Upregulation of the Cardiac Monocarboxylate Transporter MCT1 in a Rat Model of Congestive Heart Failure
Background— Cardiac metabolism becomes more dependent on carbohydrates in congestive heart failure (CHF), and lactate may be used as an important respiratory substrate. Monocarboxylate transporter 1 (MCT1) promotes cotransport of lactate and protons into and out of heart cells and conceivably flux of lactate between cells, because it is abundantly present in the intercalated disk.
Methods and Results— Six weeks after induction of myocardial infarction (MI) in Wistar rats, left ventricular end-diastolic pressures were >15 mm Hg, signifying CHF. MCT1 and connexin43 protein levels in CHF were 260% and 20%, respectively, of those in sham-operated animals (Sham), and the corresponding mRNA signals were 181% and not significantly changed, respectively. Confocal laserscan immunohistochemistry and quantitative immunogold cytochemistry showed that MCT1 density was much higher in CHF than in Sham both at the surface membrane and in the intercalated disk. In CHF, a novel intracellular pool of MCT1 appeared to be associated with cisternae, some close to the T tubules. In contrast, connexin43 particles, seen exclusively at gap junctions, were substantially fewer. Maximum lactate uptake was 107±15 mmol · L−1 · min−1 in CHF and 42±6 mmol · L−1 · min−1 in Sham cells (P<0.05). The Km values were between 7 and 9 mmol/L (P=NS).
Conclusions— In cardiomyocytes from CHF rats, (1) the amount of functional MCT1 in the sarcolemma, including in the intercalated disk, is increased several-fold; (2) a new intracellular pool of MCT1 appears; (3) another disk protein, connexin43, is much reduced; and (4) increased reliance on lactate and other monocarboxylates (eg, pyruvate) could provide tight metabolic control of high-energy phosphates.
Received January 24, 2001; revision received April 12, 2001; accepted April 17, 2001.
In the failing heart, fatty acid oxidation is reduced, and there is a greater reliance on carbohydrates, although the metabolism still is aerobic.1–3 Because the anaerobic threshold during exercise is lower, the availability of lactate may be increased.4 Thus, lactate may be a major energy resource for the failing heart, because even in normal heart muscle, lactate oxidation may account for >50% of the oxygen consumption, provided that there is sufficient supply. Because lactate is carried across the cell membrane by monocarboxylate transporters (MCTs), we hypothesized that MCT1 expression and function would be increased in a rat model of myocardial infarction (MI) and congestive heart failure (CHF). MCT1 is strongly expressed in the cardiomyocyte plasma membrane, including the intercalated disk, and is the only MCT in rat heart whose identity has been confirmed.5–7 We also compared the expression of MCT1 with that of another protein of the intercalated disk, connexin43, the channel protein of the gap junction. Expression of connexin43 is reduced in ischemic heart disease,8 but no data are available from experimental studies undisturbed by other disease states and effects of medication.
MI was induced in male Wistar rats (300 to 320 g) by left coronary artery ligation9; some rats were sham-operated. Rats were anesthetized with 30% oxygen, 70% nitrous oxide, and 1% to 4% halothane. For postoperative analgesia, injections of 0.05 mL buprenorphin (0.3 mg/mL) SC were given. After 6 weeks, the rats were anesthetized again, some of them by injection of midazolam, fentanyl citrate, and fluanisone (3.8, 0.24, and 7.5 mg/kg body wt IP, respectively), and systolic pressure and left ventricular (LV) end-diastolic pressure were measured.9 Only MI rats with an LV end-diastolic pressure >15 mm Hg were included. The study conformed to the Norwegian Animal Welfare Act.
Rabbit polyclonal antibodies were produced by immunization with a synthetic peptide corresponding to the C-terminus of MCT1.10 We also used commercial polyclonal antibodies against MCT1 (Chemicon International, Inc), connexin43 (Zymed), and the glucose transporters GLUT1 and GLUT4 (Charles River Pharmservices).
The hearts were divided into 3 groups, as follows.
1. Northern and Western blotting was performed on viable LV tissue (CHF, n=10 and Sham, n=5 rats). Northern blots were made as described with 32P-labeled cDNA probes: connexin43 (1392 bp, courtesy E. Beyer, Boston, Mass), MCT1 (1900 bp), and GAPDH (1300 bp, courtesy H. Prydz, Oslo, Norway), and slot blots were used for quantification of the signal.9 Quantitative Western blotting was carried out as described.9
2. Hearts were prepared for either immunofluorescence or immunogold electron microscopy (CHF, n=9 and Sham, n=8 rats). Surviving LV myocytes, isolated as described9 and fixed in 2% formaldehyde,11 or cryostat sections of intact heart fixed in 4% formaldehyde12 were labeled with 40 μg/mL rabbit or 18 μg/mL chicken primary antibody and visualized with a secondary antibody [F(ab0)2] conjugated to Cy3 (Jackson ImmunoResearch Laboratories, Inc) or Alexa488 (Molecular Probes Inc). Heart sections and cardiomyocytes (n=60) were viewed with a confocal microscope (Leica TCS SP, Tektronix Phaser 440). For immunogold electron microscopy, 2 different fixation procedures were used: (1) perfusion with 4% formaldehyde and 0.5% glutaraldehyde,6 and (2) a “pH shift” protocol using 4% formaldehyde and 0.2% picric acid in 0.1 mol/L acetate buffer (pH 6.0) followed by the same in 0.1 mol/L carbonate buffer (pH 10.5).13 Specimens were dissected from the posterior free wall of the LV and from the septum of CHF (avoiding scar tissue) and Sham hearts. Postembedding immunogold cytochemistry was carried out on ultrathin sections of freeze-substituted tissue samples.14 Controls included (1) preincubating the affinity-purified antibodies (anti-MCT1, 2 μg/mL; anti-connexin43, 1 μg/mL) with excess immunization peptide (0.5 and 0.25 μg/mL, respectively) or with a heterologous peptide (corresponding to amino acids 679 to 697 of protein kinase C-γ at the same concentration as the immunizing peptide) and (2) replacement of the primary antibody with rabbit IgG (2 μg/mL) or buffer.
For gold particle analysis, in electron micrographs, 670 membrane fragments, chosen at random, were pooled into 68 groups (accumulated lengths).6 Particles within 50 nm from the membrane were counted, allowing computation of linear density (number per μm) by use of the PALIREL program.15 Membrane fragments were classified as those facing capillaries or a neighboring myocyte (“nondisk”) and as those engaged in intercalated disks, excluding desmosomes and gap junctions (“disk”). Density (particles per μm2) of the total number of gold particles within “disk fields,” defined as the area delimited by the borders between the dense filamentous masses of fascia adherens and the adjoining myofilaments, was computed with the MORFOREL program.15
3.Lactate transport into cardiomyocytes (CHF, n=4 and Sham, n=4 rats) was determined in cells loaded with BCECF acetoxymethyl ester (Molecular Probes Inc).16,17 Between 2 and 5 cells were exposed to alternating (2 Hz) 440- and 490-nm excitation wavelengths (Delta-Scan, Photon Technology International Inc). Fluorescence emission (510 nm) was recorded by a digital camera. By use of ImageMaster software (Photon Technology International), 4 pictures were averaged and fluorescence ratios obtained every 3 seconds from manually defined regions of intereSt. Rates of lactate transport (mmol · L−1 · min−1) were calculated as described previously.16
Data are presented as mean±SEM. Sigmastat and SigmaPlot (SPSS Inc) were used for calculations (Student’s t test) and graphic presentation.
Animals Show Characteristics of CHF
Six weeks after surgery, MI rats showed all signs of postinfarction CHF (Table 1), including pulmonary congestion, tachypnea, and pleural effusion (data not shown). The high LV end-diastolic pressure is closely associated with slowing of contraction of the viable posterior wall of the LV and dilation of the left atrium, as we visualized with echocardiography in this model.18
Altered Expression of Membrane Proteins
In CHF, the mRNA signal for MCT1 was 181±26% of that in Sham (P<0.05), whereas no significant difference was detected for the connexin43 mRNA level (Figure 1A and 1B). On Western blots, MCT1 protein was increased to 259±39% of that in Sham, whereas the connexin43 protein was reduced to 20±3% (Figure 1C and 1D). Of the glucose transporters, GLUT1 displayed a modest increase and GLUT4 none.
Immunofluorescence of MCT1
Confocal laserscan images of both cryostat sections and isolated cardiomyocytes from Sham hearts revealed strong MCT1 immunofluorescence along the plasma membrane and disks (Figure 2). Regular striations at 2-μm distance was indicative of labeling associated with T tubules. Cells from CHF hearts displayed strongly increased labeling of all surface membrane structures. The striation pattern was less distinct, because diffuse labeling was seen throughout the cytoplasm.
Immunogold Cytochemistry of MCT1 and Connexin43
Consistent with the light microscopic observations and our previous investigation,6 immunogold labeling of MCT1 occurred in both disk and nondisk regions of the Sham hearts. Both membrane domains showed 6 to 7 times higher linear density of gold particles in CHF, as estimated by the 2 different quantification methods (Table 2 and Figure 3). The increase in gold particle density was as pronounced in the transition zone between infarcted and preserved tissue (Figure 3D) as in the rest of the heart (including septum, Figure 3E, and the free walls of the left and right ventricles, Figure 3B and 3C). In no case were MCT1-associated gold particles observed over gap junctions (Figure 3D). The pattern of changes was the same in all animals.
CHF hearts also showed a high density of gold particles over intracellular membranes that were unlabeled in Sham and in normal hearts.6 This novel pool of MCT1 was associated with cisternae, often dilated (Figure 4A). Some labeled cisternae were seen near T tubules (Figure 4, A through C), indicating that they were part of the sarcoplasmic reticulum. Others coincided with a meshwork of tubules and vesicles beneath the plasma membrane at the site of the trans-Golgi network19 (Figure 4A). The remaining intracellular organelles (eg, mitochondria) lacked labeling above background. The extensive and novel presence of gold particles may explain the more diffuse labeling seen on the laserscan images.
In contrast, gold particles revealing connexin43 were substantially fewer in CHF than in Sham, but in both cases, particles were observed exclusively along gap junctions (Figure I, see online at http://circ.ahajournals.org).
Lactate Uptake Is Increased in Isolated Cardiomyocytes
The initial rates of lactate transport into cardiomyocytes averaged 3.49±0.46 in CHF cells and 1.18±0.19 in Sham (1:100 fluorescence ratio units/s) at 2 mmol/L l-lactate (P<0.05). At 20 mmol/L l-lactate, the corresponding values were 10.90±0.48 in CHF cells and 5.46±0.26 in Sham cells (P<0.05; Figure 5, A and B). The maximum fluorescence change at either concentration was the same for both cell types, indicating that the initial pH gradients across the sarcolemma were similar. At 2 mmol/L l-lactate, addition of 5 mmol/L α-cyano-4-hydroxycinnamate completely blocked the response, whereas at 20 mmol/L, inhibition was 90% (Figure 5C).
Initial rates of fluorescence changes at increasing l-lactate concentrations were transformed into rates of lactate uptake and fitted by nonlinear regression to the Michaelis-Menten equation7,16 (Figure 5C), giving Vmax values of 107±15 mmol · L−1 · min−1 in CHF cells and 42±6 mmol · L−1 · min−1 in Sham cells (P<0.05). The corresponding Km values were 9.1±2.9 and 7.1±2.5 mmol/L, respectively (P=NS).
Cardiomyocytes that survive MI are ≈30% larger than the Sham cells, and average cell volumes equal 22 and 29 pL, respectively.9 If it is assumed that the intracellular fluid volume into which lactate equilibrates is ≈50% of this volume, maximum transport rates per cell will average ≈26 and 8 fmol/s in CHF and Sham cells, respectively, compatible with the presence of ≈3 times more functional MCT1 transporters in CHF cells than in Sham cells.
Using a rat model of post-MI CHF, we show that the lactate transport capacity into cardiomyocytes was greatly increased, which could be ascribed to the presence of more MCT1 protein in both the disk and nondisk regions of the plasma membrane. A high rate of MCT1 protein synthesis was reflected in labeling of the trans-Golgi network, showing MCT1 molecules in transit to the cell surface. Another part of the new intracellular pool of MCT1 was associated with those sarcoplasmic membranes that face the T tubules. The functional role of these MCT1 molecules needs to be explored. In contrast to Brooks et al,20 our immunogold analysis did not reveal any MCT1 specifically associated with mitochondria.
Representativity of Tissue Specimens
Our rat MI model probably closely reflects the human condition of CHF. The MI comprised most of the free wall of the LV, but the viable tissue of the LV that survived MI was devoid of significant fibrosis (our electron microscope observations and Semb et al9). Cell shortening of the viable cardiomyocytes is reduced as in failing human hearts, and the gene expression is altered toward a fetal pattern.21 We have identified increased expression of atrial natriuretic peptide, endothelin, the Na+/Ca+ exchanger, and the α3-subunit of the Na,K-ATPase both at mRNA and peptide/protein levels.9,22 Tissue for microscopy was taken from the viable posterior free wall and from the septum.
Alterations of Disk Proteins
MCT1 in the disk region may constitute a metabolic junction at which lactate can easily flow from one cell to another.6 Because MCT1 in CHF is upregulated equally in disk and nondisk membranes, we must conclude that the upregulation does not serve the specific purpose of facilitating intercellular lactate exchange. Furthermore, we can exclude an indiscriminate control of disk proteins, because the connexin43 protein was strongly downregulated. Our data also extend the observations by Peters et al,23 because we demonstrate that the reduction of connexin43 was also apparent in the nonfibrotic and nonischemic viable tissue of the LV.
The upregulation of the MCT1 protein exceeded the increase in MCT1 mRNA by a factor of ≈2, whereas downregulation of the connexin43 proteins occurred in the absence of detectable changes of the corresponding mRNA transcript. Thus, it is possible that some control of protein expression also occurs at sites distal to transcription.
Lactate Transport and Metabolism in the Failing Heart
CHF hearts consume less oxygen than normal hearts despite the higher wall tension, but the myocardium still consumes lactate, also during exercise.24,25 The eccentric hypertrophy of the LV associated with MI is mostly due to elongation of the cardiomyocytes.9,21 Hence, capillarization and diffusion distances for nutrients and oxygen would not be different from Sham, and accordingly, the oxygenation of myoglobin is reported to be normal in failing swine hearts even during stimulation with catecholamines.3 Thus, in tissue that survives MI, the energy status is aerobic, and these hearts are able to metabolize lactate.
Because CHF hearts have switched their metabolic preference from mainly fatty acids to carbohydrates, it is of note that the MCT1 transporter and not the GLUT transporters showed a major upregulation. This could indicate that lactate has become an important substrate in these hearts. Normal hearts with a high workload exposed to elevated concentrations of lactate and fatty acids maintain a high carbohydrate oxidation in face of the blockade of the glycolysis imposed by the fatty acids. Thus, because lactate bypasses the glycolysis, it functions as a major anaplerotic substance in addition to providing ATP and allows tight metabolic control of the energy metabolism in the cell.26 Pyruvate, which is transported by MCT1, can also have this effect.27 Pyruvate increases the strength of contraction by increasing the Ca2+ load of the sarcoplasmic reticulum.28 The increased function of MCT1 in heart failure could contribute to the salutary effects of supplying pyruvate to CHF patients.29
Increased lactate concentration in the blood might provide a specific stimulus for increasing the transport capacity. This fits with upregulation of MCT1 in all parts of the myocardium and not only in the zone bordering on the infarcted tissue. Furthermore, in trained rats, MCT1 expression, together with that of citrate synthase, is increased in both heart tissue and slow twitch muscle.30 Thus, the MCT1 seems to be linked to oxidative metabolism. Two MCT isoforms are probably present in the heart, but only MCT1 has been identified.7,31 Expression of the other unidentified MCT isoform could also be changed in the CHF rats.
We conclude that cardiomyocytes from CHF rats exhibit increased synthesis of MCT1 protein and insertion of functional MCT1 transporters in the entire sarcolemma. An increased amount of transporters is also seen in the disk region, where there is a profound downregulation of the connexin43 protein. Hence, conduction of the action potential will be slower, as reported,8 whereas lactate uptake and oxidation can proceed faster.
This study was supported by Anders Jahre’s Fund for the Promotion of Sciences, the Research Council of Norway, the Rakel and Otto Chr. Bruun’s Fund, The British Heart Foundation, and the EU Biomed Program (BMH-CT9696-0851). We thank Kathrine Andersen, Morten Eriksen, Unni Lie Henriksen, Hilde Olsen, Thea Sandsbråten Solum, Gerd Torgersen, Annlaug Ødegaard, Bjørg Riber, Karen Marie Gujord, and Gunnar Lothe for expert technical assistance.
Figure I can be found online at http://www.circulationaha.org.
Bishop SP, Altschuld RA. Increased glycolytic metabolism in cardiac hypertrophy and congestive failure. Am J Physiol. 1970; 218: 153–159.
Sack MN, Rader TA, Park S, et al. Fatty acid oxidation enzyme gene expression is downregulated in the failing heart. Circulation. 1996; 94: 2837–2842.
Murakami Y, Zhang Y, Cho YK, et al. Myocardial oxygenation during high work states in hearts with postinfarction remodeling. Circulation. 1999; 99: 942–948.
Wilson JR, Mancini DM, Dunkman WB. Exertional fatigue due to skeletal muscle dysfunction in patients with heart failure. Circulation. 1993; 87: 470–475.
Jóhannsson E, Nagelhus EA, McCullagh KJA, et al. Cellular and subcellular expression of the monocarboxylate transporter MCT1 in rat heart: a high-resolution immunogold analysis. Circ Res. 1997; 80: 400–407.
Halestrap AP, Price NT. The proton-linked monocarboxylate transporter (MCT) family: structure, function and regulation. Biochem J. 1999; 343: 281–299.
Poole RC, Sansom CE, Halestrap AP. Studies of the membrane topology of the rat erythrocyte H+/lactate cotransporter (MCT1). Biochem J. 1996; 320: 817–824.
Sedarat F, Xu L, Moore ED, et al. Colocalization of dihydropyridine and ryanodine receptors in neonate rabbit heart using confocal microscopy. Am J Physiol. 2000; 279: H202–H209.
Berod A, Hartman BK, Pujol JF. Importance of fixation in immunohistochemistry: use of formaldehyde solutions at variable pH for the localization of tyrosine hydrolase. J Histochem Cytochem. 1981; 29: 844–850.
Ruud HK, Blackstad TW. PALIREL, a computer program for analyzing particle-to-membrane relations, with emphasis on electron micrographs of immunocytochemical preparations and gold labeled molecules. Comput Biomed Res. 1999; 32: 92–122.
Wang X, Levi AJ, Halestrap AP. Kinetics of the sarcolemmal lactate carrier in single heart cells using BCECF to measure pHi. Am J Physiol. 1994; 267: H1759–H1769.
Wang X, Levi AJ, Halestrap AP. Substrate and inhibitor specificities of the monocarboxylate transporters of single rat heart cells. Am J Physiol. 1996; 270: H476–H484.
Sjaastad I, Sejersted OM, Ilebekk A, et al. Echocardiographic criteria for detection of post-infarction congestive heart failure in rats. J Appl Physiol. 2000; 89: 1445–1454.
Iida H, Shibata Y. Delivery of lectin-labeled membrane to the trans-Golgi network and secretory granules in cultured atrial myocytes. J Histochem Cytochem. 1989; 37: 1885–1892.
Brooks GA, Brown MA, Butz CE, et al. Cardiac and skeletal muscle mitochondria have a monocarboxylate transporter MCT1. J Appl Physiol. 1999; 87: 1713–1718.
Wasserstrom JA, Holt E, Sjaastad I, et al. Altered E-C coupling in rat ventricular myocytes from failing hearts 6 wk after MI. Am J Physiol. 2000; 279: H798–H807.
Peters NS, Green CR, Poole-Wilson PA, et al. Reduced content of connexin43 gap junctions in ventricular myocardium from hypertrophied and ischemic human hearts. Circulation. 1993; 88: 864–875.
Braunwald E. Pathophysiology of heart failure.In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine. Philadelphia, Pa: WB Saunders Co; 1997: 394–418.
Traverse JH, Melchert P, Pierpont GL, et al. Regulation of myocardial blood flow by oxygen consumption is maintained in the failing heart during exercise. Circ Res. 1999; 84: 401–408.
Goodwin GW, Taegtmeyer H. Improved energy homeostasis of the heart in the metabolic state of exercise. Am J Physiol. 2000; 279: H1490–H1501.
Poole RC, Halestrap AP. Transport of lactate and other monocarboxylates across mammalian plasma membranes. Am J Physiol. 1993; 264: C761–C782.
Hermann HP, Zeitz O, Keweloh B, et al. Pyruvate potentiates inotropic effects of isoproterenol and Ca2+ in rabbit cardiac muscle preparations. Am J Physiol. 2000; 279: H702–H708.
Baker SK, McCullagh KJA, Bonen A. Training intensity-dependent and tissue-specific increases in lactate uptake and MCT-1 in heart and muscle. J Appl Physiol. 1998; 84: 987–994.