1062

Triple Multinuclear Probe: Expanding the Throughput and Versatility of a dDNP Polarizer

Thanh Phong Lê^1,2, Emma Wiström², Jean-Noël Hyacinthe¹, and Andrea Capozzi^2,3
¹Geneva School of Health Sciences, HES-SO University of Applied Sciences and Arts Western Switzerland, Geneva, Switzerland, ²Laboratory of Functional and Metabolic Imaging, EPFL (Swiss Federal Institute of Technology in Lausanne), Lausanne, Switzerland, ³Department of Health Technology, Center for Hyperpolarization in Magnetic Resonance, Technical University of Denmark, Kgs Lyngby, Denmark

Synopsis

Low throughput is one of dDNP main shortcomings. Here, we present the design and performance of a triple DNP probe, which not only allows multiple samples to be hyperpolarized simultaneously, but also to distinctly monitor the solid-state polarization of each sample, even when prepared with different nuclei and radicals. This versatile high throughput probe allows parallel solid-state sample characterization, as well as sequential injections of distinct tracers to probe multiple biomarkers, thus enabling more advanced in-vivo experimental designs.

Introduction

Dissolution dynamic nuclear polarization¹ (dDNP) gifts MR with unprecedented sensitivity for spectroscopy or imaging. A main shortcoming of DNP is the low sample throughput due to long polarization processes, which can be overcome by polarizing multiple samples simultaneously^2–4, or with faster polarization schemes⁵.

On the application side, there is, for instance, an interest in probing multiple enzymatic activities⁶, or in investigating in-vivo both the metabolism and perfusion of an organ^7,8. This is often performed by co-polarizing and injecting multiple tracers. However, the mixture needs to be carefully prepared for not competing along the same pathways, allowing concurrent detection and achieving efficient simultaneous polarization.

In this work, we designed a multisample DNP probe with multinuclei capabilities to not only increase the throughput of the polarizer, but also to allow the monitoring of distinct samples. This provides better versatility for both solid-state and dissolution DNP, allowing for more flexible experimental designs.

Methods

Triple DNP Probe
A DNP probe with three separate sample slots was designed for a wet 5T/1.2K DNP polarizer⁹ (Fig.1). Three custom fluid paths (CFP)¹⁰ allow loading and dissolving samples without opening the cryostat to atmospheric pressure at each time.

The three sample slots are equipped with three distinct Aldermann-Grant NMR coils, remotely tuned and matched, and interfaced to a 3Tx/3Rx console (Gecho, RS2D). A single microwave source hyperpolarizes all samples.

DNP of PA
5µl of [1-¹³C]pyruvic acid (PA) + 15mM OX063 was loaded into each slot. Optimal microwave parameters were determined by frequency (139.8-140.0GHz, 55mW) and power sweeps (1-63mW, 139.93GHz).

PA was hyperpolarized at 139.93GHz, 63mW for 2.5-3h, then dissolved in a 180°C buffer and pushed to a separator/infusion pump⁹. (753±42.6)µl of solution were injected into a 1.05T spectrometer (SpinSolve ¹³C/¹²⁹Xe, Magritek) then acquired with 1° pulses, TR=3s, 60 repetitions.

All samples were dissolved back-to-back with 15min intervals. After adding 1.5mM of Gd-DO3A-butrol, the thermal signal was measured with FA=90°, TR=3s, 1024 averages. Six measurements were performed, two per slot.

In-vivo HP MRS
[1-¹³C]PA and [1-¹³C]lactate (4.4M with 20mM OX063 in H₂O:glycerol 1:1) were simultaneously hyperpolarized at 139.93GHz.

A healthy C57BL6/J male mouse (25.1g) was scanned at 14.1T with a ¹H quadrature/¹³C linear (⌀14/11mm) head surface coil.

Lactate was dissolved first and intravenously injected using a separator/infusion pump (300µl of 80mM lactate). Cerebral ¹³C MRS was acquired with 30°BIR-4 pulses and TR=3s.

15 minutes after, PA was similarly dissolved and injected.

Simultaneous DNP of distinct nuclei/radicals
Three distinct samples were loaded:

4.4M sodium-[1-¹³C]lactate, 20mM OX063 in H₂O:glycerol (1:1) (hereafter Lac+OX063).
27.8M ¹H₂O in glycerol-d₈:PA: (3:2 v/v), ~50mM UV-generated radicals¹¹ (H₂O+UV)
2.85M ¹²⁹Xe, 30mM TEMPO in isobutanol¹² (Xe+TEMPO)

Microwave frequency sweeps (139.80-140.15GHz) were performed with 0 to 60 MHz of frequency modulation (FM). Optimal microwave settings for simultaneous buildup of all samples were determined and applied.

Results

Thermal performance
The hold time of the triple probe was about 5h at 1.30-1.35K during microwave sweep experiments (Fig.2A). Dissolutions momentarily increased both temperature and pressure and recovered after 5min (Fig.2B) without affecting the polarization of the remaining samples (Fig.2C).

DNP of PA
The three frequency sweeps show a slight shift between slot 1 and the remaining, probably due to B₀ inhomogeneity (Fig.3A). Microwave power, sufficient to saturate the radical ESR line, reaches each sample slot (Fig.3B), but the power density in slot 1 is higher, yielding to faster polarization (Fig.3C).

The liquid-state polarization 9s post-dissolution at 1.05T was (30.0±1.2)%, without differences between slots (Fig.3D). The long liquid-state T₁ results from the benchtop NMR field strength¹³, and high temperature of the pyruvate solution (~60°C).

In-vivo HP MRS
Cerebral metabolism was detected following sequential lactate and pyruvate injections (Fig.4), with a maximal SNR in single spectra of 402 and 313 respectively (lb=0). Lower SNRs were observed on lactate’s downstream metabolites due to smaller conversion compared to pyruvate¹⁴.

Simultaneous DNP of distinct nuclei/radicals
Fig.5 reports frequency sweeps on mixtures with distinct nuclei and radicals. Higher FM brought closer the positive peak of Xe+TEMPO to the negative peak of Lac+OX063, allowing simultaneous polarization (Fig.5B).

Discussion

In this work, we designed a triple DNP probe to increase the throughput of a wet dDNP polarizer. Using the CFPs, the loading/dissolution procedure does not affect the remaining samples. Despite the higher temperature compared to the single-sample probe, similar polarization levels were achieved¹⁵.

Back-to-back dDNP in-vivo experiments reduced exposure to anesthesia together with physiological parameters drifts, and successfully detected cerebral lactate and pyruvate metabolism. These substrates are potential theranostic agents in the context of stroke, providing both metabolic contrasts^14,16,17 and neuroprotection^18,19.

The triple probe allows simultaneous DNP on dissimilar nuclei and radicals, while monitoring their buildup independently. Here, we simultaneously polarized lactate, water and xenon, respectively a metabolic probe, contrast agent for angiography^20,21, and tissue-specific perfusion tracer^22,23.

The combination can be tailored to specific applications. For example, in the context of cancer, a selection providing relevant biomarkers could include [U-¹³C₆]-D-glucose to detect upregulated glycolysis^24,25, [1,4-¹³C₂]fumarate as a marker of cell necrosis²⁶, and [¹³C]bicarbonate as an extracellular pH sensor²⁷.

Conclusion

A triple DNP probe was built to upgrade a wet DNP polarizer. Distinct nuclei and radicals can be hyperpolarized, monitored simultaneously, then dissolved sequentially. The substantially higher throughput and versatility allows for new experimental designs.

Acknowledgements

This study is supported by the Swiss National Science Foundation (170155 to JN. Hyacinthe, 190547 and 193276 to A. Capozzi). The authors gratefully thank Dr. Mario Lepore, Dr. Analina Hausin and Dr. med. vet. Stefanita Mitrea for their assistance in the animal preparation, Prof. Lorenz Hirt, Dr. Mor Mishkovsky and Dr. Lara Buscemi for getting and handling the animal authorization, as well as the Center for Biomedical Imaging of the UNIL, EPFL, UNIGE, HUG and CHUV.

References

1. Ardenkjaer-Larsen, J. H. et al. Increase in signal-to-noise ratio of 10,000 times in liquid-state NMR. Proceedings of the National Academy of Sciences of the United States of America 100, 10158–63 (2003).

2. Cheng, T. et al. A multisample 7 T dynamic nuclear polarization polarizer for preclinical hyperpolarized MR. NMR in Biomedicine 33, 1–10 (2020).

3. Krajewski, M. et al. A multisample dissolution dynamic nuclear polarization system for serial injections in small animals. Magnetic Resonance in Medicine 77, 904–910 (2017).

4. Hu, S. et al. Rapid sequential injections of hyperpolarized [1-13C]pyruvate in vivo using a sub-kelvin, multi-sample DNP polarizer. Magnetic Resonance Imaging 31, 490–496 (2013).

5. Bornet, A. et al. Boosting dissolution dynamic nuclear polarization by cross polarization. Journal of Physical Chemistry Letters 4, 111–114 (2013).

6. Wilson, D. M. et al. Multi-compound polarization by DNP allows simultaneous assessment of multiple enzymatic activities in vivo. Journal of Magnetic Resonance 205, 141–147 (2010).

7. von Morze, C. et al. Investigating tumor perfusion and metabolism using multiple hyperpolarized ¹³C compounds: HP001, pyruvate and urea. Magnetic Resonance Imaging 30, 305–311 (2012).

8. Qin, H. et al. Clinical translation of hyperpolarized 13C pyruvate and urea MRI for simultaneous metabolic and perfusion imaging. Magnetic Resonance in Medicine (2021).

9. Comment, A. et al. Design and performance of a DNP prepolarizer coupled to a rodent MRI scanner. Concepts in Magnetic Resonance Part B 31B, 255–269 (2007).

10. Capozzi, A. et al. Metabolic contrast agents produced from transported solid ¹³C-glucose hyperpolarized via dynamic nuclear polarization. Communications Chemistry 4, 1–11 (2021).

11. Pinon, A. C., Capozzi, A. & Ardenkjær-Larsen, J. H. Hyperpolarized water through dissolution dynamic nuclear polarization with UV-generated radicals. Communications Chemistry 3, 1–9 (2020).

12. Capozzi, A., Roussel, C., Comment, A. & Hyacinthe, J. N. Optimal glass-forming solvent brings sublimation dynamic nuclear polarization to ¹²⁹Xe hyperpolarization biomedical imaging standards. Journal of Physical Chemistry C 119, 5020–5025 (2015).

13. Tee, S. S. et al. Sampling Hyperpolarized Molecules Utilizing a 1 Tesla Permanent Magnetic Field. Scientific Reports 6, (2016).

14. Lê, T. P. et al. Metabolism of the hyperpolarized neuroprotective agents [1-¹³C] lactate and [1-¹³C] pyruvate in a mouse model of transient ischemic stroke. in Proc. Intl. Soc. Mag. Reson. Med. 28 0689 (2020).

15. Lê Thanh Phong, Capozzi Andrea & Hyacinthe Jean-Noël. Hyperpolarized in-vivo Metabolic Imaging at 14.1T: dDNP Cryogenic Insert Redesign and Implementation. Proc. Intl. Soc. Mag. Reson. Med. 29, 3804 (2021).

16. Xu, Y. et al. Hyperpolarized ¹³C Magnetic Resonance Imaging Can Detect Metabolic Changes Characteristic of Penumbra in Ischemic Stroke . Tomography 3, 67–73 (2017).

17. Hyacinthe, J.-N. et al. Evaluating the potential of hyperpolarised [1-¹³C] L-lactate as a neuroprotectant metabolic biosensor for stroke. Scientific Reports 10, 5507 (2020).

18. Berthet, C. et al. Neuroprotective role of lactate after cerebral ischemia. Journal of Cerebral Blood Flow and Metabolism 29, 1780–1789 (2009).

19. Yi, J. S., Kim, T. Y., Kyu Kim, D. & Koh, J. Y. Systemic pyruvate administration markedly reduces infarcts and motor deficits in rat models of transient and permanent focal cerebral ischemia. Neurobiology of Disease 26, 94–104 (2007).

20. Ardenkjaer-Larsen, J. H., Laustsen, C., Bowen, S. & Rizi, R. Hyperpolarized H₂O MR angiography. Magnetic Resonance in Medicine 71, 50–56 (2014).

21. Lipsø, K. W., Hansen, E. S. S., Tougaard, R. S., Laustsen, C. & Ardenkjær-Larsen, J. H. Dynamic coronary MR angiography in a pig model with hyperpolarized water. Magnetic Resonance in Medicine 80, 1165–1169 (2018).

22. Duhamel, G. et al. Xenon-¹²⁹ MR Imaging and Spectroscopy of Rat Brain Using Arterial Delivery of Hyperpolarized Xenon in a Lipid Emulsion. (2001).

23. Kilian, W., Seifert, F. & Rinneberg, H. Dynamic NMR Spectroscopy of Hyperpolarized ¹²⁹Xe in Human Brain Analyzed by an Uptake Model. Magnetic Resonance in Medicine 51, 843–847 (2004).

24. Rodrigues, T. B. et al. Magnetic resonance imaging of tumor glycolysis using hyperpolarized ¹³ C-labeled glucose. Nature Medicine 20, 93–97 (2014).

25. Mishkovsky, M. et al. Hyperpolarized ¹³C-glucose magnetic resonance highlights reduced aerobic glycolysis in vivo in infiltrative glioblastoma. Scientific Reports 11, (2021).

26. Gallagher, F. A. et al. Production of hyperpolarized [1,4-¹³C₂]malate from [1,4-¹³C₂]fumarate is a marker of cell necrosis and treatment response in tumors. PNAS 106, 19801–19806 (2009)

27. Gallagher, F. A. et al. Magnetic resonance imaging of pH in vivo using hyperpolarized ¹³C-labelled bicarbonate. Nature 453, 940–943 (2008).

Figures

Fig.1: DNP probe designed to accommodate three custom fluid paths (CFP). Loading chamber (A) and gate valve (B). 12.6/13.0 mm stainless steel tube (C) guiding the CFP to the microwave cavity. Microwave inlet (D). Separate RF inlets (E) and stainless-steel coaxial cables (F) for each slot. Circular 3.6/4.0 mm stainless-steel waveguide (G). Gold-plated tetrahedral microwave reflector (H). Aldermann-Grant NMR coils, one for each sample (I). PTFE coil formers (J). Gold-plated microwave cavity (K).

Fig.2: (A) Temperature (blue line) and pressure (red line) in the sample space during a microwave frequency sweep experiment (A). A temperature of 1.30-1.35K is maintained along the duration of the experiment. The RF pulse trains cause pressure spikes. The cryostat runs out of helium after 5h30. (B) Typical dDNP experiment. Each dissolution causes a sharp temperature and pressure increase. NMR signal detected in the same dDNP experiment. The dissolution of a sample does not affect the polarization of remaining ones (C).

Fig.3: DNP of pyruvic acid doped with trityl radial. (A) Microwave frequency sweeps performed on one sample per slot. The vertical scale was normalized to the maximal value of each curve. (B) Microwave power sweeps indicate a saturation of the NMR signal; therefore, sufficient power reaches each slot. (C) Buildup of the solid-state polarization. The buildup time was shorter in the slot 1. (D) Decay of the liquid-state hyperpolarized signal at 1.05T. The liquid-state polarization was similar across all slots.

Fig.4: (A) Simultaneous hyperpolarization of lactate and pyruvate for in-vivo experiments. Both substrates were injected into the same mouse with a 15min interval. (B) Global cerebral metabolism of HP lactate, which metabolizes into pyruvate, alanine and bicarbonate. The bold line is the sum of the first 120s post-injection. lb=20Hz for display and * denotes an impurity of the stock lactate solution. (C) HP pyruvate metabolizes into lactate, alanine and bicarbonate.

Fig.5: (A) Microwave frequency sweeps with increasing microwave frequency modulation (FM) bandwidths. All sweeps were normalized to the maximal signal without FM. At 139.94 GHz and 60 MHz FM (vertical dashed line), the signal intensity of the lactate, water and xenon samples were respectively 90%, 46%, and 92% of the maximal values observed in all sweeps. The negative DNP enhancement peak of Xe+TEMPO is at 140.25GHz, and therefore not displayed. (B) Simultaneous solid-state buildup at 139.94 GHz with 60 MHz FM.

Proc. Intl. Soc. Mag. Reson. Med. 30 (2022)

1062

DOI: https://doi.org/10.58530/2022/1062