Tariquidar | Adenosinekinase Signal

Nuclear Medicine and Biology

journal homepage: www.elsevier.com/locate/nucmedbio
Nuclear Medicine and Biology 90–91 (2020) 10–14

An original radio-biomimetic approach to synthesize radiometabolites for PET imaging
Sylvain Auvity a,b, Louise Breuil a, Maud Goislard a, Michel Bottlaender a, Bertrand Kuhnast a,
Nicolas Tournier a, Fabien Caillé a,⁎
a Université Paris-Saclay, Inserm, CNRS, CEA, Laboratoire d’Imagerie Biomédicale Multimodale Paris-Saclay, 91401 Orsay, France
b Assistance Publique-Hôpitaux de Paris, Hôpital Necker – Enfants malades, Inserm, UMR-S 1144, Université de Paris, Optimisation thérapeutique en neuropsychopharmacologie, Paris, France

a r t i c l e i n f o

Article history:
Received 18 June 2020
Received in revised form 27 July 2020
Accepted 13 August 2020

Keywords: Carbon-11 Biomimetic PET imaging
Nor-buprenorphine
a b s t r a c t

To fully exploit the potential of positron emission tomography (PET) imaging to assess drug distribution and pharmacokinetics in the central nervous system, the contribution of radiometabolites to the PET signal has to be determined for correct interpretation of data. However, radiosynthesis and extensive study of radiometabolites are rarely investigated and very challenging for complex drugs. Therefore, an original radio- biomimetic (RBM) approach was developed to rapidly synthesize radiometabolites and non-invasively investi- gate their kinetics with PET imaging. This method enabled the challenging radiosynthesis of [11C]nor- buprenorphine ([11C]nor-BUP), the main metabolite of buprenorphine (BUP) which has been identified as a sub- strate of the P-glycoprotein (P-gp) transport function at the blood-brain barrier (BBB). Biomimetic conditions using cytochromes P450 3A4 to convert BUP into nor-BUP were optimized taking into account the short half- life of carbon-11 (t1/2 = 20.4 min). Those conditions afforded 32% of conversion within 20 min and were applied to the biomimetic radiosynthesis of [11C]nor-BUP from [11C]BUP. Automated radiosynthesis of [11C]BUP accord- ing to a procedure described in the literature followed by optimized RBM conditions afforded [11C]nor-BUP in 1.5% decay-corrected radiochemical yield within 90 min and 90 ± 15 GBq/μmol molar activity. HPLC quality con- trol showed chemical and radiochemical purities above 98%. To demonstrate the applicability of the RBM ap- proach to preclinical studies, brain PET images in rats showed a drastic lower uptake of [11C]nor-BUP (0.067 ± 0.023%ID/cm−3) compared to [11C]BUP (0.436 ± 0.054%ID/cm−3). P-gp inhibition using Tariquidar increased the brain uptake of [11C]nor-BUP (0.557 ± 0.077%ID/cm−3).

1. Introduction

Positron emission tomography (PET) is currently the most advanced minimally invasive and quantitative technique to study drug distribu- tion to the brain and provide insight into the interaction of drugs with central nervous system (CNS) and peripheral targets in humans. PET- assisted pharmacokinetic imaging of drugs mainly relies on the feasibil- ity of their isotopic labelling using positron emitters such as carbon-11 or ﬂuorine-18. Development of target-specific PET probes for quantita- tive imaging is often a challenge. Availability of such probes is essential since it provides an elegant and pragmatic way to non-invasively ad- dress the ability of a drug to interact with intended and/or off targets within the human brain. PET probes or isotopologues of a drug are therefore unique tools to accelerate the drug development and allow for early assessment of in vivo biodistribution, target engagement and optimal dose-response of investigational compounds in humans [1,2].

* Corresponding author.
E-mail address: [email protected] (F. Caillé).
A major pitfall in the use of radiopharmaceuticals for PET imaging is the unpredictable impact of radiolabelled metabolites on the quantification and interpretation of kinetic data. Indeed, PET signal associated with radiometabolites, i.e. metabolites that carry the positron emitting iso- tope, cannot be discriminated from the PET signal associated with the parent radioactive compound in tissues. Indeed, radiometabolites may show different and/or overlapping distribution to the target tissue and have different affinity for the molecular target than the parent drug, and therefore “contaminate” the PET signal and impair the pharmacoki- netic modelling of data [3,4]. Radiometabolites can be easily discrimi- nated from the parent compound in plasma using separative analytical techniques. To this end, radio-chromatography methods are performed on plasma sample obtained at selected time-points. These data provide information regarding the extent of peripheral metabolism and allow for determination of the metabolite-corrected input function. Molecular identification of radiometabolites found in plasma is also possible but radiometabolite analysis in the brain is only possible in rodents where radio-chromatography of brain lysate removed at selected times or brain microdialysis experiments can be performed.
However, these
https://doi.org/10.1016/j.nucmedbio.2020.08.001 0969-8051/© 2020 Elsevier Inc. All rights reserved.
approaches do not provide relevant information regarding the ability of radiometabolites to accumulate in target tissues, interact with targets, and account for the PET signal. These data, only provided by radiometabolite PET imaging, are therefore of prime interest to quantify PET images of the initial radiotracer more accurately in the tissues of interest.
An elegant approach to non-invasively and quantitatively assess the impact of radiometabolites is to synthesize them and study their kinet- ics using PET imaging [5,6]. This approach is rarely investigated and ex- tremely challenging to apply to chemically complex drugs from synthesis and radiosynthesis perspective. The most straightforward method to synthesize metabolites is to follow a biomimetic route where the metabolite of interest is obtained directly from the parent compound using the enzyme responsible for its in vivo transformation. Biocatalysis is an emerging approach to incorporate PET radionuclides such as carbon-11, nitrogen-13 or ﬂuorine-18 into molecules of biolog- ical interest [7]. However, this method has not been applied to the pro- duction of radiometabolites by the cytochromes P450 family (CYP). CYPs predominantly mediate the oxidative metabolism of small mole- cule radiotracers used in pharmacological imaging studies [8]. In vitro metabolism experiments are inexpensive and readily carried out. How- ever, duration of incubation has to be limited given the short radioactive half-life of radioisotopes such as carbone-11 (t1/2 = 20.4 min), the most relevant positron emitter for isotopic labelling of drugs.

We describe herein an original straightforward method to synthe-
size radiometabolites using a radio-biomimetic (RBM) approach for PET imaging purpose. The parent molecule was first isotopically radiolabelled with carbon-11 and exposed to CYPs to allow for the bio- transformation reaction using relevant enzymes. As a proof-of-concept, we applied the RBM approach to the first radiosynthesis of [11C]nor- buprenorphine ([11C]nor-BUP), the major metabolite of buprenorphine (BUP), which brain distribution remains to be elucidated in vivo as P- glycoprotein (P-gp, ABCB1) transport protein may restrict the brain penetration of nor-BUP [11–13]. BUP is a substrate of CYP3A4 in the liver which mediates its N-dealkylation into nor-BUP (Fig. 1) [9,10]. The isotopic radiolabelling of [11C]BUP has already been described in the litterature [14,15] and used in monkeys to predict the feasibility of opioid receptors PET imaging in humans [16] but [11C]nor-BUP has never been radiolabelled with carbon-11 due to the chemical complex- ity of this thebaine derivative. We have established an optimized proto- col for the bioinspired synthesis of nor-BUP using human CYP3A4 (hCYP3A4) allowing for an effective biotransformation of BUP into nor-BUP in a time compatible with the short half-life of carbon-11. This protocol was applied under radioactive conditions to set up the RBM approach and successfully synthesize [11C]nor-BUP.

As a proof of concept for the applicability of the RBM approach to PET imaging, the first in vivo brain images of [11C]nor-BUP were realized in the absence and in the presence of tariquidar (TQD), a well-characterized P-gp in- hibitor [12]. This study demonstrates the ability of the RBM approach to produce and study the brain kinetics of radiometabolites that are hardly amenable for direct radiolabelling.
hCYP3A4 assisted metabolism of BUP into nor-BUP by N-dealkylation. Both BUP and nor-BUP display a methoxy moiety for isotopic labelling with carbon-11.

2. Experimental section

2.1. Chemistry

Unless otherwise stated, all chemicals were purchased from Sigma- Aldrich and used as received. Nor-BUP analytical samples were kindly provided by Pr. S. Cisternino (UMR-S 1144, Université de Paris). Human supersomes™ (mixture of CYP3A4, oxidoreductase and CYPb5) were purchased from Corning Life Sciences, aliquoted and kept at
−80 °C when received. 3-O-Trityl-6-O-desmethyl-buprenorphine (pre-
cursor 1) was purchased from ABX chemicals.

2.1.1. General procedure for the non-radioactive biotransformation of BUP into nor-BUP
Variable quantities (100 μL, 200 μL or 300 μL) of human
supersomes™ (1 nmol/mL of CYP3A4, 8.5 nmol/mL of CYPb5) were mixed with nicotinamide adenine dinucleotide phosphate (NADP, 25 mg/nmol of CYP3A4), glucose-6-phosphate (G6P, 25 mg/nmol of CYP3A4), glucose-6-phsophate dehydrogenase (G6PDH, 50 U/nmol of CYP3A4) and 17 μL of a solution of BUP (0.59 mM in H2O/CH3CN 1/1 v/v) and the volume of the reaction was completed to 1 mL using phos- phate buffer (0.1 M, pH 7.4). The mixture was stirred at 37 °C. At differ- ent time points (5, 10, 15, 20, 30 and 40 min), 100 μL of the reaction mixture were withdrawn and CH3CN (50 μL) was added to precipitate the CYPs before centrifugation (2054 g, 5 min). The supernatant was an- alyzed by HPLC using a 717plus Autosampler system equipped with a 1525 binary pump and a 2996 photodiode array detector (Waters, USA). The system was monitored with the Empower 3 (Waters) soft- ware. HPLC were realized on a reverse phase analytical Symmetry C18 (150 × 3.9 mm, 5 μm, Waters) column using a mixture of H2O/CH3CN/ PicB7® (75/25/0.25 v/v/v, 2 mL/min) as eluent. UV detection was per- formed at 215 nm. The conversion rate was calculated by comparing the area under the curve of each compound.

2.2. Radiochemistry

All radiochemical yields and molar activities are decay-corrected from the end of bombardment.

2.2.1. [11C]BUP radiosynthesis
[11C]BUP was synthesized from 3-O-trityl-6-O-desmethylbupren orphine (ABX) in two steps using a TRACERlab™ FX C Pro (GE Healthcare, Uppsala, Sweden) module according to the method described by Luthra et al. [14] with slight modifications. No carrier added [11C]CO2 (45–55 GBq) was produced via the 14N(p,α)11C nuclear reaction by irra- diation of a [14N]N2 target containing 0.15–0.5% of O2 on a cyclone 18/9 cyclotron (IBA, Belgium). [11C]CO2 was subsequently reduced to [11C] CH4 and iodinated to [11C]CH3I following the process described by Larsen et al. [17] and finally converted to [11C]CH3OTf according to the method of Jewett [18]. [11C]CH3OTf was bubbled into a solution of 3-O-trityl-6-O- desmethylbuprenorphine 1 (1 mg) and sodium hydride (1 mg) in DMF (300 μL) at −20 °C for 3 min The mixture was heated at 100 °C for 2 min and HCl (3 M, 200 μL) was added. The mixture was further heated at 100 °C for 1 min After cooling to 60 °C, a mixture of ammonium formate (0.1 M) and acetonitrile (0.5 mL, 55/45 v/v) was added. Purification was realized by semi-preparative reverse phase HPLC (Waters Symmetry® C18 7.8 × 300 mm, 7 μm) with ammonium formate (0.1 M)/acetonitrile (55/45 v/v, 5 mL/min) as eluent. UV (λ = 220 nm, K2501, Knauer, Germany) and gamma detection (LB513, Berthold, France) were per- formed (see Fig. S6 in supporting information). The purified compound was diluted with water (20 mL) and passed through a Sep-Pak® C18 car- tridge (Waters). The cartridge was eluted with ethanol (2 mL) and the sol- vent was evaporated to dryness at 120 °C under vacuum for 3 min to afford [11C]BUP (1.8 ± 0.4 GBq, 15% RCY, n = 10) within 45 min with ra- diochemical and chemical purities above 98% and a molar activity of 110 ± 20 GBq/μmol (n = 10).

2.2.2. Radio-biomimetic synthesis of [11C]nor-BUP
To [11C]BUP (1.8 ± 0.4 GBq) was added a solution of 0.1 M phos- phate buffer (1 mL, pH 7.4) containing human supersomes™ (0.2 nmol of CYP3A4 and 1.7 nmol of CYPb5), NADP (5 mg), G6P (5 mg) and G6PDH (10 U). The RBM reaction was carried out at 37 °C for 20 min with gentle stirring. Acetonitrile (500 μL) was added to stop the reaction and the mixture was centrifuged (500g, 30 s). The su- pernatant was purified by semi-preparative reverse phase HPLC (Wa- ters Symmetry® C18 7.8 × 300 mm, 7 μm) with ammonium formate (0.1 M)/acetonitrile (55/45 v/v, 5 mL/min) as eluent using both UV (λ = 220 nm, K2501, Knauer) and gamma detection (LB513, Berthold). The purified compound was diluted with water (20 mL) and passed through a Sep-Pak® C18 cartridge light (Waters). The cartridge was eluted with ethanol (1 mL) which was evaporated to ca. 100 μL before dilution with 0.9% NaCl solution (1 mL). Ready-to-inject [11C]nor-BUP (35 ± 10 MBq, 10% RCY from [11C]BUP, 1.5% from the end of bombard- ment, n = 10) was obtained within 45 min from the end of the [11C]BUP radiosynthesis (90 min from the end of bombardment) in above 98% ra- diochemical and chemical purities and a molar activity of 90 ± 15 GBq/ μmol (n = 10).

2.2.3. Quality Control
Quality control was performed by HPLC on three consecutive runs for each batch of both [11C]BUP (Fig. S7 in supporting information) and [11C]nor-BUP (Fig. S8 in supporting information) following the same protocol as for the analysis of the non-radioactive biotransforma- tion of BUP into nor-BUP. Identification of the peak was assessed by comparing the retention time of the labelled compound with the reten- tion time of the non-radioactive reference (tRref). For acceptance, the re- tention time must be within the tRref ± 10% range. Radiochemical and chemical purities were calculated as the ratio of the area under the curve (AUC) of the peak of BUP or nor-BUP over the sum of the AUCs of all other peaks on gamma and UV chromatograms respectively. Ra- diochemical and chemical purities are the mean values of three consec- utive runs. Molar activity was calculated as the ratio of the activity of the collected peak of [11C]BUP or [11C]nor-BUP (Capintec®, Berthold) over the molar quantity of BUP of nor-BUP respectively, determined using calibration curves. Molar activity is calculated as the mean value of three consecutive runs.

2.3. PET imaging

2.3.1. Animals
PET imaging was performed in 7 male Sprague-Dawley rats (355 ± 184 g) housed in a controlled environment (22 ± 3 °C; 55 ± 10% rela- tive humidity) and a 12 h dark/light cycle, with access to food and tap water ad libitum. All animal use procedures were in accordance with the recommendations of the European Community for the care and use of laboratory animals (2010/63/UE) and the French National Com- mittees (French decret 2013-118).

2.3.2. PET scans
One rat was included per experimental condition: rat 1 received
47.8 MBq of [11C]BUP, rat 2 received 12.1 MBq of [11C]nor-BUP andrat 3 received 13.6 MBq of [11C]nor-BUP 15 min after tariquidar ad- ministration (i.v, 8 mg/kg). Tariquidar dimesylate hydrate (10 mg, ChemScene LLC) was dissolved in aqueous glucose solution (10% w/ v, 0.5 mL) and water for injection (1.5 mL) was added. A volume of0.5 ± 0.3 mL of this solution was injected as a bolus in the rat tail vein 15 min prior to the radiotracer injection. PET scans were per- formed on a μPET camera (Inveon®, Siemens, Germany) dedicated for small animals. Anesthesia was induced and thereafter maintained using 1.5–3% isoﬂurane in O2. After radiotracer injection in the tail vein, PET images were acquired during 15 min, and then recon- structed with the FORE + OSEM2D algorithm including normaliza- tion, attenuation, scatter, and random corrections. PET images were summed over the 15 min of acquisition. Mean brain radioactive con- centrations were obtained using Pmod® software (version 3.9, Switzerland). The volume of interest (VOI) was manually drawn using Pmod® so that the extracted volume only includes the brain volume, avoiding any partial volume effect due to the signal of adja- cent regions. The extracted VOI is the same across tested conditions. Results were corrected for the injected activity and expressed as the percentage of injected dose per cubic centimetre (% ID/cm−3).

3. Results and discussion

3.1. Optimization of the biotransformation of BUP into nor-BUP

One of the major challenge of the RBM approach with carbon-11 is the short half-life of the radioisotope (t1/2 = 20.4 min). Incubation conditions to allow for biotransformation using hCYP3A4 were therefore optimized to meet a compromise between the maximum conversion of BUP into nor-BUP in the shortest duration of incuba- tion time (Table 1 and Figs. S1 to S5 in supporting information). Incu- bation conditions were optimized starting from nanomolar concentrations of BUP to be consistent with the amount of labelled [11C]BUP obtained after radiolabelling (typically [11C]BUP is ob- tained with a molar activity of 110 GBq/μmol). 18% of BUP were bio-transformed into nor-BUP using 0.1 nmol of hCYP3A4 and
0.85 nmol of cytochrome b5 (CYPb5) at 37 °C for 20 min. 2% of other unknown metabolites were formed (entry 1, Table 1). After 20 min incubation, the conversion rate started to slow down to reach only 26% after 40 min of incubation. An incubation duration of 20 min, which corresponds to approximatively one radioactive half-life of carbon-11, was therefore considered for optimal RBM conditions. Higher incubation temperature of 50 °C slightly increased the con- version rate for short incubation times (less than 10 min) but com- peted with the degradation of cytochromes observed after longer time (entry 2, Table 1). A 2-fold increase in the amount of hCYP3A4 and CYPb5 (0.2 and 1.7 nmol respectively) improved the conversion up to 32% of nor-BUP after 20 min of incubation at 37 °C (entry 3, Table 1). Further increase in the amount of enzymes tented to de- crease the conversion of nor-BUP while promoting the formation of other metabolites (entry 4, Table 1). Finally, the presence of CYPb5 was found essential for effective formation of nor-BUP (entry 5, Table 1).

Table 1
Conditions investigated for the optimization of the non-radioactive biotransformation of BUP into nor-BUP.

Entry hCYP3A4a
(nmol) CYPb5 (nmol) T (°C) Nor-BUP conversion at 20 min Nor-BUP conversion at 40 min Other metabolites at 20 min Other metabolites at 40 min
1 0.1 0.9 37 18% 26% 2% 2%
2 0.1 0.9 50 17% 21% 1% 1%
3 0.2 1.7 37 32% 46% 3% 4%
4 0.3 2.6 37 22% 43% 6% 15%
5 0.2 0 37 1% 5% 0% <1%
a All reactions were carried out in the presence of nicotinamide adenine dinucleotide phosphate (25 mg per nmol of hCYP3A4), glucose-6-phosphate (25 mg per nmol of hCYP3A4) and glucose-6-phsophate dehydrogenase (50 U per nmol of hCYP3A4) as cofactors.

Two-step radiosynthesis of [11C]nor-BUP from commercially available precursor 1 using the RBM approach. The biotransformation step was realized following the optimized conditions depicted in Table 1.

3.2. Radiosynthesis of [11C]nor-BUP by the RBM approach

The optimized conditions were applied to the biotransformation of [11C]BUP into [11C]nor-BUP. [11C]BUP was radiosynthesized from cyclotron-produced [11C]CO2 using a TRACERlab FX C Pro module fol- lowing a two-step procedure described in literature [14] using [11C] methyl triﬂate instead of [11C]methyl iodide (Scheme 1). Brieﬂy, com- mercially available precursor 1 was radiomethylated in the presence of [11C]CH3OTf and sodium hydride in DMF at 100 °C for 2 min followed by deprotection of the trityl (Trt) moiety with 3 M hydrochloric acid. After HPLC purification (see Fig. S6 in supporting information), [11C] BUP (1.8 ± 0.4 GBq, n = 10) was obtained in 15% decay corrected radio- chemical yield (RCY) and 110 ± 20 GBq/μmol (n = 10) decay-corrected molar activity (MA) within 45 min [11C]BUP was formulated in ethanol (2 mL) which was further evaporated prior to the biotransformation re- action with hCYP3A4. Indeed, CYP are sensitive to the presence of or- ganic solvents. González-Pérez et al. demonstrated that only 1 or 2% of most organic solvents significantly decreased the activity of hCYP3A4 [19]. Ethanol represents the best compromise between minimizing the deleterious action on CYP and the possibility for rapid evaporation. The biotransformation reaction was carried out according to the previ- ously determined optimal conditions (Scheme 1).

Biotransformation was stopped by addition of acetonitrile to precipitate the cytochromes. After centrifugation, reverse phase HPLC purification was performed followed by solid-phase extraction formulation. Ready-to-inject [11C] nor-BUP (35 ± 10 MBq, n = 10) was obtained in 10% RCY from [11C] BUP (1.5% from [11C]CO2) and 90 ± 15 GBq/μmol (n = 10) MA, decay-corrected from the end of bombardment, within 45 min from the end of [11C]BUP radiosynthesis. The identity of [11C]nor-BUP was
confirmed by HPLC during control quality according to the European Pharmacopeia recommendations. A conversion of 26% of [11C]BUP into [11C]nor-BUP was observed and is consistent with results observed starting from non-radioactive buprenorphine (see Fig. S8 in supporting information). About 10% of the activity remained in the pellets after cen- trifugation, which may be explained by hydrophobic interactions of both BUP and nor-BUP with the CYP proteins. Despite moderate overall yields explained by the partial conversion of incubated [11C]BUP, suffi- cient amounts of [11C]nor-BUP were obtained to perform PET imaging experiments in rats. This demonstrates the feasibility of the RBM ap- proach to produce radiometabolites and study their brain kinetics in vivo.

3.3. Brain PET imaging

Brain PET imaging was performed in male Sprague-Dawley rats using an Inveon® μPET scanner (Fig. 2). Animals were injected with ei- ther [11C]BUP (47.8 MBq, n = 1, see Fig. S7 for HPLC chromatogram of [11C]BUP) or [11C]nor-BUP (12.1 MBq, n = 1). P-gp was inhibited using TQD (8 mg/kg) injected 15 min before [11C]nor-BUP (13.6 MBq, n = 1). Mean brain radioactive concentrations were 0.436 ± 0.054, 0.067 ± 0.023 and 0.557 ± 0.077%ID/cm−3 for [11C]BUP, [11C]nor-BUPand [11C]nor-BUP + TQD respectively. Four other animals were injected with [11C]BUP and [11C]nor-BUP without TQD and results are presented as supporting information data (Table S1). [11C]BUP showed significant brain penetration with predominant binding in cortical regions and the striatum, consistent with the expression of opioid receptors in these re- gions (Fig. 2A). [11C]nor-BUP poorly entered the brain when P-gp was fully functional (Fig. 2B). TQD was injected at 8 mg/kg 15 min before

PET summed images from 0 to 15 min after injection of A) [11C]BUP (47.8 MBq), B) [11C]nor-BUP (12.1 MBq) and C) TQD (8 mg/kg) followed by [11C]nor-BUP (13.6 MBq) 15 min later. The dotted white circles represent the VOI manually drawn using Pmod® and used to extract the brain radioactive concentration.
[11C]nor-BUP injection to inhibit the P-gp at the rat BBB [12]. P-gp inhi- bition drastically increased up to 7-fold the brain uptake of [11C]nor- BUP(Fig. 2C), a result in accordance with the known avidity of nor- BUP for the P-gp. These preliminary results suggest that the brain PET signal associated with [11C]BUP reﬂects the tissue concentration of the parent compound with negligible contribution of [11C]nor-BUP and seems to confirm that the P-gp is a major rate limiting factor for the brain distribution of nor-BUP. Although these experiments have been conducted on an insufficient number of animals to draw statistical con- clusions, our approach may be relevant to explore the neuropharmaco- logical fate of metabolites which is often neglected as it may account for overall therapeutic efficacy and toxicity of drugs.

4. Conclusion

The RBM concept developed herein is an original and efficient method to synthesize radiometabolites for PET imaging. This approach was successfully applied to a small molecule radiotracer labelled with carbon-11, a short-lived isotope, demonstrating the feasibility of the method. The RBM approach was applied to the concrete case of [11C] nor-BUP and proved to offer sufficient amount of radiotracer to be injected in rats to record the first in vivo PET images of [11C]nor-BUP. This strategy brought the first in vivo brain PET images of [11C]nor- BUP, demonstrating the applicability of the method. The RBM approach could therefore be extended to the synthesis of other radiometabolites, regardless the radioisotope, as part of the in vivo validation of candidate radiotracers. Moreover, availability of kinetic data of radiometabolites in the brain may be used to enrich pharmacokinetic models and improve PET images quantification for existing or newly developed radioligands. The case of BUP is somehow ideal because almost only one compound is formed through metabolic transformation in the liver. However, with the possibility to perform radioHPLC at the end of RBM synthesis, the method could be applied to more complex biostransformations, usually performed by one or a combination of cytochromes of the P450 family. The RBM approach could therefore be applied to radiotracers presenting radiometabolites like [18F]DPA714 or [11C]UCB-J among many others.

Acknowledgments

This work was supported by the CEA intramural funding and France Life Imaging. The authors also thank Pr. S. Cisternino (UMR-S 1144, Université de Paris) for kindly providing nor-BUP reference samples.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi. org/10.1016/j.nucmedbio.2020.08.001.

References
[1] Bergström M, Långström B. In imaging in drug discovery and early clinical trials. Progress in drug research; Birkhäuser Basel; 2005. p. 279–317.
[2] Aboagye EO, Price PM, Jones T. In vivo pharmacokinetics and Tariquidar pharmacodynamics in drug development using positron-emission tomography. Drug Discov Today. 2001; 6:293–302.
[3] Saba W, Peyronneau MA, Dollé F, Goutal S, Bottlaender M, Valette H. Difficulties in dopamine transporter radioligand PET analysis: the example of LBT-999 using [18F] and [11C] labelling: part I: PET studies. Nucl Med Biol. 2012;39:227–33.
[4] Peyronneau MA, Saba W, Dollé F, Goutal S, Coulon C, Bottlaender M, et al. Difficulties in dopamine transporter radioligand PET analysis: the example of LBT-999 using [18F] and [11C] labelling: part II: metabolism studies. Nucl Med Biol. 2012;39: 347–59.
[5] Gourand F, Emond P, Bergström JP, Takano A, Gulyás B, Guilloteau D, et al. A radiometabolite study of the serotonin transporter PET radioligand [11C]MADAM. Nucl Med Biol. 2014;41:501–6.
[6] Andrée B, Halldin C, Pike VW, Gunn RN, Olsson H, Farde LJ. The PET radioligand [car- bonyl-11C]desmethyl-WAY-100635 binds to 5-HT1A receptors and provides a higher radioactive signal than [carbonyl-11C]WAY-100635 in the human brain. J Nucl Med. 2002;43:292–303.
[7] da Silva ES, Gómez-Vallejo V, López-Gallego F, Llop J. Biocatalysis in radiochemistry: enzymatic incorporation of PET radionuclides into molecules of biomedical interest. J Label Compd Radiopharm. 2018;61:332–54.
[8] Jia L, Liu X. The conduct of drug metabolism studies considered good practice (II): in vitro experiments. Curr Drug Metab. 2007;8:822.
[9] Kobayashi K, Yamamoto T, Chiba K, Tani M, Shimada N, Ishizaki T, et al. Human buprenorphine N-dealkylation is catalyzed by cytochrome P450 3A4. Drug Metab Dispos. 1998;26:818–21.
[10] Picard N, Cresteil T, Djebli N, Marquet P. In vitro metabolism study of buprenorphine: evidence for new metabolic pathways. Drug Metab Dispos. 2005; 33:689–95.
[11] Alhaddad H, Cisternino S, Declèves X, Tournier N, Schlatter J, Chiadmi F, et al. Respi- ratory toxicity of buprenorphine results from the blockage of P-glycoprotein- mediated efﬂux of norbuprenorphine at the blood-brain barrier in mice. Crit Care Med. 2012;40:3215–23.
[12] Pottier G, Marie S, Goutal S, Auvity S, Peyronneau MA, Stute S, et al. Imaging the im- pact of the P-glycoprotein (ABCB1) function on the brain kinetics of metoclopramide. J Nucl Med. 2016;57:309–14.
[13] Tournier N, Chevillard L, Megarbane B, Pirnay S, Scherrmann JM, Declèves X. Interac- tion of drugs of abuse and maintenance treatments with human P-glycoprotein (ABCB1) and breast cancer resistance protein (ABCG2). Int J Neuropsychopharmacol. 2010;13:905–15.
[14] Luthra SK, Brady F, Turton DR, Brown DJ, Dowsett K, Waters SL, et al. Appl Radiat Isot. 1994;45:857–73.
[15] Luthra SK, Pike VW, Brady F, Horlock PL, Prenant C, Crouzel C. Automated radiosyntheses of [6-O-methyl-11C]diprenorphine and [6-O-methyl-11C] buprenorphine from 3-O-trityl protected precursors. Int J Rad Appl Instrum A. 1987;38:65–6.
[16] Galynker I, Schlyer DJ, Dewey SL, Fowler JS, Logan J, Galley SJ, et al. Opioid receptor imaging and displacement studies with [6-O-[l1C]methyl]buprenorphine in baboon brain. Nucl Med Biol. 1996;23:325–31.
[17] Larsen P, Ulin J, Dahlstrom K. A new method for production of 11C-labelled methyl iodide from 11C-methane. J Label Compd Radiopharm. 1995;37:73–5.
[18] Jewett DM. A simple synthesis of [11C]methyl triﬂate. Int J Rad Appl Instrum A. 1992; 43:1383–5.
[19] González-Pérez V, Connolly EA, Bridges AS, Wienkers LC, Paine MF. Impact of organic solvents on cytochrome P450 probe reactions: filling the gap with (S)-warfarin and midazolam hydroxylation. Drug Metab Dispos. 2012;40:2136–42.