Background
An inhibitory G-protein coupled receptor of the neurotransmitter noradrenaline (NA), the α
2C-adrenoceptor (α
2C-AR) subtype, has attracted considerable interest as a therapeutic target to treat CNS disorders [
1]. The α
2C-AR may be involved in mediation of the fine-tuning effects of NA on central neurotransmission, particularly during stressful conditions. Results obtained with gene-targeted (knock-out) mice indicate that manipulation of α
2A-AR and α
2C-AR activation yields differential behavioural effects in nonclinical tests that are commonly used for assessing antidepressant, antipsychotic or pro-cognitive properties of drugs [
1]. This has led to the proposition that selective α
2C-AR antagonism might be a promising approach for the treatment of neuropsychiatric symptoms, potentially across a wide range of CNS disorders, with an improved therapeutic profile compared to non-selective α
2-AR antagonists [
1].
The availability of the novel α
2C-AR antagonist ORM-12741 and the α
2C-AR positron emission tomography (PET) tracer [
11C]ORM-13070 now provides novel opportunities to investigate the roles and possible therapeutic utility of α
2C-AR modulation in CNS disorders [
1]. Of all known α
2C-AR antagonists, ORM-12741 is the most advanced molecule in terms of data on human exposure; it is rapidly absorbed after oral dosing and has shown acceptable tolerability [
2]. Direct evidence supporting drug target engagement is a key element for establishing confidence in proof of concept evaluation in nonclinical and human studies [
3]. [
11C]ORM-13070 as a PET tracer has provided a valuable probe for specifically investigating α
2C-AR subtype functions as well as brain receptor occupancy in experimental animals and humans. Its application as a PET tracer has been validated and established in several studies, with acceptable test–retest reproducibility [
4‐
7]. Furthermore, additional work has shown that [
11C]ORM-13070 binding is sensitive to changes in extracellular NA concentrations in the human brain, provoked by physiological or pharmacological interventions, indicating that it may be a valuable tracer for the investigation of alterations in noradrenergic tone [
8,
9].
Methods
The current translational investigation was aimed at demonstrating target receptor engagement for ORM-12741 as well as establishing the utility of [11C]ORM-13070 as a suitable PET tracer for assessment of α2C-AR occupancy in rat and human brain.
α2-AR subtype binding and antagonist characteristics in vitro
Receptor binding assays were performed at Cerep (Celle l’Evescault, France) according to their standard procedures, using stably transfected cell lines. Inhibition constants (
Ki) were calculated using the Cheng–Prusoff equation [
10]. CHO cells transfected to express human α
2-AR subtypes were used to determine antagonist properties of ORM-12741 in a calcium ion based fluorescent assay as described previously [
11]. Adrenaline and noradrenaline were used as agonists, and changes in intracellular calcium were monitored with a FLEXstation bench top scanning fluorometer equipped with an integrated fluid transfer workstation (Molecular Devices, San Jose, CA, USA) and SOFTmax PRO version 3.2 software. ORM-12741 (Orion Pharma, Espoo, Finland; 10
–2 M) was dissolved in DMSO and subsequently diluted in Probenecid Ringer buffer.
Radiosynthesis of [11C]ORM-13070
[
11C]ORM-13070 was synthesized at Turku PET Centre Radiopharmaceutical Laboratory, Turku, Finland as described previously [
4] and was dissolved in a mixture of propylene glycol/ethanol/0.1 M phosphate buffer (7/3/45, v/v/v), pH 7.4 [
4].
Rat brain ex vivo autoradiography
An ex-vivo autoradiography method, as described in [
4], based on specific displacement of [
11C]ORM-13070 binding in the caudate-putamen nucleus, was used to determine α
2C-AR occupancy in rat brain. Male Sprague–Dawley rats (
n = 4–6/group) were treated with vehicle (PEG 300/5% glucose) or ORM-12741 (dissolved in PEG 300 and diluted with 5% glucose solution; 2, 10, 50 or 1000 µg/kg, s.c.) 10 min before injection of [
11C]ORM-13070 (38–81 MBq) into the tail vein. Body temperature was kept stabile using a heating mattress. At 10 or 30 min after [
11C]ORM-13070 administration, the rats were stunned by CO
2 asphyxiation and terminal blood samples were taken for determination of plasma levels of ORM-12741. Brains were frozen by immersion in isopentane chilled on CO
2 ice. Cryosections (40 µm) of the brain were prepared and regions of interest (caudate-putamen/cerebellum) were analysed by autoradiography as described previously by the Aida 2D densitometry program [
4]. Occupancy calculations were done similarly to the clinical study described below but instead of the baseline, average values in the vehicle group were used.
Animal care complied with the guidelines of the International Council of Laboratory Animal Science. The Animal Experiment Board of the Province of Southern Finland approved the methodologies used in this study.
α2C-AR occupancy in human brain in vivo
The clinical trial was an open label, single dose, uncontrolled study performed at a single centre. The primary objective of the study was to determine the extent of brain α2C-AR occupancy after different single oral doses of ORM-12741 and to describe the relationship of α2C-AR occupancy as a function of ORM-12741 dose and drug concentration in plasma. The study design involved dose ranging with adaptive selection of doses and assessment time points. The study protocol was approved by the Ethics Committee of the Hospital District of Southwest Finland and the Finnish Medicines Agency (EudraCT 2008-004929-42), and the trial was registered in the ClinicalTrials.gov database (NCT00829907).
Healthy male volunteers were enrolled after informed consent. Concomitant medications that could have affected the outcome of the study were prohibited within 2 weeks prior to the first PET scan or less than 5 times the elimination half-life of the medication. The use of nicotine containing products were forbidden during the stay at the study centre. Drug abuse and alcohol breath test were performed prior to PET scans and had to be negative. Each subject had three visits: a screening visit, a treatment visit and an end-of-study safety visit. A brain MRI scan was obtained for an individual anatomical reference map. All included subjects had a baseline PET scan ([
11C]ORM-13070 alone) and 1–3 scans at set time points after different doses of ORM-12741 (Table
1).
Table 1
Numbers of subjects in each dose group and PET scan time points after oral dosing with the α2C-adrenoceptor antagonist ORM-12741
0.3 | 2 | 1 | 1 | 1 | | 1 |
1 | 3a | 1 | 1 | 1 | | 1 |
10 | 5 | 3 | 3 | 2 | | 2 |
30 | 5 | 3 | 3 | 2 | | 2 |
60 | 4 | 4 | 4 | | 4 | |
Soft gelatin capsules containing ORM-12741 (0.1 mg, 1 mg and 10 mg) were produced by Orion Pharma (Espoo, Finland), and each study subject received a single oral dose (0.3, 1, 10, 30 or 60 mg) (Table
1).
Each [
11C]ORM-13070 dose (target radioactivity 550 MBq; < 10 μg of ORM-13070) was given as a rapid intravenous bolus injection (1–10 mL) at the start of the PET scan. PET imaging was performed as described previously [
7]. In brief a high-resolution research tomograph (HRRT; Siemens Medical Solutions, Knoxville, TN) with the subject’s head fixed in a head holder with an individually prepared thermoplasticmask was used. In addition, head movements were recorded with an infrared camera (Vicra®; Northern Digital Inc., Waterloo, ON, Canada). Slices of approximately 1.22 mm thickness covered the whole brain (axial field of view 25.3 cm). The camera was used in 3-D mode with scatter correction. The HRRT achieved transaxial and axial spatial resolution (full-width at half-maximum) of 2.5 mm. Before each PET scan, a transmission scan was done for attenuation correction with a
137Cs rotating point source. Regions-of-interest (ROIs) were manually drawn on the co-registered MRI scans using Imadeus software (version 1.1, Forima, Turku, Finland), checked to match the summated PET images and then transferred onto the dynamic PET image, from which regional time-activity curves were obtained for the following selected regions of the left and right brain hemispheres: caudate nucleus, cerebellar cortex and putamen as described previously [
7].
Tracer uptake in the ROIs was described with areas under the curves (AUC) in the scan time window of 5–30 min after tracer injection. As the cerebellum has been reported to be devoid of α
2C-ARs, it was used as a reference region for correction of non-specific uptake. A binding parameter (BiP) was calculated for each ROI as the ratio of specific binding (AUC
region − AUC
cerebellar cortex) and the AUC in the cerebellar cortex. Receptor occupancy by [
11C]ORM-13070 was negligible at the employed tracer doses (< 10 μg) [
7]. Receptor occupancy in the target regions was calculated according to the equation:
$$\% \;{\text{Receptor}}\;{\text{occupancy}} = \left( {1 - \frac{{{\text{BiPdrug}}}}{{{\text{BiPbaseline}}}}} \right) \times 100\%$$
where BiP
baseline = pre-drug baseline BiP value, BiP
drug = BiP value following ORM-12741.
Left- and right-side receptor occupancy estimates were averaged to a single value for each ROI.
Liquid chromatography-tandem mass spectrometry was used for the determination of concentrations of ORM-12741 in plasma extracted from venous blood samples that were collected before and 10 min, 40 min, 60 min 90 min, 2 h, 3.5 h, 4 h, 6 h, 6,5 h,12 h, 12.5 h and 24 h after ORM-12741 dosing. Plasma PK variables (Cmax: peak concentration, tmax: time to peak concentration, AUCt: area under the drug plasma concentration–time curve from time zero to the last observed concentration, AUC∞: area under the drug plasma concentration–time curve from time zero to infinity, t1/2: terminal half-life) for ORM-12741 were calculated by non-compartmental analysis using the WinNonlin® Professional software package version 5.0.1 (Pharsight Corporation, Mountain View, CA, USA). The actual time points for blood sampling were used in the PK calculations.
Nonlinear regression analysis was used to evaluate the relationships between ORM-12741 plasma levels and receptor occupancy (Sigmoid
Emax model):
$${\text{Occupancy}} = \frac{{E_{\max } *C^{h} }}{{{\text{EC}}_{50}^{h} + C^{h} }}$$
where Emax is a maximum receptor occupancy estimate, EC50 is a half maximal effective concentration estimate and h is a slope factor. Temporal occupancy patterns were estimated with a regression model. Statistical analyses were performed with SAS® for Windows (SAS Institute Inc., Cary, NC, USA) on observed cases only.
The safety of the subjects was evaluated by recording of adverse events (AEs), supine heart rate and blood pressure, 12-lead electrocardiogram, laboratory safety assessments and physical examination findings.
Discussion
The current translational investigation provides evidence supporting ORM-12741 as a selective, high-affinity antagonist of α2C-ARs with sufficient penetration of the blood–brain barrier to occupy α2C-ARs in the human brain, confirming its primary mode of action.
Receptor binding analysis demonstrated that ORM-12741 has high affinity for the cloned human α
2C-AR (
Ki: 0.08 nM) and lower affinity for the α
2A-AR (
Ki: 8.3 nM) and α
2B-AR (
Ki: 0.8 nM) subtypes, i.e. approximately 100- and 10-fold receptor subtype selectivity. ORM-12741 also antagonized intracellular calcium responses mediated by cloned human α
2C-AR activated with adrenaline (
Kb: 0.04 nM) or noradrenaline (
Kb: 0.01 nM) with potency estimates consistent with its binding affinity. Its relative α
2-AR subtype selectivity was somewhat higher in the functional assay compared to the receptor binding assay, with 4100- and 560-fold higher potency at the α
2C-AR compared to α
2A-AR (
Kb: 41 nM) and α
2B-AR (
Kb: 5.6 nM). In a general selectivity screen with 126 additional receptors and drug binding sites (GPCRs, ion channels, transporters, enzymes), binding of ORM-12741 to the α
1A-AR (
Ki estimate, 46 nM) was most notable, but this represented an approximately 575-fold affinity ratio when compared with α
2C-AR. ORM-12741 had much lower affinity (α
2C-AR selectivity at least > 2000 fold) against all other targets tested (unpublished data, Orion Pharma). Overall, these results confirm that ORM-12741 is a selective, high-potency antagonist of human α
2C-ARs. Since the α
2A-AR is the most prevalent and widely distributed α
2-AR subtype in humans, high selectivity over this target should reduce the potential for peripheral (e.g. cardiovascular) or central (e.g. anxiety) side-effects that are commonly observed with subtype non-selective α
2-AR antagonists [
12,
13].
[
11C]ORM-13070 has previously been validated as a selective PET ligand for assessing α
2C-AR expression and occupancy in rat [
4] and human brain [
7]. The current results extend and support these previous findings, confirming that [
11C]ORM-13070 shows similar regional distribution patterns in rat and human brain, with the most intense signal in the striatum. In an ex vivo autoradiography experiment pretreatment of rats with ORM-12741 inhibited [
11C]ORM-13070 binding in a dose- and exposure-related manner with significant effects at 10 µg/kg (s.c.). This dose was associated with a
Cmax in rat plasma of 3–6 nM, and a protein-unbound free drug concentration of 0.015–0.03 nM (free fraction 5% in rat plasma), which is in line with the affinity of ORM-12741 for α
2C-AR in vitro. Furthermore, similar exposure levels have been associated with the pharmacodynamic effects of ORM-12741 seen in the rat forced swim test (FST) and the phencyclidine-induced prepulse inhibition (PPI) model at doses of ≥ 16 µg/kg (s.c.) and ≥ 10 µg/kg (s.c.), respectively [
14]. Consistently, gene-targeted α
2C-AR knock-out mice have shown reduced immobility in the FST [
15,
16]. In addition, other α
2C-AR antagonists have shown similar effects in the FST and PPI models [
1,
11,
17]. Collectively, the accumulated in vitro and in vivo receptor-level evidence, together with the phenotypic pharmacodynamic signals observed in the FST and PPI models, formed the basis for this translational study in human subjects to validate the engagement of brain α
2C-ARs by ORM-12741.
The current PET study further confirmed the previously reported [
11C]ORM-13070 uptake and distribution pattern in the human brain, with the strongest binding signal being observed in the caudate nucleus and putamen [
6,
7]. This is also in line with the known distribution of α2C-adrenoceptors in post-mortem human brain samples, i.e. high in the caudate nucleus and putamen, low in cortex and neglible in cerebellum [
18]. Given the very small mass (average, 0.4 µg) of ORM-13070 delivered with the target radioactivity, the PET tracer was unlikely to compromise receptor availability for the occupancy analysis. These features of [
11C]ORM-13070 together with acceptable PK properties and good test–retest reliability make it a feasible tracer for PET-based receptor occupancy analysis. The results obtained with ORM-12741 in the present investigation provide further support for this notion, for the first time employing a subtype-selective α
2C-AR antagonist. Dosing with ORM-12741 decreased the specific binding of [
11C]ORM-13070 in the caudate nucleus and putamen in a time- and exposure-dependent manner, indicating occupancy of α
2C-ARs.
Significant α2C-AR occupancy was detectable in the human brain after ≥ 10 mg oral doses of ORM-12741. The peak receptor occupancy and the time course of occupancy were in agreement with drug concentrations in plasma, in terms of e.g. Cmax and tmax. The observed mean Cmax of ORM-12741 in plasma after 10 mg doses was 62.6 ng/mL, corresponding with a protein-unbound free drug concentration of 0.2 nM, which is close to its in vitro Ki estimate (0.08 nM) for the α2C-AR. Based on the measured plasma concentrations after 30 mg and 60 mg doses of ORM-12741, and taking into account an approximately 0.1% free fraction in human plasma, these doses yielded approximately 0.4 nM and 0.5 nM free concentrations of ORM-12741 in plasma, respectively. These estimates are broadly consistent with the results obtained in vitro with cloned human α2C-AR, indicating that 1 nM ORM-12741 produces 93% inhibition of (–)adrenaline binding. Furthermore, 1 nM ORM-12741 did not affect (–)adrenaline binding to the α2A-AR, suggesting that the doses used in the current PET study are likely to reflect selective antagonism of α2C-ARs. The low doses of 0.3 and 1 mg of ORM-12741 resulted in average Cmax of 1.2 ng/mL and 2.7 ng/mL, respectively, which provided free drug concentrations (4–9 pM) well below its α2C-AR Ki, explaining the lack of displacement of [11C]ORM-13070 after these doses. At the 10–60 mg dose levels, the basal ganglia occupancy estimates reached their maximum at about 1 h after dosing and then declined at the 6 and 12 h time points towards minimal residual occupancy. The maximum occupancy was increased in a dose-related fashion up to the 30 mg dose level (about 70% in the caudate nucleus), but increasing the dose further to 60 mg did not increase occupancy at 1 h. Still, the occupancy estimates at 3.5 h were somewhat higher after 60 mg than after 30 mg.
The relationship of α
2C-AR occupancy in the caudate nucleus and putamen with plasma ORM-12741 concentrations was best described by a sigmoidal
Emax model, in concordance with classical receptor binding to a single population of receptors. The analysis of the relationship of α
2C-AR occupancy with ORM-12741 concentrations in plasma was limited by the paucity of PET scanning data at higher plasma concentrations of ORM-12741 than 125 ng/mL. Thus, further increases in regional brain α
2C-AR occupancy with increasing concentrations of ORM-12741 in plasma can therefore not be excluded. Therefore, the
Emax, EC
90 and EC
50 estimates should be viewed as preliminary estimates. However, the maximum occupancy estimates achieved in the present study were in the same range as previous results where [
11C]ORM-13070 occupancy was measured in healthy human subjects after administration of the subtype-nonselective α
2-AR antagonist atipamezole [
7], and are also in line with the rat ex vivo autoradiography data. Issues to be considered in this context include the relative receptor binding specificity of the two competing α
2C-AR ligands, the tracer and the test drug ORM-12741, and the contribution of a putative radioactive tracer metabolite that may have interfered with the occupancy estimation. The results of an earlier validation study of the α
2C-AR PET tracer [
11C]ORM-13070 [
7], supported by nonclinical observations [
4], indicated that a radioactive tracer metabolite may enter the brain but appears to exhibit no specific binding to α
2C-ARs. Therefore, a negative bias may be present in the BiP estimates. Longer scan times would be expected to lead to even greater bias in BiP due to the accumulation of the metabolite in the brain. It is also noteworthy that it was not possible to include time as a factor into the model. Thus, it seems plausible that the occupancy estimates follow plasma ORM-12741 concentrations relatively closely.
At the time of the human PET study a validated simplified activity ratio method was used for receptor occupancy determination. This was based on extensive validation in a series of earlier studies with the same PET ligand [
6‐
9]. However, it is important to emphasize that this may be a significant limitation as the activity ratio method used does not allow to take into account any medication related changes in blood flow. Using the reference tissue approach for determination of binding potential may have been more effective in this respect. In addition, due to unfortunate circumstances MRI images were not available for additional analyses any more.
Species differences in the binding of ORM-12741 to plasma proteins are likely to explain the difference in total exposure levels required for effects in humans and rats; the free fraction is ~ 50-fold higher in rat plasma compared to human plasma. Overall, the PET study results provide direct evidence to support the primary mode of action of ORM-12741, involving α2C-AR occupancy in the human brain.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.