Dopamine Transport Sites Selectively Labeled by a Novel Photoaffinity Probe: L*%DEEP

The Journal of Neuroscience, August 1989, g(8): 2664-2670

Dopamine Transport Sites Selectively Labeled by a Novel Photoaffinity Probe: l*%DEEP

D. E. Grigoriadis,’ A. A. Wilson,2 R. Lew,’ J. S. Sharkey,3 and M. J. Kuharl ‘Neuroscience Branch, Addiction Research Center, National Institute on Drug Abuse, Baltimore, Maryland 21224, ‘Radiation Health, Johns Hopkins University School of Hygiene, Baltimore, Maryland 21205, and 3Department Clinical Neurosciences, Western General Hospital, Edinburgh, Scotland EH4, 2XU, United Kingdom

The dopamine transporter was labeled using a photosen- selective ligands that could label the site irreversibly. Photoaf- sitive compound related to GBR-12909, 125l-1 -[2-(diphenyl- finity labels are ligands that possessintrinsic selective affinity methoxy)ethyl]-4-[2-(4-azido-3-iodophenyl)ethyl]piperazine for a particular binding site and also contain a photolabile func- (Y-DEEP). ‘*%DEEP bound reversibly and with high affinity tional group that is capable of forming a covalent bond when to the dopamine transport protein in the absence of light and exposed to UV light. For example, the development of pho- could be covalently attached to the protein following ex- toaffinity probes as covalent labels for alpha- and beta-adren- posure to UV light. In rat striatal homogenates, Y-DEEP was ergic receptors (Hess et al., 1983; Leeb-Lundberg et al., 1984; found to incorporate covalently into a protein with apparent Stiles et al., 1984) as well as dopamine receptors(Amlaiky and molecular weight of 58,000 Da. The properties of this binding Caron, 1985; Lew et al., 1985; Redouaneet al., 1985; Niznik protein were characteristic of the dopamine transporter since et al., 1986; Grigoriadis et al., 1988) have served to elucidate covalent attachment could be inhibited by dopamine-uptake the molecular structure and characteristics of these receptors. blockers with the proper pharmacological rank order of po- In order to study the molecular componentsof the dopamine tencies. Covalent binding was also inhibited in a stereospe- transport site, it was usefulto develop a high-affinity photolabel cific manner by (+) and (-) cocaine, as well as other cocaine that could be usedto further characterize the dopamine trans- analogs. The protein was not found in the cerebellum. The porter. We now report the synthesisof a novel photoaffinity li- dopamine transporter appears to exist in a glycosylated form gand, T- 1-[2-(diphenylmethoxy)ethyl]-4-[2-(4-azido-3-iodo- since photoaffinity-labeled transport sites could adsorb to phenyl)ethyl]piperazine (lZSI-DEEP),and its ability to selectively wheat germ-agglutinin and could be specifically eluted from photolabel the dopamine transporter in rat striatum. the column by B-iV-acetylglucosamine. Materials and Methods Dopamine reuptake is thought to be the mechanismof inacti- GBR- 12909, GBR- 12935, and other closely related compounds are the vation of releaseddopamine at the synapse.Dopamine uptake most selective inhibitors of dopamine uptake known (Van der Zee et al., 1980; Berger et al., 1985; Bonnet et al., 1986; Janowsky et al., 1986). has beencharacterized by several investigators (for reviews, see Accordingly, these compounds were examined for possible sites of al- Snyder, 1970; Iversen, 1973; Kuhar, 1973; Horn, 1978; Raiteri teration to introduce an iodo-azido moiety, thus creating a selective, et al., 1978). More recently, the transporter has been studied by high-affinity photoaffinity radiolabeled probe. As shown by Van der Zee in vivo binding (Bonnet and Costentin, 1986; Andersen et al., et al. (1980) phenyl alkyl substitutions on the piperazine ring of derivatives of 1-[2-(diphenylmethoxy)ethyl]piperazine yield potent inhibitors 1987; Benmansour et al., 1987; Chagraoui, et al., 1987; Kil- of dopamine uptake. Several iodo-azido compounds were synthesized bourn, 1988), as well as by in vitro binding techniques(Javitz and tested for uptake inhibition in the dark. The most potent analog et al., 1983; Berger et al., 1985; Bonnet et al., 1986; Janowsky tested was IZSI-DEEP, and it was therefore selected as the ligand for et al. 1986; Andersen et al., 1987). Recently, Ritz et al. (1987) subsequent studies. and Sharkey et al. (1987) have provided evidence that the 3H- mazindol binding site and the 3H-GBR-12935 binding site on Synthesis of 1z51-DEEP the transporter have the properties of a cocainereceptor related All new compounds gave satisfactory spectral (NMR and IR) and ele- mental (C, H, N) analyses. Purifications and analyses of ‘2SI-containing to the reinforcing properties of cocaine. radioactive mixtures were performed on an HPLC system composed of While considerable progresshas been made in the biochem- a Rheodyne 7125 injector, 2 Waters 5 10 EF pumps, a UV detector ical and pharmacological characterization of the dopamine (Waters 48 l), and an Ortec flow radioactivity detector. Peak areas were transporter using tritiated selective dopamine-uptake blockers measured using Hewlett-Packard 3990A recording integrators. Sodium in binding studies(Van der Zee et al., 1980; Berger et al., 1985; 1251-iodide was obtained from Amersham Corp. (IMS-30 or IMS-300). 1-[2-(Diphenylmethoxy)ethyl/-4-[2-(4-nitropheny~ethy~piperazine. A Bonnet and Costentin, 1986; Bonnet et al., 1986; Janowsky et solution of 4-nitrobenzyl bromide (2.0 gm, 8.7 mmol) 1-[2-(diphenyl- al., 1986) attempts to study the molecular componentsof the methoxy)ethyl]-piperazine (2.58 gm, 8.7 mmol), and tetramethyl pi- transport site have been hindered by the lack of specific and peridine (3 ml) in acetonitrile (20 ml) was stirred under reflux for 6 hr. Ether (50 ml) was added to the cooled solution and the mixture filtered. The filtratewas evaporated to drynessand the residuechromatographed on silica [EtOAc:Pet.ether:Et,N, 400:100:25 (vol/vol)] to give a pale Received Oct. 11, 1988; revised Jan. 9, 1989; accepted Jan. 13, 1989. yellow oil (2.7 gm, 70%). Correspondence should be addressed to Dr. Michael J. Kuhar, NIDA Addiction 1-[2-(Diphenylmethoxy)ethyl]-4-[2-(4-aminophenyl)ethyl]piperazine Research Center, P.O. Box 5 180, Baltimore, MD 21224. (Fig. 1, I). A mixture of 1-[2-(diphenylmethoxy)ethyl]-4-[2-(4-nitro- Copyright 0 1989 Society for Neuroscience 0270-6474/89/082664-07%02.00/O phenyl)ethyl]piperazine (2.5 gm, 5.6 mmol) and 5% palladium on char- The Journal of Neuroscience, August 1989. 9(E) 2665 coal (0.4 gm) in ethanol (30 ml) was stirred under 1 atmosphere of hydrogen for 3 hr and then filtered. The filtrate was evaporated to dryness to leave a colorless oil (2.3 1 gm, 99.2%). The trihydrochloride salt was recrystallized from ethanol. - 1-/2-(diuhenvlmethoxv~ethvll-4-/2-(4-amino-3-iodoohenvllethvll- pipe~azz%A solution of’iodine monochloride (0.24 gm, 1.45 m&l) in acetic acid (1.3 ml) was added slowly to a stirred solution of 1-[2- (diphenyl-methoxy)ethyl]-4-[2-(4-aminophenyl)ethyl]piperazine (0.6 gm, 1.44 mmol) in acetic acid (3.3 ml) under argon at room temperature. After 15 min, the reaction was quenched by the addition of saturated aqueous Na,CO,. The mixture was extracted with ether (2 x 20 ml), a) Na”‘I, pH 5 which was washed with 20 ml aqueous Na,S,O, (1 M) and 20 ml brine, chloramine-T dried (NaSO,), and filtered. The resultant red oil was chromatographed on silica [EtOAc:Pet.ether:Et,N, 350: 100:25 (vol/vol)] to give the prod- b) NaNO,, AcOH uct as a pale yellow oil (0.5 1 gm, 65%). Iz71-DEEP 1-[2-(diphenylmethoxy)ethyl]-4-[2-azido-3-iodophen- c) NaN, yl)ethyl]piperazine(‘z71-DEEP). An ice-cold stirred solution of l-[2-(diphenylmethoxy)ethyl]-4-[2-(4-amino-3-iodophenyl)ethyl]piperazine t (0.37 gm, 0.68 mmol) in aqueous acetic acid (28 ml, 3 M) and methanol (8 ml) was treated with a solution of sodium nitrite (65 mg, 0.94 mmol) in water (2 ml) dropwise. The mixture was stirred for 20 min and a solution of sodium azide (63 mg, 0.97 mmol) in water (2 ml) added dropwise. The solution was stirred at room temperature for 20 min and cont. aqueous NH,OH (8 ml) was added cautiously. The mixture was extracted with CH,Cl, (2 x 20 ml), dried (Na,SO,), and filtered. Evap- oration of the solvent left an oil that was chromatographed on silica [EtOAc:Pet.ether:Et,N, 250:250:25 (vol/vol)] to give the product as a pale brown oil (0.3 gm, 77%). ‘2sI-I-[2-(diphenylmethoxylethyl]-4-[2-(4-azido-3-iodophenyl)ethyl]- Figure 1. Synthesis of Y-DEEP. ‘Y-l-[2-(diphenylmethoxy)ethyl]- DiDerazine(/‘2sIlDEEPI (Fias. 1. 2). This was vrevared from l-12-(di- 4-[2-(4-azido-3-iodophenyl)ethyl]piperazine (‘Y-DEEP; 2) was pre- phenylmeti;ox;)ethyl]:4:[21(4-amjnophenyl)ety~]piperazine using a pared from the corresponding 4-aminoethyl compound (I) by a 3-step, 3-step, 1-pot method for the conversion of an aminophenyl moiety into l-pot method for the conversion of an aminophenyl moiety into an 12sI- an ‘Y-labeled phenylazide. Semipreparative reverse-phase HPLC was labeled phenylazide. Semipreparative reverse-phase HPLC was used to used to isolate 1251-DEEP. The radiochemically pure product (>99%) isolate IY-DEEP. The radiochemicallv pure product (>99%) was ob- was obtained in 45-50% isolated radiochemical yield from starting lzsI- tained in 45-50% isolated radiochemical-yields from starting lisI-iodide iodide and had identical chromatographic (HPLC and TLC) properties and had identical (HPLC and TLC) properties to authentic 1251-DEEP. to authentic Y-DEEP. Specific activities were determined by an HPLC Specific activities were determined by an HPLC method (Wilson et al., method (Wilson et al., 1986) to be between 800 and 1500 Ci/mmol 1986) to be between 800 and 1500 Ci/mmol. (refer to Fig. 1). 3H-GBR- 12935 was obtained through New England Nuclear (Boston, MA). All other standard chemicals and reagents were obtained through either Sigma Chemical Co. (St. Louis, Mb) or Bio-Rad (Richmond, bard, 1980). Protein determinations were carried out by the method of CA). Wheat germ aaglutinin-Sevharose 6B was vurchased from Phar- Lowry et al. (195 1) with BSA as the standard. ma&a LKB (PleasantHill, CA).- Covalent photoa$inity labeling of striatal membranes Membrane preparation For the photoaffinity labeling of rat striatal homogenates, membranes Male Sprague-Dawley rats (Harlan Ind, IN) were killed by decapitation, were incubated with l*$I-DEEP (final concentration, l-2 nM) as de- and the brains were quickly removed on ice. Striata were dissected and scribed above for the reversible studies. Following the 60 min incubation quick-frozen in liquid nitrogen and stored whole at -70°C until day of at 4°C 1 ml aliquots were placed in 35 x 10 mm petri dishes, placed assay. On the day-of assay,striata were thawed, weighed, and homog- 11 cm below an 85 W Hg lamp (Thomas Scientific, NJ), and exposed enized in 10 ml buffer containing 50 mM Tris-HCl. 120 mM NaCl. 5 to direct UV light for 40 set without stirring (path length or depth of mM KCl, 1 mM phenylmethylsuifonyl fluoride (PI&F), and 1 &ml membrane suspension was approximately 3-4 mm). Membranes were leupeptin (assay buffer) using a Brinkmann polytron (setting 6) for 20 then transferred directly into clean eppendorf tubes, washed twice with sec. Membranes were centrifuged at 20,000 rpm for 10 min at 4°C the assay buffer by centrifugation, and processed for electrophoresis as de- supernatant discarded, and the pellets resuspended in 10 ml buffer and cribed below. centrifuged again at 20,000 rpm for 10 min at 4°C. The final pellets were resuspended in the above buffer to a working concentration of 20 SDS-PAGE mg/ml original wet weight (o.w.w.) and kept on ice until use. Samples for electrophoresis were resuspended in 120 ~1 SDS-sample buffer containing 50 mM Tris-HCl. 2% SDS. 10% elvcerol. 5% a-mer- Reversiblebinding of lzsI-DEEP to rat striatal homogenates captoethanol, and 0.005% bromphenol blue (pH 6.8, 22°C) and incu- Reversible binding of 1251-DEEP was performed in the dark using a bated for 60 min on a vortex mixer. Following the incubation, 100 ~1 centrifugation binding assay. Typically, 100 ~1 of the membrane prep- ofthe soluble sample was loaded on a discontinuous slab gel (6% stacking aration was added to tubes containing either )H-GBR-12935 (final con- and 10 or 12% running) according to the method of Laemmli (1970). centration, 1 nM) or 1251-DEEP (final concentration, l-2 no), as well as An equivalent amount of protein was loaded onto each lane (200-400 100 ~1 of the competing ligands in a final volume of 1.O ml. Incubations pg) and typically run overnight at a constant current of 10-20 mA. were carried out on ice (04°C) for 60 min and the reaction terminated Boiling the samples for 10 min prior to electrophoresis did not alter the by centrifugation in a Beckman microfuge for 10 min at 12,000 rpm. labeling pattern of Y-DEEP to membrane homogenates (data not Aliquots of the supematant (100 ~1) were removed and monitored for shown). Prestained molecular-weight protein standards (Sigma) were radioactivity in order to determine the amount of “free” ligand present included on each gel and used to calculate the apparent molecular weights in the tubes. The pellets were washed twice by centrifugation and count- of the labeled species. Following electrophoresis, the gels were dried (for ed for bound radioactivity in an LKB gamma-counter at 80% efficiency autoradiography) using a Bio-Rad slab gel drier or cut into lanes, sliced for ‘251-experiments or a Beckman 3801 liquid-scintillation counter at into 2 mm slices and monitored for radioactivitiy. For autoradiography, 50% efficiency for tritium. Data analysis was performed using the non- dried gels were exposed to Kodak X-AR5 film using Lightning-Plus linear least-squares curve-fitting program LIGAND (Munson and Rod- intensifying screens at -70°C for a period of 5-10 d. 2666 Grigoriadis et al. * Dopamine Transporter

Figure 2. Competition of ‘H-GBR- 12935 and “‘I-DEEP in rat striatal homogenates. Y-DEEP competed for 3H- GBR- 12935 (final concentration, I nM) under nonphotolyzing conditions. Rat striatal homogenates (final concentration, 2 mg/ml wet weight) were incubated for 60 min at OqC in a final volume of I ml, and the reaction was terminated by centrifugation. Aliquots (100 bl) of the supematant were counted to determine the “free” concentration of ligand present in each tube. Pcl- lets were washed twice without rcsuspcnsion with ice-cold buffer by centrifugation and counted for bound radioactivity. Total binding was typically 20.000 DPM, while nonspecific binding was approximately 7000 DPM, representing a 65% signal to noise ratio. IC,,, values were estimated by the non- linear least-squares curve-fitting program LIGAND (Munson and Rodbard, 1980) and are given in the text. The 0 above experiment was performed in ,o-~i “‘;‘;;-I’ ““;‘;-I” “‘1’;;-9’ ‘“‘ii-a’ “‘1’6-7’ “‘$6’ “70-5 triplicate and is representative of 2 independent determinations with similar results. 12’1-DEEP Concentration (M)

Wheat germ aggiurinin f WGA) afinitv chromatography binding (1 nM) in rat striatal homogenates using GBR- 12909 as Affinity chromatography on WGA-Sepharose was performed on mem- a blank, under nonphotolyzing conditions (in the dark). 12’1- branes that were photolabeled as described above. Following photolysis, DEEP was found to bind with high affinity to the dopamine the membranes were washed by centrifugation and resuspended in assay transport protein in rat striatal homogenates with an apparent buffer containing digitonin (Waco, TX) at a final concentration of 1%. Samples were incubated at 4°C for 60 min and centrifuged at 110,000 x g for 60 min. The supematant was passed through a Millex 0.22 pm filter (Bedford, MA), loaded onto 20 ml of WGA-Scpharose beads in a column equilibrated with assay buffer plus 0.1% digitonin and allowed to incubate on the column for 45 min at 4°C. Following the incubation, the column was washed in 100 ml (5 column volumes) assay buffer plus 0. I% digitonin before clution. Elution was performed in the same buffer as above with 200 mM /3-N-acetylglucosamine (GlcNac) at 1 ml/min 8 300 flow rate and fractions collected and monitored for radioactivity. Results 2 .E 200 Reversiblebinding analysis of DEEP \+Ythrat striatal membranes 6 NON-SPECIFIC In order to determine the binding characteristics of DEEP, un- 100 labeled (‘*‘I-DEEP) DEEP was competed for jH-GBR-12935

0 I Table 1. Inhibitory order of potencies of drugs on 1>51-DEEP binding 0 10 20 30 40 50 0 to rat striatal membranes SLICE NUMBER

Figuw 3. SDS-PAGE labeling pattern of photolabeled dopamine Compound ZM) transporter using ‘Y-DEEP. Membranes wcrc incubated in the dark (final concentration, 2 mg’ml o.w.w.) with l-2 nM ‘lSI-DEEP for 60 min GBR- I2909 36.3 at 0-4”C. Nonspecific binding was determined in the presence of I PM Mazindol 20.1 GBR- 12909. Following the incubation, I ml aliquots were exposed to Nomifcnsine 138.2 IJV light for 40 set and centrifuged for 15 min at 12,000 rpm. The Citalopram > I0,000 pellets were washed twice by centrifugation, solubilized in SDS sample Desipramine > 10,000 buffer, and processed for electrophoresis. Gels were typically run overnight at a constant current of 10-20 mA and the lanes cut into 2 mm The data show that competition of various compounds for the reversible bmdmg slices and monitored for radioactivity in a gamma counter. Typically, ol‘ ‘:‘I-DEEP (final concentration, I-2 nM) exhibits the proper pharmacological 25,000-30,000 cpm were loaded onto each lane. The efficiency ofcova- rank-order profile for the dopaminc transporter and corresponds with the lent labeling was calculated to be approximately 5-6% of specific bound pharmacology observed on the 58,000 Da protein covalently labeled and DEEP. Prestained molecular-weight standards were run on each gel and demonstrated in Figure 4. All compounds inhibited the binding of “?I-DEEP to striaral membrane homogenates in a monophasic manner, demonstrating binding used to estimate the molecular weights of the bound species. In rat to a single class of binding sites. All curves were performed in triplicate using the striatum, a major protein with an apparent molecular weight of 58,000 centrifugalion assay described in Materials and Methods and analyzed using the Da was labeled and could be inhibited by the addition of I PM GBR- nonhnear least-squares curve-fittmg program IKAND (Munson and Kodbard, 1980). 12909. The Journal of Neuroscience, August 1989, 9(8) 2667

12’ I-DEEP Mr (k)

180-r, 116a

84+

Figure4. Pharmacologicalspecificity of “‘I-DEEP labeledprotems. Mem- braneswere preparedas describedin Materialsand Methodsand Figure 3, and incubatedm the presenceof various compounds.The resultsdcmon- stratethat the labeledprotein migrating at 58,000Da representsthe dopamine- 36+ transporterbinding site. Other nonspe- cificallylabeled proteins were apparent but were not inhibited by any of the dopaminergichgands tested. Similarly, the serotonin-uptakeblocker citalo- 26+ pram and the norepmephnne-uptake inhlbitordesipraminc could not inhibit the covalent incorporation of “‘I-DEEP to any of the protems labeled in striatal membranehomogenates. The majorla- beledband at 58,000 Da wasdefined by both rank-order potencm and ste- reospecificityto bethe dopammetrans- porter. This experimentIS rcpresenta- tive of 4 independent studies that 1lOpM] y yieldedidentical results. affinity (K,) of IO-20 nM (seeFig. 2). Figure 2 also demonstrates polyacrylamide gelsas describedin Materials and Methods. lZsI- that the binding was monophasicand thus representedbinding DEEP was mainly incorporated into a protein of molecular to a single classof sites. These data are in agreementwith the weight 58,000 Da as observed by direct measurementof radio- valuesobtained from direct saturationisotherms using lZ51-DEEP activity in the gel (Fig. 3). Binding to this protein was blocked (data not shown). by the addition of GBR-12909, a selective dopamine-uptake blocker. Autoradiograms of gels also revealed that ‘151-DEEP Pharmacological spec(ficity of ‘“‘I-DEEP binding wasbeing incorporated into other proteins but to a lesserextent; Various ligandswere competed for lZSI-DEEPbinding (1-2 nM theseproteins had apparentmolecular weights of 70,000,45,000, final concentration), again under nonphotolyzing conditions, and 30,000, and <20,000 Da (Figs. 4, 5). their apparent K, valuesare listed in Table 1. Reversible.binding In order to further demonstratethat the protein band labeled of “‘I-DEEP revealed that GBR-12909, mazindol, and nomi- at 58,000 had pharmacologicalproperties characteristic of the fcnsine could inhibit the binding with K, values similar to those dopamine transport protein, various compoundswere incubat- reported for -‘H-GBR- 12935binding to the dopamine transport ed with ‘251-DEEPprior to photolysis and’the membranessol- site (Andersen, 1987). The serotonin transport blocker citalo- ubilized in SDS-sample buffer and processedfor electropho- pram. as well as the adrencrgic-uptake inhibitor desipramine, resis. As clearly shown in Figure 4, in the absence of any had very little effect on the binding of Y-DEEP even at high competing ligands (total), 12%1-DEEPwas incorporated into a concentrations. maj’or labeled protein migrating with an apparent molecular weight of 58,000 Da. In the presenceof 10 PM GBR-I 2909, 10 Covalent photoincorporation of “‘I-DEEP to rat striatal PM mazindol, or 10 PM dimethocaine, photoincorporation into membranes this protein was completely inhibited. (+)Cocaine, 10 PM, did lZSI-DEEP was photoincorporated into rat striatal membrane not inhibit the incorporation of ‘>‘I-DEEP, while the samecon- homogenatesby exposure to UV light. The membraneswere centration of (-)cocainc was sufficient to block the covalent solubilized in SDS samplebuffer and elcctrophoresedon SDS- attachment. Similarly WIN-35,065-2 could inhibit the incor- 2668 Grigoriadis et al. l Dopamine Transporter

STRIATUM CEREBELLUM Mr (W I

180+ 116+ 84+

58+

48+

Figure 5. Covalent incorporation of i251-DEEPin striataland cerebellar homogenates.Rat striatal and cerebellar homogenates were prepared in parallel, 36.cc incubated with l-2 nWSI-DEEP in the presence or absence of the compounds listed above. Assays were conducted for 26-, 60 min at 0-4”C andphotolyzed as de- scribedunder Materials and Methods. The autoradiogram shown is overex- posed to bring out minor peaks of radioactivity in orderto morethoroughly assessspecificity; the bulk of radioactivity, however,is in the 58,000 Da band,as shownin Figure 3. Although identical photoincorporationof Y- DEEPto manynonspecific proteins was evidentin bothtissues, there was a conspicuousabsence of any labeledproteins in the cerebellum with an apparent molecular weight of 58,000 Da.

poration of ‘*Y-DEEP, while WIN-35065-3 wasineffective. In other experiments, we also found that 1 I.LM GBR-12909, 1 PM Adsorption of IzsI-DEEP labeledprotein to mazindol, and 1 KM nomifensine inhibited photoincorporation WGA-Sepharose6B of ‘ZSI-DEEPinto the 58,000 Da protein (data not shown).These To further elucidate some of the molecular properties of the data demonstrate that the covalent labeling of the dopamine dopamine transport site, photolabeled proteins were solubilized transport protein by 12*1-DEEPwas stereospecific and could be in 1% digitonin and adsorbed onto WGA-Sepharose as de- inhibited by the proper pharmacological rank-order profile of scribed in Materials and Methods. We found that covalently compounds (Ritz et al., 1987)expected for the dopamine trans- labeledproteins could be adsorbedand specifically eluted from porter. The serotonergicuptake blocker citalopram and the ad- WGA using200 mM of the sugar,6-N-acetylglucosamine. Figure renergic uptake inhibitor desipraminewere ineffective in block- 6 illustratesthe completewash and elution profiles of 1251-DEEP- ing the covalent attachment of 1251-DEEPto the dopamine labeledtransport proteins from WGA. The figure demonstrates transport binding site (Fig. 4). that following 100 ml (5 column volumes) of wash, there was In order to further establish that 1251-DEEPwas specifically very little detectable radiolabel in the fractions, although 90- labeling the dopamine transport protein, we prepared mem- 95% of the radioactivity loaded onto the column was removed brane homogenatesfrom rat cerebellum, photolyzed them in during the wash. Following the onset of elution with 200 mM the presenceof 1251-DEEP(l-2 nM final concentration) with or GlcNac, there was robust and rapid desorption of labeled ma- without competing ligands, and compared the labeling pattern terial off the column, with more than 90% of the protein re- on SDS gels to that observed in the rat striatum. Figure 5 dem- moved in one column volume (20 ml) following the exclusion onstratesthat although identical nonspecifically labeledproteins of the first 10 ml of elution. This indicates that the protein appearboth in the cerebellumas well asthe striatum, the 58,000 contains either N-acetylglucosamineresidues or terminal sialic Da protein that exhibits a pharmacological profile of the do- acid residues.Further studieswill be required to determine the pamine transport protein is not evident in the cerebellum. It is quantity and nature of the glycosylation siteson the dopamine interesting to note that the nonspecifically labeled proteins in transport protein. the cerebellum could not be inhibited by either the serotonergic uptake blocker citalopram or the adrenergicuptake blocker de- Discussion sipramine (Fig. 5), thus strengthening the evidence presented GBR 12909 and related compounds have been shown to be that ‘*Y-DEEP was not labeling other monoaminergic uptake highly specificinhibitors of the dopaminetransport system(Van systems. der Zee et al., 1980; Heikkila and Manzino, 1984; Bonnet et The Journal of Neuroscience, August 1989, 9(8) 2669

2x105 - Ii 200 mM O-GlcNac

Figure 6. Wheat germ agglutinin-af- WASH ELUTION fmity chromatographyof i2SI-DEEP-la- E 1x10-6- beled proteins. Membrane homoge- nateswere prepared and photolabeled asdescribed. Following photolysis, the 2 membraneswere washed twice by centrifugation and resuspendedin assay E buffer(final concentration, 100 ma wet weight/ml)with 1%digitonin and sol- 2 TOTAL ubilizedfor 45 min at 0-4”C in a final volumeof 10 ml. Solubilizedproteins &x10+ wereloaded onto a 20 ml columnof J WGA-Sepharoseand allowedto incubate on the columnfor 45 min. The I columnwas washed at a flow rate of 1 1 pM GBR-12909 ml/minwith 5 columnvolumes ofbuff- er (100ml) containing0.1% digitonin. The columnwas eluted at a flow rate of 0.5 ml/min with assaybuffer con- O- •l taining0.1% digitonin and 200 mM /3-N- acetylglucosamine.lZ51-DEEP labeled 0 20 40 60 80 100 120 140 160 180 200 proteinswere specifically eluted in the first 10ml of elution. Thisexperiment FRACTION NUMBER wasrepeated with identicalresults.

al., 1986; Andersen, 1987; Chagraoui et al., 1987). 3H-GBR- membraneproteins weresubjected to SDS-PAGE. The gelswere 12935has beenshown to be a potent and highly selectiveligand sliced into 2 mm slicesand monitored for incorporated radio- for binding to dopamine transporters (Berger et al., 1985; Jan- activity. The major peak of incorporation occurred at an ap- owsky et al., 1986; Andersen et al., 1987). Since it has been parent molecular weight of 58,000 Da, and 1 PM GBR-12909 determined that the piperazine-containing portion of the GBR inhibited this incorporation, suggestingthat the protein labeled compoundscould be altered and yet retain high-affinity binding at 58,000 Da was the dopamine transporter. Conclusive evi- to and specificity for the transporter (van der Zee et al., 1980) dence that the 58,000 Da protein did indeed representthe do- we utilized the strategy of developing an iodo-azido moiety at pamine transporter wasobtained from the pharmacologicalpro- this portion of the molecule. After examining several iodo-azi- file of the photoincorporated proteins. do-GBR derivatives, lZSI-DEEPwas selectedas a high-affinity In the presenceof various compounds used to identify the specific ligand for the dopamine transporter. dopamine-uptakesite, the incorporation of lZSI-DEEPcould be Under reversible conditions, 12’I-DEEPinhibited the binding inhibited by compoundssuch as GBR- 12909, mazindol, nom- of 3H-GBR- 12935 with a K, value of 1O-20 nM that correspond- ifensine, dimethocaine, (-)cocaine, and the cocaine analog ed closely to the affinity of a related molecule, GBR- 12909. The WIN-35,065-2. Furthermore, the 58,000 Da covalently labeled affinity of lZ51-DEEPdetermined by direct saturation isotherms protein exhibited stereoselectivity with respect to (+) and was found to be slightly higher K, = 2 nM) than the K, value (-)cocaine, as well as the stereoisomersof the cocaine analogs obtained from competition experiments. The pharmacological WIN-35,065-2 and WIN-35,065-3. In addition to the phar- rank order of potencies of compounds that inhibited DEEP macological data, photolyzed membranes of rat cerebellum binding was found to be characteristic of the dopamine trans- showed a conspicuousabsence of any labeled protein with a porter. GBR- 12909 had approximately the sameaffinity as ma- molecular weight of 58,000 Da, strongly suggestingthat the zindol, while both had greater affinity than nomifensine. The protein labeled in the rat striatum centered at approximately serotonin-uptake blocker citalopram and the noradrenergicup- 58,000 Da was the dopamine-transport site. These data taken take inhibitor desipraminewere both without effect. Thesedata together demonstratethat the 58,000 Da protein has the same indicate that so long as the integrity of the azido moiety remains pharmacologicalcharacteristics and regional localization asso- intact (in the dark), 1251-DEEPinteracts with the dopaminetrans- ciated with the dopamine transporter. Other laboratories have port site with the samereversible characteristics as any of the solubilized other neurotransmitter transporters such as those establishedGBR analogs. for GABA (Radian et al., 1986) and serotonin (Rehavi et al., Following photolysis, when the incubation mixtures were ex- 1982; Habert et al., 1986). posed to UV light under conditions in which specific binding To further elucidate some of the biochemical properties of of lZSI-DEEP to dopamine transporters was occurring, we ob- the dopamine-transport site, photolabeled proteins were ad- served incorporation of radiolabeled drug into various mem- sorbed onto wheat germ agglutinin and were specifically eluted brane proteins. In order to obtain an initial estimate of the from the lectin with the appropriate sugar. Since wheat germ molecular weight of the dopamine transporter, photolabeled agglutinin recognizesand bindsto N-acetylglucosamineresidues 2670 Grigoriadis et al. * Dopamine Transporter as well as terminal sialic acid residues on carbohydrate chains brane and affinity purified receptors. Proc. Natl. Acad. Sci. USA 80: (Goldstein et al., 1965: Bhavanandan and Katlic, 1979), these 2102-2106. Horn, A. S. (I 978) Characteristics of neuronal dopamine uptake. Adv. data indicate that the dopamine transporter (or the portion of Biochem. Psychopharmacol. 19: 25-34. the transport complex that was photolabeled) contained either Iversen, L. L. (1973) Catecholaminc uptake processes. Br. Med. Bull. sugars of the N-linked type or terminal sialic acid residues. 29: 130-135. Further studies are required in order to determine the exact Janowsky, A., P. Bergcr, F. Vocci, R. Labarca, P. Skolnick, and S. M. nature and amount of glycosylation of the dopamine-transport Paul (I 986) Characterization ofsodium-dependent [‘H]GBR-I 2935 binding in brain: A radioligand for selective labeling of the dopamine site. transport complex. J. Neurochem. 46: 1272-1276. In summary, we have described a novel iodinated photoaf- Javitz, J. A., R. 0. Blaustcin, and S. H. Snyder (1983) [‘H]Mazindol finity ligand (lZSI-DEEP) that can specifically label the dopamine binding associated with neuronal dopamine uptake sites in corpus transporter or some component of it. This protein has an ap- striatum membranes. Eur. J. Pharmacol. 90: 46 l-462. Kilboum, M. R. (1988) In vivo binding of [‘“F]GBR-I31 I9 to the parent molecular weight of about 58,000 Da and has all the brain dopaminc uptake system. Life Sci. 42: 1347-1353. pharmacological and biochemical characteristics of the well- Kuhar, M. J. (1973) Neurotransmitter uptake: A tool in identifying established dopamine-transporter site previously described. The neurotransmittcr-specific pathways. Life Sci. 13: 1623-1634. protein appears to exist in a glycosylated form, and further work Laemmli. U. K. (1970) Cleavage of structural proteins during the is necessary for fully characterizing the type and extent of gly- assembly of the head of bacteriophage T4. Nature 227: 680-685. Leeb-Lundbera. L. M. F.. K. E. G. Dikinson. S. L. Heald. J. E. S. cosylation. As a photoafftnity ligand, ‘251-DEEP should prove Wikberg, P.-G. Hagen, J. F. Debernardis, M. Winn. D. L. Anderscn, to be a useful tool in the further characterization. isolation, and R. J. Lclkowitz, and M. G. Caron (1984) Photoafftnity labeling of purification of the dopamine transporter. mammalian alpha,-adrenergic receptors: Identification of the ligand binding subunit with a high affinity radioiodinated probe. J. Biol. Chem. 259: 10909-10915. Lcw, J. Y.. E. Mcller, and M. Goldstein (I 985) Photoaffinity labeling References and purification of solubilized D, dopaminc receptors. Eur. J. Phar- Amlaiky, N.. and M. G. Caron (1985) Communication: Photoaffinity ma&l. 113: 145-146. labeling of the D, dopamine receptor using a novel high-affinity ra- Lowrv. 0. H.. N. J. Roscbroueh. A. L. Farr. and R. J. Randall (195 I) dioiodinated probe. J. Biol. Chcm. 260: 1983-1986. Protein measurement with tbe’folin reagent. J. Biol. Chem. IYj; 2651 Andersen, P. H. (1987) Biochemical and pharmacological character- 275. ization of [‘H]GBR- I2935 binding in vitro to rat striatal membranes: Munson. P. J., and D. Rodbard (I 980) LIGAND: A versatile approach Labeling of the dopamine uotake comolex. J. Neurochem. 48: l887- for characterization of ligand-binding systems. Anal. B&hem. 107: 1896. - 220-239. Andersen, P. H., J. Aas Janscn, and E. B. Nielsen (1987) [)H]GBR- Nirnik, H. B., D. Grigoriadis, and P. Seeman (1986) Photoaffinity 12935 binding in vivo in mouse brain: Labelling of a piperazine labelingofdopamine D, receptors by [3H]-azidomethylspipcrone. FEBS acceptor site. Eur. J. Pharmacol. 144: l-6. Lett. 201: 7 l-76. I&mm&sour, S., J.-J. Bonnet, P. Protais, and J. Costentin (1987) So- Radian, R., A. Bcndahan, and B. 1. Kanner (1986) Purification and dium indenendence of the bindina of I’HIGBR- 12783 and other do- identification of the functional sodium- and chloride-coupled a-ami- pamine uptake inhibitors to the dipamincuptake complex. Neurosci. nobutyric acid transport glycoproteins from rat brain. J. Biol. Chem. Lett. 77: 97-102. 261: 15437-15441. Bergcr, P., A. Janowsky, F. Vocci, P. Skolnik, M. M. Schwcri, and S. Raiteri, M., F. Cerrito, A. M. Cervoni, R. del Carmine, M. T. Ribera, M. Paul (I 985) ]‘H]GBR-I 2935: A specific high affinity ligand for and G. Levi (1978) Studies on dopaminc uptake and release in labeling the dopamine transport complex. Eur.-J. Pharmacol. 107: synptosomes. Adv. Biochem. Psychopharmacol. 19: 35-56. 289-290. Redouanc. K.. P. Sokoloff. J. C. Schwartz. P. Hamdi. A. Mann. C. G. Bhavanandan, V. P., and A. Katlic (1979) The interaction of wheat Wermuth, J: Roy, and J.’ L. Morgat (I 985) Photoaffinity labeling of germ agglutinin with sialoglycoproteins. The role of sialic acid. J. D-2 dopamine binding subunits from rat striatum, anterior pituitary Biol. Chem. 254: 4000-4008. and olfactory bulb with a new probe [lH]azidosulpiridc. Biochem. Bonnet. J. J., and J. Costentin (1986) GBR-12783. A potent and Biophys. Rcs. Commun. 130: 1086-1092. selective inhibitor of dopamine uptake: Biochemical studies in vivo Rehavi, M., P. Skolnick, and S. M. Paul (1982) Solubilization and and ex vivo. Eur. J. Pharmacol. 121: 199-209. partial purification of the high affinity [‘Hlimipramine binding site Bonnet, J. J., P. Protais, A. Chagraoui, and J. Costentin (I 986) High from human platelets. FEBS Lett. ISO: 5 14-5 18. affinity >H-GBR- I2783 binding to a specilic site associated with the Ritz, M. C.. R. J. Lamb, S. R. Goldberg, and M. J. Kuhar (1987) neuronal dopaminc uptake complex in the central nervous system. Cocaine receptors on dopamine transporters are related to self-admin- Eur. J. Pharmacol. 126: 21 l-222. istration of cocaine. Science 237: I2 19-l 223. Chagraoui. A., J.-J. Bonnet, P. Protais, and J. Costcntin (1987) In Sharkey. J., M. C. Ritz, and M. J. Kuhar (1987) The cocaine binding vivo binding of [‘H]GBR-12783, a selective dopamine uptake inhib- site associated with dopamine uptake inhibition as labeled by ‘H- itor, in mouse striatum. Ncurosci. Lctt. 78: 175-l 79. GBR 12935. Sot. Neurosci. Abstr. 13: 144. Goldstein, I. J., C. E. Hollerman. and E. E. Smith (1965) The lectins: Snyder, S. H. (I 970) Putative neurotransmitter in the brain: Selective Carbohydrate binding proteins of plants and animals. Biochemistry neuronal uptake. subcellular localization and interactions with cen- 4: 876-883. trally acting drugs. Biol. Psychiatry 2: 367-389. Grigoriadis, D., H. B. Niznik, and P. Seeman (1988) Glycoprotein Stiles, G. L., M. G. Caron, and R. J. Lellcowitr (1984) P-adrenergic nature of D, dopamine receptors. FEBS Lett. 227: 220-224. receptors: Biochemical mechanisms ofphysiological regulation. Phys- Habcrt, E., D. Graham, and S. Z. Langer (1986) Solubilization and iol Rev. 64: 66 l-743. characterization of the 5-hydroxytryptamine transporter complex from van der Zee, P., H. S. Koger, J. Gootjes, and W. Hespc (1980) Aryl rat cerebral cortical membranes. Eur. J. Pharmacol. 18: 197-204. I .4-dialk(en)ylpiperazines as selecbvc and very potent inhibitors of Heikkila, R. E., and L. Manzino (I 984) Behavioral properties ofGBR- donamine uptake. Eur. J. Med. Chem. 15: 363-370. 12909, GBR- I3069 and GBR-I 3098: Specilic inhibitors ofdopamine Wilson, A. A.: R. F. Dannals, H. T. Ravert, H. D. Bums, and H. N. uptake. Eur. J. Pharmacol. 103: 241-248. Wagner, Jr. (I 986) I- I25 and I- I23 labelled iodobenzyl bromide, a Hess. H. J., R. M. Graham, and C. J. Homey (1983) Photoaffinity useful alkylating agent for radiolabelling biologically important mo- label for the alpha,-adrenergic receptor: Synthesis and effect on mem- leculacs. J. Labcllcd Comp. Radiopharm. 23: 83-93.