A ratiometric ¹⁹F MR-based method for the quantification of Ca²⁺ using responsive paramagnetic probes

Abstract

We present a method for assessing the extracellular calcium concentration using ¹⁹F chemical shift imaging. Specifically, a custom made calcium-responsive and lanthanide-based ¹⁹F MRI probe that undergoes a strong and highly specific modulation of its signal upon coordination with calcium ions was developed and its performance is presented.

---

Magnetic resonance imaging (MRI) is a powerful and widely used technique, able to provide three-dimensional anatomical images with excellent spatial resolution. However, it proves challenging to extrapolate further information about the voxel-wise composition of the signal and the specific nuclei from which it originates. On the other hand, magnetic resonance spectroscopy (MRS) allows acquisition of localized mono-dimensional (1D) MR spectra, while with chemical shift imaging (CSI) it is possible to obtain a spatial distribution of chemical shifts in a three-dimensional sample.^1–3 Using these techniques, a very detailed image of the investigated tissue can be obtained, without losing any spatial information. To this end, ¹H MRS has significantly contributed in providing meaningful insights into metabolic activity, particularly in the brain.⁴

By providing the localized 1D proton spectra of different metabolites with their specific chemical shifts,⁵ this technique enables the following of biological processes by monitoring the distribution and concentration of such metabolites in the tissue. However, because of the narrow spectral width of the proton frequencies, signals of metabolites usually result in complex spectra that are difficult to resolve, in particular in the region between 2 and 3 ppm. Other nuclei, such as ³¹P and ¹³C, are also exploited in MRS.⁶ Yet, both present major drawbacks, such as lower sensitivity compared to proton nuclei (both ³¹P and ¹³C), low isotopic abundance (¹³C), or absence in many biomolecules (³¹P). Moreover, the MRS methods mentioned above are suitable only for organic biomolecules and metabolites that contain ¹H, ¹³C or ³¹P nuclei, while they are unsuitable for endogenous metal ion targets.

Exploiting ¹⁹F represents an excellent alternative with high potential for in vivo MRI, MRS and CSI.⁷ Primarily, its high natural isotopic abundance and relative sensitivity are comparable to those of protons. Although the negligible endogenous concentration of ¹⁹F prevents its use in the direct investigation of biological processes, the lack of background signal is advantageous and can be exploited to develop tracers that yield high signal-to-noise ratios (SNR). Following this idea, emulsions of highly fluorinated molecules have been successfully used for labelling and subsequently tracking cells in vivo using MRI.^8–10 Furthermore, ¹⁹F is extremely sensitive to changes in its microenvironment on a chemical and electronic level, with an NMR resonance frequency range that is quite wide. Thus exploiting the relaxation properties, chemical shift and chemical exchange variations led to the development of responsive systems to measure tissue pO₂, and detect ions, pH and enzymatic activity via various triggering mechanisms.^7,11,12

Monitoring of fluctuations in the concentration of endogenous metal ions is a particularly interesting application because of their role and relevance in the signalling and regulation of numerous biological processes. By using responsive,¹³ or so-called smart contrast agents (SCAs), a functional MRI (fMRI) method that operates at the ¹H frequency demonstrated considerable potential for the assessment of neuronal activity by detecting the intracellular and extracellular calcium concentrations.^14,15 Importantly, the capability of Ca-sensitive SCAs to monitor cerebral ischemia has also been presented, thus showing significant recent advancements in this field and great prospects of molecular fMRI for contemporary biomedical research and applications.¹⁶

In parallel, the susceptibility of ¹⁹F nuclei has promoted the development of probes that can detect metal ions through changes in their chemical shift.^17,18 The ¹⁹F chemical exchange saturation transfer (CEST) investigations built upon this early work to result in an attractive methodology that allows multi-ion detection,^19,20 while self-assembled, ¹⁹F-containing peptide amphiphiles were also used to quantify Ca²⁺ concentrations that are biologically relevant.²¹ For ¹⁹F MRI applications, however, high concentrations and shorter relaxation times of a fluorinated probe would be required to obtain sufficient signals within reasonable acquisition times.⁷ One strategy to overcome this has been provided by combining fluorinated probes and paramagnetic ions at a short distance, usually within the same molecule.²² The ability of paramagnetic ions to affect the relaxation properties and chemical shifts of the ¹⁹F nuclei via paramagnetic relaxation enhancement (PRE) and pseudocontact shifts (PCS), respectively, has granted the possibility to exploit such phenomenon in the design of responsive systems for ¹⁹F MRI and MRS.^11,23,24 Benefiting from these effects, a few SCAs have been designed and assessed. For instance, SCAs for the detection of enzymatic activity showed their great potential;^25–27 however, the response of these probes is irreversible due to the cleavage of the enzyme-specific linker. On the other hand, SCAs that are able to reversibly detect pH, Ca²⁺ or citrate anions have also been reported.^28–30 However, their sensitivity was partially compromised by the presence of multiple signals at different frequencies due to the presence of molecular isomers, in the case of cyclen-based complexes,³¹ or a lower number of ¹⁹F spins in the responsive molecule.

With the intention to take advantage of the current knowledge in probe design and to capitalize on the favourable features of ¹⁹F MRI and MRS, we designed the ligand L that, when chelating paramagnetic lanthanide ions, can act as a calcium-responsive SCA (Fig. 1). To establish it as a probe suitable for future applications, we implemented a few significant improvements in the design of the sensor molecule. First, we used a lanthanide-cage based on the macrocyclic ligand 1,4-bis(carboxymethyl)-6-[bis(carboxymethyl)]amino-6-methylperhydro-1,4-diazepine (AAZTA).³² By selecting this chelator in the design of the SCA, we attempted to avoid additional ¹⁹F NMR resonances; to date no isomeric distribution in the NMR-timescale has been reported for such compounds. To improve the sensitivity of the probe, we used perfluorinated tert-butylether as the fluorine-bearing group, attached by a flexible propyl chain to the calcium-chelating moiety (Fig. 1). By doing so, we functionalized our molecule with a remarkably high number of spectroscopically equivalent ¹⁹F nuclei per paramagnetic ion (nine), while maintaining the excellent water solubility of the complex. Finally, we selected the calcium chelator based on ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid (EGTA) as the part of the molecule responsible for the coordination of Ca²⁺, due to its well investigated high specificity and affinity towards these ions when incorporated within the SCA.^30,33 The fluorinated ligand was obtained starting with the amide coupling of the amine 1 with bromoacetic acid (Scheme S1 in ESI†). The obtained bromide 2 was used to alkylate the secondary amine of 3. Meanwhile, fluorinated amine 6 was coupled with bromoacetyl bromide to obtain fluorinated bromide 7. Following hydrogenation of 4, 7 was used to alkylate the newly obtained secondary amine of 5, yielding a protected ligand 8. Once the chelating molecule L was prepared by hydrolyzing tert-butyl esters with formic acid, we selected dysprosium as the paramagnetic ion for incorporation in the SCA, due to the high PCS effect and the optimal PRE properties of the ions.

Fig. 1 Chemical structure of the reported complexes DyL and YL (top) and graphical illustration of the working principle of the responsive agent (bottom).

Advantageously, the ¹⁹F NMR spectra of DyL showed a single resonance at −72.7 ppm (Fig. 2b). Moreover, the obtained SNR value was as much as 5 times higher in comparison to those of compounds reported in the literature for the same concentration of metal ([Dy³⁺] = 3.0 mM), due to the larger amount of fluorine atoms per lanthanide ion.³⁰ Subsequently, we assessed the response of the system to Ca²⁺ by titrating DyL with a solution of CaCl₂ and recording the ¹⁹F NMR spectra (Fig. 2b and c). The obtained SNR of the ¹⁹F NMR signal undergoes a massive 10-fold decrease under these conditions. Such a powerful response is due to the PRE effect, resulting in the significant shortening of the longitudinal and transverse ¹⁹F relaxation times; the analogous changes in the ¹⁹F relaxation times for YL were not pronounced (Table 1).

Fig. 2 ¹⁹F NMR titration experiments with DyL and YL. (a) ¹⁹F NMR spectra of YL (left) and DyL (right) in the presence of increasing [Ca²⁺] (from 0 to 2.0 equiv.) measured at 7 T and 25 °C. (b) Signal intensity values for YL(A_Y, red area in the left panel) and DyL (A_Dy, blue area in the right panel) plotted as a function of increasing [Ca²⁺]. (c) Average values for A_Dy/A_Y plotted against the normalized [Ca²⁺] for a set of three samples containing DyL and YL in ratios of 50 : 50, 75 : 25 and 90 : 10 ([Dy] = 2.0 mM), measured at 7 T and 25 °C.

Table 1 ¹⁹F T₁ and T₂ relaxation times for DyL and YL in the absence and presence of Ca²⁺ (1 equiv.) at 7 T, 25 °C and pH 7.4 (50 mM HEPES)

Complex	T 1 (ms)	T 2 (ms)
DyL only/+ Ca^2+	157/127	17/6
YL only/+ Ca^2+	1320/1190	543/454

---

Concurrently, this effect is accompanied by the PCS and further shifting of the resonance frequency of the Ca²⁺-bound complex (ca. −1 ppm). We hypothesize that both these effects are the result of a contraction of the distance between the perfluorinated group and the paramagnetic centre, as a consequence of the structural rearrangement that occurs upon coordination of the Ca²⁺ ion. This assumption is corroborated by characterizing the ¹⁹F NMR signal behaviour of the diamagnetic Y³⁺ analogue YL. Indeed, by exchanging the paramagnetic ion for a diamagnetic one, we observed virtually no effect on the coordination of the Ca²⁺ ion on the chemical shift (<0.1 ppm) and signal intensity of the ¹⁹F NMR signal (Fig. 2a). Also, DyL did not show any interaction with Mg²⁺, while it proved able to coordinate Zn²⁺ (Fig. S2 in the ESI†). However, this phenomenon presents no problem for this methodology, because of the much lower concentration of Zn²⁺ in the brain extracellular space.³⁴

By obtaining Ca²⁺-independent or -dependent ¹⁹F NMR signals for YL and DyL, respectively, we were able to quantify variations in [Ca²⁺] by using the signal of the former as the internal reference for changes in the signal of the latter probe. To this end, we titrated a mixture of DyL and YL in three different ratios (50 : 50, 70 : 30 and 90 : 10, respectively) with CaCl₂, while maintaining the concentration of [DyL] (2.0 mM). We recorded the ¹⁹F NMR spectra after every addition of Ca²⁺, and integrated the regions between −70.0 and −70.5 ppm (A_Y) and between −72.5 and −73.0 ppm (A_Dy) to obtain the signal intensity of YL and DyL, respectively. Subsequently, we calculated the A_Dy/A_Y for every titration point, and plotted the normalized values of the three titrations against the [Ca²⁺] (Fig. 2c and Fig. S3–S6 in ESI†). Importantly, the behaviour of the complexes is very consistent and independent of their ratio and their total concentration.

To further explore the potential of this method, we prepared a set of phantom tubes containing a mixture of DyL and YL in 10 : 1 ratio (5.5 mM total [DyL + YL]) in the presence of 0.0, 0.2, 0.4, 0.6, 0.8 and 1.0 equivalents of Ca²⁺, and performed CSI measurements in a 7 T MRI scanner (Fig. S7 in the ESI†).

This allowed us to complete a voxel-wise analysis of the obtained ¹⁹F MR spectra and images, thus revealing the potential application in functional molecular imaging studies (Fig. 3 and 4). We first averaged the ¹⁹F spectra for each sample from the CSI image. In analogy to the ¹⁹F NMR experiments, the signal intensity of the DyL resonance decreased with increasing [Ca²⁺]; however, it remained constant under all conditions at the YL frequency (Fig. 3).

Fig. 3 Averaged ¹⁹F MRS spectra of phantom tubes 1–6 [DyL] = 5.0 mM/[YL] = 0.5 mM in 50 mM HEPES pH 7.4, [Ca²⁺]_eq = 0.0 (1), 0.2 (2), 0.4 (3), 0.6 (4), 0.8 (5), and 1.0 (6); obtained by recording a three-dimensional CSI dataset at 7 T.

Fig. 4 CSI on the phantom tubes containing a mixture of DyL and YL in 10 : 1 ratio (5.5 mM total [DyL + YL]) in the presence of 0.0, 0.2, 0.4, 0.6, 0.8 and 1.0 equivalents of Ca²⁺ in tubes 1–6, respectively. (a) Normalized CSI image of A_Dy/A_Y; the side bar shows the A_Dy/A_Y values. (b) Quantitative [Ca²⁺] map obtained by fitting the ratiometric data; the side bar shows the obtained [Ca²⁺] in mM.

We then processed the acquired CSI dataset by computing the A_Dy and A_Y values to provide a voxel-wise map of their ratio for applying a linear fit to the data obtained from the ¹⁹F NMR titration experiments (Fig. 2c and Fig. S6 in the ESI†), and we could correlate the concentration of Ca²⁺ according to eqn (1), where A_Dy/A_Y is the normalized ratio of DyL and YL signals, a and b are values obtained from the fit, and [Ca²⁺]_eq is the concentration of Ca²⁺ expressed in equivalents. Subsequently, we multiplied the calculated [Ca²⁺]_eq value with [DyL + YL] used in this measurement (5.5 mM) for each voxel. We obtained a voxel-wise concentration map of the absolute [Ca²⁺] expressed in mM, which was in excellent agreement with the actual Ca²⁺ concentration used in each sample (Fig. 4b and Table S3 in the ESI†).

In turn, this method proved to be fully quantitative under all conditions where [DyL] ≥ [Ca²⁺] and [YL] is detectable. Specifically, due to the ability of ¹⁹F MRI to quantify the amount of ¹⁹F spins,¹⁰ the method presented here can achieve the determination of overall probe concentration via the fluorine resonance in YL, while the resonance of DyL and its variable signal will aid the quantification of the Ca²⁺ concentration. This specific and very important feature of the methodology presented here could turn into an important tool for the absolute quantification of essential endogenous ions such Ca²⁺ or Zn²⁺ that exist in μM to mM concentrations,³⁴ since the abnormalities in their concentration may be indicative of various pathological conditions.^35,36 Also the majority of practical aspects for the implementation of this methodology should be fulfilled: despite the lower sensitivity of ¹⁹F MRS/MRI, high amounts of ¹⁹F nuclei in the reported SCA ensure common duration of the CSI recordings on the order of tens of minutes. Namely, the tests in the scanner were made with the acquisition time of 1 hour for a quite large field of view (80 × 80 mm) and a very good signal-to-noise ratio (SNR) of >60 was obtained for 10% of YL in each sample. Considering this high SNR, the CSI acquisition could be reduced a few times with still improved spatial resolution, while providing good quality data to support the necessary conclusions.

We report a methodology that is highly advantageous for potential in vivo utilization. Generally, variations in the concentration of the probe and the inability to quantify it in biological tissues currently represent a major hindrance for the application of responsive CAs in vivo. However, our method has the capacity to circumvent most of these obstacles: it deals with a pair of highly specific Ca-variable and -invariable signals, the latter allowing easy quantification of the SCA, while not requiring the exact binding affinity constants to successfully determine the [Ca²⁺]. The two ¹⁹F signals are generated by two chemically different probes, which should have virtually identical biocompatibility and behaviour in vivo (i.e. diffusion rate, stability, interaction with tissue and excretion). Moreover, the high number of fluorine atoms per SCA and the acquisition of a single resonance can ensure signal detection after acquisition times at the level of minutes. Additionally, the developed mechanism could be employed for the design of SCAs suitable for other metal ions and molecular targets, provided that the sensor or recognition moiety is appropriately adjusted. Overall, the approach shown here represents an incredibly attractive perspective to overcome some of the main obstacles to the use of SCAs in vivo. The possibility to obtain spatially resolved maps of [Ca²⁺] in a tissue would represent a great leap towards a better understanding of numerous pathological and biological processes such as ischemia and neural activity, defining a substantial step forward for contemporary molecular functional imaging studies.

The financial support from the German Research Foundation (DFG, grant AN 716/7-1), the German Federal Ministry of Education and Research (BMBF, e:Med program: FKZ: 01ZX1503) and the German Academic Exchange Service (DAAD, PhD fellowship to T. G.) is gratefully acknowledged. Open Access funding provided by the Max Planck Society.

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. L. Brateman, Am. J. Roentgenol., 1986, 146, 971–980.
2. T. R. Brown, B. M. Kincaid and K. Ugurbil, Proc. Natl. Acad. Sci. U. S. A., 1982, 79, 3523–3526.
3. A. Haase, J. Frahm, W. Hanicke and D. Matthaei, Phys. Med. Biol., 1985, 30, 341–344.
4. J. Frahm and F. Hanefeld, Neuropediatrics, 1996, 27, 64–69.
5. P. S. Allen, R. B. Thompson and A. H. Wilman, NMR Biomed., 1997, 10, 435–444.
6. P. C. M. Van Zijl and P. B. Barker, Ann. N.Y. Acad. Sci., 1997, 820, 75–96.
7. J.-X. Yu, R. R. Hallac, S. Chiguru and R. P. Mason, Prog. Nucl. Magn. Reson. Spectrosc., 2013, 70, 25–49.
8. J. M. Janjic and E. T. Ahrens, WIREs Nanomed. Nanobiotechnol., 2009, 1, 492–501.
9. M. Srinivas, A. Heerschap, E. T. Ahrens, C. G. Figdor and I. J. M. D. Vries, Trends Biotechnol., 2010, 28, 363–370.
10. M. Srinivas, P. Boehm-Sturm, C. G. Figdor, I. J. de Vries and M. Hoehn, Biomaterials, 2012, 33, 8830–8840.
11. K. L. Peterson, K. Srivastava and V. C. Pierre, Front. Chem., 2018, 6, 160.
12. M. Yu, B. S. Bouley, D. Xie, J. S. Enriquez and E. L. Que, J. Am. Chem. Soc., 2018, 140, 10546–10552.
13. M. C. Heffern, L. M. Matosziuk and T. J. Meade, Chem. Rev., 2014, 114, 4496–4539.
14. S. Okada, B. B. Bartelle, N. Li, V. Breton-Provencher, J. J. Lee, E. Rodriguez, J. Melican, M. Sur and A. Jasanoff, Nat. Nanotechnol., 2018, 13, 473–477.
15. A. Barandov, B. B. Bartelle, C. G. Williamson, E. S. Loucks, S. J. Lippard and A. Jasanoff, Nat. Commun., 2019, 10, 897.
16. T. Savić, G. Gambino, V. S. Bokharaie, H. R. Noori, N. K. Logothetis and G. Angelovski, Proc. Natl. Acad. Sci. U. S. A., 2019, 116, 20666–20671.
17. G. A. Smith, R. T. Hesketh, J. C. Metcalfe, J. Feeney and P. G. Morris, Proc. Natl. Acad. Sci. U. S. A., 1983, 80, 7178–7182.
18. H. L. Kirschenlohr, J. C. Metcalfe, P. G. Morris, G. C. Rodrigo and G. A. Smith, Proc. Natl. Acad. Sci. U. S. A., 1988, 85, 9017–9021.
19. A. Bar-Shir, A. A. Gilad, K. W. Y. Chan, G. S. Liu, P. C. M. van Zijl, J. W. M. Bulte and M. T. McMahon, J. Am. Chem. Soc., 2013, 135, 12164–12167.
20. A. Bar-Shir, N. N. Yadav, A. A. Gilad, P. C. M. van Zijl, M. T. McMahon and J. W. M. Bulte, J. Am. Chem. Soc., 2015, 137, 78–81.
21. A. T. Preslar, L. M. Lilley, K. Sato, S. R. Zhang, Z. K. Chia, S. I. Stupp and T. J. Meade, ACS Appl. Mater. Interfaces, 2017, 9, 39890–39894.
22. K. H. Chalmers, E. De Luca, N. H. M. Hogg, A. M. Kenwright, I. Kuprov, D. Parker, M. Botta, J. I. Wilson and A. M. Blamire, Chem. – Eur. J., 2010, 16, 134–148.
23. I. Bertini, C. Luchinat and G. Parigi, Prog. Nucl. Magn. Reson. Spectrosc., 2002, 40, 249–273.
24. P. Harvey, I. Kuprov and D. Parker, Eur. J. Inorg. Chem., 2012, 2015–2022.
25. S. Mizukami, R. Takikawa, F. Sugihara, Y. Hori, H. Tochio, M. Walchli, M. Shirakawa and K. Kikuchi, J. Am. Chem. Soc., 2008, 130, 794–795.
26. S. Mizukami, H. Matsushita, R. Takikawa, F. Sugihara, M. Shirakawa and K. Kikuchi, Chem. Sci., 2011, 2, 1151–1155.
27. A. Keliris, I. Mamedov, G. E. Hagberg, N. K. Logothetis, K. Scheffler and J. Engelmann, Contrast Media Mol. Imaging, 2012, 7, 478–483.
28. A. M. Kenwright, I. Kuprov, E. De Luca, D. Parker, S. U. Pandya, P. K. Senanayake and D. G. Smith, Chem. Commun., 2008, 2514–2516.
29. P. Harvey, K. H. Chalmers, E. De Luca, A. Mishra and D. Parker, Chem. – Eur. J., 2012, 18, 8748–8757.
30. P. Kadjane, C. Platas-Iglesias, P. Boehm-Sturm, V. Truffault, G. E. Hagberg, M. Hoehn, N. K. Logothetis and G. Angelovski, Chem. – Eur. J., 2014, 20, 7351–7362.
31. K. J. Miller, A. A. Saherwala, B. C. Webber, Y. Wu, A. D. Sherry and M. Woods, Inorg. Chem., 2010, 49, 8662–8664.
32. S. Aime, L. Calabi, C. Cavallotti, E. Gianolio, G. B. Giovenzana, P. Losi, A. Maiocchi, G. Palmisano and M. Sisti, Inorg. Chem., 2004, 43, 7588–7590.
33. G. Angelovski, P. Fouskova, I. Mamedov, S. Canals, E. Toth and N. K. Logothetis, ChemBioChem, 2008, 9, 1729–1734.
34. G. G. Somjen, Ions in the brain: normal function, seizures, and strokes, Oxford University Press, New York, Oxford, 2004.
35. P. Marambaud, U. Dreses-Werringloer and V. Vingtdeux, Mol. Neurodegener., 2009, 4, 20.
36. M. C. McCord and E. Aizenman, Front. Aging Neurosci., 2014, 6, 77.

---

Footnote

† Electronic supplementary information (ESI) available: Synthetic procedures and additional ¹⁹F NMR and CSI experimental data. See DOI: 10.1039/c9cc09977h

---