Design, Synthesis, and Characterization of [¹⁸F]mG2P026 as a High-Contrast PET Imaging Ligand for Metabotropic Glutamate Receptor 2

Abstract

An array of triazolopyridines based on JNJ-46356479 (6) were synthesized as potential positron emission tomography radiotracers for metabotropic glutamate receptor 2 (mGluR2). The selected candidates 8–10 featured enhanced positive allosteric modulator (PAM) activity (20-fold max.) and mGluR2 agonist activity (25-fold max.) compared to compound 6 in the cAMP GloSensor assays. Radiolabeling of compounds 8 and 9 (mG2P026) was achieved via Cu-mediated radiofluorination with satisfactory radiochemical yield, >5% (non-decay-corrected); high molar activity, >180 GBq/μmol; and excellent radiochemical purity, >98%. Preliminary characterization of [¹⁸F]8 and [¹⁸F]9 in rats confirmed their excellent brain permeability and binding kinetics. Further evaluation of [¹⁸F]9 in a non-human primate confirmed its superior brain heterogeneity in mapping mGluR2 and higher affinity than [¹⁸F]6. Pretreatment with different classes of PAMs in rats and a primate led to similarly enhanced brain uptake of [¹⁸F]9. As a selective ligand, [¹⁸F]9 has the potential to be developed for translational studies.

Graphical Abstract

INTRODUCTION

As the major excitatory neurotransmitter in the central nervous system (CNS) of vertebrates, glutamate is involved in numerous physiological and behavioral processes.^1,2 It modulates synaptic responses by activating two different classes of receptors, the ionotropic glutamate receptors (iGluRs) and the G-protein coupled metabotropic glutamate receptors (mGluRs).³ Among the mGluRs, the metabotropic glutamate receptor 2 (mGluR2) is highly enriched in the forebrain and predominantly localized on the presynaptic nerve terminals.^4,5 Activation of mGluR2 reduces the glutamate level and GABA release; therefore, it is hypothesized as a therapeutic intervention toward neurological diseases that involve increased glutamate transmission.^6,7 Activation of mGluR2 has shown efficacy in multiple preclinical animal models of schizophrenia,^8–12 anxiety,⁷ and depression.¹³ LY2140023 (1), the prodrug of mGluR2/3 agonist LY404039, achieved clinical validation in phase II clinical trials in schizophrenic human subjects (Figure 1).¹¹ However, it was discontinued due to the lack of efficacy in the following three phase II or phase III clinical trials.^14,15 Subsequent studies suggested mGluR2 but not mGluR3 mediated the antipsychotic effect of the mGluR2/3 agonist.^16,17

Figure 1.

Structures of typical mGluR2 PAMs and PET radioligands.

To improve the selectivity of mGluR2 binding over mGluR3, recent mGluR2-based drug discovery is leaning toward the development of agents that bind to the allosteric sites within the hydrophobic seven-transmembrane region (7-TM) instead of the highly conserved orthosteric glutamate Venus flytrap domain.¹⁸ Besides enhanced binding selectivity, these ligands feature enhanced brain permeability and reduced liability of receptor desensitization than the amino acid-based orthosteric ligands.^19–21 Numerous structurally diverse mGluR2 positive allosteric modulators (PAMs) have been reported, for example, BINA (2),²² AZD8529 (3),²³ and JNJ-40411813²⁴ (4, Figure 1). Compound 3 entered phase II clinical trials for schizophrenic patients in 2009 but was halted in 2011 because of lack of efficacy.²⁵ Compound 4 advanced to a phase IIa study in schizophrenia, but its efficacy was not ideal.^26,27 Scaffold-hopping of the pyridone core of compound 4 led to a series of 1,2,4-triazolopyridines as potent and selective mGluR2 PAMs with improved drug properties.^28–30 In this class of PAMs, [¹¹C]JNJ42491293 ([¹¹C]5, Figure 1) was the first disclosed positron emission tomography (PET) radioligand that was characterized in humans.³¹ Despite the satisfactory in vitro, ex vivo, and in vivo imaging results in rats, [¹¹C]5 showed apparent off-target binding in human myocardium, precluding it as an useful imaging tool for mGluR2.³¹ As a non-invasive imaging technique, PET would enable quantification of the biodistribution, expression, and modulation of a protein target under normal and disease conditions. We have therefore developed [¹⁸F]JNJ-46356479 ([¹⁸F]6)^32,33 and [¹¹C]mG2P001 ([¹¹C]7)^34,35 as alternative PET imaging tools (Figure 1). Compound 6 belongs to 1,2,4-triazolopyridines²⁸ and has been employed as an mGluR2-selective PAM to study the binding selectivity of [¹¹C]5.³¹ Our studies have confirmed that both [¹⁸F]6 and [¹¹C]7 are suitable ligands for imaging mGluR2 in the brain of rodents and non-human primates.^32,34 Meanwhile, we also found apparent white matter binding of [¹⁸F]6 at the later stage of the dynamic scan (60–120 min) in the primate studies.³² Although mGluR2 is known to express in fiber bundles of white matter,³⁶ we suspect part of the binding was off target.³²

To enhance the in vivo binding specificity of [¹⁸F]6, we intend to improve its binding affinity and optimize its physicochemical properties. Although a vast chemical space of 1,2,4-triazolopyridines has been covered by patent applications and research articles,^28–30,37 their use as PET imaging ligands has not been explored. Herein, we focused on four close analogues of compound 6 to fine-tune its binding and physicochemical properties in compounds 8–11 (Scheme 1). Moreover, we explored the potential of modifying the core scaffold of 6 in producing new classes of mGluR2 PAMs as described for compounds 12 and 13 (Scheme 1). In addition to the mGluR2 functional potency, the synthesized ligands were also characterized for their binding toward other mGluRs, especially their agonist activity against mGluR2/3. Furthermore, molecular docking was used to probe the binding interactions at the allosteric site. Once labeled with fluorine-18, radioligands [¹⁸F]8 and [¹⁸F]9 were evaluated by PET imaging studies using rats and/or a non-human primate to examine their feasibility as an in vivo imaging tool for mGluR2.

Scheme 1. Syntheses of Compounds 8–13^a.

^aReagents and conditions: (a) TEA, MgSO₄, NaBH(OAc)₃, rt, 12 h; (b) NH₂NH₂, 1,4-dioxane, 80 °C, 16 h; (c) CyPrCOCl, Et₃N, CH₂Cl₂, rt, 16 h; (d) POCl₃, 1,2-dichloroethane, 150 °C, 10 min, microwave; I 4,4,5,5-tetramethyl-2-vinyl-1,3,2-dioxoborolane, Pd(PPh₃)₃, 1,4-dioxane, sodium bicarbonate, 150 °C, 15 min, microwave; (f) NaIO₄, OsO₄, 1,4-dioxane/H₂O, rt, 2 h; (g) 1-(2,4-difluorophenyl)piperazine, TEA, MgSO₄, NaBh(OAc)₃, rt, 12 h; (h) EtOH, reflux, 16 h; (i) POCl₃, reflux, 16 h; (j) (i) NaI, acetyl chloride, CH₃CN, 50 °C, 4 h; (ii) CuI, methyl 2,2-difluoro-2-(fluorosulfonyl)acetate, DMF, 100 °C, 6 h; (k) LiOH, THF/H₂O, rt, 12 h; and (l) 1-(2,4-difluorophenyl)piperazine, HATU, DIPEA, DMF, rt, 16 h.

RESULTS AND DISCUSSION

Chemistry.

We aimed to synthesize PET radioligands with favorable physicochemical properties, enhanced binding profiles, and ease to incorporate fluorine-18. As Scheme 1 shows, the structural variations based on 6 were focused on (1) replacing the piperazine linker with the more basic piperidine in compound 8, (2) fine-tuning the fluorine substituents at the tail arene of compounds 9–11, and (3) modifying the core scaffold to give compounds 12–13. Compound 12 is a region-analogue of 6, where the positions of aryl-CF₃ and cyclopropyl methylene groups on 1,2,4-triazolopyridine were changed. Compound 13 changes to a pyrazolopyrimidine analogue by altering the nitrogen atoms and substituents on the 1,2,4-triazolopyridine core. The amide group of 13 was not further reduced to maintain a lower lipophilicity (cLogP = 3.89, Table 1).

Table 1.

In Vitro Properties of Compound 6 and Its Analogues 8–13

comp.	MW (g/mol)	cLogPa	LogD7.4	mGluR2 PAM EC50 (nM)b	mGluR2 agonist EC50 (nM)b	mGluR3 agonist EC50 (nM)b	other mGluRs EC5 (nM)c
6	451.45	4.74	3.20	166	2089	9082	>8000
8	450.46	5.21	3.11	20	136	3025	>10,000
9	467.90	5.32	3.38	11	84	3757	>10,000
10	467.90	5.32	4.01	8.3	82	4162	>10,000
11	433.45	4.54	2.84	NDd	371	1542	>10,000
12	451.45	4.74	3.71	913	2353	3948	>10,000
13	465.43	3.89	3.09	7,317	7278	7131	>1000

^aCalculated using ChemDraw 16.0.

^bValues are the mean of at least two experiments.

^cSee the Supporting Information.

^dNot determined.

Syntheses of compounds 8–11 were achieved under the same reductive amination conditions as those of compound 6.^28,32 The coupling reactions occurred between aldehyde 14 and the corresponding amines 15–18 as shown in Scheme 1. Compound 14 was prepared via the literature method.²⁸ Compound 12 was obtained in a similar manner as 6 but using 2,3-dichloro-5-(trifluoromethyl)pyridine (19) as a starting material. 8-Chlorotriazolopyridine 22 was obtained from 19 following sequential reactions of chlorine displacement, acylation, and cyclodehydration. Aldehyde intermediate 24 was prepared from 22 via a Suzuki coupling to incorporate the vinyl group and a subsequent oxidative reaction with osmium tetroxide. A final reductive amination between 24 and 1-(2,4-difluorophenyl)piperazine led to compound 12. Synthesis of compound 13 began with the condensation reaction between compounds 25 and 26 to get compound 27.³⁸ Compound 27 was then subjected to oxidative cyclization, halogen exchange with the CF₃ source, and hydrolysis to give intermediate 30. The final coupling reaction between 30 and 1-(2,4-difluorophenyl)piperazine led to compound 13.

In Vitro Properties.

Compounds 8–13 were characterized for their functional potency as mGluR2 PAMs and their selectivity against mGluR1–6 and 8 following our previous methods.^32,34 The mGluR2 PAM assay evaluates G_i/o G-protein-induced changes in the intracellular cAMP concentration using the cAMP GloSensor assay kit (Promega) with the mGluR2 stably expressing cells (HEK 293 derived Flp-In t REx-293 cells). In a preliminary PAM assay, where EC₂₀ l-glutamate (10 μM) was added to a serial dilution of the experimental compound, compounds 8–10 showed enhanced PAM activity (EC₅₀s, 8–20 nM) than compound 6 (EC₅₀ = 166 nM, Figure 2A and Table 1). However, the region-analogue 12 showed significantly decreased mGluR2 PAM activity (EC₅₀ = 913 nM), and pyrazolopyrimidine 13 failed to retain this functional potency (EC₅₀ = 7317 nM). We then tested compounds 8–10 and the monoaryl fluoride 11 in the binding cooperativity assay with l-glutamate, where subagonist doses of compounds 8–10 and agonist doses of compound 11 up to 300 μM were added to serial dilutions of l-glutamate to further assess the PAM activity of these compounds on mGluR2. Data were fitted according to the Black Leff Ehlert equation using Prism 9.0. In this assay, the binding cooperativity factor, alpha, and efficacy cooperativity factor, beta, provide estimates of the effect of the allosteric modulator on binding and efficacy, respectively, of the orthosteric agonist.³⁹ These can be expressed together as α*β or composite cooperativity. From our assay, compounds 8–10 are agonists with possibly a small amount of modulation that is difficult to fit and, thus, quantify. For compound 11, allosteric modeling calculations of alpha*beta showed an increase in sensitivity (Supporting Information Figure S1). Moreover, in these assays, compounds 6–11 did not exhibit non-specific activity in the untransfected HEK293T cells at or below 10 μM in the presence of 0.1 μM isoproterenol (Supporting Information, Figure S2).

Figure 2.

Pharmacological properties of compounds 8–13. (A) Dose–response curves of compounds 8–10, 12–13 as mGluR2 PAMs (in the presence of EC₂₀ amount of l-glutamate (10 μM)); (B) functional profiles of compounds 8–13 in agonist mode for mGluR2 in the absence of l-glutamate and (C) mGluR3 in the absence of l-glutamate. The relative luminescence units in the Y axis in A, B, and C were normalized to their maximum response values. Data are presented as the mean of three independent assays performed in triplicate, with error bars representing mean ± S.E.M. Graphs were made in Prism 9.0.

In addition, by screening their mGluR2/3 agonist activity, compounds 8–11 showed an enhanced mGluR2 agonist activity, ranging from 82 to 371 nM. Therefore, these compounds not only increase the potency and efficacy of glutamate but also induce receptor activation by themselves, making them mGluR2 ago-PAMs. They also had a greater selectivity for mGluR2 over mGluR3 than compound 6 (Figure 2B,C and Table 1). Besides, compounds 8–13 showed no obvious functional interactions with other mGluRs (Supporting Information, Table S1). Additionally, although compounds 6 and 8–13 have relatively high cLogP values, their experimentally determined LogD_7.4 values were in the range of 1.0–4.0 for CNS penetrants (Table 1).^40–42

As reported in the literature, there are probe-dependent effects for different classes of mGluR2 PAMs regarding their modes of action.^43–45 Moreover, it was found that an mGluR2 agonist could affect the binding event of a tritiated PAM radioligand by increasing the radioligand binding affinity toward mGluR2^45,46 and/or increasing the mGluR2 receptor density⁴⁷ in vitro. As described in our previous work, compound 7 showed no apparent mGluR2/3 agonist activity (EC₅₀ > 10 μM) in the same assays.³⁴ Therefore, the added feature of potent mGluR2 agonist activity in compounds 8–11 might result in unique features in characterizing the PET radioligands in vivo.^43,44

Structural Insights of Compounds 8–11.

Compounds 8–11 were docked into our previously described mGluR2 homology model,³⁴ and their ligand–protein binding interactions were compared with that of compound 6. As shown in Figure 3A, compounds 6 has its heterocyclic core buried at the bottom of the allosteric binding site, which is surrounded by interacting residues of F623, F643, Y647, L732, R735, V736, L739, W773, and V798. The 1,2,4-triazolopyridine core of 6 forms π–π stacking interactions with residues of F623, F643, and W773. Compounds 8–11 occupy the same binding site and adopt similar binding poses as that of compound 6. An overlaid binding image of 6 and 8–11 is shown in Figure 3B. The docking scores for compounds 6 and 8–11 were more than 10.0 kcal/mol, indicating their potent binding toward mGluR2. Although functional potencies are not necessarily consistent with binding affinities,^48,49 the predicted binding modes provide insights into the ligand–protein binding events. The detailed binding mode for each ligand and the corresponding docking score are included in the Supporting Information (Figure S3 and Table S2).

Figure 3.

Molecular docking results of 6 and 8–11. (A) Binding pose of compound 6; (B) binding poses of 8–11 overlaid with 6 at the allosteric site. The GluR2 homology model was built from the crystal structures of mGluR1 complexed with glutamate (PDB ID: 1EWK),⁵⁰ mGluR5 complexed with glutamate (PDB ID: 3LMK),⁵¹ and metabotropic glutamate receptor 5 apo form (PDB ID 6N52)⁵² as previously described.³⁴ Images are rendered in Pymol 2.3.3.

Radiochemistry.

Compounds 8 and 9 were selected for radiolabeling with fluorine-18 due to their improved PAM affinities than compound 6. Moreover, the para-fluoride in compounds 8 and 9 provides less steric hindrance when introducing a bulky leaving group and/or during S_NAr replacement with [¹⁸F]F⁻ ions than the ortho-fluoride at compounds 10 and 11. Both [¹⁸F]8 and [¹⁸F]9 were synthesized by the previously described method using our modified alcohol-enhanced Cu-mediated radiofluorination of organoboranes as [¹⁸F]6 (Scheme 2).^32,53 The boronic acid pinacol esters (Bpin) 37 and 38 were prepared via the reductive amination between 14 and the corresponding fragments 35 and 36, respectively. Fragment 35 was prepared over three steps. The intermediate 34 was obtained via the Suzuki coupling between compounds 31 and 32 and the subsequent Miyaura borylation of 33.⁵⁴ Compound 34 was then hydrogenated and deprotected to give compound 35. The order of the synthetic steps was important. Otherwise, if compound 33 was hydrogenated first using Pd/C (10 wt %), the bromide group would be cleaved under the hydrogenation conditions. Alternatively, switching to PtO₂ as a catalyst seems to have avoided this issue according to a recent publication.⁵⁵ The automated synthetic procedures used for [¹⁸F]6⁵³ were applied for the syntheses of both [¹⁸F]8 and [¹⁸F]9 in the TRACERLab FX_F-N platform. In the reaction, tetraethyl ammonium bicarbonate (TEAB) was used as a base and phase transfer agent, n-BuOH/dimethylacetamide (DMA) were used as solvents, and [Cu(OTf)₂py₄] was used as a catalyst. The same stoichiometry for each reagent was applied. [¹⁸F]8 was isolated with a radiochemical yield (RCY) of 8 ± 2% (non-decay-corrected, n = 5) and a molar activity (A_m) of 273 ± 124 GBq/μmol (n = 2) at the end of synthesis (EOS, t = 55 min). [¹⁸F]9 was prepared with a RCY of 5 ± 3% (non-decay-corrected, n = 6) and a A_m of 348 ± 144 GBq/μmol (n = 3) at the end of synthesis (t = 55 min). At last, both radiotracers, in the range of 1.85–7.4 GBq radioactivity prepared from ~74 GBq [¹⁸F]F⁻, were formulated into 10% ethanolic saline solution with excellent radiochemical purity (>98%) at the time of injection (Supporting Information, Figures S4–S9).

Scheme 2. Syntheses of [¹⁸F]8 and [¹⁸F]9^a.

^aReagents and conditions: (a) Na₂CO₃, Pd(PPh₃)₂Cl₂, dimethyl ether/water (5:1, v/v), 85 °C, 2 h; (b) KOAc, Pd(PPh₃)₂Cl₂, 1,4-dioxane, 80 °C, overnight; (c) (i) H₂, Pd/C (10 wt %), 43 psi, rt, overnight; (ii) 4 N HCl in dioxane, rt, 1 h; (d) TEA, MgSO₄, NaBH(OAc)₃, rt, 12 h; and (e) 37 or 38 (6.2 μmol), TEAB (14.1 μmol), [Cu(OTf)₂py₄] (13.3 μmol), DMA (0.8 mL), n-Butanol (0.4 mL), 130 °C, 10 min.

PET Imaging in Rats.

The radioligands [¹⁸F]8 and [¹⁸F]9 were first characterized in rat models (male, Sprague–Dawley). In each study, rats were anesthetized, and 60 min dynamic PET scans were performed following the tail vein injection of the radiotracer. Pretreatment agents were administered 10 min before radioactivity. Representative PET images of cumulative volumetric distribution of [¹⁸F]8 and [¹⁸F]9 are shown in Figure 4A. The summed images at a time interval of 2–20 min are displayed on coronal, axial, and sagittal levels. Both tracers were brain penetrant and clearly outlined the biodistribution of mGlu2 receptors in regions of striatum, thalamus, cortex, hippocampus, and cerebellum. Time–activity curves (TACs) of [¹⁸F]8 and [¹⁸F]9 from their representative baseline studies showed rapid brain uptake and time-dependent radioactivity accumulation in different brain regions (Figure 4B). [¹⁸F]8 peaked at 2.5 min with an SUV_max of 1.9, and the accumulations were similar across the regions of interest. On the other hand, [¹⁸F]9 had an SUV_max value of 1.7 at 2.5 min, and the accumulations were similar in the striatum, cerebellum, cortex, and hippocampus but relatively lower in the thalamus. The binding kinetics of [¹⁸F]9 is more favorable for diagnostic purposes than that of [¹⁸F]8 because it shows faster washout from all investigated brain areas than [¹⁸F]8 resulting in a lower radiation dose response. The estimated remaining radioactivity 60 min after the injection was 37% of the maximum binding using [¹⁸F]9, while it was 58% using [¹⁸F]8.

Figure 4.

In vivo characterization results of [¹⁸F]8 and [¹⁸F]9 in rat brain. (A) Regional distribution of [¹⁸F]8 and [¹⁸F]9 at a time interval of 2–20 min overlaid by color-coded regions of interest; (B) representative TACs from the baseline studies in different brain areas; (C) radioactivity uptake after pretreatment with different classes of mGluR2 PAMs using different doses at the time window of 2–20 min after administration of the radioligand. *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001 (vs baseline). Graphs were made using Prism 9.0.

To test the in vivo tracer-binding selectivity and the pharmacological effects of mGluR2 PAMs toward radioligand binding, pretreatment experiments were carried out for [¹⁸F]8 via self-pretreatment and for [¹⁸F]9 via different classes of mGluR2 PAMs, including compounds 2, 7, and 9, at different doses. As shown in Figure 4C, all the pretreatment experiments resulted in an enhanced radioactivity uptake in the brain regions of interest at the time window of 2–20 min. Self-pretreatment of [¹⁸F]8 at a dose of 4 mg/kg increased its uptake by 20.3–40.9%, and a significant enhancement was achieved in all the cortex, hippocampus, and cerebellum. On the other hand, [¹⁸F]9 had the most elevated radioactivity uptake from the self-pretreatment studies (51.1–81.6%). No dose dependency was observed for self-pretreatment at the current doses of 2.0 versus 4.0 mg/kg. Compound 7 elicited similar potentiation effects at a dose of 4.0 mg/kg compared to the self-pretreatment, ranging from 45.1 to 63.1%, and a significant enhancement was observed in all the investigated brain areas. This is consistent with our previous finding that compound 7 was able to enhance the brain uptake of the mGluR2 PAM-based radiotracer of [¹⁸F]6 in non-human primates.³² Preadministration of BINA (2) resulted in a significantly enhanced binding of [¹⁸F]9 in the thalamus, cingulate and prefrontal cortices, hippocampus, and cerebellum at a dose of 2 mg/kg, ranging from 16.7 to 29.3%. However, pretreatment of BINA at a lower dose of 1 mg/kg resulted in negligible radioactivity enhancement at the brain areas investigated.

These results confirm that both [¹⁸F]8 and [¹⁸F]9 are suitable radioligands for imaging mGluR2 in the rat brain, and their brain uptake can be enhanced by the pretreatments of mGluR2 PAMs. For diagnostic purpose, [¹⁸F]9 has a better brain heterogeneity and more favorable binding kinetics than [¹⁸F]8.

PET Imaging in a Non-Human Primate.

The feasibility of [¹⁸F]9 as an mGluR2 PET imaging ligand was further evaluated in a cynomolgus monkey. In each study, a 120 min dynamic PET acquisition was performed following the injection of [¹⁸F]9. Compound 2 was used as a pretreatment agent in the monkey study, and it was injected 10 min before tracer injection. Parallel to this process, arterial blood was sampled to measure the tracer metabolism, plasma (PL)-free fraction (f_p), and the time courses of [¹⁸F]9 concentration in the whole blood (WB) and PL. As shown in Figure 5A, the WB/PL ratio of [¹⁸F]9 was consistent across the studies stabilizing after 1 min post-tracer injection with a mean value of 1.12 ± 0.09. RadioHPLC measurements indicated the presence of primarily polar and moderately polar metabolites as shown by the radiochromatograms (Figure 5B). Percent parent in PL measurements revealed a rather slow rate of metabolism with 49.3 ± 1.4% of PL activity attributable to unmetabolized [¹⁸F]9 at 90 min (Figure 5C). Figure 5D shows the corresponding individual metabolite-corrected [¹⁸F]9 SUV time courses in PL. The PL-free fraction was 0.12 ± 0.06 across studies (range: 0.106–0.076).

Figure 5.

[¹⁸F]9 analysis in arterial blood. (A) WB/PL ratio; (B) representative radioHPLC chromatogram of PL samples; (C) individual time course of [¹⁸F]9 percent parent in PL (%PP); and (D) individual metabolite-corrected [¹⁸F]9 SUV time courses in PL.

[¹⁸F]9 entered the monkey brain readily and peaked at 6 min post-tracer injection (SUV > 3.5). Brain uptake and kinetics revealed heterogenous distribution of [¹⁸F]9 across brain regions. Figure 6A,B shows the selected brain TACs along with compartmental model fits for the cerebellum non-vermis, striatum, thalamus, frontal cortex, hippocampus, and whole brain from the baseline condition and after pretreatment with BINA (2, 0.5 mg/kg, iv.) 10 min prior to PET data acquisition. The TACs along with compartmental model fits for another 14 brain regions under baseline and pretreatment conditions are included in the Supporting Information (see Figure S10). According to the Akaike information criteria (AIC),⁵⁶ the preferred model was a reversible 2T model with a fixed vascular contribution v included as a model parameter (2T4k1v). This model provided a stable regional total volume of distribution V_T estimates under both baseline and pretreatment conditions. The modeling parameters and regional V_T values derived from 2T4k1v and Logan graphical analysis were provided in the Supporting Information for different brain regions in a baseline study (Supporting Information, Tables S3 and S4). K₁ values, reflecting tracer delivery, were ~0.25 mL/min/cc in the whole brain based on the 2T model, indicating moderate to high brain penetration (Supporting Information, Table S3). Logan plots linearized very well by a t* of 30 min (Supporting Information, Figure S11). The Logan V_T estimates from this representative study demonstrated enhanced [¹⁸F]9 uptake in brain areas with high tracer uptake as the thalamus (~3.3%) after pretreatment with 0.5 mg/kg compound 2 (Figure 6C). Figure 6D shows the corresponding Logan V_T images.

Figure 6.

[¹⁸F]mG2P026 (9) in the primate brain. [¹⁸F]9 kinetics and 2T4k1v model fits in six brain areas, obtained for a baseline study (A) and after pretreatment with 0.5 mg/kg compound 2 (B); (C) Logan V_T bar graph of [¹⁸F]9 at different brain regions; and (D) structural MRI (MEMPRAGE) and [¹⁸F]9 Logan V_T images at t*30 min for the baseline (middle panel) and pretreatment conditions (bottom panel). Images are presented in the NIMH Macaque Template (NMT)⁵⁷ space.

Direct Comparison between [¹⁸F]9 and [¹⁸F]6 Using a Graphical Method.

[¹⁸F]9 displayed an enhanced heterogeneity in kinetics across brain regions compared to [¹⁸F]6 based on its TACs (Supporting Information, Figure S12). We further compared the performance of [¹⁸F]9 as an mGluR2 PET imaging ligand with [¹⁸F]6 using the graphical method reported by Guo et al.⁵⁸ As Figure 7 shows, a comparison of the regional Logan V_T values of [¹⁸F]9 and [¹⁸F]6 demonstrates a strong linear relationship (r² = 0.9, p < 0.0001), indicating that the two ligands bind to the same target. The in vivo K_D ratio between [¹⁸F]6 and [¹⁸F]9 was estimated with the graphical analysis using the f_p data (f_p [¹⁸F]6 = 0.15 ± 0.04; f_p [¹⁸F]9 = 0.12 ± 0.06) and the estimated slope (slope = 1.061, CI95% [0.8857–1.237]) and gave a result of 1.326. The in vivo K_D ratio suggested [¹⁸F]9 has a higher in vivo binding affinity than [¹⁸F]6. The y-intercept (y = −0.054, CI95% [−0.8793 to 0.7704]) of this linear regression suggested in vivo specific binding of the two tracers is equivalent.

Figure 7.

Scatter plots comparing the Logan V_T values obtained for [¹⁸F]9 and [¹⁸F]6. Paired comparison is plotted for the regional Logan V_T estimates in amygdala (1), anterior cingulate cortex (2), caudate nucleus (3), cerebellum grey (4), cerebellum whole Non-vermis (5), cortex all (6), frontal cortex (7), hippocampus (8), insular cortex (9), nucleus accumbens (10), occipital cortex (11), parietal cortex (12), posterior cingulate cortex (13), prefrontal cortex (14), putamen (15), striatum (16), temporal cortex (17), thalamus (18), white matter (19), and whole brain (20). Black line represents the linear regression line. Graph was made using Prism 9.0.

CONCLUSIONS

We have developed a high-contrast PET radioligand [¹⁸F]-mG2P026 (9) for imaging mGluR2 in the brain. Compared to compound 6, the ago-PAM 9 had a 16-fold increase of the PAM activity (EC₅₀ = 11 nM) and an apparent mGluR2 agonist activity (EC₅₀ = 84 nM). Radiolabeling of compound 9 was achieved under the same conditions as [¹⁸F]6 with similar RCY and molar activity. [¹⁸F]9 readily entered the rat brain with better imaging characteristics than the piperidine analogue [¹⁸F]8. Distinct from the orthosteric radioligands, pretreatment with mGluR2 PAMs BINA (2) and compounds 6 and 9 led to increased [¹⁸F]9 uptake in the rat brain, where the self-pretreatment had the most enhancement. Further characterization of [¹⁸F]9 in a non-human primate revealed moderate metabolism stability, fast binding kinetics, and feasibility in mapping the mGlu2 receptors. The V_T values derived from the Logan model demonstrated a trend of radioactivity enhancement following the pretreatment with compound 2 (0.5 mg/kg, iv), although the statistical significance was not achieved under the investigated conditions. As a high-contrast imaging ligand, [¹⁸F]9 was further compared with [¹⁸F]6 in the same monkey using a graphical method, where [¹⁸F]9 showed enhanced heterogeneity and a higher binding affinity although its white matter binding was similar to that of [¹⁸F]6. Therefore, [¹⁸F]9 is a promising PET imaging ligand for mGluR2. In the future, the imaging characteristics of [¹⁸F]9 will be further evaluated with different classes and/or doses of drugs to explore its selectivity and sensitivity as an mGluR2 PET imaging ligand. Additionally, whole body PET imaging studies in primates will be done to determine their radiation-dose distribution as well as possible off-target binding before potential translation to human studies.

EXPERIMENTAL SECTION

Animal Procedures.

Animal studies were approved and performed following the guidelines of the Subcommittee on Research Animals of the Massachusetts General Hospital and Harvard Medical School in accordance with the Guide of NIH for the Care and Use of Laboratory Animals.

Chemistry.

All chemical reagents and solvents were purchased from the commercial sources and used without further purification. Thin layer chromatography (TLC) was performed on silica gel TLC aluminum foils (Supelco). Flash column chromatography was carried out on silica gel with a particle size of 60 Å, 230–400 mesh (Supelco). Microwave-assisted reactions were done using a CEM Discover microwave synthesizer. Nuclear magnetic resonance (NMR) spectra were collected with a JEOL 500 MHz spectrometer using chloroform-d (CDCl₃) and methanol-d₄ (CD₃OD) as solvents. Chemical shifts (δ) are assigned as parts per million (ppm) downfield from tetramethylsilane. Coupling constants (J) are expressed in hertz. Splitting patterns are described as s (singlet), d (doublet), t (triplet), q (quartet), or m (multiplet). Liquid chromatography–mass spectrometry (LCMS) was performed using a 1200 series HPLC system (Agilent Technologies, Canada) comprising a multiwavelength UV detector, a model 6310 ion trap mass spectrometer (Santa Clara, CA), and an analytical column (Agilent Eclipse C8, 150 mm × 4.6 mm, 5 μm). The samples were subjected to a gradient elution with the mobile phase of 0.1% formic acid solution of water (A) and acetonitrile (B) over 7 min unless otherwise noted at a flow rate of 0.7 mL/min before the MS spectrometer. Purities of all new compounds were determined via the analytical reverse phase HPLC in the LCMS system, using the area percentage method on the UV trace scanning under a wavelength of 254 nm. All compounds are >95% pure by HPLC analysis. High-resolution mass spectrometry (HRMS) was carried out at the Harvard Center for Mass Spectrometry (Cambridge, MA). The electrospray ionization (ESI) technique was carried out with a Thermo_q-Exactive_Plus_I mass spectrometer.

General Procedure for Reductive Amination.^28,29

To a solution of amine (1.1 equiv) in anhydrous dichloromethane were added triethylamine (4.0 equiv), magnesium sulfate (10.0 equiv), and aldehydes (1.0 equiv). The reaction mixture was stirred at room temperature for 30 min under argon and then sodium triacetoxyborohydride (1.5 equiv) was added. The mixture was stirred at room temperature for another 16 h and was diluted with dichloromethane and washed with water. The aqueous layer was washed two times with dichloromethane, and the combined organic layers were dried over magnesium sulfate. The solvent was reduced in vacuo, and the residue was purified by flash column chromatography to give the desired product.

3-(Cyclopropylmethyl)-7-((4-(2,4-difluorophenyl)piperidin-1-yl)-methyl)-8-(trifluoromethyl)-[1,2,4]triazolo[4,3-a]pyridine (8).

Starting from 14 (50 mg, 0.19 mmol) and 4-(2,4-difluorophenyl)-piperidine (15, 41.2 mg, 0.21 mmol) and following the general procedure described for the reductive amination, compound 8 was obtained as a white solid (40.0 mg, 46.7%). LCMS: m/z 451.2 [M + H]⁺, t_R = 2.95 min. ¹H NMR (500 MHz, CD₃OD): δ ppm 8.52 (d, J = 7.2 Hz, 1H), 7.53 (d, J = 7.2 Hz, 1H), 7.29–7.30 (m, 1 H), 6.84–6.89 (m, 2H), 3.80 (s, 2H), 3.09 (d, J = 6.9 Hz, 2H), 2.96 (d, J = 10.9 Hz, 2H), 2.84–2.85 (m, 1H), 2.28 (t, J = 11.4 Hz, 2H), 1.76–1.84 (m, 4H), 1.24–1.26 (m, 1H), 0.58–0.61 (m, 2H), 0.33–0.35 (m, 2H). ¹³C NMR (125 MHz, CD₃OD): 166.0 (dd, J = 12.0, 99.0 Hz), 164.0 (dd, J = 12.0, 100.2 Hz), 152.0, 150.2, 147.2, 132.7 (d, J = 3.6 Hz), 132.5 (m), 129.8, 127.3 (q, J = 274.7 Hz), 119.4, 118.6 (q, J = 33.8 Hz), 114.7 (d, J = 18.0 Hz), 107.0 (t, J = 26.3 Hz), 61.6, 58.0, 38.9, 35.9, 32.3, 11.9, 8.0. HRMS (ESI⁺) for C₂₃H₂₄F₅N₄⁺ [M + H]⁺ requires m/z = 451.1916; found, 451.1920.

7-((4-(2-Chloro-4-fluorophenyl)piperazin-1-yl)methyl)-3-(cyclo-propylmethyl)-8-(trifluoromethyl)-[1,2,4]triazolo[4,3-a]pyridine (9).

Starting from 14 (70 mg, 0.26 mmol) and 1-(2-chloro-4-fluorophenyl)piperazine (16, 62.3 mg, 0.29 mmol) and following the general procedure described for the reductive amination, compound 9 was obtained as a white solid (35.0 mg, 28.8%). LCMS: m/z 468.1 [M + H]⁺, t_R = 3.44 min. ¹H NMR (500 MHz, CD₃OD): δ ppm 8.53 (t, J = 7.0 Hz, 1H), 7.51 (t, J = 7.0 Hz, 1H), 7.12–7.19 (m, 2H), 6.99–7.04 (m, 1H), 3.84 (s, 2H), 3.10 (t, J = 7.0 Hz, 2H), 2.98–3.04 (m, 4H), 2.66–2.70 (m, 4H), 1.23–1.27 (m, 1H), 0.58–0.62 (m, 2H), 0.33–0.35 (m, 2H). ¹³C NMR (125 MHz, CD₃OD): 158.5 (d, J = 243.6 Hz), 148.1, 146.2, 145.9, 142.8, 129.5 (d, J = 10.6 Hz), 125.9, 123.4 (q, J = 274.9 Hz), 121.4 (d, J = 9.0 Hz), 117.2 (d, J = 25.9 Hz), 115.6, 114.9 (q, J = 33.9 Hz), 114.0 (d, J = 21.8 Hz), 57.4, 53.2, 51.4, 28.3, 7.9, 4.1. HRMS (ESI⁺) for C₂₂H₂₃ClF₄N₅⁺ [M + H]⁺ requires m/z = 468.1573; found, 468.1578.

7-((4-(4-Chloro-2-fluorophenyl)piperazin-1-yl)methyl)-3-(cyclo-propylmethyl)-8-(trifluoromethyl)-[1,2,4]triazolo[4,3-a]pyridine (10).

Starting from 14 (70 mg, 0.26 mmol) and 1-(4-chloro-2-fluorophenyl)piperazine (17, 62.3 g, 0.29 mmol) and following the general procedure described for the reductive amination, compound 10 was obtained as a white solid (25.0 g, 20.6%). LCMS: m/z 467.9 [M + H]⁺, t_R = 3.89 min. ¹H NMR (500 MHz, CDCl₃): δ ppm 8.04 (d, J = 7.2 Hz, 1H), 7.38 (d, J = 7.2 Hz, 1H), 7.03–7.05 (m, 2H), 6.84 (t, J = 9.1 Hz, 1H), 3.80 (s, 2H), 3.08–3.10 (m, 6H), 2.67–2.68 (m, 4H), 1.18–1.19 (m, 1H), 0.61–0.65 (m, 2H), 0.33–0.36 (m, 2H). ¹³C NMR (125 MHz, CD₃OD): 159.3 (d, J = 248.8 Hz), 152.1, 150.2, 146.6, 143.0 (d, J = 8.7 Hz), 130.6 (d, J = 9.8 Hz), 128.3, 126.2, 123.8, 120.2 (d, J = 24.8 Hz), 119.5, 118.9 (q, J = 33.9 Hz), 61.3, 56.9, 54.3, 32.3, 11.9, 8.0. HRMS (ESI⁺) for C₂₂H₂₃ClF₄N₅⁺ [M + H]⁺ requires m/z = 468.1573; found, 468.1578.

3-(Cyclopropylmethyl)-7-((4-(2-fluorophenyl)piperazin-1-yl)-methyl)-8-(trifluoromethyl)-[1,2,4]triazolo[4,3-a]pyridine (11).

Starting from 14 (0.10 g, 0.37 mmol) and 1-(2-fluorophenyl)-piperazine (18, 64.5 μL, 0.41 mmol) and following the general procedure described for the reductive amination, compound 11 was obtained as a white solid (55.4 mg, 34.6%). LCMS: m/z 434.4 [M + H]⁺, t_R = 6.28 min (15 min gradient elution). ¹H NMR (500 MHz, CD₃OD): δ ppm 8.53 (dd, J = 7.5, 9.5 Hz, 1H), 7.51 (dd, J = 7.5, 9.5 Hz, 1H), 6.93–7.06 (m, 4H), 3.84 (d, J = 8.3 Hz, 2H), 3.08–3.12 (m, 6H), 2.67–2.69 (m, 4H), 1.24–1.26 (m, 1H), 0.59–0.63 (m, 2H), 0.33–0.36 (m, 2H). ¹³C NMR (125 MHz, CD₃OD): 155.8 (d, J = 244.9 Hz), 148.1, 146.2, 142.7, 140.0 (d, J = 8.5 Hz), 126.6, 124.4 (d, J = 2.9 Hz), 123.4 (q, J = 274.9 Hz), 122.6 (d, J = 7.9 Hz), 118.9 (d, J = 1.9 Hz), 115.6 (d, J = 20.6 Hz), 115.5, 114.9 (q, J = 33.7 Hz), 57.4, 53.1, 50.5, 28.3, 7.9, 4.1. HRMS (ESI⁺) for C₂₂H₂₄F₄N₅⁺ [M + H]⁺ requires m/z = 434.1962; found, 434.1966.

3-Chloro-2-hydrazineyl-5-(trifluoromethyl)pyridine (20).

To a solution of 2,3-dichloro-5-(trifluoromethyl)pyridine (19, 1.0 g, 4.63 mmol) in 1,4-dioxane (7.0 mL) was added hydrazine monohydrate (2.25 mL, 27.8 mmol). The reaction was heated in a sealed reactor at 80 °C for 16 h. The reaction mixture was then cooled to room temperature, quenched with NH₄OH (32% aqueous solution), and reduced in vacuo. The resulting residue was taken up by ethanol and heated to reflux. After filtering off the solid while hot, the filtrate was concentrated in vacuo to give compound 20 as a white solid (0.3 g, 30.6%). LCMS: m/z 211.9 [M + H]⁺, t_R = 2.93 min (8 min gradient elution). ¹H NMR (500 MHz, CDCl₃): δ ppm 8.33 (s, 1H), 7.65 (d, J = 1.8 Hz, 1H), 6.55 (s, 1H), 4.04 (s, 2H). ¹³C NMR (125 MHz, CDCl₃): 157.1, 143.6 (d, J = 4.5 Hz), 133.1 (d, J = 2.9 Hz), 123.6 (q, J = 271.2 Hz), 117.3 (d, J = 33.5 Hz), 114.4. HRMS (ESI⁺) for C₆H₆ClF₃N₃⁺ [M + H]⁺ requires m/z = 212.0197; found, 212.0200.

N′-(3-Chloro-5-(trifluoromethyl)pyridin-2-yl)-2-cyclopropylace-tohydrazide (21).

To a solution of 20 (0.3 g, 1.40 mmol) in anhydrous dichloromethane (4.0 mL) were added triethylamine (0.3 mL, 1.28 mmol) and cyclopropylacetyl chloride (0.2 g, 1.70 mmol) at 0 °C. The reaction was stirred at room temperature for 16 h and then quenched with saturated NaHCO₃ solution. The resulting solution was extracted three times with dichloromethane, and the combined organic layers were dried over magnesium sulfate and reduced in vacuo to give 21 as a white solid (0.17 g, 41.4%). LCMS: m/z 294.0 [M + H]⁺, t_R = 4.45 min (8 min gradient elution). ¹H NMR (500 MHz, CDCl₃): δ ppm 8.32 (s, 1H), 8.29 (d, J = 4.5 Hz, 1H), 7.74 (d, J = 1.8 Hz, 1H), 7.59 (d, J = 4.9 Hz, 1H), 2.31 (d, J = 7.2 Hz, 2H), 1.08–1.11 (m, 1H), 0.66–0.69 (m, 2H), 0.28–0.31 (m, 2H). ¹³C NMR (125 MHz, CDCl₃): 170.7, 154.5, 143.5 (d, J = 4.0 Hz), 134.0, 123.3 (q, J = 271.4 Hz), 119.3 (q, J = 33.8 Hz), 115.2, 39.6, 6.8, 4.8. HRMS (ESI⁺) for C₁₁H₁₁ClF₃N₃O⁺ [M + H]⁺ requires m/z = 294.0616; found, 294.0619.

8-Chloro-3-(cyclopropylmethyl)-6-(trifluoromethyl)-[1,2,4]-triazolo[4,3-a]pyridine (22).

To a solution of compound 21 (0.113 g, 0.39 mmol) in anhydrous 1,2-dichloroethane (1.0 mL) was added POCl₃ (0.073 mL, 0.77 mmol) in a microwave reactor. The reaction mixture was heated at 150 °C for 10 min in the CEM microwave reactor and then quenched with a solution of saturated NaHCO₃. The organic layer was isolated, and the aqueous layer was extracted twice with dichloromethane. The combined organic layers were dried with magnesium sulfate, and the volatiles were reduced under vacuo. The residue was purified by flash column chromatography to give 22 as a brown oil (54.0 mg, 50.9%). LCMS: m/z 276.0 [M + H]⁺, t_R = 4.39 min (8 min gradient elution). ¹H NMR (500 MHz, CDCl₃): δ ppm 8.28–8.29 (m, 1H), 7.42 (d, J = 1.1 Hz, 1H), 3.12 (d, J = 6.8 Hz, 2H), 1.19–1.22 (m, 1H), 0.66–0.70 (m, 2H), 0.35–0.38 (m, 2H). ¹³C NMR (125 MHz, CDCl₃): 149.8, 147.7, 124.6, 122.4 (dd, J = 261.5, 554.0 Hz), 121.3, 120.4 (q, J = 5.6 Hz), 118.4 (q, J = 35.0 Hz), 29.6, 8.4, 5.4. HRMS (ESI⁺) for C₁₁H₁₀ClF₃N₃⁺ [M + H]⁺ requires m/z = 276.0510; found, 276.0513.

3-(Cyclopropylmethyl)-6-(trifluoromethyl)-8-vinyl-[1,2,4]-triazolo[4,3-a]pyridine (23).

In a microwave reactor, 22 (82.0 mg, 0.30 mmol), vinylboronic acid pinacol ester (61.0 μL, 0.36 mmol), Pd(PPh₃)₄ (17.3 mg, 0.015 mmol), and saturated aqueous NaHCO₃ solution (0.62 mL) were mixed in 1,4-dioxane (3.2 mL). The reaction mixture was heated at 150 °C for 15 min in the CEM microwave reactor. After cooling to room temperature, the reaction mixture was diluted with ethyl acetate/water. The aqueous layer was extracted three times with ethyl acetate. The combined organic layers were washed with brine and dried over magnesium sulfate. The solvent was reduced in vacuo, and the residue was purified by flash column chromatography to give 23 as a yellow solid (47.0 mg, 58.6%). LCMS: m/z 268.0 [M + H]⁺, t_R = 4.73 min (8 min gradient elution). ¹H NMR (500 MHz, CDCl₃): δ ppm 8.22 (s, 1H), 7.25 (s, 1H), 6.99 (d, J = 6.0 Hz, 2H), 5.84 (t, J = 6.1 Hz, 1H), 3.11 (d, J = 6.8 Hz, 2H), 1.21–1.25 (m, 1H), 0.65–0.68 (m, 2H), 0.34–0.37 (m, 2H). ¹³C NMR (125 MHz, CDCl₃): 148.2, 130.3, 128.0, 124.5, 123.1 (q, J = 271.6 Hz), 119.9, 119.8, 119.7, 118.5 (q, J = 34.1 Hz), 29.4, 8.5, 5.3. HRMS (ESI⁺) for C₁₃H₁₃F₃N₃⁺ [M + H]⁺ requires m/z = 268.1056; found, 268.1061.

3-(Cyclopropylmethyl)-6-(trifluoromethyl)-[1,2,4]triazolo[4,3-a]-pyridine-8-carbaldehyde (24).

To a solution of 23 (0.54 g, 2.02 mmol) in water (5.4 mL) and 1,4-dioxane (21.6 mL) were added sodium periodate (1.30 g, 6.08 mmol) and osmium tetroxide (2.5% in tert-butanol, 1.03 mL, 0.08 mmol). The reaction mixture was stirred at room temperature for 2 h. After the reaction was completed, the mixture was diluted with ethyl acetate/water. The aqueous layer was extracted three times with ethyl acetate. The combined organic layers were washed with brine and dried over magnesium sulfate. The solvent was reduced in vacuo, and the residue was purified by flash column chromatography to give 24 as a brown oil (0.224 g, 41.2%). LCMS: m/z 270.0 [M + H]⁺, t_R = 3.94 min (8 min gradient elution). ¹H NMR (500 MHz, CDCl₃): δ ppm 10.78 (s, 1H), 8.55 (d, J = 1.1 Hz, 1H), 7.99 (d, J = 1.3 Hz, 1H), 3.17 (d, J = 6.8 Hz, 2H), 1.23–1.24 (m, 1H), 0.69–0.72 (m, 2H), 0.37–0.40 (m, 2H). ¹³C NMR (125 MHz, CDCl₃): 186.3 (d, J = 4.2 Hz), 148.7, 147.7, 125.7, 124.1, 123.9, 122.6 (q, J = 271.7 Hz), 118.3 (q, J = 35.2 Hz), 29.5, 8.4, 5.5. HRMS (ESI⁺) for C₁₂H₁₁F₃N₃O⁺ [M + H]⁺ requires m/z = 270.0849; found, 270.0853.

3-(Cyclopropylmethyl)-8-((4-(2,4-difluorophenyl)piperazin-1-yl)-methyl)-6-(trifluoromethyl)-[1,2,4]triazolo[4,3-a]pyridine (12).

Starting from 24 (50 mg, 0.19 mmol) and 4-(2,4-difluorophenyl)-piperidine (41.2 mg, 0.21 mmol) and following the procedure described for the reductive amination, compound 12 was obtained as a white solid (22.0 mg, 25.6%). LCMS: m/z 452.1 [M + H]⁺, t_R = 4.28 min (8 min gradient elution). ¹H NMR (300 MHz, CD₃OD): δ ppm 9.16 (s, 1H), 7.90 (s, 1H), 7.02–7.10 (m, 1H), 6.82–6.93 (m, 2H), 4.02 (s, 2H), 3.07–3.10 (m, 4H), 2.84 (d, J = 7.0 Hz, 2H), 2.75–2.79 (m, 4H), 1.21–1.31 (m, 1H), 0.53–0.59 (m, 2H), 0.28–0.33 (m, 2H). ¹³C NMR (75 MHz, CD₃OD): 168.2, 151.5, 136.6 (d, J = 9.0 Hz), 127.3, 126.3 (d, J = 5.4 Hz), 124.6 (d, J = 2.7 Hz), 123.4 (d, J = 270.5 Hz), 119.8 (dd, J = 4.1, 9.3 Hz), 117.5 (d, J = 34.7 Hz), 110.5 (d, J = 3.8 Hz), 110.2 (d, J = 3.8 Hz), 103.8 (d, J = 51.6 Hz), 103.9, 55.9, 53.0, 50.6, 32.8, 9.2, 3.8. HRMS (ESI⁺) for C₂₂H₂₃F₅N₅⁺ [M + H]⁺ requires m/z = 452.1868; found, 452.1873.

Diethyl 2-(((1-(Cyclopropylmethyl)-1H-pyrazol-5-yl)amino)-methylene)malonate (27).

A solution of diethyl 2-(ethoxymethylene)malonate (25, 1.62 mL, 8.02 mmol) and 1-(cyclopropylmethyl)-1H-pyrazol-5-amine (26, 1.0 g, 7.29 mmol) in ethanol was heated to reflux overnight. The solvent was removed in vacuo. The residue was purified via flash column chromatography to give 27 as a pale yellow solid (3.94 g, 88.0%). LCMS: m/z 308.0 [M + H]⁺, t_R = 3.58 min. ¹H NMR (500 MHz, CDCl₃): δ ppm 11.02 (d, J = 12.8 Hz, 1H), 8.16 (d, J = 12.8 Hz, 1H), 7.41 (d, J = 1.9 Hz, 1H), 6.06 (d, J = 1.9 Hz, 1H), 4.31 (q, J = 7.1 Hz, 2H), 4.22 (q, J = 7.1 Hz, 2H), 3.96 (d, J = 6.9 Hz, 2H), 1.37 (t, J = 7.1 Hz, 3H), 1.30 (t, J = 7.1 Hz, 3H), 1.22–1.26 (m, 1H), 0.59–0.63 (m, 2H), 0.39–0.43 (m, 2H). ¹³C NMR (125 MHz, CD₃Cl₃): 169.1, 165.1, 154.1, 139.2, 138.9, 95.46, 94.31, 60.9, 60.4, 53.2, 14.4, 14.3, 11.2, 4.0. HRMS (ESI⁺) for C₁₅H₂₂N₃O₄⁺ [M + H] requires m/z = 308.1605; found, 308.1608.

Ethyl 4-Chloro-1-(cyclopropylmethyl)-1H-pyrazolo[3,4-b]-pyridine-5-carboxylate (28).

A solution of compound 27 (1.57 g, 5.10 mmol) in POCl₃ (30.0 mL) was refluxed at 120 °C overnight. The reaction mixture was quenched with water at 0 °C and extracted with ethyl acetate. The aqueous layer was further extracted with two times extraction with ethyl acetate. The combined organic layers were washed with brine and dried over magnesium sulfate. The solvent was reduced in vacuo, and the residue was purified by flash column chromatography to give 28 as a colorless waxy solid (1.35 g, 67.5%). LCMS: m/z 280.0 [M + H]⁺, t_R = 5.54 min (8 min gradient elution). ¹H NMR (500 MHz, CDCl₃): δ ppm 9.01 (s, 1H), 8.19 (s, 1H), 4.45 (q, J = 7.1 Hz, 2H), 4.38 (d, J = 7.2 Hz, 2H), 1.43 (t, J = 7.1 Hz, 3H), 1.36–1.40 (m, 1H), 0.53–0.57 (m, 2H), 0.44–0.47 (m, 2H). ¹³C NMR (125 MHz, CD₃Cl₃): 164.5, 151.7, 151.1, 139.7, 132.6, 118.4, 116.2, 61.8, 52.4, 14.4, 11.3, 4.0. HRMS (ESI⁺) for C₁₃H₁₅ClN₃O₂⁺ [M + H]⁺ requires m/z = 280.0847; found, 280.0852.

Ethyl 1-(Cyclopropylmethyl)-4-(trifluoromethyl)-1H-pyrazolo-[3,4-b]pyridine-5-carboxylate (29).

To a solution of 28 (0.10 g, 0.36 mmol) in acetonitrile (5.7 mL) were added sodium iodide (0.108 g, 0.72 mmol) and acetylchloride (0.57 mL). The reaction mixture was stirred at 50 °C overnight. The reaction mixture was diluted with dichloromethane and washed with saturated aqueous NaHCO₃ solution and brine. The organic layer was dried over magnesium sulfate, and the solvent was reduced under vacuo. The resulting residue was dissolved in DMF (1.4 mL) and mixed with fluorosulfonyl-difluoro-acetic acid methyl ester (51.0 μL, 0.40 mmol) and copper(I) iodide (76.0 mg, 0.40 mmol) in a sealed tube. The mixture was stirred at 100 °C for 6 h. After cooling, the mixture was diluted with water and extracted with ethyl acetate. The organic layer was washed with brine and reduced under vacuo. The resulting residue was purified by flash column chromatography to give 29 as a white solid (65.0 mg, 57.6% over two steps). LCMS: m/z 313.9 [M + H]⁺, t_R = 5.85 min (8 min gradient elution). ¹H NMR (300 MHz, CDCl₃): δ ppm 8.96 (s, 1H), 8.26 (q, J = 2.3 Hz, 1H), 4.42–4.49 (m, 4H), 1.42 (t, J = 7.1 Hz, 3H), 1.36–1.41 (m, 1H), 0.54–0.58 (m, 2H), 0.47–0.50 (m, 2H). ¹³C NMR (75 MHz, CDCl₃): 165.4, 151.2, 150.2, 132.6 (q, J = 4.0 Hz), 131.0 (q, J = 35.7 Hz), 122.7 (d, J = 275.1 Hz), 119.6, 110.9 (d, J = 2.4 Hz), 62.4, 52.2, 13.9, 11.2, 3.9. HRMS (ESI⁺) for C₁₄H₁₅F₃N₃O₂⁺ [M + H]⁺ requires m/z = 314.1111; found, 314.1116.

1-(Cyclopropylmethyl)-4-(trifluoromethyl)-1H-pyrazolo[3,4-b]-pyridine-5-carboxylic Acid (30).

To a solution of 29 (50 mg, 0.16 mmol) in tetrahydrofuran/water (2.0 mL/0.15 mL, v/v) was added lithium hydroxide monohydrate (20.2 mg, 0.48 mmol). The reaction mixture was stirred at room temperature for 12 h. After the reaction was completed, 1 N HCl was added to quench the reaction and the mixture was extracted with ethyl acetate. The organic layer was washed with brine and dried over magnesium sulfate. The solvent was removed under vacuo, and the residue was purified by flash column chromatography to give 30 as a white solid (40.0 mg, 87.7%). LCMS: m/z 286.0 [M + H]⁺, t_R = 4.70 min (8 min gradient elution). ¹H NMR (500 MHz, CDCl₃): δ ppm 9.16 (s, 1H), 8.33 (q, J = 2.4 Hz, 1H), 4.47 (d, J = 7.2 Hz, 2H), 1.40–1.45 (m, 1H), 0.56–0.60 (m, 2H), 0.48–0.51 (m, 2H). ¹³C NMR (125 MHz, CD₃OD): 166.6, 151.0, 150.4, 132.0 (d, J = 3.8 Hz), 130.4 (q, J = 35.8 Hz), 122.9 (q, J = 274.3 Hz), 120.4, 110.6, 51.8, 10.7, 2.9. HRMS (ESI⁺) for C₁₂H₁₁F₃N₃O₂⁺ [M + H]⁺ requires m/z = 286.0798; found, 286.0802.

(1-(Cyclopropylmethyl)-4-(trifluoromethyl)-1H-pyrazolo[3,4-b]-pyridin-5-yl) (4-(2,4-difluorophenyl)piperazin-1-yl)methanone (13).

A solution of 30 (0.12 g, 0.42 mmol) in DMF (4.5 mL) were added HATU (0.16 g, 0.42 mmol) and DIPEA (0.15 mL, 0.84 mmol). After stirring 10 min at room temperature, 4-(2,4-difluorophenyl)piperidine (93.0 mg, 0.465 mmol) was added and stirred for 16 h. The reaction was then quenched with water and extracted with ethyl acetate. The organic layer was dried over magnesium sulfate, and the solvent was removed under vacuo. The residue was purified by flash column chromatography to give 13 as a yellow solid (0.15 mg, 76.9%). LCMS: m/z 466.2 [M + H]⁺, t_R = 4.36 min. ¹H NMR (500 MHz, CDCl₃): δ ppm 8.52 (s, 1H), 8.20 (d, J = 1.6 Hz, 1H), 6.87–6.92 (m, 1H), 6.79–6.84 (m, 2H), 4.44–4.45 (m, 2H), 4.05–4.09 (m, 1H), 3.96–4.01 (m, 1H), 3.39–3.41 (m, 2H), 3.11–3.14 (m, 2H), 2.92–2.94 (m, 2H), 1.40–1.43 (m, 1H), 0.56–0.58 (m, 2H), 0.47–0.49 (m, 2H). ¹³C NMR (125 MHz, CD₃Cl₃): 165.6, 158.5 (d, J = 233.0 Hz), 155.9 (d, J = 249.7 Hz), 150.4, 147.1 (d, J = 30.4 Hz), 136.0, 131.4, 128.0 (d, J = 35.3 Hz), 122.9 (t, J = 137.5 Hz), 120.2 (d, J = 9.4 Hz), 110.9, 110.8 (d, J = 58.4 Hz), 105.1 (d, J = 52.3 Hz), 105.0, 52.4, 51.0, 50.7, 47.6, 42.3, 11.2, 4.1. HRMS (ESI⁺) for C₂₂H₂₁F₅N₅O⁺ [M + H]⁺ requires m/z = 466.1661; found, 466.1665.

4-(2-Fluoro-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-phenyl)piperidine (35).

To a solution of 34 (0.30 g, 0.74 mmol) in ethanol (10.0 mL) was added Pd/C (0.05 g, 10 wt %). The resulting solution was subjected to a hydrogen environment at 43 psi at room temperature overnight. The Pd/C was filtered out, and the solvent was reduced under vacuo to give a yellow oil. LCMS: m/z 350.1 [M + H]⁺ (t-butyl cleaved mass was observed), t_R = 5.17 min. The yellow oil was then taken up by a solution of 4 N HCl in dioxane (2.67 mL) and stirred at room temperature for 1 h. The solvent was removed under vacuo, and the residue was used for the next step without further purification. LCMS: m/z 306.1 [M + H]⁺, t_R = 2.78 min.

7-((4-(2-Chloro-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-phenyl)piperazin-1-yl)methyl)-3-(cyclopropylmethyl)-8-(trifluoromethyl)-[1,2,4]triazolo[4,3-a]pyridine (37).

Starting from 14 (78.0 mg, 0.29 mmol) and 1-(2-chloro-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)piperazine·HCl salt (35, 99.1 mg, 0.29 mmol) and following the general procedure described for the reductive amination, compound 37 was obtained as a waxy white solid (54.0 mg, 33.3%). LCMS: m/z 559.2 [M + H]⁺, t_R = 3.32 min. ¹H NMR (500 MHz, CDCl₃): δ ppm 8.04 (d, J = 7.2 Hz, 1H), 7.52 (d, J = 7.5 Hz, 1H), 7.42–7.47 (m, 2H), 7.23–7.26 (m, 1 H), 3.75 (d, J = 1.4 Hz, 2H), 3.09 (d, J = 6.7 Hz, 2H), 2.90–2.92 (m, 3H), 2.27–2.32 (m, 2H), 1.79–1.83 (m, 4H), 1.33 (s, 12H), 1.15–1.19 (m, 1H), 0.62–0.64 (m, 2H), 0.34–0.36 (m, 2H). ¹³C NMR (125 MHz, CDCl₃): 160.6 (d, J = 246.2 Hz), 147.7, 147.1, 146.7, 141.5, 135.9, 135.8, 130.6, 127.2, 124.3, 124.1, 116.8, 115.1 (m), 84.1, 57.9, 54.5, 35.5, 32.0, 29.3, 24.8, 8.5, 5.3. HRMS (ESI⁺) for C₂₉H₃₆BF₄N₄O₂⁺ [M + H]⁺ requires m/z = 559.2862; found, 559.2871.

3-(Cyclopropylmethyl)-7-((4-(2-fluoro-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)piperidin-1-yl)methyl)-8-(trifluoromethyl)-[1,2,4]triazolo[4,3-a]pyridine (38).

Starting from 14 (100.0 mg, 0.37 mmol) and 4-(2-fluoro-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)piperidine (36, 150.8 mg, 0.42 mmol) and following the general procedure described for the reductive amination, compound 38 was obtained as a waxy brown solid (20.0 mg, 9.4%). LCMS: m/z 576.2 [M + H]⁺, t_R = 4.32 min. ¹H NMR (500 MHz, CDCl₃): δ ppm 8.04 (d, J = 7.3 Hz, 1H), 7.78 (d, J = 1.4 Hz, 1H), 7.63 (dd, J = 1.4, 8.0 Hz, 1H), 7.40–7.45 (m, 1H), 6.97–7.04 (m, 1H), 3.81 (d, J = 1.5 Hz, 2H), 3.10–3.13 (m, 4H), 3.09 (d, J = 6.7 Hz, 2H), 2.70 (t, J = 4.5 Hz, 4H), 1.32 (s, 12H), 1.17–1.20 (m, 1H), 0.61–0.64 (m, 2H), 0.34–0.36 (m, 2H). ¹³C NMR (125 MHz, CD₃OD): 151.6, 148.2, 146.1, 136.4, 135.8, 134.1, 133.2, 127.7, 126.1, 123.3 (q, J = 274.7 Hz), 119.6, 115.7, 115.2 (d, J = 34.5 Hz), 83.9, 57.3, 53.1, 50.5, 28.3, 23.8, 7.9, 4.1. HRMS (ESI⁺) for C₂₈H₃₅BClF₃N₅O ⁺₂ [M + H]⁺ requires m/z = 576.2519; found, 576.2530.

Radiochemistry.

The fully automated radiosyntheses of [¹⁸F]8 and [¹⁸F]9 followed the same procedures described for [¹⁸F]6 in GE TRACERLab FX_FN platform.^32,53 Briefly, no carrier added [¹⁸F]-fluoride ion was produced via the ¹⁸O(p,n)¹⁸F reaction by irradiating ¹⁸O-enriched water (Isoflex Isotope, San Francisco, CA) in a GE PETtrace 16.5 MeV cyclotron (GE Healthcare, Waukesha, WI, USA). The [¹⁸F]fluoride solution was enriched with a QMA Sep-Pak Cartridge (Sep-Pak plus light, Waters, Milford, MA) before released into the reactor by a solution of tetraethylammonium bicarbonate (TEAB, 2.7 mg, 14.1 μmol) in acetonitrile/water (0.7 mL/0.3 mL). Azeotropic drying of the [¹⁸F]fluoride ion was performed at 80 °C for 10 min and then 100 °C for 3 min with addition of another 1.0 mL anhydrous acetonitrile (1 mL). Then, 0.4 mL n-BuOH (0.4 mL) was added to the reactor, followed by a solution of 37 or 38 (6.2 μmol) and [Cu(OTf)₂Py₄] (9.0 mg, 13.3 μmol) in DMA (0.8 mL). The reaction mixture was heated at 130 °C for 10 min and then quenched by adding 4.0 mL water at 50 °C. [¹⁸F]8 or [¹⁸F]9 was then separated with a semipreparative HPLC system equipped with an XBridge BEH C₁₈ OBD column (130 Å, 5 μm, 10 × 250 mm) and eluted with acetonitrile/0.1 M ammonium formate solution [55:45 (v/v)] at 6 mL/min. The fraction containing [¹⁸F]8 or [¹⁸F]9 was collected, diluted with 25 mL high purity water, and passed through a C18 cartridge (light Sep-Pak, Waters, Milford, MA). The cartridge was then washed with 10 mL water and eluted with 0.6 mL ethanol into a product collection vial. Then, 5.4 mL saline was passed the cartridge into the collection vial to afford the formulated solution. The radiochemical identity, molar activity (A_m), and purity of the injected radioligands were determined using the radio-HPLC system (Waters 4000) using an XBridge analytical column (C18, 3.5 μm, 4.6 × 150 mm) eluted with acetonitrile/0.1 M ammonium formate solution [60:40 (v/v)] at 1 mL/min and a UV wavelength of 254 nm.

Molecular Modeling.

The molecular docking experiments of compounds 6 and 8–11 in the allosteric site were done as previously described for compound 7 using the in-house made mGluR2 homology model.³⁴ The ligands were prepared and optimized in Avogadro 1.2⁵⁹ before docking.

Molecular docking was performed into the homology model using Extra Precision/Induced Fit Docking in Glide.^60–62

Partition Coefficient.

LogD_7.4 was determined via a literature method.⁴⁰ Briefly, a test compound (0.1 mg) was mixed with n-octanol (1.0 mL) and PBS buffer (1.0 mL) at pH 7.4 in an Eppendorf tube. The tube was vortexed for 1 min before incubated at 37 °C overnight. The amount of the test compound was measured in each phase from the area under the peak at a wavelength of 254 nm using the HPLC system (UltiMate 3000). LogD_7.4 was calculated as log([compound in octanol]/[compound in PBS]). The assays were repeated at least three times for each compound.

In Vitro Functional Assays.

The preliminary PAM activity of test compounds toward mGluR2 and their selectivity against other mGluRs were determined as previously described for compounds 6 and 7.^32,34 Briefly, HEK 293-derived Flp-In T REx-293 cells stably transfected with mGluR1, mGluR2, mGluR3, mGluR4, mGluR6, or mGluR8 in a tetracycline-inducible manner were maintained and passaged in DMEM medium containing 10% FBS, 100 U/mL penicillin, 100 μg/mL streptomycin (Gibco-Thermo Fisher), 100 μg/mL hygromycin B (Omega Scientific), and 15 μg/mL blasticidin (InvivoGen) in a humidified atmosphere at 37 °C and 5% CO₂. The ligand activation toward mGluR1 and mGluR5 was measured by using Ca²⁺ mobilization assays. In these assays, the mGluR1 or mGluR5 stable cell lines were plated into poly-l-lysine (PLL)-coated 384-well black clear bottom cell culture plates with complete Basal Medium Eagle (BME) buffer. The density was 20,000 cells in 40 μL per well. The ligand activation of the G_i/o coupled receptors mGluR2, mGluR3, mGluR4, mGluR6, and mGluR8 was determined using the split luciferase-based GloSensor cAMP biosensor assay (Promega) measuring the changes of intracellular cAMP concentration. The preliminary PAM functional activity was tested in the presence of an EC₂₀ amount of l-glutamate (10 μM). Further PAM characterization in the binding cooperativity assay was performed by adding subagonism doses of compounds 8–10 (and agonist doses of compound 11) to serial dilutions of l-glutamate up to 300 μM. Each assay was repeated at least three times.

Data Analysis.

The concentration–response curves for each compound were generated using the Prism GraphPad software 9.0 (Graph Pad Inc., San Diego, CA, U.S). The curves were fitted to a four-parameter logistic equation to determine the EC₅₀/IC₅₀ values. The EC₅₀/IC₅₀ are the concentration of a compound that causes a half-maximal potentiation or suppression of the corresponding response in each assay. For the binding cooperativity assay, Schild plots of the data were analyzed and characteristics approximated using the Black Leff Ehlert equation using Prism 9.0.

In Vivo Characterization in Rats.

The PET imaging studies in rats were performed using the Triumph II Preclinical Imaging System (Trifoil Imaging, LLC, Northridge, CA). During the scan, rats were anesthetized with a flow of 1.0–1.5% isoflurane in oxygen (11.5 L/min), and the tail vein was catheterized to inject the radioligands. A dynamic 60 min dynamic acquisition was started from the injection of a radioligand (2041 MBq iv). The vital signs were monitored throughout the imaging. A CT scan was performed after each PET acquisition to obtain anatomical information and correction for attenuation.

Altogether, five Sprague–Dawley rats (male, 275–500 g) were used in eight studies to investigate in vivo imaging characteristics of [¹⁸F]8. Three rats had control studies followed by the “pretreatment” studies 1 week later and one rat had only a control or pretreatment study. The unlabeled compound 8 was used to investigate selectivity and sensitivity of [¹⁸F]8 for imaging mGluR2. Compound 8 was formulated in a solution of 10% DMSO, 5% Tween-20, and 85% PBS with a pH under 5.5 and was administered (4 mg/kg, iv) 10 min before the radioactivity. Similarly, six Sprague–Dawley rats (male, 275–500 g) were used in 14 studies to investigate in vivo imaging characteristics of [¹⁸F]9. Four rats had control studies followed by ten “pretreatment” studies with additional two rats in the pretreatment studies. Imaging studies in each animal were repeated with a 1 or 2 week interval to minimize body-weight-induced changes and still provide enough time to recover from the previous procedure. Pretreatments with compounds 2 (three studies), 7 (four studies), and 9 (three studies) were used to investigate the tracer selectivity and sensitivity. All of these compounds were formulated in 10% DMSO, 5% Tween-20, and 85% PBS with a pH of 5.5 and injected 10 min before the tracer injection. Compound 2 was injected at 1.0 mg/kg (one study) and 2.0 mg/kg (two studies) doses, compound 7 was administered at 4 mg/kg dose (four studies), and compound 9 was studied at 2 mg/kg (one study) and 4 mg/kg doses (two studies).

The PET imaging data were corrected for uniformity, scatter, and attenuation and processed by using the maximum-likelihood expectation-maximization algorithm with 30 iterations to dynamic volumetric images (9 × 20″, 7 × 60″, 6 × 300″, 2 × 600″). The CT data were reconstructed via the modified Feldkamp algorithm using matrix volumes of 512 × 512 × 512 and a pixel size of 170 μm. The ROIs, that is, striatum, thalamus, cortex, hypothalamus, hippocampus, cerebellum, and whole brain, were drawn onto coronal PET slices based on the rat brain atlas. PMOD 3.2 (PMOD Technologies Ltd., Zurich, Switzerland) was used to create the corresponding TACs (time–activity curves) for each tracer. The percent changes in selected brain regions between the baseline and pretreatment studies were calculated at the time window of 2–20 min after radiotracer injection. The statistical significances between the regional binding values of the control and pretreatment studies were calculated based on the mean binding value, standard deviation, and sample size of each group using the two-sample Welch’s t-test (Figure 4C).

In vivo Characterization of [¹⁸F]9 in a Primate.

PET Monkey Scans.

[¹⁸F]9 was characterized in a non-human primate (female, Cynomolgus fascicularis) via the same protocol as described for [¹⁸F]6 and [¹¹C]7.³² Briefly, the PET/CT imaging studies were conducted in the same monkey using a Discovery MI (GE Healthcare) PET/CT scanner. Before the scan, the animal was anesthetized with ketamine/xylazine (10/0.5 mg/kg IM) and then maintained in anesthesia using isoflurane (1–2% in 100% O₂). At baseline, PET acquisition started immediately prior to a 3 min intravenous tracer infusion (138.9 MBq, 375.4 GBq/μmol), and was acquired in 3D list mode for 120 min. Under pretreatment conditions, compound 2 (0.5 mg/kg) was administered intravenously 10 min before [¹⁸F]9 (187.4 MBq, 477.6 GBq/μmol) in the same monkey 2 weeks later.

In each study, a CT scan was carried out to center the head in the imaging field and acquire data for attenuation correction. Arterial blood samples were drawn for radiometabolite analysis, PL-free fraction determination, and radiometabolite-corrected arterial input function. Images were reconstructed via a fully 3D time-of-flight iterative reconstruction algorithm using 3 iterations and 34 subsets. The raw PET data were framed into dynamic series of 6 × 10, 8 × 15, 6 × 30, 8 × 60, 8 × 120, and 18 × 300 s frames. All PET images were corrected for photon attenuation and scatter, system dead time, radioactive decay, random coincident events, and detector inhomogeneity.

During the PET image processing and analyses, MATLAB and FSL⁶³ were used for PET data processing and registration, respectively. A 3D structural T1-weighted magnetization-prepared rapid gradient-echo (MEMPRAGE) imaging from a 3T Biograph mMR (Siemens Medical Systems) was used as anatomical reference during the registration to obtain regional time–activity curves.³²

Regional total volume of distribution (V_T) was estimated for the extracted TACs by reversible one-(1T) and two-(2T) tissue compartment model configurations with the metabolite-corrected arterial PL input function. The Logan graphical analysis technique was also investigated to generate V_T estimates.

Analysis of the Radiometabolite.

The arterial blood sampling and processing of [¹⁸F]9 was performed with the same procedures as described previously.³² Briefly, 22 arterial blood samples were drawn during the 120 min PET scans. The PL samples were obtained as the supernatants following vigorous centrifugation of the WB samples. The radiometabolite profile of [¹⁸F]9 was mapped using the selected PL samples of 3, 5, 10, 15, 30, 60, 90, and 120 min. The percent parent in PL (%PP) was determined via the same automated column switching radioHPLC system as described for [¹⁸F]JNJ-46356479 (6). Similarly, the PL free fraction f_p of [¹⁸F]9 was determined via ultracentrifugation.

The concentration of radioactivity in WB and PL was measured in a well counter and was expressed as kBq/cc. The time courses of % PP(t) were fitted to a sum of two decaying exponentials plus a constant. By multiplying the resulting model fits with the time course of total PL radioactivity concentration, the metabolite-corrected arterial input function was generated for tracer kinetic modeling.

ABBREVIATIONS

AIC

Akaike information criteria

A _m

molar activity

BBB

blood–brain barrier

BME

basal medium Eagle

Bpin

boronic acid pinacol

DMA

dimethylacetamide

DMEM

Dulbecco’s modified Eagle’s medium

EOS

end of synthesis

f _p

plasma-free fraction

% ID/g

percentage of injected dose per gram of wet tissue

LCMS

liquid chromatography–mass spectrometry

MEMPRAGE

magnetization-prepared rapid gradient-echo

MLEM

maximum-likelihood expectation-maximization

NMT

NIMH macaque template

PAM

positive allosteric modulator

PDB

Protein Data Bank

PL

plasma

PLL

poly-l-lysine

RCY

radiochemical yield

SUV

standardized uptake value

TAC

time–activity curve

TEAB

tetraethyl ammonium bicarbonate

7-TM

seven transmembrane

USP

United States Pharmacopeia

VFTD

Venus flytrap domain

V_T

regional total volume of distribution

WB

whole-blood