Translational Preclinical PET Imaging and Metabolic Evaluation of a New Cannabinoid 2 Receptor (CB2R) Radioligand, (Z)- N-(3-(2-(2-[18F]Fluoroethoxy)ethyl)-4,5-dimethylthiazol-2(3 H)-ylidene)-2,2,3,3-tetramethylcyclopropane-1-carboxamide

Abstract

We have previously developed seven fluorinated analogues of A-836339 as new PET tracers for cannabinoid type 2 receptor (CB₂R) imaging, among which (Z)-N-(3-(2-(2-[¹⁸F]fluoroethoxy)ethyl)-4,5-dimethylthiazol-2(3H)-ylidene)-2,2,3,3-tetramethylcyclopropane-1-carboxamide ([¹⁸F]FC0324) displayed high affinity and selectivity for CB₂R in healthy rats. In the present study, we have further evaluated the imaging and metabolic properties of [¹⁸F]FC0324 in a rat model of human CB₂R overexpression in the brain (AAV-hCB₂) and in non-human primates (NHPs). Autoradiography with AAV-hCB₂ rat brain sections exhibited a signal of [¹⁸F]FC0324 8-fold higher in the ipsilateral region than in the contralateral region. Blocking with NE40, a CB₂R-specific agonist, resulted in a 91% decrease in the radioactivity. PET experiments showed a signal 7-fold higher in the ipsilateral region, and the specificity of [¹⁸F]FC0324 for hCB₂R in vivo was confirmed by the 80% decrease after blocking with NE40. In NHPs, brain time-activity curves displayed a fast and homogeneous distribution followed by a rapid washout, in accordance with the low amount of CB₂Rs in healthy brain. Whole-body PET-CT suggested a high and specific uptake of the radiotracer in the spleen, a CB₂R-rich organ, and in the organs involved in metabolism and excretion, with a low bone uptake. In vitro metabolism with monkey liver microsomes (MLMs) led to the formation of six main hydroxylated metabolites of FC0324. Five of them were produced by human liver microsomes, being much less active than MLMs. In vivo, in NHPs, the main radiometabolite was likely to result from further oxidation of hydroxylated compounds, and parent [¹⁸F]FC0324 accounted for 8 ± 3% of plasma radioactivity (at 120 min p.i.) with a low level of potential interfering radiometabolites. Furthermore, this metabolism should be significantly reduced in humans due to species differences. In conclusion, [¹⁸F]FC0324 appears to be a promising candidate for further human studies with suitable kinetics, selectivity, and metabolic profile for CB₂R PET imaging.

Figure 1

Autoradiography images of AAV-hCB2 rat brain sections incubated with [¹⁸F]FC0324. (A) Brain section of a rat injected with the AAV virus expressing the human CB₂ receptors in the right striatum, (B) blocking with CB₂ agonist NE40 (100 μM), and (C) quantitative comparison (n = 2 for control AAV-hCB₂ and n = 2 for blocking conditions) of the tracer signal in the ipsilateral and contralateral regions of each brain section. Data are expressed as digital light units per square millimeter (DLU/mm²).

Figure 2

Brain PET imaging in AAV-hCB₂ rats injected with [¹⁸F]FC0324 (33 MBq): (A) summed image (0-120 min) of [¹⁸F]FC0324 accumulation in control AAV-hCB₂ rat brain (baseline) and (B) time–activity curves (0-120 min) of [¹⁸F]FC0324 accumulation in the right and left striata of AAV-hCB₂ rats without (baseline) and with blocking with NE40 (1 and 10 mg/kg).

Figure 3

Representative profile of the MLM incubation extract (30 min, 37 °C): (A) full scan MS (with m/z values) and (B) UPLC-UV (RT: retention time) and (C) percentages of parent FC0324 and its metabolites following a 60 min incubation with MLMs or HLMs (values were expressed as mean % ± SD, n = 3 independent experiments performed in triplicates).

Figure 4

Representative radio-HPLC profile of acetonitrile extracts of NHP plasma at 5, 10, and 90 min p.i. in baseline conditions.

Figure 5

(A) Mean time-course of parent [¹⁸F]FC0324 in plasma under baseline conditions (n = 8 independent experiments with 4 NHPs) and (B) mean plasma metabolite-corrected arterial input functions (baseline conditions, n = 8 independent experiments with 4 NHPs) expressed as SUVs.

Figure 6

Brain PET imaging study in NHPs: representative PET axial slice [averaged from 0 to 20 min (left) and from 40 to 90 min (right)], expressed as SUVs and overlaid with its own t₁-weighted MRI scan.

Figure 7

(A) Time–activity curve of [¹⁸F]FC0324 in the whole brain of NHPs (baseline n = 8 experiments). Data were expressed as mean SUV ± SD and (B) regional AUC_0–120 min measured in the brain of NHPs (baseline n = 8 experiments).

Figure 8

(A) Time–activity curve of [¹⁸F]FC0324 in the spleen, left myocardium, whole brain, and lung from the control (straight line) and blocking (dashed lines) experiments of the same animal (n = 1). Data are expressed in SUVs. (B) Tissue-to-blood ratios (left ventricle) in control and saturation with NE40 experiments for several organs (n = 1).