Radiosynthesis and Evaluation of - American Chemical Society

PET Centre, Centre for Addiction and Mental Health and Departments of Psychiatry and Pharmacology, University of Toronto,. Toronto, Ontario M5T 1R8, ...
0 downloads 0 Views 124KB Size
Download PDF
J. Med. Chem. 2005, 48, 4153-4160

4153

Radiosynthesis and Evaluation of [11C]-(+)-4-Propyl-3,4,4a,5,6,10b-hexahydro-2H-naphtho[1,2-b][1,4]oxazin-9-ol as a Potential Radiotracer for in Vivo Imaging of the Dopamine D2 High-Affinity State with Positron Emission Tomography Alan A. Wilson,* Patrick McCormick, Shitij Kapur, Matthaeus Willeit, Armando Garcia, Doug Hussey, Sylvain Houle, Philip Seeman, and Nathalie Ginovart PET Centre, Centre for Addiction and Mental Health and Departments of Psychiatry and Pharmacology, University of Toronto, Toronto, Ontario M5T 1R8, Canada Received February 17, 2005

In vivo imaging of dopamine D2 receptors with agonist (as opposed to the more commonly employed antagonist) radiotracers could provide important information on the high-affinity (functional) state of the D2 receptor in illnesses such as schizophrenia, movement disorders, and addictions. We report here the radiosynthesis and evaluation of the potent D2 agonist (+)-4-propyl-3,4,4a,5,6,10b-hexahydro-2H-naphtho[1,2-b][1,4]oxazin-9-ol, (+)-3, labeled with carbon-11, as a potential radiotracer for imaging the high-affinity state of dopamine D2 receptors with positron emission tomography (PET). [11C]-(+)-3 was reliably synthesized in the quantities and at the specific activities and radiochemical purities required for human PET studies. Ex vivo biodistribution studies in rat brain demonstrated that [11C]-(+)-3 crossed the blood-brain barrier readily and had an appropriate regional brain distribution for a radiotracer that maps dopamine D2 receptors. The binding of [11C]-(+)-3 was saturable and demonstrated an excellent signal-to-noise ratio as measured by its striatum-to-cerebellum ratio of 5.6, 60 min postinjection. The binding was highly stereospecific, and blocking and displacement studies were consistent with selective and specific binding to the dopamine D2 receptors. Further, [11C]-(+)-3 showed marked and appropriate sensitivity to both increases and decreases in the levels of endogenous dopamine. Brain radioactive metabolite and physicochemical measurements are in full accord with the desired properties of a neuroreceptor imaging agent for PET. All of the above, coupled with the documented full D2 agonistic properties of (+)-3, strongly indicate that [11C]-(+)-3 is a leading candidate radiotracer for the imaging of the dopamine D2 high-affinity state using PET in human subjects. Introduction An increasingly important and intriguing application of positron emission tomography (PET) is its use as a modality to measure endogenous levels of neurotransmitters in vivo in both man and animal.1 This technique hinges on the competition between the endogenous neurotransmitter and the administered radiotracer for the same binding site (neuroreceptor), which is reflected in changes in the measured binding of the radiotracer. While several neuroreceptor systems have been studied in this context, a substantial portion of the research has focused on the D2 subtype of the dopamine (DA) receptor.2-4 Important clinical applications of the measurement of radiotracer binding sensitivity to endogenous neurotransmitter have already appeared in both the PET and single photon emission computed tomography (SPECT) literature such as the finding that drug-naive schizophrenics are more sensitive to DA manipulation.5,6 A recent comprehensive review7 has summarized both progress and limitations in this area of in vivo imaging. Many G-protein-coupled receptors including the D2 receptor have been postulated to exist in interconvert* To whom correspondence should be addressed. Phone: 416 979 4286. Fax: 416 979 4656. E-mail [email protected].

able high- and low-affinity states with respect to agonist binding,8 and this concept has been incorporated into models of receptor binding.9-12 Such models state that antagonists (and thus antagonist radiotracers) do not distinguish between the high- and low-affinity states of the receptor, while agonists preferentially bind to the high-affinity state, the state coupled to G proteins. By competing directly for the same high-affinity sites as DA, a full agonist radiotracer should exhibit increased sensitivity to DA levels and should also show full displacement at very high levels of endogenous DA. In addition, since the high-affinity state of the DA receptor is believed to be the functional state,13 in vivo imaging with an agonist radiotracer should provide more relevant information about the neuroreceptor than an antagonist radiotracer because the latter fails to discriminate between the functional and nonfunctional states of the receptor. For these reasons many groups have pursued a suitable D2 selective ligand that fulfills all the criteria required for a successful PET radiotracer and that is also a demonstrable full agonist. Two major classes of D2 agonists, apomorphine analogues14-18 and hydroxyaminotetralin analogues,19-21 have been explored as potential imaging agents (Chart 1).

10.1021/jm050155n CCC: $30.25 © 2005 American Chemical Society Published on Web 05/19/2005

4154

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 12

Chart 1. Examples of 11C- and 18F-Labeled Putative D2 Agonist Radiotracers for PET That Have Been Tested ex Vivo or in Vivo

Wilson et al.

sessed it as a new DA agonist radiotracer, [11C]-(+)-3, the in vitro and ex vivo results from which are reported here. Our findings strongly suggest that [11C]-(+)-3 is a highly suitable candidate radiotracer to image the high-affinity state of the D2 receptor using PET. Experimental Section

Even though many of the candidates had high affinity for the D2 receptor and some gave promising in vitro autoradiographic images,19,20 at best only modest levels of specific binding were obtained in ex vivo dissection or in vivo imaging studies (with one exceptionsvide infra). With hindsight some of the failures are readily explicable; e.g., N-fluoropropylnorapomorphine (2b) probably failed because of the detrimental influence of the fluorine substitution on the basicity of the proximal nitrogen22 and some of the aminotetralin analogues such as 1c and 1d are likely to be too lipophilic resulting in high levels of nonspecific binding when employed in vivo.23 However, it is often unclear why other candidates, e.g., 2-methoxy-N-propylnorapomorphine (2d), were disappointing.16 Recently a breakthrough in this field was reported by Hwang and co-workers. They showed that N-propylnorapomorphine (NPA, 2e), labeled with 11C, fulfilled most of the requirements of a potential radiotracer for imaging the high-affinity state of the D2 receptor including good brain penetration and a moderate signal-to-noise ratio ex vivo in rats and in vivo using PET in nonhuman primates.17,18,24 The fate of this tracer in humans is not yet known however. (+)-4-Propyl-3,4,4a,5,6,10b-hexahydro-2H-naphtho[1,2-b][1,4]oxazin-9-ol [(+)-PHNO, (+)-3, Figure 1] is a

Figure 1. Structure of (+)-3.

naphthoxazine first synthesized in 1984 and reported as the most potent known D2 agonist at that time.25 Compound (+)-3 has subnanomolar affinity for the D2 receptor with a eudismic ratio of more than 10 000 26 and is a full agonist in many in vitro and in vivo assays.27,28 In canine striatal homogenates, guanilylimidodiphosphate (Gpp[NH]p) markedly inhibited [3H]-(+)3 binding, suggesting that [3H]-(+)-3 was binding primarily due to the high-affinity state of DA D2 receptors.26 In vitro quantitative autoradiography on rat brain sections, using [3H]-(+)-3, demonstrated high levels of binding in striatal regions that could be completely removed by incubation with either sulpiride (a selective D2 antagonist) or Gpp[NH]p (conversion of D2 high-affinity state to low-affinity state).29 We have radiolabeled compound (+)-3 with carbon-11 and as-

A Scanditronix MC 17 cyclotron was used for radionuclide production. Purifications and analyses of radioactive mixtures were performed by high-performance liquid chromatography (HPLC) with an in-line UV (280 nm) detector in series with a NaI crystal radioactivity detector (radiosynthesis and QC) or a Bioscan Flowcount coincidence radioactivity detector (metabolite analysis and QC). Isolated radiochemical yields were determined with a dose calibrator (Capintec CRC-712M). 3,4,4a,5,6,10b-Hexahydro-2H-naphtho[1,2-b][1,4]oxazin-9-ol hydrochloride (both enantiomers) were purchased from Toronto Research Chemicals (Toronto, Canada). All other chemicals were obtained from Aldrich or Fisher. Tetrahydrofuran (THF) was freshly distilled under nitrogen from lithium aluminum hydride (LiAlH4), while dimethylformamide (DMF) was distilled from barium oxide and stored over 4 Å molecular sieves prior to use. Diisopropylethylamine (DIPEA) was stored over 4 Å molecular sieves prior to use. Phthaloyl dichloride was distilled under reduced pressure and stored in a desiccator at 3 °C. Radiosyntheses were performed using a robotic arm system (ANATEC, Sweden) in conjunction with an automated in-house module controlled by Labview software. Elemental analyses were carried out by Atlantic Microlab (GA). All animal experiments were carried out under humane conditions, with approval from the Animal Care Committee at the CAMH and in accordance with the guidelines set forth by the Canadian Council on Animal Care. (+)-4-Propyl-3,4,4a,5,6,10b-hexahydro-2H-naphtho[1,2b][1,4]oxazin-9-ol Hydrochloride (+)-3. A stirred suspension of (+)-3,4,4a,5,6,10b-hexahydro-2H-naphtho[1,2-b][1,4]oxazin-9-ol hydrochloride (100 mg, 0.41 mmol) and DIPEA (250 µL) in acetonitrile (5 mL) was treated with n-propyl iodide (200 µL, 2 mmol). The mixture was heated to reflux for 4 h and then left at ambient temperature for 3 days. Upon evaporation of the acetonitrile, the residue was taken up in ethyl acetate (20 mL), washed with saturated aqueous sodium bicarbonate, dried (Na2SO4), and filtered. The solid obtained from evaporation of solvent was dried under high vacuum overnight. Treatment with ethereal HCl (1 N) afforded the hydrochloride salt as a white solid, which was recrystallized from ethanol (63.9 mg, 55% yield). A second crop (27.1 mg, 23%) was obtained upon treating the mother liquor with ether: mp 280282 °C (lit.30 255 °C (dec)). Anal. (C15H22ClNO2) C, H, N. [11C]Propionyl Chloride. The method of Luthra31 with some modifications was used. [11C]-CO2, produced by the 14N(p,R)11C nuclear reaction, was concentrated from the gas target in a stainless steel loop cooled to -178 °C. Upon warming, the [11C]-CO2 was passed through a NOx trapping column32 and a drying column of P2O5 into a solution of ethylmagnesium bromide (0.4 mL, 0.5 N in anhydrous diethyl ether/THF 50/ 50) by a flow of N2 (14 mL/min) at ambient temperature. When transfer of radioactivity was complete (4 min), the N2 flow was stopped and phthaloyl dichloride in THF (0.5 mL, 2 M) was added, followed by 0.5 mL of a solution of 2,6-di-tert-butylpyridine (2.1M) and DMF (0.65M) in THF. The mixture was incubated for 120 s and then heated to 130 °C, and the formed [11C]propionyl chloride swept into a receiving vial along with the THF via Teflon tubing by a N2 flow of 40 mL/min. A further 0.7 mL of THF was added to the reaction vial in two portions upon distillation of the initial THF. and N2 flow continued at 80 mL/min until radioactivity levels plateaued in the collection vial. [11C]-(+)-3. [11C]Propionyl chloride/THF (prepared as described above) was trapped in a 5 mL V-vial containing (+)3,4,4a,5,6,10b-hexahydro-2H-naphtho[1,2-b][1,4]oxazin-9-ol, hydrochloride (8 µmol, (4)), DIPEA (50 µL), and THF (50 µL) at less than -30 °C. When levels of radioactivity in the vial peaked, the vial was immersed in an oil bath at 85 °C until

Radiotracer for in Vivo Imaging

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 12 4155

the internal temperature reached 60 °C (about 2.5 min). One minute later the vial was then cooled in an ethanol/dry ice bath until the internal temperature was less than -30 °C, at which point LiAlH4 in THF was added (0.2 N, 0.6 mL). The vial was then reimmersed in the oil bath, and THF was removed by a flow of N2 (80 mL/min) through the vial. Upon evaporation of all the THF, aqueous HCl (0.6N, 0.8 mL) was added followed after 30 s by 1 mL of HPLC eluent. The reaction mixture was purified by reverse-phase HPLC (Phenomenex Luna C18(2) 250 mm × 10 mm, 6 mL/min, 20% CH3CN 80% H2O + 0.1 N ammonium formate at pH 4). The desired fraction (tR((+)-3) ) 6.5 min) was collected and evaporated to dryness under vacuum at 70 °C, and the residue was taken up in 10 mL of sterile saline. The saline solution of [11C]-(+)-3 was passed through a sterile 0.22 µm filter into a sterile, pyrogenfree bottle containing aqueous sodium bicarbonate (1 mL, 8.4%). Aliquots of the formulated solution was used to establish the chemical and radiochemical purity and specific activity of the final solution by analytical HPLC, Phenomenex Prodigy C18 10 µm (250 mm × 4.5 mm, 20% CH3CN/80% H2O + 0.1 N ammonium formate at pH 4, 3 mL/min). The identity of the product was confirmed by HPLC and radio-thin-layer chromatography (radio-TLC). Radio-TLC of the formulated product was carried out on silica plates using both ethyl acetate/ triethylamine (95/5, Rf ) 0.57) and ethyl acetate/triethylamine/ methanol (90/5/5, Rf ) 0.35). HPLC was performed with a variety of normal and reverse-phase columns using different solvents and pHs. [11C]-(-)-4-Propyl-3,4,4a,5,6,10b-hexahydro-2H-naphtho[1,2-b][1,4]oxazin-9-ol Hydrochloride ([11C]-3. This was prepared in an identical manner to the (+)-enantiomer using (-)-3,4,4a,5,6,10b-hexahydro-2H-naphtho[1,2-b][1,4]oxazin-9ol as precursor. Results were comparable between enantiomers. [11C]-2e. This was prepared in an manner similar to that for [11C]-(+)-3 using normethylapomorphine hydrochloride as precursor.18 Radiochemistry results were comparable. Measurements of log P. The partition coefficients of [11C](+)-3 and [11C]-2e, between 1-octanol and 0.02 M phosphate buffer at pH 7.4, were determined by a previously described method.33 Biodistribution in Rats. Rats (male, Sprague-Dawley, 230-280 g) were kept on a reversed 12 h light and 12 h dark cycle and allowed food and water ad libitum. A previously described method34 was used to determine the regional brain uptake of radioactivity. Briefly, rats, in a restraining box, received 25-30 MBq (at the time of first injection) of high specific activity radiotracer (1-2 nmol) in 0.3 mL of buffered saline via the tail vein, vasodilated in a warm water bath. Rats were killed by decapitation at various time intervals after radiotracer administration, and the brain was removed and stored on ice. Brain regions were excised, blotted, and weighed, while blood was collected (from the trunk). Radioactivity in tissues was assayed in an automated γ counter and backcorrected to time of injection, using diluted aliquots of the initial injected dose as standards. Pharmacological Challenge Studies in Rats. For most pharmacological studies, groups of rats (n ) 4 or 5) received an injection of a solution (1 mL/kg body weight in saline iv unless otherwise stated) of the pharmacological agent 25-30 min prior to the radiotracer injection. Control animals (a control group was run each time) received only saline. For the DA depletion studies, 20 rats were divided into four groups of five. One group received reserpine (2 mg/kg, sc) 24 h prior to [11C]-(+)-3 administration, followed by R-methyl-p-tyrosine (AMPT, 250 mg/kg, ip) 1 h prior to [11C]-(+)-3 administration). The second group received reserpine and then saline, while the third group received saline and then AMPT. The fourth group (controls) received only saline administered twice. In all studies, animals were killed 60 min after radiotracer injection and brain regions were dissected and counted as described above. Metabolism Studies. A rat (male, Sprague-Dawley, 260 g) received 50 MBq of high specific activity (+)-[11C]-3 (2.5

nmol) in 0.5 mL of buffered saline via the tail vein. The rat was killed at 40 min postinjection by decapitation, blood was collected from the trunk in a heparinized tube, and the whole brain was surgically removed from the skull and stored on ice. A control animal, which received no radiotracer, was also killed, blood was collected, and the brain was excised and stored on ice. An amount of 1 MBq of (+)-[11C]-3 in 10 µL of buffered saline was added to the control brain and control blood only. Both brains were homogenized (Polytron, setting 7) in 4 mL of ice-cold 70% aqueous ethanol and centrifuged (17 000 rpm, 15 min). Aliquots of the supernatants and the pellets were counted for radioactivity, and the supernatants were then diluted 4:1 (v/v) with water for HPLC analysis. Blood was centrifuged to separate the plasma, which was used directly for HPLC analysis. HPLC analysis of plasma and brain extracts were performed by minor modifications of the method described by Hilton.35 Briefly, samples were loaded onto a 5 mL HPLC injector loop (Valco Texas) and injected onto a small capture column (4.6 mm × 20 mm) packed inhouse with OASIS HLB 30 µm (Waters, NJ). The capture column was eluted with 1% aqueous acetonitrile (2 mL/min) for 4 min and then back-flushed (25% acetonitile/75% H2O + 0.1 N ammonium formate) onto a Phenomenex 10 µProdigy C18 column (250 mm × 4.6 mm). Both column effluents were monitored through a flow detector (Bioscan Flow-Count) operated in coincidence mode. After 14 min, the HPLC eluent was changed to 60% acetonitile/40% H2O + 0.1 N ammonium formate, pH 4, at 2 mL/min to monitor for strongly retained (lipophilic) metabolites. All radioactivity data were corrected for physical decay and integrated using a PC. Hold-up of radioactivity in the HPLC system was less than 2% of applied radioactivity, while the pellet had less than 10%. The experiment was repeated once. In Vitro Binding Assays. A commercial service (NOVASCREEN central nervous system (CNS) receptor screen, Biosciences Corporation, Hanover, MD) was employed to explore the in vitro affinity of (+)-3 for a broad selection of receptors, ion channels, and enzymes. Plasma Protein Binding. Binding of [11C]-(+)-3 to human plasma proteins was performed using an ultrafiltration technique described previously.36,37 Briefly, human blood samples (n ) 6, in triplicate) were centrifuged and the plasma was spiked with [11C]-(+)-3. The plasma was then filtered through an ultrafiltration unit (AmiconCentrifree; Millipore), and the ultrafiltrate and plasma were counted for radioactivity and weighed. Binding to the filtration apparatus (99%. Co-injection of the radioactive product with an authentic standard of (+)-3 under different conditions (solvents, pH) with different analytical HPLC columns (Waters, Alltech, Phenomenex) further established the identity of the radiotracer, as did radio-TLC. Under all conditions [11C](+)-3 cochromatographed with authentic unlabeled (+)3. The formulated radiotracer displayed no radiolysis for at least 60 min postsynthesis and required no stabilizing agents such as ascorbate. The substitution of the 2° amine precursor (4) in Scheme 1 by either its (-)-enantiomer or normethylapomorphine enabled the radiosynthesis of [11C]-(-)-3 and [11C]-2e, respectively, with comparable outcomes.

4156

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 12

Wilson et al.

Scheme 1. Radiosynthesis of [11C]-(+)-(3) Using [11C]Propionyl Chloride

In Vitro Binding Studies. A broad screen commercial assay (>60 targets) for a variety of CNS receptors showed that 1 µM solutions of (+)-3 had an affinity only for D2, D3, R-2 adrenergic, σ, and 5-HT1A receptors. Previous studies had already demonstrated that (+)-3 had an affinity for D2,26 D3,38 and R-2 adrenergic receptors.39 Ex Vivo Biodistribution Studies with [11C]-(+)3 in Rats. Upon iv radiotracer administration, high levels of radioactivity were observed in rat brain (Figure 2A), indicating good blood-brain barrier penetration of [11C]-(+)-3 with 2% injected dose per gram wet tissue (%ID/g) present in the striatum 5 min postinjection. Uptake of radioactivity was highest and washout was slowest from the striatum, an area of brain known to be rich in DA D2 receptors.40 All other brain regions had radioactivity levels comparable to the cerebellum, a brain region known to contain very few DA D2 receptors.40 Specific binding ratios, defined as (%ID/ gstriatum - %ID/gcerebellum)/%ID/gcerebellum, increased with time, from 1.1 at 5 min to 2.8 at 15 min to 3.9 at 30 min, reaching 4.6 at 60 min postinjection. The enantiomeric radiotracer, [11C]-(-)-3, displayed no preferential uptake in any rat brain region; instead, a homogeneous distribution was found throughout the brain with specific binding ratios of 0-0.1 at all time points in all regions (Figure 2B). The regional brain uptake of [11C]-2e was also determined to allow a direct comparison between the radiotracers. Results were very similar to those reported in the literature.18 Five minutes postinjection, 1.3% %ID/g was found in the striatum with specific binding ratios of 0.7, 1.7, 2.9, and 3.3 at 5, 15, 30, and 60 min postinjection, respectively. Effect of Pharmacological Challenges on Specific Binding of [11C]-(+)-3. The effects of pretreatment of groups of rats with a variety of CNS active drugs were explored to determine the specificity and saturability of binding of [11C]-(+)-3 in brain regions (Table 1). In addition, treatments (RTI-32 and amphetamine) designed to enhance the levels of endogenous DA were performed to determine the sensitivity of [11C](+)-3 specific binding to DA levels (Table 1). In contrast, Table 2 shows the results of paradigms that aimed to

Figure 2. Time-activity curves of regional brain uptake following injection of (A) [11C]-(+)-3 and (B) [11C]-(-)-3 into rat tail vein. Note the homogeneous distribution of radioactivity in all brain regions except in the striatum in (A).

deplete DA levels (reserpine, AMPT) on [11C]-(+)-3 specific binding. Metabolism and Physicochemical Studies. HPLC analysis of rat plasma showed that 74% radioactivity in the rat plasma was polar metabolites, with 26% unchanged [11C]-(+)-3, 40 min after radiotracer injection. In contrast, only 5% of the radioactivity in the rat brain extract was attributable to polar metabolites (Figure 3). No lipophilic metabolites were detected in either brain extract or plasma, and no (95% of the radioactivity distilled and trapped was in the form of [11C]propionyl chloride. Formation of the [11C]amide (5) was rapid, requiring only 1 min once the precooled reaction mixture had reached 60 °C. Apart from [11C](+)-3, the only other significant radioactive product detected by HPLC upon reduction was a faster-eluting (less lipophilic) peak. This was probably [11C]propanol, formed by LiAlH4 reduction of unreacted [11C]propionyl chloride, reduction of an O-acylated product, reductive cleavage of the intermediate amide (5), or a combination thereof.46 It was observed in some pilot runs that this byproduct was formed in significant amounts (up to 80%) during the reduction of the [11C]amide (5) with LiAlH4. Cooling the reaction mixture to below -30 °C before LiAlH4 addition minimized this side reaction. Reproducible and efficient HPLC purification of the crude reaction mixture required that (a) all the THF be evaporated and (b) the residue be briefly hydrolyzed by mineral acid before addition of buffer and application of the reaction mixture to the chromatography column. The regional rat brain biodistribution of [11C]-(+)-3 (Figure 2A) demonstrated that this radiotracer has an appropriate distribution for a D2 receptor imaging agent with high binding in the D2-rich striatum. In stark contrast, the enantiomeric radiotracer, [11C]-(-)-3, displayed no preferential uptake in any rat brain regions; instead, a homogeneous distribution was found throughout the brain (Figure 2B). Such stereospecificity in receptor binding was anticipated because (+)-3 and (-)-3 have a eudismic ratio of over 10 000 for affinity to the D2 receptor in homogenate binding studies.26 Compared to [11C]-2e, [11C]-(+)-3 had a higher signal, faster washout from receptor-poor brain regions such as the cerebellum, and higher specific binding ratios at all time points examined. These findings are in accord with the somewhat higher affinity of (+)-3 for the D2 receptor than 2e (Ki values of 0.14 versus 0.30 nM in the same assay)26 and the lower measured lipophilicity of (+)-3 (log P ) 2.14) compared to 2e (log P ) 2.54) at pH 7.4.

4158

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 12

Wilson et al.

A broad screen binding assay was employed to determine the in vitro receptor selectivity of (+)-3. The screen was designed to reveal even modest receptor affinities because a relatively high concentration of 1000 nM (+)-3 was used in the assays and a hit was defined as >25% inhibition. The results were used as a guide in the planning of the ex vivo blocking studies designed to assess the ex vivo specificity of [11C]-(+)-3 binding (vide infra). The pharmacological specificity of the binding of [11C](+)-3 in rat brain was scrutinized by preadministering specific drugs before radiotracer injection. The drugs were chosen in part on the basis of the in vitro broad screen assay results. Drugs known to bind to the DA D2 receptor such as haloperidol and raclopride reduced the specific binding ratios in a dose-dependent manner with the highest doses resulting in specific binding ratios near zero. “Cold” (+)-3 itself also reduced specific binding ratios to near zero, indicating that the binding of [11C]-(+)-3 in striatum was saturable. In contrast, drugs such as SCH 23390 (DA D1 antagonist), WAY 100635 (5-HT1A antagonist), clonidine (R-2 adrenergic agonist), or MK 912 (R-2 adrenergic antagonist) had no significant effect on specific binding ratios (Table 1). The extent of striatal binding to D3 receptors could not be evaluated because no D3 ligand could be found that displayed the necessary selectivity over D2 receptors. However, given the negligible density of D3 receptors in rat caudate and putamen,47 it is unreasonable to expect that any significant portion of the striatal binding of [11C]-(+)-3 is to the D3 receptor. In addition, [3H](+)-3 binding in striatum has been shown to be sensitive to Gpp[NH]p,29 which is uncharacteristic for ligand binding to the D3 receptor.48 The selective DA reuptake inhibitor, RTI-32, induced a significant dose-dependent decrease on [11C]-(+)-3 specific binding (73% at 10 mg/kg, Table 1), suggesting that, as anticipated, binding of the agonist radiotracer is sensitive to endogenous levels of DA. Previous studies have reported a dose-dependent decrease in [11C]raclopride binding of 10-25% following administration of the selective DA reuptake inhibitor GBR 12909, accompanied by a concomitant increase in DA levels in striatal extracellular fluid.49 Further evidence for [11C]-(+)-3’s endogenous DA sensitivity comes from the reduced binding, in a dose-dependent manner, upon pretreatment of the rats with the DA releaser amphetamine. However, the maximum effect size (38% reduction) was smaller than we anticipated, given the large doses and route of administration of amphetamine (up to 4 mg/kg iv), and only comparable to the effect of iv amphetamine on raclopride binding in rat.50 Thus, in this paradigm at least, we find no evidence that an agonist radiotracer is measurably more sensitive to DA release by amphetamine than the antagonist radiotracer raclopride. Amphetamine-induced changes in blood flow, metabolism, and/or protein binding may be obscuring our findings, and full kinetic modeling will be required to resolve the issue in a more quantitative manner. The opposite effect on the specific binding of [11C](+)-3, i.e., an increase, was seen in those experiments where the rats were pretreated by the DA depleting agents, reserpine and AMPT, either on their own or in combination. Here, we see the expected increases (30-

34%) in specific binding ratios resulting from the removal of the competing endogenous DA from the DA neuron milieu. The presence of significant quantities of radioactive metabolites in the imaging areas of analysis during a PET scan may confound the measurements and can present an insurmountable barrier to proper quantification of the images and interpretation of the results. While [11C]-(+)-3 is quickly metabolized in rat plasma, the polar radioactive metabolites formed do not appear to cross the blood-brain barrier. Greater than 98% of the radioactivity in the rat brain tissue, 40 min postinjection, is unmetabolized [11C]-(+)-3 at a time when most (74%) of the radioactivity in plasma arises from metabolites. Metabolism of [11C]-(+)-3 in humans may follow a different route from that found in rat because rat and human liver microsomes have been found to metabolize (+)-3 and structurally related D2 agonists in a different fashion.51 The optimal physicochemical properties of a potential PET radiotracer for imaging neuroreceptors have yet to be determined. In general, a balance in lipophilicity is sought between sufficient lipophilicity to ensure good blood-brain barrier penetration and low lipophilicity to try to minimize nonspecific binding. The log P at pH 7.4 of [11C]-(+)-3 was determined using the octanol/ buffer shake-flask method33 and was found to be 2.14, which is in the middle of the range of a variety of other useful neuroreceptor PET radiotracers.23,33 For comparison purposes, the value for [11C]-2e (2.57) was also measured. The determination and extent of plasma protein binding of a radiotracer or rather its inverse, the free fraction unbound to plasma proteins, is thought to be an important parameter in PET kinetic modeling and in determining the image quality obtained.37,52 Some radiotracers, e.g., [11C]-(+)-McNeil 5652, have too low (