Synthesis of [18F]GBR13119, a presynaptic dopamine uptake antagonist

Appl. Radiat. lsot. Vol. 39, No. 4, pp. 279-282, 1988 Int. J. Radiat. Appl. lnstrum. Part A

0883-2889/88 $3.00 + 0.00 Copyright @ 1988 Pergamon Press plc

Printed in Great Britain. All rights reserved

Synthesis of [18F]GBR13119, a Presynaptic Dopamine Uptake Antagonist MICHAEL

R. K I L B O U R N

and MICHAEL

S. H A K A

Division of Nuclear Medicine, Department of Internal Medicine, University of Michigan Medical Center, Ann Arbor, MI 48109, U.S.A. (Received 15 October 1987)

[I8F]GBRI3119 (l-[(4-[JSF]fluorophenyl)-(phenyl)methoxy]ethyl-4-(3-phenylpropyl) piperazine) has been prepared in no carrier added form by a four-step synthesis from [tSF]fluoride. Isolated yields are 7 10% (uncorrected) in a synthesis time of 120min. The product is obtained in high specific activity (> 1000 Ci/mmol) and high radiochemical purity (> 99%) without chromatographic purification. Small amounts of chemical impurity, identified as the nitro-substituted analog by independent synthesis, can be removed by HPLC. [18F]GBR13119 is proposed as a new radiotracer for the presynaptic dopamine uptake system.

promising results in preliminary PET studies in primates and humans. Nomifensine, however, exhibits Alterations in the dopaminergic neurons have been poor selectivity between adrenergic and dopaminerdemonstrated in neurological diseases (e.g. Parkingic uptake systems, and the hydroxylated metabolites son's disease) and implicated in psychiatric disorders of nomifensine exhibit considerable binding affinity (e.g. schizophrenia). Many investigators involved in for the dopamine receptor (Jacob et al., 1981). Positron Emission Tomography (PET) have apMore recently, a series of N, Nl-disubstituted proached the in vivo study of the dopamine system piperazines were reported (Van der Zee et al., 1980) through carbon-ll or fluorine-18 labeled dopamine which exhibited high affinity (nanomolar ICs0 values) receptor ligands (butyrophenone neuroleptics (Kilto the dopamine uptake system and good selectivity bourn and Zalutsky, 1985; Kilbourn and Welch, (DA/NE ratios of 30-100). One of these com1986; Swart and Korf, 1987) and raclopride (Farde et pounds, GBR12935 (l-[2-(diphenylmethoxy)ethyl]-4al., 1985) or labeled precursors of dopamine (3-phenylpropyl)piperazine), is available in tritiated ([1-HC]DOPA, 6-fluoroDOPA)). PET studies with form and has been used for numerous in vitro studies. these agents in Parkinson's patients (Leenders et al., Using rat brain tissue slices and quantitative in vitro 1986), schizophrenics (Wong et al., 1986) and MPTPautoradiography, the regional distribution of poisoned individuals (Perlmutter et al., 1987) have [3H]GBRI2935 (and thus, presumably, the dopamine demonstrated changes which may reflect the inuptake system distribution) was demonstrated (Dawvolvement of the dopaminergic system in these son et al., 1986). The affinity of [3H]GBR12935 has diseases. been thoroughly examined in synaptosomal memThe in vivo synthesis of dopamine from DOPA and brane preparations from rat brain and post-mortem the post-synaptic dopamine receptor are only two human brain tissues (Anderson, 1987; Janowsky et parts of the dopaminergic system. A third facet is the al., 1987~ Berger et al., 1985). Finally, and most presynaptic dopamine uptake system. A number of synthetic compounds (nomifensine, mazindol, encouraging for the use of a dopamine uptake antagmethylphenidate), have been found to be potent onist and positron emission tomography, binding of inhibitors of the dopamine uptake system, and would [3H]GBRI2935 was reported decreased in synbe candidates for radiolabeling for use in PET. aptosomal membrane preparations from patients Nomifensine (8-amino-4-phenyl-2-methyl-1,2,3,4- with Parkinson's disease (Janowsky et al., 1987). A number of fluorine-containing derivatives of tetrahydroisoquinoline) was recently labeled with GBR12935 were originally described by Van der Zee carbon-ll (Leenders et al., 1987) and has shown (Van der Zee et al., 1980), and all retained considerable biological activity and selectivity. In this Correspondence should be addressed to: Dr Michael R. paper, the synthesis of one of these compounds, Kilbourn, Cyclotron/PET Facility, 3480 Kresge III Bldg, 1-[(4-[ t~F]fluorophenyl) (phenyl)methBox 0552, The University of Michigan, Ann Arbor, MI GBR13119 oxyethyl-4-(3-phenylpropyl) piperazine, in no carrier 48109-0552, U.S.A.

Introduction

279

280

MICHAEL R. K1LBOU~N and MICHAEL S. HAKA

added, high specific activity fluorine-18 labeled form is described, hi vivo studies with this new fluorine-18 labeled radiotracer are reported elsewhere (Kilbourn, submitted for publication).

Experimental Materials and methods 4~Nitrobenzophenone, lithium aluminum hydride (LAH, 1M in THF), N-hydroxyethylpiperazine, I-bromo-3-phenylpropane, thionyl chloride, and dimethylsulfoxide were obtained from Aldrich Chemical Co. Thin layer chromatography (TLC) was done using Merck silica gel plastic backed TLC plates. N-(3-Phenylpropyl)-N-hydroxyethylpiperazine (4), GBRI2935, and GBRI3119 were prepared by literature methods (Van der Zee, 1980) and purified by silica gel preparative layer chromatography. HPLC was performed using Varian SI-10 (4mm × 30cm) and Phenomenex Maxsil 5 C8 ( 4 . 6 m m × 15cm) columns.

Preparation of tetrahutylarnmonium [lSF]fluoride [>F]Fluoridc ion was produced by proton irradiation of oxygen-18 enriched water (86% isotopic enrichment: Mound Laboratories) held in an allsilver cyclotron target (I mL target volume). The aqueous [~*F]fluoride was converted to a solution of tetrabutylammonium [~SF]fluoride by evaporation in a Vacutainer ~ (Becton-Dickinson # 6434)essentially according to the procedure of Brodack et al. (1986). Resolubilization efficiencies were 75-88%. 4-[l*F]Fluorobenzophenone (I). To a solution of [ISF]TBAF (1 102 mCi) in DMSO (100-200/tL) were added 2 mg of 4-nitrobenzophenone, and the solution heated (155- 165 C) for 25 min. Analysis of an aliquot by TLC (silica gel, 8/2 pentane/diethyl ether) showed an average ~*F incorporation of 52% (range 19- 87%). The [~F]fluorobenzophenone could be isolated in 15-85% yield and 99% radiochemical purity by water--ether partitioning of the products. Typically, the product was not isolated but immediat.ely reduced. 4-[ISF]Fluorobenzhydrol (2). The DMSO solution of crude ketone 1 was placed in an ice water bath but not allowed to freeze. LAH (300/,L of 1 M solution) was added, the vessel shaken, and then removed from the cold bath. After 1 min the vessel was returned to the cold bath and I mL of 6 N H:SOa added dropwise. The aqueous mixture was extracted with 1.5 mL of diethyl ether, and the ether layer separated and dried (NaeSO4) to yield a solution of crude alcohol 2. TLC analysis indicated quantitative reduction to the alcohol (silica gel, 8/2 pentane/diethyl ether, &. alcohol = 0.15, Rf ketone = 0.55). Yields of alcohol were 95% (from 1). The product was not further isolated but carried on to the chlorination step. 4-[~*F]Fluorobenzhydryl chloride (3). To the ether solution of alcohol 2 was added 300/~L of thionyl chloride, the vessel tightly capped, and the solution

heated (100:C) for 10min. TLC analysis showed conversions of 85-100% to the chloride (silica gel, 8/2 pentane/diethyl ether, chloride Rf = 0.5). [~*F]GBR 13119 (5). The solution of chloride 3 was evaporated (heat, N2 flow) to yield a small volume ( ~ 10/~ L) of yellow oil. To this was added a solution of 1-(2-hydroxyethyl)-4-(3-phenylpropyl)piperazine (10rag) in toluene (3 drops). The vessel was tightly capped and heated (160-170 C) for 20rain. After cooling, water (1 mL) was added and the aqueous mixture extracted with diethyl ether (2 x I mL portions). TLC analysis of the crude extract showed the desired condensation product in 65% yield (silica gel, 19/1 CHCI~/CH~OH, R~ chloride 3 = 0.74. RI 4 = 0.13, R~ 5 = 0.45). The ether was then extracted with 1 mL of 2 N H,SO4, the aqueous layer washed with I mL of ether, and the other layers discarded. The aqueous layer was neutralized (solid KeCO0 and extracted with diethyl ether. The organic layer was evaporated to yield the desired product 5. TLC analysis showed an average radiochemical purity of 99% . HPLC analysis (silica gel, 95,'5 CH,CI,/CH~OH, 1 mL/mm, R r = 6.4 min: C~, 604t) CH~CH/H20, ImL/min, R ~ = l l . 5 m i n ) showed a single radioactive product which co-eluted with authentic GBRI3119,

Results and Discussion The synthesis of [~F]GBRI3119, shown in Fig. 1, is based on the original synthesis of Van der Zee et ul. (1980) with modifications to allow incorporation of the radioisotope. Nucleophilic aromatic substitution of [~SF]fluoride-for-nitro in 4-nitrobenzophenone was straightforward: radiochemical yields (average 52%) are consistent with previously reported substitutions using pSF]fluoride ion (Berridge and Tewson, 1986). Reduction of the ketone in DMSO solution using LAH is interesting: the ketone reduction is quantitative and probably instantaneous, whereas the reduction of DMSO to dimethyl sulfide is slower and does not appreciably occur in the first minute. This unusual use of a hydride reagent in DMSO solution allows for a one-pot synthesis of the 18F-labeled alcohol 2 in overall yields of 50% (average). Alternatively, the ketone can be isolated via liquid liquid extraction or C~ SepPak bonded phase chromatography and subsequently reduced in ethereal solution, but there were no advantages in yields in that lengthy procedure. The alcohol 2 was isolated in diethyl ether solution but not purified before chlorination: unreacted [~SF]fluoride ion was effectively removed by extraction into the aqueous washing. Rapid chlorination of the alcohol 2 was effected using a large excess of thionyl chloride in ether solution. Chlorination is essentially quantitative to yield the desired [l~F]fluorobenzhydryl chloride and no side products. Again, the product was not isolated: prior to the final coupling with the

Synthesis of [~SF]GBR13119 0

OH

II

C

•~'I/ ~..~NO 2

281

C i,ii

~s F

~aF

Fig. 1. Synthesis of [18F]GBR13119. Reagents and conditions: (i) Bu4N~SF, DMSO, 160"C, 25 min; (ii) LAH, (~5°C, lmin, the H30+; (iii) SOCI,, Et20 , 100°C, 13min; (iv) N-(2-hydroxyethyl)N~-(3-phenylpropyl) piperazine, toluene, 160"C, 20 min. amino-alcohol 4, solvent and thionyl chloride are aration of [18F]GBR13119 for clinical human studies simply removed by evaporation. The volatility of would require removal of the nitro derivative; we SOC12 make it a convenient reagent in this regard. have successfully isolated [18F]GBRI3119 in such The final product, [~SF]GBRi3119, is formed by purified form by HPLC using a silica gel column heating a toluene solution of the chloride 3 and the (95 : 5 CH2CIjCH3OH eluant). amino-alcohol 4. At the high temperature used The synthesis described here has not been opti(160°C) the toluene boils away, leaving a dark brown mized but at present provides sufficient gum from which the product can be extracted. No [~SF]GBR13119 for animal studies. For example, added base is needed to catalyze the condensation from 102 mCi of tetrabutylammonium [tSF]fluoride reaction, and the product is obtained as the free base. were prepared 9.4 mCi of [18F]GBR! 3119, with a 2 h The desired product 5 can be simply separated from synthesis time. Attempts to shorten the synthesis by the neutral, hydrophobic by-products (unlabeled and performing an acid catalyzed condensation of ~SF-labeled) by simple acid-base extraction tech- 4-[18F]fluorobenzhydrol and aminoalcohol 4 were unniques, and the piperazine 5 obtained in > 9 9 % successful, and yielded only a high Rf (silica gel TLC) radiochemical purity without chromatography. product. Similarly, we have observed that poor conThe product [~8F]GBRI3119 is obtained along with version of the [~SF]fluorobenzhydrol (2) to chloride 3 a trace amount (micrograms) of the corresponding • (i.e. significant remaining alcohol) leads to very poor nitro substituted GBR derivative, as detected by TLC yields in the subsequent condensation reaction with 4. and HPLC. No cold fluoro compound could be Based on the sensitivity of our u.v. detector for detected, consistent with a synthesis starting with HPLC, we conservatively estimate that the specific high specific activity, no-carrier-added [lSF]fluoride. activity of the product [~8F]GBRI3119 is in excess of Formation of the nitro compound is similar to the 1000Ci/mmol. This is consistent with the use of situation with the butyrophenone neuroleptics pre- no-carrier-added [18F]fluoride ion as the synthetic pared from a nitroaromatic precursor and precursor, and similar to that reported for other [~8F]fluoride ion (Shiue et al., 1985)• In the synthesis syntheses involving aromatic nucleophilic substiof [XSF]GBR13119, however, most of the nitro- tution (Shiue et al., 1985). benzophenone precursor does not end up as the The basic structure of these GBR compounds disubstituted piperazine, but rather as other prod- offers numerous options for labeling with fluorine-18 ucts, possibly due to reduction of the nitro group in or carbon-I 1. Labeling with ~SF in the single phenyl the LAH step, or side product formation due to the group might be possible using nucleophilic aromatic higher reactivity of a p-nitrobenzyl halide (possibly substitution of the corresponding p-nitrophenyl proN-alkylation). We have independently synthesized piophenone analog, followed by reduction of the the nitro derivative of GBR13119 by the conden- ketone, approaches similar to those taken with the sation of 4-nitrobenzylhydryl chloride and N-(3- butyrophenone neuroleptics (Hamacher et al., 1986) phenylpropyl)-N-hydroxyethylpiperazine: under and the opiate fentanyl (Hwang et al., 1986). Labelreaction conditions identical to those used for ing in an aliphatic position within the propyl chain [18F]GBR131.19 preparation, this condensation pro- might be possible by alkylation of l-(diphenylceeds in very low yields with considerable formation methoxy)-2-(N-piperidyl) ethane with l-bromoof other products. The affinity of a nitro substituted 2-[JSF]fluoro-3-phenyl propane, for which a nopiperazine for the dopamine uptake system is un- carrier-added synthesis has previously been described known• Although suitable for animal studies, a prep- (Chi et al., 1986). Labeling with carbon-I 1 would be

282

MI('HAEL R. KILBOURN and MICHAEL, S. H,~.KA

feasible using the c o n d e n s a t i o n reaction in Fig. I and c a r b o n - I 1 labeled d i p h e n y l m e t h a n o l , which is again a literature c o m p o u n d ( D i s c h i n o el al., 1983). Alternatively, 3 - p h e n y l p r o p i o n i c acid could be labeled via ~[C-carboxvlation o f p h e n e t h y h n a g n e s i u m halide, a n d used in an acylation r e d u c t i o n s e q u e n c e as r e p o r t e d by Van der Zee c t a l , (1980). T h e s e alternatives in radioIabeling m a y p r o v e i m p o r t a n t : the time course o f the agents in civo r e m a i n s largel5 unknov~n, and the routes o f m e t a b o l i s m (and levels o f lipophilic r a d i o l a b e l e d m e t a b o l i t e s ) have i)ot yet been d e ! e r m i n e d . Summary A high specific a c t M t y , n o - c a r r i e r - a d d e d s y n t h e s e s o f a new r a d i o t r a c e r , [I~F]GBR13119, has been developed. T h e p r o d u c t can be o b t a i n e d in high radiochemical purily ( > 9 9 % ) , suitable l\)r m tit-o animal studies, w i t h o u t c h r o m a t o g r a p h i c purification. A n H P L C isolation system has been d e v e l o p e d for purification, and [ I q ; ] G B R 13119 o f sufficient radiochemical and chemical purities lbr h u m a n use can be o b t a i n e d F u r t h e r e v a l u a t i o n o['[~SF]GBRI3119 u s a ligand for P E T i m a g i n g o f the d o p a m i n e u p t a k e syslem in u n d e r w a y . "{'hiswork was supported by National Institutes of tlealth Grant 2-P01-NSI5655, and Nltt Trainmg Grant I-T32-NS07222-06. Aehnowh,d, eements

References Anderson I'. H. (1987) Biochemical and pharmacological characterization of [~H]GBRI2935 binding m vitro to rat striatal membranes: labeling of the dopamme uptake complex. J. Ncuro{'hem. 46, 1887. Berger P.. Janowsky A., Vocci F., Skolnick P., Schweri M. M. and Paul S M. I1985)[~H]GBRI2935: a specific high affinity ligand for labeling the dopamme transport complex. Eur. J. Pharma{ ol, 107, 289. Berridgc M. S and Yewson T J. 11986) Chemistr.', of lluorinc-18 radiopharmaceutical,. Ap[)l. Radiat. L~ot. 3% 685. Brodack J. W.. Kilbourn M. R., Welch M. J. and Katzenellenbogen J A. (1986) i6-[F-18]Fluoroestradiol: the effect of reaction vessel on fluorine-18 resolubilization. product yield, and effecti,,e specific acti\it). AppI. Radial. lsot. 37. 21"7. ('hi D. 5., Kiescv, etter D. O., Katzenellenbogen J. A., Kilbourn M. R. and Welch M. J. (1986) ttalofluorination of olelins: elucidation of the reaction characteristics and applications m labeling 'aith fluorine-18. J. Fluorine ('hem. 31, 99. Dawson T. M.. Gehlert D. R. and Wamsley J. K. (1986} Quantitative autoradiographic localization of the dopamine transport complex in the rat brain: use of a highly selective radioligand, pH]GBRI2935. Eur. J. Pharnlaeol. 126, 171. Dischino D. D., Welch M. J., Kilbourn M. R. and Raichle M. E. (1983) The relationship between lipophilicity and the brain extraction of carbon-ll radiopharmaeeuticals. d. Nuel. Med. 24, 1030. Farde L., Ehrin E., Eriksson I,., Grietz T., Hall H., ttedstrom C - G , Litlon J-E. and Sedvall G. (1985) Substituted

benzamides as ligailds for visualization of dopamine receptor binding in the human brain by positron Pro~ Natl. Acad. S~i. 1[2S'A 82, 3863. Hamacher K., Coenen H. H. and Stocklm G. (1986) N ( A radiofluorination of spiperone and .~-methyl-spiperone via ammopolyether supported direct nucleophilic substitution. J. l,ahelh, d (',mpd. Radiopharm. 23, 1047 [twang D-R., Fehu A [., Wolf A. P.. MacGregor R. R.. Fowler J. S , Arnelt C. D.. ttolland M. J.. Carr K. and Simon E. J. (1986) Synthesi:, and evaluation of fluorinated derivatives of fcntan)l as candidates for opiate receptor studies using positron emission lomogi'aphy ,/ Labelled ( ),npd. RadioptIarm. 23, 277 Jacob J, N.. Nichols D. E,, Kohli J. D. and Glock D (1981} Dopamine agonist properties of N-alkyl-4-{3,4-dihydroxyphenyl)-l.2,3,4-tetrahydroisoquinolines. ,I ~led. (Ttem. 24, I013. ,lanowsky A., Bergcr P., Vc.cci F , Labarca R.. Skoinick P. and Paul S. M. {[98,5) Characterization of sodiumdependent [~tt]GBRI2935 binding in brain: a radioligand for selective labelling of the dopamine transport complex J Vcurochem. 46, 1272 Janov.skv A., Vocci f.. Bcrger P.. Angel 1., Zelnik N., Kleinman ,I, E., Skolnick P. and Paul S. M. (1987l [;tt]GBRI2935 binding I{/ the dopamine transporter is decreased in tile caudatc ntlclcus in Parkinson's disease ,1..'¢{,urochem 49, 617. Kilhourn M, R., Welch M. J.. Donee C. S.. Tewson T. J., Saji tt. and Maeda M. (1984) Carrier-added and nocarrier-added syntheses of IF- 18]spiroperidol and IF- I 8] haloperidol. Int. ,I. Appl. Radial. l,~ot, 35, 591, Kilbourn M. R. and Zalutsky M. R, (1985) Research and clinical potential of receptor based radiopharmaceuticals. ,I. Nuel. Mcd. 26, 655 Kilbourn M. R. and Welch M. J. (1986) tiuorme-18 labeled receptor based radi{,pharmaceuticals..Ii}pI. Radial ls,t 37, 677 LeendersK. l , , P a h n c r A J , Q u i n n N.,Clark J . C . , F i r n a u (i.,Garnett E. S . , N a h m i a s ( ' , J o n e s T and MarsdenC. D. (1986) Brain dopaminc metabolism in patients with Parkinson's disease measured with positron emission tomography. J, \:eurol. Neurosurg. P~'vchiat. 49, 853. Leenders K. L., Aquilonms S-M., Eckernas S-A., Hartvig P.. Lundqvist H. and Langstrom B, (1987) Brain dopaminergic nerve terrninals assessed in viro using {[~C)nomifensinc and positron emission tomography (PET). J. Cerehral Blood Fh},' Metah. 7 {Suppl. I), $354. Perlmutter J. S., Kilbourn M R., Raichle M. E. and Welch M. J, (1987) MP'l-P-Induced up-regulation c,f in vh'o dopaminergic radioligand-receptor binding in man. Neurology 37, 1575. Shiue C-Y., Fowler J. S.. Wolf A. P.. Watanabe M. and Arnett C. D. {1985) Syntheses and specific activity determinations of no-carrier-added (NCAI ~SF-labeled butyrophenone neuroleptics benperidol, haloperidol, spiroperidol, and pipamperone. J. NueL Med. 26, 181. Swart J. A. A. and Korf J. (1987} In cii'o dopamme receptor assessment l\~r clinical studies using positron emission tomography. Biochem. Pharmaeol, 36, 2241. Van der Zee P., Koger H. S,, Gootjes J. and ttespe W. (1980) Aryl 1,4-dialk(en)yl-piperazines as selective and very potent inhibitors of dopamine uptake. Eur..l. Med. Chem. 15, 363. Wong D. F.. Wagner H N. Jr, Tune [. E., Dannals R. F., Pearlson G. D., Links J. M., Tamminga C. A., Broussolle E. P., Ravert H. T., Wilson A. A , Toung J. K. T., Malat J., Williams J. A., O'Tauma L. A., Snyder S. H., Kuhar M. J. and Gjedde A. (1986) Positron emission tomography reveals elevated Dz dopamine receptors in drugnaive schizophrenics. Science 234, 1558.