We describe an experimental setup for administrating hyperpolarized 13C-labeled metabolites in continuous perfusion mode to an isolated perfused mouse heart. A dedicated 13C-NMR acquisition approach enabled the quantification of metabolic enzyme activity in real-time, and a multiparametric 31P-NMR analysis enabled the determination of the tissue ATP content and pH.
Metabolism is the basis of important processes in cellular life. Characterizing how metabolic networks function in living tissues provides crucial information for understanding the mechanism of diseases and designing treatments. In this work, we describe procedures and methodologies for studying in-cell metabolic activity in a retrogradely perfused mouse heart in real-time. The heart was isolated in situ, in conjunction with cardiac arrest to minimize the myocardial ischemia and was perfused inside a nuclear magnetic resonance (NMR) spectrometer. While in the spectrometer and under continuous perfusion, hyperpolarized [1-13C]pyruvate was administered to the heart, and the subsequent hyperpolarized [1-13C]lactate and [13C]bicarbonate production rates served to determine, in real-time, the rates of lactate dehydrogenase and pyruvate dehydrogenase production. This metabolic activity of hyperpolarized [1-13C]pyruvate was quantified with NMR spectroscopy in a model free-manner using the product selective saturating-excitations acquisition approach. 31P spectroscopy was applied in between the hyperpolarized acquisitions to monitor the cardiac energetics and pH. This system is uniquely useful for studying metabolic activity in the healthy and diseased mouse heart.
Alterations in cardiac metabolism are associated with a variety of cardiomyopathies and often form the basis of the underlying pathophysiological mechanisms1. However, there are numerous obstacles to studying metabolism in living tissues, as most biochemical assays require the homogenization of the tissue and cell lysis and/or radioactive tracing. Therefore, there is a pressing need for new tools to investigate myocardial metabolism in living tissues. Magnetic resonance (MR) of hyperpolarized 13C-labeled substrates allows for real-time measurements of metabolism in living tissues2, without the use of ionizing radiation, by increasing the MR signal-to-noise (SNR) ratio of the labeled site(s) by several orders of magnitude3. Here, we describe an experimental setup, an acquisition approach, and an analytical approach for studying the rapid metabolism in the isolated mouse heart and, in parallel, present indicators of general tissue energetics and acidity. The cardiac pH is a valuable indicator, as the acid-base balance is disrupted in the early stages of cardiac diseases and conditions such as myocardial ischemia, maladaptive hypertrophy, and heart failure6.
Hyperpolarized [1-13C]lactate and [13C]bicarbonate production from hyperpolarized [1-13C]pyruvate helps in determining the production rates of lactate dehydrogenase (LDH) and pyruvate dehydrogenase (PDH). Most of the previous studies performed using hyperpolarized substrates in the isolated rodent heart either used complex kinetic models to derive the enzymatic activity of LDH and PDH, or reported the signal intensity ratios of the hyperpolarized product to a substrate without calculating the actual enzyme activity rates2,4,5,6,7,8,9,10,11,12,13,14. Here, we used the product selective saturating-excitations approach15, which allows for the monitoring of the enzyme activity in a model-free manner15,16. In this way, the absolute enzymatic rates (i.e., the number of moles of product produced per unit of time) were determined. 31P spectroscopy was utilized to observe the signals of inorganic phosphate (Pi), phosphocreatine (PCr), and adenosine triphosphate (ATP). A multi-parametric analysis was used to characterize the pH distribution of the heart, as demonstrated by the heterogeneous chemical shift in the Pi signal of the tissue.
The retrogradely perfused mouse heart (Langendorff heart)17,18,19 is an ex vivo model for the intact beating heart. In this model, the heart viability and pH are preserved for at least 80 min20, and it has shown potential for recovery following a prolonged ischemic injury21,22. Nevertheless, inadvertent variability during micro-surgery may lead to variability in the tissue viability across hearts. Previous studies have reported on the deterioration of this heart over time19; for example, a reduction in contractile function of 5%-10% per hour has been observed18. The adenosine triphosphate (ATP) signal has previously been shown to report on the myocardial energetic status and viability23. Here, we noted that the perfused heart may occasionally show unintentional variability in viability levels, as demonstrated by the ATP content, despite the fact that we had an uninterrupted perfusion and oxygen supply. We demonstrate here that normalizing the LDH and PDH rates to the ATP content of the heart reduces the inter-heart variability in these rates.
In the following protocol, we describe the surgical procedure used for heart cannulation, isolation, and consequent perfusion in the NMR spectrometer. Of note, other surgical approaches aimed at isolating and perfusing the mouse heart have been described before24,25.
The methodologies used for acquiring data related to enzymatic rates in the beating heart (using 13C spectroscopy and hyperpolarized [1-13C]pyruvate) and the heart's viability and acidity (using 31P NMR spectroscopy) are described as well. Finally, the analytical methodologies for determining metabolic enzyme activities and tissue viability and acidity are explained.
The joint ethics committee (IACUC) of the Hebrew University and Hadassah Medical Center approved the study protocol for animal welfare (MD-19-15827-1).
1. Krebs-Henseleit buffer preparation
2. Perfusion system preparation
3. Calibration and preparation of the NMR spectrometer for acquisition
4. Animal preparation, surgical procedure, and perfusion of the heart in the NMR tube
5. Acquiring data for cardiac energetics and pH
6. DNP spin polarization and dissolution
7. Hyperpolarized 13C spectroscopy
8. Determination of the tissue wet weight and volume
9. ATP content quantification
10. Resolving the Pi signal of the heart
NOTE: In order to evaluate the tissue pH, it is first necessary to deconvolve the heart's Pi signal from that of the total Pi signal (Pit). This is done by omitting the signal of the KHB Pi (PiKH) from that of the Pit.
11. Multi-parametric pH analysis
12. Calculation of the LDH and PDH activities
NOTE: The production rates of the hyperpolarized metabolites [1-13C]lactate and [13C]bicarbonate are used to calculate the LDH and PDH activities, respectively. In the product selective saturating-excitation approach15, only newly synthesized hyperpolarized metabolites are detected by each selective excitation.
The 31P spectra recorded from a mouse heart perfused with KHB and from the buffer alone are shown in Figure 1A. The signals of α-, β-, and γ-ATP, PCr, and Pi were observed in the heart. The Pi signal was composed of two main components: in the higher field (left side of the signal), the Pi signal was mostly due to the KHB at a pH of 7.4; in the lower field (right side of the signal), the Pi signal was broader and less homogeneous due to the more acidic environment. The latter pattern arises from the cardiac tissue. The cardiac tissue Pi signal is extracted by subtracting the Pi signal of the KHB (Figure 1B) and then converted from the ppm scale to the pH scale (Figure 1C). The pH is investigated using a multi-parametric analysis of the tissue Pi signal by calculating the weighted mean, weighted median, global maximum, and skewness (Figure 1C).
Figure 1: Typical 31P spectra, the distinction between the Pi signal of the KHB and the heart, conversion from the chemical shift to pH axes, and the statistical parameters of the pH distribution. (A) The upper panel displays a typical 31P NMR spectrum obtained from a mouse heart perfused with KHB in the spectrometer, while the lower panel shows a spectrum obtained from KHB alone. The dash-marked spectral region is shown enlarged in panel B. Abbreviations: Pi = inorganic phosphate; PCr = phosphocreatine; ATP = adenosine triphosphate; ADP = adenosine diphosphate. (B) The original spectrum is shown in black (Pit); the dashed curve shows the Pi signal from the buffer only (Pib). The latter was obtained by fitting a Lorentzian line shape with its center at the corresponding KHB component of the Pit signal and adjusting for the amount of KHB in the probe following the heart insertion (according to Eq. 1A). The Pi signal attributed to the heart (Pih) is shown in orange, and this is obtained by subtracting the Pib signal from the Pit signal (according to Eq. 1B). (C) Conversion of the Pih signal chemical shift distribution shown in (B) to a pH distribution and multi-parametric pH analysis. The non-linearity between the chemical shift and the pH scales was corrected as previously described32. The results of the multi-parametric pH analysis are marked by the vertical lines. For this specific distribution, the values of the statistical parameters are provided. Please click here to view a larger version of this figure.
With the hyperpolarized product selective saturating-excitations acquisition approach15, it is possible to perform absolute quantification of the LDH and PDH enzyme activities. Figure 2 summarizes the acquisition and processing steps required for this determination. Table 1 shows the ATP content and the LDH and PDH rates in five different hearts. Normalizing the LDH and PDH activities to the ATP content of each heart reduced the variability in the LDH and PDH measurements across the group, as can be seen in the enzyme activity columns in Table 1, which are expressed in units of nmol/s/µmol ATP.
Figure 2: Hyperpolarized 13C MRS processing and analysis of [1-13C]pyruvate metabolism. (A) Representative 13C NMR spectra acquired with the product selective saturating-excitations approach during an injection of 14 mM of hyperpolarized [1-13C]pyruvate into the perfused mouse heart. Signal assignment: 1 = [1-13C]lactate (183.2 ppm); 2 = [1-13C]pyruvate (171 ppm); 3 = [13C]bicarbonate (161.1 ppm); * = impurities arising from the [1-13C]pyruvate formulation33. (B) Time course of the hyperpolarized signal intensities displayed in A. The integrated intensities of [1-13C]pyruvate (grey) were adjusted for T1decay and RF pulsation with a Teff time constant of 32 s (black), and this yielded the expected flow dynamics for the substrate (wash-in, plateau, wash-out). Further analysis was performed on the data points within the time window in which the adjusted [1-13C]pyruvate signal showed a constant and maximal [1-13C]pyruvate concentration in the NMR tube (highlighted in light blue). The production rates of LDH and PDH ( and
) for the selected time points were calculated according to Eq. 4A and Eq. 4B and then averaged. The mean values for this injection (
and
) are provided in units of nmol/s. For display purposes, the adjusted [1-13C]pyruvate, the [1-13C]lactate, and the [13C]bicarbonate were multiplied by 0.143, 10, and 80, respectively, relative to the [1-13C]pyruvate signal. Please click here to view a larger version of this figure.
Heart No. | ATP (µmol) | LDH rate (nmol/s) | PDH rate (nmol/s) | LDH rate (nmol/s/µmol ATP) | PDH rate (nmol/s/µmol ATP) |
1 | 0.49 | 7.61 | 0.59 | 15.4 | 1.19 |
2 | 0.25 | 3.66 | 0.32 | 14.42 | 1.26 |
3 | 0.51 | 6.01 | 0.66 | 11.81 | 1.3 |
4 | 0.53 | 9.27 | 1.09 | 17.34 | 2.04 |
5 | 0.64 | 9.38 | 0.6 | 14.77 | 0.94 |
Mean (SD) | 0.49 (0.13) | 7.19 (2.15) | 0.65 (0.25) | 14.75 (1.78) | 1.35 (0.37) |
SD % Mean | 26.00% | 30.00% | 38.30% | 12.10% | 27.50% |
Table 1: ATP content and LDH and PDH rates in five hearts. Abbreviations: SD = Standard deviation; SD % Mean = the percentage of the standard deviation from the mean.
We demonstrate an experimental setup that is designed to investigate hyperpolarized [1-13C]pyruvate metabolism, tissue energetics, and pH in an isolated mouse heart model.
The critical steps within the protocol are as follows: 1) ensuring that the pH of the buffer is 7.4; 2) ensuring that all components of the buffer are included; 3) avoiding blood clotting in the cardiac vessels by heparin injections; 4) avoiding ischemic damage to the heart by reducing the metabolic activity (KCl injection and ice-cold buffer); 5) avoiding the introduction of air bubbles into the heart at any point of the procedure; 6) validating successful cannulation of the aorta by the color of the tissue, which changes from dark red to light pink; 7) making sure that the perfusion is continuous once the heart is moved to warm buffer; 8) closely monitoring the temperature inside the NMR tube in the spectrometer next to the heart and keeping it at 37 °C; 9) checking the magnetic field homogeneity throughout the measurement; and 10) using accurate and updated calibration for the RF pulses.
Modifications and troubleshooting of the technique
We note a couple of modifications that were made in this study: 1) the initiation of perfusion with ice-cold KHB further to surgery was employed to protect the heart; and 2) the validation of the buffer pH was performed throughout to make sure that this parameter was under control and was not a source of variation in the results. This was done in the various steps of preparation and experimentation using a pH meter, pH indicator strips, and the Pi signal on the 31P spectrum.
Sources of variability and how to correct for them
The mouse heart was chosen to fit into a 10 mm NMR tube. However, this small heart provides low signals, and the delicate surgical procedure for isolating and perfusing the heart is challenging. Overall, this leads to variable tissue viability and metabolic activity. To account for this variability, we normalized the metabolic rates of LDH and PDH to the ATP content (i.e., mole/s/ATP). A similar use of the ATP content as a reference was previously reported in xenograft breast tumor slices29. This type of normalization is beneficial because it allows for comparison with the results of other researchers and with other conditions. In addition, by using a conversion factor from the ATP amount to the tissue mass, one may derive the enzymatic activity per tissue weight (for the tissue in which metabolism occurs).
In addition to the variability between different isolated heart preparations (i.e., from different animals), the tissue Pi (Pih) chemical shift showed a non-homogeneous distribution in this study. Lutz et al.32 demonstrated a similar non-homogeneous distribution in xenograft tumors, and this distribution was analyzed using a multi-parametric approach. In this work, this methodology was implemented in the perfused mouse heart. We note that it is likely that the Pih signal reports predominantly on the intracellular pH, as indicated previously16.
Uncertainty
An underlying assumption in the quantification of and
is that the hyperpolarized [1-13C]pyruvate solution is filling the entire volume detected by the NMR probe (Vp, Eq. 4A and Eq. 4B). The hyperpolarized [1-13C]pyruvate solution flows through the coronary arteries of the heart to fill the extracellular and intracellular compartments. Therefore, the heart volume is assumed to be filled with the hyperpolarized solution to the same extent as the rest of the Vp.
Using the hyperpolarized product selective saturating-excitations approach15,29, the enzymatic activity can be quantified independently of any variations in the T1 of the metabolites and the reversibility of the enzymatic reaction. This determination is robust and immune to variations in T1 and excitation profiles and is likely more reproducible across labs compared to area under the curve analyses, which are dependent on the specific acquisition conditions and setup in each lab.
Investigating [1-13C]pyruvate metabolism, which is at the crossroad of aerobic and anaerobic metabolism, is of great value for studying hypoxia, ischemia, reperfusion injury, starvation, and alteration in cardiac metabolism, such as in diabetic cardiomyopathy. Hyperpolarized 13C-labeled pyruvate analogs have been tested clinically34,35,36,37,38,39,40,41,42,43, and, therefore, the results of such studies are likely translational. Most importantly, our approach enables the quantification of in-cell enzyme activity in the whole organ.
The isolated rodent heart preparation and the procedures and methodologies involved in this research will help to further the understanding of the effect of a variety of stressors on cardiac function, energetics, and metabolism, as the measured parameters in this model are attributed solely to the heart. As opposed to the rat heart, the mouse heart is suitable for studying transgenic models.
The authors have nothing to disclose.
This project received funding from the Israel Science Foundation under grant agreement No. 1379/18; the Jabotinsky Scholarship of the Israeli Ministry of Science and Technology for Applied and Engineering Sciences for Direct PhD Students No. 3-15892 for D.S.; and the European Union's Horizon 2020 research and innovation program under grant agreement No. 858149 (AlternativesToGd).
Equipment | |||
HyperSense DNP Polariser | Oxford Instruments | 52-ZNP91000 | HyperSense, 3.35 T, preclinical dissolution-DNP hyperpolarizer |
NMR spectrometer | RS2D | NMR Cube, 5.8 T, equiped with a 10 mm broad-band probe | |
Peristaltic pump | Cole-Parmer | 07554-95 | |
Temperature probe | Osensa | FTX-100-LUX+ | NMR compatible temprature probe |
Somnosuite low-flow anesthesia system | Kent Scientific | ||
Lines, tubings, suture | |||
Platinum cured silicone tubes | Cole-Parmer | HV-96119-16 | L/S 16 I.D. 3.1 mm |
Thin polyether ether ketone (PEEK) lines | Upchurch Scientific | id. 0.040” | |
Intravenous catheter | BD Medical | 381323 | 22 G |
Silk suture | Ethicon | W577H | Wire diameter of 3-0 |
Chemicals and pharmaceuticals | |||
[1-13C]pyruvic acid | Cambridge Isotope Laboratories | CLM-8077-1 | |
Calcium chloride | Sigma-Aldrich | 21074 | CAS: 10043-52-4 |
D-(+)-Glucose | Sigma-Aldrich | G7528 | CAS: 50-99-77 |
Heparin sodium | Rotexmedica | HEP5A0130C0160 | |
Hydrochloric acid 37% | Sigma-Aldrich | 258148 | CAS: 7647-01-0 |
Insulin aspart (NovoLog) | Novo Nordisk | ||
Isoflurane | Terrel | ||
Magnesium Sulfate | Sigma-Aldrich | 793612 | CAS: 7487-88-9 |
Potassium chloride | Sigma-Aldrich | P4504 | CAS: 7447-40-7 |
Potassium phosphate monobasic | Sigma-Aldrich | P9791 | CAS: 7778-77-0 |
Sodium bicarbonate | Gadot Group | CAS: 144-55-8 | |
Sodium chloride | Sigma-Aldrich | S9625 | CAS: 7647-14-5 |
Sodium hydroxide | Sigma-Aldrich | 655104 | CAS: 1310-73-2 |
Sodium phosphate dibasic | Sigma Aldrich | S7907 | CAS: 7558-79-4 |
Sodium phosphate monbasic dihydrate | Merck | 6345 | CAS: 13472-35-0 |
TRIS (biotechnology grade) | Amresco | 0826 | CAS: 77-86-1 |
Trityl radical OX063 | GE Healthcare AS | NC100136 | OX063 |
NMR standards | |||
13C standard sample | Cambridge Isotope Laboratories | DLM-72A | 40% p-dioxane in benzene-D6 |
31P standard sample | Made in house | 105 mM ATP and 120 mM phenylphosphonic acid in D2O | |
Software | |||
Excel 2016 | Microsoft | ||
MNova | Mestrelab Research |