4286
Polarisation, T1 and Synthetic Impurities in Hyperpolarised Acetoacetate: Density Functional Theory to In Vivo Application1Wolfson Brain Imaging Centre, University of Cambridge, Cambridge, United Kingdom, 2Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, United Kingdom, 3Oxford Centre for Clinical Cardiac Magnetic Resonance Research, University of Oxford, Oxford, United Kingdom, 4Department of Chemistry, University of Oxford, Oxford, United Kingdom, 5University Hospitals Bristol NHFS Foundation Trust, Bristol, United Kingdom, 6Department of Physics, University of Oxford, Oxford, United Kingdom
Synopsis
Alterations in ketone body metabolism are implicated in disease. Several studies have observed metabolism in hyperpolarised sodium acetoacetate, but as a hyperpolarised probe it has a short T1 (28s at 7T), limiting polarisation of 7-8%, and is often chemically impure as it spontaneously decarboxylates at neutral pH at 300K. We have studied selective deuteration of [1-13C] and [1,3-13C]acetoacetate together and evaluated the effect of salt solvation on hyperpolarisation in the discretised Borghini model of thermal mixing. Li+[2,2-2H2,1,3-13C2]acetoacetate is observed to have a higher limiting polarisation and substantially longer T1 than Na+[1-13C]acetocetate.
Purpose
Recent work has shown that modulation of ketone body oxidation rates may be implicated in the cardiovascular protective effect of empagliflozin, a novel antidiabetic SGLT2 inhibitor that increases the circulating concentration of ketone bodies. Hyperpolarisation via Dynamic Nuclear Polarisation side-steps the fundamental thermodynamic limitations of NMR, and hyperpolarised acetoacetate has been shown to be a safe and effective metabolic probe in the rat heart,[1] demonstrating altered ketone flux in diabetes.
Previous work on hyperpolarised acetoacetate has predominantly studied the sodium salt of either [1-13C]–, [1,3-13C], or [3-13C]acetocetate,[1–5] which, owing to its chemical instability at 300K at neutral pH, is synthesised in-house via the addition of NaOH to an ethyl-acetoacetate precursor followed by freeze-drying, a process that results in visible hyperpolarised impurities.[2,4,5] However, earlier work has proposed LiOH catalysed hydrolysis with improved purity and yield compared to NaOH.[6] As lithium is toxic with a low therapeutic window, we wished to repeat this synthesis, chelate Li+, and additionally deuterate acetoacetate to improve its T1.
Methods
Briefly, ethyl-[1,3-13C2]acetoacetate was hydrolysed via LiOD/D2O at 40$$$^\circ\text{C}$$$ followed by rotary evaporation, lyophilisation and purification via methanol/ether recrystallisation, with deuteration provided by ketone/enol tautomerization with a yield of $$$>65\%$$$. For DNP, 30 mg aliquots were mixed with 4.8 μL of a 20 mM EPA/10 mM Gd3+/dotarem mixture, neutralised in D2SO4, and frozen as 15 μL spheres in liquid N2 prior to hyperpolarisation at 3.35 T. Dissolution was performed with PBS buffered D2O with an equimolar quantity of 12-crown-4 ether as a lithium chelator (product pH$$$\approx7$$$) in a prototype hyperpolariser prior to 1 mL injection into either a phantom or fasted anaesthetised (Isoflurane, $$$2\%\,\text{in}\,\text{O}_2$$$) Wistar rats. Cardiac slab-selective spectra were obtained via a transmit/receive array following injection (FA=10$$$^\circ$$$, 10$$$\,$$$kHz bandwidth, 10$$$\,$$$mm thick, 1s TR). Purity and yield were assessed via thermal equilibrium NMR and IR spectroscopy under basic conditions.
Plasma lithium and ketone concentrations were measured post injection via clinically validated analysers (SmartLyte/Accuchek). Multicoil data were summed in phase and quantified via a custom Matlab implementation of AMARES.
To quantitatively understand differences in polarisation seen, Density Functional Theory analysis was performed in Spartan to predict the low temperature unit cell of labelled Na+ or Li+ acetoacetate. Additionally, the discretised Borghini model of thermal mixing in the regime of weak electron nucleon contact as proposed by Serra et al.[7] was solved under conditions corresponding to the elongated solid state T1, assuming a OX063 radical at 3.35 T.
Results and Discussion
We found that Li+[2,2-2H2,1,3-13C2]acetoacetate is feasible for preclinical imaging and obtains approximately 20% solid state polarisation. In contrast to Na+[1-13C]acetoacetate, Li+[2,2-2H2,1,3-13C2 ]acetoacetate had no visible hyperpolarised impurity peaks (Fig. 1), a purity of $$$>93\%$$$, and a more than doubled liquid state T1 compared to Na+[1-13C]acetoacetate ($$$76\pm3\,\text{s}$$$ vs $$$28\pm3\,\text{s}$$$ Fig. 2) . The injection was well tolerated, and after chelation, lithium was not detectable in plasma at the limit of sensitivity of the analyser used (0.2 mM). Blood ketone concentration was increased after injection but was not supraphysiological (max 1.9 mM).
To quantitatively explain the increased solid state polarisation obtained with deuteration we numerically solved a model for thermal mixing.[7] This did not indicate that the increase in polarisation arising from an increase in nuclear T1 predicted at low temperature was significant, which is consistent with reports that deuteration decreases the limiting polarisation reached as it effectively demands a greater cooling capability of the electron Zeeman system.[8]
To investigate this phenomenon, we predicted the low temperature structure of acetoacetate with both sodium and lithium salts used via the quantum chemistry package Spartan. We note that the decreased ionic radius of lithium results in a profoundly different conformation with respect to the two carbons of interest, forming an $$$I=3/2$$$ moiety spatially closer to the $$$^{13}$$$C of interest, and at 3.35T, the large quadrupole moment of naturally abundant $$$^7$$$Li, $$$^{35}$$$Cl and $$$^{37}$$$Cl permit multiple quantum transitions that potentially facilitate spin diffusion during DNP.[9]
Conclusion
We have shown the feasibility of a simple synthesis method for the in-house production of [2,2-$$${}^{2}\mathrm{H}_2$$$,1,3-$$${}^{13}\mathrm{C}_2$$$]acetoacetate with minimal impurities. The deuteration provides a significant increase in liquid state T1 and a corresponding improvement of downstream metabolites in vivo. Unexpectedly, we also observed a significantly higher solid state polarisation despite the increased cooling requirement of the electron system due to the presence of the low-gamma 2H, which we hypothesise can be explained by salt arguments. This probe is of interest for future work investigating the mechanistic causes and potential amelioration of diabetic cardiomyopathy, and future work will investigate both its physical chemistry and biomedical utility.Acknowledgements
BYD acknowledges receipt of a Gates Cambridge Trust scholarship, and a British Heart Foundation Vacation Studentship. JJM would like to acknowledge a Novo Nordisk Postdoctoral Fellowship. All authors would like to acknowledge the support of the British Heart Foundation, and the University of Oxford BHF Centre for Research Excellence. CTR is funded by a Sir Henry Dale Fellowship from the Welcome Trust and the Royal Society [098436/Z/12/B].References
[1] J. J. Miller, D. R. Ball, A. Z. Lau, D. J. Tyler,NMR Biomed.2018,31, DOI10.1002/nbm.3912.
[2] C. von Morze, M. A. M. Ohliger, I. Marco-Rius, D. M. D. Wilson, R. R. Flavell, D. Pearce, D. D. B.Vigneron, J. Kurhanewicz, Z. J. Z. Wang,Magn. Reson. Med.2018,79, 1862–1869.
[3] D. Abdurrachim, X. Q. Teo, C. C. Woo, W. X. Chan, J. Lalic, C. S. Lam, P. T. H. Lee,Diabetes Obes.Metab.2018, DOI10.1111/dom.13536
[4] W. Chen, C. Khemtong, W. Jiang, C. R. Malloy, D. Sherry in Proc. Intl. Soc. Mag. Reson. Med. 24,2016, p. 0672.
[5] J. X. Wang, M. E. Merritt, D. Sherry, C. R. Malloy,Magn. Reson. Chem.2016,54, 665–673.
[6] B. W. C. Kennedy, M. I. Kettunen, D.-E. Hu, S. E. Bohndiek, K. M. Brindle in Proc. Intl. Soc. Mag. Reson.Med. 20,2012, p. 4326.
[7] S. C. Serra, A. Rosso, F. Tedoldi,Phys. Chem. Chem. Phys.2013,15, 8416–28.
[8] P. Niedbalski, C. Parish, A. Kiswandhi, Z. Kovacs, L. Lumata,J. Phys. Chem. A2017,121, 3227–3233.
[9] J. C. Paniagua, V. Mugnaini, C. Gabellieri, M. Feliz, N. Roques, J. Veciana, M. Pons,Phys. Chem.Chem. Phys.2010, DOI10.1039/c003291n.
Figures
