Periodic Reporting for period 1 - PARA-FLUOR (Paramagnetic, fluorinated and water-soluble metal complexes for 19F MRI)
Reporting period: 2022-10-01 to 2024-09-30
19F MRI has emerged recently as a complementary technique to classical, 1H-based MRI. Given the lack of background signal, 19F MRI is very attractive for various biomedical applications, including in vivo cell tracking. However, it also has many limitations, mainly related to low solubility, low stability, low sensitivity and relatively long 19F relaxation times of current 19F MRI probes, which are mostly based on polyfluorinated compounds. This project aimed to overcome these drawbacks by developing highly water-soluble, stable and kinetically inert metal complexes. They consist of small molar weight (< 1000 Da), 19F containing ligands and paramagnetic metal ions (Mn(II) or Fe(III)) which largely shorten 19F relaxation times, thereby maximizing the 19F MRI signal to noise ratio achievable in a given timeframe.
Building on the optimal relaxation properties of paramagnetic iron and manganese ions and on the strong expertise of the candidate and the host group in Mn(II) and Fe(III) coordination chemistry, the project’s goal was to specifically address the major current bottlenecks of 19F MRI by developing 19F probes which (i) avoid solubility, stability and biodistribution problems associated with nanoemulsions, and (ii) maximize 19F MRI signal to noise ratio by using efficient relaxation agents. We aimed at combining, for the first time, advantageous paramagnetic properties of Mn(II)/(III) or Fe(II)/(III) and small molar weight molecules which have sufficient 19F content for MRI detection to create water soluble, highly stable and inert complexes endowed with short 19F relaxation times and good biocompatibility. This requires structural optimization by integrating 19F atoms in the complex at an optimized distance from the metal ion in order to allow for fast relaxation without extensive broadening of the 19F NMR peaks that would cause signal vanishing in MRI. Additionally, we also aimed at designing probes, which are efficient not only in 19F MRI, but can combine detection capabilities in both 19F and classical 1H MRI.
Building on the optimal relaxation properties of paramagnetic iron and manganese ions and on the strong expertise of the candidate and the host group in Mn(II) and Fe(III) coordination chemistry, the project’s goal was to specifically address the major current bottlenecks of 19F MRI by developing 19F probes which (i) avoid solubility, stability and biodistribution problems associated with nanoemulsions, and (ii) maximize 19F MRI signal to noise ratio by using efficient relaxation agents. We aimed at combining, for the first time, advantageous paramagnetic properties of Mn(II)/(III) or Fe(II)/(III) and small molar weight molecules which have sufficient 19F content for MRI detection to create water soluble, highly stable and inert complexes endowed with short 19F relaxation times and good biocompatibility. This requires structural optimization by integrating 19F atoms in the complex at an optimized distance from the metal ion in order to allow for fast relaxation without extensive broadening of the 19F NMR peaks that would cause signal vanishing in MRI. Additionally, we also aimed at designing probes, which are efficient not only in 19F MRI, but can combine detection capabilities in both 19F and classical 1H MRI.
In the objective of exploring the potential of stable, inert, highly fluorinated, and small molecular weight Mn(II) chelates as dual 19F and 1H MRI agents, a novel ligand, L1, was designed and its synthesis was successfully performed in two steps from commercial cyclohexyldiamine-tetraacetic acid (CDTA). Ligand purification was achieved by chromatography (HPLC). Protonation constants of the ligand as well as protonation and stability constants of the Mn, Mg, Ca and Zn complexes were determined by pH potentiometric techniques. Kinetic inertness of the MnL1 complex, fundamental for in vivo safety, was investigated through Cu(II) and Zn(II) exchange reactions, which were monitored by UV-visible spectrophotometry and 1H relaxometry, respectively. Based on these results, we could conclude that this MnL1 complex possesses high stability (log KMn(II)L= 12.51; pMn = 8.17 calculated as -log [Mn2+]free at pH = 7.4 cLig = cMn2+ = 10 µM) and high resistance to dissociation (dissociation half-life, t1/2= 1285 h, was estimated at pH= 7.4 and 25 °C).
Three isomers of the MnL1 complex could be separated and their interconversion was found to be very slow (approximately 10 % in 5 days), allowing for their individual investigation. Their 1H relaxation properties were investigated by fast field cycling relaxometry and 17O NMR measurements were performed on an aqueous solution of MnL1 isomers to assess water exchange. Thanks to the high proton relaxivity, MnL1 provides excellent contrast enhancement on 1H MRI images (r1p= 5.36 and 5.26 mM-1s-1 for Isomer 1 and 3 at 20 MHz and 25 °C, respectively). Finally, the combined analysis of the 1H NMRD and 17O NMR data allowed for the determination of all key parameters defining proton relaxation efficiency, such as water exchange rate, rotational correlation time and number of water molecules coordinated to the metal ion.
In order to characterize the capacity of the complex to act as an 19F MRI agent, 19F relaxation times were determined for the different MnL1 isomers and for the ligand at various field strengths. Due to the presence of the highly paramagnetic Mn(II) metal ion, the T1 and T2 relaxation times of the fluorine in the ligand were greatly reduced to the milliseconds range (T1: 2.2 – 2.8 ms, T2: 1.5 – 1.9 ms at 9.4 T and 25 °C) allowing for higher number of MRI scans, thus better signal intensity, in a given timeframe. The analysis of the field-dependent relaxation rates allowed for calculating the Mn-F distance (rMnF = 8.2±0.2 Å). The hydrophobicity of the complex was characterized by determining the logP value by the shake-flask method combined with 19F NMR determination of the concentrations. The lack of cytotoxicity has been evidenced on different cell lines. Indeed, MnL1 did not show cytotoxic effect up to 2 mM complex concentration on HeLa and K-562 lymphoblast cells. Also, MnL1 was found to undergo limited cellular internalization in K-562 lymphoblasts.
1H and 19F phantom MRI studies and in vivo MRI acquisitions were performed to assess the MRI performance of the probe. In vivo MRI acquisitions following intramuscular injection of MnL1 into the mouse hind leg (c= 2.1 mM, 40 µl) were done by using fast acquisition MRI techniques adapted to fast relaxation and yielded excellent signal to noise ratios in a short time (9 min) and with good resolution (128×128 matrix size).
Hypoxia is related to different pathological states, thus detection of hypoxic tissues is crucial for early diagnosis allowing early disease treatment. In this context, fluorinated porphyrin complexes have been also investigated as potential redox sensors detectable in both 1H and 19F MRI. They are based on the MnII/MnIII redox switch, which can be activated by biological reducing and oxidizing agents, such as ascorbate and peroxides, respectively. The fluorinated porphyrin-based manganese complex Mn-TPP-p-CF3 is one example of such a potential dual redox responsive contrast agent in 1H and 19F MRI. Although Mn-TPP-p-CF3 is insoluble in pure aqueous media (we have to improve the solubility of these kinds of compounds), it is a useful model compound to better understand the structural features required in particular for 19F redox reporters based on manganese porphyrins. Therefore, we have characterized the redox properties of Mn-TPP-p-CF3 by cyclic voltammetry. It has a reversible Mn(II)/Mn(III) redox potential 0.574 V vs. NHE in HEPES/THF solution under oxygen free condition. The reduction of Mn(III)-TPP-p-CF3 in the presence of ascorbate is slow at pH 7.4 but fully reversed in the presence of air oxygen, as monitored by UV-Vis spectrometry and 19F NMR. The moderately broad 19F NMR signals of Mn(III)-TPP-p-CF3 disappear in the presence of 1 equivalent ascorbate and are replaced by a shifted and further broadened 19F NMR signal originating from Mn(II)-TPP-p-CF3. 19F relaxation times of the Mn(III) and Mn(II) complexes are remarkably different (T1: 11.5 and 6.5 ms, T2: 7.7 and 1.9 ms at 9.4 T and 25 °C, respectively) due to the difference in the paramagnetic nature of the two oxidation states. Phantom 19F MR images in DMSO show a MRI signal intensity decrease and a signal shift upon reduction of Mn(III)-TPP-p-CF3, retrieved upon complete re-oxidation in air within ~24 h.
Three isomers of the MnL1 complex could be separated and their interconversion was found to be very slow (approximately 10 % in 5 days), allowing for their individual investigation. Their 1H relaxation properties were investigated by fast field cycling relaxometry and 17O NMR measurements were performed on an aqueous solution of MnL1 isomers to assess water exchange. Thanks to the high proton relaxivity, MnL1 provides excellent contrast enhancement on 1H MRI images (r1p= 5.36 and 5.26 mM-1s-1 for Isomer 1 and 3 at 20 MHz and 25 °C, respectively). Finally, the combined analysis of the 1H NMRD and 17O NMR data allowed for the determination of all key parameters defining proton relaxation efficiency, such as water exchange rate, rotational correlation time and number of water molecules coordinated to the metal ion.
In order to characterize the capacity of the complex to act as an 19F MRI agent, 19F relaxation times were determined for the different MnL1 isomers and for the ligand at various field strengths. Due to the presence of the highly paramagnetic Mn(II) metal ion, the T1 and T2 relaxation times of the fluorine in the ligand were greatly reduced to the milliseconds range (T1: 2.2 – 2.8 ms, T2: 1.5 – 1.9 ms at 9.4 T and 25 °C) allowing for higher number of MRI scans, thus better signal intensity, in a given timeframe. The analysis of the field-dependent relaxation rates allowed for calculating the Mn-F distance (rMnF = 8.2±0.2 Å). The hydrophobicity of the complex was characterized by determining the logP value by the shake-flask method combined with 19F NMR determination of the concentrations. The lack of cytotoxicity has been evidenced on different cell lines. Indeed, MnL1 did not show cytotoxic effect up to 2 mM complex concentration on HeLa and K-562 lymphoblast cells. Also, MnL1 was found to undergo limited cellular internalization in K-562 lymphoblasts.
1H and 19F phantom MRI studies and in vivo MRI acquisitions were performed to assess the MRI performance of the probe. In vivo MRI acquisitions following intramuscular injection of MnL1 into the mouse hind leg (c= 2.1 mM, 40 µl) were done by using fast acquisition MRI techniques adapted to fast relaxation and yielded excellent signal to noise ratios in a short time (9 min) and with good resolution (128×128 matrix size).
Hypoxia is related to different pathological states, thus detection of hypoxic tissues is crucial for early diagnosis allowing early disease treatment. In this context, fluorinated porphyrin complexes have been also investigated as potential redox sensors detectable in both 1H and 19F MRI. They are based on the MnII/MnIII redox switch, which can be activated by biological reducing and oxidizing agents, such as ascorbate and peroxides, respectively. The fluorinated porphyrin-based manganese complex Mn-TPP-p-CF3 is one example of such a potential dual redox responsive contrast agent in 1H and 19F MRI. Although Mn-TPP-p-CF3 is insoluble in pure aqueous media (we have to improve the solubility of these kinds of compounds), it is a useful model compound to better understand the structural features required in particular for 19F redox reporters based on manganese porphyrins. Therefore, we have characterized the redox properties of Mn-TPP-p-CF3 by cyclic voltammetry. It has a reversible Mn(II)/Mn(III) redox potential 0.574 V vs. NHE in HEPES/THF solution under oxygen free condition. The reduction of Mn(III)-TPP-p-CF3 in the presence of ascorbate is slow at pH 7.4 but fully reversed in the presence of air oxygen, as monitored by UV-Vis spectrometry and 19F NMR. The moderately broad 19F NMR signals of Mn(III)-TPP-p-CF3 disappear in the presence of 1 equivalent ascorbate and are replaced by a shifted and further broadened 19F NMR signal originating from Mn(II)-TPP-p-CF3. 19F relaxation times of the Mn(III) and Mn(II) complexes are remarkably different (T1: 11.5 and 6.5 ms, T2: 7.7 and 1.9 ms at 9.4 T and 25 °C, respectively) due to the difference in the paramagnetic nature of the two oxidation states. Phantom 19F MR images in DMSO show a MRI signal intensity decrease and a signal shift upon reduction of Mn(III)-TPP-p-CF3, retrieved upon complete re-oxidation in air within ~24 h.
First of all, we have pioneered a fundamentally novel concept to create highly efficient 19F MRI agents which can yield excellent signal to noise ratios in 19F MR images in very short acquisition times. So far, the combination of strong relaxation agents with fluorinated ligands was considered to lead to 19F signal broadening and thus non-detectable MRI signals. We have evidenced for the first time that, small, fluorinated Mn2+ complexes can be potential alternatives of perfluorinated nanoparticles in 19F MRI applications. Towards this goal, we have successfully made the first steps of the rational optimization of the molecular structure and we have shown that it is possible by using highly rigid chelators that allow for a fine control of the 19F distance. We have provided the first detailed investigation of a Mn(II) complex in the context of 19F MRI, including in vivo detection. In addition, this probe allows, for the first time, coupling efficient 19F MRI with classical 1H detection. While this first system requires further optimization (e.g. the Mn-F distance or delivery to biological targets, etc), our work provided proof of concept data, which will certainly inspire further research in the quickly growing field of 19F MRI.
In parallel, we have also extended our interest towards the detection of redox states in combined 19F and 1H MRI. The first example proposed in this context is based on a manganese-porphyrin chelate. Again, the first steps in rationalizing the structure-efficacy relationships have been made. The major limitation of this system remains its aqueous solubility, which needs to be improved for in vivo use.
In parallel, we have also extended our interest towards the detection of redox states in combined 19F and 1H MRI. The first example proposed in this context is based on a manganese-porphyrin chelate. Again, the first steps in rationalizing the structure-efficacy relationships have been made. The major limitation of this system remains its aqueous solubility, which needs to be improved for in vivo use.