RTI‐4793‐14, a new ligand with high affinity and selectivity for the (+)‐MK801‐insensitive [3H]1‐[1‐(2 thienyl)cyclohexyl]piperidine binding site (PCP site 2) of guinea pig brain

[3H]TCP, an analog of the dissociative anesthetic phencyclidine (PCP), binds with high affinity to two sites in guinea pig brain membranes, one that is MK‐801 sensitive and one that is not. The MK‐801‐sensitive site (PCP site 1) is associated with NMDA receptors, whereas the MK‐801‐insensitive site (PCP site 2) may be associated with biogenic amine transporters (BAT). Although several “BAT ligands” are known that bind selectively to PCP site 2 and not to PCP site 1 (such as indatraline), these compounds have low affinity for site 2 (Ki values > 1 μ). Here we demonstrate that the novel pyrrole RTI‐4793‐14 is a selective, high affinity ligand for PCP site 2. We determined the IC50 values of RTI‐4793‐14 and several reference compounds [PCP, (+)‐MK801 and indatraline] for PCP site 1 (assayed with [3H](+)‐MK801), PCP site 2 (assayed with [3H]TCP in the presence of 500 nM (+)‐MK801) and a variety of BAT‐related measures ([3H]CFT binding to the DA transporter, [3H]nisoxetine binding to the norepinephrine transporter, [3H]dopamine uptake, [3H]serotonin uptake). In addition, we determined the ability of RTI‐4793‐14 to block NMDA responses in cultured hippocampal neurons under voltage clamp. (+)‐MK801 had high affinity for PCP site 1 (4.6 nM) and potently inhibited NMDA‐induced responses, but was much less potent in the BAT‐related measures (IC50s > 10 μ). PCP had high affinity at PCP site 1 (IC50 = 92 nM) and PCP site 2 (IC50 = 117 nM), and was moderately potent in all BAT‐related measures except [3H]nisoxetine binding. Indatraline was potent in BAT‐related measures (IC50s, 2 to 5 nM), but weak in other measures (IC50s > 1 μ). In contrast, RTI‐4793‐14 had high affinity for PCP site 2 (38 nM), low affinity for PCP site 1 (> 36 μ), moderate IC50s for all BAT‐related measures, and negligible activity at NMDA receptors. Viewed collectively, these data indicate that RTI‐4793‐14 binds with high affinity and selectivity to PCP site 2 and provide further support for an association between PCP site 2 and the BATs. © 1994 Wiley‐Liss, Inc.


INTRODUCTION
[3H]1-[1-(2-thienyl)cyclohexyl]piperidine (TCP) and [3H]MK-801 have been extensively used to label the binding site associated with NMDA re-site that is insensitive to MK-801 (PCP site 2) and [3H]TCP in the presence of MK-801 can be used to label this site. Several lines of evidence suggest that this site may be associated with biogenic amine transporters (Rothman et al., 1989;Akunne et al., 1991;Akunne et al., 1992;Rothman et al., 1992). First, PCP, which inhibits the reuptake of dopamine (DA) and serotonin (5-HT) (Smith et al., 1977), has high affinity for site 2, whereas (+ I-MK801, which is a very weak DA reuptake blocker (Snell et al., 1988)) does not bind to this site. Second, as reported for the guinea pig (Rothman et al., 1989) and human brain (Akunne et al., 19911, high affinity serotonergic (fluoxetine) and dopaminergic (GBR12909 and BTCP) reuptake inhibitors bind selectively to site 2. Third, fluoxetine, GBR12909, and BTCP bind nearly irreversibly to PCP site 2 consistent with a high affinity interaction (Akunne et al., 1991). Fourth, MPTP-induced lesions of the dopaminergic innervation to the caudate nucleus decrease [3H]TCP binding in this region (Akunne et al., 1992). The latter observation suggests a partial presynaptic localization for at least some of these sites. Based upon the association of PCP site 2 with BATs, drugs with high affinity for PCP site 2 might be expected to inhibit the reuptake of biogenic amines. In the course of studying the interaction of a series of pyrroles with PCP binding sites, we observed that (2RS,3aSR,8bRS)-1,2,3,3a,4,8b-hexahydro-2-benzyl-lmethylindeno-[1,2-blpyrrole resorcylate  was a potent and selective ligand for PCP site 2 (Carroll et al., in press). The availability of this compound has enabled us to obtain further evidence supporting the concept that PCP site 2 is associated with BATs.

MATERIALS AND METHODS
Preparation of membranes Since PCP site 1 and PCP site 2 are readily measured in guinea pig brain membranes (Rothman et al., 1989(Rothman et al., , 1992, the present study was conducted using this preparation. Large batches of frozen membranes were prepared with minor modifications of published procedures (Rothman et al., 1989). Frozen guinea pig brains with cerebellum (20-30) were thawed for 15 min and homogenized with a Polytron in ice-cold 5 mM Tris- HCl, pH 8.0 (10 mhrain). The homogenate was centrifuged at 37,000 x g for 10 min, and the pellet was washed by resuspension in the same volume of buffer followed by recentrifugation. The pellets were resuspended in an equal volume of 5 mM Tris-HC1, pH 8.0, and the concentration was adjusted to 50 mM by the addition of 1 M Tris-HC1, pH 8.0. The homogenate was centrifuged for 10 min at 37,000 x g. The pellets were then washed three times by resuspension and centrifugation using 50 mM Tris HC1, pH 8.0. The final pellets were resuspended in the Tris-HC1 buffer (0.5 mlibrain), pooled and 1 ml aliquots were distributed to microfuge tubes, which were stored at -80°C for assay. Membranes for the [3H]nisoxetine assay were prepared as described .
For [3H12~-carbomethoxy-3p-(4-fluorophenyl)tropane (CFT, WIN35,428) binding assays (see below), male Sprague-Dawley rats (200-300 g) (Charles River) were anesthetized with C02 gas and decapitated. Striata were dissected using glass manipulators, placed in small plastic containers, and then allowed to freeze by placing the container in dry ice. Striata collected in this way were stored at -80" C. On the day of the assay, each striatum was placed in ice-cold binding buffer (BB: 55.2 mM sodium phosphate buffer, pH 7.4, 5 ml per caudate) and homogenized while still frozen with a Polytron. The homogenate was centrifuged for 10 min at 30,000 x g, and the pellet was resuspended in an equal volume of BB. The homogenate was recentrifuged, and the pellet was resuspended in an equal volume of BB. An aliquot was saved for determination of protein (0.5 ml), and the remaining homogenate was brought up to a final volume of 50 mustriaturn using ice-cold BB. Typical final protein concentrations, determined with the method of Lowry et al., (19511, were 100 pg/ml. Initial experiments showed that at these protein concentrations, the specific binding was directly proportional to protein and that < 10% of the radioligand was bound (data not shown). Aliquots of membranes were then used in the radioligand binding assays (see below).

Binding assays
The [3H]TCP binding assay for PCP site 2 was conducted with minor modifications of published protocols (Rothman et al., 1989). Briefly, 12 x 75 mm polystyrene test tubes were prefilled with 100 pl of [3HlTCP (2 nM final concentration) in a protease inhibitor/ antioxidant cocktail (PIC) containing 5,000 nM (+)-MK801, 100 pl of distilled H,O or drug (in distilled H,O), 50 pl of buffer (5 mM Tris-HC1, pH 8.01, and 750 p1 of membrane (0.5-1.0 mg/ml protein, in 5 mM Tris-HC1, pH 8.0). The (+ I-MK801 (500 nM final concentration) was included to block I3H1TCP binding to PCP site 1. The protease inhibitorlanti-oxidant cocktail was composed of 25 Fg/ml leupeptin, 25 pg/ml chymostatin, 0.1 mM ethylenediaminetetraacetic acid (EDTA), and 0.1 mM ethyleneglycol-bis-( P-aminoethyl-ether)-N,N,N',N'-tetraacetic acid (EGTA). The incubation time was for 18-24 hours at 4" C (steady state) in a final volume of 1 ml. Triplicate samples were filtered with an MR24 Brandel Cell Harvester and washed with two 5 ml aliquots of ice-cold buffer. Whatman GF/B filters were presoaked in buffer containing 2% polyethylenimine. The tritium retained on the filters was measured using a Taurus scintillation counter at 44% efficiency after an overnight extraction into ICN Cytoscint cocktail. Nonspecific binding was determined using 10 pM of TCP. Protein was determined using the method of Lowry et al. (1951). The [3H1(+)-MK801 assay (2 nM final concentration) proceeded as described above for L3H1TCP except that (+)-MK801 was not included as a blocker, the incubation time was for 4-6 hours at 4" C, and nonspecific binding was determined with 1 pM TCP.
The [3H]nisoxetine assay for the norepinephrine transporter (NE) was carried out as described (Tejani-Butt et al., 1990;Rothman et al., 1993). DA transporters were labeled with [3H]CFT with minor modifications of published protocols (Madras et al., 1989). Briefly, 12 x 75 mm polystyrene test tubes were prefilled with 100 p1 of drug, 100 p1 of radioligand (1 nM final concentration), 50 p1 of buffer. Drugs were made up in BB containing 1 mg/ml bovine serum albumin (BB/BSA). Radioligands were made up in a protease inhibitor cocktail containing 1 mgml BSA {BB containing chymostatin (25 pg/ml), leupeptin (25 pg/ml), EDTA (100 pM) and EGTA (100 pM)). The assay was initiated by the addition of 750 p1 of membranes, prepared as described above. Brandell cell harvesters were used to filter the samples over Whatman GF/B filters, which were presoaked in wash buffer (ice-cold 10 mM TRIS-HCl, pH 7.4 containing 150 mM NaC1) containing 2% polyethylenimine. In this procedure, samples were filtered and then washed with two 4 ml aliquots of ice-cold wash buffer. Nonspecific binding was defined using 10 pM (final concentration) GBR12909. All subsequent steps were identical to those described above.
Biogenic amine uptake assays Synaptosomal uptake assays were carried out as described . Briefly, synaptosomes were prepared by homogenization of rat caudate (for [3H]DA reuptake) or whole rat brain minus cerebellum (for [3H]5-HT reuptake) in ice-cold 10% sucrose, using a Potter-Elvehjem homogenizer. After a 1,000 x g centrifugation for 10 min at 4" C, the supernatants were retained on ice. The uptake assays were initiated by the addition of 100 p1 of synaptosomes to 12 x 75 mm polystyrene test tubes, which were prefilled with 750 p1 of [3H]ligand (2 nM [3H]5-HT or 5 nM [3H]DA) in a Krebsphosphate buffer (pH 7.4), which contained ascorbic acid (1 mg/ml) and pargyline (50 pM) (buffer), 100 pl of test drugs made up in buffer containing 1 mg/ml bovine serum albumin, and 50 pl of buffer. The [3H]5-HT reuptake experiments were conducted in the presence of 100 nM nomifensine and 100 nM GBR12935 in order to block any possible reuptake into NE and DA nerve terminals. Although these agents were routinely included in the L3H15-HT reuptake assays, control studies showed that they had no discernible effect on [3H]5-HT reuptake (data not shown). The nonspecific uptake of each l3H1ligand was measured by incubations in the presence of 1 p M GBR12909 (f3H1DA) and 10 pM fluoxetine (l3H15-HT). The incubations were terminated after a 15 min (r3H1DA) or 30 min (L3H15-HT) incubation at 25" C by adding 4 ml of wash buffer (10 mM TRIS-HC1, pH 7.4 containing 0.9% NaCl at 25" C), followed by rapid filtration over Whatman GFh3 filters and one additional wash cycle. Control studies indicated that specific uptake was (1) linear with time up to 30 min; and (2) was directly proportional to protein in the protein range used here. The Krebs-phosphate buffer contained 154.5 mM NaC1, 2.9 mM KC1, 1.1 mM CaCl,, 0.83 mM MgCl,, and 5 mM glucose. The tritium retained on the filters was counted, in a Taurus beta counter, after an overnight extraction into ICN Cytoscint cocktail.
Electrophysiological studies Whole-cell recordings of NMDA-induced currents in cultured rat hippocampal neurons were carried out as described previously (Subramaniam et al., 1992). Hippocampal neurons grown in primary culture from 19day-old Sprague-Dawley rat embryos were used 7-14 days after plating. The extracellular recording solution consisted of (in mM): 140 NaC1,5 KC1,O.l CaCl,, and 10 HEPES (osmolality, 315-325 mOsm; pH, 7.4). Tetrodotoxin (1 pM) and strychnine (1 pM) were added to block Na+ channel currents and glycine-activated C1currents, respectively. Patch pipettes (2-5 MR) were prepared from filament-containing thin wall glass capillary tubes and filled with an intracellular solution that consisted of (in mM): 145 CsC1,2 MgCl,, 5 HEPES, 0.1 CaCl,, and 1 EGTA (osmolality 310 mOsm; pH, 7.4). Drugs were dissolved in recording solution and applied via a rapid gravity-fed perfusion system. Flow of the drug solutions was regulated by a microvalve operated by a programmable controller. Whole-cell recordings were performed with an Axopatch 1C patch-clamp amplifier (Axon Instruments, Burlingame, CA) and displayed on a high speed ink pen recorder (Gould Electronics, Cleveland, OH).
In vivo microdialysis experiments Male Sprague-Dawley rats (350-400 g) were anesthetized with chloral hydrate (400 mg/kg; 100 mg/ml) and supplemented as required (90 mg/kg) to abolish the corneal reflex. Animals were then placed in a stereotaxic frame and dialysis probes (CMA 12, 2 mm) were -  implanted into the nucleus accumbens (AP:+2 mm, LAT:+1.5 mm, H:-7.5 mm relative to bregma) according to the atlas of Paxinos and Watson (Paxinos, 1982). The dialysis probes were perfused with a physiological medium containing 145 mM NaCl, 4 mM KC1, 1.2 mM CaCl,, and 2 mM Na2HP04, the final pH of the medium being pH 7.4. The probes were dialyzed at a flow rate of 2.34 mumin and samples were collected every 15 mins. The dialysis probes were allowed to equilibrate for 1-2 hr prior to collection of the baseline samples. Baseline samples were then collected until three consecutive samples did not differ significantly. RTI 4793-14 (1-100 bM) was then administered focally via the dialysis probe for a 15-min sampling period at the end of which normal dialysis medium was dialysed through the probe. Sampling continued for a further three samples before application of the subsequent concentration of RTI-4793-14. In a subsequent group of animals, RTI-4793-14 was administered systemically (5 mgkg; 5 mgl ml) via a femoral vein catheter.

TABLE I . Interaction of RTI-4793-14 and reference compounds with NMDA receptor-and biogenic amine transporter-related indices
The samples were analyzed for DA using HPLC with electrochemical detection. The mobile phase [75 mM NaH2P0,, 1.5 mM sodium dodecyl sulphate, 20 mM EDTA, 15% acetonitrile, and 12% methanol, pH 5.61 was filtered and degassed before pumping at a rate of 0.75 mVmin through a HR-80 column (3 m BDS, 80 x 4.6 mm). Electrochemical detection was performed using a Coulochem I1 with Guard cell: + 350mV, Det 1: -75mV and Det 2: +220mV.
Data analysis Inhibition curves were fit to the two parameter logistic equation for the best-fit estimates of the IC50 using nonlinear least squares curve fitting methods as previously described (Jones and Rogawski, 1992;Rothman et al., 1993).
In vivo microdialysis data were analyzed by using a one-way ANOVA and posthoc Dunnetts t-test. In the focal RTI experiments in which three drug concentrations were administered the baseline value for dopamine was defined as the sample preceding the application of each drug concentration. Analysis of any drug effect was therefore made by comparing with the baseline value prior to the drug administration. PCP was obtained from the NIDA Addiction Research Center pharmacy. RTI-4793-14 was synthesized as described (Blough et al., 1993;Carroll et al., in press). Frozen guinea pig brains were purchased from Pel-Freeze Laboratories (Rogers, AR). The sources of equipment and reagants required for the in vivo microdialysis studies are as published (Rothman et al., 1991). Rats were cared for according to NIH guidelines.

RESULTS
The IC,, values of RTI-4793-14, PCP, (+)-MK801, and indatraline in the binding assays are presented in Table I. As previously observed, PCP had moderate and about equal affinity for PCP site 1 and PCP site 2 (Rothman et al., 1989). Also as observed by others (Smith et al., 19771, PCP inhibited ["IDA (1C5,, 347 nM) and L3H15-HT (IC5*, 1424 nM) uptake with moderate potency. Consistent with its reported low affinity at the DA transporter labeled with [3H]mazindol (Kuhar et al., 1990), PCP also had lower affinity at the [3H]CFT binding site than would be predicted on the basis of its IC,, for inhibition of DA uptake. PCP also had very low affinity for the NE transporter as labeled by [3Hlnisoxetine. In contrast, (+ )-MK801 had high affinity for PCP site 1 and negligible affinity for PCP site 2 and BATrelated measures. Consistent with its 25-fold greater potency at PCP site 1, (+ )-MK801 has been shown to be much more potent than PCP in inhibiting NMDA-induced current responses (ffrench-Mullen and Rogawski, 1989;Jones and Rogawski, 1992).

DISCUSSION
The results presented in this study demonstrate that RTI-4793-14 binds with high affinity and selectivity to the (+)-MK801-insensitive r3H]TCP binding site (PCP site 2). Although most BAT ligands such as indatraline are selective for PCP site 2, they generally have low  Table I; the steepness parameter n was 0.83 and 0.96 for RTI-4793-14 and indatraline, respectively.  (Rothman et al., 1992). , which is at least 1,000-fold selective for PCP site 2 vs. PCP site 1, had the highest affinity for PCP site 2 of any of the drugs we tested and is the first ligand to be described that clearly distinguishes between the two PCP binding sites. Like PCP, which has moderate po- tency in BAT-related measures, RTI-4793-14 inhibited [3H]CFT binding and L3H1 5-HT and [3H1DA uptake with moderate potency, but unlike PCP was a very low potency inhibitor of NMDA-induced current responses. Thus, RTI-4793-14 has a pharmacological profile similar to that of PCP, except that it is largely devoid of PCP site 1 activity and has considerably higher affinity at the NE transporter.
In accord with previous reports (Arnt et al., 1985;Hyttel and Larsen, 1985); indatraline was a potent inhibitor of [3H]CFT binding, [3HJnisoxetine binding and 13H]5-HT and L3H1DA uptake (Rothman et al., 1992). However, as is typical of other previously investigated BAT ligands, the drug bound relatively weakly to PCP site 2. Although they both interact with biogenic amine transporters, RTI-4793-14 was distinguished from indatraline in having substantially higher binding affinity for PCP site 2.
DA uptake blockers, including PCP (Carboni et al., 1989) and indatraline (Hurd and Ungerstedt, 1989b), are known t o increase extracellular DA levels. Consistent with its in vitro profile as a BAT inhibitor, RTI-4793-14 increased extracellular DA upon local or systemic administration in rats. The magnitude of the effect obtained with a 5 mgkg IV dose was similar to that previously observed with a 1 mgkg IV dose of cocaine, a potent DA uptake blocker (Hurd and Ungerstedt, 1989a). Thus, RTI-4793-14 acts in vivo as well as in vitro as a biogenic amine transport blocker.
The data presented here suggest that DA uptake inhibitors may fall into two classes: (1) classical BAT ligands, which exhibit high affinity Na+-dependent binding, potent inhibition of 13H]DA uptake, and weak but selective binding to PCP site 2, and ( 2 ) PCP site 2 ligands, which have Na+-inhibited high affinity for PCP site 2, moderate potency for inhibiting [3H]DA up-take, and moderate potency at the classical BAT binding site. It is of interest to note that the phencyclidine derivative BTCP, which potently inhibits DA uptake in vitro (Vignon et al., 1988) and in vivo (Maurice et al., 1992), falls into the category of classical BAT ligands. [3H]BTCP exhibits high affinity Naf -dependent binding to the DA transporter both in vitro (Vignon et al., 1988;Cerruti et al., 1991) and in vivo (Maurice et al., 1989). Like indatraline, it has low affinity and high selectivity for PCP site 2 (Rothman et al., 1989). Consistent with this neurochemical profile, BTCP has eocaine-like, not PCP-like, behavioral effects (Koek et al., 1989). Thus, although BTCP is chemically derived from PCP, it has the characteristics of a classical BAT ligand rather than a PCP site 2 ligand.
The results of the present study demonstrate that RTI-4793-14 is a potent and selective ligand for PCP site 2 and that it also has activity as a biogenic amine transport blocker. Unlike classical BAT ligands such as indatraline, but similar to PCP, RTI-4793-14 has only moderate activity as an uptake blocker. Nevertheless, RTI-4793-14 and PCP bind with high affinity to PCP site 2 and, as noted in the Introduction, there is substantial evidence linking this site with biogenic amine transporters. Thus, RTI-4793-14 may bind to a site on biogenic amine transporters that is distinct from the binding site for classical BAT ligands. If this is the case, RTI-4793-14 could represent the first of a novel class of ligands that bind with high affinity to biogenic amine transporters but are relatively less potent amine reuptake blockers than classical BAT ligands.
The high affinity and selectivity of RTI-4793-14 for PCP site 2 suggest that it or an analog may be useful in radioligand binding studies to label this site with greater precision than is currently possible. It should be possible to test the hypothesis that PCP site 2 is associated with biogenic amine transporters using the recently cloned DA and 5-HT transporters (Blakely et al., 1991;Kilty et al., 1991;Shimada et al., 1991). If the association is confirmed, RTI-4793-14 may represent a useful new tool for investigating biogenic amine transporters.
ACKNOWLEDGMENT This work was supported in part by Grant DA06302 from NIDA.