|Home | About | Journals | Submit | Contact Us | Français|
An investigation was conducted to determine if the 5-fluoro analog of TPI (5-chloro-6-[(2-iminopyrrolidin-1-yl)methyl]uracil), a potent inhibitor of human thymidine phosphorylase (TP), has an IC50 in a range that might allow to use it labeled for imaging of TP expression in vivo. The previously unreported fluoro analog, TPI(F), was prepared and tested against TPI and TPI(Br) using an inhibition assay of [H-3]thymidine cleavage. An assay, performed in the presence of 0.4 mg/ml of human TP, yielded IC50 values of 2.5 nM, 2.7 nM and 9.0 nM for TPI, TPI(Br) and TPI(F), respectively. The results indicate that further studies to develop 18F-labeled TPI(F) as a potential radiopharmaceutical for PET imaging of TP expression in vivo are warranted.
Thymidine phosphorylase (TP) is an essential enzyme involved in endogenous nucleotide salvage [1,2]. It specifically cleaves the glycosidic bond in thymidine to produce thymine and 2-deoxyribose-1α-phosphate (2dR-1P) (SCHEME 1). The reaction is non-energy dependent, reversible under physiologic conditions and is strongly coupled to thymine degradation. As a result, TP generally functions to clear thymidine from blood, which keeps plasma concentrations low. However, there is growing interest in the potential role TP plays in tumor biology since it was discovered that 2dR-1P (or 2dR) promotes neovacularization, in vitro and in experimental tumor models [3–8].
Our research group’s aim is to investigate tumor biology, particularly lung cancer, using PET imaging and to actively develop novel agents for that purpose. Recently, we reported a radiosynthesis and in vitro characterization of 5′-deoxy-5′-[18F]fluorothymidine (DFT, Figure 1) as a targeted radiopharmaceutical for imaging TP in vivo . This fluorinated thymidine analog is a TP substrate and was actively metabolized to 2,5-dideoxy-5-[18F]fluororibose-1α-phosphate (ddFR-1P) in cells. We hypothesized that labeled ddFR-1P would be trapped in cells due to its phosphate charge and, consequently, accumulate in proportion to intracellular TP activity. Unfortunately, ddFR-1P proved too unstable to be useful as a PET imaging agent.
The instability of ddFR-1P prompted us to consider an alternative strategy that directly targeted TP with a potent 18F-labeled enzyme inhibitor. That goal required a 18F-labeled compound with exceptional binding potency, since the imaging agent and target would have a 1:1 stoichiometry and there would be no metabolic turnover to feed a labeled metabolite pool. Fortunately, a TP inhibitor, TPI (5-chloro-6-[(2-iminopyrrolidin-1-yl)methyl]uracil) (Figure 1), has been previously shown to have an IC50 of 35 nM . Moreover, TPI is a weak inhibitor of the related enzyme uridine phosphorylase (IC50 > 1000 nM) .
The high TPI inhibitory binding with TP is similar to that found in other receptor-binding 18F-labeled molecules that have been successfully imaged in vivo [11,12], so a fluorinated analog was of interest. Replacement of the chlorine atom in TPI with a fluorine atom was obvious, but the critical question was whether that analog retained the high inhibitory potency of TPI. Reported herein is a preliminary investigation that focused on the synthesis of 5-fluoro-6-[(2-iminopyrrolidin-1-yl)methyl]uracil (TFI(F)) and determination of its IC50 value with TP. The goal was to determine if TPI(F) had adequate potency to justify its fluorine-18 labeling and evaluation as an imaging agent of TP expression in vivo.
TPI(F) was prepared by alkylating 2-iminopyrrolidine with 6-chloromethyl-5-fluorouracil [13, 14] in methanol, promoted by 1,8-diazabicyclo[5.4.0]undec-7-ene . TPI and TPI(Br) were also prepared to directly compare the relative 50% inhibitory values (IC50) with human TP. TPI(F) was an obvious choice for a fluorinated TPI analog, given that a fluorine atom represents minimal structural modification of the parent drug. A recent synthesis and evaluation of 5-fluoro-6-[(1H-imidazol-1-yl)methyl]uracil (TPI(F)-(IM), Figure 1)) supports that view [13, 14]. However, in that case, the additional impact of the imidazole ring, as the pendent amine moiety, was unclear.
The relative potencies (IC50 values) for TPI and TPI(Br) and TPI(F) were determined to be 2.5 nM and 2.7nM and 9.0 nM, respectively. The inhibition curves are shown in Figure 2. However, our values for TPI and TPI(Br) are lower than previously reported . This may be attributed to the specific enzyme concentrations used in the respective enzyme assays. Regardless, the trend for TPI, TPI(Br) and TPI(F) is clear. Fluorination, alone, does not improve potency. Nevertheless, the potency of TPI(F) still falls within the range that is typical of many PET radiotracer ligands. Thus, radiolabeling of TPI(F) with high specific activity [18F]fluoride ion (1–2 Ci/μmol) and evaluation of its potential for imaging regional TP expression in vivo is warranted.
All chemicals were purchased from Sigma-Aldrich and used as received. NMR spectra: 1H (300 MHz, δ, TMS); 13C (75 MHz, δ, TMS); 19F (282 MHz, δ, CCl3F), were recorded using a Bruker AV301 multi-nuclear instrument. Mass spectrometry (MS) was performed using a Micromass Quattro Premier XE instrument and electrospray ionization (ES). Characteristic data is given for selected synthetic intermediates and products, which were not available.
Thymidine phosphorylase activity was assayed in vitro using a modification of previously described methods . 100 μl aliquots of 0.4 mg/ml human recombinant thymidine phosphorylase (Sigma-Aldrich Catalog No. T9319) in 0.5% bovine serum albumin were placed on ice. At time zero minutes the reaction was initiated by addition of 300 μl of assay buffer (4°C) containing 76 mM sodium phosphate, pH 7.4 and 1.6 mM [3H-5′]-thymidine (Moravek Radiochemicals Catalog No. MT-846W) plus or minus the appropriate concentrations of TPI compounds, serially diluted in double distilled water. The mixture was vortexed and placed in a 37°C gently shaking water bath. Time zero minute controls were terminated immediately without being removed from ice. After 30 min at 37°C, reactions were terminated by the addition of 400 μl of a 4°C slurry of 5 % charcoal (Sigma-Aldrich Catalog No. 242276, “Darco G-60”) in 10 % trichloroacetic acid. The terminated mixture was centrifuged (10,000 × g/10 minutes/room temperature) and 0.4 ml of the supernate was counted in 5 ml of EcoScint A (National Diagnostics) in a Beckman LS5000C liquid scintillation counter. For an individual experiment values for each condition were determined from the mean of two separate determinations after subtracting the time zero control values. IC50 values were extrapolated from the averaged data from three separate experiments and represent the concentration of inhibitor that inhibited TP activity by 50 %.