A new positron emission tomography imaging agent for the serotonin transporter: synthesis, pharmacological characterization, and kinetic analysis of [11C]2-[2-(dimethylaminomethyl)phenylthio]-5-fluoromethylphenylamine ([11C]AFM).

The synthesis, radiolabeling, and in vitro and in vivo evaluation of a new positron emission tomography (PET) radioligand for the serotonin transporter (SERT), [(11)C]2-[2-(dimethylaminomethyl)phenylthio]-5-fluoromethylphenylamine ([(11)C]AFM) is reported. AFM was prepared from 4-chloro-3-nitrobenzyl acetate and thiosalicylic acid in a five-step synthetic sequence. In binding studies in vitro with cloned human transporters, AFM displayed high binding affinity (Ki 1.04 nmol/L for hSERT) and good selectivity (Ki 664 nmol/L for hNET and >10,000 nmol/L for hDAT) for SERT. The radiolabled compound [(11)C]AFM was prepared in 30-37 minutes from its monomethylamine precursor by reaction with high specific activity [(11)C]iodomethane. Radiochemical yield was 12.3 +/- 8.1% based on [(11)C]iodomethane and specific activity was 1733 +/- 428 Ci/mmol at end of synthesis (EOS, n = 14). Radiochemical and chemical purity of the final product was >97%. Biodistribution studies in rats indicated that [(11)C]AFM entered the brain readily and localized in regions known to contain high concentrations of SERT, with high specific to nonspecific binding ratios. Furthermore, binding of [(11)C]AFM in SERT-rich regions was blocked by the cold compound AFM and the selective serotonin reuptake inhibitor citalopram but not by the selective norepinephrine reuptake inhibitor nisoxetine or the selective dopamine reuptake inhibitor GBR 12935. At 30 minutes after injection, >95% of the brain activity corresponded to the parent compound, indicating the absence of radiolabeled metabolites in the rat brain. PET imaging experiments in baboons showed a brain distribution pattern of [(11)C]AFM consistent with the regional concentrations of SERT, with the highest levels of radioactivity detected in the midbrain and thalamus, moderate levels in the hippocampus and striatum, and the low levels in the cortical regions. Pretreatment of the baboons with citalopram (4 and 6 mg/kg, intravenously) reduced regional brain distribution volumes to low and homogeneous levels, thus underlining the binding specificity of [(11)C]AFM for SERT in vivo. Analysis of blood samples indicated a fast metabolism of the radioligand into more hydrophilic components, as well as the absence of radiolabeled lipophilic metabolites. Regional time-activity curves were analyzed with kinetic and graphical analysis methods using the arterial concentrations as input function. Both methods returned similar kinetic parameters and documented high specific to nonspecific equilibrium coefficients (V(3)") for [(11)C]AFM. Identical V(3)" values were also derived with the simple reference tissue method, indicating that quantification of SERT with [(11)C]AFM can be achieved without arterial blood sampling. In summary, [(11)C]AFM appears to be an excellent PET radioligand for the visualization and reliable quantification of SERT in vivo.