US 20050033522A1 (19) United States (12) Patent Application Publication (10) Pub. No.: US 2005/0033522 A1 Wainer et al. (43) Pub. Date: Feb. 10, 2005

(54) COMPUTER-BASED MODEL FOR Related U.S. Application Data IDENTIFICATION AND CHARACTERIZATION OF (63) Continuation-in-part of application No. 10/411,206, NON-COMPETITIVE INHIBITORS OF filed on Apr. 11, 2003. NCOTINIC ACETYLCHOLINE RECEPTORS AND RELATED LIGAND-GATED ON Publication Classification CHANNEL RECEPTORS (51) Int. Cl." ...... G06F 19/00; G01N 33/48; (76) Inventors: Irving W. Wainer, Washington, DC GO1N 33/50 (US); Krzysztof Jozwiak, Abingdon, (52) U.S. Cl...... 702/19 MD (US); Ruin Moaddel, Germantown, MD (US); Sarangan (57) ABSTRACT Ravichandran, Frederick, MD (US); A computer readable medium holding data of a molecular Jack R. Collins, Frederick, MD (US) model of a ligand-gated ion channel receptor and/or a computer System for modeling Said receptor are provided by Correspondence Address: the instant invention. The molecular model can be used to BRCH STEWART KOLASCH & BRCH design novel compounds having activity as non-competitive PO BOX 747 inhibitors of the ion channel. A preferred embodiment of the FALLS CHURCH, VA 22040-0747 (US) invention relates to nicotinic acetylcholine receptorS. Com pounds having activity as non-competitive inhibitors of (21) Appl. No.: 10/820,809 ligand-gated ion channel receptors and methods for inhib iting the receptor and treating diseases or disorderS mediated (22) Filed: Apr. 9, 2004 by function of the receptor are also disclosed. Patent Application Publication Feb. 10, 2005 Sheet 1 of 16 US 2005/0033522 A1

ONE SUBUNT OF nAChR

- N NH2 UGAND-BINDING STE

Soffee CROSS-SECTION OF FME ASSEMBLED SUBUNITS (2xa + 3x p) OF nAChR i: T. IONSGolf PASS CYTOPLASMIC Epo FOUR HYDROPHOBC S 5 4. TRANSMEMBRANE . . BINDING STE FOR DOMAINS (M1-M4) NEUROTRANSMITTER 5

B SUBUNIT O SUBUNIT 7 ACELULAR u?

CYTOPLASMIC

CHANGE IN INTRACEULAR ON CONCENTRATION

FIG.1 Patent Application Publication Feb. 10, 2005 Sheet 2 of 16 US 2005/0033522 A1

(Z–)ONIJONSYldOIX0 ?J?GYIMISV0W|}}}}}/W» (91)ONIHENOTEIl (II)ONIHENOBI (8)ONIHENIMES (*)ONIHEN|NOBAHI „(ZZ)ONIHMINTEOVNIXG|—À (GI)ONIHENTWA|<--> (1)ONIHTIMEWARINI\<--->

Patent Application Publication Feb. 10, 2005 Sheet 3 of 16 US 2005/0033522 A1

FIG.3a

Patent Application Publication Feb. 10, 2005 Sheet 4 of 16 US 2005/0033522 A1

Patent Application Publication Feb. 10, 2005 Sheet 5 of 16 US 2005/0033522 A1

3 3

s S Patent Application Publication Feb. 10, 2005 Sheet 6 of 16 US 2005/0033522 A1

Patent Application Publication Feb. 10, 2005 Sheet 7 of 16 US 2005/0033522 A1

Patent Application Publication Feb. 10, 2005 Sheet 9 of 16 US 2005/0033522 A1

Patent Application Publication Feb. 10, 2005 Sheet 10 of 16 US 2005/0033522 A1

Patent Application Publication Feb. 10, 2005 Sheet 11 of 16 US 2005/0033522 A1

Patent Application Publication Feb. 10, 2005 Sheet 12 of 16 US 2005/0033522 A1

- DEXTROMETHORPHAN - LEVOMETHORPHAN

O 20 40 60 80 100 RETENTION TIME (min) FIG.11 Patent Application Publication Feb. 10, 2005 Sheet 13 of 16 US 2005/0033522 A1

FIG.12b Patent Application Publication Feb. 10, 2005 Sheet 14 of 16 US 2005/0033522 A1

2.4 - S / 2.2 | CLOZAPINE 20 | 1.8 / y : LOGK VERAPAML VERAPAML 1.6 | Q Q EuroPh. 8 DLTHAZEN": S 1.4 p LEVOETHORPHAN METHADONE 9 ETRORPHAN 1.2 w \::i- " :: us;5People.1,953OMF:

O 5-1 QMK-801 N. -76 A MENANTINE \ -8.0 ECAMANNEó oBUPROPON -8,4- Eurt. EHONO -88 N. f gouro Patent Application Publication Feb. 10, 2005 Sheet 15 of 16 US 2005/0033522 A1

FIG.14 Patent Application Publication Feb. 10, 2005 Sheet 16 of 16 US 2005/0033522 A1

250000

200000 150000

100000

50000

LOGNC FIG.15 US 2005/0033522 A1 Feb. 10, 2005

COMPUTER-BASED MODEL FOR Torpedo marmorata has been examined by Unwin and IDENTIFICATION AND CHARACTERIZATION OF coworkers using cryo-electron microscopy and revealed the NON-COMPETITIVE INHIBITORS OF NCOTINC conical shape of the channel portion of the receptor and the ACETYLCHOLINE RECEPTORS AND RELATED relationship of the membrane-Spanning helices to each other LGAND-GATED ON CHANNEL RECEPTORS (see FIG. 1). In spite of these unprecedented advances in 0001. The present application is a Continuation-In-Part resolving the Structures of transmembrane proteins, the of U.S. application Ser. No. 10/411,206, filed Apr. 11, 2003, detailed, atomic resolution, Structure of the entire nAChR the entire contents of which are hereby incorporated by family remains unresolved. reference and for which priority is claimed under 35 U.S.C. 0007 Muscular nAChRs are located at the nerve-muscle S 120. junctions and are responsible for triggering motor motion, 0002 The present application includes an appended and neuronal nAChRs, widely distributed in the nervous Sequence Listing of 15 amino acid Sequences and Appen System, are involved in the fast Synaptic transmission of dices 1 to 5 providing computer programming Scripts, inter-neuronal communication. It is known that these recep parameter files and atomic coordinates of computer models tors are structurally similar in their overall composition but of the luminal channel portion of the ligand-gated ion differ in the exact make-up of the protein Subunits forming channel Subtypes. the receptorS. 0008 The nicotinic acetylcholine receptor (nAChR) is FIELD OF THE INVENTION presently the best characterized member of the ligand-gated ion channel Superfamily. The nicotinic receptors are of great 0003. The present invention relates to a computer system therapeutic importance. The Subunits assemble combinato for generating molecular models of ligand-gated ion chan rily to form a variety of pentameric transmembrane protein nels and in particular, molecular models of the inner lumen Subtypes. of a ligand-gated ion channel and associated binding pock ets. The present invention further relates to a computer 0009. Each receptor is formed by bringing together five System Simulating interaction of the computer-based model Separate trans-membrane proteins, each containing a large of the ligand-gated channel and non-competitive inhibitor extra-cellular N-terminal domain, four membrane Spanning compounds for identification and characterization of non alpha helices (M1, M2, M3, and M4) and a small C-terminal competitive inhibitors and to inhibitor compounds So dis domain (see FIG. 1). Two, homologous, neurotransmitter covered. The present invention also relates to methods for binding sites are formed by the N-terminal domains where treating various disorders related to ligand-gated ion channel cholinergic agonist and competitive antagonists bind, and receptor function. The invention also provides a way to are the usual targets for drug design. The ion channel is examine compounds for “off-target activity that may cause formed by a pentameric arrangement of the M2 helical undesirable side effects to a desired target activity or that segments contributed by the five proteins (see FIG. 2). The may represent a new therapeutic activity for a known channel Specificity, characteristic of each receptor Subtype, compound. is controlled by the identity of each of the M2 helices. 0010 Neuronal nicotinic acetylcholine receptors BACKGROUND OF THE INVENTION (nAChRs) are the class of ligand-gated ion channels of the 0004 Ligand gated ion channels (LGICs) are currently central and peripheral nervous System that regulate Synaptic very important targets for drug discovery in the pharmaceu activity. The basic structure of the nAChR is shown in tical industry. The Superfamily is separated into the nicotinic FIGS. 1 and 2. Referring to FIG. 1, nAChR consists of five receptor Superfamily (muscular and neuronal nicotinic, transmembrane Subunits 1, 2, 3, 4, 5 oriented around a GABA-A and -C, glycine and 5-HT3 receptors), the exci central pore 6 permeable to cations. Cations flow through the tatory amino acid Superfamily (glutamate, aspartate and pore is regulated by ligand binding. The Subunits in nAChR kainate receptors) and the ATP purinergic ligand gated ion are typically C. Subunits and B Subunits. channels. These families only differ in the number of trans 0011. At present, 12 different homologous subunits have membrane domains found in each Subunit (nicotinic-4 been identified in neuronal nAChRs, 9 C. Subunits (C2-C-10) transmembrane domains, excitatory amino acid recep and 3 B subunits (B2-34). The major difference between C. tors-3 transmembrane domains, ATP purinergic LGICs-2 and B Subunits is the presence and location of the disulfide transmembrane domains). bond formed by two adjacent cysteines in the C. Systems, the 0005 Nicotinic acetylcholine receptors (nAChRs) are a absence of this feature distinguishes non-C. Subunits. This family of ligand gated ion channels that control the fast disulfide bond located on the extracellular domain plays an permeation of cations through the postsynaptic cell mem important role in neurotransmitter binding as well as the brane when stimulated by acetylcholine. Physiologically, mechanism of channel opening. These Subunits combine to nAChRS are key targets in drug discovery for a number of form multiple nAChR Subtypes and predominant Stoichiom diseases, including Alzheimer's and Parkinson's disease, etry is (C)(B), however pentamers containing only C. and have been widely discussed and investigated. Subunit are also known e.g., (C7). In case of muscular nAChR the Stoichiometry is more complicated, the muscular 0006 Structural and functional studies of nAChRs have led to the elucidation of three physiological States of the nAChR receptor is predominantly described as (C)(3öy. receptor: 1) resting (channel closed); 2) acetylcholine stimu 0012. The nAChRs are very complex systems with doz lated (channel open); and 3) a desensitized State where the ens of potential different binding domains for different ion flux is inhibited even in the presence of neurotransmitter. classes of compounds of both endo- and exogenous origin The overall Structure of nicotinic acetylcholine receptor of (Arias H. R., (1997) Topology of ligand binding sites on the US 2005/0033522 A1 Feb. 10, 2005 nicotinic acetylcholine receptor. Brain Res. Rev. 25: 133 hand, designing ligands that Specifically interact with this 91). Two primary cholinergic binding sites are located on the Site can be part of the development of new treatments of extracellular side 7 (refer to FIG. 1, approximately 30-35 A Alzheimer's and Parkinson's diseases, for example by iden above the membrane) in the pocket at the interface between tifying compounds likely to exhibit Side effects through the C. and B subunits. The nAChR contains several other non-competitive inhibition of a LGIC. Furthermore, the classes of binding Sites at which non-competitive inhibitors compounds identified as NCIS by the present method are (NCIs) bind (Arias H. R. (1998) Binding sites for exogenous likely to find use in treating Tourette's Syndrome and cog and endogenous non-competitive inhibitors of the nicotinic nitive disorders, Schizophrenia, pain See, Lloyd, G. K. and acetylcholine receptor. Biochim. Biophys. Act. 1376: 173 Williams, M. (2000) J. Pharmacol. Exper. Ther. 292, 461 220). One, so-called “luminal high affinity” NCI binding 467., anxiety, depression, neurodegeneration and addictions domain is located on the Surface of the internal lumen caused by an overactive LGIC receptor, especially diseases forming the ion channel. This site is a highly polar and in which nicotine agonist activity against a neuronal nAChR negatively charged domain, which primarily plays the role is part of the etiology (e.g. Smoking addiction). The inven as a cation Selector. In general, an NCI compound does not tion can also be used to evaluate cardiovascular toxicity of compete with the neurotransmitter ligand of the receptor for a compound mediated by non-competitive inhibition of a binding to the neurotransmitter ligand binding site of the LGIC receptor, e.g. arrythmia and GI spasming or diarrheal receptor located on the external Surface both C. Subunits in Side effects of a compound caused by inhibition of a musclar a pocket approximately 30-35 A from the transmembrane nAChR. portion of the Subunit (that is, above the Surface membrane when the receptor is expressed on in a cell), as described by 0018 Classical methods of NCI identification are time Arias Arias, H. R. (2000) Localization of agonist and consuming and not effective in rapid Screening of chemical competitive antagonist binding sites on nicotinic acetylcho libraries of drug candidates. line receptors Neurochem. Int 36,595-645). 0019. Several different molecular models of the nAChR 0013 Such drugs as mecamylamine, ketamine, bupro transmembrane domain have been reported (Capener C E, pion or barbiturates bind in the narrowest region of the Kim H J, Arinaminpathy Y, Sansom M S (2002) Ion chan channel on the cell membrane level. Inhibitors acting there nels: structural bioinformatics and modelling. Hum Mol are mainly amines. It is believed that the ligands bind into Genet 11:2425-33). However, none of those models were this region and Sterically plug the channel, blocking the flux used to investigate interaction with channel blockers. A of ions. computer based model for in silico simulations of NCI interactions with the luminal domain of LGICs is needed to 0.014 “Non-luminal” sites are the population of 10-30 better understand the phenomenon of the receptor's inhibi binding Sites located mostly at the lipid-protein interface for tion by NCIs. which an allosteric mechanism of non-competitive inhibi 0020) Furthermore, in drug discovery, the potential tion was proposed. Agents of different origin (Steroids, fatty adverse effects of drug candidates are of great importance. acids, alcohols, local anesthetics etc.) can bind to those sites In-depth understanding of mechanistic interaction of lumi and modulate nAChR activity. nal NCIS with different Subtypes of LGICs, especially of 0.015. Other classes of ligand-gated ion channels include nAChRS, is required to remove potential unwanted Side GABA (Johnston G. A. (2002) Medicinal chemistry and effects at this site. In this respect, a rapid Screening tech molecular pharmacology of GABA(C) receptors. Curr Top nology that would identify NCIS of LGICs, and especially of Med Chem 2,903-13), 5HT3 (D. C. Reeves, S. C. Lummis, nAChRs would be greatly desired. (2002) The molecular basis of the structure and function of the 5-HT3 receptor: a model ligand-gated ion channel 0021. The functional determination and characterization (review). Mol. Membr. Biol. 19, 11-26), AMPA (T. B. of a NCI of a LGIC is very complex and time consuming. Stensbol, U Madsen, P. Krogsgaard-Larsen, (2002) The One approach is affinity chromatography based on immo AMPA receptor binding site: focus on agonists and com bilized receptor protein. This is a versatile tool for investi petitive antagonists. Curr. Pharm. Des. 8, 857-72) and gation of intermolecular interactions of a receptor with its NMDA (K. A. Macritchie, A. H. Young, (2001) Emerging ligands. The chemometric approach of affinity chromatog targets for the treatment of depressive disorder. Expert Opin. raphy can be employed for determination of reliable relative Ther. Targets 5, 601-612) receptors, etc. Although the affinities of ligands as well as kinetic characterization, which molecular structure of these receptors differ Significantly, it otherwise would be inaccessible, for a large Set of com is believed that the luminal domains are homologous to the pounds (Kaliszan R., Wainer I. W. (1997) Combination of luminal domain of nAChRs. There are five (or occasionally Biochromatography and Chemometrics: A Potential New four) transmembrane helices forming the wall of the channel Research Strategy in Molecular Pharmacology and Drug with "rings” of polar amino-acids exposed on the pre Design. In Chromatographic Separations Based on Molecu forming Surface and the same non-competitive inhibition lar Recognition. K. Jinno, editors Wiley-VCH). phenomenon can be observed. 0022 Methods using nAChR and other receptors immo bilized on a chromatographic Support have been elaborated 0016. In summary, the luminal high affinity NCI binding (U.S. Pat. Nos. 6,387,268, 6,139,735, provisional applica domain is located on the Surface of the internal lumen tion No. 60/337,172). It was shown that the obtained sta forming the ion channel. Drugs of different origin bind in tionary phases worked as Selective binding materials for this region and Sterically plug the channel blocking the flux competitive cholinergic ligands and can be used for high of ions. throughput Screening of various competitive agonists and 0017 Non-competitive inhibition of the nAChR can be antagonists (R. Moaddel, I. W. Wainer, (2003) Immobilized responsible for severe adverse drug effects. On the other nicotinic receptor Stationary phases: going with the flow in US 2005/0033522 A1 Feb. 10, 2005 high-throughput Screening and pharmacological Studies J. ture for ethidium: the only compound permanently ionized Pharm Biomed Anal. 30, 1715-24). The usefulness of such and that binds to its Specific Site. columns based on immobilized nAChR for investigations and modeling of NCI affinity has also been demonstrated. 0030 FIG. 8 shows the mecamylamine binding to the Using a novel non-linear chromatography approach off and luminal domain of C.3f4. on kinetics of ligand interaction with the receptor can be 0031 FIG. 9 shows the MK-801 binding to the luminal determined. (K. Jozwiak, J. Haginaka et al., (2002) Dis domain of C3 B4. placement and nonlinear chromatographic techniques in the investigation of interaction of noncompetitive inhibitors 0032 FIG. 10 shows a correlation of log k' (chromato with an immobilized C3f4 nicotinic acetylcholine receptor graphic) with log (1/k) (docking simulation). liquid chromatographic Stationary phase. Anal Chem 74: 0033 FIG. 11 shows the enantioselectivity of the dex 4618-4624). tromethorphan/levomethorphan pair determined in chro matographic experiments. Dextromethorphan had a longer BRIEF DESCRIPTION OF THE DRAWINGS retention time and the profile was more asymmetric. 0034 FIG. 12 shows an overlay of the most stable 0023 The features of the invention may be better under docked orientations of dextromethorphan (grey) and stood by reference to the drawings described below. levomethorphan (magenta) complexes with the C3f4 0024 FIG. 1 and FIG. 2 schematically show the general nAChR luminal domain (FIG. 12a) and the C3,32-naChR Structure of a neuronal nicotinic acetylcholine receptor luminal domain (FIG. 12b). Only two helices (one a3 and (nAChR). one f) are presented for clarity. The Serine residues (position 0025 FIG. 3 shows the luminal domain of C3 B2 and 8, and also in position 12 in FIG. 12a) are rendered as CPK O3f4 channels. Red-negatively charged (e.g., aspartic models and colored in orange. The cleft formed by the valine acid, glutamic acid), orange-polar (e.g., Serine or threo to phenylalanine Substitution in the B4 Subtype helix is nine), green-hydrophobic (e.g., leucine, Valine), light indicated by the arrow in FIG. 12a and its absence is blue-positively charged (e.g., lysine), dark blue-aromatic similarly indicated in FIG. 12b. (e.g., phenylalanine). FIG. 3a shows a model of the C3,32 0035 FIG. 13 illustrates cluster analysis according to the luminal domain having five helices forming the wall of the present invention. Three clusters were identified among ion channel. FIG. 3b shows the O.3f2 luminal domain model tested NCIS. Cluster 1-red, cluster, 2-magenta, cluster (in ribbon and CPK rendering) in perpendicular view. Only 3-green. 3 helices are shown for clarity. FIG.3c shows a model of the Cl3(B4 luminal domain having five helices forming the wall 0036 FIG. 14 shows the synthetic scheme for novel of the ion channel. FIG. 3d shows the C3f4 luminal domain compound DM-01; i-1-chloroethyl chloroformate; model (in ribbon and CPK rendering) in perpendicular view. i-methanol; iii-1-chloroacetone. Again, only 3 helices are shown for clarity. The cleft formed 0037 FIG. 15 shows the NCI activity in the Rb" efflux by the substitution of phenylalanine for valine at position 15 assay of compound DM-01. in the helix is noted by the arrow. DETAILED DESCRIPTION OF THE 0.026 FIG. 4 is a schematic representation of a computer INVENTION System useful in the practice of the invention. 0038. The present invention results from understanding 0027 FIG. 5 is a model of luminal domain of C3 B4 of the interactions between a particular Subtype of the Subtype of nAChR illustrating its electroStatic potential of neuronal nAChRs and molecules that inhibit the flow of ions the inner Surface of the channel. The Figure particularly through the cell membrane. A first Step in this understanding shows the electronegative potential of the cation Selector is to characterize the composition of the membrane-Span region of the channel. Negative potentials are shown in red, ning M2 helices. So far, twelve distinct M2 helices (known and the positive potentials are shown in blue. as Subunits of the channel), nine labeled alpha (O2-C.10) and 0028 FIG. 6 shows a two cluster interaction of the ligand three labeled beta (B2-?34), have been shown to form chan PCP with C3 B4. Generally NCIs bind into the small pocket nels of a wide variety of both homomeric and heteromeric formed on the apolar domain (Phenylalanine/Valine rings). subtypes of neuronal nAChRs. The most common subunit Tested Structures primarily entered a hydrophobic pocket stoichiometry has been determined to be (CX)(BY), (X=2- formed between the C3 and B4 helices and Subsequently 4, Y=2-4), respectively for heteromeric Subtypes and (CZ), interacted with protein Side chains forming hydrogen bonds. (Z=7-10) for the homomeric subtypes. However, other, more Ligands most likely form two Separate clusters on two complex, combinations have also been reported. These Symmetrical active Sites. Estimated free energies of docking various subtypes have been found in different locations of are in the range of experimental ICso of tested inhibitors. the central and peripheral nervous System and can be assigned to different functions. For instance: the C.432 and 0029 FIG. 7 shows example compounds tested by chro a4B4 Subtypes appear to play a role in cognition, neurode matography on an OB nAChR affinity column. Among the generation, pain, anxiety and depression; the a3f32 Subtype tested drugs are aliphatic amines like mecamylamine, aman in dopamine release and Parkinson's disease; the C7 in tadine, memantine and Such compound like bupropion, GABA release; the C.9 in auditory function and develop ketamine and mk-801. Also, Some examples of more com ment; and the C3f4 in norepinephrine release, cardiovascu plicated Structures include clozapine, pcp, methadone and lar and gastrointestinal action. In addition, NCIS of ligand Verapamil. Further, the Structures of two enantiomers dex gated ion channels are expected to have therapeutic benefit tromethorphan and levomethorphan. Finally, there is a struc in treatment of cognitive dysfunction/attentional disorders US 2005/0033522 A1 Feb. 10, 2005

Such as ADHD, neurodegenerative diseases Such as Alzhe the traditional cholinergic mechanism of receptor regulation, imer's disease, Schizophrenia, depression, epilepsy, and ligands inhibiting the receptor in this way are called Tourette's Syndrome and in Smoking cessation. non-competitive inhibitors (NCIs) or channel blockers. Non competitive action on the neuronal nAChR has been 0039) Non-competitive inhibition of nAChRs may be assigned to a large number of marketed drugs and their responsible for many of the adverse effects attributed to drug metabolites and can be responsible for many toxic side therapy. For example, the impairment of cardiovascular effects of various therapies. For example clinical side effects function observed during ketamine anesthesia has been observed during ketamine anesthesia (i.e., the impairment of asSociated with the inhibitory action of ketamine on gangli the cardiovascular function, etc.) have been associated with onic nAChRs. The administration of Such drugs as metha the inhibitory action of ketamine on ganglionic nAChRS done (opioid antagonist) mecamylamine or Verapamil (anti (Friederich, P.; Dybek, A.; Urban, B. W. Stereospecific hypertensive agents) often results in gut motility impairment interaction of ketamine with nicotinic acetylcholine recep and constipation, which has been associated with their NCI tors in human sympathetic ganglion-like SH-SY5Y cells. activity on the C3f4 nAChR. Anesth. 2000, 93,818-24). Thus, there is a need to develop 0040. The pentameric bundle of M2 helices forms the models to identify ligands that might be NCIS of nAChRs. “lumen” (FIGS. 1-3), the central surface of the narrowest 0042. The present invention lies in part in a computer part of the channel, which takes part in channel gating and System that generates molecular models of ligand-gated ion Selection. The amino acid residues forming the Surface neuronal receptors and a method of using the same. The of the lumen are quite conserved acroSS different Subunits, computer System generates a computer-based model of the and form distinct regions of the channel, or “rings” (See inner lumen of a ligand-gated ion channel having binding FIG. 2 and Table 2). These rings are important for proper pockets for non-competitive inhibitors. The computer Sys function and selectivity of the neuronal nAChRs and are tem simulates interaction of Structures from chemical librar common to all Subtypes. An illustration of this importance is ies or of any desired compound with the generated com the fact that even a Single point mutation in this domain can puter-based model of the ligand-gated ion channel. The lead to a variety of Serious diseases e.g., autosomal dominant Simulation can Serve to predict and describe the pharmaco nocturnal frontal lobe epilepsy, associated with a Serine logical importance of the interaction. Thus, the invention (S)->phenylalanine (F) mutation in the M2 segment of the constitutes a System for drug discovery and for Screening of C.4 subunit of nAChR (Steinlein, O. K. Nicotinic acetylcho a drug candidate for unexpected Side effects and toxicities. line receptors and epilepsy. Curr. Drug Target CNS Neurol. 0043. In an embodiment of the present invention, as Disord. 2002, 1, 443-448). Therefore, the sequence and shown in FIG. 4, the computer System comprises a memory, structure of the M2 subunits forming the luminal domain are e.g. disk 105, Storing positional data of the atomic coordi important for understanding disease States associated with nates of the transmembrane portion of at least one Subunit of nAChRs. a ligand-gated neurotransmitter receptor protein, and a pro 0041. The luminal domain of the ion channel has been ceSSor 101 generating a molecular model having a three identified as a high affinity binding Site for a large number dimensional shape representative of the pore portion of the of exogenous and endogenous Substances in both the open ligand-gated neurotransmitter receptor based on positional and desensitized State. Many drugs, particularly ionizable data. During execution of the process for generating the amines, can elicit deleterious Side effects by binding to the molecular model, it is understood that the positional data Surface of the lumen, Sterically plugging the channel and would be stored in, for example, RAM 102, or other memory blocking the flux of ions. This mechanism is distinct from readily accessible by the processor 101.

TABLE 1.

Residue. No 1 2 3 4 5' 6" 8' 9' 10 11 12 13 14 15 16 17 18, 19 20 21

delta E K M S T A S W L L A G A V F L L L T S G R SEO ID NO: 1 gamma? O K C T L S S W L L A Q T I F L F L I A O K SEO ID NO: 2 alpha 1 E K M T L S. S W L L S L T V F L L V V E L. SEO ID NO:3 alpha 33) E K W T L C S W L L S L T V F L L V I T E T SEO ID NO: 4 alpha 4 E K T L C S W L L S L T V F L L L I T E I SEO ID NO: 5 alpha 58) E K I C L C S V L V S L T V F L L V E E SEO ID NO: 6 alpha 6) E K W T L C SEO ID NO: 7 alpha 7° E K I S. L. G. SEO ID NO: 8 alpha 98) E K V S L G L SEO ID NO: 9 alpha 107) E K V S L G SEO ID NO: 10 beta 1 E K M G L S. SEO ID NO: 11 US 2005/0033522 A1 Feb. 10, 2005

TABLE 1-continued

Residue. No 1 2 3 4 5' 6" 7", 8 9' 10 11 12 13 14 15 16 17 18, 19 20 21 22 23 beta 29 E K M T L C I S W L L A L T V F L L L I S K I SEO ID NO: 12 beta 3 E K L S L S T S W L V S L T V F L L V E E I SEO ID NO: 13 beta 4) E K M T L C I S W L L A L T F F L L L I S K I SEO ID NO: 14 epsilon O K C T V S I N V L L A O T V F L F F L I A Q SEO ID NO: 15 1)S. J. Opella, F. M. Marassi, et al., Nat. Struc. Biol., 1999, 6,374-379. °Hucho, F: Tsetlin, V. I. Machold, J. Eur. J. Biochem. 1996, 239, 539-55. J. C. Webster, M. M. Francis, et al., Brit. J. Pharmacol., 1999, 127, 1337–1348. M. W. Francis, R. W. Pazquez, et al., Mol. Pharmacol. 2000, 58, 109-119. O. K. Steinlein, A Magnusson, et al., Hum. Mol. Genet., 1997, 6,943-947. E Bertacini, JR Trudell, Protein Eng. 2002, 15, 443-453. 7)AB Elgoyhen, DE Vetter et al., Proc Natl Acad Sci USA 2001, 98, 3501-6. ENTREZ protein databank at the US National Library of Medicine 0044) The memory, in particular, stores data of the atomic tural and energetic properties of molecules. Comp. PhyS. coordinates of at least an O. chain and a B chain of a nicotinic Comm. 91, 1-41). The AMBER package contains a set of acetylcholine receptor. The data of the atomic coordinates molecular mechanical force fields for the Simulation of can include atomic coordinates of at least one polypeptide biomolecules and a package of molecular simulation pro having an amino acid Sequence Selected from the group grams. In particular, the model is preferably refined using consisting of the polypeptides shown in Table 1 (SEQ ID the “SANDER" program (for Simulated Annealing with NOS: 1-15). The data of the atomic coordinates should NMR-Derived Energy Restraints) was used. SANDER is the include atomic coordinates of the portion of the transmem main program used for molecular dynamics simulations. brane portion of the Subunit consisting of at least the amino SANDER allows for NMR refinement based on NOE acid sequence of residues 8 to 19 of SEQ. ID NOS: 1-15. derived distance restraints, torsion angle restraints, and 004.5 The processor 101 can generate a molecular model penalty functions based on chemical shifts and NOESY of the luminal domain portion, especially the pore, of a Volumes. ligand-gated neurotransmitter receptor having a Subunit Sto ichiometry ranging from (C)(B) to (C)(B)s. For example, 0049. Once the model has been refined using the the Subunit Stoichiometry can include(C)(B) useful for SANDER program of AMBER, the final model is evaluated. modeling the neuronal nAChR regulating cardiovascular A preferred Software package for evaluating the final model and GI actions. The model may be generated with only four is the PROCHECK package, e.g. version 3.5.4 (Laskowski helices to model other LGIC families. R A, MacArthur M. W. Moss D S & Thornton J. M., (1993) 0046) Modeling Step: PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Cryst., 26, 283-291). 0047. In generating a molecular model and simulating its PROCHECK checks the stereochemical quality of a protein interaction with various molecules, the computer System of Structure, producing a number of PostScript plots analyzing the present invention first generates a molecular model of its overall and residue-by-residue geometry. the receptor channel based on a template Structure deter mined in an NMR investigation of synthetic channel model 0050. In order to construct subtype-specific molecular (Opella S.J., Marassi F. M., Gesell J. J., Valente A. P., Kim models, the primary Structures of the particular Subtypes are Y., Oblatt-Montal M., Montal M., (1999) Structures of the required. Different Subtypes can be found in different region M2 channel-lining Segments from nicotinic acetylcholine of the human brain and peripheral nervous System and are and NMDA receptors by NMR spectroscopy. Nat. Struct. responsible for Specific functions. Subtype-specific models Biol. 6:374-9). Using this model, the molecular structures of of the nAChR luminal domain can be utilized in designing all of the neuronal subtypes of nAChR can be built. All subtype-specific NCIS. Subtypes of nAChR share Several common Structural arrangements in the luminal domain, which makes it pos 0051. The procedure of building the luminal model can Sible to build the model of a particular Subtype using a be easily adopted to constrain models of luminal domain of homology modeling approach. other subtypes of the nAChR and with some modification to constrain lumen models of other classes of ligand-gated ion 0.048. Once a molecular model is generated, the model is channels. The procedure is basically explained in the mod refined. A preferred Software package for refining the eling step of Example 1. The model of the C3f4-nAChR can molecular model is the AMBER molecular modeling pack Serve as the template to constrain other neuronal and mus age, e.g. AMBER version 7, (D. A. Pearlman, D. A. Case, J. cular Subtypes: Since those Subtypes are very homologous W. Caldwell, W. S. Ross, T. E. Cheatham III, S. De Bolt, D. (see Table 1). Only a few residues need to be modified in M. Ferguson, G. L. Seibel and P. A. Kollman, (1995) order to obtain new subtype. The new model after residue AMBER, a package of computer programs for applying modification must be Subjected to energy minimization by molecular mechanics, normal mode analysis, molecular AMBER procedures described previously and finally should dynamics and free energy calculations to Simulate the Struc be evaluated using PROCHECK. Elaborated docking pro US 2005/0033522 A1 Feb. 10, 2005

cedures can be applied to those models and the entire phenotype. In other words, in AutoDock local Search is used approach can be used in detailed molecular characterization to update the fitness associated with an individual in the of the luminal domain of specific subtypes of nAChR and genetic algorithm Selection. moreover, Subtype specific interaction with different classes of NCIS. 0057 The Lamarckian genetic algorithm uses as input a grid data set produced by the AutoGrid module. The Auto 0.052 More complicated procedures must be applied if Grid module is used to create 3-dimensional maps over the one want to obtain a model of the domain formed by other host protein using Several atom Specific and electronic classes of ligand-gated ion channels (GABA, NMDA, 5HT3 probes at each grid point. etc). First, amino-acid sequence alignment modeling is performed. An example and detailed description of Such 0058 Results of these simulations allow the classification analysis can be found in the paper by Bertaccini and Trudel of tested compounds in terms of free energy of binding, E. Bertaccini and J. R. Trudell, (2002) Predicting the which leads to the identification of ligands that may be transmembrane Secondary Structure of ligand-gated ion potent inhibitors. The same approach can be used to design channels Protein Eng. 15, 443-453). Thus, homologous new compounds with high affinity binding properties to a parts of the ion channel can be found and a new model of Specific Subtype of the nAChR. A compound that is identi transmembrane domain LGIC can be made. For some fied as a non-competitive inhibitor of a LGIC is one having LGICs, the transmembrane domain is formed by four trans a AG less than -6 kcal/mol, preferably less than -7 kcal/mol, membrane helices instead of five as in the case of nAChR. Still more preferably one having a AG less than -10 kcal/ In Such case one of the helices must be removed and the mol. remaining four need to be properly repositioned in order to 0059. The ligand structures used in docking simulations form the channel structure. Then the model can be relaxed are preferably made using the HyperChem package (of and refined in AMBER procedures and finally evaluated in HyperCube Inc., Gainsville, Fla.). In particular, it is pre PROCHECK. In case of such distinct models the docking ferred that the AM1 semiempirical method implemented in procedures need be parameterized by initial Studies as HyperChem be used to minimize the System energy and to described in the simulation step of Example 1 below. The calculate atomic charges in final structures (J. J. P. Stewart, values of the size of the grid box, the dielectric constant and Semiempirical molecular orbital methods, in: K. B. Lipkow the ganum evals must be optimized, since the size and itz, D. B. Boyd (Eds.), Reviews in Computational Chemistry, environment of the channel would have been changed Significantly. vol. 1, VCH, New York, 1990, pp. 45-81). 0060. The in silico approach described above can be 0.053 Using the modeling method of the invention, it has supported by examining the NCI-nAChR interaction by been discovered that there are NCI binding sites at the affinity chromatography (Jozwiak K, Haginaka J, Moaddel interface between C. and B helices of LGICs, especially of R and Wainer I W (2002) Displacement and nonlinear nicotinic AchRS. Among modeled candidate NCIS, the com chromatographic techniques in the investigation of interac pound enters into a Small hydrophobic pocket formed by tion of noncompetitive inhibitors with an immobilized nico residues 12, 15 and 18 of the transmembrane domains of the tinic acetylcholine receptor liquid chromatographic Station receptor subunits (e.g. SEQ ID NOS: 1-15, Table 1). A ary phase. Anal Chem 74: 4618-4624), preferably in an hydrophobic group of the NCI compound will interact with iterative fashion. Chromatographic affinity Screening can this portion of the NCI binding site. A polar group (e.g. an provide experimental data that is then employed for proper amino group) of a putative NCI can interact by hydrogen parameterization of the computer-based molecular simula bonding with Surrounding polar residues (e.g. residues 12 tion. Alternatively, the results of computer-based simulation and 14 of SEQ ID NOS: 1-10). can be related and evaluated by further chromatographic and 0054 Simulation Step: functional experiments. 0055. After generating the molecular model, the final 0061 Until recently, the screening of drug candidates for molecular model is used as a target protein for docking activity as NCIS was not a Standard procedure in the drug Simulation for compounds that may be potential inhibitors. development process. However, the present invention will A preferred Software package for docking Simulation is the permit pharmaceutical companies to rapidly Screen their AutoDock package, e.g. version 3.5. AutoDock allows dock potential drugs for NCI properties. In addition, the luminal ing of a flexible ligand into a rigid structure of the target domain of nAChR can be used as a target in drug discovery protein using genetic algorithms as the Search method. programs, which represents a new therapeutic approach to the treatment of diseases Such as Alzheimer's and Parkin 0056. A particular genetic algorithm included in the Son's diseases and for treatment of drug and tobacco depen AutoDock package is the Lamarckian genetic algorithm. dency, which are related to LGIC functions, especially to The Lamarckian genetic algorithm was preferably used with nAChR functions. local Search in order to improve efficiency. The Lamarckian genetic algorithm works in a reverse order compared to 0062) The nAChR, for example, was found to contain typical genetic algorithms. In particular, new traits in an two cholinergic agonist binding sites located at the interface organism develop because of a need created by the envi between the C. and B Subunits and on the extracellular ronment and these acquired characteristics are transmitted to N-terminal of the a Subunits. These Sites are key targets for its offspring. In AutoDock the ligands atomic coordinates drug discovery in a variety of diseases, including Alzhe represent a genotype and fitness is represented by interaction imer's disease (Cfa), Parkinson's disease (CB), cardio free energy with the proteins. Genotypes are found through vascular and GI actions (CB), anxiety and depression interations of the local Search and then the atomic coordi (C.f3), short term memory (C7) and auditory function and nates are translated into the ligand's State coordinates as the development (C.9). US 2005/0033522 A1 Feb. 10, 2005

0.063 Candidate NCI compounds discovered by the com CaCl2, and 11 mM glucose at pH 7.4) to remove Rb"free putational modeling method of the invention can be con in the medium. After washing, 1 ml of buffer with or without firmed by in vitro experimental methods. Two preferred the drugs under Study was added to each well, and the cells methods are by binding experiments or by functional assayS. were incubated for 2 min. The incubation buffer was then Either of these methods may employ the target LGIC, a collected, after which the cells were lysed in 0.1 N NaOH. population of LGICS representing the target receptor and The radioactivity in the buffer samples and cell lysates was receptors that the compound should preferably not inhibit (to measured by liquid scintillation counting. The total Rb" avoid side effects), or a population of LGICs representing a loaded into the cells (after washing) was calculated as the group of target receptors (with or without a group repre sum of the buffer samples and the cell lysates from each senting LGICs that the compound should preferably not well, and the amount of Rb" efflux was then expressed as inhibit). The LGICs for the in vitro functional assays can be a percentage of the total Rb" loaded (fractional release). present either as expression products in cells, as partially Stimulated efflux was defined as the difference between purified proteins, e.g. membrane preparations made as efflux in the presence and absence of nicotine (i.e., total known in the art, or as isolated proteins. If isolated proteins minus basal efflux). The maximum Rb" efflux, found at a are used in binding experiments, the proteins are preferably nicotine concentration of -300 uM or higher, was ~45% of immobilized. the amount loaded and was independent of the amount of 0064. A preferred binding assay is a displacement assay Rb" loaded into the cell. In studies to determine the performed as described by Jozwiak et al. Jozwiak K, inhibition of nicotine-stimulated Rb" efflux by the drugs Haginaka J, Moaddel R and Wainer I W (2002) Displace under Study, data were expressed as a percentage of control ment and Nonlinear chromatographic techniques in the values measured with 100 uM nicotine. investigation of interaction of noncompetitive inhibitors 0067. In these assays a compound is identified as a NCI with an immobilized C3f4 nicotinic acetylcholine receptor that inhibits the ion channel activity of the ligand-gated liquid chromatographic Stationary phase. Anal Chem 74: neurotransmitter receptor in nicotine stimulated Rb" efflux 4618-4624. Using this assay, a compound is identified as a with an ICs lower than 50 uM. A more preferred NCI non-competitive inhibitor of the ligand-gated neurotransmit compound is one that inhibits ion efflux with an ICso lower ter receptor as one that specifically binds to the ligand-gated than 5 uM. Even more preferable compounds are those that neurotransmitter receptor with a k" value greater than 8, inhibition efflux with an ICs lower than 500 nM. One of preferably with a k' value greater than 9 or even more skill in the art will recognize that compounds that are preferably a k" value greater than 10. effective at even lower concentrations are still more prefer 0065 Specificity of NCI binding to particular LGICs can able, and ICs of 50 nM, or even 5 nM might be observed. be shown by displacement of compounds that are Selective 0068. In some instances as described above, it might be for the pore portion of the desired LGIC. Specificity of the preferred to have a NCI that is selective for a particular binding to a nicotinic AChR and homologous receptors can LGIC. By “selective” is meant that the NCI inhibits the be shown by displacement by mecamylamine. Displacement target LGIC with an ICs that is at least 5-fold higher than of mecamylamine at a concentration of 10 um (of mecamy the ICs of the one or more LGICs that it is desired not to lamine) indicates good specific binding, ability to displace inhibit. The degree of selectivity is preferably 10-fold, more mecamylamine at a concentration of 40 uM indicates Strong preferably 20- to 50-fold, and still more preferably 100- to Specific binding. Preferably it is possible to displace 500-fold or more. mecamylamine at a concentration of 100 uM. Thus, a 0069. On the other hand, the binding assays or functional compound that is a preferred NCI of a nicotinic AChR is one assays also can be used to provide initial data that can be that exhibits a k" of greater than 8 in a chromatographic used to constrain the in Silico modeling method described binding experiment and can be displaced by mecamylamine above. Alternatively, the in Silico modeling and the in Vitro at a concentration of 10 to 100 uM. assays can be run iteratively to converge upon NCI com 0.066 Preferred functional ion channel activity assays are pounds that have desired properties. described by Hernandez et al. Hernandez S C, Bertolino M, Xiao Y, Pringle KE, Caruso F S and Kellar KJ (2000) 0070 Methods for synthesis of compounds of the inven Dextromethorphan and its metabolite dextrorphan block tion are considered within the skill of the ordinary synthetic Cl3(B4 neuronal nicotinic receptors. J Pharmacol Exp Ther chemist. Preferred NCI compounds have the above struc 293: 962-967) and by Jozwiak et al. K. Jozwiak, S C tural features and exhibit activity of inhibiting the ion Hernandez, K J Kellar, I W Wainer (2003) The Enantiose channel activity of a ligand-gated neurotransmitter receptor lective Interactions of Dextromethorphan and Levomethor in nicotine stimulated Rb" efflux with an ICso lower than phan with the C3,34-Nicotinic Acetylcholine Receptor: 100 uM or other activities as set forth in detail above. Comparison of Chromatographic and Functional Data Sub 0071 Dosage of compounds used for treatment of a mitted to J Pharmacol Exp Ther. In brief, 1-ml aliquots of subject can be easily determined by the ordinarily-skilled cells in growth medium were plated onto 24-well plates pharmacologist using known pharmacokinetic and pharma coated with poly(D-lysine). The plated cells were grown at codynamic assays and calculations from ICso data obtained 37° C. for 16 to 18 huntil reaching 90 to 100% confluence. by the inventive method. Doses of from 100 lug to 500 mg On the day of the experiment, the growth medium was per dose are typical. Formulation and administration of aspirated and the cells were incubated in fresh medium compounds useful for treatment is also well-known in the containing 2 uCi/ml RbCl for 4 h at 37° C. After this art. For example, many of the compounds listed in Table 2 loading procedure, the medium was aspirated and the cells have been administered therapeutically and it is expected were washed three times with 1-ml aliquots of buffer (15 mH that compounds of the invention can be similarly formulated HEPES, 140 mM NaCl, 2 mM KC1, 1 mM MgSO, 1.8 mM and administered. US 2005/0033522 A1 Feb. 10, 2005

EXAMPLE 1. ried-out in Stages by relaxing i) only hydrogen atoms, ii) hydrogen--side-chain atoms, or iii) all atoms except alpha Modeling of the Lumen of a C.3f4 nAChR and carbons. Finally, a restrained minimization was also per Docking of a Putative NCI formed on the alpha-carbons of all the chains/residues of the 0072 The molecular model of a 6-M2-nAChR trans model. This was to relax the structure but keep it near the membrane channel determined by frozen state NMR was initial position of the known template structure (PDB acces used as the template for further modification (atomic coor Sion no. 1 EQ8). Respective Scripts used to run model dinates were found in Protein Data Bank-PDB id: 1EQ8). refining with AMBER are presented in Apendix 1. This model represents a channel that mimics the transmem brane arrangement of known LGICs (Opella S.J., Marassi 0.077 Using PROCHECK to evaluate the model it was F. M., Gesell J. J., Valente A. P., Kim Y., Oblatt-Montal M., found that the whole luminal domain is constrained fully by Montal M., (1999) Structures of the M2 channel-lining C.-helix Secondary Structure. Along the lumen model Seven Segments from nicotinic acetylcholine and NMDA receptors rings of residues exposed to the center of the channel can be by NMR spectroscopy. Nat. Struct. Biol. 6:374-9). The found; three polar residues (E,T and S) and then three apolar model channel consisted of 5 uniform polypeptides oriented residues (L, V/F and LL) and the last polar residue (E/K). around a central pore. The amino-acid Sequence of this polypeptide is analogous to the Sequence of transmembrane 0078. It is believed that apolar rings in the middle of the M2 segment of 8 subunit of nAChR found in Torpedo Structure form the actual "gate' of the channel and play a Californica. role in conformational change of the receptor from a closed to an open state. Polar residues on both sides of the “gate' 0073. In the 8-M2-nAChR transmembrane channel, the Spatial arrangement of polypeptide helices conserves five participate in the cation Selective function of the receptor. An fold Symmetry, with certain residues exposed to the center of important Structural parameter found in the obtained model the pore. These residues (predominantly polar) form an is the change in position from Valine in the C3 Sequence to explicit surface of the channel. This is consistent with the phenylalanine in the f4 Sequence (see residue 15 in Table 1). concept of the presence of amino acid rings distributed along This provides the formation of small pockets between C3 the pore and is a common property found in all Subtypes of and B4 subunits, found during the simulation of NCI-C.3f4 nAChR and also other ligand-gated ion channels Changeux nAChR interactions. The developed model of C3 B4-nAChR J. P., Galzi J. L., Devillers-Thiery A., Bertrand D., (1992) luminal domain can be used as a template to constrain The functional architecture of the acetylcholine nicotinic homologous Systems of other nicotinic receptors, especially receptor explored by affinity labelling and Site-directed neuronal nicotinic receptors, and other ligand-gated ion mutagenesis. O. Rev. BiophyS. 25: 395-432). channels. 0.074. With respect to the spatial arrangement of five 0079 The resulting atomic coordinates represent the final helices in the luminal domain, distribution of certain amino model. FIG. 5 illustrates the electrostatic potential of the acid rings along the channel is a common property of all inner Surface of the ion channel, and especially the elec Subtypes of nAChR. Since primary Sequences acroSS differ tronegative potential of the cation Selector of the channel. ent Subtypes are predominantly homologous as presented in FIG. 3c shows an example of the resulting luminal domain Table 1, and essential (exposed) residues are highly con model having five helices forming the wall of the ion Served, a Subtype specific model of the luminal domain can channel. FIG. 3d shows a luminal domain model in perpen be built using homology modeling techniques. dicular view with residue rings. 0075 Based on the sequence comparison presented in Table 1, the initial model was modified by exchange of 8 0080. In order to perform docking simulations, the Auto helix residues into C3 and B4 using the SYBYL 6.8 molecu Grid module was first used to create 3-dimensional maps lar modeling system (Tripos Inc., 1699 South Hanley Road, over the host protein using Several atom Specific and elec St. Louis, Mo., 63144, USA). Therefore, the channel con tronic probes at each grid point. An example parameteriza taining C3, 34, C3, B4 and B4 helices, respectively, was tion file for the AutoGrid module used in this example can constrained. be found in Appendix 2. The optimal size of constrained grid 0.076 The model was further refined by energy minimi maps was a 22.5x22.5x45 A box (i.e., a grid of 60x60x120 zation using the Sander Classic module of AMBER 6.0 points, each separated by 0.375 A). This allowed exploration Software. Both termini of each helix were blocked in a of the whole internal Space of the lumen domain but pre Standard AMBER procedure: acetyl beginning groups vented ligands from being bound on the external Side. The (ACE) and N-methylamine ending group (NME) groups grid-box size can be altered in the 3" dimension (along the were attached, respectively, to each helix. The AMBER 94 lumen) in order to explore interaction with a particular force field (Cornell, W. D., Cieplak, P., Bayly, C. I., Gould, Segment of the lumen or to calculate the interaction profile I. R., Merz, Jr. K. M., Ferguson, D. M., Spellmeyer, D.C., along the model. Fox, T., Caldwell, J. W., Kollman, P.A., (1995).J. Am. Chem. 0081. An important parameter to properly explore elec Soc. 117, 5179-5197) parameters were used for energy tronic interaction in ligand receptor complexes is the dielec minimization with the convergence criterion of the root tric constant value (d) used to calculate the electronic grid mean-Square of the gradient to be less than 1.0E-4 kcal/mole map. During the initial evaluation tests, the Standard distant A. Each minimization run was started with the Steepest dependent dielectric constant did not produce proper results: descent followed by the conjugate gradient method. A dis the electrostatic interaction were almost Zero. The Simula tance-dependent dielectric function was used to evaluate the tion did not discriminate between neutral and protonated electroStatic energy. The energy minimization run was car ligands. US 2005/0033522 A1 Feb. 10, 2005

TABLE 2 AG (kcal/mol) of best docked conformation obtained in different dielectric environment. Diel. const. Dist. dependent 40 3O 2O 15 1O MCM -6.23 -6.22 -6.22 -6.23 -6.26 -6.43 -6.76 -19.52 MCM- -6.60 -7.56 -8.13 -9.29 -11.01 -14.47 -27.25 -138.2 DMT -8.65 -8.66 -8.68 -8.71 -8.73 -8.81 -8.99 DMT -8.74 -9.46 -9.77 -10.39 -11.87 -15.72 -28.OO LMT -8.31 -8.33 -8.34 -8.38 -8.40 -8.49 -8.70 LMT- -8.95 -9.53 -9.81 -10.59 -12.28 -15.69 -27.85

0082) A detailed test of severald values was carried out of test Simulations on Several ligands including conforma using three pairs of ligands and the results are presented in tionally flexible and rigid Systems was performed. It was Table 2. Table 2 shows an unexpected diminished difference found that a ga num evals value of at least 5 million is between neutral and protonated Systems when distant-de required to assure obtaining a Statistically significant number pendent d was used; differences gradually increase with of lowest energy complexes. In the case of bigger ligand decreasing d. Simultaneously the increase in electroStatic molecules with more than 2 rotatable bonds, the optimal impact in the ligand receptor interaction was noticed when value should be at least 50 million. Higher values are a low dielectric value was used. However, a very low value acceptable; however higher values may dramatically (ds 10) produced unrealistic AG values. Finally, as a mater increase the time of each Simulation. of compromising these two effects, d=15 was chosen for final calculations as the value producing Suitable electronic 0087. The optimal number of docking search runs was properties of the ligand-receptor complex in the transmem found to be 50. Again the number of docking Search runs can brane ion channel System. This approach is in agreement be higher, but would take more time for Simulation and have with values of the dielectric constant in transmembrane no effect on the final result. pores obtained by theoretical calculations (Cheng et al., 0088 AutoDock 3.5 implemented a free-energy scoring (1998) Eur: Biophys J., 27 105-112 and Gutman et al., function that is based on a linear regression analysis, the (1992) Biochim. Biophys Acta 1109: 141-148) where it was AMBER force field, and a large set of diverse protein-ligand found that the actual dielectric constant in transmembrane complexes with known inhibition constants (e.g. See Web channels remains low and ranges from 25 to 5 depending on Site at Scripps.edu/pub/olson-web/doc/autodock/). This the structure. Thus, in the case of the NCI-nAChR docking function was employed to estimate the free energy change of simulations d value can vary from 10 to 20. the NCI-nAChR complex and eventually lead to an esti mated inhibition constant of a particular ligand. Docking 0.083. The resulting ligand 3D structure was loaded into Simulations allow quantitative classification of the Stability the AutoDock System and was iteratively Sampled over of the NCI-nAChR complexes formed by tested ligands in previously created grid-maps in order to find optimal posi terms of free energy of binding, which eventually lead to the tions and the lowest energy of interaction. An example identification of ligands exerting potent inhibitory proper parameterization file for the AutoDock module used in this ties. It was found that molecular Systems forming the example can be found in Appendix 3. complex with AG value lower than -6.0 kcal/mol should be 0084. The Lamarckian genetic algorithm with local considered as potential NCIS. Lower AG values represent Search was used from the AutoDock package. Atomic coor more potent NCI compounds. Preferred NCI compounds dinate files of ligands were transformed into a format exhibit a AG value lower than -7.0 kcal/mol; more preferred suitable to AutoDock using the HIN2PDBQ script (Johans compounds exhibit a AG value lower than -10.0 kcal/mol. son M. (2002) Some computational chemistry related 0089. Detailed exploration of the spatial arrangement of python conversion scripts. See Web site helsinki.fi/%7Emp ligand-receptor conformations leads to building a pharma johans/python?). cophore model of a subtype specific NCI-nAChR. Simula 0085. The ligand structures used in the docking simula tions on the C3 B4 model showed that NCIS bind predomi tions were made using the HyperChem Software package. nantly in the channel in the apolar domain (F/V ring). Tested Further, the AM1 semiempirical method implemented in Structures primarily entered a Small hydrophobic pocket HyperChem was used to minimize the System energy and to formed between C3 and B4 subunits and Subsequently calculate atomic charges in final Structures. interacted with protein side chains, forming hydrogen bonds. It is expected that this is a type of interaction that would not 0.086 An initial simulation was performed in order to be found in those receptor Subtypes that lack the bulky optimize the docking Settings. Since previously described phenylalanine residue in this position. Since there are two docking space Seemed to be large in the model of C3f4 quasi-symmetrical pockets between C3 and B4 helices in the nAChR active site (22,781.25A) it was important to model, ligands most likely form two separate clusters on optimize the maximum number of energy evaluations these two symmetrical sites (e.g., FIG. 6) at which the energy (ga num evals) required in each Search run. It was found of interaction does not significantly differ. Estimated free that too low a value of ganum evals could result in energies of docking are in the range of experimental ICso of finishing the Simulation too quickly, and the global mini tested inhibitors and also can be related to experimental mum of the complex conformation may not be found. A Set affinity chromatography results. The model can be applied to US 2005/0033522 A1 Feb. 10, 2005 a variety of compounds and is useful for in Silico designing asSociation and dissociation of the drug-protein complex are of new drugs with particularly high non-competitive inhibi the main cause of band broadening and asymmetry of the tory activity. peak profile. The chromatographic peak profiles were ana lyzed using PeakFit v4.11 for Windows Software (SPSS EXAMPLE 2 Inc., Chicago, Ill.). The mathematical approach used was the non-linear chromatography (NLC) model derived from Chromatographic Assay of NCI Activity Impulse Input Solution Wade J. L., Bergold AF and Carr P W (1987) Theoretical description of nonlinear chromatog 0090 Chromatographic studies based on immobilized raphy, with applications to psychochemical measurements in nAChRS were performed to characterize ligand binding for affinity chromatography and implications for preparative broad groups of compounds. In order to further understand scale separations. Anal Chem 59:1286-1295. and described the mechanistic action of NCIS on the molecular level, the by Equation 1 (PeakFit User's Manual, p. 8-25): model of the transmembrane domain of the C3 B4 nAChR was built and used for computer simulations of docking inhibitors into the receptor. The entire approach allowed the Eqn. 1 classification of NCIS in terms of their functional effective CSS. al ?y. sell) 0091 FIG. 7 presents compounds tested on an C3 B4 nAChR column. The chemicals can be divided into several Subgroups. The first group contains drugs from different origin, which are well known as non-competitive inhibitors of nAChRs. The second group is of the dextromethorphan 0097 where: family, levomethorphan, dextromethorphan and its ana 0098 y-intensity of signal, logues, and the final group is Verapamil, its congeners, and metabolites. In order to properly assess the influence of 0099 x-reduced retention time, non-specific retention, five other chemicals (acetanilide, acetaminophen, 2,4-dinitrobenzoic acid, 3,4-dimethoxyben Zoic acid and phenylbutazone) were tested as negative T(u, v) = exp(-v) exp(-i).(2vvi di controls. The affinity of ligands was investigated by non O linear chromatography on an O.3f4 nicotinic receptor affinity column. 0092) 10° Cells from the KXO3(B4R2 cell line were 01.00) I() and I () are Modified Bessel functions suspended in Tris-HCl 50 mM, pH 7.4 (buffer A), homog 01.01 alo-area parameter, enized for 30 sec, and centrifuged at 35,000xg for 10 min at 4 C. The pellet was resuspended in 2% cholate in buffer A 0102) a-center parameter, reveal to true thermo and stirred for 2 h. The mixture was centrifuged at 35,000xg dynamic capacity factor, for 30 min, and the Supernatant containing Cl3(B4 nAChR 0103) a-width parameter, cholate solution was collected. 200 mg of the IAM station ary phase was added to the C3f4 nAChR-cholate solution. 01.04] a-distortion parameter. Subsequently the solution was stirred for 1 h. The suspen 0105 Experimental chromatograms obtained by single sion was dialyzed against 2x1 L buffer A for 24 h at 4 C. injection of ligand into the chromatographic column with The IAM liquid chromatographic Support containing the immobilized receptor were processed with PeakFit v4.11 Cl3(B4-nAChR was packed into a HR5/2 glass column to Software. After Standard linear baseline Subtraction, each form a chromatographic bed of 20 mmx5 mm i.d. The peak profile was fitted to the NLC function. The set of NLC Cl3(B4-nAChR column was then placed in the chromato parameters (ao, a, a and as) was collected for each profile graphic System and used. and used for the calculation of descriptors of the kinetic 0.093 Aqueous solutions 10 uM of each compound interactions with the immobilized nAChR, dissociation rate were prepared and 20 ul aliquots were injected into column. constant (off); equilibrium constant (K), association rate The mobile phase was composed of ammonium acetate 10 constant (k) real thermodynamic capacity factor (k’), mM, pH 7.4 modified with methanol in the ratio 85:15 according to the following equations: (v/v). The flow rate was 0.2 ml/min and the experiments were carried out at ambient temperature. 0094) DM and LM were monitored in the positive ion mode (ESI+). The compounds were detected using single k 1 Eqn. 3 ion monitoring at m/z=272 {MW+H"ion. The chromato of a., grams were recorded and processed using MassLynx v. 3.5. K = (3. Eqn. 4 (Micromass). Co 0.095 The non-linear chromatography approach was used ko = koff K Eqn. 5 to determine kinetics of the NCI-nAChR interaction in affinity chromatography Studies. 0096. The mathematical model assumes limited (and a 0106 where: to is the dead time of a column (time needed relatively low) number of active sites on the column. Slow by non-retained Substance to reach the detector); Co is a US 2005/0033522 A1 Feb. 10, 2005 11 concentration of Solute injected multiplied by a width of the 0109 Values of logK and k' presented in Table 3 can be injection pulse (as a fraction of column dead volume). regarded as a measure of relative affinity of tested NCI 0107 Thus, by analyzing the ligand in an immobilized compounds for the nicotinic AChR. Among tested com receptor System four descriptors can be collected: retention pounds, ethidium, clozapine, Verapamil and Some of its (k), association rate constant (k), dissociation rate con congeners (PR-22, nor-Verapamil and galapamil) have the Stant (k) and equilibrium constant (logK). It was found highest affinities towards the Cl3(B4 nicotinic receptor col that ligands which are non-competitive inhibitors have k" umn as reflected by both logK and k". Both Verapamil and greater than 8, k greater than 10x10 M's (preferred nor-Verapamil were tested for enantioSelectivity of binding inhibitors have k of greater than 15x10 M's espe towards nicotinic affinity column but chromatographic cially potent inhibitors have k greater than 30x10 M's experiments as well as NLC data did not exhibit noticeable 1), k, smaller than 15s' (preferably lower than 2 s') and differences between enantiomers. Interestingly, dex logK greater than 5.9 (preferably greater than 6.5). tromethorphan exhibited markedly increased affinity com 0108. The k value obtained in chromatographic experi pared to the optical enantiomer levomethorphan. ments is the one which is closely correlated with ICso values 0110. The NLC approach allows estimating the kinetic from functional in vitro or in vivo experiments. In the docking Simulation, it is preferred that AG be lower than -6 rates of the complex formation and dissociation, k. and k, kcal/mol (preferably less than -7 kcal/mol, most preferably respectively. The well-known and potent NCIS mecamy less than -10 kcal/mol). In functional nicotine Stimulated lamine, ketamine, ethidium and bupropion had high asso Rb+ efflux experiments, the ICs value is preferrably lower ciation constant rates. Ketamine, methamphetamine, aman than 100uM (preferred inhibitors exhibit an ICso<10 uM). tadine and mecamylamine dissociated markedly quicker than other tested ligands. The lowest dissociation constant TABLE 3 rates were exhibited by ethidium, clozapine and Verapamil congenerS. Detailed chromatographic characterization of tested non competitive inhibitors k" - retention capacity factor, k and ki are association and dissociation constant rates, respectively TABLE 4 (kinetics of formation and disformation of the complex in chromatographic system), logK is chromatographic equilibrium QSAR models build on of chromatographic data cOnstant. In general, it was found that different molecular properties correlated with chromatographic experimental characteristics: kon 10 koff log K non specific bulkiness (Vol) and lipophilicity (log P; TASA) konLC) M-1s-1 Is M descriptors, specific shape descriptors (like Xien), tested drugs another group were electronic properties: ability to form hydrogen bonds, EHoMo and Norder. Two latter descriptors can be amantadine 8.98 30.8 6.73 6.66 associated with ability of a ligand molecule to protonation. Both bupropion 12.97 28.7 5.14 6.75 bulkiness/shape and Hbond/protonation seem to be important in chlorpromazine respect of known mechanism of inhibition by luminal NCIS: clozapine 155.17 24.8 0.55 7.65 ligand must enter the polar pore, interact with negative and polar dilthiazem 43.53 26.8 1.60 7.22 surface (primarily designated for cation selection) and eventually ketamine 8.25 38.4 8.5O 6.65 block or inhibit the flux of ion during receptor's open stage. laudanosine 22.87 25.0 2.18 7.06 mecamylamine 10.89 40.1 5.96 6.83 Equation R F memantine 16.71 18.8 345 6.74 methadone 44.45 15.9 1.37 7.06 log k = 5.328(+0.745) + 0.00633(+0.000715) 961 63021 26 methamphetamine 8.38 29.1 6.81 6.63 Volume + 0.519 (+0.0740)EM - -0.165 MK-8O1 19.10 27.1 3.48 6.89 (+0.0317)Hbondacceptors - 0.2087(+0.0538)Node. phenylcyclidine 24.06 23.2 2.69 6.94 quinacrine log k = 4.152(+0.595) + 1.474(+0.483)RASA + .802 9.441 26 ethidium 191.82 35.9 O18, 8.30 2.383 (+0.499)XY + 0.117 dextromethorphan 61.30 23.7 1.01 7.37 (+0.033)Node +0.0486(+0.0161)Hbondacceptors levomethorphan 35.81 18.6 1.55 7.08 log k = -3.440(+0.653) - 0.00654(+0.000635) .969 80.326 26 dextrorphan 26.79 20.7 2.3O 6.95 Volume - 0.507 (+0.0657) Eco + 0.168 3 MM 56.47 18.8 1.OO 7.28 (+0.0281)Hbondsets + 0.2308(+0.0478)N ce. 3OM 26.45 14.3 1.97 6.86 log K = 9.830(+0.752) + 0.00321 (+0.000652) .908 34.654 26 verapamil-R 96.86 31.O O.68 7.66 TASA + 0.3982(+0.0794)EM - 0.057 verapamil-S 96.32 30.6 O66 7.66 (t0.022)Xiength nor-verapamil-R 97.99 16.0 O.58 7.44 nor-verapamil-S 97.86 15.6 O.61 7.40 galapamil 75.93 2O.O O.74 7.43 D-617 22.22 15.O 2.72 6.74 D-62O 17.72 11.6 3.43 6.53 0111 Examples of complexes resulting from simulations PR-22 99.29 16.0 O.53 7.48 are provided in FIGS. 6, 8 and 9. FIG. 6 shows a two cluster PR-25 1942 10.6 2.52 6.63 interaction of the ligand PCP with C3 B4. FIG. 8 shows the control compounds mecamylamine luminal domain of C3 B4. FIG. 9 shows the acetaminophen 5.30 8.4 17.17 5.69 MK-801 luminal domain of C3 B4. acetanilide 5.95 8.2 25.54 5.51 dimethoxybenzoic ac. 4.46 9.8 18.21 5.73 0112 Quantitative results of simulated docking affinities dinitrobenzoic acid 7.77 9.1 12.12 S.87 phenylbutazone 6.29 8.7 22.22 5.59 were related to experimental results from chromatographic Studies. Using AutoDock’s Scoring function, estimated inhi bition constant were calculated. These values exhibited very US 2005/0033522 A1 Feb. 10, 2005 12 good correlations with affinity data from NLC calculations rphan is less likely to form Such interaction. This makes a (FIG. 10). This correlation can be illustrated by equation: difference instabilities of two complexes by ca. 0.3 kcal/mol determined by both docking and chromatographic analysis log k'-0.418(+0.037)log(1/K)-0.89(+0.19) (FIG. 11). r=0.930 F-127.7 n=22 0114. Furthermore, the estimated inhibition constant TABLE 5 obtained during the simulations is very well correlated with equilibrium measures obtained in affinity chromatographic the collection of dextromethorphan (DM)/(LM) levomethorphan experiments. characterization by different approaches (chromatographic and docking were explained above), functional in vivo is nicotine stimulated Rb+ efflux experiments: it was found that DM EXAMPLE 3 has significantly longer recovery time than LM, which was predicted by chromatographic and docking modeling. QSAR-3D Clustering Technique The ICso does not significantly differ. 0115 Classical methods for the identification and char Descriptor DM LM acterization of non-competitive inhibitors to ligand gated ion channels are time consuming. They are not applicable to the functional in vivo rapid Screening of chemical libraries for potential new drug ICso LuM 10.10 (+1.10) 10.90 (+1.08) candidates nor can they be routinely used in the new drug % recovery after 49.83 (+5.16) 79.00 (+3.50) development process. An important advancement in the area 7 min. washout is the development of a method of identification of potent % recovery after 82.09 (+3.64) 94.09 (+4.43) 4 h. washout NCIS. The method is based of the chemometric processing chromatographic of the chromatographic data obtained using a Stationary (NLC and van't Hoff) phase modified by immobilization of particular subtype of the receptor. The non-linear chromatography approach K 61.30 (+0.27) 35.81 (+0.15) K. LuM"sec' 23.66 (+0.61) 18.61 (+0.38) allows description of the NCI-receptor interactions in terms Konsec' 1.01 (+0.01) 1.549 (+0.002) of real thermodynamic capacity factor (k), equilibrium K, LuM" 23.40 (+0.36) 12.01 (+0.23) constant for binding (KC) and kinetics rate constants for logKa 7.37 7.08 AH kcal mol –6.92 (+0.19) –6.59 (+0.18) association (k) and dissociation (k). We have determined AS cal mol"T" -15.70 (+0.7) -15.20 (0.6) that a strong correlation exists between the drug ka param AG kcal mol" -2.33 (+0.4) -2.04 (+0.4) eter obtained in affinity chromatography experiments and docking the relative length of the effect of this drug in functional AG kcal mol" -8.73 -8.40 studies (nicotine stimulated efflux of Rb", from cells Edocked kcal mol"l -8.84 -8.52 expressing the target naChR) (K. Jozwiak, J. Haginaka et KM 3.98 * 10-07 6.91 * 10-07 al., Anal. Chem., 2002, 74, 4618-4624. and K. Jozwiak, S. logK -6.40 -6.16 C. Hernandez et al., J. Chromatogr. B. 2003, 797,423-431). 0116 Astrong relationship between the chromatographic 0113) Enantiomers have identical physiochemical prop rate constant and the length of the functional effect was erties and, therefore, all possible non-specific interactions found. However, more than chromatographic affinity has between the enantiomers of a chiral NCI and an immobilized been found necessary to predict the ICs value for NCI nAChR stationary phase should be equivalent. Any differ activity. The non-linear chromatographic parameters deter ences in the chromatographic retention between the enanti mined in these Studies were obtained in a dynamic System omers will be due to Specific binding interactions with the but under Simplified conditions when compared to a func active site of the protein. FIG. 11 shows chromatograms of tional assay (i.e. no neurotransmitter stimulation, no trans dextromethorphan (DM) and its enantiomer-levomethor membrane potential, etc.). Thus, the efficacy of the NCI's phan (LM). The pair of enantiomers was further investigated expressed as ICso values were not directly correlated with by chromatographic, docking and functional Studies (Table the calculated non-linear chromatographic parameters. 5). It was learned from the chromatographic experiments 0117 Quantitative Structure-Activity Relationship that the drug dextromethorphan (DM) exert higher affinity (QSAR) analysis provided models of the chromatographic on C3 B4-nAChR than its enantiomer levomethorphan (LM) affinity (Table 6). Each of the derived equations contains a and the difference in AG of the complexes was 0.3 kcal/mol. descriptor related to the electronic properties of the NCI's, These data were valuable in evaluating parameter Selection Eoo (Energy of the Highest Occupied Molecular during initial tests of the docking simulations to optimally Orbital), TPSA (Total Polar Surface Area) or a number of choose the channel dielectric constant or evaluate the use hydrogen bond acceptors. These models are consistent with fulness of the Scoring function for calculating estimated AG the fact that NCI's bind at the internal Surface of the nAChR implemented in AutoDock. The docking Simulations give ion channel, which is highly polar and negatively charged. insights into chiral recognition on the molecular level (FIG. Three of the four equations also contain a shape descriptor 12). In binding to the C3f4 luminal domain, both molecules (Shadow-YZ), which is consistent with the fact that the interact initially with a hydrophobic pocket on the border NCI’s bind within a defined space on the receptor. Thus, the between the C3 and B4 helices (FIG. 12a). This binding QSAR analyses describe a chromatographic and, as dis determines the positions of the terminal amine group (blue) cussed above, NCI-receptor proceSS where the primary differently for dextromethorphan (grey) than levomethor driving force is electrostatic interactions between positively phan (magenta). The amine group of dextromethorphan can charged ligands and a negatively charged nAChR, which easily form Secondary interaction hydrogen bonds with take place in the Structurally defined central pore of the neighboring polar residues (orange balls), while levometho receptor. US 2005/0033522 A1 Feb. 10, 2005

8 compounds (dextromethorphan analogs, clozapine, TABLE 6 laudanosine and phencyclidine) with high values of log k. and EM but moderate YZ (mean values/range: 1.61/1.3 to QSAR equation describing affinity chromatography parameters. 2.2, -8.64/-9.0 to -7.7; and 50.0/45 to 60, respectively); and log k = 5.255(+0.942) + 0.491 (+0.092) Eqn. 1 Cluster 3 contained 7 compounds (MK-801, adamantadine, Eoo + 0.0118(+0.0049)YZ bupropion, ketamine, mecamylamine, memantine, metham r = 0.894, s = 0.168, F = 27.929, n = 17 log k = 7.693(+0.111) - 0.00787(+0.00257)YZ + Eqn. 2 phetamine) with low values of log k", Eomo and YZ 0.0700(+0.0237)Hbonde - -0.00276 parameters (mean value/range: 1.06/0.9 to 1.3; -9.45/-9.8 to (+0.00118)TPSA r = 0.762, s = 0.0883, F = 5.997, Outlier: n = 17 mecamylamine -9.1 and 37.9/25 to 45, respectively). log k = -3.096(+0.926) - 0.454(+0.090)Eoo - Eqn. 3 0.0128(+0.00471)YZ 0123 The analysis Segregates the compounds by size and r = 0.891, s = 0.165, F = 26.961, n = 17 charge, with the Smaller, more electronegative compounds log K = 11.412(+0.604) + 0.492(+0.0669)EHoo Eqn. 4 appearing in Cluster 3. This division reflects a pharmaco r = 0.885, s = 0.135, F = 54.130, n = 17 logical reality Since compounds contained in Cluster 3 can rapidly and deeply penetrate the luminal pore of the nAChR producing a high percentage of blockade per concentration 0118. A 3-dimensional scatterplot of the variables asso ciated with Eqn. 1, i.e. log k", EM and YZ, Suggested that of molecules. This would be reflected in lower ICs values. the whole cassette of tested NCI's could be Subdivided into 0124. The ICs values have been established for 4 in three separate clusters, FIG. 13. Cluster 1, 4 of the 8 in Cluster 2 and 6 of the 8 in Cluster 0119) The parameter k' is derived from chromatographic 3 using the Rb+efflux assay described above using cell lines experiments using the non-linear chromatography approach expressing the relevant receptor. KXa3b4R2 is a line of described by Jozwiak et al. (Jozwiak K, Haginaka J, Moad human embryonic kidney 293 cells stably transfected with del R., Wainer I W. Displacement and nonlinear chromato rat neuronal nicotinic acetylcholine receptor (nAChR) C3 graphic techniques in the investigation of interaction of and 34 subunit genes. This cell line can be obtained from Dr. noncompetitive inhibitors with an immobilized alpha3beta4 Kenneth Kellar-Department of Pharmacology, George nicotinic acetylcholine receptor liquid chromatographic Sta town University, Washington, D.C. K177 is a line of human tionary phase. Anal. Chem. Sep. 15, 2002;74(18):4618-24) embryonic kidney 293 cells stably transfected with human and applied in Example 2 above. Eomo is given in electron neuronal nicotinic acetylcholine receptor (nAChR) C3 and volts (eV) and is the highest occupied molecular orbital B4 subunit genes. These cells can be obtained from Dr. energy. EcoMo is an electronic descriptor of the molecule Daniel Bertrand, Dept. of Physiology, University of Geneva, obtained in molecular Simulation. In the present Example Switzerland. SH-SY5Y cells are a human neuroblastoma E. was calculated using the MOPAC module in Cerius’ clonal Subline of the neuroepithelioma cell line SK-N-SH software (Cerius2 v. 4.8. Accelrys Inc., San Diego, Calif.). from the bone marrow. This cell line can be obtained from Additional information about the EoNo parameter can be the European Collection of Cell Cultures (ECACC), cata found in J. M. Goodman, Chemical application of Molecu logue no. 94030304. PC-12 Rat adrenal gland pheochro lar Modeling, c. Royal Society of Chemistry. 1998. p. 139. mocytoma cells are available from the American Type 0120 “Shadow YZ” is a surface area projection descrip Culture Collection, ATCC Number CRL-1721. Results are tor-the molecular Surface is projected the YZ plane (deter shown in (Table 7). mined by principal axis of inertia of the molecule) and the shadow is calculated in A. In the present example, we used 0.125 When these values were considered in relationship the QSAR+ module of Cerius' software (Cerius2 v. 4.8. to the compounds in Clusters 2 and 3, 3 of the 4 compounds Accelrys Inc., San Diego, Calif.). More information about in Cluster 2 had ICs valuese 10 uM while 5 of the 6 the Surface area projection descriptor can be found in compounds in Cluster 3 had ICso valuess 10 uM. Rohrbaugh et al. (Rohrbaugh RH, Jurs PC. Molecular shape and the prediction of high-performance liquid chromato TABLE 7 graphic retention indexes of polycyclic aromatic hydrocar The ICsa values of Rb" efflux of various compounds. bons. Anal Chem. Apr 1, 1987:59(7):1048-54). Cluster 0121 A cluster analysis based on the three properties was Ligand ICso Number Cell line carried out using K-mean clustering method of variables and methadone 1.9 (+0.2) 1. KXCI.334R2 the results confirm that there are 3 distinct clusters. K-mean Verapamil 8.1 (+1.3) 1. KXCI.334R2 clustering is a Standard clustering method that determines a nor-verapamil 2.6 (+1.0) 1. KXCI.334R2 user-specified number of clusters with the goal of minimiz Dilthiazem 2.26 (+1.0) 1. KXCI.334R2 dextromethorphan 8.9 (+1.1) 2 KXCI.334R2 ing within-cluster variability while maximizing between 10.1 (+1.10) KXCI.334R2 cluster variability. In the present example, the method was levomethorphan 10.9 (+1.08) 2 KXCI.334R2 implemented as in Statistica (STATISTICA v. 6.0. Statsoft dextrorphan 29.6 (+5.7) 2 KXCI.334R2 phencyclidine 7.0 (+1.3) 2 KXCI.334R2 Inc., Tulsa, Okla.). 5.9 SH-SYSY MK-8O1 26.6 (9.6) 3 KXCI.334R2 0122) Cluster 1 was formed by four compounds (dilthi mecamylamine 1.0 (+0.04) 3 KXCI.334R2 azem and methadone Verapamil and nor-Verapamil) and can memantine 6.60 (+0.92) 3 K177 (c4|B2) be characterized by high values of log k" and EM and YZ amantadine 3.44 (+0.67) 3 K177 (c4|B2) parameters (mean values/range: 1.645/1.3 to 2.2, -8.93/-9.2 Bupropion 1.4 3 SH-SYSY to -8.6 and 64.5/60 to 70, respectively); Cluster 2 included US 2005/0033522 A1 Feb. 10, 2005 14

enantioselectivity between dextroemthorphan and TABLE 7-continued levomethorphan (Jozwiak, K.; Hernandez, S.C.; Kellar, K. J.; Wainer, I. W. The enantioselective interactions of dex The ICso values of Rb" efflux of various compounds. tromethorphan and levomethorphan with the Cl3(B4-nicotinic Cluster acetylcholine receptor: comparison of chromatographic and Ligand ICso Number Cell line functional data. J. Chromatogr. B. 2003, 797,423-431). Ketamine 5.2 (+0.5) 3 PC-12 0.130. The clefts are associated with the presence of 1.4 SH-SYSY phenylalanine in B4 M2 domain and will not exist in other, non-f4 subtypes of the nAChR. Interestingly, the results from a chromatographic Study which utilize an immobilized 0.126 The method of NCI clustering using Equation 1, Cl3(B2-n AChR column showed enantioselectivity for dex above, identifies potent NCIS, i.e. those with low ICs. tromethorphan and levomethorphan Significantly dimin Compounds belonging to cluster 3 are considered as poten ished as compared to the immobilized C.3B4-nAChR column tial NCIs and are expected to be effective in functional tests. (Table 7), further supporting the conclusion that the cleft is Compounds in the cluster 2 are expected to express weaker a feature of B4 Subtype receptors that can be important for inhibition properties. The compounds of cluster 1, which NCI activity. consists of large, bulky compounds with Strong chromato graphic affinity, are also expected to be potent NCIS. Ini tially, the ICso value of only one of the four compounds in TABLE 7 cluster 1 was known, methadone. The cluster analysis pre Comparison of enantioselectivity of DM/LM pair of enantiomers dicted that Verapamil, nor-Verapamil and diltiazem should on two different nAChR systems. Experimental data from affinity be effective NCIS of the C3f4 nAChR and functional studies chromatography - Selectivity factor (C. confirmed this prediction. C.3B4 Cl3(B2 0.127) Functional studies were carried out using a nico tine-stimulated Rb"efflux assay on KXC334R2 cell line kfM 1.62 1.03 expressing C.3f4 subtype of neuronal nAChR. The studies ki revealed that the ICso values of dilthiazem, Verapamil and AAG = -RTInc. -0.29 kcal/mol -0.02 kcal/mol nor-verapamil are 2.3 uM, 8.2uM and 2.1 uM respectively. Thus all compounds in cluster 1 are strong inhibitors. The cluster analysis technique is applied in Example 6 below to 0131 Based on these observations, a molecular model of identify compounds with high potency as NCIS, which was Cl3(B2 luminal domain was constructed. The main difference identification was further verified by functional studies of in the Structure of the B2-type channel is the exchange of Rb efflux. phenylalanines from 34 helices for valines associated with 0128. The technique of cluster analysis using Eqn. 1 also f2 helices (See Table 1). A graphic representation of the Suggests that high NCI potency can be attributed to two model is presented in FIGS. 3a and 3b. FIG.3b. shows the Structurally different groups of compounds. It can be specu residues forming the surface of the channel. The distribution lated that the two groups of compounds may express their of the particular rings along the channel can be easily inhibitory properties by two different molecular mecha noticed. The rings are distributed as follows (from top to nisms. bottom): extracellular ring, leucine ring, valine ring, leucine ring, Serine ring, threonine ring and intermediate ring (con EXAMPLE 4 sisting of glutamic acid residues) and this is consistent with general considerations. The Cl3(32 model revealed Some Investigation of C3B2 nAChR Subtype important differences when compared with the C3 B4 chan nel. The most important is the lack of clefts formed on the 0129. The C3 B4 subtype of the nAChR is extensively apolar surface of the lumen. The C3 B2 model is considered characterized, easily accessible in Stably transfected cell the more general of the two and represents the shape of the lines (e.g., KXO3(34R2) and widely tested in functional channel associated with majority of Subtypes of neuronal Studies. Moreover, functional Studies of this Subtype are nAChR, since there is no Substantial difference in the relatively easy. However, the transmembrane domain of Sequence of the exposed residues along the channel when Cl3(34-Subtype has Some unique features not found in other compared with other Subunits. Only the B4 Subunit possesses Subtypes. AS previously Stated, the general Structure of the a significant mutation of Val->Phe in the valine ring. There luminal domain is believed to be fairly well conserved fore, the new model of the C3B2 subtype is more homolo among the subtypes of the nAChRs. However, the M2 gous to other important subtypes of nAChR than the C3f4 transmembrane part of B4 Subunit has one critical mutation (phenylalanine (F) in the B4 subunit at position 15 while model and should be considered as a general template for most other subunits have a valine (V) in this position (Table detailed Studies of other nAChRS and in Some perspective 1). As a result the nAChRs containing the f4 Subunit may other members of the ligand gated ion channel Superfamily. display Significantly different properties than would other 0132) The different structure of the luminal domain of the Subunit types and may exhibit differences in the interaction Cl3(32 channel produces a profound change in the docking of the nAChR channel with NCIS. The introduction of the interaction of NCIS. Since there is no apolar cleft on the phenylalanine moieties on the B4 Subunits produces Small Surface of the channel, the NCI molecules must find alter clefts in the Surface of the luminal domain of the channel. native interaction in the binding Site. The case of Special These clefts play an important role in in the binding of NCIS, interest are docking Simulations: dextromethorphan and as described further below. The cleft explains the observed levomethorphan. FIGS. 12a and 12b present the overlaid US 2005/0033522 A1 Feb. 10, 2005

lowest energy conformations of these two enantiomers proximity to (i.e., approximately 5 to 10 A from, preferably docked onto the model of C3f2 luminal domain. Two from 5 to 8 A from, more preferably less than 7 A from) said molecules adopt different orientations compared to docking hydrophobic moiety. The amino group can be directly onto Cl3(B4 model: the ligands binds primarily to the apolar bonded to the bulky hydrophobic moiety or can be linked by part of the lumen with nitrogen atom interacting with Serine a Spacer moiety, Such as, but not limited to, a short hydro ring. However, in contrast to C3f4 docking, there is no carbon chain. The amino group can be Substituted defined cavity on the Surface and only Side interactions are (-NRR, where RandR are the same or different and are possible which results in the Significantly weaker energy of Selected from the group consisting of H, C-C alkyl, C-C, interaction and, what is even more important, the AG alkoxy, dialkyl keto). The substituent is preferably one that difference between DM-C3f2-nAChR complex and retains a hydrogen-bonding potential; a preferred Substituent LM-C3f2-nAChR complex is significantly diminished is a keto-group, for example a dialkyl keto group, especially when compared to simulations on Cl3(B4-nAChR model CH(C=O)CH. Another preferred substituent is a hydroxyl (Table 8). or alkoxyl (-CH-OH) group, e.g. a C-C normal or branched alkoxyl group. Preferred Substituted amino groups 0133. The presence of the hydrophobic cleft in the C3f4 are a dialkyl keto amino group (e.g., HNCH (C=O)CH), a receptor Subtype and its absence from the C3f2 Subtype hydroxyl amino group or a methoxy amino group. An presents a target for designing of compounds that are spe example of Such a compound is 3-methoxy-17-propane-2- cific for one subtype over the other. one 9C, 13C, 14C, morphinan. TABLE 8 0.137. A preferred compound designed using the method of the above considerations is one comprising a hydrophobic Enantiospecificity of dextromethorphan and levomethorphan in group. A preferred hydrophobic group comprises at least one docking simulation studies of the C3f4 and C332 luminal channels. ring that includes at least two conjugated unsaturated bonds, Said ring optionally being fused to additional rings to form a ring System and Said additional rings optionally including AGDM kcal/mol) -8.73 -7.10 one or more hetero atoms. Alternatively, the hydrophobic AGLM kcal/mol) -8.40 -6.93 group can be a hydrocarbon chain or Saturated cyclic com AAG kcal/mol) -0.33 -0.17 pound. The hydrocarbon chain can be linear or branched and preferably contains from 4 to 10 carbon atoms, more pref erably from 4 to 7 carbon atoms. The hydrocarbon chain can further include alkenyl or alkynll unsaturations at one or two EXAMPLE 5 positions. Designing of New NCI Molecules 0.138. The compound will also preferably contain a hydrogen bond accepting group, which more preferably is 0134) The molecular models, clustering analysis and Selected from the group consisting of a keto group, a dynamic chromatographic method of the invention can be nitrogen-containing heterocyclic group and a guanidinium used to design molecules that possess enhanced activity as group. Typically, the ring or ring System and Said hydrogen NCIS of nAChRs. The molecular model of the NCI binding bond accepting group are joined by a linker comprising 1 to Site and docking Studies provide an understanding of the 4 carbon atoms and optionally containing an oxygen or mechanism of non-competitive inhibition. Using the dock Sulfur atom. ing orientations of molecular NCI-nAChR complexes, we have designed modifications of known molecules to more 0139. One consideration for design of effective NCI Strongly accommodate the active site and as a result molecules is that the molecule will preferably Span portion obtained new compounds that express Stronger NCI activity. of the luminal domain from the hydrophobic region defined Such new molecules are of interest in the pharmaceutical by the leucine and/or valine rings to the more polar region industry as new treatments of disorders associated with defined by the serine and/or threonine rings (FIG. 2). nAChR overactivity, e.g., as aids in Smoking cessation. 0140. The compound will preferably have activity as a 0135) In general, the docking orientation of a putative non-competitive inhibitor of Rb" efflux of a ligand-gated NCI is such that the molecule occupies a position within the neurotransmitter ion channel receptor with an ICs of less luminal channel and exhibits a AG of about -8.5 kcal/mol. than 10 uM. The molecule will generally be designed to have molecular 0141 AS an example of Such an approach, we undertook contacts with at least one, preferably 2, 3 or 4 of the Side modifications of dextromethorphan (DM). DM possesses chains of the amino acids lining the luminal channel. strong activity as a NCIS (IC=10 uM on C3 B4-nAChR Molecular contacts that are useful in providing high binding with a greatly prolonged duration of the NCI action), it energies (i.e. negative AG), include hydrogen bonds and pi easily passes blood-brain barrier and unlike as its enanti orbital overlaps. omer levomethorphan expresses little action on opioid 0.136 A structure-activity relation for a NCI of a LGIC receptors. The docking orientation of the DM-C3 B4-nAChR has been derived using the above-described methods. Thus, complex is presented in FIG. 15. The molecule occupies the a compound having a bulky hydrophobic moiety (e.g., a valine/phenylalanine cleft on the border between C.3-M2 phenyl or napthyl ring System or other fused aromatic ring helix and B4-M2 helix with an amino- group exposed for System, cyclopentyl or cyclohexyl ring System, a fused ring interaction with polar residues below this cleft. The inter System including but not limited to bicyclo 2.2.1]heptane, action would be even Stronger if the amino group could bicyclo 2.2.2]octane, morphinan and dibenzo 1.4 diaz interact with the Serine residues forming the Serine ring, but epine) and a primary, Secondary or tertiary amino group in this interaction is prevented by the distance of ca. 5 A US 2005/0033522 A1 Feb. 10, 2005

Separating two moieties. Study of the complex showed that the energy of interaction should be significantly enhanced if TABLE 9-continued a methyl group attached to nitrogen would be exchanged Molecules designed to inhibit C3f4-nAChR followed by AG values into a longer moiety with hydrogen bond acceptors in order obtained in docking simulations (reference AG of dextromethorphan = to allow forming Strong hydrogen bonds and therefore -8.73 kcal/mol). Stabilizing the complex. Several possible patterns of dex Compound Formula AG tromethorphan modification were designed and these are shown in Table 9. Those molecules are based on the inter DM-03 -10.31 kcal/mol action with Cl3(B4-nAChR and may be possibly selective O blockers for this subtype. Compound DM-01 was synthe sized and tested for activity in a Rb" efflux assay. The synthesis of DM-01 is described in FIG. 14 and the data are presented in FIG. 15.

TABLE 9

Molecules designed to inhibit C3f4-nAChR followed by AG values obtained in docking simulations (reference AG of dextromethorphan = -8.73 kcal/mol).

Compound Formula AG DM-04 HN1\ -9.39 kcal/mol dex- -8.73 kcal/mol NH trometorphan

DM-01 O -9.09 kcal/mol DM-05 1N -10.18 kcal/mol HN NH

N ( ) s

-O DM-02 O -9.35 kcal/mol EXAMPLE 6 Prediction of Side Effects of Compounds Mediated by Non-Competitive Inhibition of Ligand-Gated Ion Channels 0142. The analytic methods of the present invention can also be applied to assessment of NCI activity of compounds compounds, both known drugs and novel compounds, to predict Side effects. For example, the drugs Verapamil, nor-Verapamil and dilthiazem are commonly administered for treatment of high blood pressure. An undesirable side effect of these drugs is constipation. 0.143 AS explained above, the C3f4 nAChR subtype plays a role in regulation of gut motility and the Side effects US 2005/0033522 A1 Feb. 10, 2005 of Verapamil, nor-Verapamil and dilthiazem on gut motility have been related to NCI activity of these compounds TABLE 10-continued against the nAChR. AS an example of application of the Results of the cluster analysis characterization of tested calcium analytic methods of the present invention to the investiga channel blockers. tion of drug Side effects, we applied the cluster analysis method to predict the NCI activity of various compounds Ligand logk EHoMo YZ Cluster No. ICs uM used as calcium channel blockers for treatment of high blood amlodipine 2 -8.7228 62.93.348 1. 5.8 pressure, or their metabolites (MA-M6 and D-620) and nifedipine 1.27 -8.6323 58.42415 2 24.7 listed in Table 10. NCI activity is predicted if the compound falls into Cluster 1. Predicted NCI activity (or lack thereof was then confirmed using the Rb" efflux assay. Results are 0148 Compounds named in Table 3, Table 7 or Table 10, shown in Table 10. or specifically named in FIG. 9 or FIG. 13, and bupropion, 0144. The ranges defining clusters are as above and are: ketamine, laudanosine, mecamylamine, methadone, Cluster 1 (low ICso values) log k from 1.3 to 2.2 and EHoo MK-801, phenylcylclidine, ethidium, and dextromethorphan from -9.2 to -8.6 and YZ from 60 to 70; Cluster 2 (high ICso are compounds known in the prior art and So are not values) log k" from 1.3 to 2.2 and E from -9.0 to -7.7 considered to be inventive compounds per Se within the and YZ from 60 to 45; and Cluster 3 (low ICso values) log scope of the present invention. Methods of the invention for k' from 0.9 to 1.3; E from -9.8 to -9.1 and YZ from non-competitively inhibiting a LGIC, especially a nicotinic 45 to 25. AChR, or for treatment of a disease mediated by overactivity of a nicotinic AChR, exclude the use of bupropion, ket 0145 The chromatographic method using Cl3(B4 nAChR amine, laudanosine, mecamylamine, methadone, MK-801, column was used to obtain experimental affinity for 13 phenylcylclidine, ethidium, and dextromethorphan. Structures (diltiazem and 5 of its metabolites; Verapamil and 3 of its metabolites, nicardapine, nifedipine and amlo 014.9 The invention may be embodied in other specific dipine). The computational method of the invention was use forms without departing from the Spirit or essential charac to calculate EM and YZ descriptors and the data are teristics thereof. The foregoing embodiments are therefore presented in Table 10. Based on this data all compounds to be considered in all respects illustrative rather than were assigned to respective clusters (Table 10). limiting on the invention described herein. Scope of the 0146). After the prediction has been done based on clus invention is thus indicated by the appended claims rather tering the actual values of ICso were determined using the than by the foregoing description, and all changes that come nicotine-stimulated Rb" efflux assay on cell line within the meaning and range of equivalency of the claims KXO.3f4R2. These data are also presented in Table 10 and are intended to be embraced therein. the comparison of cluster method prediction with actual activity gives very good agreement-all tested ligands falls 0150 All patent and literature references cited herein are into proper categories. hereby incorporated by reference in their entirety and for all purposes, including the following references: 0147 As it can be seen from Table 10 all of the tested compounds could be assigned to either cluster 1 or to cluster 0151) 1. Wainer I W, Zhang Y, Xiao Y, Kellar KJ (1999) 2, which Segregate them into two groups: Very effective Liquid chromatographic Studies with immobilized neuronal NCIS (ICso<10 uM-cluster No. 1) and less effective NCIS nicotinic acetylcholine receptor Stationary phases: effects of (ICso>10 uM-cluster No. 2). The cluster analysis properly predicted the NCI activity of all 13 drugs and metabolites. receptor Subtypes, pH and ionic Strength on drug-receptor Furthermore, the results Suggest that the cardiovascular interactions. J Chromatogr B Biomed Sci Appl 724:65-72. benefit attributed to calcium channel blocking activity may 0152 2. Zhang Y, Xiao Y, Kellar KJ, Wainer IW (1998) derive at least in part from previously unrecognized activity Immobilized nicotinic receptor Stationary phase for on-line of inhibition of ligand-gated ion channels. liquid chromatographic determination of drug-receptor TABLE 10 affinities. Anal Biochem 264:22-5. Results of the cluster analysis characterization of tested calcium 0153. 3. Barrantes F J. (2002) Lipid matters: nicotinic channel blockers. acetylcholine receptor-lipid interactions (Review). Mol Menbir Biol 19:277-84. Ligand logk EHoMo YZ Cluster No. ICs uM Dilthiazem 1.64 -8.66306 62.297.32 1. 2.2 0154 4. Morris GM, Goodsell DS, Halliday RS, et al. MA 1.61 -8.6465 66.535 1. 4.2 M1 1.6 -8.5788 62.40288 2 30.4 (1998) Automated docking using a Lamarckian genetic M2 1.61 -8.58 57.4339 2 77.6 algorithm and empirical binding free energy function. M4 1.45 -8.4067 6184228 2 73.2 19:1639-62. M6 1.48 -8.6465 58.1142 2 63.1 Verapamil 1.99 -9.05746 64.8O137 1. 8.1 0155 Appendix 1 AMBER Scripts for Stepwise Refining Nor-verapamil 1.99 -9.12446 64.82286 1. 2.6 galapamil 1.88 -9.04879 66.25418 1. 6.4 the Model D-62O 1.25 -9.34918, 48.52653 2 48.9 nicardapine 2.33 -8.8397 65.2O084 1. 2.5 0156 All Runs were made in AMBER 6.0. The computer used was SGI Octane. US 2005/0033522 A1 Feb. 10, 2005 18

0157 SGI Octane information is given below: -continued IOC3 serial port: ty2 IOC3 parallel port: p1p1 1195 MHZ IP3O Processor Graphics board: SI CPU: MIPS R10000 Processor Chip Revision: 2.7 FPU: MIPS R10010 Floating Point Chip Revision: 0.0 Integral Fast Ethernet: efo, version 1, pci 2 Main memory size: 1536 Mbytes Iris Audio Processor: version RAD revision 12.0, number 1 Xbow ASIC: Revision 1.3 Instruction cache size: 32 Kbytes Data cache size: 32 Kbytes Secondary unified instruction/data cache size: 1 Mbyte Integral SCSI controller 0: Version QL1040B (rev. 2), single ended 0158. The Amber runs were made on a potassium channel Disk drive: unit 1 on SCSI controller 0 receptor model that was built using the template Structure of Disk drive: unit 2 on SCSI controller 0 Integral SCSI controller 1: Version QL1040B (rev. 2), single ended PDB entry 1EQ8 on Sybyl 6.8. Amber 6.0 was used to refine IOC3 serial port: ty1 the structure that was built in Sybyl 6.8. Scripts used to do the energy minimization are attached below:

#++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ SCRIPT 1 #------#Script for running Sander Classic in AMBER 6.0 #Relaxing only Hydrogen atoms ----Ravi (May 23, 2002)

#ALL Hs are relaxed IBELLY OPTION, \epsilon(r) #------timlim=36000., imin=1, nimropt=0, ntX=1, irest=0, intrX=1, ntxo=1, intpr=10, intwr=0, intwx=50, ntwv=0, intwe=50, ntwxm=0, intwVm=0, in twen=0, ioutfm=0, intwprt=0, ntf=1, intb=0, idiel=0, dielc=4.0, cut-9.0, intmb=1, msnb=25, ntid=0, scnb=2.0,scee=1.2, cut2nd=0.0 ichdna=0, isftrp=0, rwell=0.0, ipol=0, ibelly=1, intr=0, maxcyc=5000, incyc=550, intmin-1, dx0=0.01, dxm=0.05, &end GROUPNUMBER 1 FIND : H : :

: HA : :

SEARCH RES 1 125 END END #++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ SCRIPT 2 #------#Script for running Sander Classic in AMBER 6.0 #Relaxing Hydrogen + Side-Chains ----Ravi (May 23, 2002) #

#------# Channel ALL H+SC are moving # &cntrl timlim=36000., imin=1, nimropt=0, intX=1, irest=0, intrX=1, ntxo=1, intpr=5, ntwr=0, intwx=50, ntwv=0, intwe=50, ntwxm=0, intwVm=0, intwen=0, ioutfm=0, intwprt=0, ntf=1, intb=0, idiel=0, dielc=4.0, cut-9.0, intmb=1, msnb=25, ntid=0, scnb=2.0,scee=1.2, cut2nd=0.0 ichdna=0, isftrp=0, rwell=0.0, US 2005/0033522 A1 Feb. 10, 2005 19

-continued ipol=0, ibelly=1, intr=0, maxcyc=5000, incyc=250, intmin-1, dx0=0.01, dxm=0.05, &end GROUPNUMBER 1 FIND

* CAB * * CAS * * OHS * * SHS *

: : E : SEARCH RES 1 125 END END #++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ SCRIPT 3: #------#Script for running Sander Classic in AMBER 6.0 #Relaxing everything except alpha-Carbons #----Ravi (May 23, 2002) # #

#---- # Except Alpha C, all other atoms move &cntrl timlim=36000., imin=1, nimropt=0, intX=1, irest=0, intrX=1, ntxo=1, intpr=5, ntwr=0, intwx=50, ntwv=0, intwe=50, ntwxm=0, intwVm=0, intwen=0, ioutfm=0, intwprt=0, intf-1, intb=0, idiel=0, dielc=4.0, cut-9.0, intmb=1, msnb=25, ntid=0, scnb=2.0,scee 1.2, cut2nd=0.0 ichdna=0, isftrp=0, rwell=0.0, ipol=0, ibelly=1, intr=0, maxcyc=5000, incyc=250, intmin-1, dx0=0.01, dxm=0.05, &end GROUPNUMBER 1 FIND : : 3 : : : B : : : S : : : E :

CCM * CH3 CTM *

SEARCH RES 1 125 END END #++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ SCRIPT 4: #------#Script for running Sander Classic in AMBER 6.0 #Restrained minimization of the alpha-Carbons of the channel #----Ravi (May 23, 2002)

#------# Restrained minimization of the alpha-Carbons &cntrl timlim=36000., imin=1, nimropt=0, intX=1, irest=0, intrX=1, ntxo=1, intpr=5, ntwr=0, intwx=50, ntwv=0, intwe=50, ntwxm=0, intwVm=0, intwen=0, ioutfm=0, intwprt=0, intf-1, intb=0, idiel=0, dielc=4.0, cut-9.0, intmb=1, msnb=25, ntid=0, scnb=2.0,scee=1.2, cut2nd=0.0 ichdna=0, US 2005/0033522 A1 Feb. 10, 2005

-continued isftrp=0, rwell=0.0, ipol=0, ibelly=0, intr=1, maxcyc=2000, incyc=250, intmin-1, dx0=0.01, dxm=0.05, &end GROUPNUMBER 1 1.O.O FIND CA : : : SEARCH RES 1 125 END END #++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

0159) Appendix 2 AutoGrid Parameter File -continued nbp r eps 1.90 0.3280000 2 10 # N-H b receptor M3.pdbqs # macromolecule model nbp r eps 1.90 0.3280000 2 10 # N-H b gridfid M3.maps.fld # grid data file sol par 0.000.0000 # N atomic fragmental volume, mpts 60 60 120 # num.grid points in xyz solvation parameters spacing 0.375 # spacing(A) constant 0.000 # N grid map constant energy gridcenter 0.009 0.026 -0.172 # xyz-coordinates or auto map M3.O.map # atom-specific affinity map types CANOH # atom type names smooth 0.5 # store minimum energy Wfin rad(A) nbp r eps 3.60 0.0257202 26 # O-C li map M3.C. map # atom-specific affinity map nbp r eps 3.35 0.0265667 26 # O-N li nbp r eps 4.00 0.0222750 12 6 # C-Cl nbp r eps 3.20 0.0297000 26 # O-O li nbp r eps 3.75 0.023002612 6 # C-N nbp r eps 3.60 0.0297000 26 # O-S li nbp r eps 3.60 0.0257202 12 6 # C-Ol nbp r eps 1.90 0.3280000 2 10 # O-Hib nbp r eps 4.00 0.0257202 12 6 # C-Sl nbp r eps 1.90 0.3280000 2 10 # O-Hib nbp r eps 3.000.008137812 6 # C-H nbp r eps 1.90 0.3280000 2 10 # O-Hib nbp r eps 3.000.0081378 12 6 # C-H nbp r eps 3.000.008137812 6 # C-H sol par 0.000.0000 # O atomic fragmental volume, sol par 12.77 0.6844 # C atomic fragmental volume, solvation parameters constant 0.236 energy S olvation parameters # O grid map constan constant 0.000 # C grid map constan energy map M3.H.map # atom-specific affinity map map M3.A.map # atom-specific affinity map nbp r eps 3.000.0081378 nbp r eps 4.00 0.0222750 12 6 # A-C li nbp r eps 1.90 0.3280000 nbp r eps 3.75 0.023002612 6 # A-N nbp r eps 1.90 0.3280000 nbp r eps 3.60 0.0257202 12 6 # A-Ol nbp r eps 3.000.0093852 nbp r eps 4.00 0.0257202 12 6 # A-S li nbp r eps 3.000.008137812 6 # A-Hl nbp r eps 2.000.0029700 nbp r eps 3.000.008137812 6 # A-Hl nbp r eps 2.000.0029700 nbp r eps 3.000.0081378 12 6 # A-Hl nbp r eps 2.000.0029700 sol par 10.80 0.1027 # A atomic fragmental volume, sol par 0.000.0000 # H atomic fragmental volume, S olvation parameters solvation parameters constant 0.000 # A grid map constan energy constant 0.118 # H grid map constan energy map M3.N.map # atom-specific affinity map elecmap M3.e. map # electrostatic potential map nbp r eps 3.75 0.0230.026 2 6 # N-C dielectric 15.0 # <0, distance-dep.diel;>0, constant nb is 3.5O O.O2376OO 2 6 # N-Nil fmap M3.fmap # floating point potential gridmap nb eps 3.35 0.0265667 2 6 # N-O nb ebs 3.75 O.O265667 26 # N-S li nb eps 1.90 O.328OOOO 2 10 # N-H hb 0160 Appendix 3 AutoDock Parameter File

seed pid time # seeds for random generator types CANOH # atom type names fld M3.maps.fld # grid data file map M3.C. map # atom-specific affinity map map M3.A.map # atom-specific affinity map map M3.N.map # atom-specific affinity map map M3.O.map # atom-specific affinity map map M3.H.map # atom-specific affinity map map M3.e. map # electrostatics map move DMT.out.pdbq # small molecule about -0.088 0.126 0.069 # small molecule center tranO random # initial coordinates/A or random quat) random # initial quaternion US 2005/0033522 A1 Feb. 10, 2005 21

-continued indihe 1 # number of active torsions dihed random # initial dihedrals (relative) or random tstep 2.0 # translation step/A qstep 50.0 # quaternion stepfdeg SO.O # torsion stepfdeg s dof 1 0.3113 # torsional degrees of freedom and coeffiecent bp reps 4.00 0.0222750 12 6 # C-C li bp r eps 4.00 0.0222750 12 6 # C-Ali bp reps 3.75 0.0230026 12 6 # C-N li bp reps 3.60 0.0257202 12 6 # C-O li bp reps 3.000.0081378 12 6 # C-H li bp r eps 4.00 0.0222750 12 6 # A-Ali bp reps 3.75 0.0230026 12 6 # A-N li bp reps 3.60 0.0257202 12 6 # A-O li bp reps 3.000.0081378 12 6 # A-H li bp reps 3.50 0.0237600 12 6 # N-N li bp reps 3.35 0.026566712 6 # N-O li bp reps 2.75 0.0084051 12 6 # N-H li bp reps 3.20 0.0297000 12 6 # O-O li bp reps 2.600.0093852. 12 6 # O-H li intnbp reps 2.000.0029700 12 6 # H-H li outlev 1 # diagnostic Output level rimstol 0.5 # cluster tolerance/A extnrg 1000.0 # external grid energy eOmax O.O 1OOOO # max initial ernergy; max number of retries ga pop size 50 # number of individuals in population ga num evals 5000000 # maximum number of energy evaluations ga num generations 27000 # maximum number of generations ga elitism 1 # number of top individuals to survive to next generation ga mutation rate 0.02 # rate of gene mutation ga crossover rate 0.8 # rate of crossover ga window size 10 # ga cauchy alpha 0.0 # Alpha parameter of Cauchy distribution ga cauchy beta 1.0 # Beta parameter Cauchy distribution Set ga. # set the above parameters for GA or LGA sw max its 300 # iterations of Solis & Wets local search Sw max succ 4 # consecutive successes before changing rho sw max fail 4 # consecutive failures before changing rho sw rho 1.0 # size of local search space to sample sw b rho 0.01 # lower bound on rho ls search freq 0.06 # probability of performing local search on individual set psw1 # set the above pseudo-Solis & Wets parameters ga run 50 # do this many hybrid GA-LS runs analysis # perform a ranked cluster analysis

0.161 Appendix 4: Atomic Coordinates of the Luminal Channel of a Cl3(B4 nAChR Ion Channel

pdb file of the C3f4 nAChr model ATOM 1 CA ACE 9.270 5.413 -18.665 0.00 O.OO ATOM 2 C ACE 9.064 4.649 -17.364 O.OO O.OO ATOM 3 O ACE 9.286 5.198 - 16.285 O.OO O.OO ATOM 4 N GLU 8.656 3.377 -17.484 O.OO O.OO ATOM 5 H GLU 8.496 3.014 -18.412 O.OO O.OO ATOM 6 CA GLU 8.345 2.482 -16.365 0.00 O.OO ATOM 7 CB GLU 7.78 1. 1164 -1693O O.OO O.OO ATOM 8 CG GLU 7.27 4 0.164 -15.874 0.00 O.OO ATOM 9 CD GLU 6.3 1. O.782 - 14.852 O.OO O.OO ATOM 10 OE1 GLU 6.52 5 0.540 -13.643 0.00 O.OO ATOM 11 OE2 GLU 5.38 1. 1.495 - 15.291 O.OO O.OO ATOM 12 C GLU 9.541 2.246 -15.422 0.00 O.OO ATOM 13 O GLU 9.344 1832 -14.284 O.OO O.OO ATOM 14 N LYS 10.771 2.539 -15.863 O.OO O.OO ATOM 15 H. LYS 10.867 2.901 -16.8OO O.OO O.OO ATOM 16 CA LYS 11.990 2.351 -15.083 O.OO O.OO ATOM 17 CB LYS 13.218 2.415 -16.010 0.00 O.OO ATOM 18 CG LYS 13.496 1.12O -16.797 O.OO O.OO ATOM 19 CD LYS 12.434 O.760 -17.851 O.OO O.OO ATOM 20 CE LYS 12.843 -0.456 -18.690 O.00 O.OO

US 2005/0033522 A1 Feb. 10, 2005 51

-continued pdb file of the C3f2 model ATOM 1054 C ILE 124 -O.378 -9.710 16.388 O.OO O.OO ATOM 1055 O ILE 124 -1293 -10.446 16.758 O.OO O.OO ATOM 1056 CB ILE 124 -1.737 -8.470 14.570 O.OO O.OO ATOM 1057 CG1 ILE 124 -1.402 -9.547 13.510 O.OO O.OO ATOM 1058 CG2 ILE 124 -2.030 -7.099 13.925 O.OO O.OO ATOM 1059 CD1 ILE 124 -2.384 -10.723 13496 O.OO O.OO ATOM 1060 H ILE 124 1160 -7.964 14834 1.OO 99.99 ATOM 1061 H ILE 124 1160 -7.964 14834 1.OO 99.99 HETATM 1062 N NME 125 O.912 -1O.O26 16.573 O.OO O.OO HETATM 1063 H NME 125 162O -9.385 16.248 O.OO O.OO HETATM 1064 CA NME 125 1.375 -11.257 17.196 O.OO O.OO CONECT 210 211 212 CONECT 423 424 425 CONECT 635 636 637 CONECT 848 849 850 CONECT 1062 1063 1064 SPDBVT 1.OOOOOOOOOO O.OOOOOOOOOO O.OOOOOOOOOO SPDBVT O.OOOOOOOOOO 1.OOOOOOOOOO O.OOOOOOOOOO SPDBVT O.OOOOOOOOOO O.OOOOOOOOOO 1.OOOOOOOOOO SPDBVT O.OOOOOOOOOO O.OOOOOOOOOO O.OOOOOOOOOO SPDBVT O.OOOOOOOOOO O.OOOOOOOOOO O.OOOOOOOOOO SPDBVV default; SPDBVV 8.228955557850 1111.571236645O71 2O.OOOOOOOOOOOO SPDBVV O.9458,704269 -O.251O394381 -O.2O56899O28 SPDBVV -O.2749012337 -O.9566045536 -O.O96628.3586 SPDBVV -O.1725063688 O.1479423148 -O.973835OO87 SPDBVV -O.254OOOOO81 O.O17OOOOOO9 -0.6704999804 SPDBVV O.OOOOOOOOOO O.OOOOOOOOOO O.OOOOOOOOOO SPDBV 3 3 3 3 33 3 3 3 3 3 3 33 3 3 3 3 3 3 SPDBV 3 3 3 3 33 3 3 3 3 3 3 33 3 3 3 3 3 3 SPDBV 3 3 3 3 33 3 3 3 3 3 3 33 3 3 3 3 3 3 SPDBV 3 3 3 3 33 3 3 3 3 3 3 33 3 3 3 3 3 3 SPDBV 3 3 3 3 33 3 3 3 3 3 3 33 3 3 3 3 3 3 SPDBV 3 3 3 3 33 3 3 3 3 3 3 33 3 3 3 3 3 3 SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBV OO OO OO 1.OOOO OO OO 1.OOOO OO OO OO SPDBVb O.OO O.OO O.2O END US 2005/0033522 A1 Feb. 10, 2005 52

0164)

SEQUENCE LISTING

<160> NUMBER OF SEQ ID NOS : 15 <210> SEQ ID NO 1 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Delta Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 1 Glu Lys Met Ser Thr Ala Ile Ser Val Leu Leu Ala Gly Ala Val Phe 1 5 10 15 Leu Lleu Lieu. Thir Ser Gly Arg 2O

<210> SEQ ID NO 2 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) . . (23) <223> OTHER INFORMATION: Table 1 Gamma Sequence - Transmembrane domain of ligand gated ion channel subunit

<400 SEQUENCE: 2 Gln Lys Cys Thr Leu Ser Ile Ser Val Leu Leu Ala Gln Thr Ile Phe 1 5 10 15

Leu Phe Lieu. Ile Ala Glin Lys 2O

<210> SEQ ID NO 3 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Alpha 1 Sequence - Transmembrane domain of ligand gated ion channel subunit

<400 SEQUENCE: 3 Glu Lys Met Thr Leu Ser Ile Ser Val Leu Leu Ser Leu Thr Val Phe 1 5 10 15

Leu Lieu Val Ile Wall Glu Lieu 2O

<210> SEQ ID NO 4 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Alpha 3 Sequence - Transmembrane domain of ligand gated ion channel subunit US 2005/0033522 A1 Feb. 10, 2005 53

-continued

<400 SEQUENCE: 4 Glu Lys Val Thr Leu Cys Ile Ser Val Leu Leu Ser Leu Thr Val Phe 1 5 10 15

Leu Leu Wall Ile Thr Glu Thr 2O

<210 SEQ ID NO 5 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Alpha 4 Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 5 Glu Lys Ile Thr Leu Cys Ile Ser Val Leu Leu Ser Leu Thr Val Phe 1 5 10 15

Leu Lleu Lieu. Ile Thr Glu Ile 2O

<210> SEQ ID NO 6 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Alpha 5 Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 6 Glu Lys Ile Cys Leu Cys Thr Ser Val Leu Val Ser Leu Thr Val Phe 1 5 10 15

Leu Lieu Val Ile Glu Glu Ile 2O

<210 SEQ ID NO 7 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Alpha 6 Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 7 Glu Lys Val Thr Leu Cys Ile Ser Val Leu Leu Ser Leu Thr Val Phe 1 5 10 15

Leu Leu Wall Ile Thr Glu Thr 2O

<210 SEQ ID NO 8 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Alpha 7 Sequence - Transmembrane domain of ligand gated ion channel subunit US 2005/0033522 A1 Feb. 10, 2005 54

-continued

<400 SEQUENCE: 8 Glu Lys Ile Ser Leu Gly Ile Thr Val Leu Leu Ser Leu Thr Val Phe 1 5 10 15

Met Leu Lieu Wall Ala Glu Ile 2O

<210 SEQ ID NO 9 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Alpha 9 Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 9 Glu Lys Val Ser Leu Gly Val Thr Ile Leu Leu Ala Met Thr Val Phe 1 5 10 15

Gln Leu Met Wall Ala Glu Ile 2O

<210> SEQ ID NO 10 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Alpha 10 Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 10 Glu Lys Val Ser Leu Gly Val Thr Val Leu Leu Ala Leu Thr Val Phe 1 5 10 15

Glin Lieu. Ile Leu Ala Glu Ser 2O

<210> SEQ ID NO 11 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Beta 1 Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 11 Glu Lys Met Gly Leu Ser Ile Phe Ala Leu Leu Thir Leu Thr Val Phe 1 5 10 15 Leu Lleu Lleu Lieu Ala Asp Lys 2O

<210> SEQ ID NO 12 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Beta 2 Sequence - Transmembrane domain of ligand gated ion channel subunit US 2005/0033522 A1 Feb. 10, 2005 55

-continued

<400 SEQUENCE: 12 Glu Lys Met Thr Leu Cys Ile Ser Val Leu Leu Ala Leu Thr Val Phe 1 5 10 15 Leu Lleu Lieu. Ile Ser Lys Ile 2O

<210> SEQ ID NO 13 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Beta 3 Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 13 Glu Lys Leu Ser Leu Ser Thr Ser Val Leu Val Ser Leu Thr Val Phe 1 5 10 15

Leu Lieu Val Ile Glu Glu Ile 2O

<210> SEQ ID NO 14 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Beta 4 Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 14 Glu Lys Met Thr Leu Cys Ile Ser Val Leu Leu Ala Leu Thr Phe Phe 1 5 10 15 Leu Lleu Lieu. Ile Ser Lys Ile 2O

<210 SEQ ID NO 15 &2 11s LENGTH 23 &212> TYPE PRT <213> ORGANISM: Unknown &220s FEATURE <221 NAME/KEY: misc feature <222> LOCATION: (1) ... (23) <223> OTHER INFORMATION: Table 1 Epsilon Sequence - Transmembrane domain of ligand gated ion channel subunit <400 SEQUENCE: 15 Gln Lys Cys Thr Val Ser Ile Asin Val Leu Leu Ala Gln Thr Val Phe 1 5 10 15

Leu Phe Phe Leu. Ile Ala Glin 2O

What is claimed is: ii) determining the energy of the highest occupied 1. A method for identifying a compound as a non- molecular orbital of the compound in eV; channelcompetitive receptor inhibitor comprising: of a ligand-gated neurotransmitter ion iii)- - - determining the area in A of a plane projection of the i) determining the log k" for binding- of the compound to compound; and the ligand-gated neurotransmitter ion channel receptor, iv) identifying as an effective non-competitive inhibitor a or to a subunit thereof; compound exhibiting all of log k from 1.1 to 1.9, an US 2005/0033522 A1 Feb. 10, 2005 56

energy of the highest occupied molecular orbital from a hydrogen bond accepting group Selected from the group –8.6 to -9.2 eV and a YZ shadow of 25 to 50 A. consisting of a keto group, a nitrogen-containing het 2. A method for making a pharmaceuctical composition erocyclic group and a guanidinium group; comprising admixing a compound identified by the method of claim 1 with a pharmaceutically acceptable carrier. a linker joining Said hydrophobic group and Said hydro 3. The method of claim 1, in which k' is determined using gen bond accepting group and comprising 1 to 4 carbon docking of a computational model of the compound to a atoms and optionally containing an oxygen or Sulfur computational model of the luminal channel of the ligand atom, gated neurotransmitter ion channel receptor. 4. The method of claim 1, in which k' is determined by the compound having activity as a non-competitive chromatography of the compound on an affinity matrix inhibitor of Rb" efflux of a ligand-gated neurotransmit comprising at least one Subunit polypeptide of the ligand ter ion channel receptor with an ICso of less than 10 gated neurotransmitter ion channel receptor. tM. 5. The method of claim 1, in which the ligand-gated 12. The compound of claim 11, in which the ring is a neurotransmitter receptor has a Subunit Stoichiometry rang planar, aromatic ring System. ing from (C)(B) to (C)(B), or (C) oy. 13. The compound of claim 11, in which the ring is 6. The method of claim 5, in which the stoichiometry is Selected from the group consisting of a phenyl ring, a (C)(B)s. napthyl ring, morphinan and dibenzo 1.4 diazepine. 7. The method of claim 6, in which the model of the 14. A computer System comprising: luminal channel of the ligand-gated neurotransmitter ion channel receptor has the atomic coordinates of the C3f4 i) a memory storing positional data of the atomic coor Subtype receptor or the C3f2 Subtype receptor shown in dinates of the transmembrane portion of at least one Appendix 4 or AppendiX 5, respectively. Subunit of a ligand-gated neurotransmitter receptor 8. A compound that is a derivative of dextromethorphan protein; and having the nitrogen-bound methyl group Substituted by a C, alkyl group bearing a hydrogen-bond accepting group. ii) a processor generating a molecular model having a 9. The compound of claim 8, in which the hydrogen bond three dimensional shape representative of a luminal accepting group is a keto group, a guanidinium group or a domain portion of the ligand-gated neurotransmitter nitrogen-containing heterocyclic group. receptor having a stochiometry of (C)(B) based the 10. The compound of claim 9, in which the nitrogen positional data. containing heterocyclic group is a pyrollidine, imidazoli 15. The computer system of claim 14, in which the a dine, piperidine, hexahydropyrimidine or pyrimidine group. subunits are C3 Subtype and the B subunits are B2 or 34 11. A compound comprising Subtype. a hydrophobic group comprising a Saturated or unsatur 16. A method for treating Tourette's Syndrome, Schizo ated alkyl chain containing 4 to 10 carbon atoms, a phrenia, a cognitive disorder, pain, anxiety, depression, Saturated hydrocarbon ring containing 5 or 6 carbon neurodegeneration or an addiction caused by an overactive atoms, or at least one ring that includes at least two ligand-gated ion channel receptor, comprising administering conjugated unsaturated bonds, Said ring optionally to a Subject an amount of a compound of claim 11 effective being fused to additional rings to form a ring System to inhibition flux through Said ligand-gated ion channel. and Said additional rings optionally including one or more hetero atoms,