Impaired Long-Chain Fatty Acid Utilization by Cardiac Myocytes Isolated From Mice Lacking the Heart-Type Fatty Acid Binding Protein Gene
Abstract—Heart-type fatty acid binding protein (H-FABP), abundantly expressed in cardiac myocytes, has been postulated to facilitate the cardiac uptake of long-chain fatty acids (LCFAs) and to promote their intracellular trafficking to sites of metabolic conversion. Mice with a disrupted H-FABP gene were recently shown to have elevated plasma LCFA levels, decreased cardiac deposition of a LCFA analogue, and increased cardiac deoxyglucose uptake, which qualitatively establishes a requirement for H-FABP in cardiac LCFA utilization. To study the underlying defect, we developed a method to isolate intact, electrically stimulatable cardiac myocytes from adult mice and then studied substrate utilization under defined conditions in quiescent and in contracting cells from wild-type and H-FABP−/− mice. Our results demonstrate that in resting and in contracting myocytes from H-FABP−/− mice, both uptake and oxidation of palmitate are markedly reduced (between –45% and –65%), whereas cellular octanoate uptake, and the capacities of heart homogenates for palmitate oxidation and for octanoate oxidation, and the cardiac levels of mRNAs encoding sarcolemmal FA transporters remain unaltered. In contrast, in resting H-FABP−/− cardiac myocytes, glucose oxidation is increased (+80%) to a level that would require electrical stimulation in wild-type cells. These findings provide a physiological demonstration of a crucial role of H-FABP in uptake and oxidation of LCFAs in cardiac muscle cells and indicate that in H-FABP−/− mice the diminished contribution of LCFAs to cardiac energy production is, at least in part, compensated for by an increase in glucose oxidation.
Long-chain fatty acids (LCFAs) serve important roles as fuel molecules and as components of phospholipid membranes and triacylglycerol stores. Because of their hydrophobic nature, LCFAs are virtually insoluble in aqueous environments and therefore rely on specific mechanisms for efficient vascular transport, cellular entry, and intracellular movement. This is especially apparent in heart muscle, which depends on an unimpeded supply of blood-borne fatty acids (FAs) to energize its contractile function.1
LCFAs are delivered to the heart by albumin, a plasma and interstitial protein with multiple high-affinity binding sites for hydrophobic molecules. Although the molecular mechanism by which FAs cross the cellular membrane is largely unknown, evidence is accumulating that in heart muscle, as well as in other organs, FAs enter parenchymal cells both through passive diffusion and through a carrier-mediated mechanism.2 3 4 5 6 7 Inside the cell, cytoplasmic proteins with high affinity for LCFAs are proposed to be indispensable for efficient transcytoplasmic trafficking of LCFAs.8 These small (14- to 15-kDa) cytoplasmic FA binding proteins (FABPs) are members of the family of intracellular lipid binding proteins that is currently known to comprise at least 13 different proteins.8 9 10
Because of their abundance in tissues with a high rate of FA metabolism, such as the heart, FABPs are thought to facilitate plasmalemmal uptake of FAs and promote subsequent intracellular transport to metabolizing organelles. Until recently, however, only circumstantial evidence has been provided for this role of FABP in cellular FA handling,8 11 eg, (1) correlation between heart-type FABP (H-FABP) content and oxidative capacity observed during development and among different muscle types,12 13 (2) enhanced transfer of FA derivatives between model membranes by FABPs,14 and (3) augmented uptake and/or metabolism of FAs in cell lines transfected with liver-type or intestinal-type FABP cDNA.15 16
To study the role of H-FABP in a physiological environment and to test the hypothesis that H-FABP is required for an adequate rate of cardiac LCFA utilization, mice with a disrupted H-FABP gene were created.17 Initial studies revealed that plasma LCFA levels were elevated in H-FABP nullizygous mice. Moreover, systemic administration of a radiolabeled FA analogue (125I-labeled BMIPP [125I-BMIPP]) resulted in diminished accumulation of this compound in cardiac and skeletal muscle of H-FABP−/− mice. Taken together, these findings indicated that FA utilization was hindered in tissues normally expressing H-FABP.17
This previous work, however, was mainly qualitative and did not address the metabolic fate of FAs in hearts of H-FABP nullizygous mice. Moreover, because of the in vivo approach, possible involvement of, for instance, hemodynamic factors in the diminished deposition of 125I-BMIPP in H-FABP−/− hearts could not be excluded. The present study aimed at characterizing cardiac FA metabolism in H-FABP−/− mice in more detail. Specifically, the cellular uptake and/or metabolism of LCFAs by isolated cardiac myocytes was determined, and we investigated whether the absence of H-FABP in hearts of H-FABP−/− mice invoked adaptive pathways, such as expression of other, including unknown, FABP types or usage of alternative energy substrates.
Our results show that the absence of H-FABP markedly limits myocardial LCFA utilization. In hearts of H-FABP−/− mice, the diminished availability of LCFAs is most likely compensated for by an augmented usage of glucose for energy production. To the best of our knowledge, and in support of the previous in vivo data,17 the present findings are the first physiological demonstration of a role of FABP in FA oxidation in cardiac myocytes.
Materials and Methods
The following chemicals were obtained from the indicated suppliers: collagenase (type I, Life Technologies); [8-3H]octanoic acid and [9,10-3H]palmitic acid (Arc Inc); [1-14C]octanoic acid (DuPont NEN); [1-14C]palmitic acid, [1-14C]oleic acid, 2-deoxy-d-[1-3H]glucose, and d-[U-14C]glucose (Amersham Life Science); octanoic acid, palmitic acid, 2-deoxy-d-glucose, phloretin, and BSA (fraction V) (Sigma); and a 10% (wt/wt) substituted hydroxyalkoxypropyl derivative of Sephadex G-25 (Lipidex 1000; Packard Instruments BV). The FA content of the BSA batch used was analyzed by gas chromatography and found to be <0.10 mol FA per mol BSA (data not shown). All other chemicals were of analytical grade and were purchased from various suppliers.
Mice with a targeted deletion of the entire H-FABP locus were created as previously described and were shown to be viable and fertile and to exhibit no gross abnormalities.17 Adult mice (4 to 10 months of age) of either H-FABP−/− or H-FABP+/+ genotype and of a mixed 129×BALB/c background were used in all experiments. Mice were bred at the Animal Facility of the Max Delbrück Center in Berlin. For isolation of cardiac myocytes and other experimental procedures, mice were transferred to the Centralized Animal Facility of Maastricht University, where they had unrestricted access to food (SRM-A, a standard commercial diet [Hope Farms BV]) containing 27.5% protein, 42.5% carbohydrate, and 7.5% fat) and water. This study was approved by the local Ethical Committee on Animal Experimentation.
Studies With Isolated Cardiac Myocytes
Mice were anesthetized by intraperitoneal injection of pentobarbital combined with heparin or were killed directly by cervical dislocation. Hearts were quickly removed and placed in a dish containing ice-cold buffer A containing the following (in mmol/L): NaCl 115, KCl 2.6, MgSO4 1.2, KH2PO4 1.2, NaHCO3 10, HEPES 10, taurine 0.4, and glucose 11, pH 7.4 (at 37°C), saturated with carbogen (95% O2, 5% CO2). The aorta was mounted on a cannula (diameter 0.9 mm), and hearts were perfused in a Langendorff apparatus at 37°C for 5 minutes with buffer A followed by a 7.5-minute perfusion with buffer A supplemented with 25 μmol/L CaCl2, 0.06% collagenase, 0.7% BSA, and 15 mmol/L butanediene monoxime. A constant flow rate of 2.1 mL/min was applied. All solutions used during perfusion and subsequent steps were continuously aerated with carbogen. At the end of the perfusion period, hearts were removed from the perfusion apparatus, carefully opened, and incubated in 15 mL of medium A supplemented with 0.03% collagenase, 25 μmol/L CaCl2, 1.4% BSA, and 7.5 mmol/L butanediene monoxime for 10 minutes in a shaking waterbath (37°C, 200 rpm) (model WTR, HT Infors). Incubation proceeded for an additional 5 minutes (37°C, 100 rpm) while the Ca2+ concentration was raised in 200 μmol/L intervals to 1.0 mmol/L. Next, the cell suspension was filtered through 0.2 mm nylon mesh, and cardiac myocytes were pelleted by centrifugation (2 minutes, 23g). The cell pellet was resuspended in medium B (medium A supplemented with 1.0 mmol/L CaCl2 and 2.0% BSA [=0.3 mmol/L BSA]), washed once, and suspended in an appropriate volume of medium B yielding ≈3 to 5 mg cellular wet mass per milliliter. In cases in which cardiac myocytes were used to determine uptake and/or oxidation of (deoxy)glucose, cells were suspended in medium B devoid of glucose.
After isolation, the percentage of cells with rod-shaped cell morphology was determined by counting a minimum of 250 cells. For this, a representative sample of the cell suspension was taken at the start of the experiment, fixated in 2.0% glutaraldehyde, and analyzed after the experiment had ended. Rod-shaped cell morphology ranged between 25% and 70%, the average being independent of H-FABP genotype. All rod-shaped cells excluded trypan blue, which indicates that these cells were structurally and functionally intact, whereas rounded-up cells were damaged myocytes. Only a minor percentage of rod-shaped cells showed spontaneous contractions. For determination of cellular wet mass, duplicate aliquots of the cell suspension were centrifuged in a microcentrifuge (5 seconds, 10 000g), and the weight of the cell pellet was determined.
At the beginning of our experiments, only a few reports had appeared on the isolation of cardiac myocytes from mice.18 19 20 Published methods did not, however, report the yield or the quality of the cell preparation obtained. Moreover, the isolated cardiac myocytes were mainly used for electrophysiological and mechanical studies that suffice with relatively few intact cells. The method we have developed typically yields ≈40 to 50 mg wet mass per single heart of ≈125 mg and contains on average 45% trypan blue–excluding cells with rod-shaped morphology. Cell suspensions consist mainly of myocytes (>95%). To the best of our knowledge, this is the first reported use of mouse cardiac myocytes for metabolic studies.
Cellular Uptake of Octanoate, Palmitate, and Deoxyglucose by Cardiac Myocytes
For determination of the cellular uptake of various substrates, 0.9 mL of a suspension of freshly isolated cardiac myocytes was incubated in a capped 20-mL glass vial in a shaking waterbath (37°C, 120 rpm). A carbogen atmosphere was maintained in the incubation vials. After thermal equilibration, 0.3 mL of substrate solution was added, and incubation proceeded for 3 minutes. Next, 1.0 mL of the incubation mixture was rapidly transferred to a tube containing 9.0 mL of ice-cold stop solution (medium A supplemented with 0.1% BSA, 1.0 mmol/L CaCl2, and 200 μmol/L phloretin).2 Cells were pelleted by centrifugation (2 minutes, 75g, 4°C) and washed twice with 7.5 mL of ice-cold stop solution. Cell-associated radioactivity was assayed after the final cell pellet was dissolved in liquid scintillation fluid. All measurements were performed in duplicate, and controls were included to correct for extracellular-associated radioactivity. Substrate solutions consisted of (1) 0.4 mmol/L [8-3H]octanoate and 0.4 mmol/L [1-14C]palmitate complexed to 0.3 mmol/L BSA in medium A supplemented with 1.0 mmol/L CaCl2 or (2) 0.4 mmol/L [2-3H]deoxyglucose and 0.4 mmol/L [1-14C]palmitate complexed to 0.3 mmol/L BSA in medium A lacking glucose but supplemented with 1.0 mmol/L CaCl2.
Pilot studies with cardiac myocytes isolated from either wild-type or H-FABP−/− mice indicated that uptake of octanoate, palmitate, and deoxyglucose was linear up to at least 3 minutes.
Oxidation of Palmitate and of Glucose by Cardiac Myocytes
For determination of oxidation of palmitate and glucose, incubations were performed as described above, with the following modifications. Substrate solutions consisted of either of the following: (1) 0.4 mmol/L [1-14C]palmitate complexed to 0.3 mmol/L BSA in medium A supplemented with 1.0 mmol/L CaCl2 or (2) 0.4 mmol/L [U-14C]glucose in medium A with 1.0 mmol/L CaCl2. Incubations were carried out for 30 minutes and were stopped by addition of perchloric acid to a concentration of 0.6 mol/L. Radiolabeled CO2 was trapped in 200 μL ethanolamine/ethylene glycol (1:2 vol/vol). When oxidation of palmitate was measured, acid-soluble oxidation intermediates were determined by extracting 900 μL of the acid-soluble supernatant (2 minutes, 10 000g) with 4.5 mL chloroform/methanol (2:1 vol/vol).2
Electrical Stimulation of Cardiac Myocytes
For determination of cellular uptake and/or oxidation of substrates by electrically stimulated cardiac myocytes, incubations were performed in 20-mL incubation vials using 1.5 mL cell suspension and 0.5 mL substrate solution as described above. Electric stimuli were generated by a pulse generator (Strotmann Laboratory Supplies) and were delivered to the cardiac myocyte suspension by 2 platinum electrodes immersed in the incubation medium. For this, the platinum electrodes were mounted in the cap of the incubation vial. To prevent electrolysis of cells, pulses were biphasic. The pulse parameters (frequency, 2 Hz; 200 V top to top; monophasic pulse duration, 250 μs; phase-shift duration, 10 μs) were empirically set to trigger contraction of cardiac myocytes without causing excessive cellular damage.20a Electric stimulation for up to 10 minutes did not affect the percentage of rod-shaped cells (data not shown).
Preparations consisting solely of trypan blue–positive, rounded-up cells did not show measurable uptake and/or oxidation of substrates (data not shown). Moreover, pilot studies with cardiac myocytes isolated from either wild-type or H-FABP−/− mice indicated that oxidation products of palmitate were not yet detectable after 3 minutes. Hence, uptake data need not be corrected for loss of radiolabeled CO2. Palmitate oxidation is given as the sum of radiolabeled CO2 and the amount of acid-soluble oxidation intermediates, whereas glucose oxidation is expressed as the amount of 14CO2 produced. As a linear relationship was observed between percentage cells with rod-shaped morphology and the measured parameters (data not shown), all results were normalized to a 100% rod-shaped cell suspension.
For analysis of GLUT4 expression, isolated cardiac myocytes were dissolved (5% mass/volume [m/v]) in 1% SDS. Total protein (30 μg) was separated on a 10% denaturing polyacrylamide gel, electroblotted onto nylon membrane, and stained with a polyclonal anti-rat GLUT4 antibody combined with chemiluminescent detection as described.21 Signals were quantified by densitometric scanning of x-ray films. For detection of H-FABP, total protein (30 μg) was separated on a 15% denaturing polyacrylamide gel, transferred to nylon membrane, and immunostained with monoclonal anti-human H-FABP antibodies, as reported.22
Studies in Heart Homogenates
Homogenates (5% m/v) of adult hearts (n=4 to 5) were prepared in 0.25 mol/L sucrose, 10 mmol/L Tris, and 0.1 mmol/L EDTA (pH 7.4) using a Potter-Elvejhem tissue homogenizer and 3 pestles with increasing diameter.23 Oxidation of [1-14C]octanoate was measured as described by Veerkamp et al,24 and [1-14C]palmitate oxidation was performed according to Glatz and Veerkamp.23 Oxidation rates are calculated as the sum of 14CO2 produced and the amount of radiolabeled acid-soluble oxidation intermediates.
For determination of enzyme activities and H-FABP levels, the above-mentioned homogenates were sonicated.25 Activities of citrate synthase, β-hydroxyacyl–coenzyme A (CoA) dehydrogenase, and lactate dehydrogenase were assayed as described earlier.25 H-FABP content was determined (n=7) with a sensitive ELISA of the sandwich type, according to the method of Wodzig et al.26 For the ELISA, anti-human H-FABP monoclonal antibodies cross-reactive with mouse H-FABP were used, and recombinant mouse H-FABP, a kind gift of Dr T. Börchers (Institute for Biochemical Sensor Research, Münster, Germany), was used as a standard.
Determination of Cytosolic FA Binding Capacity and Radiochromatography
Hearts were collected from adult mice (n=3). To remove extracellular albumin, hearts were perfused for 5 minutes with buffer A supplemented with 1.0 mmol/L CaCl2 as described above, under Studies With Isolated Cardiac Myocytes. Hearts were homogenized in ice-cold PBS (5% m/v) by successive ultra-Turrax treatment and sonication.25 Cellular debris was pelleted by brief centrifugation (10 minutes, 600g, 4°C), and the supernatant was subjected to ultracentrifugation (90 minutes, 105 000g, 4°C). The resulting cytosolic preparations were delipidated using Lipidex-1000 extraction27 and were dealbuminized by passage through columns containing Blue Sepharose CL-6B (Pharmacia Biotech). For measurement of cytosolic FA binding capacity, 20 to 30 μg of protein was incubated with [1-14C]oleate (1 μmol/L) for 30 minutes at 37°C. Protein-associated FA was determined using the Lipidex-1000 assay as previously described.27
For radiochromatography, a 105 000g supernatant was prepared from isolated cardiac myocytes. Cytosolic protein (5 to 10 mg/mL) was incubated with [3H]palmitate for 30 minutes at 37°C. Thereafter, 100 μL of the incubation was loaded on a Superdex 75 HR 10/30 column (Pharmacia) and eluted with PBS at a flow rate of 0.65 mL/min. Absorbance was monitored at 280 nm, and radioactivity was determined online with a radiochromatography detector (type A-250, Radiometric Instruments and Chemical Co). Using this system, the lower limit of radioactive detection corresponded to the amount of [3H]palmitate bound to 3 μg recombinant rat H-FABP. Globular proteins with known molecular masses were used as retention time markers. Recombinant rat H-FABP22 complexed with [3H]palmitate was used to determine the time delay between absorbance and radioactivity detection.
RNA Isolation and Northern Analysis
Total RNA was isolated from hearts (n=5) homogenized in Trizol reagent (Life Technologies). RNA (15 μg) was separated on formaldehyde-containing agarose gels and transferred to Hybond-N+ membranes (Amersham). Filters were hybridized with 32P-labeled cDNA fragments, washed, and exposed to phosphor imager screens and/or x-ray films. Signals were quantified using ImageQuant software (Molecular Dynamics) and were normalized against the 18S rRNA signal, which was obtained by reprobing stripped filters with a radiolabeled 18S rRNA probe. cDNA probes for detecting cardiac expression of FABP types besides H-FABP were as described.17 The other cDNA fragments used for hybridization were the following: mouse FA-transport protein (FATP; 1250-bp XbaI-BglII fragment), rat FA translocase (FAT; 2.4-kb EcoRI-EcoRI fragment), rat acyl-CoA binding protein (450-bp EcoRI-EcoRI fragment), rat mitochondrial aspartate aminotransferase/plasmalemmal FABP28 (FABPpm; 1.6-kb NcoI-HindIII fragment), rat H-FABP (676-bp EcoRI-BamHI), rat acyl-CoA synthetase (520-bp EcoRV-HindIII fragment), rat GLUT4 (1.6-kb XbaI-XhoI fragment), rat hexokinase type II (1.5-kb NcoI-XhoI), and mouse 18S rRNA (750-bp EcoRI-EcoRI fragment). The respective cDNAs were kindly provided by Dr J. Schaffer (Washington University, St. Louis, MO), Dr N. Abumrad (State University of New York, Stony Brook), Dr J. Knudsen (Odense University, Denmark), Dr A. Iriarte (University of Missouri, Kansas City), Dr P. Brecher (Boston University, MA), Dr T. Yamamoto (Tohoku University, Sendai, Japan), Dr D. James (Washington University), Dr J. Wilson (Michigan State University, East Lansing), and Dr I. Oberbäumer (Max Planck Institute for Biochemistry, Martinsried, Germany).
For electron and light microscopy, hearts (n=2) were perfused for 5 minutes with medium A supplemented with 1.0 mmol/L CaCl2 as described above, under Studies With Isolated Cardiac Myocytes. To better preserve heart ultrastructure, hearts were perfused with a fixed hydrostatic pressure of 60 cm H2O. Heart tissue was fixated by subsequent perfusion for 10 minutes with 2.5% glutaraldehyde in 0.1 mol/L phosphate buffer (pH 7.4). Pieces of left ventricular tissue were then embedded in epoxy resin.29 For electron microscopy, coupes of ≈100 nm were prepared and stained using standard procedures. For light microscopy, 1-μm coupes were stained with toluidine blue.
Cardiac glycogen and triacylglycerol contents were determined in perfused hearts (n=3) as described before.25 Protein content was determined with the bicinchoninic acid method (Pierce), with BSA serving as a protein standard.
Data Presentation and Analysis
Data are expressed as mean±SD. The number (n) of wild-type and H-FABP−/− mice used in the experiments and measurements is indicated in parentheses. Mann-Whitney testing was used to evaluate statistical difference of measured parameters in wild-type and H-FABP−/− mice. P<0.05 was considered to denote a statistically significant difference.
Absence of H-FABP in Hearts of H-FABP−/− Mice
The absence of H-FABP in cardiac tissue derived from H-FABP nullizygous mice was confirmed by Western blotting (data not shown) and Northern blotting (Figure 1A⇓), as previously reported,17 and by direct measurement of H-FABP levels using a sensitive ELISA. In wild-type mice, H-FABP expression in cardiac muscle amounted to 1.9±0.2 mg/g wet mass.
Although H-FABP−/− mice were reported previously to be normal on histological sections,17 we now explored whether disruption of the H-FABP gene caused ultrastructural adaptations, paying special attention to mitochondrial localization. For this, heart sections were studied using electron (Figure 2⇓) and light (data not shown) microscopy. In cross sections of left ventricular heart tissue derived from either wild-type or H-FABP−/− mice, linear arrays of mitochondria lining the sarcomeres were observed, and mitochondria appeared evenly distributed throughout the sarcoplasm of the cardiac myocyte. There were no indications for an altered subcellular localization of mitochondria in cardiac myocytes of H-FABP−/− mice.
Cytosolic FA Binding Capacity of Hearts
The possible compensatory expression of other FABP types in cardiac muscle of H-FABP−/− mice was studied specifically by Northern analysis of known FABP types and functionally by determining cytosolic FA binding capacity and by radiochromatographic analysis. As reported,17 neither mRNA coding for brain-type FABP (B-FABP), epidermal-type FABP (E-FABP), adipocyte lipid binding protein (ALBP), nor liver-type FABP (L-FABP) was detectable in cardiac muscle of either wild-type or H-FABP−/− mice, whereas clear signals were observed in the respective control tissues. In cytosolic preparations derived from heart tissue of H-FABP−/− mice, the capacity to bind oleic acid was reduced by 55%, as determined using the Lipidex-1000 assay (2.1±0.1 versus 1.0±0.1 nmol oleate/mg protein for wild-type and H-FABP−/− heart cytosol, respectively). However, radiochromatographic analysis of the preparations used for this assay revealed that residual binding capacity in hearts of H-FABP−/− mice may, in part, be explained by incomplete removal of albumin.
Expression of other FABP types in cardiac myocytes of H-FABP−/− mice was further scrutinized by radiochromatography. A single peak of radioactivity, eluting at the same time as recombinant rat H-FABP (Figure 3A⇓), was observed when cytosolic proteins from wild-type cardiac myocytes were incubated with [3H]palmitate and then subjected to gel filtration (Figure 3B⇓). However, proteins capable of binding FAs were not detected in cytosol derived from isolated cardiac myocytes of H-FABP nullizygous mice (Figure 3C⇓).
Expression of Putative Sarcolemmal FA Transporters and FA Metabolizing Enzymes
Cardiac expression of several proteins involved in cellular uptake, metabolism, and intracellular transport of metabolites of FAs was examined by Northern blotting. Levels of mRNA coding for (1) FATP,30 FAT,31 and FABPpm,28 3 membrane-associated proteins putatively involved in sarcolemmal FA uptake; (2) acyl-CoA synthetase, a mitochondrial and sarcoplasmic reticular membrane-bound enzyme catalyzing the first step in FA metabolism; and (3) acyl-CoA binding protein, a cytoplasmic protein involved in binding and transport of acyl-CoA esters32 were all not different in heart tissue from wild-type and H-FABP−/− mice (Figure 1A⇑; data not shown).
Enzyme Activities and Oxidation of FAs in Heart Homogenates
The metabolic capacity of hearts was assessed by measuring FA oxidation rates and determining enzyme activities in tissue whole homogenates. Heart homogenates from wild-type and H-FABP−/− mice showed equal capacities to oxidize octanoate, a medium-chain FA, and palmitate, a LCFA (Table⇓). In addition, the activities of the mitochondrial enzymes citrate synthase and β-hydroxyacyl–CoA dehydrogenase were comparable in heart homogenates from wild-type and H-FABP−/− mice (Table⇓). As judged from cardiac triacylglycerol content, the storage of FAs as neutral lipids was comparable in hearts of wild-type and H-FABP−/− mice (Table⇓).
Uptake and Oxidation of FAs by Isolated Cardiac Myocytes
The initial uptake rate of palmitate, a ligand for H-FABP, was diminished by 45% in cardiac myocytes isolated from H-FABP−/− mice (33.7±7.9 versus 18.5±3.9 nmol · min–1 · g–1 wet mass for wild-type and H-FABP−/− cells, respectively). Under these conditions, the uptake of octanoate, which is not bound by H-FABP, was identical in cardiac myocytes from wild-type and H-FABP−/− mice (4.5±1.0 versus 4.2±1.2 nmol · min−1 · g−1 wet mass for wild-type and H-FABP−/− cells, respectively) (Figure 4A⇓). The rate of palmitate oxidation by cardiac myocytes was 45% lower in cells derived from H-FABP−/− mice (10.7±4.4 versus 5.9±0.8 nmol · min−1 · g−1 wet mass for wild-type and H-FABP−/− cells, respectively) (Figure 4B⇓).
When cardiac myocytes were made to contract by electrical stimulation (2 Hz), the initial uptake rates of palmitate and octanoate were comparable with those measured in quiescent cells (data not shown). Electrical stimulation, however, clearly affected the rate of palmitate oxidation by cardiac myocytes isolated from wild-type mice, which showed a 2-fold higher value (10.7±4.4 versus 20.9±8.5 nmol · min–1 · g–1 wet mass for quiescent and stimulated cells, respectively), but did not alter the rate of palmitate oxidation by H-FABP−/− cardiac myocytes (5.9±0.8 versus 7.7±1.8 nmol · min–1 · g–1 wet mass for quiescent and stimulated cells, respectively). In electrically stimulated cardiac myocytes, oxidation of palmitate was 65% lower in cells derived from H-FABP−/− mice when compared with wild-type cells (Figure 4B⇑).
Cardiac Glucose Utilization in H-FABP−/− Mice
To examine whether the diminished cellular uptake and oxidation of FAs was compensated for by a metabolic switch toward increased carbohydrate utilization, cellular entry and metabolism of glucose were studied. The mRNA (data not shown) and protein (Figure 1B⇑) levels of GLUT4, the main sarcolemmal glucose transporter, were comparable in hearts of wild-type and H-FABP−/− mice. Likewise, the cardiac mRNA level of hexokinase type II (data not shown), the enzyme catalyzing the first step in glucose metabolism, and the activity of the glycolytic enzyme lactate dehydrogenase (Table⇑) did not change in H-FABP−/− mice. Interestingly, the cardiac glycogen content was found to be ≈60% higher in H-FABP−/− mice (Table⇑).
Isolated cardiac myocytes were used to study the uptake of deoxyglucose and the oxidation of glucose. The initial rate of deoxyglucose uptake did not differ between cardiac myocytes isolated from wild-type or H-FABP−/− mice (12.8±5.6 versus 14.6±6.1 nmol · min–1 · g–1 wet mass for wild-type and H-FABP−/− cells, respectively) (Figure 4A⇑). However, glucose oxidation by quiescent cardiac myocytes was 80% higher when cells were isolated from H-FABP−/− mice (1.8±0.6 versus 3.2±0.9 nmol · min–1 · g–1 wet mass for wild-type and H-FABP−/− cells, respectively). Electrical stimulation (2 Hz) resulted in a 2-fold increase in the rate of glucose oxidation by cardiac myocytes from wild-type mice but did not further augment glucose oxidation by H-FABP−/− cardiac myocytes (3.6±1.8 versus 3.5±0.8 nmol · min–1 · g–1 wet mass for wild-type and H-FABP−/− cells, respectively) (Figure 4C⇑).
Cardiac LCFA uptake is a complex process requiring passage of LCFAs across several cell barriers, yet it proceeds with high efficiency.1 By an unknown mechanism, blood-borne albumin-bound LCFAs pass the capillary vessel wall and enter the interstitial fluid.33 In addition to passive diffusion, protein-assisted translocation of LCFAs across the sarcolemma is known to occur. In the heart, 3 membrane-associated proteins have been implicated in this process, namely, FATP, FAT, and FABPpm.11 34 For subsequent delivery to metabolizing organelles such as mitochondria, LCFAs most likely associate with cytoplasmic H-FABP, which can be envisioned as the intracellular counterpart of albumin. Although the collective findings from in vitro experiments8 and theoretical considerations35 36 support a prominent role of H-FABP in intracellular FA trafficking, direct proof for such a function has been lacking.
Recently, it was observed that in mice with a targeted disruption of the H-FABP gene, tissue deposition of a radiolabeled FA analogue (125I-BMIPP) was severalfold lower in cardiac tissue and moderately reduced in (resting) skeletal muscle of H-FABP nullizygous mice, indicating that utilization of LCFAs was obstructed in tissues normally expressing H-FABP.17 To allow the measurement of substrate utilization under defined conditions and to evade the potential effect of the endothelium as a barrier in overall FA utilization,33 the use of isolated cardiac myocytes, rather than intact hearts, is desired. To this end, we developed a method to isolate intact, electrically stimulatable cardiac myocytes from adult mice hearts and then studied FA uptake and oxidation in both quiescent and contracting myocytes from wild-type and H-FABP nullizygous mice.
Given the predominant reliance on FAs for cardiac energy production, we have systematically investigated whether absence of H-FABP invoked adaptations in hearts of H-FABP nullizygous mice. Possible compensatory mechanisms include (1) intracellular shift in localization of mitochondria toward the sarcolemma resulting in shortened diffusion distances, (2) expression of other FABP types in cardiac tissue, (3) altered cellular capacity to metabolize FAs, and (4) changes in expression of sarcolemmal LCFA transporters.
First, as judged from microscopic studies, the subcellular localization of mitochondria appeared similar in left ventricular tissue from H-FABP−/− and wild-type mice. Hence, diffusion distances for exogenous FAs between sarcolemmal and mitochondrial membranes are most likely not different in hearts of wild-type and H-FABP−/− mice. In addition, the activity of citrate synthase, a marker enzyme for mitochondrial density, was not different in heart homogenates from wild-type and H-FABP−/− mice.
Second, as H-FABP is a member of a large family of intracellular lipid binding proteins,8 expression of other FABP types in cardiac tissue was investigated. Two experimental approaches demonstrate that in cardiac myocytes of H-FABP nullizygous mice, no other FABP types are expressed. Northern analysis using specific probes failed to detect expression of B-FABP, E-FABP, ALBP, or L-FABP mRNA in hearts of H-FABP−/− mice.17 More importantly, using a functional test (FA binding capacity of cytosolic proteins), it was found that no substantial expression of other FABP types occurs in hearts of H-FABP nullizygous mice. In contrast, disruption of the gene encoding ALBP resulted in an elevated expression of E-FABP in adipose tissue. This compensation probably explains the apparent lack of a phenotype of ALBP nullizygous mice.37
Third, the metabolic capacity of H-FABP−/− hearts was assessed by measuring the rate of FA oxidation in homogenates. In such cell-free preparations, the delivery of FAs to intact mitochondria for subsequent oxidation is independent of the presence of H-FABP, because FAs are provided complexed to albumin. It appeared that octanoate and palmitate each were oxidized at similar rates in heart homogenates of wild-type and H-FABP−/− mice, indicating that hearts of H-FABP−/− mice had equal capacity to oxidize FAs. Supporting this notion are the observations that the activity of β-hydroxyacyl–CoA dehydrogenase, an enzyme involved in the β-oxidative pathway, and the mRNA level of acyl-CoA synthetase, the enzyme catalyzing the first step in FA metabolism, were comparable in hearts of wild-type and H-FABP−/− mice.
Fourth, the expression of membrane-associated FA transporters was evaluated by Northern blotting and revealed comparable levels of the mRNAs coding for FATP, FAT, and FABPpm in hearts of wild-type and H-FABP−/− mice. It cannot be excluded, however, that protein levels and/or the localization of these FA transporters is altered. Membrane-bound acyl-CoA synthetases catalyze the conversion of FAs to acyl-CoA esters. The latter compounds are bound with high affinity by acyl-CoA binding protein for subsequent delivery to metabolic sites.32 Acyl-CoA synthetase and acyl-CoA binding protein mRNA levels were not altered in hearts of H-FABP−/− mice. Other transcripts measured in our previous report,17 ie, long-chain acyl-CoA dehydrogenase and carnitine palmitoyl transferase I, were also maintained, in agreement with the conclusion of the present study that the capacity for LCFA metabolism is not altered in hearts of H-FABP−/− mice.
Having established that the potential for sarcolemmal FA translocation and FA metabolism most likely is maintained in hearts of H-FABP−/− mice and that apparently no other mechanisms operate to directly compensate for the absence of H-FABP, we then determined uptake and oxidation of FAs by intact cells. Cardiac myocytes of H-FABP−/− mice showed a significantly reduced uptake of the LCFA palmitate, whereas the uptake of the medium-chain FA octanoate was unaffected. Moreover, the rate of palmitate oxidation by cardiac myocytes derived from H-FABP nullizygous mice was markedly lower. Theoretically, this could be caused by preferential targeting of FAs to other metabolic pathways, such as incorporation into triacylglycerols. However, the comparable (steady-state) triacylglycerol content of hearts of wild-type and H-FABP−/− mice argues against this possibility.
Although diminished, cardiac myocytes of H-FABP−/− mice apparently were still able to take up and oxidize palmitate. Perhaps in quiescent cardiac myocytes from H-FABP−/− mice, cytoplasmic diffusion of non–protein-bound FAs suffices to meet the cellular energy requirement. As uptake of LCFAs is thought to be driven by the maintenance of a steep FA gradient between sarcolemma and, eg, respirating mitochondria, an enhanced energy demand would imply an elevated transcytoplasmic flux of FAs. To test this and to explore whether cardiac myocytes from H-FABP−/− mice were able to attain such increased intracellular flux of FAs, cardiac myocytes were electrically stimulated to evoke rhythmical contractions of these cells. Although this treatment did not influence the initial rate of cellular uptake of palmitate or octanoate, it clearly augmented the rate of palmitate oxidation by cardiac myocytes from wild-type mice. Presumably, electrical stimulation diverted FAs from incorporation into esterified lipids toward oxidation in mitochondria. However, cardiac myocytes of H-FABP−/− mice were unable to achieve a higher rate of palmitate oxidation, thus demonstrating that LCFA utilization is impaired in cardiac myocytes from H-FABP−/− mice. Taking these results together, we conclude that in the cardiac myocyte, H-FABP is required for bulk and rapid transport of LCFAs from the extracellular compartment to the mitochondria for oxidation.
In isolated rat cardiac myocytes, the initial rate of palmitate uptake (22 nmol · min–1 · g–1 wet mass)2 is in the same order of magnitude as values calculated from in vivo arteriovenous differences in FA concentration (≈100 nmol · min–1 · g–1 wet mass).1 33 The rate of palmitate uptake by rat and mouse cardiac myocytes is comparable and reflects physiologically relevant values. Thus, the observed impairment of LCFA utilization by cardiac myocytes of H-FABP−/− mice could well be of physiological significance.
In view of their importance for cardiac energy production, we subsequently investigated whether the impaired utilization of LCFAs in H-FABP nullizygous mice resulted in a metabolic shift from LCFAs toward carbohydrate utilization. Previous work revealed that plasma glucose levels were lowered in H-FABP−/− mice under all conditions, especially in starvation and with a high-fat diet, and that cardiac deposition of deoxyglucose, a nonoxidizable glucose analogue, was elevated in fasted H-FABP−/− animals.17 In the present study, we found that, despite similar levels of GLUT4 protein and hexokinase type II mRNA and a comparable cellular uptake rate of deoxyglucose, the rate of glucose oxidation by resting H-FABP−/− cardiac myocytes was elevated to a level that would require electrical stimulation of wild-type cells. Apparently, glucose oxidation by H-FABP−/− cardiac myocytes proceeded already at maximum rates, as electrical stimulation did not further affect this rate, suggesting that either cellular uptake or subsequent metabolism may have become rate limiting. The cardiac glycogen content was found to be elevated in H-FABP−/− mice and may serve to store glucose for use during periods of increased workload.38 Together these findings suggest that the impaired utilization of LCFAs in H-FABP−/− mice results in an increased reliance on carbohydrates to meet the cardiac energy demand.
In conclusion, we showed that cardiac myocytes isolated from H-FABP−/− mice have a markedly reduced ability to take up and oxidize palmitate, which supports the proposed role of H-FABP in mediating cellular uptake and intracellular transport of LCFAs. Thus, in line with our previous in vivo studies,17 the present findings provide firm evidence for a substantial role of H-FABP in maintaining high rates of LCFA utilization by cardiac myocytes.
We are greatly indebted to Yvonne de Jong, Will Coumans, Maurice Pelsers, Theo Roemen, and Peter Willemsen for expert technical assistance and to Dr Peter Frederik and Paul Bomans for performing electron and light microscopy. We thank Dr Joost Luiken for pioneering work on electrical stimulation of cardiac myocytes. Quantification of GLUT4 protein levels by Dr Yvan Fischer (RWTH Aachen, Germany) is greatly appreciated. We thank Dr Torsten Börchers (Institute for Chemical and Biochemical Sensor Research, Münster, Germany) for providing recombinant mouse H-FABP.
- Received January 18, 1999.
- Accepted June 3, 1999.
- © 1999 American Heart Association, Inc.
van der Vusse GJ, Glatz JFC, Stam HC, Reneman RS. Fatty acid homeostasis in the normoxic and ischemic heart. Physiol Rev. 1992;72:881–940.
Luiken JJFP, van Nieuwenhoven FA, America G, van der Vusse GJ, Glatz JFC. Uptake and metabolism of palmitate by isolated cardiac myocytes from adult rats: involvement of sarcolemmal proteins. J Lipid Res. 1997;38:745–758.
Sorrentino D, Stump D, Potter BJ, Robinson RB, White R, Kiang CL, Berk PD. Oleate uptake by cardiac myocytes is carrier mediated and involves a 40-kD plasma membrane fatty acid binding protein similar to that in liver, adipose tissue, and gut. J Clin Invest. 1988;82:928–935.
Bonen A, Luiken JJFP, Liu S, Dyck DJ, Kiens B, Kristiansen S, Turcotte LP, van der Vusse GJ, Glatz JFC. Palmitate transport and fatty acid transporters in red and white muscles. Am J Physiol. 1998;275:E471–E478.
Fitscher BA, Elsing C, Riedel HD, Gorski J, Stremmel W. Protein-mediated facilitated uptake processes for fatty acids, bilirubin, and other amphipathic compounds. Proc Soc Exp Biol Med. 1996;212:15–23.
Zakim D. Fatty acids enter cells by simple diffusion. Proc Soc Exp Biol Med. 1996;212:5–14.
Hamilton JA. Fatty acid transport: difficult or easy? J Lipid Res. 1998;39:467–481.
van Nieuwenhoven FA, Verstijnen CP, Abumrad NA, Willemsen PH, van Eys GJ, van der Vusse GJ, Glatz JFC. Putative membrane fatty acid translocase and cytoplasmic fatty acid-binding protein are co-expressed in rat heart and skeletal muscles. Biochem Biophys Res Commun. 1995;207:747–752.
Storch J, Bass NM. Transfer of fluorescent fatty acids from liver and heart fatty acid-binding proteins to model membranes. J Biol Chem. 1990;265:7827–7831.
Murphy EJ. L-FABP and I-FABP expression increase NBD-stearate uptake and cytoplasmic diffusion in L cells. Am J Physiol. 1998;275:G244–G249.
Binas B, Danneberg H, McWhir J, Mullins L, Clark AJ. Requirement for the heart-type fatty acid-binding protein in cardiac fatty acid utilization. FASEB J. 1999;13:805–812.
Wolska BM, Solaro RJ. Method for isolation of adult mouse cardiac myocytes for studies of contraction and microfluorimetry. Am J Physiol. 1996;271:H1250–H1255.
Dorn GW II, Robbins J, Ball N, Walsh RA. Myosin heavy chain regulation and myocyte contractile depression after LV hypertrophy in aortic-banded mice. Am J Physiol. 1994;267:H400–H405.
He H, Giordano FJ, Hilal-Dandan R, Choi DJ, Rockman HA, McDonough PM, Bluhm WF, Meyer M, Sayen MR, Swanson E, Dillmann WH. Overexpression of the rat sarcoplasmic reticulum Ca2+ ATPase gene in the heart of transgenic mice accelerates calcium transients and cardiac relaxation. J Clin Invest. 1997;100:380–389.
Luiken JJFP, van der Vusse GJ, Glatz JFC. Fatty acid uptake and oxidation by isolated rat cardiac myocytes is increased by electrical stimulation. J Mol Cell Cardiol.. 1999;31:A90. Abstract.
Fischer Y, Thomas J, Sevilla L, Munoz P, Becker C, Holman G, Kozka IJ, Palacin M, Testar X, Kammermeier H, Zorzano A. Insulin-induced recruitment of glucose transporter 4 (GLUT4) and GLUT1 in isolated rat cardiac myocytes: evidence of the existence of different intracellular GLUT4 vesicle populations. J Biol Chem. 1997;272:7085–7092.
van der Vusse GJ, Dubelaar ML, Coumans WA, Seymour AM, Clarke SB, Bonen A, Drake-Holland AJ, Noble MI. Metabolic alterations in the chronically denervated dog heart. Cardiovasc Res. 1998;37:160–170.
Wodzig KW, Pelsers MMAL, van der Vusse GJ, Roos W, Glatz JFC. One-step enzyme-linked immunosorbent assay (ELISA) for plasma fatty acid-binding protein. Ann Clin Biochem. 1997;34:263–268.
Stump DD, Zhou SL, Berk PD. Comparison of plasma membrane FABP and mitochondrial isoform of aspartate aminotransferase from rat liver. Am J Physiol. 1993;265:G894–G902.
Luft JH. Improvements in epoxy resin embedding methods. J Biophys Biochem Cytol. 1961;9:409–414.
Abumrad NA, el-Maghrabi MR, Amri EZ, Lopez E, Grimaldi PA. Cloning of a rat adipocyte membrane protein implicated in binding or transport of long-chain fatty acids that is induced during preadipocyte differentiation: homology with human CD36. J Biol Chem. 1993;268:17665–17668.
Hotamisligil GS, Johnson RS, Distel RJ, Ellis R, Papaioannou VE, Spiegelman BM. Uncoupling of obesity from insulin resistance through a targeted mutation in aP2, the adipocyte fatty acid binding protein. Science. 1996;274:1377–1379.
Goodwin GW, Ahmad F, Doenst T, Taegtmeyer H. Energy provision from glycogen, glucose, and fatty acids on adrenergic stimulation of isolated working rat hearts. Am J Physiol. 1998;274:H1239–H1247.