Comparison of three novel radiotracers for GluN2B-containing NMDA receptors in non-human primates: (R)-[11C]NR2B-Me, (R)-[18F]of-Me-NB1, and (S)-[18F]of-NB1

Abstract

The NMDA receptor GluN2B subunit is a target of interest in neuropsychiatric disorders but to date there is no selective radiotracer available to quantify its availability in vivo. Here we report direct comparisons in non-human primates of three GluN2B-targeting radioligands: (R)-[¹¹C]NR2B-Me, (R)-[¹⁸F]OF-Me-NB1, and (S)-[¹⁸F]OF-NB1. Plasma free fraction, metabolism, tissue distribution and kinetics, and quantitative kinetic modeling methods and parameters were evaluated in two adult rhesus macaques. Free fraction in plasma was <2% for (R)-[¹¹C]NR2B-Me and (R)-[¹⁸F]OF-Me-NB1 and higher for (S)-[¹⁸F]OF-NB1 (15%). All radiotracers showed good brain uptake and distribution throughout grey matter, with substantial (>68%) blockade across the brain by the GluN2B-targeting drug Co-101,244 (0.25 mg/kg), including in the cerebellum. Time-activity curves were well-fitted by the one-tissue compartment model, with volume of distribution values of 20-40 mL/cm³ for (R)-[¹¹C]NR2B-Me, 8-16 mL/cm³ for (R)-[¹⁸F]OF-Me-NB1, and 15-35 mL/cm³ for (S)-[¹⁸F]OF-NB1. Estimates of regional non-displaceable binding potential were in the range of 2-3 for (R)-[¹¹C]NR2B-Me and (S)-[¹⁸F]-OF-NB1, and 0.5-1 for (R)-[¹⁸F]OF-Me-NB1. Altogether, each radiotracer showed an acceptable profile for quantitative imaging of GluN2B. (S)-[¹⁸F]OF-NB1 has particularly promising imaging characteristics for potential translation into humans. However, the source of unexpected displaceable binding in the cerebellum for each of these compounds requires further investigation.

Keywords: (R)-[11C]NR2B-Me; (R)-[18F]OF-Me-NB1; (S)-[18F]OF-NB1; GluN2B subunit; NMDA receptor; Positron emission tomography; glutamate.

Figure 1.

Structural diagrams and brain distribution in non-human primate of GluN2B-targeting radiotracers. Top, chemical structures of (R)-[¹¹C]NR2B-Me, (R)-[¹⁸F]OF-Me-NB1, and (S)-[¹⁸F]OF-NB1. Middle and bottom rows, representative mean PET SUV images in template space from 60–90 minutes at baseline or following pre-treatment with the GluN2B antagonist Co-101,244 (0.25 mg/kg).

Figure 2.

Time-activity curves and kinetic modeling results. a, example regional time-activity curves for each radiotracer with 1TCM fits. b, comparison of V_T values derived from 2TCM vs. 1TCM (top row) and MA1 vs. 1TCM (bottom row).

Figure 3.

Blocking and displacement studies. a, Occupancy plots showing reduction in specific binding of GluN2B-targeting radiotracers following pre-treatment with Co-101,244 (0.25 mg/kg) relative to baseline in each animal. b, Time-activity curve showing bolus plus constant infusion study of (R)-[¹⁸F]OF-Me-NB1 with sequential administration of Co-101,244 (0.75 mg/kg, +75 min) and FTC-146 (0.125 mg/kg, +225 min). Reduced signal in tissue is seen following administration of Co-101,244 but no further displacement is apparent following the addition of FTC-146. The fitted input function is shown in red.

Figure 4.

Guo plots comparing V_T across ROIs during baseline scans for two animals. In these plots, y-intercept > 0 indicates that the tracer on the x-axis is expected to have higher binding potential values. Thus, BP_ND values follow the rank order (R)-[¹¹C]NR2B-Me > (S)-[¹⁸F]OF-NB1 > (R)-[¹⁸F]OF-Me-NB1.