Radiosynthesis, In Vitro and In Vivo Evaluation of a 6-(2-[18F]Fluoroethoxy)-7-methoxy-pyrrolidinylquinazoline for PET Imaging of Phosphodiesterase 10A in Brain

1. Introduction

Phosphodiesterase 10A (PDE10A) is a key enzyme that mediates neural signal transduction by regulating intracellular concentration of the cyclic nucleotides adenosine (cAMP) and guanosine monophosphate (cGMP) and is mainly present in brain. Particularly high PDE10A expression and activity was observed in brain regions of dopaminergic and GABAergic neurotransmission. As these relations are not fully understood and PDE10A hypofunction is supposed to correlate with neuropsychiatric disorders such as schizophrenia, obsessive-compulsive disorders, Parkinson’s disease, and Huntington’s disease, there is an ongoing interest in PDE10A as target for molecular imaging by PET.
For the pharmacological treatment of schizophrenia, a highly PDE10A inhibiting, selective and brain penetrable 6,7-dimethoxy-4-pyrrolidinylquinazoline has been designed (Ki=4 nM [1]). Based on this structure we developed fluorine-18 labeled derivatives as potential PET radiotracers for imaging PDE10A in brain [2]. Herein we report on the radiosynthesis and radiotracer properties of the 6-[18F]fluoroethoxy derivative [18F]2 (Figure 1).

2. Materials & Methods

The non-radioactive reference compound 2 and the 7-tosyloxy precursor 1 were prepared in multi-step syntheses, and 2 was screened for its PDE10A inhibitory potency (Ki=32 nM) as well as selectivity in enzyme activity studies [2]. The automated radiosynthesis of [18F]2 was carried out on a TRACERlabTM FX F-N synthesizer. Aqueous n.c.a. [18F]Fluoride was transferred to MeCN via a Chromafix® 30-PS-HCO3 cartridge, dried azeotropically and converted to its K[18F]F-K2.2.2-carbonate complex. Nucleophilic 18F-for-OTs substitution was performed with ~2 mg of 1 in 750 µL MeCN at 85°C within 15 min. Purification of [18F]2 was carried out by SPE (SepPak®Plus C-18, MeCN), followed by semi-preparative RP-HPLC (e.g. on ReproSil-Pur® C18 AQ, 7 µm; 5010 + 15010 mm, 42% MeCN, 20 mM NH4OAc, 4 mL/min; tR=30.3 min). Formulation of [18F]2 was done by SPE, removal of organic eluent and dissolution in physiological saline.
Radioligand stability (EtOH, MeCN, physiological saline and phosphate buffer solution, 40 and 80°C, 5-120 min) and lipophilicity (logD7.0-7.4, shake-flask) were determined. Further characterization in vitro of [18F]2 included the determination of stability in rat plasma (37°C, 30 and 60 min), PDE10A affinity (KD, PDE10A transfected SF21 cells) and autoradiographic imaging of sagittal female rat brain slices, incubated with [18F]2 alone, together with 2 or highly PDE10A-specific MP-10, respectively. Evaluation in vivo of [18F]2 in female CD-1 mice was carried out by the determination of biodistribution and brain uptake as well as metabolism studies and ex vivo brain autoradiography, with validation of specificity by pre-treatment with MP-10 (1mg/kg at 15 min before radiotracer).

Figure 1. Radiosynthesis of [18F]2 and its binding at a sagittal rat brain slice in vitro.

3. Results

Radiosynthesis of [18F]2 resulted in labeling efficiencies of 76-94%, a radiochemical yield of 41.210.3% (n=6, 2 h, based on [18F]F-), a radiochemical purity of ≥99% and specific activities of 80-1030 GBq/mol. [18F]2 remained stable during heating in organic solvents (97% of [18F]2, 120 min, 80°C), and showed moderate stability in aqueous buffer solutions (95% of [18F]2, 60 min, 40°C) and rat plasma (94% of [18F]2, 60 min, 37°C). A logD7.0-7.4 of ~2.5, was determined, and by homologous competition a KD of 24 nM (n=2) was estimated. By autoradiography in vitro, a heterogeneous distribution of [18F]2 in rat brain was observed (Figure 1), which was partially inhibited by MP-10. Biodistribution studies revealed an initial brain uptake of 1.6%ID/g at 5 min p.i.. Striatal uptake at 60 min p.i. was not inhibited by MP-10, which was confirmed by ex vivo autoradiography. Only 41%, 64% and 20% of the radioactivity measured in plasma, brain and liver, respectively, at 30 min p.i. corresponded to parent radioligand. An evidence for slight defluorination was observed because in femur uptake values increased by 25% at 60 min p.i. after removal of the bone marrow.

4. Discussion/Conclusion

A convenient radiosynthesis and satisfactory radiochemical results, as well as moderate lipophilicity and PDE10A affinity initially indicated [18F]2 to be a suitable radiotracer. However, the heterogeneous but non-displaceable binding of [18F]2 in vitro provides evidence for binding affinity to another target. Reasonable high uptake of [18F]2 in mice brain was found, which was non-specific and non-selective. The radioligand stability in vivo is comparably low and the presence of brain metabolites is inappropriate for molecular imaging.
In conclusion, a structural revision of our pharmaceutical lead is needed to improve the in vivo properties and to develop an applicable radiotracer for neuroimaging of PDE10A with PET.

Research Support: European Regional Development Fund (ERDF).

References: [1] Chappie TA, Humphrey JM, Allen MP et al. [2007] J.Med.Chem. 50: 182-185.
[2] Nieber K, Erdmann S, Briel D et al. [2010] Patent Appl. P1014WO.