which had both high enzyme activity and a coat color marker for convenient identification of transgenic mice, was chosen for further studies. All subsequent data were obtained from this line. We noted no differences in fertility or body weight between MyADH 239 mice and FVB mice. Tissue specificity of the transgene was examined at the level of RNA. Multiple Northern blots of different transgenic tissues identified expression of the transgene only in the heart (a representative blot is shown in Fig. 2). At least two of three MyADH bands observed in this Northern blot may be due to the presence of multiple polyadenylation sites in the transgene: one in the ADH cDNA and another derived from the rat insulin II gene.

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Expression of the MyADH transgene. Enzyme activity in heart extracts from FVB mice and three transgenic lines. The transgenic line and number of hearts assayed for that transgenic line, in parentheses, is indicated below the bars. Vertical lines indicate the S.E.

View larger version (32K): [in this window] [in a new window]   Fig. 2.   MyADH transgene expression is specific to the heart in line 239. RNA was isolated from the indicated tissues and probed with a fragment specific to the transgene product. As explained in the text, multiple bands in the heart may be due to multiple polyadenylation sites in the transgene. The positions of ribosomal RNAs are indicated at the left. Hrt, heart; Liv, liver; Brn, brain; Lng, lung; and Msl, muscle.

Most important to our original aim, we found that the elevation in cardiac ADH activity significantly increased cardiac acetaldehyde levels after the administration of an i.p. injection of 3 g/kg alcohol. Thirty minutes after injection, mice were sacrificed for rapid removal of the heart and quantification of acetaldehyde through gas chromatography and flame ionization detection. As shown in Fig. 3, acetaldehyde levels measured in heart tissue of transgenic mice were four times higher than those in control mice. We also measured systemic acetaldehyde levels with the use of a new subcutaneous sampling procedure (manuscript in preparation) designed to avoid artifactual production of acetaldehyde that plagues plasma and blood sample analysis (Eriksson et al., 1984). Our preliminary assays indicate that systemic acetaldehyde levels were significantly higher in transgenic mice (p < .02 by Student's t test, n = four per group).

View larger version (31K): [in this window] [in a new window]   Fig. 3.   The ADH transgene elevates cardiac acetaldehyde 30 min after a 3 g/kg i.p. alcohol injection. *p < .02, different from control by Student's t test (n = 6/point).

Because the transgene satisfied our primary aim of increasing cardiac exposure to acetaldehyde, we started these mice on a chronic alcohol diet. Animals were initiated on a 1% alcohol (by volume) liquid diet using the Lieber DeCarli formulation (Dyets Inc.); the diet was gradually increased to 4% alcohol. The 4% maximum was chosen because control mice fed higher-percentage alcohol diets demonstrated very high mortality rates in preliminary studies. After 10 weeks on the alcohol diet, mice of each group were sacrificed to assess the onset of cardiomyopathy by measuring changes in gene expression. Increased expressions of SkActin and ANF have been reported in several cardiomyopathies (Jones et al., 1996; Colbert et al., 1997), and this provides a sensitive indicator of cardiac damage. As shown in Fig. 4, levels of these two mRNAs were significantly increased in alcohol-treated MyADH mice relative to alcohol-treated control mice. Without alcohol treatment, MyADH mice did not have increased expression of these genes. Control mice on alcohol also did not show a statistically significant increase in expression of either mRNA. However, a possible trend toward increased ANF expression was observed in alcohol-treated control mice (Fig. 4B). These results, obtained after only 10 weeks of alcohol treatment, indicated that the MyADH transgene sensitized the heart to the deleterious effects of alcohol.

View larger version (32K): [in this window] [in a new window]   Fig. 4.   Expression of SkActin and ANF mRNAs are increased in MyADH mice treated with alcohol for 10 weeks. Levels of RNA expression are normalized to levels of GAPDH expression in each sample. Separate samples were obtained from three to six animals for each point. *p < .05 by Student's t test, SkActin or ANF expression levels are higher in alcohol-treated ADH mice than in alcohol-treated FVB mice.

Human hearts are enlarged due to AHMD. Normal mouse hearts were also enlarged after 5 months of an alcohol diet (Fig. 5). However, the MyADH transgene dramatically increased this effect. Five months of the alcohol diet produced an increase in heart-to-body weight ratio of 18% in control mice and 80% in transgenic mice. All alcohol-treated transgenic hearts were severely dilated, as is seen in human AHMD, and most alcohol-treated transgenic hearts had a large thrombus in the left atrium. This is also a common finding in human AHMD (Ferrans, 1989).

View larger version (59K): [in this window] [in a new window]   Fig. 5.   The ADH transgene exacerbates cardiac enlargement due to chronic alcohol. Hearts from age-matched transgenic or control FVB mice were examined after 5 months of alcohol diet or standard laboratory chow diet. *, ADH alcohol hearts are larger than FVB alcohol hearts and **, FVB alcohol hearts are larger than hearts without alcohol (p < .05 by ANOVA, n = 7/group).

Electron microscopy revealed morphological effects of alcohol treatment that were exacerbated by the transgene. Myocardium in non-alcohol-treated control (Fig. 6A) and non-alcohol-treated transgenic (Fig. 6C) animals showed that both exhibit fine structural characteristics considered typical for heart muscle (McNutt and Fawcett, 1974). In contrast, alcohol-treated control (Fig. 6B) and alcohol-treated transgenic (Fig. 6D) animals demonstrated fine structural abnormalities that were exacerbated in the latter. Alcohol-treated control animals showed focal lesions similar to those seen in early stages of human AHMD (Ferrans, 1989); these included intracellular edema, fragmentation and loss of myofibrils, and glycogen accumulation. Focal lesions were common, and in some areas, ultrastructural changes characteristic of chronic AHMD were noted (Urbano-Marquez et al., 1989), including interstitial fibrosis, loss of myofibrils, and mitochondria with disorganized cristae.

View larger version (144K): [in this window] [in a new window]   Fig. 6.   Transmission electron micrographs of mouse left ventricular myocardia. Non-alcohol-treated FVB (A) and transgenic (C) animals show normal myocardial ultrastructural features, whereas alcohol-treated FVB (B) and transgenic (D) mice show alterations that are exacerbated in the latter. Alcohol-treated FVB animals (B) show focal sarcoplasmic edema and enlarged mitochondria, with loss and disorganization of contractile elements within sarcomeres. Alcohol-treated transgenics (D) are more globally affected and show significant intracellular edema, loss of sarcoplasmic reticulum, further degeneration of myofibrillar structures, and smaller mitochondria with sparse matrix and disorganized cristae. M, mitochondrion; S, sarcomere; Z, Z-line; arrow, T-tubule; arrowhead, sarcoplasmic reticulum. All original magnifications, ×18,500.

Electron microscopic observations of the myocardium in alcohol-treated transgenic mice (Fig. 6D) showed they were the most severely affected. This degree of damage was typical of most of the myocardium in all alcohol-treated transgenic hearts. Significantly, the alterations in these animals closely mirrored those seen in chronic AHMD (Bulloch et al., 1972; Urbano-Marquez et al., 1989). In these animals, the most striking change was a loss of myofibrils and a concomitant increase in edematous sarcoplasm. Mitochondria were reduced in number and smaller and less dense than in nonalcoholic transgenic animals. In addition, loss and disorganization of contractile elements, loss of sarcoplasmic reticulum, and mitochondria swelling were evident. Interestingly, our study showed no significant change in lipid inclusions in alcohol-treated animals, which is consistent with studies of myocardium in human chronic alcoholics (Bulloch et al., 1972).

Decreased contractility is a characteristic of human AHMD. Eighteen weeks of the alcohol diet markedly reduced contractility in both transgenic and control animals (Fig. 7). However, the residual contractility of control hearts was three times greater than the contractility of MyADH hearts. Without alcohol treatment, there was no detectable difference between control and transgenic contractility.

View larger version (32K): [in this window] [in a new window]   Fig. 7.   The MyADH transgene exacerbates the detrimental effect of alcohol on cardiac contractility. Contractility was measured in Langendorff perfused hearts as described in Materials and Methods. *p < .05, different from control. **p < .05, by ANOVA, different from alcohol-treated FVB.

We considered the possibility that the cardiac damage in transgenic mice was due to depletion of NAD⁺, an essential cosubstrate in the enzymatic conversion of ethanol to acetaldehyde. To examine this, cardiac NAD⁺ and NADH were measured 30 min after the i.p. administration of 0.4 g of ethanol/kg b.wt. in seven MyADH and seven control mice. This dose of alcohol was chosen because it produced levels of blood alcohol that were considerably higher than that of the chronically alcohol-treated transgenic mice. The NAD⁺/NADH ratios were similar in transgenic and control hearts (3.89 ± 0.38 versus 3.31 ± 0.79 for transgenic and control mice, respectively; p > .7 by Student's t test). These results do not indicate that depletion of NAD⁺ was an adequate factor to explain the cardiac damage observed in transgenic mice.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

To determine the role of acetaldehyde in AHMD, we developed a new transgenic model designed to increase cardiac exposure to acetaldehyde. Cardiac ADH activity was elevated 40-fold by the MyADH transgene in line 239. The resultant increase in cardiac alcohol metabolism produced a 4-fold increase in acetaldehyde exposure. Transgene expression was confined to the heart; therefore, cardiac effects are not secondary to transgene actions in other organs. The innate stability and specificity of the transgene provide a more stable and specific method than do pharmacological approaches (Preedy and Richardson, 1994) for increasing acetaldehyde exposure.

By several criteria, it was clear that the MyADH transgene produces a very rapid, alcohol-dependent cardiomyopathy. After only 10 weeks of the alcohol diet, transgenic hearts demonstrated changes in gene expression for ANF and SkActin that are characteristic of cardiomyopathy (Colbert et al., 1997). After 18 weeks on the alcohol diet, hearts were grossly enlarged, the cardiac ultrastructure was disrupted, and contractility was reduced. The very rapid onset of cardiac damage emphasizes the potency of acetaldehyde and provides alcohol researchers with a savings of many months for the induction of full-blown AHMD.

The damage in transgenic hearts appeared to be an exacerbation of typical alcohol damage seen in control mice. In normal mice, alcohol enlarged the heart and produced focal ultrastructural damage. In transgenic animals, hearts were even larger and ultrastructural damage was more global. Also, the morphological changes seen in the alcohol-treated transgenic mice are similar to many of those seen in human AHMD (Bulloch et al., 1972; Klein and Harmjanz, 1975; Ferrans, 1989; Urbano-Marquez et al., 1989). At the gross level, these changes include enlarged hearts, dilated ventricles, and intra-atrial thrombi. At the level of ultrastructure, they include loss of contractile element-sarcoplasmic reticulum, disorganization of myofilaments/sarcomeres, sarcoplasmic edema, and swollen disrupted mitochondria. Functional damage to the heart was also more severe in transgenic mice. Consistent with previous studies, chronic alcohol significantly reduced cardiac contractility in control animals (Thomas et al., 1994; Brown et al., 1998), and this decrement was more severe in the transgenic mice. More specific contractile deficits that have been characterized in chronic alcohol studies in several species (Thomas et al., 1994) were not addressed in the present study.

These results support the hypothesized role of acetaldehyde in alcohol-induced cardiac damage. In our transgenic mice, cardiac damage was secondary to acetaldehyde exposure. Alcohol was required for the effect of the transgene. Changes in NAD⁺/NADH ratios were not significant, and organs outside of the heart were unaffected by the transgene. Therefore, the most parsimonious explanation for cardiac damage was direct exposure to acetaldehyde. Encouraged by these data, we are producing additional transgenic mice that will reduce cardiac exposure to acetaldehyde by overexpressing acetaldehyde dehydrogenase. We predict that this new transgene will protect mouse hearts from the damaging effects of alcohol and establish the hypothesis that acetaldehyde plays a crucial role in the development of AHMD.