Fast Ion-Chelate Dissociation Rate for In Vivo MRI of Labile Zinc with Frequency-Specific Encodability

Abstract

Fast ion-chelate dissociation rates and weak ion-chelate affinities are desired kinetic and thermodynamic features for imaging probes to allow reversible binding and to prevent deviation from basal ionic levels. Nevertheless, such properties often result in poor readouts upon ion binding, frequently result in low ion specificity, and do not allow the detection of a wide range of concentrations. Herein, we show the design, synthesis, characterization, and implementation of a Zn²⁺-probe developed for MRI that possesses reversible Zn²⁺-binding properties with a rapid dissociation rate (k_off = 845 ± 35 s^-1) for the detection of a wide range of biologically relevant concentrations. Benefiting from the implementation of chemical exchange saturation transfer (CEST), which is here applied in the ¹⁹F-MRI framework in an approach termed ion CEST (iCEST), we demonstrate the ability to map labile Zn²⁺ with spectrally resolved specificity and with no interference from competitive cations. Relying on fast k_off rates for enhanced signal amplification, the use of iCEST allowed the designed fluorinated chelate to experience weak Zn²⁺-binding affinity (K_d at the mM range), but without compromising high cationic specificity, which is demonstrated here for mapping the distribution of labile Zn²⁺ in the hippocampal tissue of a live mouse. This strategy for accelerating ion-chelate k_off rates for the enhancement of MRI signal amplifications without affecting ion specificity could open new avenues for the design of additional probes for other metal ions beyond zinc.

Figure 1

¹⁹F-NMR of fluorinated chelates designed for Zn²⁺ binding studies. (a) The chemical structure of the synthesized fluorinated chelates 1, 2, and 3 with the fluorine substituent at the 6, 3, and 5 positions of the pyridine ring, respectively. (b) Schematic illustration of the dynamic exchange process between the free and Zn²⁺-bound chelate and the obtained ¹⁹F-NMR spectrum of 1, 2, and 3 in the presence of Zn²⁺ at 25 °C (3 mM chelate and 0.6 mM ZnCl₂ at 100 mM Hepes buffer, pH = 7.2, 9.4 T NMR). Shown are the chemical shift offsets (Δω) between the peak of the free chelate (set at 0.0 ppm) and the peak Zn²⁺-chelate complex.

Figure 2

Zn²⁺-chelate exchange dynamics as a function of the chelate structure. (a) Synthetic route used for the synthesis of 3 and its methylated derivative 4. (b) Synthetic route used for the synthesis of 5 and its methylated derivative 6. (c) ¹⁹F-NMR spectra of 3 mM fluorinated chelates (3–6) in the presence of 0.6 mM Zn²⁺ at 37 °C and the obtained Δω between the peak of the free ligand (set at 0.0 ppm) and the Zn²⁺-bound ligand. (d) Representative ¹⁹F-iCEST spectra obtained for an aqueous solution of 3 mM of either of the chelates (from left to right: 3, 4, 5, or 6) in the presence of 30 μM Zn²⁺ at 37 °C. All NMR data were performed on aqueous solutions (100 mM Hepes buffer, pH = 7.2) at 37 °C with a 9.4 T NMR spectrometer. Reaction conditions: (i) 2-aminoethan-1-ol, NaBH(OAc)₃; (ii) PPh₃, CBr₄; (iii) 2-aminoethan-1-ol, K₂CO₃.

Figure 3

Zn²⁺-chelate exchange dynamics as a function of the chelate structure. (a) Evaluated exchange rates k_ex (s^–1) between Zn²⁺-bound and free chelate as determined for 3, 4, and 5. All NMR data were performed on aqueous solutions (100 mM Hepes buffer, pH = 7.2) at 37 °C with a 9.4 T NMR spectrometer. The X-ray crystal structures of the Zn²⁺-chelate complexes are shown for 3-Zn²⁺ (b), 4-Zn²⁺ (c), and 5-Zn²⁺ (d).

Figure 4

¹⁹F-iCEST Zn²⁺ sensitivity and selectivity using 5. (a) ¹⁹F-iCEST effect for 3 mM 5 as a function of Zn²⁺ concentration. (b) ¹⁹F-iCEST profile of 1 mM 5 and 500 nM Zn²⁺. (c) ¹⁹F-NMR spectra of 3 mM 5 in the presence of 0.6 mM Na⁺, K⁺, Mg²⁺, Ca²⁺, Mn²⁺, Fe²⁺, Co²⁺, Ni²⁺, Cu²⁺, and Zn²⁺ at 25 °C. (d) ¹⁹F-iCEST effect (Δω = 3.2 ppm) of 3 mM 5 and 30 μM (in 100 mM Hepes buffer, pH = 7.2) of s-block (Na⁺, K⁺, Mg²⁺, Ca²⁺) and d-block (Mn²⁺, Fe²⁺, Co²⁺, Ni²⁺, Cu²⁺, Zn²⁺) metal ions obtained at 37 °C on a 9.4 T NMR spectrometer. (e) ¹⁹F-iCEST MRI: (i) schematic representation of the studied phantom composed of seven tubes containing 7 mM 5 and 100 μM cation, i.e., Ca²⁺ (#1), Cu²⁺ (#2), Mg²⁺ (#3), Na⁺ (#4), K⁺ (#5), and Zn²⁺ (#7). Tube #6 contained only 5; (ii) ¹H-MRI; (iii) ¹⁹F-MRI obtained with a presaturation pulse applied at Δω = −3.2 ppm; (iv) ¹⁹F-MRI obtained with a presaturation pulse applied at Δω = +3.2 ppm; (v) ¹⁹F-iCEST map obtained by the subtraction of the image in (iv) from that in (iii) overlaid on ¹H-MRI.

Figure 5

In vivo¹⁹F-iCEST maps of labile Zn²⁺ pools in the mouse brain. Shown results for two regions of the brain: (a) CA3 in the hippocampus (zinc-rich ROI) or (b) the thalamus (TH, zinc-poor ROI). From left-to-right are the schematic illustration of the setup used to deliver 5 to either CA3 or TH, the¹H-MRI, the¹⁹F-MRI S^–Δω (presaturation pulse applied at Δω = −3.2 ppm, i.e., “off-resonance”), the ¹⁹F-MRI S^+Δω (presaturation pulse applied at Δω = +3.2 ppm, i.e., “on-resonance”), and the ¹⁹F-iCEST contrast (Zn²⁺ map) obtained from subtracting ¹⁹F-MRI S^+Δω from ¹⁹F-MRI S^–Δω overlaid on the ¹H-MRI. MRI scans were performed at 15.2 T. Infusion rate was set to 0.25 μL/min (of 10 mM 5 in PBS), and iCEST data acquisition started 90 min from the onset of the infusion of 5.

Figure 6

In vivo¹⁹F-iCEST quantification plot: The average percentile of ¹⁹F-iCEST contrast (S^Δω+/S^Δω–) as quantified in CA3 at Δω = 3.2 ppm (N = 7 mice) or Δω = 18 ppm (N = 7 mice), or in the TH (N = 7) at Δω = 3.2 ppm. Error bar denotes SEM, *p-value < 0.05, **p-value < 0.001, unpaired Student’s t test.