In Vivo Alterations in Cardiac Metabolism and Function in the Spontaneously Hypertensive Rat Heart

Introduction

The World Health Organisation estimates that in 2004, 7.2 million people died worldwide from coronary heart disease¹. A major risk factor for coronary heart disease is hypertension. This is characterized by an increase in arterial blood pressure and a subsequent increase in left ventricular (LV) wall stress. To protect the hypertensive heart, pro-hypertrophic factors increase LV wall thickness and equalize wall stress via the LaPlace law². However, if left untreated the heart will ultimately fail due to an inability to maintain normal cardiac output².

In the healthy heart, β-oxidation contributes 60-80% of ATP^3,4. The remaining 20-40% is derived from carbohydrate metabolism, principally the metabolism of glucose and lactate. However β-oxidation and glucose oxidation are dynamically regulated to provide the most efficient energy production with the most abundant fuel⁵. This ensures the heart can continue to function under changing metabolic environments. Recent evidence suggests that many forms of heart disease are linked to alterations in cardiac metabolism⁶. For example, it has been suggested that in the hypertrophied heart there are changes in the profile of substrate utilization, limiting energy provision and eventually leading to a deterioration in heart function^6,7.

Spontaneously hypertensive rats (SHRs) are used as a model of hypertension and hypertrophy^8,9. The combination of hypertension and hypertrophy result in continuing adaptation of cardiac metabolism^8,9. During early development of hypertension (5-10 weeks), the SHR heart displays compensated alterations in glycolytic intermediates and high energy phosphates, which lead to a normalized myocardial energy state⁹. At 15 weeks of age SHRs represent an early compensatory disease model of hypertrophy with elevated blood pressure and concentric hypertrophy¹⁰.

The SHR has a number of genetic mutations which contribute to the hypertensive phenotype. A major mutation is a loss of function of fatty acid translocase (FAT/CD36), a sarcolemmal fatty acid transporter¹¹. FAT/CD36 is crucial for the transport of long chain fatty acids (LCFA) across the plasma membrane. Studies in the CD36-deficient mouse have revealed a significant increase in basal glycolysis and glucose uptake in the heart¹². Fatty acid uptake and oxidation are also severely impaired in these mice, resulting in elevated plasma free fatty acids (FFA) and triglycerides¹². This is mimicked in the SHR rat at 15 weeks where ¹³C magnetic resonance spectroscopy (MRS) on the perfused ex vivo SHR heart has been used to track a reduction in ¹³C label incorporation into the TCA cycle from fatty acids¹³. Furthermore, using the radioactive fatty acid analogue, BMIPP (¹²³I-β-methylp-iodophenylpentadecanoic acid), a significant reduction in fatty acid uptake and oxidation was observed in the SHR, similar to the CD36^−/− mouse¹⁴. This in vitro evidence demonstrates a major shift in cardiac metabolism away from β-oxidation in the 15 week old SHR heart^13,14.

At this age, the SHR also shows a significant 60% reduction in the inducible pyruvate dehydrogenase (PDH) kinase (PDK)4 mRNA expression¹⁵. PDK4 is one of the key regulatory enzymes of PDH in the heart. In type-1 diabetes and high fat feeding, elevated mRNA and protein expression of PDK4 has been shown to decrease both in vitro PDH activity and in vivo PDH flux^16-19. Thus, given the decrease in PDK4 in the SHR heart, we would hypothesise that in vivo PDH flux would be increased.

Progression into sustained hypertension and hypertrophy shows that by 25 weeks the SHR heart has a reduced lactate/pyruvate ratio and altered high energy phosphate levels, indicative of perturbed energy homeostasis⁹. As the SHR heart ages to 16-18 months, the animals display severe insulin resistance and increasingly compromised cardiac energetics²⁰.

The nature of hypertension and hypertrophy mean that there are a multitude of factors affecting disease progression and severity. This is particularly difficult to mimic using ex vivo models. Only the in vivo setting provides the correct balance of hormonal, functional and regulatory responses to fully characterize the disease. ¹³C MRS is particularly suited to the study of metabolism in the heart due to the extensive range of metabolites that can be observed¹³, but suffers from an inherently low sensitivity which has limited in vivo applications. However, this low sensitivity can be overcome using hyperpolarized ¹³C MRS^21-23. Hyperpolarization, via dynamic nuclear polarization (DNP) increases the sensitivity of ¹³C MRS upwards of 10,000-fold over conventional methods²¹. A hyperpolarized compound can be injected in vivo and its metabolism visualized in real time²¹. For example, in vivo hyperpolarized [1-¹³C]pyruvate flux through PDH into [¹³C]bicarbonate and ¹³CO₂ has been shown to accurately reflect PDH activity¹⁶. [1-¹³C]pyruvate is also processed into [1-¹³C]lactate and [1-¹³C]alanine, via lactate dehydrogenase (LDH) and alanine aminotransferase (AAT), respectively. In contrast to [1-¹³C]pyruvate, [2-¹³C]pyruvate is able to probe the entry of pyruvate derived acetyl-CoA into the TCA cycle via visualization of [1-¹³C]acetylcarnitine, [1-¹³C]citrate and [5-¹³C]glutamate^24,25.

In this study our aim was to determine in vivo flux through PDH in the SHR heart at a compensated stage of hypertrophy to investigate the balance between anaerobic glycolytic metabolism and glucose oxidation in vivo. The findings of this work were supported by in vivo cine-MRI measurements of heart function and mass, and in vitro biochemical assays of enzyme activity and protein expression.

Methods

[1-¹³C] and [2-¹³C]pyruvic acid were obtained from Sigma Aldrich (Sigma-Aldrich Company Ltd. Dorset, UK). Fifteen week old male spontaneously hypertensive rats (SHR) (~300g, n=13) and male Wistar (275-325g, n=11) control rats were obtained from Harlan UK. All animals were housed on a 12:12-hour light–dark cycle and studies were performed between 7a.m.-1p.m., during the early absorptive (fed) state. All investigations conformed to Home Office Guidance on the Operation of the Animals (Scientific Procedures) Act (HMSO) of 1986, to institutional guidelines and the Directive 2010/63/EU of the European Parliament. Experiments were also approved by a local university ethics review board.

Animal handling

Anaesthesia was induced by 2.5-3% isofluorane in oxygen and nitrous oxide (4:1, total of 2l/min). Adequacy of anaesthesia was initially based on loss of pedal response and was then continuously monitored through evaluation of the heart and respiration rate. Anaesthesia was maintained by means of 2% isofluorane delivered during the experiment (O₂:NO₂ 4:1, total of 2l/min). When required, a tail vein cannulation for intravenous administration of hyperpolarized solutions was performed before the animals were subsequently placed in a home-built MR animal-handling system²⁶. ECG, respiration rate and body temperature were monitored throughout the experiment and air heating was provided to maintain body temperature at 37°C as previously described¹⁷. Animals fully recovered following anaesthesia for serial MR experiments measurements.

Cine magnetic resonance imaging

Animals were initially imaged to assess left ventricular mass (LVM) and cardiac function. They were positioned in an 11.7T vertical bore MR scanner (Magnex scientific, Oxon UK) interfaced to a Bruker Avance console (Bruker Medical, Ettlingen Germany) and a shielded gradient system (Magnex scientific, UK). A 60mm birdcage transmit/receive RF coil was used to obtain MR images (Rapid biomedical, Rimpar Germany). Experiments were carried out as previously described²⁷. Sequences were ECG-triggered and 28–40 frames were collected per cardiac cycle. Left ventricular volumes were derived using the free-hand draw function in ImageJ (NHI, USA). For each heart, LVM, ejection fraction, stroke volume and cardiac output were calculated. Hypertrophy was assessed as an increase in LVM compared to control animals.

A high temporal resolution (HTR) cine-MRI sequence was used to measure diastolic peak ejection and filling rates in SHR and control animals²⁸. Briefly, after acquisition of the conventional cine-MRI sequences, a mid-papillary slice was selected and imaged using HTR cine-MRI over two consecutive cardiac cycles. Chamber volume was measured in every frame after threshold adjustment, using ImageJ. This was then plotted as a function of time and early and late peak filling rates were calculated.

Pyruvate polarization and dissolution

Approximately 40mg of [1-¹³C]pyruvic acid or [2-¹³C]pyruvic acid, doped with 15mM trityl radical (OXO63, GE Healthcare) and 3μl Dotarem (1:50 dilution, Guerbet, Birmingham, UK), was hyperpolarized in a polarizer, with 45min of microwave irradiation as previously described²¹. The sample was subsequently dissolved in a pressurized and heated alkaline solution, containing 2.4g/L sodium hydroxide and 100mg/L EDTA dipotassium salt (Sigma-Aldrich), to yield a solution of 80mM hyperpolarized sodium [1-¹³C]pyruvate or [2-¹³C]pyruvate with a polarization of ~30% or ~20% respectively^17,24.

Hyperpolarized ¹³C MRS protocol

On two separate days animals received either a [1-¹³C]- or [2-¹³C]pyruvate injection and MR scan, with at least one day between scans. A home-built ¹H/¹³C butterfly coil (loop diameter, 20mm) was placed over the rat chest, localizing signal from the heart¹⁷. Rats were positioned in a 7T horizontal bore MR scanner interfaced to an Inova console (Varian Inc, Yarnton, UK). Correct positioning was confirmed by the acquisition of an axial proton fast low angle shot image. An ECG-gated shim was used to reduce the proton linewidth to ~120Hz. One ml of either hyperpolarized [1-¹³C]- or [2-¹³C]pyruvate was injected over 10s into the anesthetised rat. Sixty individual ECG-gated ¹³C MR pulse-acquire cardiac spectra were acquired over one min after injection (TR, 1s; excitation flip angle, 5°; sweep width, 13,593Hz; acquired points, 2,048; frequency centred on the C1 pyruvate resonance).

¹³C-MRS data analysis

All cardiac ¹³C spectra were analysed using the AMARES algorithm in the jMRUI software package²⁹. Spectra were DC offset-corrected based on the last half of acquired points. The peak areas of [1-¹³C]pyruvate, [1-¹³C] lactate, [1-¹³C]alanine, [¹³C]carbon dioxide (¹³CO₂) and [¹³C]bicarbonate (for [1-¹³C]pyruvate) and [2-¹³C]pyruvate, [1-¹³C]acetylcarnitine, [1-¹³C]citrate and [5-¹³C]glutamate (for [2-¹³C]pyruvate) at each time point were quantified and used as input data for a kinetic model. The kinetic model developed for the analysis of hyperpolarized [1-¹³C] and [2-¹³C]pyruvate MRS data is based on a model initially developed by Zierhut et al. (2010) and further developed by Atherton et al. (2011)^30,31.

The peak areas of in vivo ¹³CO₂ and [¹³C]bicarbonate were also used to estimate intracellular pH (pH_i) in the heart. Using the method described by Schroeder et al. (2010)³², two single 1s spectra were summed to increase signal to noise. The first 16 time points were then plotted to estimate pH_i³².

Tissue collection

One day after hyperpolarized scans, animals were anaesthetized using an intraperitoneal injection of pentobarbital sodium (150mg/kg body weight, Euthatal, Merial, UK). The beating heart was rapidly removed and placed in ice cold phosphate buffered saline (PBS, Sigma-Aldrich). A subset of hearts were perfused in the Langendorff mode to assess cardiac energetics using ³¹Phosphorus (³¹P) MRS and were therefore placed in ice cold Krebs buffer, instead of PBS. The remaining hearts were blotted (to remove liquid), freeze clamped and stored at −80°C for subsequent biochemical analysis.

³¹P MRS

Six control and six SHR hearts were perfused using a retrograde Langendorff perfusion mode at 85 mmHg³³. A dual tune ¹H/³¹P volume coil was used for ³¹P MRS on an 11.7T vertical bore MR scanner. Fully relaxed scans were performed to assess phosphocreatine (PCr), ATP, P_i and pH_i. Cardiac ³¹P MR spectra were analysed using the AMARES algorithm in the jMRUI software package²⁹. Absolute ³¹P metabolite concentrations were calculated by assigning ATP concentrations ([ATP]) of 10.6mmol/L to the initial γ-ATP peak area for Wistar control hearts³⁴ and expressing all other ATP, PCr and P_i peak areas relative to this area^32,35. pH_i was calculated from the P_i chemical shift³⁵, and this was compared with the in vivo pH_i derived from the ratio of ¹³CO₂ and [¹³C]bicarbonate³².

Western blotting

Protein expression of PDK (1, 2 and 4) and PDP (1 and 2) were measured in total heart homogenates from both SHR and Wistar control animals using SDS-PAGE and Western blotting ³³. Consistent protein loading was ensured by measuring protein concentration beforehand and confirmed using Ponceau staining of the membrane. All samples were run in duplicate on separate gels to confirm results. All data were normalized to control hearts.

Pyruvate dehydrogenase activity assay

PDH activity was measured using a spectrophotometric assay as previously described¹⁸. Briefly, this assay measures the proportion of the PDH enzyme in the active form and the maximal PDH activity. Activity is shown as μmoles NADH reduced per minute per gram wet weight (g.w.w.).

Statistics

All results are expressed as the mean ± SEM. Analysis for statistical significance was performed using an unpaired two-tailed Student’s t-test assuming equal variance (SPSS, IBM, USA). Significance was taken at p<0.05.

Results

Hypertrophy and adaptive changes in cardiac function in the SHR heart

Cine-MRI was used to confirm the hypertrophic phenotype in the SHR animals. Left ventricular mass (LVM) was significantly increased by 56% in the SHR versus control (Table-1). This translated to a significant increase in the LVM to body weight ratio (53%), indicative of hypertrophy (Table-1, Figure-1). Cine-MRI was also used to assess in vivo cardiac function. Stroke volume and end diastolic volume were significantly increased by 19% and 18%, respectively, whilst heart rate was reduced by 18% in the SHR. This resulted in no overall change in cardiac output or ejection fraction indicative of compensated hypertrophy.

Table-1. Function data from cine-MRI.

	LVM (mg)	Body weight (g)	LVM/BW (mg.g−1)	Heart Rate (BPM)	Stroke Volume (μl)	End diastolic volume (μl)	Cardiac Output (ml.min−1)	Ejection fraction (%)
Control	540 ± 20	306 ± 10	1.77 ± 0.04	419 ± 7	261 ± 16	325 ± 19	105 ± 6	81 ± 1
SHR	840 ± 20 ***	312 ± 5	2.71 ± 0.06 ***	343 ± 14 ***	310 ± 14 *	382 ± 19 *	104 ± 5	81 ± 2

A significant decrease in heart rate was observed in the SHR, with an increase in stroke volume. Ejection fraction and cardiac output were unchanged between the SHR and control hearts. Values are mean data with S.E.M. Two-tailed Student t-test

^*p<0.05,

^***p<0.001.

Figure-1.

A typical mid-papillary MR image of the control and SHR heart, during diastole and systole. Significant hypertrophy of the SHR heart can be seen compared to the control heart.

A mid-papillary slice was selected to assess diastolic function using HTR-cine MRI. HTR-cine is able to assess filling of the LV following end systole²⁸. Whilst the peak early filling rate was unaltered between groups (Figure-2A), the peak late filling rate, which represents atrial filling of the LV, was significantly reduced by ~0.3ml.s⁻¹ in the SHR heart (p<0.05, Figure-2B).

Figure-2. High Temporal Resolution cine-MRI was used to assess diastolic function.

A) No change is seen in the peak early filling rate. B) A significant decrease in peak late filling rate (panel B) was seen in SHRs. * : p<0.05

In vivo PDH flux was increased in the SHR heart

¹³C label incorporation into CO₂ and bicarbonate pools (PDH flux) was significantly increased by 85% in the SHR heart compared to control (Figure-3A). ¹³C label incorporation into alanine was also significantly increased by 28%, whilst lactate labelling was unchanged between groups. To confirm the increase in ¹³C label incorporation into CO₂ and bicarbonate, an in vitro biochemical assay was performed to measure PDH activity. Activity of the active PDH fraction in the SHR heart is significantly elevated from 0.8±0.2 to 1.8±0.3μmol.min⁻¹.g.w.w.⁻¹, an increase of 135% compared to control (Figure-3B). Maximal PDH activity was unchanged between groups (3.6±0.8 and 5.5±1.0μmol.min⁻¹.g.w.w.⁻¹, for controls and SHR hearts, respectively). This indicates PDH flux is increased in the SHR heart in vivo and that changes to ¹³C label incorporation into CO₂ and bicarbonate are due to more of the enzyme being active rather than an increase in maximal PDH activity and content.

Figure-3. Assessed flux using [1-¹³C]pyruvate and PDH activity.

A) Hypertensive rats show an increase in ¹³C label incorporation through PDH (CO₂ + bicarbonate) and through AAT (alanine), whilst there is no change to LDH (lactate). B) PDH activity assay, showing activity of active PDH fraction. Significant increase is seen in the activity of PDH in the SHR compared to control. *: p<0.05, **: p<0.01

In vivo assessment of TCA cycle intermediates in the SHR heart

In the SHR heart, [2-¹³C]pyruvate incorporation in the SHR heart was significantly increased by 72% into citrate, 118% into acetylcarnitine and 81% into glutamate compared to control (Figure-4A). As there was a significant increase in PDH flux, this increased ¹³C label incorporation may be a result of more labelled acetyl-CoA being processed in to the TCA cycle. To correct for this increase in ¹³C label flux, [2-¹³C]pyruvate data was normalized to [1-¹³C]pyruvate incorporation into bicarbonate+CO₂. Following normalization, no change to ¹³C label incorporation was detected into either acetylcarnitine, glutamate or citrate pools (Figure-4B). Therefore the observed increase in [2-¹³C]pyruvate label incorporation was relative to increased PDH flux and not greater flux in the TCA cycle.

Figure-4. Assessed flux using [2-¹³C]pyruvate.

A) Hypertensive rats show an increase in label incorporation into the citrate, glutamate and acetylcarnitine pools. *: p<0.05, **: p<0.01, ***:p<0.001. B) [2-¹³C]pyruvate data normalized to label flux through PDH.

Normal phosphate energetics and pH_i in the in vitro SHR heart

To assess high energy phosphate energetics in the heart, ³¹P MRS was used performed on the perfused control and SHR hearts. In the control hearts, the average [ATP] was 10.6±1.2mmol/L, [PCr] was 16.4±2.7mmol/L and [P_i] was 1.9±0.3mmol/L (Figure-5). In the SHR hearts, the average [ATP] was 9.3±0.5mmol/L, [PCr] was 14.7±1.7mmol/L and [P_i] was 3.6±0.9mmol/L. [ATP], [PCr] and [P_i] were not significantly different between groups. This led to an average PCr/ATP ratio of 1.8±0.1 and 1.6±0.1, for control and SHR hearts respectively.

Figure-5. ³¹P MRS of the in vitro heart.

High energy phosphate energetics in the SHR and control heart are unaltered as represented by a normal [ATP], [PCr], [P_i] and PCr/ATP ratio.

The estimated in vivo pH_i from the ¹³C-MRS data was 7.12±0.02 and 7.14±0.02 for control and SHR, respectively. The pH_i measured from ³¹P MRS was 7.06±0.01 and 7.06±0.02 for control and SHR, respectively, these data showed that there was no significant difference in cardiac intracellular pH between groups.

Increases in PDH flux are a result of changes in regulatory enzymes

To assess the regulatory state of PDH, western blots were performed on the key regulators of PDH in the heart: PDP(1 and 2) and PDK(1, 2 and 4). Assessment of PDK1 and 4 protein expression, showed that in the SHR heart there was a 26% and 25% reduction in protein expression, respectively, compared to control hearts (Figure-6). PDK2 was also reduced but not significantly (p=0.15). PDP1 protein expression showed a non-significant decrease (p=0.07), whilst PDP2 was significantly reduced by 29% between controls and SHR hearts (Figure-6).

Figure-6. Altered protein expression of the PDKs and PDP2 in the SHR heart.

To assess regulatory state of PDH, PDK1, PDK2, PDK4, PDP1, and PDP2 protein expression was measured. PDK1 and PDK4 expression was significantly reduced in the SHR heart compared to control. No significant alteration in protein expression was observed in PDP1 (p=0.07), however PDP2 was significantly decreased. *: p<0.05 **: p<0.01, data displayed as relative expression compared to control (arbitrary units).

Discussion

Extensive studies of the SHR heart have shown that after 15 weeks they develop concentric hypertrophy due to pressure overload^15,36. In vivo measurements in this study using cine-MRI, have confirmed a 56% increase in left ventricular mass (LVM) and 53% increase in LVM to body weight ratio. Also in vivo cardiac functional data indicates that the 15 week SHR is an early disease model with hyperfunctional compensatory remodelling³⁷. There are several key compensatory changes in function in the SHR compared with controls. Heart rate is significantly lower in the SHR, whilst stroke volume and end diastolic volume are significantly elevated. The result is that cardiac output and ejection fraction remain unchanged between groups, indicative of compensatory remodelling in response to the hypertension³⁷. To accompany this, HTR cine-MRI showed that there was a significant reduction in peak late filling rate. This is hypothesised to be the rate of filling of the LV from the left atrium²⁸. These data indicate a compensated hypertrophy disease stage in the 15 week SHR, with normalized cardiac function.

Further to compensated functional changes in the SHR, there are also several metabolic alterations which maintain normal energetics. In vivo assessment of SHR cardiac metabolism was performed using hyperpolarized ¹³C MRS and highlighted several alterations. Firstly we demonstrate that the SHR heart has increased modulation of both in vivo PDH flux and AAT. This was detected as a significant increase in ¹³C label incorporation between [1-¹³C]pyruvate and [¹³C]bicarbonate+¹³CO₂ (PDH) and [1-¹³C]alanine (AAT) in the in vivo hypertensive heart. An in vitro PDH activity assay was also employed to determine PDH activity, which was significantly elevated in the SHR heart, confirming hyperpolarized PDH flux data.

¹³C label incorporation was also significantly increased into alanine, via AAT. These data suggest that in the SHR heart there is either a larger pool of free alanine in the tissue or an increase in AAT activity. There are several possible explanations for this increased ¹³C label incorporation into alanine, which include export of cardiac alanine to the liver for the glucose-alanine cycle³⁸ or increased alanine production as would be required for the increase in protein synthesis characteristic of hypertrophy. Also, several papers have linked the presence of alanine in the blood/urine to increases in blood pressure, so the observed increase here could be a consequence of the underlying hypertension^39,40.

The increase in PDH flux was coupled with an increase in ¹³C label incorporation into citrate, acetylcarnitine and glutamate pools. No change was observed in ¹³C label incorporation into the TCA cycle when label flux was normalized to flux through PDH. This implies that the apparent increases in citrate, acetylcarnitine and glutamate are derived from increased production of [2-¹³C]acetyl-CoA, consistent with an increase in glucose oxidation. Furthermore, it was expected that the label incorporation from [1-¹³C]pyruvate to [1-¹³C]lactate would increase in the SHR heart, which would have implied a shift to a glycolytic phenotype, however this was not the case. Both these pieces of evidence highlight that the SHR has compensated cardiac metabolism, with a switch to greater glucose oxidation.

A defect in the LCFA transporter (FAT/CD36) has previously been identified in the SHR¹¹, which leads to a reduction in β-oxidation. Labarthe et al. (2005) have shown a significant reduction in ¹³C label incorporation from [1-¹³C]oleate, a LCFA, into the TCA cycle in the SHR¹³. They also stressed the importance of medium CFA (MCFA) in the SHR to compensate for the reduced contribution of LCFA to β-oxidation^13,41. The defect in FAT/CD36 does not however, lead to a change in energetics in the heart. ³¹P MRS can be used to determine cardiac energetics where significant alterations in [ATP] or/and [PCr] can indicate a mismatch in energy production in the heart. As in vivo cardiac ³¹P MRS is particularly difficult due to sensitivity issues and cross contamination from other organs, the perfused heart was used to assess high energy phosphate energetics. Within the SHR heart [ATP] and [PCr] were not significantly different from controls hearts. This resulted in no significant difference in the PCr/ATP ratio, a measure of cardiac energetics within the heart. The SHR heart will be accustomed to higher arterial pressures and as the perfusion pressure was matched between the SHR and control animals, this may account for the absence of any difference in energetics ¹³. However a previous study, measuring absolute concentrations of PCr and other high energy phosphates, observed no difference at 15 weeks of age in the SHR heart, but this balance was disturbed at the later timepoint of 25 weeks ⁹. ³¹P and ¹³C MRS were also used here to determine the intracellular pH in the in vitro and in vivo heart, respectively³². These results indicate that the intracellular pH is unaltered between the SHR and control heart and the pH_i measurements were consistent with previously published data from the control Wistar heart³². This would imply normal aerobic metabolism and fits with no significant increase in the [1-¹³C]lactate flux.

To further understand the mechanism behind the increase in PDH activity several key regulatory proteins were probed. PDKs are one of the major regulatory factors of PDH activity in the heart^19,42. The PDK isoforms are activated by several mechanisms including hypoxia, increases in the end-products acetyl-CoA and NADH, and increases in β-oxidation^42-44. PDKs can be activated short term by these factors or long term by transcriptional up regulation^19,44. Here we observe a significant reduction in protein expression of PDK1 and PDK4 in the SHR heart compared to controls. Bowker-Kinley and colleagues have previously shown that PDK1 displays the highest specific activity and is highly sensitive to changes in the NAD⁺:NADH ratio, which will be disturbed in the SHR, due to the reduction in LCFA oxidation^11,42. Our data indicates that the reduction in PDK1 and 4 protein expression might be sufficient to elicit the large increase in PDH flux and activity demonstrated here.

Other mechanisms for PDH modulation also exist including increased PDP activity and end product inhibition. Interestingly, protein expression data here show that PDP1 content did not increase but rather there was potentially a reduction, although not significant (p=0.07). This was accompanied by a significant decrease in the calcium insensitive PDP2 in the SHR heart. Whilst it seems counter intuitive that PDP1 and PDP2 protein expression would be decreased in the SHR heart, there may be compensatory alterations in activity. Elevated intracellular calcium ([Ca²⁺]_i) and other cofactors (i.e. magnesium) have been shown to activate PDP, concurrently increasing PDH flux and the proportion of PDH in the active fraction⁴⁵. Previous data from the left ventricle of SHRs, have shown a significantly elevated concentration of [Ca²⁺]_i compared to Wistar rats³⁶. The final mechanism to consider is that the PDH increase is mediated by a reduction in end product inhibition. In the SHR heart there would be a reduction in acetyl-CoA and NADH derived from LCFA¹³, this would reduce end product inhibition on PDH. Resulting in an increase in PDH flux. The current evidence suggests that increased PDH activity in the SHR heart is modulated by changes to PDK1 and 4 expression as well as a reduction in acetyl-CoA end product inhibition due to the reduced oxidation of LCFA.

The results of this study highlight the unique ability of hyperpolarized MRS to investigate in vivo alterations in cardiac metabolism seen in the diseased heart. Hyperpolarized MRS also offers the possibility to monitor animals at different time points during the development of disease and permits in vivo assessment where previously terminal techniques would have been required, i.e. biochemical analysis or the perfused heart^13,46. The demonstration of in vivo metabolic alterations in animal models of cardiovascular diseases is made more relevant by the potential for future clinical application of this technology⁴⁷. Clinical application of hyperpolarized MRS would allow direct, in vivo assessment of alterations in cardiac metabolism in a range of cardiovascular diseases. However the limitations of this technique should also be considered, for example the rapid decay of the hyperpolarized signal requires fast tracer production and injection. There is also a general requirement for injection of supraphysiological concentrations of the tracer. Although in the case of pyruvate, as used here, the physiological effect of the injection has been shown to be minimal with the increase in the plasma pyruvate concentration limited to 0.25mM, only slightly above the normal physiological level^16,24,48.

Conclusions

As expected, 15 week SHRs had significant hypertrophy which was accompanied by compensatory remodelling. Functional alterations in stroke volume, heart rate and the late diastolic filling rate were consistent with compensated hypertrophy. The use of in vivo hyperpolarized spectroscopy techniques in this study showed a significant increase in ¹³C label incorporation through PDH into [¹³C]bicarbonate+¹³CO₂ and into the TCA cycle. Taken together with the [1-¹³C]lactate data, this would suggest that there is an increase in glucose oxidation through PDH rather than an increase in anaerobic metabolism in the SHR heart. Our work indicates a switch away from the classical, predominantly β-oxidation driven cardiac metabolism, matching evidence from previous in vitro data¹³. The genetic predisposition to reduced β-oxidation due to loss of CD36, may therefore drive the metabolic alterations seen in the SHR heart.