Synthesis and Characterization of A New Positron Emission Tomography Probe for Orexin 2 Receptors Neuroimaging

Abstract

The orexin receptors (OXRs) have been involved in multiple physiological and neuropsychiatric functions. Identification of PET imaging probes specifically targeting OXRs enables us to better understand the OX system. Seltorexant (JNJ-42847922) is a potent OX₂R antagonist with the potential to be an OX₂R PET imaging probe. Here, we describe the synthesis and characterization of [¹⁸F]Seltorexant as an OX₂R PET probe. The ex vivo autoradiography studies indicated the good binding specificity of [¹⁸F]Seltorexant. In vivo PET imaging of [¹⁸F]Seltorexant in rodents showed suitable BBB penetration with the highest brain uptake of %ID/cc = 3.4 at 2 minutes post-injection in mice. The regional brain biodistribution analysis and blocking studies showed that [¹⁸F]Seltorexant had good binding selectivity and specificity. However, pretreatment with unlabelled Seltorexant and P-gp competitor CsA observed significantly increased brain uptake of [¹⁸F]Seltorexant, indicating [¹⁸F]Seltorexant could interact P-gp at the blood-brain barrier. Our findings demonstrated that [¹⁸F]Seltorexant is a potential brain OX₂R PET imaging probe, which paves the way for new OX₂R PET probes development and OX system investigation.

Graphical Abstract

Introduction

Orexins (also known as hypocretins), which are endogenous neuropeptides secreted by the hypothalamus, contains two subtypes of orexin-A and orexin-B.^{1, 2} Orexin-A and orexin-B have been found to interact with G protein-coupled receptors orexin 1 receptor (OX₁R) and orexin 2 receptor (OX₂R) respectively to regulate a variety of physiological functions, including energy homeostasis, sleep-wake cycle, stress response, and brain reward mechanisms.^2–4 Studies have shown that abnormalities in orexin signaling have a close relationship with many diseases, especially sleep disorders.⁵ Further research revealed that OX₁R and OX₂R have different expressions and distributions in the brain, suggesting their distinct physiological functions. For example, OX₁Rs are mainly located in the limbic system, paraventricular thalamic nucleus, and locus coeruleus regulate emotion responsible for reward and autonomic regulation.⁶ Conversely, OX₂Rs are exclusively expressed in the controlling arousal regions, which play a critical role in regulating sleep and wakefulness.⁷

Several lines of evidence indicate that the blockade of orexin receptors shows therapeutic potential for various diseases, primarily for insomnia treatment.^{4, 8–12} Subsequently, the development of OXR antagonists has aroused widespread interest in this field of research. In the past decade, a number of small-molecule OXR antagonists have been discovered. One of the most representative OXR antagonists is Suvorexant (MK-4305, Figure 1), a dual OXR antagonist (DORA), which has been approved by the U.S. Food and Drug Administration (FDA) for insomnia treatment.^{13, 14} Notably, an increasing number of selective OX₁R or OX₂R antagonists have been reported in recent years (Figure 1).¹⁵ Compared with dual OXR antagonists, selective OXR antagonists exhibited advantages such as better efficacy and fewer adverse effects. However, to date, the biological mechanism of OX₁R and OX₂R in CNS-related diseases has not been fully elucidated.^{15, 16}

Figure 1.

Chemical structures and binding affinities of representative dual orexin receptor antagonists (DORA) and selective orexin 1/2 receptor antagonists.

Molecular imaging is an advanced technique that enables us to obtain valuable biological information at the molecular level.¹⁷ For instance, in vitro autoradiography of [³H]EMPA, a selective OX₂R radioligand, has indicated the biodistribution and density of OX₂R in the rat brain.¹⁸ Positron emission tomography (PET) imaging provides a powerful tool for visualization and quantification of pharmacological targets in living subjects in a non-invasive way.¹⁹ PET imaging of orexin receptors with a suitable OXR radioligand could contribute to elucidate the biological function of the orexin system as well as facilitate novel selective OXR antagonists discovery. Although a series of OXR PET probes were synthesized and characterized in the past few years, limited success has been achieved. For example, Watanabe et al. reported F18-labeled tetrahydroisoquinoline derivatives OX₁R PET tracers with low brain uptake in the brain of mice.²⁰ The carbon-11 labeled OX₂R PET radioligands, including [¹¹C]CW4²¹, [¹¹C]EMPA²², and [¹¹C]9a²³ exhibited either limited brain uptake or unexpected non-specific binding in rodents and non-human primates (NHP) studies, which impeded their further translation for OX₂R imaging. Recently, we have reported a non-selective OXRs PET radioligand [¹¹C]CW24 with good brain uptake in rodents and NHP.²⁴ However, the moderate binding affinity and specificity of [¹¹C]CW24 for OXRs still need to be optimized. As a result, so far, no PET radioligand specifically targets OX₂R for neuroimaging in preclinical or clinical application.

Radiolabeling the potent known OX antagonist with short half-life isotopes (e.g., carbon-11 (t_1/2 = 20 minutes) and fluorine-18 (t_1/2 = 109 minutes)) is a practical and effective strategy to develop PET imaging probes specifically targeting OXR, such as [¹¹C]MK-1064²⁵, [¹¹C]CW4, and [¹¹C]EMPA. As part of our continuing effort for the OXRs PET radioligand development, we screened for the known selective OXR antagonists that can be labeled with radioisotopes. There are some criteria that we use to choose OXR antagonist candidates for radiolabelling. First, the binding affinity and selectivity for OXR must be good enough to measure the target precisely in vivo. Additionally, as our interested organ is the brain, the candidates should possess appropriate physicochemical properties to penetrate the blood-brain barrier (BBB). Seltorexant (JNJ-42847922), a selective OX₂R antagonist, was first discovered by Janssen Pharmaceutical Research & Development, LLC and is being in clinical studies for the treatment of insomnia and major depressive disorder.²⁶ In the preclinical studies^26–28, Seltorexant exhibited favorable pharmacological properties with good binding affinity and selectivity for OX₂R, good brain uptake, and suitable pharmacokinetic properties, which is an ideal candidate for radiolabeling and OX₂R PET imaging in the brain.

In this study, we describe the synthesis of the precursor for [¹⁸F]Seltorexant preparation. Comprehensive evaluations of [¹⁸F]Seltorexant, including binding specificity and brain permeability, were carried out by in vitro autoradiography and in vivo dynamic PET imaging in rodents, which provide significant information for future OX₂R PET radiotracers’ development.

Materials and methods

All commercially available chemical reagents and solvents were of ACS–grade purity or higher and directly used without further purification. The unlabelled Seltorexant, orexin 2 antagonist compound 30 and competitive P-gp inhibitor Cyclosporin A were purchased from MedChemExpress. Anhydrous Dimethylsulfoxide (DMSO) was purchased from Acros Organics. No-carrier-added ¹⁸F-fluoride was produced from water 97% enriched in ¹⁸O (Sigma-Aldrich®) by the nuclear reaction ¹⁸O(p, n)¹⁸F with a Siemens Eclipse HP cyclotron and a silver-bodied target at Athinoula A. Martinos Center for Biomedical Imaging.

All animal studies were carried out at Massachusetts General Hospital (PHS Assurance of Compliance No. A3596–01). The Subcommittee on Research Animal Care (SRAC) serves as the Institutional Animal Care and Use Committee (IACUC) for the Massachusetts General Hospital (MGH). SRAC reviewed and approved all procedures detailed in this paper.

Chemistry

2-iodo-6-nitrobenzoic acid (2) To a solution of Compound 1 (1.0 g, 5.49 mmol) in concentrated HCl (3 mL) and water (3 mL) was added NaNO₂(0.46 g, 6.59 mmol) in H₂O (3 mL) below 5 °C. Then the mixture was stirred at 5°C for 0.5 h and followed by adding KI (1.83 g, 10.98 mmol) in H₂O (3 mL), heated to 90°C for 16 h. The reaction mixture was quenched with saturated NaHSO₃ (50 mL), extracted with ethyl acetate (25 mL × 3). The aqueous layer was adjusted to pH = 2 by HCl aqueous (2 N), extracted with ethyl acetate (25 mL × 3). The combined organic layer was dried over Na₂SO₄, filtered, and concentrated afford the crude product. The crude product was purified using CombiFlash (0%–10% MeOH in DCM) to yield compound 2 (0.9 g, 56%). ¹H NMR (400 MHz, DMSO-d₆) δ 14.00 (br, 1H), 8.352 − 8.330 (dd, J = 0.80, 8.80 Hz, 1H), 8.233 − 8.210 (dd, J = 0.80, 8.00 Hz, 1H), 7.495 − 7.454 (t, J = 8.4, 8.0 Hz, 1H). LC-MS [M+H]⁺ : 293.9. 2-nitro-6-(2H-1,2,3-triazol-2-yl)benzoic acid (3) To a mixture of compound 2 (0.9 g, 3.07 mmol) and K₂CO₃ (1.06 g, 7.68 mmol) in THF (15 mL) was added CuI (0.06 g, 0.30 mmol) and N1,N2-Dimethylethane-1,2-diamine (0.06 g, 0.61 mmol). 2H-1,2,3-triazole (0.43 g, 6.14 mmol) in THF (2 mL) was added. The reaction mixture was stirred at 65 °C for 16 h. After the complete consumption of compound 2, the solvent was removed, the residue was acidified to pH = 3 before 200 mL water was added. The aqueous layer was extracted with ethyl acetate (25 mL × 2). The combined organic layers were dried over Na₂SO₄ and evaporated. The crude product was purified using CombiFlash (0%–10% MeOH in DCM) to obtain compound 3 (0.35 g, crude, 49%). LC-MS [M+H]⁺: 235.1. (5-(4,6-dimethylpyrimidin-2-yl)hexahydropyrrolo[3,4-c]pyrrol-2(1H)-yl)(2-nitro-6-(2H-1,2,3-triazol-2-yl)phenyl)methanone (4) To a solution of Compound 3 (350 mg, 1.49 mmol) in DMF (5 mL) was added DIPEA (580 mg, 4.48 mmol), HATU (853 mg, 2.24 mmol). The reaction mixture was stirred at room temperature for 30 min. Then 2-(4,6-dimethylpyrimidin-2-yl)octahydropyrrolo[3,4-c]pyrrole (358 mg, 1.64 mmol) was added. Then the reaction mixture was stirred at room temperature for 16 h. The resulting mixture was quenched with brine (15 mL) and extracted with ethyl acetate (25 mL × 2). The combined organic layers were dried over Na₂SO4, filtered, concentrated in vacuum. The residue was purified by column chromatography eluting with DCM / MeOH (20 : 1) to yield the precursor 4 (120 mg, 18.5%). ¹H NMR (400 MHz, CDCl₃) δ 8.320 − 8.239 (m, 3H), 8.003 (s, 1H),7.931 − 7.885 (m, 1H), 6.417 − 6.407 (d, J = 4.00 Hz, 1H), 3.764 − 3.319 (m, 6H), 3.147 − 3.055 (m, 1H), 3.046 − 2.690 (m, 3H), 2.259 − 2.231(m, 6H). LC-MS [M+H]⁺: 435.2.

Log D Determination.

The general procedure for Log D determination assays as described in our previous literature.^{29, 30} Briefly, 10 μL Seltorexant DMSO solution (10 mM) was mixed with 100 μL octanol in PBS buffer (pH 7.4) and followed by shaking in a rotator for 1 hour at 30 rpm. After incubations, the concentrations of Seltorexant in octanol and water were measured by LC-MS (Agilent 6310 ion trap mass spectrometer). Then the Log D_7.4 was calculated by Log [the ratio between the amount of test compound in n-octanol and PBS]. The assay was performed in triplicate.

Radiochemistry

The ¹⁸F-Fluoride solution (97% ¹⁸O enriched water, Sigma-Aldrich) was generated from the cyclotron (Siemens Eclipse HP) and directly loaded onto a QMA anion exchange cartridge (Chromafix^® 30-PS-HCO₃⁻). The cartridge with [¹⁸F]fluoride was washed with water and eluted into a vial with 3.5 mg/ml Kryptofix 2.2.2 (K₂₂₂) solution in MeCN (1.2 mL) and 2.5 mg/ml K₂CO₃ solution in water (0.3 mL). Removal of the water was accomplished by azeotropic evaporation with anhydrous MeCN (1 ml each) under N₂ three times. After the [¹⁸F]fluoride solution dried down, precursor 4 (2 mg) in DMSO (1 mL) was added to the vial and it was heated to 150 °C for 20 min. After adding 1.0 mL water, it was purified by injecting the solution to a semi-preparative HPLC (Agilent Eclipse XDB-C18 (5 μm, 250 mm × 9.4 mm), with a mobile phase of 68% H₂O + 0.1% TFA/32% CH₃CN, at the flow rate of 5.0 mL/min). [¹⁸F]Seltorexant was further reformulated by loading onto a solid-phase exchange (SPE) C-18 SepPak cartridge, rinsed with 5 mL water, 1 mL ethanol, and diluted by saline solution (0.9%, 9 mL). The radiochemical yield of [¹⁸F]Seltorexant was 7 – 9% (non-decay corrected, n = 6) and purity > 97% (measured with HPLC equipped with a UV detector and a gamma detector), and the specific activity (A_s) of [¹⁸F]Seltorexant was 110 – 130 GBq/μmol (EOS).

In vitro Autoradiography

The in vitro autoradiography assay has been described in our previous report.³¹ Briefly, the mice brain sections (sagittal, 20 μM) were pre-incubated with Tris-HCl buffer (50 mM) solution for 20 min, followed by incubation with [¹⁸F]Seltorexant (1 mCi/L, 50 mM Tris-HCl buffer). For blocking studies, unlabeled Seltorexant (10 μM) or compound 30 (10 μM) with radiotracer mixed in the incubation solution. Following incubation, mice brain sections were washed in ice-cold buffer and then dipped in ice-cold distilled water. Then the mice brain sections were dried at ambient temperature. An imaging plate (BAS-MS2025, GE Healthcare, NJ, USA) was exposed to the dried brain sections. Autoradiograms were obtained, and ROIs were carefully drawn with the reference of naked-eye observation. The gray value of images was measured by ImageJ software.

Rodent PET/CT acquisition

The general procedure for rodent PET/CT imaging studies was described previously.^{30, 32} Briefly, male C57BL/6 mice (6-month, n = 4 for each test group) were arranged in a Triumph PET/CT scanner (Gamma Medica, Northridge, CA) with inhalational isoflurane anesthesia (2%, Patterson Vet Supply, Inc., Greeley, CO, USA). The mice were administrated with 0.2 mL of saline with 5% [¹⁸F]Seltorexant via a lateral tail vein catheter at the beginning of PET acquisition. For the blocking study, mice were pretreatment with unlabeled Seltorexant (1.0 mg/kg) or compound 30 (1.0 mg/kg), or Cyclosporin A (0.5 mg/kg) before [¹⁸F]Seltorexant injection. Each PET scan lasted for 60 minutes and was followed by computed tomography (CT). PET data were reconstructed using a 3D-MLEM method resulting in full width at a half-maximum resolution of 1 mm.

Rodent PET/CT Image Analysis.

The reconstructed PET and CT images were processed by PMOD software (PMOD Technologies Ltd., Zurich, Switzerland). Anatomical volumes of interest (VOIs) were generated manually in spheres under the guide of high-resolution CT structural images. Time-activity curves (TACs) were exported as decay-corrected activity per unit volume. The TACs were expressed as percent injected dose per unit volume (%ID/cc) for analysis.

Results and discussion

Pharmacological and Physicochemical Properties of Seltorexant

For a CNS PET radiotracer, binding potential (BP, the ratio of receptor’s density (B_max) to ligand-receptor binding affinity (K_i)) is crucial for visualization and quantitation of the objective target. Generally, the BP of an ideal CNS PET radioligand needs to be greater than 5.³³ Seltorexant has been reported to exhibit high binding affinity and selectivity against OX₂R, with K_i value of 10 nM and 80-fold selectivity for OX₂R over OX₁R.²⁸ The B_max of OX₂R in the rat brain was tested within 40–140 nM,²¹ suggesting Seltorexant can be a potential PET imaging agent for OX₂R imaging. Several physicochemical properties such as molecular weight, total polar surface area (tPSA), and lipophilicity can be used to predict the brain permeability of a candidate compound prior to radiolabeling. The tPSA and cLogP values of Seltorexant were determined in silico prediction with ChemBioDraw 16.0, and the Log D value was evaluated by means of the “shake flask method” (Figure 2). These physicochemical properties of Seltorexant are within the favorable range for sufficient brain permeability. Indeed, the brain permeability and binding specificity of Seltorexant were also evaluated by Bonaventure et al. by in vitro autoradiography study with [³H]EMPA.²⁶ After orally administrated 30 mg/kg Seltorexant, high brain concentration and OX₂R occupancy were observed in rat brain. As such, the pharmacokinetic and physicochemical properties of Seltorexant make it a potential PET probe candidate for OX₂R imaging in the brain.

Figure 2.

The pharmacological and physicochemical properties of Seltorexant. ^acLogP value of Seltorexant is calculated by ChemBioDraw 16.0; Log D values were quantified in n-octanol/phosphate buffer (pH 7.4) by the shake-flask method. ^bLetavic et al. reported the in vitro binding affinity of Seltorexant.^{28 c}Bonaventure et al. reported the pharmacokinetics of Seltorexant; single oral administration (p.o.) of Seltorexant at 30 mg/kg showed high concentration in rat brain.²⁶

Chemistry and Radiochemistry

Aromatic nucleophilic radiofluorination is a commonly used method for the preparation of fluorine-18 labeled PET radiotracers.^{34, 35} The presence of fluorine atom in the structure of Seltorexant indicated it could be directly labeled with fluorine-18 without altering its molecular structure. Hence, for the radiosynthesis of [¹⁸F]Seltorexant, the radiolabeling precursor was synthesized by substituting the native fluorine atom of Seltorexant with a nitro group (as the leaving group), with amenability for fluorine-18 nucleophilic substitution. Precursor 4 was prepared by three steps (Scheme 1): 1) intermediate 2 was obtained by diazotization and iodization of starting material 1 in the presence of hydrochloric acid, sodium nitrite, and potassium iodide in 56% yield; 2) intermediate 2 was then reacted with 2H-1,2,3-triazole to give compound 3; 3) the nitro substitution precursor 4 for [¹⁸F]Seltorexant labeling was synthesized by condensation of intermediate 3 with 2-(4,6-dimethylpyrimidin-2-yl)octahydropyrrolo[3,4-c]pyrrole.

Scheme 1.

Synthesis of precursor 4. Reagents and conditions: (i) NaNO₂, KI, in H₂O, 90 °C, 16 h; (ii) 2H-1,2,3-triazole, N1,N2-Dimethylethane-1,2-diamine, K₂CO₃, CuI, THF, 90 °C, 16 h; (iii) 2-(4,6-dimethylpyrimidin-2-yl)octahydropyrrolo[3,4-c]pyrrole, DIPEA, HATU, DMF, room temperature, 16 h.

As outlined in Scheme 2, radiosynthesis of [¹⁸F]Seltorexant was carried out via reacting precursor 4 with [¹⁸F]fluoride in the presence of K₂CO₃ and Kryptofix K₂₂₂ and DMSO at 150 °C for 20 minutes. The reaction mixture was then injected into a reverse semi-preparative HPLC for purification. The desired fraction [¹⁸F]Seltorexant was collected and reformulated for animal injection. The validation of [¹⁸F]Seltorexant was performed by coinjection with the unlabeled Seltorexant (purchased for MedChemEpxress) on an analytical HPLC (Figure S1). The radiochemical yield (RCY) was 7 – 9% (decay corrected to end of cyclotron bombardment (EOB), n = 6) after 90 minutes of average synthesis time; the molar activity was 110 – 130 GBq/μmol, the radiochemical purity was ≥97%.

Scheme 2.

Radiosynthesis of [¹⁸F]Seltorexant. Radiolabeling condition: 4 (precursor, 3.0 mg), DMSO, K₂₂₂, K₂CO₃, [¹⁸F]F⁻, 150°C, 20 min.

Ex Vivo Autoradiography Study

The ex vivo autoradiography studies of [¹⁸F]Seltorexant for mice brain sections were carried out to evaluate the OX₂R binding specificity of [¹⁸F]Seltorexant. The structurally distinct OX₂R antagonist, 9-(4,6-dimethylpyrimidin-2-yl)-2-((5-methoxy-1H-indol-3-yl)methyl)-2,9-diazaspiro[5.5]undecan-1-one (Compound 30, the structure showed in Figure 1), was used for the blocking study. The representative baseline and blocking (co-incubation with 10 μM compound 30) autoradiography images of sagittal mice brain sections with [¹⁸F]Seltorexant are outlined in Figure 3. Heterogeneous distribution of [¹⁸F]Seltorexant signal in the baseline was detected. In the whole brain, the binding was decreased by 27% in the presence of OX₂R antagonist compound 30. Notably, a significant decrease in specific binding of [¹⁸F]Seltorexant in the cortical layer, hippocampus, thalamus, and striatum in the blocking was observed, indicating the specific binding sites for [¹⁸F]Seltorexant in these brain regions. The promising ex vivo results prompted us to move [¹⁸F]Seltorexant forward to an in vivo PET investigation.

Figure 3.

(A) Representative in vitro autoradiographic images of mice brains (sagittal); (B) the relative radioactive uptake in baseline and blocking. The baseline mice brain sections were treated with [¹⁸F]Seltorexant only, and the blocking mice brain sections were co-incubation with OX₂R antagonist compound 30 (10 μM). The gray value data were expressed as mean ± SD, n = 4; Asterisks indicate statistical significance. *p < 0.05, **p ≤ 0.01, and ***p ≤ 0.001.

Rodents PET-CT imaging of [¹⁸F]Seltorexant

Next, PET-CT imaging in mice (C57BL/6 male mice, 6-month old) was conducted to evaluate the in vivo properties of [¹⁸F]Seltorexant as an OX₂R PET probe. 100–150 μCi (0.1–0.15 mL, 31–46 pmol) of [¹⁸F]Seltorexant was administered by intravenous bolus injection and then followed a 60 minutes dynamic PET imaging scanning and 10 minutes computed tomography (CT). We first examined the brain permeability of [¹⁸F]Seltorexant. In general, for a CNS PET radiotracer, the required brain uptake (determined by injected dose per cc (%ID/cc)) within 5 min of injection in rodents is at least 0.1%.³³ In vivo PET imaging of [¹⁸F]Seltorexant in mice demonstrated that [¹⁸F]Seltorexant had a suitable BBB penetration with a maximum %ID/cc of 3.4 at 2 minutes post-injection in the whole brain (Figure 4B). The time−radioactivity curve (TAC) showed that [¹⁸F]Seltorexant can rapidly penetrate the brain, bind with OX₂R, and wash out gradually, suggesting reasonable brain kinetics. Additional biodistribution analysis of [¹⁸F]Seltorexant in mice brain was performed with FUSION module in image analysis software PMOD (PMOD 4.01, PMOD Technologies Ltd., Zurich, Switzerland). Relatively high radioactivity uptake was observed in OX₂R-rich regions, including the cortex and hippocampus. In contrast, low radioactivity uptake was found in the cerebellum, where OX₂R was less expressed (Figure 4C), corresponding with previous results on OX₂R imaging.²³ The regional brain biodistribution analysis results of [¹⁸F]Seltorexant demonstrated the OX₂R biodistribution in the brain regions.

Figure 4.

(A) Representative baseline and blocking PET/CT images of [¹⁸F]Seltorexant focus on the mice brain (0–60 min, the baseline is shown as horizontal, coronal, and sagittal plane; the blockings are shown as a sagittal plane); (B) the baseline and blocking time-activity curves of [¹⁸F]Seltorexant in the mice whole brain; (C) Time-radioactivity curves for [¹⁸F]Seltorexant in the hippocampus, cerebellum, thalamus, cortex, and striatum (n = 3); (D) the normalized baseline and blocking time-activity curves [¹⁸F]Seltorexant (normalized the brain uptake curves with the highest radioactivity in the blood at each time point).

In the blocking studies, mice were pre-administrated unlabeled Seltorexant (1.0 mg/kg) and compound 30 (1.0 mg/kg) 5 minutes before radiotracer administration to confirm the specific binding of [¹⁸F]Seltorexant. The TACs showed a remarkably increased brain uptake of [¹⁸F]Seltorexant when pretreated with blocking agents (Figure 4B). The blocking effects may cause the increased free radiotracer in plasma, resulting in increased total uptake of radiotracer in the brain compared with baseline. Consequently, we normalized the brain uptake with the highest radioactivity in the blood at each time point. From the normalized TAC curves, blocking effects were observed in both Seltorexant and compound 30 pretreated mice groups (Figure 4D). Together with in vitro autoradiography studies, these results indicated the binding specificity of [¹⁸F]Seltorexant for OX₂R.

The significantly increased brain uptake of [¹⁸F]Seltorexant in the self-blocking mice may be attributed to its interaction with efflux transporters, such as the P-glycoprotein (P-gp), located at the BBB. To verify this hypothesis, we carried out PET imaging studies of [¹⁸F]Seltorexant in mice pretreatment with the competitive P-gp inhibitor Cyclosporin A (CsA, 0.5 mg/kg). As we assumed, compared with baseline, significantly increased uptake of [¹⁸F]Seltorexant was observed in the CsA pretreatment mice brain, which indicated that Seltorexant could interact with P-gp at the BBB (Figure 5). Of the previously reported OXR PET probes, most had low brain penetration and failed to further translation for OX₂R imaging due to their interaction with P-gp, such as [¹¹C]EMPA²² and [¹¹C]9a²³. Though [¹⁸F]Seltorexant appears to interact with P-gp, it maintained good brain uptake and binding specificity, which could guide further OX₂R PET probes optimization and development.

Figure 5.

The representative baseline and Cyclosporin A pretreatment PET/CT images with [¹⁸F]Seltorexant focused on mice brain and time-activity curves of [¹⁸F]Seltorexant in the whole brain of mice.

Conclusion

In this work, we screened known selective OX₂R antagonists and found that Seltorexant has the potential to be an OX₂R imaging probe. We successfully synthesized the labeling precursor and obtained [¹⁸F]Seltorexant. Ex vivo autoradiography studies for mice brain sections indicated the good binding specificity of [¹⁸F]Seltorexant toward OX₂R. In vivo PET imaging in rodents showed that [¹⁸F]Seltorexant had a suitable BBB penetration for OX₂R imaging in the brain. The regional brain biodistribution analysis and blocking studies showed that [¹⁸F]Seltorexant had good binding selectivity and specificity. Pretreatment with competitive P-gp inhibitor CsA increased the brain uptake of [¹⁸F]Seltorexant, suggesting that [¹⁸F]Seltorexant might be the substrate of the efflux transporter. In conclusion, our preliminary results showed [¹⁸F]Seltorexant is a potential PET probe for OX₂R imaging in the brain, which could assist further OX₂R PET probes development.

• OX₂R PET imaging probe [¹⁸F]Seltorexant was successfully synthesized and characterized.

• [¹⁸F]Seltorexant exhibited good binding specificity for OX₂R in ex vivo autoradiography studies.

• In vivo PET imaging of [¹⁸F]Seltorexant in rodents showed good brain uptake and specificity for OX2R.