Pentafluorosulfanyl (SF5) as a Superior 19F Magnetic Resonance Reporter Group: Signal Detection and Biological Activity of Teriflunomide Derivatives

Fluorine (19F) magnetic resonance imaging (MRI) is severely limited by a low signal-to noise ratio (SNR), and tapping it for 19F drug detection in vivo still poses a significant challenge. However, it bears the potential for label-free theranostic imaging. Recently, we detected the fluorinated dihydroorotate dehydrogenase (DHODH) inhibitor teriflunomide (TF) noninvasively in an animal model of multiple sclerosis (MS) using 19F MR spectroscopy (MRS). In the present study, we probed distinct modifications to the CF3 group of TF to improve its SNR. This revealed SF5 as a superior alternative to the CF3 group. The value of the SF5 bioisostere as a 19F MRI reporter group within a biological or pharmacological context is by far underexplored. Here, we compared the biological and pharmacological activities of different TF derivatives and their 19F MR properties (chemical shift and relaxation times). The 19F MR SNR efficiency of three MRI methods revealed that SF5-substituted TF has the highest 19F MR SNR efficiency in combination with an ultrashort echo-time (UTE) MRI method. Chemical modifications did not reduce pharmacological or biological activity as shown in the in vitro dihydroorotate dehydrogenase enzyme and T cell proliferation assays. Instead, SF5-substituted TF showed an improved capacity to inhibit T cell proliferation, indicating better anti-inflammatory activity and its suitability as a viable bioisostere in this context. This study proposes SF5 as a novel superior 19F MR reporter group for the MS drug teriflunomide.


Dose response model, including outlier detection and IC 50 calculation for DHODH and CFSE data
The analysis of raw data from DHODH inhibition assays and CFSE cell proliferation assays was performed in R (version 3.6.1, R Foundation) using the drc R package for

Plate design and dilutions
For assessing the cytotoxicity of compounds, we performed a cytotoxicity test in HepG2 cells. Experiments were performed in 384 well plates (cellbind plate, Corning) in triplicates at seven concentrations ranging from 1.5 µM to 100 µM (DMSO concentration was 0.53 % in all wells). 5-FU (1 mM) was used as positive toxicity control and DMSO alone as negative control.

Cell culture
HepG2 cells were cultured in RPMI medium (10 % FCS) with 1800 cells per well.
Compounds were added on the following day, and plates were incubated for 72 h.

Assay
Following the incubation period, cells were fixed, stained, and measured using an automated widefield fluorescence microscope (Arrayscan XTI/Thermo). Hoechst 3342 (Sigma Aldrich, final concentration 1 µM, nuclear staining) and Yo-Pro-1 (Invitrogen, final concentration 0.1 µM, apoptosis assay) was used for live cell staining (1 h incubation). The live cell measurement microscopy was performed on two fluorescence channels at 20x magnification. The number of non-proliferating (Hoechst 3342 + ) and dead (Yo-Pro-1 + ) cells were quantified. Following fixation, viable cells were counted by microscopy, using 10x magnification.

Comparison of growth inhibition and toxicology
We performed a growth inhibition assay and a cytotoxicity screening of TF and its derivatives in HepG2 cells to study their impact on liver cells as drug entrance and metabolic compartment. All compounds showed inhibitory activity in the concentration range of 1.5 µM to 100 µM. Cytotoxicity was assessed by quantifying damaged cells by apoptosis markers. TF did not show any cytotoxic effect in this concentration range compared to untreated negative controls. Equally, CF 3 O-TF did not display cytotoxic effects up to 100 µM. di-CF 3 -TF showed no cytotoxicity up to    Figure S7. 19 F MR spectroscopy of a phantom containing TF dissolved in DMSO.

F MR spectral line splitting of SF 5 -TF in DMSO
The doublet at 67.4 ppm is shown separately from the pentet at 91.7 ppm.

Biological and 19 F MR reporter activity of TF and SF 5 -TF
We administered TF and SF 5 -TF orally to C57BL/6 mice (n = 6) using the same molar concentrations: TF =12.15 mg/ml (n = 3) and SF 5 -TF = 10 mg/ml (n = 3). 19 F MR UTE MR images were then acquired using optimized parameters. Peak 19 F SNR values for each compound were calculated for all 6 mouse stomachs (Table S4).
These values were calculated from the 19 F MR images of TF and SF 5 -TF ( Figure   S8).   Table S4. SNR is indicated by the color bars.