Most metabolic studies with hyperpolarized ¹³C media have used [1-¹³C]pyruvate polarized in its acid form with trityl radicals.¹ In vivo studies with hyperpolarized sodium lactate and acetate have used either trityl or TEMPO-type nitroxyl radicals, but these polarization methods have not been directly compared in vivo. In developing a hyperpolarized metabolic model of myocardial ischemia in the intact rat,² we compared pyruvate polarized as the sodium salt with 4-hydroxy-TEMPO (TEMPOL) with the trityl-polarized acid.

Pyruvate preparation and polarization. Sodium [1-¹³C]pyruvate and TEMPOL (final concentrations 2.7 M and 50 mM) were dissolved in HOD and glycerol (18% v/v) and frozen into glassy beads (~10 µl) in liquid nitrogen. A sample cup was charged with 12 beads, polarized at 7 T and 1.0 K with microwave irradiation (197.25 GHz) in a custom-built dissolution-DNP polarizer³ and then dissolved in 6 ml D₂O and rapidly transferred to a remotely-controlled phase separator/infusion pump.⁴ Neat [1-¹³C]pyruvic acid doped with 21 mM OX063 (17 µl) was frozen in a sample cup along with a separately frozen stoichiometric equivalent of 10 M NaOH for neutralization, negatively polarized at 196.80 GHz, and dissolved with 6 ml of D₂O containing 47 mM sodium phosphate, 100 mM NaCl, 2.7 mM KCl, 0.3 mM EDTA, pH 7.4. Dissolved pyruvate concentrations were determined by high-resolution ¹³C NMR spectroscopy.

Animal preparation. Male Wistar rats were anesthetized with isoflurane, intubated for mechanical ventilation, and catheters were installed in the femoral arteries, for sampling and blood pressure measurement, and in a femoral vein for infusion of the hyperpolarized solution. A left thoracotomy was performed and a suture thread was installed under the left coronary artery, threaded through tubing to create an occluding snare, but was left untightened. Pancuronium bromide (0.2 mg/kg) was administered to suppress irregular chest motion that interfered with respiration-triggered gating.

MR data acquisition and analysis. A ¹H/¹³C surface coil was positioned over the heart and the rat was placed inside a 9.4 T Varian / Magnex small animal scanner. Coil placement was confirmed by ¹H gradient-echo MRI shimming performed with FAST(EST)MAP using a voxel containing the anterior wall of the left ventricle. Infusion of 1.0 mL of the hyperpolarized [1-¹³C]pyruvate solution was started 3 s after triggering dissolution, along with the spectral acquisition, which consisted of a series of 40 pulse- and respiration-triggered ¹³C MRS scans using a 30º BIR-4 excitation pulse and a TR of ~3 s. [1-¹³C]Pyruvate, [1-¹³C]pyruvate hydrate and [1-¹³C]lactate, [1-¹³C]alanine and [¹³C]bicarbonate metabolite spectral signals were fit with Bayes and areas under the curve calculated using the spectral time series. Statistical significance was evaluated by 2-way ANOVA with Bonferroni correction for multiple comparisons.

Metabolites were readily detected using either preparation of pyruvate, including the lower-intensity bicarbonate signal. The signal-to-noise ratio (SNR) of TEMPOL-polarized pyruvate and its metabolites was lower (on average by 3.9- and 2.6- fold, respectively), reflecting the lower degree of polarization. The lactate signal, expressed as a fraction of the sum all the spectral signals (“total carbon signal”), was on average higher with TEMPOL-polarized pyruvate (Figure 1A, 0.149 ± 0.023 SD, n=9) compared to the trityl-polarized case (0.077 ± 0.026 SD, n=21, p < 0.0001). Bicarbonate and alanine signals were similarly found at a higher proportion of the total signal using TEMPOL-polarized pyruvate (Figure 1C/D), but not to a significant level. The ratio of the bicarbonate and lactate signals was on average nearly the same with either pyruvate formulation (Figure 1B). The greater metabolite signal fraction observed with the TEMPOL formulation partly compensates for its overall lower signal and demonstrates that the composition of the hyperpolarized pyruvate infusion can affect its metabolism in the heart, with a particularly clear effect on the lactate signal. Several factors may account for this difference. Pyruvate and pyruvic acid readily dimerize to form parapyruvate,^5,6 which inhibits lactate dehydrogenase⁷ and 2-oxoglutarate dehydrogenase.⁸ TEMPO can react with endogenous free radicals,⁹ and the trityl-polarized pyruvate solution has a higher osmolarity due to the buffer and salt in the dissolution medium. Glycerol, used as a glassing agent in the TEMPOL formulation, may increase perfusion by reducing blood viscosity.¹⁰ Identification of the component(s) responsible for the metabolic differences is underway. The fact that all the metabolites appear to be affected implies an effect at the tissue level, i.e. on pyruvate delivery or uptake in the myocardium.This study clearly demonstrates that the composition of the infused hyperpolarized pyruvate solution can significantly affect its metabolism in vivo.