Supranormal Myocardial Creatine and Phosphocreatine Concentrations Lead to Cardiac Hypertrophy and Heart Failure
Insights From Creatine Transporter–Overexpressing Transgenic Mice
Background— Heart failure is associated with deranged cardiac energy metabolism, including reductions of creatine and phosphocreatine. Interventions that increase myocardial high-energy phosphate stores have been proposed as a strategy for treatment of heart failure. Previously, it has not been possible to increase myocardial creatine and phosphocreatine concentrations to supranormal levels because they are subject to tight regulation by the sarcolemmal creatine transporter (CrT).
Methods and Results— We therefore created 2 transgenic mouse lines overexpressing the myocardial creatine transporter (CrT-OE). Compared with wild-type (WT) littermate controls, total creatine (by high-performance liquid chromatography) was increased in CrT-OE hearts (66±6 nmol/mg protein in WT versus 133±52 nmol/mg protein in CrT-OE). Phosphocreatine levels (by 31P magnetic resonance spectroscopy) were also increased but to a lesser extent. Surprisingly, CrT-OE mice developed left ventricular (LV) dilatation (LV end-diastolic volume: 21.5±4.3 μL in WT versus 33.1±9.6 μL in CrT-OE; P=0.002), substantial LV dysfunction (ejection fraction: 64±9% in WT versus 49±13% in CrT-OE; range, 22% to 70%; P=0.003), and LV hypertrophy (by 3-dimensional echocardiography and magnetic resonance imaging). Myocardial creatine content correlated closely with LV end-diastolic volume (r=0.51, P=0.02), ejection fraction (r=−0.74, P=0.0002), LV weight (r=0.59, P=0.006), LV end-diastolic pressure (r=0.52, P=0.02), and dP/dtmax (r=−0.69, P=0.0008). Despite increased creatine and phosphocreatine levels, CrT-OE hearts showed energetic impairment, with increased free ADP concentrations and reduced free-energy change levels.
Conclusions— Overexpression of the CrT leads to supranormal levels of myocardial creatine and phosphocreatine, but the heart is incapable of keeping the augmented creatine pool adequately phosphorylated, resulting in increased free ADP levels, LV hypertrophy, and dysfunction. Our data demonstrate that a disturbance of the CrT-mediated tight regulation of cardiac energy metabolism has deleterious functional consequences. These findings caution against the uncritical use of creatine as a therapeutic agent in heart disease.
Received July 4, 2005; revision received July 29, 2005; accepted August 29, 2005.
Heart failure is characterized by a high prevalence and poor prognosis.1 Although medical therapy (eg, with angiotensin-converting enzyme inhibitors and β-blockers) significantly improves survival, mortality has remained high, and new, more effective forms of treatment are urgently needed. The recent failure of treatment trials with endothelin-1 receptor blockers and cytokine antibodies has fueled the search for alternative options in long-term pharmacological therapy to further improve the limited prognosis of heart failure patients.
The concept of the failing heart being energy starved was first proposed several decades ago.2 Results of large clinical heart failure trials have consistently shown that energy-costly treatments, such as positive inotropic agents (β-receptor mimetic drugs and phosphodiesterase inhibitors),3 increase mortality, whereas energy-sparing treatments, such as angiotensin-converting enzyme inhibitors,4 angiotensin II blockers,5 or β-receptor blockers,6 reduce mortality.7,8 A number of studies in both humans and animal models have all shown that phosphocreatine (PCr) and free creatine levels decrease by up to 60% to 70% in heart failure, independent of the underlying etiology but in relation to the severity of contractile dysfunction.9–15 Impairment of high-energy phosphate availability in creatine kinase (CK)–16 or in guanidinoacetate-N-methyltransferase17–knockout mice has adverse consequences for myocardial function. Failing hearts with an impaired CK system are unable to maintain low levels of free ADP and thus, high phosphorylation potential during inotropic stimulation.18
Clinical Perspective p 3139
If, as these studies suggest, cardiac high-energy phosphate metabolism does indeed contribute to contractile dysfunction in heart failure, then new forms of treatment specifically directed toward improving cardiac energetics may constitute a new therapeutic principle in heart failure. A recent review of the subject concluded that “What is probably most needed is a strategy for improving energy metabolism in the failing heart. Novel methods to augment [ATP] or [PCr] stores are critically needed.”8 However, to date, no intervention has succeeded in achieving increased cardiac high-energy phosphate levels. ATP synthesis by the mitochondria is subject to tight feedback regulation,19 and therefore, ATP levels have never been observed to increase to above-normal levels under any circumstance. PCr levels in the heart are mainly determined by the size of the total creatine pool, approximately two thirds of which, in the absence of ischemia, is present in phosphorylated form as PCr. Creatine is not synthesized in the heart but is taken up by cardiomyocytes against a large concentration gradient via the Na+-creatine cotransporter (creatine transporter, CrT).20 Most likely, downregulation of creatine transport is the major mechanism leading to creatine and PCr depletion in the failing heart.21 We have previously attempted to increase myocardial creatine and PCr levels in normal and failing hearts by feeding rats a diet containing 3% creatine.22 However, myocardial creatine levels did not increase because the CrT was downregulated in response to increased extracellular creatine content.23,24 Thus, to date, no therapeutic intervention has been reported to increase myocardial creatine and PCr levels.
To achieve this goal, we created 2 lines of CrT-overexpressing (CrT-OE) mice, which exhibited up to 4-fold increases in myocardial creatine concentrations. Surprisingly, we report that an increase of creatine and PCr to supranormal levels has detrimental effects, because the heart is unable to keep the augmented creatine pool adequately phosphorylated, resulting in increased free ADP levels, left ventricular (LV) hypertrophy, and dysfunction. Thus, this article demonstrates the importance of the tight regulation of myocardial high-energy phosphate stores in the heart.
Construction of CrT-Transgenic Mice
The 1.9-kb open reading frame of rabbit CrT cDNA (accession No. X67252; a gift from Dr M. Kilimann, Bochum, Germany) was amplified with the use of Turbo Pfu (Stratagene). Polymerase chain reaction (PCR) products, generated with primers for CrT (forward, 5′-gga att cac gca gcc acc atg gcg aag aag agc gc-3′, which contains a flanking EcoRI restriction site and a Kozak consensus sequence, and reverse, 5′-gct gct cta gat cac aga tcc tct tct gag atg agc ttc tgt tcc atg aca ctc tcc acc acg ac-3′, which contains the human c-myc epitope tag, a translational stop signal, and an XbaI restriction site), were digested with EcoRI and XbaI overnight and ligated into the EcoRI/XbaI site of a construct comprising pBluescript II, a 2.1-kb KpnI/EcoRI fragment of the murine ventricular myosin light chain 2 (MLC2v) promoter, and a 590-bp MluI/NotI fragment of pBK-cytomegalovirus (simian virus 40 intron and polyA addition signal sequence; a gift from Dr Hend Farza, Oxford University, Oxford, England). The resulting construct, MLC-CrT-myc, was confirmed by nucleotide sequencing. Functionality of the transporter was verified by subcloning CrT-myc sequences into pcDNA3.1 (Invitrogen) and measuring creatine uptake in transiently transfected HEK293 cells.
The 4.7-kb linear DNA insert of MLC-CrT-myc was excised by digestion with BssHII and gel purified. Insert DNA was diluted to a concentration of 2 ng/μL in 5 mmol/L Tris and 0.1 mmol/L EDTA, pH 7.5, and ≈2 pL was microinjected into male pronuclei of fertilized C57BL/6J mouse oocytes before transfer into pseudopregnant females. Tail DNA from 3-week-postpartum pups was screened for the presence of the transgene by BamHI digestion and Southern blot analysis by using a biotin-labeled 708-bp SacI fragment of rabbit CrT-myc as a probe. All transgenic mice (CrT-OE) and their wild-type (WT) littermates used in the experiments were generated by mating male transgenic mice with C57BL/6 females. Animals were genotyped by PCR (CrTtg forward, 5′-gcatcttcatcttcaacatcgtgta-3′ and reverse, 5′-tcacagatcctcttctgagatgag-3′; glyceraldehyde 3-phoaphate dehydrogenase, forward 5′-ccttcattgacctcaactacatgg-3′ and reverse, 5′-ggc agcaccagtggatgcagg-3′) under standard conditions.
Mice were kept in cages with a 12-hour light/dark cycle and controlled temperature (20°C to 22°C) and were fed creatine-free chow and water ad libitum. All studies were performed in 5- to 9-month-old mice and were conducted in accordance with the UK Home Office Animals (Scientific Procedures) Act of 1986.
Quantitative Assessment of CrT-myc Gene Expression
Total RNA was extracted from heart tissue with the use of the RNeasy kit (Quiagen). RNA was treated with proteinase K and DNase I to prevent protein and genomic DNA contamination. Real-time reverse transcription (RT)–PCR was performed (Quantitect SYBR green RT-PCR kit, Quiagen) to measure CrT-myc mRNA (RTtg forward, 5′-gtgacagatggcgggatgtat-3′ and reverse, 5′-cggcttgtagtacacgatgttg-3′) relative to β-actin mRNA (RT actin forward, 5′-gacaggatgcagaaggagattact-3′ and reverse, 5′-tgatccacatctgctggaaggt-3′) in transgenic mice with use of the Rotor-Gene system (Corbett Research Ltd). Real-time RT-PCRs were performed in duplicate with 2 different RNA concentrations for each sample. Standard curves to quantify CrT-myc and β-actin mRNAs (in arbitrary units) were generated from serial dilutions of linear MLC-CrT-myc plasmid and WT RNA, respectively.
Quantitative Assessment of ANP and β-MHC Gene Expression
Atrial natriuretic peptide (ANP) and β-myosin heavy-chain (β-MHC) mRNA expression was measured in LV heart tissue as described earlier with previously described primers.25 Results, in arbitrary units, were normalized to the expression levels of the housekeeping gene MLN51.26
CrT-OE (n=20) mice were compared with WT littermate controls (n=10) at 5 to 9 months of age. A detailed description and validation of our 3D echocardiographic technique have been published.27 In brief, mice were anesthetized with 1.25% to 1.5% isoflurane in 100% O2 administered via nose cone and placed in the left lateral decubitus position on a heated platform designed with micrometer-precision rotation and translation capabilities. An Agilent Sonos 5500 system was equipped with a 7/15-MHz linear-array probe secured in a holder device with 360° freedom of movement and needle electrodes for ECG and respiratory gating. Multiple, consecutive, short-axis views were acquired for the duration of a cardiac cycle by translating the mouse platform in 500-μm increments along the long axis of the heart. All images were stored digitally and subsequently reassembled in a 3D matrix with custom-designed software developed at our University (MV1). LV surfaces were reconstructed, and end-systolic and end-diastolic volumes (LVESV and LVEDV, respectively) were calculated at each time point. A single echocardiographer blinded to mouse genotype performed all image analyses.
To confirm our findings, a second CrT-OE line of mice (n=7) was compared with WT controls (n=6) by standard 2D echocardiography. Equipment and anesthesia were the same as used previously, but a single short-axis view was obtained at the level of the papillary muscles, and subsequently, LV ejection fraction (LVEF) was estimated from the fractional change in cavity cross-sectional area.
In Vivo LV Hemodynamics
These measurements were obtained, as previously described,17 immediately after the 3D echocardiogram under the same general anesthesia, which was maintained with 2% isoflurane during cannulation. The right carotid artery was dissected and cannulated with a 1.4F microtipped pressure cannula (SPR-839, Millar Instruments), which was advanced retrogradely into the LV cavity under echocardiographic guidance. Isoflurane was reduced to 1.25% to 1.5%, followed by a period of equilibration until hemodynamic indices were stable for >15 minutes. Measurements were recorded on a Powerlab 4SP data recorder (AD Instruments) and stored digitally for later analysis with PVAN software, version 3.0 (Millar Instruments). At the end of the experiment, mice were killed by cervical dislocation, and organs were blotted and weighed. The LV was snap-frozen in LN2 for molecular and biochemical analyses.
In Vivo MRI
High-resolution murine magnetic resonance cine imaging (cine MRI) was performed in vivo to assess LV mass and volumes, as previously described.28 Eight to 10 short-axis slices, covering the heart from base to apex, were acquired (spatial resolution, 100 μm×100 μm×1 mm; temporal resolution, 4.6 ms) with a cardiac-triggered and respiration-gated fast low-angle-shot sequence on a 11.7-T high-field MR system (see section on 31P magnetic resonance spectroscopy [MRS] for system details).
31P-MRS Experiments in Perfused Hearts
Isolated, perfused heart experiments were performed as previously described17 (constant pressure of 80 mm Hg at 37°C) with Krebs-Henseleit buffer containing (in mmol/L) 11 d-glucose, 4.5 pyruvate, and 0.5 lactate as substrates. Experiments were performed in a vertical-bore, 11.7-T (500-MHz) MR system (Magnex Scientific) with a Bruker Avance console (Bruker Medical) as previously described,17 a pulse-and-collect sequence at a repetition time of 2 seconds, and a flip angle of 60°. 31P-MR spectra of 150 averages each (5 minutes per spectrum) were acquired and added manually in postprocessing, where applicable, and were corrected for partial saturation. PCr, ATP, and inorganic phosphate (Pi) were quantified with respect to the signal intensity of a phenylphosphonic acid standard (PPA; Sigma-Aldrich) solution in the intraventricular balloon. Because of a low signal-to-noise ratio, Pi and pHi could be assessed only in summed spectra for all hearts for each genotype. After termination of the protocol, the balloon volume was increased twice with a known amount of PPA, and a 31P spectrum was recorded after each increase to relate signal intensity to the amount of PPA. Finally, hearts were freeze-clamped, and total creatine was measured by high-performance liquid chromatography and corrected for perfusion. PPA, PCr, ATP, and Pi were quantified by the AMARES algorithm and MRUI software. To calculate baseline concentrations during control perfusion, 15-minute spectra were used. pHi was determined from the chemical shift between the Pi and PCr peaks from the equation pH=6.72−log[(δ−5.72)/(3.17−δ)]. The free cytosolic ADP concentration (in μmol/L) and the free-energy change of ATP hydrolysis (ΔG; in kJ/mol) were calculated from the CK equilibrium assumption as previously described.29 Given that Pi and pHi could be assessed only in summed spectra for all hearts studied for each genotype by MRS, group mean values for Pi and pH were used for the calculation of individual heart ADP and ΔG values.
Total creatine content was measured by high-performance liquid chromatography as previously described.14 To calculate the intracellular creatine concentration, we assumed the total heart protein content to be 0.17 g protein per gram wet weight and the intracellular volume to be 0.5 mL/g wet weight.14 Total CK activity and CK isoenzyme composition were measured as previously described.14
All data are presented as mean±SD. Statistical significance between groups was assessed by Student’s t test, and Pearson correlation calculations were used to determine correlation coefficients. Differences were considered significant at P<0.05.
Characteristics of CrT-OE Mice and WT Littermate Controls
Six transgenic founders were identified. Analysis by real-time RT-PCR showed that 2 lines had high but variable levels of transgene expression. The remaining 4 lines had low or undetectable levels of expression and were not investigated further. In the 2 lines with high transgene expression, we observed up to a 4-fold increase in intracellular cardiac creatine levels (WT mean, 66±6 nmol/mg protein; range, 58 to 74; Tg55 mean, 133±52; range, 65 to 240; Tg46 mean, 100±18; range, 68 to 118). Southern blot analysis of transgenic genomic DNA suggested that this increase was not due to multiple sites of transgene integration or differences in transgene copy number (data not shown). Moreover, levels of CrT-myc mRNA measured in cardiac tissue by real-time RT-PCR were correlated closely with intracellular creatine (n=26, r=0.75, P<0.0001), suggesting that transcriptional regulation or mRNA stability underlay the variability in creatine levels. The transgene was inherited autosomally, and levels of CrT-myc mRNA when compared between male (n=16) and female (n=10) hearts were not significantly different (P=0.10). There were no differences in survival between WT and CrT-OE mice from weaning until the time when they were used for experiments, ie, until 8 months of age. All CrT-OE data shown subsequently are from line Tg55 mice unless specified otherwise.
In Vivo Cardiac Function
Comparisons of morphometric data, 3D echocardiography, and LV hemodynamics are shown in Table 1. There were no significant differences in body weight between CrT-OE and WT mice or in heart rates during physiological measurements. However, wet heart weight was 24% higher in CrT-OE mice, involving significant elevations of both LV and RV weights. Echocardiography showed that the LVEDV and LVESV were also significantly larger in transgenic mice. These differences were not simply due to an overall physiologically larger heart, because echocardiographically derived LVEF was significantly impaired (P<0.003) in CrT-OE mice (mean, 49±13%; range, 22% to 70%) compared with WT mice (mean, 64±9%; range, 49% to 78%), LVEDP was significantly elevated, and lung weights were also higher. In the upper quartile of CrT-OE hearts (total creatine content, 208±32 nmol/mg protein; n=5), functional changes were much more pronounced (eg, LVEF, 33±8%; LVEDP, 12±3 mm Hg). Furthermore, Pearson correlation analysis showed a significant, positive relation between LV creatine concentration and LV weight (r=0.59), RV weight (r=0.63), LVEDV (r=0.51), and LVEDP (r=0.52) and a significant negative correlation with LVEF (r=−0.74). Other hemodynamic indices showed a trend toward impaired systolic function (dP/dtmax) and diastolic dysfunction (dP/dtmin, τ) but failed to reach statistical significance owing to the heterogeneous phenotype (ranges are shown in Table 1) in the CrT-OE group. However, these parameters were all significantly correlated with LV creatine content: dP/dtmax (r=−0.69), dP/dtmin (r=0.66), and τ (r=0.71; Figure 1). Furthermore, all of these parameters were significantly changed in the upper quartile of CrT-OE hearts (dP/dtmax mean, 6028±326 mm Hg/s; dP/dtmin mean, −4184±594 mm Hg/s; mean τ, 30±5 ms). Taken together, our data clearly suggest that CrT-OE mice developed LV hypertrophy, dilatation, and systolic as well as diastolic dysfunction, resulting in congestive heart failure proportionate to the concentration of creatine in the heart. Echocardiographic findings were confirmed by in vivo cine MRI (data not shown). Examples of WT and CrT-OE cardiac MR images illustrating LV dilatation and dysfunction are shown in Figure 2.
The cardiac hypertrophy markers ANP and β-MHC, measured in arbitrary units by real-time RT-PCR and normalized to the housekeeping gene MLN51, were elevated in the CrT-OE animals (mean ANP mRNA: CrT-OE [n=24], 3.2±3.5; WT [n=11], 1.1±0.5, P=0.007; mean β-MHC mRNA: CrT-OE [n=29], 4.9±6.1; WT [n=11], 0.6±0.5, P=0.001). Moreover, elevations of ANP and β-MHC were correlated with LV creatine (ANP r=0.61, P<0.0001; β-MHC r=0.80, P<0.0001).
To exclude the possibility that the observed LV dysfunction was a nonspecific consequence of transgene overexpression (eg, because of gene disruption at the site of integration), we performed standard echocardiography in a second line of CrT-OE mice (Tg46). This line also exhibited a significantly reduced LVEF compared with WT controls (51±10% versus 61±6%, P<0.05) and an increased end-diastolic area (0.137±0.017 versus 0.116±0.014 cm2, P<0.05).
Cardiac Energetics in WT and CrT-OE Mice
31P-MR spectra from CrT-OE and WT hearts are shown in Figure 2, and mean data are given in Table 2. In WT hearts, PCr and ATP concentrations as well as PCr-ATP ratios and pHi values were normal, as previously reported.17 A major finding was that CrT-OE hearts showed PCr concentrations that were on average 1.4-fold and up to 2.4-fold higher than in WT hearts. CrT-OE hearts showed a trend toward reduced ATP levels, but this was not significant (P=0.07). Pi levels were too low to quantify reliably in individual experiments. Spectra from all experiments were therefore pooled for each of the 2 genotypes. Analysis of these spectra indicated no difference in Pi (1.3 versus 1.1 mmol/L in WT and CrT-OE hearts, respectively). Likewise, pHi was similar for both groups. Free ADP concentration, calculated from the CK equilibrium assumption, was 78±35 μmol/L in WT mice and was increased ≈2-fold, to 148±7 μmol/L (P=0.004), in the upper quartile of CrT-OE mice. Furthermore, free ADP concentration was correlated significantly with total creatine content (r=0.66, P=0.0005; Figure 3A). Accordingly, ΔGATP also was correlated inversely with total creatine content (r=−0.51, P=0.012; Figure 3B) and was reduced from −59.6±1.4 kJ/mol in WT mice to −57.7±0.5 kJ/mol (P=0.002) in the upper quartile of CrT-OE hearts. Finally, total creatine content was correlated inversely with the phosphorylated fraction of creatine (PCr/total creatine; Figure 3C; r=−0.61, P=0.001), showing that as the creatine pool accumulated, there was a decrease in the fraction that was phosphorylated. Total CK activity and CK isoenzyme distribution were unchanged in CrT-OE mice (Table 2).
Definition of the Model: Hypertrophy and Heart Failure in the CrT-OE Mouse
In this report, we describe a new animal model with supranormal levels of myocardial creatine and PCr. We achieved this by overexpressing the CrT under the control of the MLC2v promoter. We created 2 independent lines of transgenic mice (Tg55 and Tg46) and clearly demonstrated that overexpression of CrT led to increases in creatine and PCr to supraphysiological levels. Levels of creatine were closely correlated with transgene mRNA expression. Thus, transgenic overexpression of CrT allowed us to overcome the natural feedback system of the heart in which downregulation of endogenous CrT in response to increased extracellular creatine concentration keeps the myocardial creatine and PCr concentrations constant.22,23 Our model then allowed us to study the physiological effects of increased creatine and PCr levels in the heart.
The original rationale for our study was to create a mouse model that would be protected from cardiac stress by increased energy storage and transport capacity. Although CrT-OE mice with supranormal creatine and PCr levels showed normal growth curves (body weight), to our surprise, we found that such mice developed global cardiac hypertrophy, LV dilatation, and LV dysfunction. Using noninvasive in vivo echocardiography, cine MRI, invasive hemodynamics, and real-time RT-PCR, we clearly detected all signs typical of chronic heart failure: increased lung and heart weight, increased LVEDV and LVESV, reduced LVEF, and elevated expression of ANP and β-MHC. Furthermore, indices of hypertrophy and systolic dysfunction were strongly correlated with the degree of LV creatine accumulation, as were the indices of diastolic dysfunction. We also confirmed LV dysfunction in an independent, second CrT-OE mouse line. These findings in 2 lines and the strong correlations of myocardial creatine levels with parameters of LV dysfunction clearly suggest that supranormal creatine and PCr levels lead to the progressive development of chronic heart failure. The key question then arises: how do increased levels of creatine and PCr cause heart failure?
Cardiac Energy Metabolism in CrT-OE Mice
Our analysis of cardiac energy metabolism has shown that CrT-OE mice showed a trend toward reduced levels of ATP, increased levels of creatine (up to 4-fold) and PCr (up to 2.4-fold), unchanged Pi and pHi, and unchanged total CK and CK isoenzyme activities. The maximum capacity of the energy reserve via CK is best estimated as the product of total CK activity and total creatine content.30 Because total CK activity was unchanged, maximal energy reserve via CK was increased in CrT-OE mice in proportion to the degree of creatine accumulation. Thus, heart failure in these mice does not appear to be related to the function of the CK energy shuttle, which is not impaired and is indeed probably augmented in CrT-OE mice. Instead, it is likely that the functional deterioration that occurred in CrT-OE mice was related to the second crucial function of CK as an energy buffer system, which serves to keep free ADP levels low and the free energy change of ATP hydrolysis (ΔGATP) levels high.8,29 In support of this argument, we found a significant correlation between total creatine and free ADP and between total creatine and ΔGATP. In addition, we found an ≈2-fold increase in free ADP concentration (148 μmol/L compared with 78 in WT) and a lower ΔGATP (−57.7 kJ/mol, compared with −59.6 in WT) in the CrT-OE hearts with the highest total creatine levels. Thus, ADP levels were substantially increased and ΔGATP levels were reduced in CrT-OE mice. Importantly, this thermodynamic situation is similar to the one described in hypertrophy and some models of heart failure.18,31 Thus paradoxically, the effort to optimize energy storage and transport capacity in the heart by increasing total creatine stores led to a situation of energetic imbalance, which is likely to contribute to the development of heart failure in CrT-OE mice.
In WT mice, on average, 67% of the total creatine pool was phosphorylated as PCr, whereas this percentage in CrT-OE mice was only 46% in the upper quartile of CrT-OE mice. Why does the heart fail to keep an augmented creatine pool adequately phosphorylated? If the cardiomyocyte is considered a closed space that allows free mixing of all chemical components, then mitochondrial ATP synthesis would only require several seconds to produce sufficient high-energy phosphate bonds to phosphorylate the augmented creatine pool in CrT-OE hearts. However, cardiomyocytes are highly compartmentalized, and we speculate that the increased creatine levels resulted in impaired intracellular PCr and creatinefree trafficking. This may result in an increased PCr-creatinefree ratio (and thus, higher ATP/ADP) in the perimitochondrial space, thereby inhibiting ATP synthesis, and decreased PCr-creatinefree (and thus, lower ATP/ADP) near sites of ATP utilization, thus contributing to contractile failure. The relatively small changes observed for ΔG should also be interpreted in this context. Although the average decrease in ΔG was only moderate (from −59.6 to −57.7 kJ/mol) and as such insufficient to directly cause contractile dysfunction, we speculate again that ΔG in the relevant ATP-utilizing compartments may be decreased to a much larger extent, thereby contributing to contractile dysfunction.
An alternative/additional explanation could come from the fact that both ATP and PCr show significant concentration changes during the cardiac cycle.32,33 Given that the CK equilibrium constant heavily favors ATP synthesis over PCr synthesis, it is likely that such cyclical changes are exacerbated for PCr and blunted for ATP in the presence of an augmented PCr pool. It is conceivable that cyclical changes in ATP and PCr concentrations are important mediators of intracellular signal transduction mechanisms, and their alteration may have consequences for multiple signaling pathways. Although such mechanisms should be explored in future studies, with current MRS technology it is not feasible to assess cyclical changes in high-energy phosphates in the mouse heart, given that we have previously demonstrated that cyclical changes in the rat heart (which has a 10-fold larger myocardial mass than the mouse heart) are just above the MRS signal-to-noise threshold of detectability.33 It should also be noted that the immature myocardium has very low concentrations of creatine, which accumulate during maturation. Early heart failure did not seem to occur in our model, and we did not observe any elevation in fetal reabsorption or early postnatal mortality. However, elevated fetal creatine may predispose the heart to subsequent failure. Future studies incorporating fetal ultrasound techniques might clarify this point. Our data clearly indicate that the well-described phenomenon of CrT downregulation in response to increases in extracellular creatine concentration23,24 is a protective mechanism that prevents the deleterious consequences of uncontrolled cardiac creatine overload.
Implications for Derangement of Energy Metabolism in Heart Failure
What lessons can be learned from our model with regard to the role of energy metabolism in heart failure? First, CrT-OE hearts present a situation wherein free ADP is increased and ΔGATP is reduced, as is also the case in hypertrophy and some models of heart failure.18,31 Our finding that heart failure can be induced by a purely metabolic intervention that primarily alters free ADP supports the concept that energetic failure leads to contractile dysfunction in chronic heart failure. However, although increased ADP values in heart failure can occur in the presence of reduced creatine levels, an increased ADP level in CrT-OE mice was observed in the presence of increased creatine stores. Thus, creatine and PCr levels are tightly regulated and maintained within a narrow normal range in the healthy, intact myocardium. We postulate that any disturbance of this fine balance, irrespective of the direction of change, leads to energetic and subsequently functional impairment.
Importantly, our studies cannot be interpreted to suggest that a therapeutic strategy aiming at restoring reduced creatine and PCr levels in the failing heart back to normal would be deleterious. In fact, one cannot predict whether such a strategy would have beneficial or adverse effects. Some studies have suggested that the decrease of total creatine in heart failure may be a compensatory mechanism counteracting the increase in free ADP levels.8,10 On the other hand, the decrease in total creatine reduces cellular ATP transfer, ie, the energy reserve via CK, and this may impair cardiac function because of the failure to deliver sufficient high-energy phosphates to the sites of ATP utilization.9 In any case, our data clearly suggest that uncritical attempts to raise creatine levels to supranormal levels in heart failure would have to be approached with great caution.
Limitations and Future Studies
It would be important to measure and calculate PCr, creatine, free ADP, and the resulting ΔG in the relevant subcellular compartments to test our speculation regarding the exacerbation of energetic imbalance in the relevant compartments. Unfortunately, currently no methodology exists that can provide an assessment of the energetic environment in the cellular microcompartments of interest. Future studies of the CrT-OE model should establish the temporal relation between energetic imbalance and contractile dysfunction in a longitudinal time-course study. Finally, it remains to be demonstrated whether heart failure development would be altered if CrT-OE mice were subjected to chronic coronary ligation or aortic banding. Such studies should be done in mice showing a mild to moderate increase in creatine that per se does not lead to substantial LV dysfunction and would test the effect of a mildly augmented creatine pool on heart failure development.
In summary, we have shown that increases in myocardial creatine and PCr levels by transgenic overexpression of the CrT lead to the development of chronic heart failure. The most likely mechanism for this is the inability of the heart to keep the augmented creatine pool phosphorylated, thus leading to increased ADP, reduced ΔGATP levels, and subsequent contractile failure.
This study was supported by the British Heart Foundation.
Braunwald E. The Denolin lecture: congestive heart failure: a half century perspective. Eur Heart J. 2001; 22: 825–836.
Neubauer S, Clarke K. Cardiac energy metabolism in heart failure. In: Marber M, ed. Cardiac Metabolism and Ischemic Cardiomyopathy. Puteaux: Wolters Kluwer Health; 2003: 45–61.
Ingwall JS, Weiss RG. Is the failing heart energy starved? on using chemical energy to support cardiac function. Circ Res. 2004; 95: 135–145.
Liao R, Nascimben L, Friedrich J, Gwathmey JK, Ingwall JS. Decreased energy reserve in an animal model of dilated cardiomyopathy: relationship to contractile performance. Circ Res. 1996; 78: 893–902.
Nascimben L, Friedrich J, Liao R, Pauletto P, Pessina AC, Ingwall JS. Enalapril treatment increases cardiac performance and energy reserve via the creatine kinase reaction in myocardium of Syrian myopathic hamsters with advanced heart failure. Circulation. 1995; 91: 1824–1833.
Nascimben L, Ingwall JS, Pauletto P, Friedrich J, Gwathmey JK, Saks V, Pessina AC, Allen PD. Creatine kinase system in failing and nonfailing human myocardium. Circulation. 1996; 94: 1894–1901.
Zhang J, Wilke N, Wang Y, Zhang Y, Wang C, Eijgelshoven MH, Cho YK, Murakami Y, Ugurbil K, Bache RJ, From AH. Functional and bioenergetic consequences of postinfarction left ventricular remodeling in a new porcine model: MRI and 31P-MRS study. Circulation. 1996; 94: 1089–1100.
Neubauer S, Krahe T, Schindler R, Horn M, Hillenbrand H, Entzeroth C, Mader H, Kromer EP, Riegger GA, Lackner K, Ertl G. 31P magnetic resonance spectroscopy in dilated cardiomyopathy and coronary artery disease: altered cardiac high-energy phosphate metabolism in heart failure. Circulation. 1992; 86: 1810–1818.
Shen W, Asai K, Uechi M, Mathier MA, Shannon RP, Vatner SF, Ingwall JS. Progressive loss of myocardial ATP due to a loss of total purines during the development of heart failure in dogs: a compensatory role for the parallel loss of creatine. Circulation. 1999; 100: 2113–2118.
Nahrendorf M, Spindler M, Hu K, Bauer L, Ritter O, Nordbeck P, Quaschning T, Wiesmann F, Hiller KH, Wallis J, Ertl G, Bauer WR, Neubauer S. Creatine kinase knockout mice show left ventricular hypertrophy and dilatation, but unaltered remodeling post-myocardial infarction. Cardiovasc Res. 2005; 65: 419–427.
ten Hove M, Lygate CA, Fischer A, Schneider JE, Sang AE, Hulbert K, Sebag-Montefiore L, Watkins H, Clarke K, Isbrandt D, Wallis J, Neubauer S. Reduced inotropic reserve and increased susceptibility to cardiac ischemia/reperfusion injury in phosphocreatine-deficient guanidinoacetate-N-methyltransferase-knockout mice. Circulation. 2005; 111: 2477–2485.
Liu J, Wang C, Murakami Y, Gong G, Ishibashi Y, Prody C, Ochiai K, Bache RJ, Godinot C, Zhang J. Mitochondrial ATPase and high-energy phosphates in failing hearts. Am J Physiol Heart Circ Physiol. 2001; 281: H1319–H1326.
Wyss M, Wallimann T Creatine metabolism and the consequences of creatine depletion in muscle. Mol Cell Biochem. 1994; 133/134: 51–66.
Horn M, Remkes H, Dienesch C, Hu K, Ertl G, Neubauer S. Chronic high-dose creatine feeding does not attenuate left ventricular remodeling in rat hearts post-myocardial infarction. Cardiovasc Res. 1999; 43: 117–124.
Boehm E, Chan S, Monfared M, Wallimann T, Clarke K, Neubauer S. Creatine transporter activity and content in the rat heart supplemented by and depleted of creatine. Am J Physiol. 2003; 284: E399–E406.
Loike JD, Zalutsky DL, Kaback E, Miranda AF, Silverstein SC. Extracellular creatine regulates creatine transport in rat and human muscle cells. Proc Natl Acad Sci U S A. 1988; 85: 807–811.
Gaussin V, Tomlinson JE, Depre C, Engelhardt S, Antos CL, Takagi G, Hein L, Topper JN, Liggett SB, Olson EN, Lohse MJ, Vatner SF, Vatner DE. Common genomic response in different mouse models of β-adrenergic-induced cardiomyopathy. Circulation. 2003; 108: 2926–2933.
Dawson D, Lygate CA, Saunders J, Schneider JE, Ye X, Hulbert K, Noble JA, Neubauer S. Quantitative 3-dimensional echocardiography for accurate and rapid cardiac phenotype characterization in mice. Circulation. 2004; 110: 1632–1637.
Neubauer S. Cardiac magnetic resonance spectroscopy. In: Lardo A, Fayad Z, Chronos NAF, Fuster V, eds. Cardiovascular Magnetic Resonance: Established and Emerging Applications. London: Martin Dunitz; 2003: 39–60.
Tian R, Nascimben L, Ingwall JS, Lorell BH. Failure to maintain a low ADP concentration impairs diastolic function in hypertrophied rat hearts. Circulation. 1997; 96: 1313–1319.
Fossel ET, Morgan HE, Ingwall JS. Measurement of changes in high-energy phosphates in the cardiac cycle using gated 31P nuclear magnetic resonance. Proc Natl Acad Sci U S A. 1980; 77: 3654–3658.
Heart failure is associated with deranged cardiac energy metabolism, including reductions in creatine and PCr, and this phenomenon has been postulated as a mechanism contributing to contractile dysfunction. We report on the first animal model that allows the creation of supranormal levels of myocardial PCr and creatine. We achieved this by overexpressing the myocardial CrT in mice. Compared with WT littermate controls, total creatine was increased up to 4-fold in CrT-OE hearts. PCr levels were also increased, but to a smaller extent. Surprisingly, hearts with excess creatine and PCr levels developed substantial LV dilatation, dysfunction, and hypertrophy. The most likely explanation for this finding is that CrT-OE hearts showed an ≈2-fold increase in free ADP concentration; ie, the heart was incapable of keeping an augmented creatine pool adequately phosphorylated. Our data demonstrate that a disturbance of the CrT-mediated tight regulation of cardiac energy metabolism has adverse functional consequences and that therapeutic strategies to increase creatine to supranormal levels may be deleterious. However, the possibility remains that strategies to maintain creatine and PCr at normal levels in heart failure may be beneficial, and we are planning to test this hypothesis in future studies.