FLUORINE METABOLIC IMAGING (FMI) WITH 3-FLUORO-3-DEOXY-SUGARS

[0111] FIG. 1 shows the interaction of 3FDG and 3FDGal with endogenous enzymes, revealing cellular oxidation or reduction.

[0112] FIGS. 2A and 2B show schemes of phantom experiments conducted with tubes containing the 3FD-compounds (2A—the sugar 3FDG and its metabolites 3FDS and 3FDGA; 2B—the sugar 3FDGal and its metabolites 3FDGtol and 3FDGalA), verifying the feasibility of spectrally differentiating the ¹⁹F-MR signals of the sugars and their metabolites, and the ability to spatially map the sugar and metabolites with a single mCSSI-based FMI acquisition.

[0113] FIG. 3 shows the in-vivo metabolism of systemically administered 3FDG to mice.

[0114] FIG. 4 shows the in-vivo metabolism of systemically administered 3FDGal to mice.

[0115] FIG. 5 is localized ¹⁹F-MR spectroscopy of 3FDGal, overlaid on 1H-MRI data.

[0116] FIG. 6 shows the in-vivo metabolite yield of 3FDGal, concurrently reporting the spatial distribution of sugar oxidizing and reducing enzymes, generating 3FDGalA and 3FDGtol, respectively.

[0117] FIG. 7 shows the results of in-vitro incubation studies based on 3FDGal transport into different cell lines, and the metabolites produced in the cells.

[0118] FIG. 8 shows a ¹⁹F MR-generated map of athymic nude mice upon administration of 3FDGal to mice that was subcutaneously injected with A549 and LNCaP cells to induce proliferating subcutaneous tumors. Both 3FDGalA and 3FDGtol formation by A549 tumors and no 3FDGal metabolism by LNCaP tumors are observed.

[0119] FIG. 9 shows the flowchart (flow diagram) of the synthetic process of preparation of 3FDGal from 3-Deoxy-3-fluoro-1,2:5,6-bis-O-(1-methylethylidene)-ca-D-galactofuranose in a QMA cartridge.

[0120] FIG. 10 shows the ¹⁹F-NMR spectra of 3FDGal synthesis in the QMA cartridge.

[0121] FIGS. 11A-11N: Identifying the human enzymes driving the metabolic schemes of 3FDG and 3FDGal. FIG. 11A-FIG. 11B show schematic overview of the experimental design. HEK-293 cells were transfected with expression vectors (one vector at a time) that encode for the overexpression of different enzymes, to assess their potential to intracellularly reduce or oxidize 3FDG or 3FDGal (FIG. 11A). The cells were then incubated with either 3FDG or 3FDGal, followed by ¹⁹F-NMR of the intracellular (IC) content (FIG. 11B) (n=3 biologically independent experiments in all groups). FIG. 11C shows a Western blot analysis of the transfected cells confirming the successful expression of the examined proteins. FIG. 11D-FIG. 11E show representative high-resolution ¹⁹F{¹H}-NMR spectra of the IC content acquired in one control (HEK-293^NT) and two experimental conditions (HEK-293^AKR1B1 and HEK-293^ALDH1A1) following incubation with 3FDG (FIG. 11D) or 3FDGal (FIG. 11E). ¹⁹F-NMR spectra from HEK-293^NT show endogenous generation of 3FDS (but not 3FDGA) from 3FDG, and 3FDGtol (but not 3FDGalA) from 3FDGal. Spectra from HEK-293^AKR1B1 show enhanced generation of 3FDS (from 3FDG) and 3FDGtol (from 3FDGal). Spectra from HEK-293^ALDH1A1 show induced generation of 3FDGA (from 3FDG) and 3FDGalA (from 3FDGal). FIG. 11F-FIG. 11G show quantification of ¹⁹F-NMR SNR values calculated in all conditions at the relevant chemical shifts: δ=−212.4 (3FDS) and δ=−208.4 (3FDGA) in cells incubated with 3FDG (11F), and δ=−209.9 ppm (3FDGtol) and δ=−205.1 ppm (3FDGalA) in cells incubated with 3FDGal (FIG. 11G). FIG. 11H shows that after co-transfecting HEK-293 cells with AKR1B1 and SORD (schematic presentation on the top), Western blot analysis confirmed the successful expression of both proteins (bottom). FIG. 11I-FIG. 11L show representative high-resolution ¹⁹F{¹H}-NMR spectra of the IC content acquired in HEK-293^AKR1B1SORD cells show 3FDS and 3-fluoro-3-deoxy-fructose (3FDF) generation following incubation with 3FDG (FIG. 11I), versus 3FDGtol-only generation following incubation with 3FDGal (FIG. 11K). Quantification of ¹⁹F-NMR SNR values calculated at δ=−212.4 (3FDS) and δ=−207.7 (3FDF) in HEK-293^AKR1B1 versus HEK-293^AKR1B1+SORD cells incubated with 3FDG (FIG. 11J), and at δ=−209.9 ppm (3FDGtol) in HEK-293^AKR1B1 versus HEK^AKR1B1+SORD cells incubated with 3FDGal (FIG. 11L). FIG. 11M-FIG. 11N show the intracellular trapping mechanism for 3FDG (FIG. 11M) versus 3FDGal (FIG. 11N). SNR, signal-to-noise ratio. Data presented as mean values±s.d.

[0122] FIGS. 12A-12G show recording 3FDGal metabolism with ¹⁹F-NMR enables fingerprinting of cancer cells. FIG. 12A shows schematic presentation of the in vitro incubation experiment. Different human cancer cell lines were incubated with 3FDGal (6×10⁶ cells; 5 mM 3FDGal; 4 hrs; n=3 biologically independent experiments per cell line). Then, the intracellular (IC) content of the cells was analyzed on ¹⁹F-NMR, GC-MS, and Western blot. FIG. 12B shows representative high-resolution ¹⁹F{¹H}-NMR spectra of the IC content for each of the assessed cell lines. FIG. 12C-FIG. 12D show quantified SNR values calculated for the ¹⁹F-NMR peaks of 3FDGalA at δ=−205.1 ppm (top) and of 3FDGtol at δ=−209.9 ppm (bottom), represented per cell line on bar plots (FIG. 12C) and on a heatmap that illustrates the dual metabolic color-coded ¹⁹F-NMR fingerprints of the cells (FIG. 12D). FIG. 12E shows 3FDGalA (top) and 3FDGtol (bottom) concentrations, as determined by analytical GC-MS, presented per cell line on bar plots. FIG. 12F shows concentrations of 3FDGalA (top) and 3FDGtol (bottom) as a function of the SNR values calculated for the ¹⁹F-NMR peaks of 3FDGalA at δ=−205.1 ppm (top) and 3FDGtol at δ=−209.9 ppm (bottom) in all analyzed IC samples. A statistically significant excellent correlation was found between ¹⁹F-NMR-based and GC-MS-based values in both plots (Pearson's r=0.99, p-value <0.001). FIG. 12G shows Western blot of the analyzed cancer cells for ALDH1A1 and AKR1B1 expression. GC-MS, Gas chromatography-mass spectrometry. Data are presented as mean values±s.d.

[0123] FIGS. 13A-13C show imaging data of a representative mouse: ¹H-MRI anatomic axial slice at the level of the two tumors (13A) and ¹⁹F-MRI-based axial 3D-FMI-2 data from the corresponding axial location, overlaid on anatomical ¹H-MRI data (13A, right and center panels), which indicate 3FDGalA and 3FDGtol generation in the A549-derived tumor but not in the LNCaP-CL1-derived tumor. FIG. 13B show comparison of quantitative image-based parameter (SNRmax, TMV, and TTM) values as calculated in the A549- versus the LNCaP-CL1-derived tumors (data from VOIs covering the tumors at 3D-FMI-2). FIG. 13C show H&E staining and immunostaining (anti-ALDH1A1 and anti-AKR1B1) of the tumors studied in vivo.

[0124] FIGS. 14A-14G. FIG. 14A shows a heatmap that illustrates the predicted proteomic-based dual metabolic color-coded ¹⁹F-NMR fingerprints of all NSCLC cell lines included in the NCI-60 panel. FIG. 14B shows representative high-resolution ¹⁹F{¹H}-NMR spectra of the IC content for NCIH460 (top) and NCIH522 (bottom) cells after their incubation with 3FDGal. FIG. 14C shows a quantified SNR values calculated for the ¹⁹F-NMR peaks of 3FDGalA at δ=−205.1 ppm and of 3FDGtol at δ=−209.9 ppm. FIG. 14D schematically show NCIH522 and NCIH460 cells were subcutaneously injected (right and left, respectively) to induce a bilateral xenograft mouse model. FIG. 14E-FIG. 14G show 3D-FMI-2 data overlaid on anatomical 1H-MRI data at the level of the two induced tumors indicate 3FDGalA^high/3FDGtol^low and 3FDGalA^low/3FDGtol^high imaging phenotypes in NCIH522- and NCIH460-derived tumors, respectively, mirroring their ALDH1A1^high/AKR1B^low and ALDH1A1^low/AKR1B1^high proteomic phenotypes, respectively. SNRmax, maximal signal-to-noise ratio. TMV, tissue metabolite volume. TTM, total tissue metabolite. VOI, volume-of-interest. H&E, hematoxylin and eosin. NSCLC, non-small cell lung cancer. NCI-60, National Cancer Institute-60.

[0125] FIG. 15 shows a heatmap that illustrates the predicted proteomic-based dual metabolic color-coded ¹⁹F-NMR fingerprints of all RCC cell lines included in the NCI-60 panel.

[0126] FIGS. 16A-16B show a schematic presentation of the experimental procedure (FIG. 16A) and a two time points FMI of a representative mouse (FIG. 16B).

[0127] FIGS. 17A-17C show a schematic presentation of the experimental procedure (FIG. 17A) and detection of the xenografts (17B), visualizing the expected FMI phenotype (ACHN: high AKR1B1 expression, mild ALDH1A1 expression based on both proteomic and Western blot (WB) data. Quantification is shown in FIG. 17C.

[0128] FIGS. 18A-18B show a schematic presentation of the experimental procedure (FIG. 18A) and demonstrates also with 3FDG the ability to visualize uptake of 3FDG (green) followed by reduction d/t high AKR1B1 activity (conversion to 3FDS, red).

[0129] FIGS. 19A-19C show a schematic presentation of the experimental procedure (FIG. 19A) and fast FMI with 3FDGal, reducing acquisition time to 20 minute only per step, which is graphically shown in FIG. 19C.

[0130] FIGS. 20A-20D show recording of 3FDGal metabolism with ¹⁹F-NMR in two different cancer cells used to induce tumors in the brain. FIG. 20A shows a ¹⁹F-NMR spectrum of the intracellular (IC) content of A549 cells after incubation with 3FDGal. FIG. 20B shows a quantified SNR values calculated for the ¹⁹F-NMR peaks of 3FDGalA (−205.1 ppm) and of 3FDGtol (−209.9 ppm) as obtained for A549 cells (N=3). FIG. 20C shows a ¹⁹F-NMR spectrum of the intracellular (IC) content of TBDRG-05MG cells after incubation with 3FDGal. FIG. 20D shows quantified SNR values calculated for the ¹⁹F-NMR peaks of 3FDGalA (−205.1 ppm) and of 3FDGtol (−209.9 ppm) as obtained for TBDRG-05MG cells (N=3).

[0131] FIGS. 21A-21C shows a three-dimensional Fluorine Metabolic Imaging (3D-FMI) of the brain. Healthy mouse (top row, FIG. 21A) and mice inoculated with two types of tumors in their brains were studied using 3D-FMI 2 hours post-3FDGal injection. Shown are, from left to right, 4 different MRI data sets for each mouse representing an axial view of a single selected slice of the brain, anatomical ¹H-MRI, 3FDGal map (overlayed on the corresponding 1H-MRI data), 3FDGalA map (overlayed on the corresponding 1H-MRI data), and 3FDGtol map (overlayed on the corresponding 1H-MRI data). After intraperitoneal injection of 3FDGal (0.6 g/kg) at t=0, axial mCSSI-based 3D-FMI was acquired to simultaneously map 3FDGal, 3FDGalA, and 3FDGtol. Slices of 3 mm thickness and an in-plane resolution of 2 mm×2 mm were acquired. Top row: healthy mouse. FIG. 21B (middle row): mouse inoculated with 200,000 A549 (non-small cell lung cancer) cells expressing high levels of ALDH1A1 and AKR1B1, thus producing both 3FDGalA (shown as gold color) and 3FDGtol (shown as magenta) from the introduced 3FDGal in the tumor region. FIG. 21C (Bottom row): mouse inoculated with 200,000 DBTRG-05MG (glioma) cells expressing high levels of AKR1B1, thus producing 3FDGtol (shown as magenta) from the introduced 3FDGal in the tumor region. Data was acquired on a 15.2 T MRI scanner.

[0132] FIG. 22 shows two-dimensional Fluorine Metabolic Imaging (2D-FMI) of the brain following the injection of 3FDG. 3FDG (0.6 g/kg) was intraperitoneally administered to healthy C57BL/6 male mice, and coronal mCSSI-based 2D-FMI was performed repeatedly, starting at t=60 min post-injection to map 3FDG (FIG. 22A, green colored overlay on ¹H-MRI of the corresponding brain slice), 3FDGA (FIG. 22B, not shown in the brain), and 3FDS (FIG. 22C, red colored overlay on ¹H-MRI of the corresponding brain slice) simultaneously and noninvasively to visualize the real-time in vivo metabolism of 3FDG. The in vivo imaging data highlight the early, widespread uptake of the injected 3FDG, followed by metabolic reduction to 3FDGS, revealing their spatial distributions. Note that 3FDGA, the oxidized metabolite (i.e., 3FDGA) of 3FDG, was not found in the brain of healthy mice. MRI experiments were done on a 15.2 T MRI scanner with a 23 mm 1H/19F RF volume coil. A 20 mm single slice with in-plane resolution of 2 mm×2 mm was acquired at each 2D-FMI time point.

[0133] FIGS. 23A-23C shows a two-dimensional Fluorine Metabolic Imaging (2D-FMI) of the brain following the injection of 3FDGal. 3FDGal (0.6 g/kg) was intraperitoneally administered to healthy C57BL/6 male mice, and coronal mCSSI-based 2D-FMI was performed repeatedly, starting at t=60 min post-injection to map 3FDGal (FIG. 23A, green colored overlay on ¹H-MRI of the corresponding brain slice), 3FDGalA (FIG. 23B, gold colored overlay on ¹H-MRI of the corresponding brain slice), and 3FDGtol (FIG. 23C, magenta colored overlay on 1H-MRI of the corresponding brain slice) simultaneously and noninvasively to visualize the real-time in vivo metabolism of 3FDGal. The in vivo imaging data highlight the early, widespread uptake of the injected 3FDGal, followed by its disappearance and metabolic oxidation to 3FDGalA (first time point in the olfactory bulb ROI) and reduction to 3FDGtol, revealing their spatial distributions. MRI experiments were done on a 15.2 T MRI scanner with a 23 mm 1H/19F RF volume coil. A 20 mm single slice with in-plane resolution of 2 mm×2 mm was acquired at each 2D-FMI time point.

[0134] FIGS. 24A-24B show in vivo biodistribution of 3FDXyl metabolites in healthy mouse. FIG. 24A shows a chemical structures and ¹⁹F-NMR spectra of 3FDXyl and its redox derivatives (3FDXA, 3FDXtol). FIG. 24B shows a preliminary in vivo FMI after administration of 3FDXyl. The non-localized ¹⁹F-NMR acquired after probe injection shows the longitudinal transformation of 3FDXyl to its redox metabolites, and the FMI maps show the biodistribution of 3FDXtol (magenta) and 3FDXylA (gold).

[0135] FIG. 25 shows fFMI acquisition protocol improving both temporal resolution (reducing time acquisition from 80 to 20 minutes per step) and spatial resolution (decreasing voxel size from 2×2×3 mm to 1.5×1.5×3 mm, a ~44% reduction). These enhancements were achieved through bandwidth optimization: narrowing the receiver BW from 50 kHz to 10 kHz and reducing the excitation/refocusing BW from 2 kHz to 1 kHz. Data quantification (for LNCaP-CL1 tumor in the upper row and A549 tumor in the lower row) is based on a phantom tube containing known concentrations of 3FDGalA and 3FDGtol.

[0136] FIG. 26 is rate versus substrate concentration graph showing AKR1B1 enzyme kinetics with various sugar substrates.