Abstract
5-Hydroxytryptamine1A (5-HT1A) receptors have been implicated in the pathophysiology and treatment of anxiety and depression, and are a target for novel drug development. This is the first study examining the human brain in vivo occupancy by a novel, selective, silent 5-HT1A antagonist. 2-[4-[4-(7-Chloro-2,3-dihydro-1,4-benzdioxyn-5-yl)-1-piperazinyl]butyl]-1,2-benzisothiazol-3-(2H)-one-1,1-dioxide (DU 125530), a compound in clinical development, has potential applications in the treatment of anxiety and mood disorders. Positron emission tomography (PET) and [11C][O-methyl-3H]-N-(2-(4-(2-methoxyphenyl)-1-piperazinyl)ethyl)-N-(2-pyridinyl)cyclohexanecarboxamide trihydrochloride (WAY-100635), were used to assess 5-HT1Aautoreceptor and postsynaptic receptor occupancy in 12 healthy male volunteers. Over a 10- to 40-mg daily dose range, DU 125530 was well tolerated, and exhibited a dose-dependent occupancy from 0 to 72% at 2 h post the last dose. Occupancy correlated significantly with plasma levels of DU 125530, and a fitting of the data to a standard single-site binding model gave a maximum occupancy of ∼80%, and a half-saturation concentration (ED50) of ∼7 ng/ml. At 24 h after the last dose 5-HT1A occupancy was ∼50% of that achieved at 2 h. This study demonstrates that high occupancy of the human brain 5-HT1A receptor can be achieved at doses producing minimal acute side effects.
The 5-HT1A receptor is widely distributed in the human brain, with receptor concentration in postmortem human samples varying from ∼300 fmol/mg tissue in the hippocampus to ∼230 fmol/mg tissue in the midbrain raphe nuclei (RN) to negligible in the cerebellum (Burnett et al., 1997; Hall et al., 1997). It functions both as a somatodendritic autoreceptor on serotonergic neurons of the RN, and a postsynaptic receptor on neurons in cortical and limbic areas (Peroutka 1985; Pazos et al., 1988). Postsynaptic 5-HT1A receptors mediate some of the effects of 5-HT released from nerve terminals (Sinton and Fallon, 1988; Blier et al., 1990; Chaput et al., 1991; Invernizzi et al., 1991). The 5-HT1A on the RN is an inhibitory autoreceptor, the activation of which leads to decreased 5-HT neuron firing, and hence decreased 5-HT release in terminal synapses (VanderMaelen et al., 1986; Sharp et al., 1989; Adell and Artigas 1991; Invernizzi et al., 1992). The 5-HT1A autoreceptor is therefore an attractive site for therapeutic drug action, because it will influence neurotransmission through all 5-HT receptor subtypes.
The use of the mixed β-adrenergic/5-HT1A weak partial agonist pindolol has been proposed as a novel strategy for augmenting the antidepressant effects of selective serotonin reuptake inhibitors on the basis of its affinity for the 5-HT1A autoreceptor (Artigas et al., 1994). We demonstrated previously that doses of pindolol significantly higher than those used in clinical studies in depression will be needed to achieve a significant occupancy of the 5-HT1Aautoreceptor (Rabiner et al., 2000a). A similar conclusion was reached in a subsequent independent study (Martinez et al., 2000). An increased dose of pindolol will raise the risk of β-adrenergic side effects, which may limit the usefulness of this strategy in patient populations. Novel compounds with higher selectivity for the 5-HT1A receptor may therefore provide significant advantages in clinical use.
2-[4-[4-(7-Chloro-2,3-dihydro-1,4-benzdioxyn-5-yl)-1-piperazinyl]butyl]-1,2-benzisothiazol-3-(2H)-one-1,1-dioxide (DU 125530) was demonstrated to act as a full antagonist on cloned human 5-HT1A receptors (pKi = 9.1 nM), as well as in vivo in tests of drug discrimination (Mos et al., 1997), startle potentiation (Joordens et al., 1998), ultrasonic vocalizations (Olivier et al., 1998a), and stress-induced hyperthermia (Olivier et al., 1998b). DU 125530 has a 300-fold higher affinity for the 5-HT1A receptor compared with the other 5-HT receptors, as well as a 7-fold higher affinity compared with the α1-adrenergic receptor, and the D2/D3 receptor (Investigator's Brochure; Solvay Pharmaceuticals, Marietta, GA). In this study we examined the occupancy of the human 5-HT1A receptor in vivo in 12 healthy volunteers in the dose range of 10 to 40 mg p.o. daily.
Materials and Methods
Subjects
Twelve healthy male volunteers (ages 25–55) were examined. The Imperial College School of Medicine Ethics Committee and the Administration of Radioactive Substances Advisory Committee approved the studies. All subjects underwent a full medical history and examination. After giving written informed consent all subjects underwent two open-label [11C][O-methyl-3H]-N-(2-(4-(2-methoxyphenyl)-1-piperazinyl)ethyl)-N-(2-pyridinyl)cyclohexanecarboxamide trihydrochloride (WAY-100635) PET scans with a mean interval of 7 days between the scans (range 3–15 days). The first was a baseline scan, whereas the second followed treatment with DU 125530. The subjects were divided into three groups, with four subjects studied at each dose (10, 20, and 40 mg daily, Table 1). The subjects were scanned 2 h after the last dose of DU 125530. The timing of the postdrug scans was chosen to coincide with the estimatedTmax of the drug in plasma, determined in previous pharmacokinetic studies (Investigator's Brochure; Solvay Pharmaceuticals). Subjects were asked to report any adverse events, which were recorded for subsequent assessment.
In addition, four subjects (two from the 40-mg group, and one each from the 20- and the 10-mg groups) received a third PET scan, 24 h after the last dose of DU 125530, to assess the time course of occupancy.
PET Data Acquisition
PET scans were performed on an ECAT 953B PET camera (CTI/Siemens, Knoxville, TN) (Spinks et al., 1992) in three-dimensional mode with dual window scatter correction (Grootoonk et al., 1996) and a measured attenuation correction. Data were acquired over 90 min in 21 time frames. [11C]WAY-100635 was prepared at the Cyclotron Unit by 11C carboxylation of a Grignard reagent (McCarron et al., 1996), and injected intravenously as a bolus (injected dose, 326 ± 57 MBq; specific activity, 111 ± 56 MBq · ng−1; “cold” WAY-100635 injected, 1.9 ± 2.5 μg; cold WAY-100634 injected, 0.7 ± 0.6 μg). To assess the effect of DU 125530 on the metabolism of the radiotracer, venous blood samples were taken for estimation of the percentage of the parent radiotracer 1000, 2000, and 3000 s from the injection of the radiotracer, using solid phase extraction and reverse phase high-performance liquid chromatography (Osman et al., 1998). Blood samples collected during the PET scan were used to estimate plasma levels of DU 125530 at 2, 2.5, 3, and 24 h postdose. In the 40-mg group only, additional samples were collected at 4 and 8 h postdose. DU 125530 plasma levels estimation was performed by ABL (Assen, The Netherlands), via a validated method based on liquid chromatography mass spectrometry/mass spectrometry, suitable over the range of 0.17 to 85 ng/ml.
PET Data Analysis
Calculation of Binding Potential.
The [11C]WAY-100635 PET scans were analyzed using a simplified reference tissue compartmental model, with the cerebellum as a reference tissue (Lammertsma and Hume, 1996; Gunn et al., 1998). The cerebellum was chosen as a reference tissue because it has an extremely low number of 5-HT1A receptors (Hall et al., 1997). The reference tissue model allows the estimation of binding potential (BP = f2BAVAIL/KD, where f2 is the “free fraction” of the radiotracer in the nondisplaceable compartment,BAVAIL is the concentration of available binding sites, and KD is the equilibrium dissociation rate constant of the radioligand (Cunningham and Lammertsma, 1994). The reference tissue model was applied at the voxel level, using a basis function implementation (Gunn et al., 1997,1998), and parametric maps of BP were generated (BP images). Occupancy of the 5-HT1A receptor sites was inferred as a reduction of BP, and hence BAVAIL, under the assumption that f2 andKD remain constant for the two scans.
Anatomical Region Definition.
A [11C]WAY-100635 template was created from integrated images (0–90 min) in standard space (Montreal Neurological Institute space) was created as described previously (Meyer et al., 1999). Anatomical regions, including the cerebellar region, were defined on single-subject magnetic resonance imaging image in Montreal Neurological Institute space, to create a generic region of interest (ROI) map. The [11C]WAY-100635 template was warped onto each individual [11C]WAY-100635 integral (0–90 min) image, using SPM99 (Wellcome Department of Cognitive Neurology, London, UK) normalization. The transformation parameters obtained were subsequently applied to the generic ROI map, producing an individualized ROI map for each subject. The cerebellar ROI was applied to the dynamic PET image to create an input function for the calculation of parametric images. The cortical and limbic ROI map was applied to BP images to obtain regional BP values. This technique allowed the examination of 21 brain ROIs in temporal, parietal, prefrontal, and cingulate cortex, which contain postsynaptic 5-HT1A receptors.
We calculated occupancy for the midbrain RN, a small beaded structure consisting of serotonergic cell bodies with a high concentration of presynaptic 5-HT1A autoreceptors. The RN is undetectable on magnetic resonance imaging but is well defined on a [11C]WAY-100635 PET image. Therefore, an ROI for the RN was defined manually on an integral PET emission dynamic image (summated 20–90 min after injection of the radiotracer) for each individual PET scan, and then applied onto the BP images to generate the RN BP values.
A set of previously acquired test-retest data (analyzed as described above) of 15 healthy volunteers scanned on two occasions with a mean interval of 300 days (range 5–624 days), was used for comparison with the occupancy data.
Statistical Analysis
Autoreceptor and the global postsynaptic receptor occupancy at 2 h postdrug were assessed by a one-way ANOVA with the drug dose as the between-subject factor.
To determine the dose dependence of the occupancy effect, the individual subject occupancies at 2 h were fitted to the measured plasma DU 125530 concentration in the first 30 min of the PET scan (the mean of the 2- and the 2.5-h postdrug plasma levels) to derive OCCmax and EC50 as per eq. 2. The plasma concentration of DU 125530 in the first 30 min of the PET scan was used to provide an approximation of receptor levels of DU 125530 at the time of [11C]WAY-100635 injection.
The analysis of the effects of DU 125530 on the concentration of WAY-100635 in plasma over the course of the PET scan was performed by a two-way ANOVA with sample time and scan condition (baseline or postdrug) being the two within-subject factors, whereas drug dose was the between-subject factor examined.
Results
An examination of the occupancy values for cortical and limbic structures revealed no differences between the different regions; therefore, further analysis was conducted on the mean value for postsynaptic 5-HT1A receptors weighted for the region size (postsynaptic). The values for autoreceptor occupancy were calculated from the BP values for the RN.
There was a dose-dependent decrease in BP after the administration of DU 125530 in all regions of interest (ROIs) examined, producing an occupancy effect in the range from 0 to 72% (Table2). We found a significant effect of the dose of DU 125530 on the 5-HT1A receptor occupancy at both the autoreceptor and the postsynaptic receptor (one-way ANOVA: F3,22 = 15.38,p < 0.001 and F3,22 = 35.81, p < 0.001 for autoreceptor and postsynaptic receptor, respectively). Post hoc testing (Dunnet's t test, two-sided) revealed the occupancy effect to be significant at all doses at the postsynaptic receptor, and at 20 and 40 mg at the autoreceptor. Examination of the time course of occupancy revealed a 5-HT1A autoreceptor and postsynaptic receptor occupancy 24 h after the last dose of DU 125530 of approximately one-half of the occupancy at 2 h postdrug (Table 2).
Because the kinetic model used to estimate the BP (simplified reference tissue compartmental model) depends on the cerebellum as a reference region, it is important to assess changes in cerebellar binding of [11C]WAY-100635 after DU 125530 administration, to exclude changes in BP due to changes in the reference region. As can be seen from Fig. 1, DU 125530 did not affect the time-activity curve (TAC) of [11C]WAY-100635 in the reference region. In addition, there was no significant effect of DU 125530 on the plasma levels of [11C]WAY-100635 (Fig.2; repeated measures ANOVA: effect of drug dose F3,14 = 0.423,p = 0.739 and dose by time interactionF3,14 = 2.767, p = 0.081), although an examination of Fig. 2 indicates that at the highest dose there may have been an interaction.
DU 125530 was well tolerated with six of the 12 subjects reporting no acute side effects. The most frequent side effect reported by the remaining subjects, was a transient feeling of mild “tingling” throughout the body, reported by five subjects, mostly at the higher dose. Other side effects reported were headache (n = 2) and “lightheadedness” (n = 2). All side effects reported were mild, except for moderate tingling reported by two subjects. Because of the absence of a placebo group all side effects reported should be treated as anecdotal, although they were consistent with side effects experienced by subjects in previous studies with this compound (Investigator's Brochure; Solvay Pharmaceuticals).
DU 122530 plasma levels during the period around the PET examination are detailed in Table 3. The calculation of pharmacokinetic parameters (OCCmax and EC50) of DU 125530 binding to the 5-HT1A receptor produced estimates of: OCCmax = 81.2 ± 4.3% autoreceptor and 77.1 ± 10.9% postsynaptic receptor, and EC50 = 7.8 ± 2.1 ng/ml autoreceptor and 7.0 ± 4.0 ng/ml postsynaptic receptor (Fig.3).
There was no difference between the control and the drug scans in injected dose or the specific activity of the radiotracer (specific activity 114 ± 71 and 95 ± 55 MBq · ng−1, injected dose 329 ± 62 and 335 ± 47 MBq, for the baseline and postdrug scans, respectively).
Discussion
To our knowledge this is the first human study investigating the in vivo occupancy of a specific silent 5-HT1Aantagonist. The human 5-HT1A receptor occupancy by the β-adrenergic/5-HT1A partial agonist pindolol was examined previously (Andree et al., 1999; Rabiner et al., 2000a; Martinez et al., 2001), as were the occupancies by two other β-blockers, penbutolol and tertatolol (Rabiner et al., 2000a), and the azapirone buspirone (Rabiner et al., 2000b). The occupancy achieved with these compounds was relatively low at the doses used (pindolol, 20 mg, ∼40%; penbutolol, 80 mg, ∼40%; and buspirone, 20 mg, ∼20%). Drug doses needed for higher occupancy would be predicted to produce unacceptable side effects, arising from β-adrenergic or 5-HT1A activity of these compounds. In contrast, DU 125530 allowed a demonstration of high levels of 5-HT1A receptor occupancy with minimal side effects. Significant occupancy of 5-HT1A receptor by DU 125530, was present also at 24 h postdose, thus indicating that DU 125530 may be administered once daily, although its plasmat1/2 is only 10 to 12 h.
The minimal side effects, even at doses producing ∼70% occupancy of the 5-HT1A receptor, may have several explanations. The baseline “tone” at the 5-HT1A receptor may be low, and therefore the biological effects of an antagonist will only manifest themselves in situations where the tone is increased (e.g., increased synaptic 5-HT). Alternatively, the baseline tone at the receptor may be high, but a very high proportion of the receptors needs to be blocked to produce behavioral effects. Work on the displacement of [11C]WAY-100635 binding by endogenous 5-HT in the rat after fenfluramine infusion (Hume et al., 2001; Maeda et al., 2001) supports the first possibility, indicating that under baseline conditions the occupancy of the 5-HT1A receptor by 5-HT is very low.
The OCCmax derived from fitting the occupancy data to a single-site occupancy model (OCCmax = 81.2% autoreceptor and 77.1% postsynaptic receptor) suggests that DU 125530 binds to the full population of 5-HT1Areceptors. This is consistent with its preclinical characterization as a full antagonist at the 5-HT1A receptor.
The use of the simplified reference tissue model to assess drug occupancy assumes that no change occurs in the binding of the radiotracer to the nondisplaceable compartment in the reference tissue (Tref) alone (a global change in radiotracer binding to the nondisplaceable compartment, producing a change in global f2, will not violate the assumptions of the model). We cannot measuref2 in vivo, so to assess changes in the nondisplaceable compartment of the Tref, we examined the TACs of the reference tissue, scaled to the peak of each TAC. A change in the shape of the TAC would indicate a change in the binding of the radiotracer to the nondisplaceable compartment in the Tref (because the Trefshould be devoid of specific binding). We found no differences in the cerebellar TACs of subjects before and after DU 125530, and therefore the dose-dependent reduction of [11C]WAY-100635 BP is unlikely to be explained by a change in the nondisplaceable binding in the reference region. Although there was a trend toward an increase in [11C]WAY-100635 metabolism after DU 125530, an examination of Fig. 2 reveals a possible effect at the 40-mg dose, but not at the lower doses. This is unlikely to influence our results, because there was a significant occupancy at the lower doses of DU 125530, without any discernible effect on the metabolism of [11C]WAY-100635. Therefore, the change in BP can be best explained by a dose-dependent occupancy of the 5-HT1A autoreceptor and postsynaptic receptor by DU 125530.
Unlike our previous findings with pindolol (Rabiner et al., 2000a) we found no evidence of preferential autoreceptor compared with postsynaptic receptor occupancy with DU 125530. The reason for this is not clear, however, the preferential occupancy exhibited by pindolol may be related to the varying proportions of G protein-coupled and -uncoupled sites at the autoreceptor and postsynaptic receptor sites. Pindolol, being a partial agonist at the 5-HT1Areceptor, will be sensitive to the coupling state of the receptor, whereas DU 125530, being an antagonist, will not. An interesting point to note is that these results make it very unlikely that the findings of preferential binding with pindolol are due to a partial volume effect, in the raphe nuclei.
The concurrent occupancy of both the autoreceptor and the postsynaptic receptor by DU 125530 will make the expected pharmacological effect of DU 125530 depend on the endogenous tone of the serotonergic system. At conditions of high tone (high levels of 5-HT) the blocking of the 5-HT1A autoreceptor will counteract the decrease in the firing rate of the 5-HT neuron caused by 5-HT, and although the transmission through the postsynaptic 5-HT1A may be disrupted by the blockade of the postsynaptic 5-HT1A receptor, transmission through the other 5-HT receptor subtypes will be enhanced. On the other hand, in conditions of low tone, a blockade of the 5-HT1Areceptors by a full antagonist may not have a noticeable effect.
In conclusion, DU 125530 displays a dose-dependent occupancy of the 5-HT1A receptor in the human brain, reaching up to 72% in the dose range from 10 to 40 mg daily, with minimal side effects, which makes it eminently suitable for application in humans.
Acknowledgments
We thank Joanne Holmes, Andy Blyth, Dave Turton, Ray Khan, Keith Poole and team, and Safiye Osman and team for expert technical assistance.
Footnotes
-
This work was supported by Solvay Pharmaceuticals, Weesp, The Netherlands.
- Abbreviations:
- 5-HT
- 5-hydroxytryptamine
- RN
- raphe nuclei
- PET
- positron emission tomography
- BP
- binding potential
- ROI
- region of interest
- ANOVA
- analysis of variance
- TAC
- time activity curve
- OCCmax
- maximal occupancy
- DU 125530
- 2-[4-[4-(7-chloro-2,3-dihydro-1,4-benzdioxyn-5-yl)-1-piperazinyl]butyl]-1,2-benzisothiazol-3-(2H)-one-1,1-dioxide
- WAY-100635
- [O-methyl-3H]-N-(2-(4-(2-methoxyphenyl)-1-piperazinyl)ethyl)-N-(2-pyridinyl)cyclohexanecarboxamide trihydrochloride
- Received October 31, 2001.
- Accepted January 25, 2002.
- The American Society for Pharmacology and Experimental Therapeutics