Maternal-fetal exchanges characterized by dynamic hyperpolarized 13C imaging on pregnant rats
Anne Fages1, Tangi Roussel1, Marina Lysenko2, Ron Hadas2, Michal Neeman2, and Lucio Frydman1

1Chemical Physics, Weizmann Institute of Science, Rehovot, Israel, 2Biological Regulation, Weizmann Institute of Science, Rehovot, Israel

Synopsis

Dynamic nuclear polarization (DNP) enhanced 13C MRI of hyperpolarized (HP) urea and bicarbonate has been applied to monitor metabolic fluxes from the maternal blood pool to the fetuses, in pregnant rats at late gestation stage. This use of HP metabolites offers a non-invasive way to observe details of active and passive maternal-fetal exchanges.

Purpose

The mammalian fetus relies on the placenta to mediate exchanges between maternal and fetal gases and metabolites, and to excrete fetal metabolic wastes. An ability to non-invasively characterize specific maternal-fetal exchanges by MRI would be valuable to better understand fetal metabolism, developmental physiology and a variety of neo-natal diseases and malformations. In this context the combination of MRI with dissolution DNP [1-3] could allow us to monitor the transport of specific metabolites across different maternal-fetal compartments.

Methods

Hyperpolarized metabolites.13C-urea and 13C-bicarbonate were dissolved in glycerol:D2O in proportion 7:3 or in proportion 4:6 respectively with 15mM Ox63 and 0.1mM gadolinium were hyperpolarized in a Hypersense operating at 94GHz and 1.4K. A 3ml bolus of the resulting 115mM HP 13C-urea or 37mM HP 13C-bicarbonate was injected into the tail vein of the rat. 13C CSI were recorded at the end of the injection of HP urea or once 1/3 of the HP bicarbonate bolus was injected.

Pregnant rat. 7 Wistar pregnant rats at late pregnancy stage (embryonic days 17 to 21) were anesthetized with 3% of isoflurane in 1L/min of O2 and their tail vein was canulated for the HP injection.

MRI. Dynamic 13C chemical shift imaging (CSI) centric k-space ordering experiments (Figure 1) were performed on a Bruker Biospec 4.7T system using a cross-coil configuration (volume coil transmit / 20mm surface coil receive). A FOV of 5cm and a TR of 68ms was used with a matrix resolution of 10x10 or 8x8 leading to an acquisition time of 6.7s and 4.3s respectively. In addition, T1 and T2 weighted 1H anatomical images were obtained using gradient- (6.3ms TE, 615ms TR) and spin-echo (29ms TE, 5s TR) sequences with respiratory gating. 13C CSI images were reconstructed and quantified using a custom Matlab software. The images’ time resolution was increased by reconstructing interleaved images from k-space data acquired consecutively in time.

Results

1H anatomical imaging enabled the identification of all key maternal/fetal compartments including the maternal uterine artery, vena cava and the kidneys; the placenta; the fetus and its liver and heart. Series of 13C-urea images recorded with sufficient signal-to-noise ratio (SNR) for over 45s following injection (Figure 2) revealed a rapid build-up in the maternal compartments (uterine artery, kidney, vena cava) followed by a slower buildup and eventual decay in the placentas (Figure 2 & Figure 3a). In addition, weak 13C-urea signals were also observed in the fetal livers. Decays of the 13C-urea signal intensities over time followed different patterns, depending on their location: Figure 3a shows that while the signal from the kidney signal decreased exponentially, in the placenta, urea signal tended to accumulate as its highest intensity was observed in the second image, 7s after the end of the bolus injection. A different response was observed following the injection of HP bicarbonate, for which the most intense signals came from the liver of the fetus (Figure 3b). However the signal decreased rapidly due to the short bicarbonate T1.

Discussion

The notable differences in the signal intensity and their time-dependencies observed after injecting HP urea or HP bicarbonate, could be explained by distinct maternal-fetal exchanges mechanisms for these metabolites. Indeed, urea transport across the placenta is likely to occur by concentration gradient-dependent diffusion, whereas bicarbonate’s transport is highly regulated and requires active transport. This would explain the former’s slower maternal - placental - fetal progression, vis-à-vis the latter’s faster accumulation inside the fetus. To attest with better confidence if the injected urea could diffuse inside the fetal organs in this time scale, mass spectrometry measurements of metabolic extracts of the fetal organs are under progress and will be reported.

Conclusion

This study shows the ability of HP MRI to visualize non-invasively the maternal-fetal exchanges. Rodent studies of placental abnormalities are under way. DNP-enhanced 13C MRSI experiments are also in progress to characterize fetal metabolism by injecting hyperpolarized precursors of metabolic pathways.

Acknowledgements

This research was supported by DIP Project 710907, NIH grant R01HD086323, the Kimmel Institute for Magnetic Resonance and the generosity of the Perlman Family Foundation.

References

[1] Golman, K.; Zandt, R. I. T.; Thaning, M. Real-time metabolic imaging. pnas 2006, 1–6. [2] Keshari, K. R.; Wilson, D. M. Chemistry and biochemistry of 13C hyperpolarized magnetic resonance using dynamic nuclear polarization. Chem. Soc. Rev. 2014, 43, 1627. [3] Brindle, K. M. Imaging Metabolism with Hyperpolarized 13C-Labeled Cell Substrates. J. Am. Chem. Soc. 2015, 137, 6418–6427.

Figures

Figure1. Schematic representation of CSI pulse sequence (a) K-Space centric filling (b).

Figure 2. 13C images overlaid over corresponding 1H image recorded on a pregnant rat after the injection of HP 13C-urea. 13C CSI were recorded over 45 sec and normalized to the highest signal intensity throughout the images.

Figure 3. Coronal 13C CSI images acquired on a pregnant rat after the injection of HP 13C-urea (a) or HP 13C-bicarbonate (b). First 13C CSI image overlaid on the corresponding 1H anatomical image (left) and dynamic 13C signal intensity from relevant organs overtime time (right) are presented.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0668