This application is a continuation of application Ser. No. 10/411,336, filed Apr. 11, 2003, which is a continuation-in-part of prior co-pending application Ser. No. 09/852,910, filed May 11, 2001, which claims priority from prior co-pending provisional application Ser. No. 60/275,472, filed Mar. 14, 2001.

BACKGROUND OF THE INVENTION

1. Technical Field

The present invention generally pertains to the field of modulating activity of G protein-coupled receptors (GPCR) and of identifying and preparing G protein coupled receptor antagonist and agonist compounds, including direct, indirect, full, partial, inverse and allosteric agonists. The invention also encompasses compounds that bind to GPCR to stabilize a particular conformation of the GPCR. These compounds can serve as lead compounds for drug discovery purposes or for studying the GPCR three dimensional structure of specific conformations by such methods as X-ray crystallography or NMR. The invention also relates to an approach using high-throughput screening to identify small molecules that can bind to GPCRs and modulate their function by affecting the way in which they contact their cognate G protein(s). As a first step in identifying GPCR modulators, peptide analogs are identified that mimic or antagonize G proteins and bind with high affinity to the particular receptor under study. These peptides then are tested for their specificity. The most specific peptides are used in a competitive assay to screen for small molecules or other peptides that can, for example, (1) increase the binding of the high affinity peptide (â€œsuper agonistâ€ ) or (2) can decrease the binding of the high affinity peptide, presumably by competing for binding at the GPCR (â€œantagonistsâ€ ).

2. Description of the Background Art

A great number of chemical messengers exert their effects on cells by binding to G protein-coupled receptors (GPCR). GPCR include a wide range of biologically active receptors such as hormone receptors, viral receptors, growth factor receptors, chemokine receptors, sensor receptors and neuroreceptors. These receptors are activated by the binding of ligand to an extracellular binding site on the GPCR and mediate their actions through the various G proteins. The molecular interactions that occur between the receptor and the G protein are fundamental to the transduction of environmental signals into specific cellular responses.

G protein-coupled receptors have seven transmembrane helices which form, on the intracellular side of the membrane, the G protein binding domain. Experiments have suggested that activation of the receptor by ligand binding changes conformation of the receptor, unmasking G protein binding sites on the intracellular face of the receptor. The transduction of the signal from the extracellular to intracellular environments requires the actions of heterotrimeric G proteins. The molecular interactions that occur between the receptor and the G protein are fundamental to the transduction of environmental signals into specific cellular responses. Heterotrimeric G proteins are thought to interact with GPCR in a multi-site fashion with the major site of contact being at the carboxyl terminus of the GÎ± subunit. Hamm et al., Science 241:832-835, 1998; Osawa and Weiss, J. Biol. Chem. 270:31052-31058, 1995; Garcia et al., EMBO J. 14:4460-4469, 1995; Sullivan et al., J. Biol. Chem. 269:21519-21525, 1994; West et al., J. Biol. Chem. 260:14428-14430, 1985.

In the inactive state, G proteins are heterotrimeric, consisting of one Î±, one Î² and one Î³ subunit and a bound deoxyguanosine diphosphate (GDP). Following ligand binding, the GPCR becomes activated. Conformational changes in the activated receptor lead to activation of the G protein, with subsequent decreased affinity of GÎ± for GDP, dissociation of the GDP and replacement with GTP. Once GTP is bound, GÎ± assumes its active conformation, dissociates from the receptor, and activates a downstream effector. Hydrolysis of GTP to GDP, catalyzed by the G-protein itself, returns the G-protein to its basal, inactive form. Thus, the G-protein serves a dual role, as both an intermediate that relays the signal from receptor to effector and as a clock that controls the duration of the signal. A variety of studies have implicated the carboxyl terminus of G protein Î± subunits in mediating receptor-G protein interaction and selectivity.

The carboxyl terminal 11 amino acids are most important to receptor interaction and to the specificity of this interaction. Martin et al., J. Biol. Chem. 271:361-366, 1996; Kostenis et al., Biochemistry 36:1487-1495, 1997. Other regions on GÎ± also are involved in receptor contact, however. Portions of the GÎ²Î³ dimer also have been implicated in GPCR binding. See Onrust et al., Science 275:381-384, 1997; Lichtarge et al., Proc. Natl. Acad. Sci. USA 93:7507-7611, 1996; Mazzoni and Hamm, J. Biol. Chem. 271:30034-30040, 1996; Bae et al., J. Biol. Chem. 272:32071-32077, 1997. The carboxyl terminal amino acid regions of GÎ± proteins (and other GPCR binding regions of the heterotrimeric G protein) not only provide the molecular basis of receptor-mediated activation of G proteins, but also play an important role in determining the fidelity of receptor activation. Conklin et al., Nature 363:274-276, 1993; Conklin et al., Mol. Pharmacol. 50:885-890, 1996.

The involvement of the carboxyl-terminal 11 amino acids of Gt (amino acids 340-350) in interactions with the activated GPCR (R*) is suggested by many studies, including (a) the finding that Pertussis toxin catalyzes the ADP-ribosylation of Cys0347, which uncouples Gt from R*; (b) a peptide corresponding to amino acids 340-350 of Gt can uncouple R* from Gt and can itself bind to R* and mimic the effects of Gt; (c) site-directed mutagenesis; and (d) the demonstration in related G proteins that specificity of coupling to particular receptors resides in their carboxyl terminus in interacting with R*.

The G proteins play important and intricate roles in determining the specificity and temporal characteristics of the cellular response to the ligand-binding signal. Hamm and Gilchrist, Curr. Opin. Cell Biol. 8:189-196, 1996. Multiple receptors can activate a single G protein subtype, and in some cases a single receptor can activate more than one G protein, thereby mediating multiple intracellular signals. In other cases, however, interaction of a receptor with a G protein is regulated in a highly selective manner such that only a particular heterotrimer is bound.

Recognition sites are the precise molecular regions on receptors to which the activating molecules bind. An agonist is an endogenous substance or a drug that can interact with a receptor and initiate a physiological response. A drug may interact at the same site as an endogenous agonist (i.e., hormone or neurotransmitter) or at a different site. Agonists that bind to an adjacent or a different site are termed allosteric agonists. As a consequence of the binding to allosteric binding sites, the interaction with the normal ligand may be either enhanced or reduced. The conformational change which the allosteric modulators induce in receptors concerns not only the binding domain for the classical ligands, but also the domain responsible for the interaction between the receptors and the G proteins.

The visual system is an example of one in which G protein signaling is important. Rod cells of the retina make up 95% of the photoreceptors and are highly sensitive to light. Rods allow vision at night or under conditions of very dim illumination. The rod visual protein rhodopsin resides in disk membranes in the rod outer segment (ROS). Rhodopsin is a prototypical GPCR. Helmreich and Hofmann, Biochim. Biophys. Acta 1286:285-322, 1996; Menon et al., Physiol. Rev. 81:1659, 2001; Teller et al., Biochemistry 40:7761, 2001. Rhodopsin is unique among GPCRs as it is not ligand activated.

Night vision relates to the ability of the organism to discriminate between slight differences in the intensity of dim light and, when dark-adapted, to detect small changes in light. Some persons report consistent difficulties in seeing at night, even when their eyes are fully dark-adapted. They cannot detect objects readily visible to others and show both confusion and slow recovery after brief exposure to relatively bright light sources. Maneuvering in dimly illuminated spaces and driving or flying at night present serious problems to these individuals. In addition, some individuals have nyctalopia, or true night blindness, which is diagnosed on the basis of a measurement of retinal sensitivity.

No definitive data on the occurrence of nyctalopia in the population are available, since measurements have never been made on a representative sample of the population. Studies of select groups (e.g., school children, service men), show that the normal population includes a percentage of persons of low visual sensitivity whose performance will be as poor as or poorer than that of many individuals whose nyctalopia is associated with disease or degenerative processes. For example, about 2 percent of Navy men were disqualified for night duties as â€œnight blindâ€ on this basis. It is also a disease of aging. As the general population ages, incidence of night blindness increases. Night blindness also has been observed in several diseases including: (1) Retinitis pigmentosa (In the early stages of the disease, dark adaptation takes place, but at a retarded rate. As disease advances, rod function is progressively lost, and the absolute terminal threshold is elevated. More than 100,000 Americans have retinitis pigmentosa, and most people with retinitis pigmentosa are blind by the age of 40. See Farrar et al., EMBO J. 21(5):857-864, 2002; (2) Glaucoma (Early impairment and progressive loss of rod sensitivity is observed in glaucoma. Cursiefen et al., Doc. Ophthalmol. 103(l):1-12, 2001. Glaucoma is one of the leading causes of blindness in the U.S and one of the most common causes of blindness in individuals over age 60, one of the fastest growing groups in the U.S.); (3) LASIK (Recent studies indicate a significant number of patients who undergo LASIK surgery fail a night vision test (30-60%). Miller et al., CLAO J. 27:84-88, 2001; Brunette et al., Ophthalmology 107:1790-1796, 2000; (4) Side effects of drugs (Several medications can cause night blindness, including Methyltestosterone, Quinidinesis, Paramethadion and Trimethadione (anticonvulsants), Questran (cholesterol-lowering), Accutane (anti-acne), Hydroxychloroquine (anti-malarial), Videx (HIV), and Nefazodone (antidepressant)). Thus, the usefulness of a pharmaceutical approach to night blindness is clear. As the population ages, the number of affected individuals will increase.

Human dietary vitamin A deficiency can cause night blindness, and this can be reversed with vitamin A supplements. However, the night blindness associated with visual diseases such as retinitis pigmentosa (RP), cataracts, diabetic retinopathy, and glaucoma is only somewhat helped with vitamin A supplements, which do not change the course of the disease. Many of the mutations that cause retinal degeneration and visual loss are in genes that encode photoreceptor cascade proteins; others are in genes that encode photoreceptor structural proteins. Pang and Lam, Hum. Mutat. 19:189, 2002. Mutations in rhodopsin, PDEÎ², or GÎ±t have been identified in different forms of congenital stationary night blindness. Pepe, Prog. Retin. Eye Res. 20:733-759, 2001. Stationary night blindness is not associated with retinal degeneration and manifests itself in the inability to see in the dark; daytime vision is largely unaffected. Congenital stationary night blindness (CSNB) refers to a group of non-progressive retinal disorders that are characterized predominantly by abnormal function of the rod system. Clinical heterogeneity even among family members with the same mutation raises the possibility that modifying factors, either genetic or environmental, influence the severity of the disease. Gottlob, Curr. Opin. Ophthalmol. 12:378-383, 2001.

In night blindness resulting from defects in rhodopsin, GÎ±t, or PDEÎ², rod photoreceptors respond only to light intensities far brighter than normal, and the sensitivity of rods to light is similar to that of normal individuals who are not dark adapted. In fundus albipunctatus and in Oguchi disease, the rod photoreceptors can achieve normal sensitivity to dim light but only after 2 or more hours of dark adaptation, compared with approximately 0.5 hours for normal individuals. Dryja, Am. J. Ophthalmol. 130:547, 2000. In each of these forms of stationary night blindness, the poor rod sensitivity and the time course of dark adaptation correlate with the known or presumed physiologic abnormalities caused by the identified gene defects. Increasing the efficacy with which rhodopsin activates the phototransduction cascade is a possible new pharmacological approach to night blindness. Activated rhodopsin activates the rod visual G protein, Gt, which activates the visual transduction cascade. Pharmacologically increasing the effective signaling of rhodopsin can significantly impact people's ability to see and function in low light. The ability, therefore, to identify small molecule compounds that enhance the ability of G protein coupled receptors to signal would be a major benefit.

Because G proteins and their receptors influence a large number of intracellular signals mediated by a large number of different chemical ligands, considerable potential for modulation of disease pathology exists. Many medically significant biological processes are influenced by G protein signal transduction pathways and their downstream effector molecules. See Holler et al., Cell. Mol. Life Sci. 340:1012-1020, 1999. G protein-coupled receptors and their ligands are the target for many pharmaceutical products and are the focus of intense drug discovery efforts. Over the past 15 years, nearly 350 therapeutic agents targeting GPCRs have been successfully introduced into the market. Because of the ubiquitous nature of G protein-mediated signaling systems and their influence on a great number of pathologic states, it is highly desirable to find new methods of modulating these systems, including both agonist and antagonist effects. The ability to study the three-dimensional conformations of GPCRs in response to different individual ligands with different effects also is highly desirable, since these studies would aid in the search and development of drugs with particular structures which impart particular modulating effects on GPCRs.

Drug receptor theories are grounded in the law of mass action and include the concepts of affinity (the probability of the drug occupying a receptor at any given instant), intrinsic efficacy (intrinsic activity), which expresses the complex associations involving drug or ligand concentration, and activation states of receptors. Drugs classified as agonists interact with receptors to alter the proportion of activated receptors, thus modifying cellular activity. Conventional agonists increase the proportion of activated receptors; inverse agonists reduce it. Direct agonists act on receptors, while indirect agonists facilitate the actions of the endogenous agonist (the neurotransmitter itself). Allosteric modulation of receptor activation is a new approach which circumvents the development of tolerance.

Most currently available drugs affecting GPCRs act by antagonizing the binding between a G protein-coupled receptor and its extracellular ligand(s). On the other hand, receptor subtype-selective drugs have been difficult to obtain. An additional drawback to the classical approach of designing drugs to interfere with ligand binding has been that conventional antagonists are ineffective for some GPCRs such as proteinase activated receptors (PAR) due to the unique mechanism of enzymatic cleavage of the receptor and generation of a tethered ligand. In other cases, intrinsic or constitutive activity of receptors leads to pathology directly, thus rendering antagonism of ligand binding moot. For these reasons, alternative targets for blocking the consequences of GPCR activation and signaling are highly desirable. Increased understanding of the structural conformation of GPCRs under the influence of different agonists, antagonists or other ligands also allows design of compounds with highly specific effects on GPCRs.

One potential alternative target for inhibition by new pharmaceuticals has been the receptor-G protein interface on the interior of the plasma membrane. Konig et al., Proc. Natl. Acad. Sci. USA 86:6878-6882, 1989; Acharya et al., J. Biol. Chem. 272:6519-6524, 1997; Verrall et al., J. Biol. Chem. 272:6898-6902, 1997. The carboxyl terminus of GÎ± and other regions of the G protein heterotrimer conform to a binding site at the cytoplasmic face of the receptor. Sondek et al., Nature 379:311-319, 1996; Sondek et al., Nature 379:369-374, 1996; Wall et al., Science 269:1405-1412, 1996; Mixon et al., Science 270:954-960, 1995; Lambright et al., Nature 369:621-628, 1994; Lambright et al., Nature 379:311-319, 1996; Sondek et al., Nature 379:369-374, 1996; Wall et al., Science 269:1405-1412, 1996; Mixon et al., Science 270:954-960, 1995. Peptides corresponding to these binding regions or mimicking these regions can block receptor signaling or stabilize the active agonist-bound conformation of the receptor. Hamm et al., Science 241:832-835, 1988; Gilchrist et al., J. Biol. Chem. 273:14912-14919, 1998.

For example, in the case of rhodopsin, the rod photoreceptor, the GÎ± C-terminal peptide, GÎ± 340-350, stabilizes the receptor in its active metarhodopsin II conformation. Hamm et al., Science 241:832-835, 1988; Osawa and Weiss, J. Biol. Chem. 270:31052-31058, 1995. Two carboxyl terminal peptides from GÎ±S (354-372 and 384-394), but not the corresponding peptides from GÎ±i₂, evoke high affinity agonist binding to Î²₂-adrenergic receptors and inhibit their ability to activate GÎ±s and adenylyl cyclase. Rasenick et al., J. Biol. Chem. 269:21519-21525, 1994. Thus, the carboxyl terminus of GÎ± is important in mediating the specificity of G protein responses. Drug discovery approaches which take advantage of this phenomenon, however, are not available. Jones et al., Expert Opin. Ther. Patents 9(12):1641, 1999.

In general, GPCRs require agonist binding for activation. However, for some receptors basic signaling activity may occur even in the absence of an agonist (constitutive activity). In addition, modifications to the receptor amino acid sequence can stabilize the active state conformation without the requirement for a ligand. Constitutive (agonist-independent) signaling activity has been demonstrated for both mutant and wild type (or native) form receptors (Tiberi and Caron, J. Biol. Chem. 269:27925-27931, 1994; Hasegawa et al., J. Biol. Chem. 271:1857-1860, 1996). A number of GPCRs that cause disease in humans, for example, receptors for thyroid-stimulating hormone (Vassart et al., Ann N.Y. Acad Sci. 766:23-30, 1995), have been found to exhibit agonist-independent activity. An inverse agonist is an agent that binds to the receptor and suppresses this activity.

Experimentally, several single amino acid mutations have produced agonist-independent activity. Î²2 and Î±2 adrenergic receptors, for example, mutated at single sites in the third cytoplasmic loop, show constitutive activity. Ren et al., J. Biol. Chem. 268:16483-16487, 1993; Samama et al., Mol. Pharmacol. 45:390-394, 1994. In some cases, a large deletion mutation in the carboxyl tail or in the intracellular loops of GPCRs has led to constitutive activity. For example, in the thyrotropin releasing hormone receptor a truncation deletion of the carboxyl terminus or a smaller deletion in the second extracellular loop of the thrombin receptor renders the receptor constitutively active. Nussenzveig et al., J. Biol. Chem. 268:2389-2392, 1993; Matus-Leibovitch et al., J. Biol. Chem. 270:1041-1047, 1995; Nanevicz et al., J. Biol. Chem. 270:21619-21625, 1995.

These findings have led to a modification of traditional receptor theory. Samama et al., J. Biol. Chem. 268:4625-4636, 1993. It now is thought that receptors can exist in at least two conformations, an inactive conformation (R) and an activated conformation (R*), and that an equilibrium exists between these two states that markedly favors R over R* in the majority of receptors. It has been proposed that in some receptors (native and mutant) there is a shift in equilibrium in the absence of agonist that allows a sufficient number of receptors to be in the active R* state to initiate signaling. Therefore, in response to chemical or physical external stimuli, GPCRs undergo a conformational change leading to the activation of heterotrimeric G proteins which go on to initiate intracellular signaling events.

Several studies suggest that many GPCRs exhibit properties consistent with the existence of multiple conformational states. In rhodopsin, the existence of multiple conformers is evident from absorbance changes. Sakmar, Prog. Nucleic Acid Res. Mol. Biol. 59:1-34, 1998. Activation occurs by transition through intermediate conformations with the equilibrium between these forms showing a characteristic pH sensitivity. See Armis and Hoffman, Proc. Natl. Acad. Sci. USA 90:7849-7853, 1993; Vogel and Siebert, Biochemistry 41:3529-3535, 2002. Pharmacological studies suggest that the existence of distinct receptor conformers can have functional significance. Studies of fusion proteins of beta adrenergic receptor and G proteins suggest that partial agonists stabilize a conformational state distinct from that stabilized by a full agonist. Seifert et al., J. Pharmacol. Exp. Ther. 297:1218-1226, 2001.

The observation in several receptors that different agonists acting at the same receptor can direct the relative activation of downstream pathways, a phenomenon called â€œsignal trafficking,â€ also suggests the presence of multiple populations of active receptor conformers. Kenakin, Trends Pharmacol. Sci. 16:232-238, 1995; Berg et al., Mol. Pharmacol. 54:94-104, 1998; Cordeaux et al., J. Biol. Chem. 276:28667-28675, 2001; Marie et al., J. Biol. Chem. 276:41100-41111, 2001. Fluorescence studies also suggest the presence of different receptor conformational populations when complexed with functionally distinct agonists. Ghanouni et al., J. Biol. Chem. 276:24433-24436, 2001. This emerging support for the existence of distinct, functionally relevant conformers in several GPCRs suggests that, for these receptors, the molecular activation mechanism must provide the means for switching among multiple conformations. A method to study these conformers by methods such as crystallographic methods and NMR would be highly useful in the process of discovering compounds which can modulate or stabilize particular conformers.

Protein-protein interactions involved in regulatory phenomena are reversible and tend to involve only a small fraction of the protein surface. Generally, to identify peptides that block the protein-protein interactions of interest particular peptides are synthesized in an attempt to mimic sections of one of the native interacting proteins or active sequences are selected from random peptide libraries after screening. Peptides are made up of sequences of amino acids, however unlike DNA recognition, which is linearly coded into the sequence, peptide binding is dependent on three-dimensional structure.

The visual pigment, rhodopsin, is the most extensively studied member of the family of G protein receptors. Recently, the X-ray structure of crystalline bovine rhodopsin has been determined to a resolution of 2.8 â„«. This has paved the way for an understanding of the structure-function relationships of a prototypical GPCR at the molecular level. Since rhodopsin constitutes greater than 90% of the disk membrane protein, measurements made on the proteins of disk membranes predominantly reflect the properties of rhodopsin in its native environment. Rhodopsin consists of the apoprotein opsin and the chromophore 11-cis retinal. Opsin, consisting of 348 amino acids, has a molecular mass of about 40 kDa and folds into seven transmembrane helices of varying length and one short cytoplasmic helix. The retinylidene chromophore (the aldehyde of vitamin A1) is covalently bound to Lys-296 in helix 7 via a protonated Schiff base and keeps the receptor in an inactive conformation.

Light absorption causes a rapid 11-cis to all-trans isomerization of the chromophore which induces a series conformational of changes of the opsin moiety. This reaction occurs with high efficiency (quantum yield 0.67) and the primary photoproduct, photorhodopsin, is formed within a very short time (200 fs). Subsequently, photorhodopsin thermally relaxes within a few picoseconds to a distorted all-trans configuration, bathorhodopsin. On a nanosecond time scale, bathorhodopsin establishes an equilibrium with a blue-shifted intermediate before the mixture decays to form lumirhodopsin. Lumirhodopsin then is transformed into metarhodopsin I and subsequently metarhodopsin II, the active conformation for G protein coupling. Thus, there are two conformational switches in rhodopsin which are controlled by the protonation of specific amino acids of the protein: the transition from the inactive Meta I state to the active Meta II state and, in the absence of bound retinal, the transition from the inactive to the active state of opsin. According to current models, the receptor is kept in an inactive conformation by electrostatic interactions between charged groups in the protein, which are neutralized by the proton uptake involved in the transition to an active state conformation.

The active receptor species Meta II decays slowly within minutes, by hydrolysis of the Schiff base and dissociation of the receptor into the apoprotein opsin and retinal. Researchers have shown that opsin is in a pH-dependent conformational equilibrium between an active and an inactive state. During the decay of Meta II at neutral pH, most structural changes of Meta II formation are reverted and the decay product opsin eventually adopts an active conformation similar to that of Meta II.

Four distinct steps can be observed in the process of GPCR activation: (1) creation of the signal by a photon or by ligand binding; (2) transduction of the signal through the membrane; (3) interaction with the G protein; and (4) activation of the second messenger. Although the phases clearly differ in the kind of processes taking place, they are not discrete and independent. For example, allostery between ligand binding and G protein binding has been observed for several GPCRs, as well as cation-dependent allosteric regulation of agonist and antagonist binding. Wessling-Resnick and Johnson, J. Biol. Chem. 262:12444-12447, 1987; Hepler and Gilman, Trends Biol. Sci. 17:383-387, 1992; Nunnari et al., J. Biol. Chem. 262:12387-12392, 1987; Neve, Mol. Pharmacol., 39:570-578, 1991; Neve et al., Mol. Pharmacol. 39:733-739, 1991.

A number of cytoplasmic proteins interact exclusively with light-activated rhodopsin (R*). Because the crystal structure depicts the inactive form of rhodopsin as not interacting significantly with cytoplamic proteins, this structure can provide only indirect information about the R* state. In addition, two regions of the cytoplasmic surface domain of inactive rhodopsin structure (amino acid residues 236-239 and 328-333) have not been fully resolved by crystal structure analysis. Therefore, tools which can stabilize particular conformers would be useful for studying structure of GPCRs such as rhodopsin.

Negative antagonism is demonstrated when a drug binds to a receptor that exhibits constitutive activity and reduces this activity. Negative antagonists appear to act by constraining receptors in an inactive state. Samama et al., Mol. Pharmacol. 45:390-394, 1994. Although first described in other receptor systems, negative antagonism has been shown to occur with GPCRs such as opioid, Î²2-adrenergic, serotonin type 2C, bradykinin, and D1B dopamine receptors. Schutz and Freissmuth, J. Biol. Chem. 267:8200-8206, 1992; Costa and Herz, Proc. Natl. Acad. Sci. USA 86:7321-7325, 1989; Costa et al., Mol. Pharmacol. 41:549-560, 1992; Samama et al., Mol. Pharmacol. 45:390-394, 1994; Pei et al., Proc. Natl. Acad. Sci. USA 91:2699-2702, 1994; Chidiac et al., Mol. Pharmacol. 45:490-499, 1994; Barker et al., J. Biol. Chem. 269:11687-11690, 1994; Leeb-Lundberg et al., J. Biol. Chem. 269: 25970-25973, 1994; Tiberi and Caron, J. Biol. Chem. 269: 27925-27931, 1994.

That being stated, the concept of constitutively active receptors offer insights which explain pathophysiologic conditions. For example, a constitutively active receptor in a disease process such as hypertension may no longer be under the influence of the sympathetic nervous system. In hypertension, a constitutively active GPCR may be expressed in any number of areas including the brain, kidneys or peripheral blood vessels. A newly recognized class of drugs (negative antagonists or inverse agonists) which reduce undesirable constitutive activity can act as important new therapeutic agents. Thus, a technology for identifying negative antagonists (or understanding and stabilizing the conformational change in a GPCR that binding a negative antagonist compound causes) of both native and mutated GPCRs has important predictable as well as not yet realized pharmaceutical applications. Furthermore, because at least some constitutively active GPCRs are tumorigenic, the identification of negative antagonists for these GPCRs can lead to the development of anti-tumor and/or anti-cell proliferation drugs.

Mutagenesis studies of the carboxyl terminal region of GÎ±t have identified several specific amino acid residues in this binding region crucial for GÎ±t activation by rhodopsin. Martin et al., J. Biol. Chem. 271:361-6, 1996. Substitution of three to five carboxyl-terminal amino acids from GÎ±q with corresponding residues from GÎ±i allowed receptors which signal exclusively through GÎ±i subunits to activate the chimeric Î± subunits and stimulate the GÎ±q effector, phospholipase C Î². Conklin et al., Nature 363:274-276, 1993; Conklin et al., Mol. Pharmacol. 50:885-890, 1996. All of these studies suggest that GÎ± carboxyl peptide sequences are responsible for the specificity of the signaling responses of the individual G proteins. There are 16 unique GÎ± subunits (GÎ±i₁, GÎ±i₂, GÎ±i₃, GÎ±O₁, GÎ±O₂, GÎ±Z, GÎ±t, GÎ±q, GÎ±11, GÎ±14, GÎ±s, GÎ±12, GÎ±13, GÎ±15/16, GÎ±OLF and GÎ±gust) thought to mediate specific interaction with different GPCRs, several hundred of which have been cloned. Thus, peptides corresponding to G protein regions which bind the GPCR could be used as competitive inhibitors of receptor-G protein interactions. Hamm et al., Science 241-832-835, 1988; Gilchrist et al., J. Biol. Chem. 273:14912-14919, 1998. Drug discovery approaches which take advantage of this opportunity, however, are not available. Jones et al., Expert Opin. Ther. Patents 9(12):1641-1654, 1999.

Identification of potent lead compounds for use in modern high throughput screening assays and computerized design of new compounds using information about the desired three-dimensional conformation of receptor molecules, for example, are important aspects of the modern drug discovery process. One of the major challenges confronting those using these types of methods is the difficulty of identifying useful binding compounds from very large combinatorial libraries of potential candidate molecules. When literally hundreds of thousands of compounds are screened, characterizing the compounds which test positive for binding, for modulatory activity or for stabilization of a conformation (including false positives) is an expensive and time-consuming process. Hence, a method which can identify potent and useful lead compounds for high throughput screening and useful binding partners for three dimensional conformational studies and which reduce the number of false positives in the screening process would be very desirable.

SUMMARY OF THE INVENTION

Accordingly, the invention provides a method of identifying a G protein coupled receptor signaling modifying peptide, which comprises providing a peptide library based on a native G protein coupled receptor binding peptide; screening the peptide library for high affinity binding to the G protein coupled receptor; and selecting a member of the peptide library having binding to the G protein coupled receptor of higher affinity than that of the native peptide. The screening may be performed by testing for binding to an intact G protein coupled receptor or to at least an intracellular fragment of a G protein coupled receptor.

The G protein coupled receptor binding peptide may be a G protein subunit or fragment thereof which is, for example from about 7 to about 70 amino acids long or from about 7 to about 55 amino acids long or from about 8 to about 50 amino acids long or from about 9 to about 23 amino acids long, and most preferably about 11 amino acids long. The G protein subunit fragment preferably is a GÎ± subunit or a GÎ± subunit carboxyl terminal peptide but alternatively may be a GÎ²Î³ dimer.

Screening may comprise a competitive binding assay, which preferably is characterized by co-incubation of members of the peptide library with the G protein coupled receptor binding peptide, for example in an enzyme-linked immunosorbant assay wherein the peptide library members are capable of providing a detectable signal and/or wherein binding to the G protein coupled receptor is determined by measuring a signal generated from interaction of an activating ligand with the G protein coupled receptor.

The peptide library preferably is a combinatorial peptide library or a protein-peptide fusion protein library such as, for example a peptide display library or a maltose binding protein-peptide fusion protein library.

In another embodiment, the invention also provides a method of identifying a G protein coupled receptor signaling modifying compound, which comprises providing a library of candidate compounds to screen for binding to the G protein coupled receptor; providing a high affinity G protein coupled receptor binding peptide; screening the library of candidate compounds for high affinity binding to the G protein coupled receptor in competition with the high affinity G protein coupled receptor binding peptide; and identifying a member of the library of candidate compounds having binding to the G protein coupled receptor of equal or higher affinity than that of the high affinity G protein coupled receptor binding peptide or a member of the library of candidate compounds binding of which results in increased binding affinity of the high affinity G protein coupled receptor binding peptide. Screens may be performed by testing for binding to an intact G protein coupled receptor or to at least an intracellular fragment of a G protein coupled receptor.

The G protein coupled receptor binding peptide may be a G protein subunit or fragment thereof which is, for example from about 7 to about 70 amino acids long or from about 7 to about 55 amino acids long or from about 8 to about 50 amino acids long or from about 9 to about 23 amino acids long, and most preferably about 11 amino acids long. The G protein subunit fragment preferably is a GÎ± subunit or a GÎ± subunit carboxyl terminal peptide but alternatively may be a GÎ²Î³ dimer.

Screening may comprise a competitive binding assay, which preferably is characterized by co-incubation of members of the peptide library with the G protein coupled receptor binding peptide, for example in an enzyme-linked immunosorbant assay wherein the peptide library members are capable of providing a detectable signal and/or wherein binding to the G protein coupled receptor is determined by measuring a signal generated from interaction of an activating ligand with the G protein coupled receptor.

The library of candidate compounds preferably is a focused library of candidate compounds based on the structure of the high affinity G protein coupled receptor binding peptide. The library of candidate compounds may be a combinatorial library of, for example drug-like molecules or a focused small molecule library whose members, for example may be based on the chemical structure of the high affinity G protein coupled receptor binding peptide.

The invention also provides G protein coupled receptor signaling modifying peptides and compounds identified according to the methods described above, as well as methods of modifying G protein coupled receptor signaling in a cell having a G protein coupled receptor which comprise administering such compounds to the cell. Also provided are methods of inhibiting G protein coupled receptor signaling which comprise contacting a compound with the G protein coupled receptor which interferes with binding of the G protein coupled receptor to its cognate G proteins.

In a further embodiment, the invention provides a method for identifying a G protein coupled receptor signaling modifying compound, which comprises providing a peptide identified according to at least one of the methods described above, wherein the peptide is labeled to provide a detectable peptide signal; providing a library of candidate G protein coupled receptor signaling modifying compounds; contacting the peptide with the G protein coupled receptor under conditions such that the peptide binds to the G protein coupled receptor; removing unbound peptide from the G protein coupled receptor; measuring the signaling activity of the peptide-bound G protein coupled receptor and measuring the detectable peptide signal; contacting the members of the library of candidate G protein coupled receptor signaling modifying compounds with the peptide-bound G protein coupled receptor; measuring the signaling activity of the peptide bound G protein coupled receptor and measuring the detectable peptide signal; determining whether the G protein coupled receptor signaling activity is increased or decreased after contact with the candidate compound and whether G protein coupled receptor peptide binding is increased or decreased after contact with the candidate compound; and identifying compounds for which contact with the peptide-bound G protein coupled receptor results in both an increase in peptide binding to the G protein coupled receptor and an increase in G protein coupled receptor signaling, identifying compounds for which contact with the peptide-bound G protein coupled receptor results in both a decrease in peptide binding to the G protein coupled receptor and a decrease a G protein coupled receptor signaling and identifying compounds for which contact with the peptide-bound G-protein coupled receptor results in increased binding affinity of the peptide identified according to a method described above. Methods for measuring the signaling activity of the peptide-bound G protein coupled receptor may be selected from the group consisting of measuring inositol phosphate accumulation; measuring intracellular Ca²⁺ levels; measuring adenyl cyclase activity; measuring transendothelial electrical resistance; measuring stress fiber formation; measuring ligand binding; measuring receptor expression; measuring receptor desensitization; measuring kinase activity; measuring phosphatase activity; measuring nuclear transcription factors; measuring all migration (chemotaxis); measuring superoxide formation; measuring nitric oxide formation; measuring cell degranulation; measuring GIRK activity; measuring actin polymerization; measuring vasoconstriction; measuring cell permeability; measuring apoptosis; measuring cell differentiation; measuring membrane association of a protein that translocates upon GPCR activation, such as protein kinase C; measuring cytosolic accumulation of a protein that translocates upon GPCR activation, such as protein kinase C; and measuring nuclear association of a protein that translocates upon GPCR activation, such as Ran.

In yet a further embodiment, the invention provides compounds selected from the group consisting of SEQ ID NOS: 2, 4, 6, 8, 10, 12, 13, 15, 17, 21, 23, 25-27, 30, 32, 34, 36, 38, 40, 45-85, 94-111, 125-150, 160-164, 175-178 and 183-264.

In yet a further embodiment, the invention provides a method for providing a therapeutic G protein coupled receptor signaling modifier peptide to a mammal which comprises administering to the mammal an expression construct which expresses a compound selected from the group consisting of SEQ ID NOS: 2, 4, 6, 8, 10, 12, 13, 15, 17, 21, 23, 25-27, 30, 32, 34, 36, 38, 40, 45-85, 94-111, 125-150, 160-164, 175-178 and 183-264. Further, the invention provides a method for treating a disease state in which excess G protein coupled receptor signaling is a causative factor, which comprises administering a compound selected from the group consisting of SEQ ID NOS: 2, 4, 6, 8, 10, 12, 13, 15, 17, 21, 23, 25-27, 30, 32, 34, 36, 38, 40, 45-85, 94-111, 125-150, 160-164, 175-178 and 183-264.

In yet a further embodiment, the invention provides a method of identifying a G protein coupled receptor signaling enhancer, which comprises providing a peptide library based on a native G protein coupled receptor binding peptide; screening the peptide library for high affinity binding to the G protein coupled receptor; selecting a member of the peptide library having binding to the G protein coupled receptor of higher affinity than that of the native peptide; providing a library of candidate compounds to screen for binding to the G protein coupled receptor; screening the library of candidate compounds for high affinity binding to the G protein coupled receptor in competition with a member of the peptide library selected above; and identifying a member of the library of candidate compounds having binding to the G protein coupled receptor of equal or higher affinity than that of the peptide selected above or identifying a member of the library of candidate compounds binding of which results in increased binding affinity of the peptide selected above. Screening methods for use in this embodiment may include testing for binding to an intact G protein coupled receptor or testing for binding to at least an intracellular fragment of a G protein coupled receptor. The G protein coupled receptor binding peptide may be a G protein subunit or fragment thereof, for example a G protein subunit fragment from about 7 to about 70 amino acids long, from about 7 to about 55 amino acids long, from about 8 to about 50 amino acids long, from about 9 to about 23 amino acids long or most preferably about 11 amino acids long. The G protein subunit fragment preferably is a GÎ± subunit or a GÎ± subunit carboxyl terminal peptide but alternatively may be a GÎ²Î³ dimer.

Screening may comprise a competitive binding assay, which preferably is characterized by co-incubation of members of the peptide library with the G protein coupled receptor binding peptide, for example in an enzyme-linked immunosorbant assay wherein the peptide library members are capable of providing a detectable signal and/or wherein binding to the G protein coupled receptor is determined by measuring a signal generated from interaction of an activating ligand with the G protein coupled receptor.

The library of candidate compounds preferably is a focused library of candidate compounds based on the structure of the high affinity G protein coupled receptor binding peptide. The library of candidate compounds may be a combinatorial library of, for example drug-like molecules or a focused small molecule library whose members, for example may be based on the chemical structure of the high affinity G protein coupled receptor binding peptide.

Enzyme-linked immunosorbant assays for use in the inventive method may comprise the steps of immobilizing the G protein coupled receptor onto a solid support; providing a protein-peptide fusion protein display library; incubating members of the protein-peptide fusion protein display library with the immobilized G protein coupled receptor in the presence of the G protein coupled receptor binding peptide under conditions such that members of protein-peptide fusion protein display library having a binding affinity for the G protein coupled receptor at least as high as the G protein coupled receptor binding peptide bind to the immobilized G protein coupled receptor; removing unbound members of the protein-peptide fusion protein display library; incubating the bound protein-peptide fusion protein display library with antibodies which specifically recognize the protein portion of the protein-peptide fusion protein display library members under conditions such that the antibodies specifically bind to the protein-peptide fusion protein display library members; removing unbound antibodies; and detecting the bound antibodies. The protein-peptide fusion protein display library preferably is a maltose binding protein-peptide fusion protein display library and the antibodies preferably are anti-maltose binding protein antibodies. Binding to the G protein coupled receptor preferably is determined by measuring a signal generated from interaction of the signalling enhancer with the G protein coupled receptor.

The peptide library preferably is a combinatorial peptide library, for example a protein-peptide fusion protein library such as a maltose binding protein-peptide fusion protein library or any suitable peptide display library. Libraries of candidate compounds preferably are focused libraries of candidate compounds based on the structure of the compound selected above as having binding to the G protein coupled receptor of higher affinity than that of the native peptide. The library may be a peptide library or a small molecule library.

In yet a further embodiment, the invention provides compounds identified by a method as described above. In yet further embodiments, the invention provides a method for treating a disease state in which alterations in G protein coupled receptor signaling is a causative factor and a method for treating a disease state in which alterations in G protein coupled receptor signaling is a causative factor both of which comprise administering these compounds. In yet a further embodiment the invention provides a method of determining the three-dimensional structure of a G protein coupled receptor, which comprises contacting the G protein coupled receptor with a compound identified by at least one of the methods described above under conditions such that binding occurs and a conformation of the G protein coupled receptor is stabilized; co-crystallizing the G protein coupled receptor-compound binding pair; subjecting the co-crystallized binding pair to X-ray crystallography; and determining the three-dimensional structure of the co-crystallized binding pair, wherein atomic coordinates of the G protein coupled receptor are obtained. In yet a further embodiment, the invention provides a method of determining the three-dimensional structure of a G protein coupled receptor, which comprises contacting the G protein coupled receptor with a compound identified by at least one of the methods described above under conditions such that binding occurs and a conformation of the G protein coupled receptor is stabilized; subjecting the binding pair to nuclear magnetic resonance study; and determining the three-dimensional structure of the binding pair, wherein atomic coordinates of the G protein coupled receptor are obtained.

In yet a further embodiment, the invention provides a method of isolating a G protein coupled receptor binding partner, which comprises providing a solid support comprising bound compound identified by at least one of the methods described above; providing a library of candidate G protein coupled receptor binding partner compounds; contacting the library of candidate compounds with the solid support under conditions such that binding of the candidate compounds to the compound occurs; eluting unbound and nonspecifically bound candidate compounds from the solid support; and recovering bound candidate compounds from the solid support.

In yet a further embodiment, the invention provides a method of designing small molecules that modify activation of a G protein coupled receptor, which comprises determining the three-dimensional structure of a G protein coupled receptor according to at least one of the methods described above; and designing candidate structures by computer modeling based on the atomic coordinates, wherein the candidate structures are predicted to bind to the G protein coupled receptor.

In yet a further embodiment, the invention provides a nucleic acid which comprises a DNA that encodes a peptide identified by at least one of the methods described above, wherein the DNA is operably linked to a heterologous transcriptional regulatory sequence, an expression vector which comprises this nucleic acid and a cell transfected with the expression vector. The invention also provides an antibody that specifically recognizes a peptide identified by any of the methods described above, such as, for example, a monoclonal antibody, a polyclonal antibody, a humanized antibody or a single chain antibody.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGS. 1A, 1B and 1C are schematic diagrams showing the basic two-step platform.

FIGS. 2A and 2B are schematic diagrams showing an example of the basis for the affinity screening method used to separate and identify GPCR binding peptides.

FIG. 3 is a schematic diagram of vector pJS142.

FIG. 4 is a schematic diagram showing an ELISA procedure.

FIG. 5 provides results showing that the LacI-Gq fusion protein binds thrombin receptor in a concentration-dependent manner.

FIG. 6 shows data from binding assays performed on some of the clones identified using peptide 8 in the screening process.

FIG. 7 provides binding data for LacI peptide fusion proteins to PAR1 receptor. pELM6 is the MBP vector alone; pELM17 is the MBP-native Gt340-350 peptide fusion protein.

FIG. 8 is a bar graph comparing binding of high affinity fusion proteins to the high affinity peptide 8 fusion protein (MBP 8).

FIG. 9 presents ELISA results from panning CHO cells overexpressing human thrombin receptor (PAR1) using purified MBP-C-terminal fusion proteins. MBP-G11=xxxx (SEQ ID NO: 1) LQLNLKEYNLV (SEQ ID NO: 2); PAR-13=VRPS (SEQ ID NO: 3) LQLNRNEYYLV (SEQ ID NO: 4); PAR-23=LSRS (SEQ ID NO: 5) LQQKLKEYSLV (SEQ ID NO:6); PAR-33=LSTN (SEQ ID NO: 7) LHLNLKEYNLV (SEQ ID NO: 8); PAR-34=LPQM (SEQ ID NO: 9) QRLNVGEYNLV (SEQ ID NO: 10); PAR-45=SRHT (SEQ ID NO: 11) LRLNGKELNLV (SEQ ID NO:12).

FIG. 10 presents a dose-response curve of SF9 membranes (PAR1 receptor) assayed with lacI-Gq lysates.

FIG. 11 is a concentration response curve demonstrating binding of native Gq peptide-maltose binding protein to PAR1 reconstituted in lipid vesicles.

FIG. 12B is a schematic diagram showing an exemplary cDNA minigene construct. SEQ ID NOS:270 and 271 are shown in FIG. 12A.

FIG. 13 shows an agarose gel of a NcoI digest of minigene vector. Lane 1 is a 1 kb DNA ladder; lane 2 is pcDNA 3.1; lane 3 is pcDNA-GÎ±i; lane 4 is pcDNA-GÎ±iR; and lane 5 is pcDNA-GÎ±q.

FIG. 14 shows an agarose gel of PCR products showing transcription of peptide minigene RNA in transfected cells. Lane 1 contains size markers, lane 2 contains PCR products from cells transfected with pcDNA-GiR, lane 3 contains PCR products from cells transfected with pcDNA-Gi, and lane 4 contains PCR products from cells transfected with pcDNA3.1, the empty vector.

FIG. 15 is a bar graph showing the relative [³H] inositol phosphate production after thrombin stimulation normalized against the basal value.

FIG. 16 presents data showing inhibition of a GPCR mediated increase in intracellular calcium concentration in the presence or absence of a minigene vector encoding the identified high affinity peptide. FIG. 16A presents fluorescence ([Ca⁺⁺]_ilevel) increase 30 seconds after thrombin addition. FIG. 16B shows the kinetics of [Ca⁺⁺] fluorescence changes after cell stimulation with thrombin.

FIG. 17 presents data showing inhibition of a GPCR-mediated phosphoinositol (P1) hydrolysis in the presence or absence of a minigene vector encoding the identified high affinity peptide.

FIG. 18 is a bar graph indicating relative GPCR-mediated increase of MAPK activity in the presence or absence of a minigene vector encoding the identified high affinity peptide in cells expressing GPCR-binding peptides.

FIG. 19 shows reduction of thrombin-induced transendothelial electrical resistance in cells expressing GÎ±q, GÎ±i, GÎ±iR or empty vector.

FIGS. 20A, 20B, 20C and 20D are a series of photographs of cells stained for F-actin, showing the inhibition of stress fiber formation after exposure to thrombin in cells expressing pcDNA-G12 or pcDNA-G13 minigene construct.

FIGS. 21A, 21B and 21C are bar graphs showing acetylcholine (Ach) response (pA/pF) for HEK 293 cells transiently transfected with GIRK1/GIRK4 and the indicated minigene construct.

FIG. 22 demonstrates selective G protein mediated adenylyl cyclase inhibition in cells expressing minigene constructs containing GÎ± carboxyl terminal peptide inserts.

FIG. 23 presents dose-response curves of MII stabilization by Î±t340350, mutant Î±t340-350K341L and heterotrimeric Gt.

FIG. 24 shows stabilization of MII by small molecule PL_â€”0302R3C4.

FIG. 25 presents fluorescence data showing super agonists for rhodopsin have no effect on PAR1-stimulated Ca²⁺ transients.

FIGS. 26A and 26B are graphs showing light responses (as measured by a change in current) from isolated rods of dark-adapted salamander retinas in the presence of small molecule PL_â€”0302R3C4.

FIGS. 27A and 27B are graphs showing light responses (as measured by a change in current) from isolated rods of dark-adapted salamander retinas in the absence of a small molecule.

FIG. 28B is an MBP-8 binding curve with added compound PL_â€”1012R2C1, the structure of which is depicted in FIG. 28A, showing the compound's ability to enhance MBP-8 binding to EDTA-washed rhodopsin.

FIG. 29B is an MBP-8 binding curve with added compound PL_â€”0894R3C7, the structure of which is depicted in FIG. 29A, showing the compound's ability to enhance MBP-8 binding to EDTA-washed rhodopsin.

FIG. 30B is an MBP-8 binding curve with added compound PL_â€”0568R1C5, the structure of which is depicted in FIG. 30A, showing the compound's ability to enhance MBP-8 binding to EDTA-washed rhodopsin.

FIG. 31B is an MBP-8 binding curve with added compound PL_â€”0551R8C1, the structure of which is depicted in FIG. 31A, showing the compound's ability to enhance MBP-8 binding to EDTA-washed rhodopsin.

FIG. 32B is an MBP-8 binding curve with added compound PL_â€”0302R3C4, the structure of which is depicted in FIG. 32A, showing the compound's ability to enhance MBP-8 binding to EDTA-washed rhodopsin.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention involves a method of identifying compounds which can interfere with or increase binding between a G protein-coupled receptor (GPCR) and its cognate G protein(s) and compounds which stabilize a particular conformation of a GPCR for conformational study. These compounds modulate G protein-mediated signaling and thus can be used as pharmaceuticals, as lead compounds for identification of potential useful drugs, as components of assays which identify drug candidates or as binding partners in conformational studies by known methods, such as for example X-ray crystallography or nuclear magnetic resonance.

Methods for screening and drug identification use peptides that mimic the structure of the GPCR binding regions of G proteins and thus are able to modulate receptor-G protein interactions specifically or specifically bind to a given receptor with high affinity. These high affinity peptides can be delivered into cells in the context of an expression construct to act as blockers or agonists of specific receptor-mediated cellular responses in vitro and in vivo or can be administered directly to a patient. The peptides also form the basis of a screening, identification and selection process to provide traditional pharmaceutical compounds or to study structure-function relationships in binding. In particular, the invention allows one to identify high affinity analog peptides that block or mimic compounds at the receptor-G protein interface for a particular G protein and to use these high affinity analogs in a high throughput screen to identify other peptides or small molecules that likewise specifically antagonize or agonize GPCR signaling for a G protein or class of G proteins.

Small molecules can be used in analogous high throughput screening processes to identify further compounds. â€œSmall moleculeâ€ denotes any non-peptide organic compound which binds or interferes with binding to the interfacial region of a GPCR or is a candidate for such action. Such molecules that bind to and stabilize a particular conformer of a GPCR also are included in the definition of â€œsmall moleculeâ€ as used herein. Peptides or small molecules directed at the receptor-G protein interface can be designed using the inventive method to inhibit or enhance biological processes that employ signaling through a GPCR or to bind to and stabilize a particular GPCR conformer. Such compounds which bind to, interfere into binding to or stabilize a conformer of the GPCR-G protein interface (including but not limited to agonists, inverse agonists, allosteric agonists, blockers, antagonists, inhibitors, negative antagonists, partial agonists, and enhancers, as well as compounds which bind to and stabilize a particular conformer) are termed â€œmodifiersâ€ or â€œmodifyingâ€ compounds, and may include both peptides and small molecules. This approach to drug design is useful in targeting G protein-GPCR interactions for which there are no available ligands, orphan receptors the ligands of which are not known, mutant constitutively activated receptors, antibody-crosslinked irreversibly activated receptors such as TSH receptors in Graves' Disease, and proteinase activated receptors (PAR). It works equally well, however, with any GPCR-G protein interaction and more broadly, with receptor-protein interactions in general.

Because the method is useful for identifying high affinity compounds that can bind to and enhance or inhibit virtually any GPCR, the approach is useful in identifying compounds which can prevent, ameliorate or correct dysfunctions or diseases in which a specific class of G proteins is relevant. Conditions and disease states for which GPCR enhancers and inhibitors are useful include, but are not limited to: stroke; myocardial infarction; restenosis; atherosclerosis; hypotension; hypertension; angina pectoris; acute heart failure; cardiomyocyte apoptosis; cancers; infections such as bacterial, fungal, protozoan and viral infections, and particularly infections caused by HIV-1 or HIV-2; septic shock; pain; chronic allergic disorders; asthma; inflammatory bowel disease; osteoporosis; rheumatoid arthritis; Graves' disease; post-operative ileus; urinary retention; testotoxicosis; ulcers; obesity; benign prostatic hypertrophy; and psychotic and neurological disorders including anxiety, epilepsy, schizophrenia, manic depression, Parkinson's disease, Alzheimer's disease, delirium, dementia, drug addiction, anorexia, bulimia, mood disorders and sleep disorders; smoking cessation and any other disease or condition that can be treated by G protein coupled receptor inhibition and any other disease or condition that can be treated by G protein coupled receptor activation. Treatment of this diverse set of disorders is possible because the receptors to which various G proteins bind differ enough to allow the creation of a battery of analog peptides which can specifically interface with different GPCR or different classes or groups of GPCR. The relationship of G proteins and G protein signalling to various diseases and conditions such as those listed above is known in the art.

With the inventive screening methods, the sequences identified in a particular screen do not bind to all receptors, but only to the particular receptor of interest. The interaction between a G protein and a GPCR is quite specific. For example, a difference in one amino acid can substantially reduce or eliminate the ability of the GÎ±i_1/2peptide to bind the A1 adenosine G protein coupled receptor-G protein interface. Gilchrist et al., J. Biol. Chem. 273:14912-14919, 1998. Both upstream regulation of GTP/GDP exchange on G proteins and G protein-mediated effector activation may be inhibited with interfacial binding compounds. Thus, high affinity analog peptides can be designed to specifically interfere with or stabilize a particular action of one GPCR. Likewise agonizing or enhancing peptides also specifically affect one GPCR. These specifically-acting peptide analogs are useful both as pharmaceutical compounds per se, and as potent lead compounds in modern high throughput screens for other peptides and small molecule binders having the same specific GPCR interaction.

The inventive methods, in one embodiment, rely as first step on screening for small molecules that enhance or inhibit the ability of the receptor to interact with the heterotrimeric G protein. Using rhodopsin and transducin, the screen has found small molecules that significantly enhance rhodopsin's ability to form MII, the active form of rhodopsin. Such small molecules can serve as lead compounds in drug discovery efforts directed towards potential therapeutic agents to combat night blindness. Using the screen to identify small molecules, and then testing the identified compounds using in vitro and in vivo analysis will result in discovery of potent, high affinity compounds.

This invention therefore can be used to identify small molecules that enhance the ability of rhodopsin to signal. The inventive methods involve, in one embodiment, screening compound libraries to discover more molecules that increase binding of G protein peptides to activated rhodopsin. The methods also include testing these molecules in a number of assays to determine their effects on rhodopsin signaling, including MII stabilization, guanosine 5â€²-O-(3-[³⁵S]thio) triphosphate (GTPÎ³S) binding, and 3â€²,5â€² cyclic GMP phosphodiesterase (PDE) activation.

Additionally, the methods involve testing the small molecules for their specificity by measuring their effects using another GPCR, for example, human thrombin receptor (PAR1), which also has been shown to couple to Gt. Seibert et al., Vision Res. 42:517,1999. Enhancement of the sensitivity of vision in vivo can be tested according to a method of the invention using electroretinography (ERG) of wild type and mutant mice. Chang et al., Vision Res. 42:517-525, 2002. The inventive methods also optionally involve optimizing the chemical structure of enhancers or antagonists, performing pharmacokinetic, toxicological and metabolism studies of the discovered chemical entities, and large animal efficacy studies, and clinical trials for a pharmacological treatment for night blindness. Therefore, the methods of the invention can be used, for example, to identify small molecules that enhance the binding of the high affinity Gt peptides to light activated rhodopsin, determine whether the small molecules enhance rhodopsin signaling in vitro, determine if the small molecules are specific for rhodopsin, or if they can enhance other GPCR-G protein signaling events, and test the small molecules in a mouse model for stationary night blindness for increased sensitivity of vision as measured by ERG.

A first step was to identify peptides with sequences based on the C-terminus of Gt that bind with high affinity to either light-activated or dark-adapted rhodopsin. These peptide analogs were then tested for their specificity to binding to rhodopsin versus other GPCRs, as well as their ability to stabilize the MII conformation of the receptor. The methods of the invention also identify small molecules that bind light-activated rhodopsin and by doing so enhance the binding of the high affinity peptide analogs. The binding affinity (EC₅₀) of the compound is the first criterion of a successful drug candidate. The identified small molecules are tested in vitro for their ability to enhance rhodopsin signaling using assays such as MII stabilization and MII decay, GTPÎ³S binding, and PDE activation.

Upon activation, rhodopsin undergoes a conformational change that allows its interaction with and activation of Gt, leading ultimately to the stimulation of PDE. The binding of Gt to light-activated rhodopsin induces a high affinity receptor state that can be measured spectrophotometrically by stabilization of the active, signaling metarhodopsin II state of the receptor. Using a split-beam SLM Aminco DW2000 spectrophotometer, for example, one can determine if the receptor undergoes proper conformational changes following light activation. This assay shows the small molecule acting on the conformation of the receptor. If rhodopsin's active intermediate, metarhodopsin II (MII), is stabilized by the presence of the small molecule, the activation energy of the receptor is lowered. Using the inventive assay system, compounds were identified that allow the receptor to enter the active, MII conformation without hetereotrimeric G protein, which normally is required. The â€œenhancersâ€ stabilize the active (signaling) conformation of the receptor. â€œInhibitorsâ€ block the binding of transducin to rhodopsin and thus inhibit the receptor from entering the proper conformation even in the presence of agonist (light) and G protein.

Metarhodopsin II decay can be used to examine the differences in compound potencies are due to changes in MII decay. It could be postulated that differences could be due to effects of a compounds to non-specifically attach the retinal Schiff's base linkage of MII. Thus, one can compare the time dependent MII decay in the presence of the individual compounds. In the process of receptor activation, the GÎ±t subunit binds a GTP molecule. GTPÎ³S binding assays can determine the ability of the receptors to signal, with an increase in GTPÎ³S binding indicating receptor-mediated release of the GDP from the Î± subunit and subsequent binding of GTP. Conversion of inactive transducin (Gtâ€¢GDP) to the active state (Gtâ€¢GDP) is accompanied by dissociation of the GÎ± from GBÎ³. The free GÎ±tâ€¢GTP then activates cGMP phosphodiesterase (PDE) by binding to and dissociating its two inhibitory Î³ subunits. As a result, the released catalyzing Î± and Î² subunits of activated PDE (PDE*) can convert cGMP to GMP. Therefore, compounds which affect rhodopsin signaling can be tested for their affects on PDE assays.

Generally, small molecules that display an appropriate dose curve when used to compete off the high affinity peptide fusion proteins, with a resulting EC₅₀<100 Î¼M for binding to rhodopsin are suitable for continued study and are tested for the ability to stabilize MII. Preferably, those with an EC₅₀<10 Î¼M for MII stabilization are analyzed further. Further analysis may include thermal stability of rhodopsin in the presence of the small molecules (MII decay), and GTPÎ³S binding (an assay for the small molecule's effects on function). The rate of GTPÎ³S-binding is controlled by a rate-limiting GDP release of GÎ± subunits. Native GÎ±t, in the presence or absence of GÎ²Î³t, displays very slow intrinsic rates of GDP release. Therefore, an increase in guanosine 5â€²-O-(3-[³⁵S]thio) triphosphate(GTPÎ³S) binding indicates receptor mediated release of the GDP from the Î± subunit and subsequent binding of the GTP.

GTPÎ³S binding assays may be performed as follows or using any method known in the art. GÎ± subunits (1 mM) alone, or GÎ± subunits in the presence of the small molecules to be assayed are mixed with 2 mM GÎ²Î³t or 2 mm GÎ²Î³t and urea-washed ROS membranes (500 nM rhodopsin) and incubated for 3 min at 25Â° C. Binding reactions may be started by addition of 5 mm [³⁵S]GTPÎ³S (0.1 mCi). Aliquots of 50 Î¼l are withdrawn at several timepoints, mixed with 1 ml ice-cold 20 mm Tris.2HCl (pH 8.0) buffer containing 130 mM NaCl and 10 mM MgSO₄and passed through Whatmanâ„¢ cellulose nitrate filters (0.45 mm). The filters are washed three times with the same buffer (3 ml, ice-cold) and counted in a liquid scintillation counter after dissolution in 3a70B mixture. See Skiba et al., J. Biol. Chem. 271: 413-424, 1996 for exemplary methods which may be used with the invention. The skilled person will recognize variations and adjustments which may be made to the assay, and such variations are considered within the scope of this invention. The k_appvalues for the binding reactions may be calculated by fitting the data to the equation, GTPÎ³S bound (% bound)=100% (1âˆ’e^âˆ’kt). The small molecule(s) also may be tested for the ability to affect PDE activation. GtÎ± binding to PDEy relieves the inhibitory effect of the gamma subunit on the catalytic Î± and Î² subunits of PDE and allows the hydrolytic activity of these subunits to be increased almost 300 fold.

Activation of GtÎ± by rhodopsin can be monitored in the presence or absence of the small molecules using fluorescence spectroscopy at 20Â° C. as described by Cerione, Methods Enzymol. 237:409-423, 1994. This assay measures the GÎ±:GTPÎ³S (complex between a subunit of transducin and GTPÎ³S) formation rate catalyzed by wild-type rhodopsin upon illumination. The excitation wavelength is 295 nm (2 nm bandwidth), and fluorescence emission is monitored at 340 nm (12 nm bandwidth). Briefly, rhodopsin (40 nM) is added to a solution of Gt (250 nM) in a reaction mixture containing 10 mM Tris (pH 7.2), 2 mM MgCl₂, 100 mM NaCl, 1 mM DTT, and 0.01% n-dodecyl Î²-maltoside. The solution is stirred for 300 sec to equilibrate. GTPÎ³S (5 Î¼M) is added to the reaction mixture to a final concentration of 5 Î¼M, and the increase in fluorescence is followed for and additional 2000 sec. To calculate the activation rates, the slopes of the initial fluorescence increase after GTPÎ³S addition were determined through the data points covering the first 60 sec. The values in the presence of the small molecules may be normalized to the value obtained for wild-type rhodopsin with no compounds taken as 1.00. Those molecules which appear to be acting directly on rhodopsin in these assays, or variations on these assays readily apparent to the person of skill in the art are taken to the next level of testing. The small molecules also are assayed for the ability to modulate rhodopsin-transducin signaling specifically without affecting processes mediated by other GPCRs.

Preferred small molecule â€œenhancersâ€ and â€œinhibitorsâ€ are uniquely specific, not only for the receptor, but for the receptor-G protein interaction. As there are over 1000 GPCRs, and no simple way to determine the effect of compounds on each and every one of them individually, a few select and representative GPCR signaling systems may be tested. Functional coupling of the human thrombin receptor (PAR1) with Gt has been demonstrated. Seibert, Eur. J. Biochem. 266(3):911-916, 1999. Testing for effects on PAR1 may include determining if the small molecule(s) have an effect on thrombin-mediated signal transduction events such as adenylyl cyclase activity, calcium influx, and inositol phosphate accumulation. Other tests for functional coupling to PAR1 or other GPCRs are known in the art and may be used as well.

Adenylyl cyclase activity may be measured in a final volume of 50 Î¼l with [Î±-³²P]ATP (1 mM; 120-400 cpm/pmol) as the substrate and [2,8-³H]cAMP (2 mM; 200,000 cpm/pmol) to monitor recovery in an assay mixture containing 5 mM MgCl₂, 1 mM EDTA, 1 mM 2-mercaptoethanol, 100 Î¼M papaverine, 1 pg/ml bovine serum albumin, and an ATP-regenerating system consisting of 20 mM creatine phosphate and 120 units creatine phosphokinase/ml in 25 mM Tris-HCl buffer, pH 7.5. The concentration of ATP and cAMP may be determined spectrophotometrically at 259 nm, using Îµ values of 15.4 and 14.6 mM¹cm^âˆ’1, respectively. The assay may be initiated by addition of protein and after a 10 min incubation at 37Â° C. the reaction is stopped with 2 volumes stop solution (2% SDS/45 mM ATP/13 mM cAMP). The samples then may be heated (e.g. to 100Â° C. for 3 minutes) and the formed cyclic [Î±-³²P]AMP recovered. See Gilchrist et al., J. Biol. Chem . 276:25672-25679, 2001, the disclosures of which are hereby incorporated by reference. Finally, the compounds passing the previous steps may be tested in an animal model of night blindness as described by Chang et al., Vision Res. 42:517-525, 2002, the disclosures of which are hereby incorporated by reference.

To assess the effects of small molecules on photoresponses in an in vivo system, electroretinography of mice exposed to the small molecule(s) may be used to measure the amplitude of both the a- and b-waves. Plots of the amplitude against the logarithm of relative light intensities indicate if the compounds are affecting only rod signaling. The sensitivity for eliciting a threshold b wave within normal limits also may be measured. Mutant animals also may be tested to observe not only the effects of enhancers on wild type rhodopsin signaling, but also the effects on animals with night blindness.

The small molecules to be tested may be dissolved in sterile PBS and administered as eye drops. Experiments may be repeated using IV or IM injections if initial results are negative. The most promising candidates undergo the steps needed to take them from an identified compound to a lead compound. This approach identifies a pharmacological treatment for night blindness which circumvents the need for more invasive procedures such as gene therapy, laser ablation and retinal replacement.

Mapping the sites of interaction between proteins involves identifying parts of the interface between two proteins using synthetic peptides corresponding to interfacial regions. The peptides are identified because they act as competitive inhibitors of the interaction. NMR studies of peptide structures in their bound conformation using trNOESY, combined with analysis of activity of substituted peptide analogs to define the minimal structural requirements for interaction were used to understand the structural basis of rhodopsin-transducin interaction as well as G protein-effector interaction. Peptides corresponding to the C-terminus of Gt can be used to stabilize rhodopsin in its active conformation (MII) or in an inactive conformation. The 3-dimensional structures of heterotrimeric G proteins reveals that the last 7 amino acids of G protein Î± subunits are unstructured, indicating that this region of the Î± subunit is critical for binding to the cytoplasmic surface of an activated receptor with induced fit. This interaction is quite specific since a difference in a single amino acid can affect the affinity by 1000 fold.

High throughput screening is a recent technology that has been developed primarily within the pharmaceutical industry. It has emerged in response to the profusion of new biological targets and the need of the pharmaceutical industry to generate novel drugs rapidly in a changed commercial environment. Its development has been aided by the invention of new instrumentation, by new assay procedures, and by the availability of databases that allow huge numbers of data points to be managed effectively. High throughput screening combined with combinatorial chemistry, rational design, and automation of laboratory procedures has led to a significantly accelerated drug discovery process compared to the traditional one-compound-at-a-time approach. Screens may be performed manually, however robotic screening of the compound libraries is preferred as a time- and labor-saving device.

One critical aspect of the drug discovery process is the identification of potent lead compounds. A purely random selection of compounds for testing is unlikely to yield many active compounds against a given receptor. Typically, pharmaceutical companies screen 100,000 or more compounds per screen to identify approximately 100 potential lead compounds. On average, only one or two of these compounds actually produce lead compound series. Therefore, companies have been assaying larger and larger data sets in the search for useful compounds. Compound accessibility then becomes an issue: historical compound collections are limited in size and availability. In contrast, large combinatorial chemistry libraries can be synthesized on demand, but at significant technical difficulty and cost. As the library sizes expand, the difficulty becomes selecting the desired compounds from these very large combinatorial libraries. When literally hundreds of thousands of compounds are screened, it makes characterizing the candidate lead compounds an expensive and time-consuming process, particularly when many of the â€œhitsâ€ turn out to be false positives.

The multi-step approach to the drug discovery process described here provides a solution to many of these problems. One embodiment of this invention takes advantage of the properties of G protein Î± subunit carboxyl termini to identify peptides which act as high affinity, competitive inhibitors or agonists of G protein/GPCR interactions. The method, however, can be used with any specific protein-protein, protein-small molecule, protein-nucleic acid interaction or the like. In addition, peptides based on any region of a GÎ± subunit or any region of a GÎ²Î³ dimer which is involved in GPCR binding may be used in the same way. Many such GPCR binding regions are known in the art. The identification of high affinity competitors forms a first step in a screening and selection method which overcomes many of the disadvantages of high throughput screening by providing specific, high affinity lead compounds against which to test potentially useful pharmaceuticals. Because peptides selected by this method have affinity for their binding partner up to 1,000 times higher or more than the native protein, this step is one key to successfully screening and identifying useful pharmaceutical compounds.

A subsequent step of the preferred process involves high throughput screening of candidate peptide or small molecule pharmaceutical compounds against the high affinity lead peptides identified in the first step. Because the lead peptide compounds are potent and specific binders to the desired receptor, screening assays test for compounds that can decrease the binding of the peptide (â€œblockersâ€ ) or that increase the binding of the peptide (â€œagonistsâ€ ). The assay system allows one to measure both binding and function simultaneously as the peptides all serve to mimic a required step, that of specific receptor-G protein binding. By using this site, the system facilitates identification of those candidate compounds which bind not only with useful affinities (nM to Î¼M range) but by the very virtue of their selection process will affect function by either increasing or decreasing the G protein binding. The high throughput screening step of the drug discovery process is thereby greatly simplified, because the number of false positive compounds, and compounds which are identified as binders but which bind only with low affinity, is reduced or virtually eliminated. Only those compounds with a high chance of success will be identified by the screen, therefore useful compounds can be identified directly and there are many fewer compounds which need to be characterized and further studied to confirm that the compounds are specific, potent pharmaceutical compounds. In addition, the method identifies a compound through binding directly to the precise site of interest, so that the mechanism of binding and the mechanism of action of the newly identified pharmaceutical compound does not have to be discovered and confirmed later by a separate process.

The identified high affinity peptides also may be used according to the inventions to identify GPCR inverse agonists. High affinity peptides identified in a first step of the inventive method bind the receptor and stabilize it in an active or R* conformation. Screens which are used to identify potent agonists seek out compounds which can compete with this binding and also stabilize the GPCR in its R* state. Inverse agonists, on the other hand, stabilize the GPCR in an inactive or R state. Therefore, screens designed to detect dissociation of the high affinity peptide or a decrease in its affinity for the GPCR are used to identify inverse agonists.

Although this description provides examples relative to the interaction between a G protein coupled receptor and its cognate GÎ± protein, the methodology can be used to identify peptide inhibitors of most protein-protein interactions, specifically including any interaction between a GPCR and any region of a GÎ± or GÎ±Î³ G protein subunit. The high affinity peptides selected by this method may be used in high throughput screening to identify small molecules that can be used as modulators of a variety of specific biological process.

To produce very high affinity peptide GPCR blockers, the tertiary structure of a wild-type GÎ± carboxyl terminal peptide or any other GPCR binding peptide in its receptor-bound conformation may be studied, for example, using trNOESY (NMR). Dratz et al., Nature 363:276-280, 1993. Structural data derived from these types of studies of G protein regions are combined with analysis of activity of substituted peptide analogs to define the minimal structural requirements for interaction of peptides or any ligand with GPCR. The following experimental systems are examples of systems which can be used to define receptor-G protein interactions: (i) rhodopsin-transducin (GÎ±t) in retinal rod cells, (ii) Î²-adrenergic receptor-GÎ±s in C6 glioma cells, (iii) adenosine A1 receptor-GÎ±1 in Chinese hamster ovary cells, (iv) GABA_Breceptors-GÎ±1 in rat hippocampal CA1 pyramidal neurons, (v) muscarinic M2 receptor-GÎ±1 in human embryonic kidney cells, and the like. Any GPCR or group of GPCR which is convenient or desired can be used to define the interaction requirements, and skilled workers are aware of many methods to understand structure-activity relationships in receptor binding of this kind. Any of these methods are contemplated for use in these methods and may substitute for the particular methods of the exemplified embodiment.

The plasmid display method provides an efficient means of identifying specific and potent peptides that can serve as competitive inhibitors of protein-protein interactions. Using the information gleaned from structure-activity studies, a library of variant peptides encoding sequences related to a GPCR-binding region, for example the GÎ± subunit carboxyl terminus, for each of the classes of the GÎ± subtypes or GÎ²Î³ can be prepared. Exemplary native sequences upon which libraries may be based include those listed in Table III, below. Libraries advantageously contain peptides with computer-generated random substitutions within the sequence, and allow one to test a large number of peptide sequences at one time. Preferably, peptide sequences in each library are constructed such that approximately 50% of the amino acid residues are identical to the native GPCR binding region and the remaining amino acid residues are randomly selected from any amino acid. The peptides may range in size from about 7 to about 55 amino acid residues or from about 8 to about 50 amino acids long or from about 7 to about 70 amino acid residues or longer, preferably from about 9 to about 23 amino acid residues or about 9 to about 15 amino acid residues. Undecamer peptides are most preferred. Libraries may be constructed in which about 10% to about 90% of the amino acid residues unchanged from the native sequence; however, about 30% to about 70% unchanged is preferred and about 50% is most preferred.

Alternatively, a synthetic peptide library can be based on any protein known to interact with a GPCR, using randomly created overlapping regions of the protein. The peptides may be about 7-70 amino acids long or about 8-50 amino acids long or preferably about 9 to about 23 or about 9 to about 15 amino acids long and most preferably about 11 amino acids long. Oligonucleotides encoding the peptides advantageously may be cloned to the 3â€² end of the LacI gene, with a linker sequence at the N-terminus of the peptide. The linker sequence is not mandatory for successful screening, but is generally preferred. Restriction enzyme sites may be placed at either end of the peptide coding sequence for cloning purposes. See Table I below for a schematic representation of a peptide library and an example of one peptide. Additional peptides which also can be used are shown in Tables II and III, below. The oligonucleotides encoding the actual peptide sequences are synthesized with 70% of the correct base and 10% each of the remaining bases, leading to a biased peptide library with an approximately 50% chance of having the correct amino acid at any specific position along the peptide sequence. Different ratios of bases may be used to achieve the desired mutagenesis rate at any particular position in the sequence.

TABLE IExample for Construction of a Synthetic Peptide Library.(SEQ ID NO:13)â€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒQâ€ƒâ€ƒRâ€ƒâ€ƒMâ€ƒâ€ƒHâ€ƒâ€ƒLâ€ƒâ€ƒRâ€ƒâ€ƒQâ€ƒâ€ƒYâ€ƒâ€ƒEâ€ƒâ€ƒLâ€ƒâ€ƒLgaggtggt nnknnknnknnk attcgtgaaaacttaaaagattgtggtcgtttc taa ctaagtaaagcâ€ƒâ€ƒâ€ƒAâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒBâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒCâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒDâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒâ€ƒE
(SEQ ID NO:14) n =â€ƒany nucleotide base; k =â€ƒguanidine or thymidine; A =â€ƒrestriction enzyme site; B =â€ƒlinker sequence; C =â€ƒoligonucleotide encoding peptide sequence; D =â€ƒstop codon; E =â€ƒrestriction enzyme site.

TABLE IIGÎ±â€ƒSubunit Peptides and CorrespondingDNA Constructs.GÎ±SEQSubunitSequenceID NO:GtIKENLKDCGLF15atcaaggagaacctgaaagactgcggcctcttc16 Gi1/2IKNNLKDCGLF17ataaaaaat aatctaaaagattgtggtctcttc18 GRi1/2NGIKCLFNDKL19aacggcatcaagtgcctcttcaacgacaagctg20 Gi3IKNNLKECGLY21attaaaaacaacttaaaggaatgtggactttat22 Go2IAKNLRGCGLY23atcgccaaaaacctgcggggctgtggactctac24 Go1IANNLRGCGLY25attgccaacaacctccggggctgcggcttgtac26 GzIQNNLKYIGLC27atacagaacaatctcaagtacattggcctttgc28 G11LQLNLKEYNLV2ctgcagctgaacctcaaggagtacaacctggtc29 GqLQLNLKEYNAV30ctccagttgaacctgaaggagtacaatgcagtc31 GolfQRNHLKQYELL32cagcggatgcacctcaagcagtatgagctcttg33 G14LQLNLREFNLV34ctacagctaaacctaagggaattcaaccttgtc35 G15/16LARYLDEINLL36ctcgcccgctacctggacgagatcaacctgctg37 G12LQENLKDIMLQ38ctgcaggagaacctgaaggacatcatgctgcag39 G13LNDNLKQLMLQ40ctgcatgacaacctcaagcagcttatgctacag41 GsQRNHLRQYELL13cagcgcatgcaccttcgtcagtacgagctgctc42 5â€²- gatccgccgccaccatggga--tgaa-3â€²
(SEQ ID NOS:43, 44)

TABLE IIIExemplary Native C Protein Sequences forLibrary/Minigene Construction.*SEQSEQIDIDNameSequenceNO:NameSequenceNO:hGtIKENLKDCGLF15CryptoGbalLQNALRDSGIL62 hGi1/2IKNNLKDCGLF17GA3_USTLTNALKDSGIL63 G05_DROIKNNLKQIGLF45GA1_KLUIQONLKKSGIL64 GAF_DROLSENVSSMGLF46GA3_USTLTNALKDSGIL63 Gi-DROIKNNLKQIGLF45GA1_DICNLTLGEAGMIL64 hGi3IKNNLKECGLY21GA2_KLULENSLKDSGVL65 hGO-1IANNLRGCGLY25GA2_USTILTNNLRDIVL66 hGO-2IAKNLRGCGLY47Mgs-XLQRMHLPQYELL67 GAK_CAVIKNNLKECGLY21hGsQRMHLRQYELL13 GO_XENIAYNLRGCGLY48hGolfQRMHLKGYELL68 GA3_CAEELIQANLQGCGLY49GA1_COPCOLQLHLRECGLL69 GA2_CAEELIQSNLHKSGLY50GA1-SOLRRRNLFEAGLL70 GA1_CAEELLSTKLKGCGLY51GA2_SBRRRNLLEAGLL71 GAK_XENIKSNLMECGLY52GA1_SBRRRNPLEAGLL72 GA1_CANVQQNLKKSGIM53GA1_USTIQVNLRDCGLL73 hGZIQNNLKYIGLC27GA4_USTRENLKLTGLVG74 hG15LARYLDEINLL36GA1_ORYSADESMRRSREGT75 GA2_SCHPOLQHSLKEAGMF54GQ1_DROMEMQNALKEFNLG76 hG12LQENLKDIMLQ38GA2_DICTQCVMKAGLYS77 hG13LHDNLKQLMLQ40GS-SCHLQHSLKEAGMF54 GAL_DROLQRNLNALMLQ55GA-SACENTLKDSGVLQ56 GA2_YSTENTLKDSGVLQ56GA1-CEIISASLKMVGV78 hG14LQLNLREFNLV34GA2-CENENLRSAGLHE79 hG11LQLNLKEYNLV2GA3-CERLIRYANNIPV80 hGQLQLNLKEYNAV30GA4-CELSTKLKGGGLY51 GQ_DROMELQSNLKEYNLV57GA5-CEIAKNLKSMGLC81 G11_XENLQHNLKEYNLV58GA6-CEIGRNLRGTGME82 Gq_SPOSCIQENLRLCGLI59GA7-CEIQHTMQKVGIQ83 GA1_YSTIQQNLKKIGII60GA8-CEIQKNLQKAGMM84 GA1_NEUCRIIQRNLKQLIL61GAS-DICLKNIFNTIINY85
*For production of minigene constructs each nucleotide sequence should be constructed to encode the amino acids MG at the N-terminus of the peptide by using 5â€²-gatccgccgccaccatggga-(SEQ ID NO:43) and -tgaa-3â€²â€ƒ(SEQ ID NO:44).

The peptides advantageously are synthesized in a display system for convenience and efficiency of performing the binding reactions. For example, plasmid or phage display systems, as are known in the art, may be employed. While peptide display systems are preferred, any method which allows efficient contact of the peptides with a GPCR and determination of binding may be used.

A peptide display (â€œpeptides on plasmidsâ€ ) library is a convenient system for use with this invention which exploits the high affinity bond between LacI and lacO. The â€œpeptides on plasmidsâ€ display is preferred for use with this invention for two major reasons. The technique is easily set up in the laboratory. In addition, the fusion of the peptide at the carboxyl terminus of the presenting protein mimics the normal presentation for carboxyl terminal peptides during the screen. If amino terminal or interior peptides are being tested, the peptide may be cloned at the appropriate position to mimic native presentation.

The â€œpeptides on plasmidsâ€ method for testing carboxyl terminal peptides generally works as follows. Persons of skill in the art will be able to modify these methods as needed to accommodate different conditions using this general description and the examples below as a guide. A library of peptides is created by degenerate PCR based on the native GPCR-binding peptide of interest and fused to the carboxyl terminus of LacI. The peptide library is expressed via a plasmid vector carrying the fusion gene. The plasmid also contains the Lac operon (LacO), and when E. coli transcribes and translates the Lacl fusion protein, it binds back as a tetramer to the encoding plasmid through its lacO DNA binding sequence, displaying the inserted sequences of interest on the plasmid. Following transcription and translation, variant peptides encoding different sequences related to the native peptide sequence therefore are displayed as carboxyl terminal extensions of the lacl gene. Thus, a stable LacI-peptide-plasmid complex is formed which can be screened for binding to receptor. Methods described in Gates et al., J. Mol. Biol. 255:373-386, 1996, the disclosures of which are hereby incorporated by reference, are suitable. See Examples 7 and 9 for exemplary methods.

The E. coli strain used to display the peptides was ARI814, which has the following genotype: Î”(srl-recA) endA1 nupG lon-11 sulA1 hsdR17 Î” (ompT-fepC)266 Î”clpA319::kan Î”lacI lacZU118. The strain contains the hsdR17 allele that prevents restriction of unmodified DNA introduced by transformation or transduction. The ompT-fepC deletion removes the gene encoding the OmpT protease, which digests peptides between paired basic residues. The lon-11 and clpA mutations also limit proteolysis by ATP-dependent, cytoplasmic proteases. The deletion of the lacI gene prevents expression of the wild-type lac repressor, which would compete with the fusion constructs for binding to the lacO sites on the plasmid. The lacZ mutation prevents waste of the cell's metabolic resources to make Î²-galactosidase in the absence of the repressor. The endA1 mutation eliminates a nuclease that has deleterious effects on affinity purification and the recA deletion prevents multimerization of plasmids through RecA-catalyzed homologous recombination. This strain was selected also for its robust growth properties and high yields of immunocompetent cells. Transformation efficiencies of 2Ã—10¹⁰colonies per mg DNA typically were achieved. Although this strain of E. coli is preferred, those of skill in the art are aware of many alternatives which are convenient for use with the methods described. Therefore, any suitable and convenient bacterial strain known in the art is contemplated for use with this invention.

The LacI-peptide fusion protein library may be released from the bacteria by gentle enzymatic digestion of the cell wall using lysozyme. After pelleting the cell debris, the lysate then can be added directly to immobilized receptor for affinity purification or used without purification. The display library of these peptides is screened to identify those peptides which bind with high affinity to a particular GPCR. In this way, it is possible to screen for and identify high affinity peptides which bind GPCR and can interfere with or enhance activation of the pre-selected specific G protein. The library can be screened against any desired GPCR. Since the combinatorial library contains peptides based on a particular GÎ± or GÎ²Î³ subunit, any GPCR which binds to or mediates signaling through that subunit or class of subunits can be used. Multiple libraries, based on the carboxyl terminal sequences or other regions of different G protein subunits may be constructed for screening the same or different GPCR.

To screen the plasmid display library, a G protein coupled receptor of interest advantageously may be immobilized on microtiter plates for screening by ELISA. A plasmid preparation (bacterial lysate) then may be added to the wells. This screening procedure, involving allowing the peptides displayed on the library plasmids to bind receptor, is sometimes referred to as â€œpanning.â€ Sequences that bind the receptor stick to the well so that non-binding sequences can be removed by a washing step. The adherent plasmids then can be expanded and used to transform E. coli. The â€œpanningâ€ process generally is repeated 2 to 8 times. In general, however, 3 to 4 sequential screens are sufficient and preferred. In the later rounds of panning, parent peptide (wild type sequence) preferably is co-incubated with the plasmid preparation to bind receptors and serve as a competitive inhibitor. In this way, only high affinity sequences on the display library are captured by the immobilized receptor. The same competitive inhibition advantageously may be performed using a high affinity peptide or small molecule which has already been identified, rather than the native peptide. See FIG. 1 for a schematic diagram generally describing the â€œpanningâ€ procedure and Example 7 for a specific embodiment. The selection process in this embodiment preferably is carried out in low salt buffers because high salt concentrations destabilize the Lacl-lacO complex, and could lead to peptides becoming associated with the incorrect plasmid. For the same reason, the panning buffers preferably contain lactose, which causes the Lacl to bind more tightly to lacO.

The selection process of this invention allows the identification of peptide sequences with higher and higher affinity binding with each round of panning. For example, diversity in an unpanned library may look like the sequences given in Table IV, below, i.e. highly randomized. After successive rounds of selection, the selected adherent peptides would look more like those given in Table V, below.

TABLE IVDiversity in Unpanned Gq Library.SEQ. ID NO.NativeLQLNLKEYNLV2 clone #1LLLQLVEHTLV86 clone #2HRLNLLEYCLV87 clone #3EQWNMNTFHMI88 clone #4SQVKLQKGHLV89 clone #5LRLLL*EYNLG90 clone #6RRLKVNEYKLL91 clone #7LQLRLREHNLV92 clone #8HVLNSKEYNQV93

TABLE VSelection in Panned GÎ±ll Library.SEQ ID NO.NativeLQLNLKEYNLV2Round 11MKLNVSESNLV942LQTNQKEYDMD953LQLNPREDKLW964RHLDLNACNMG975LR*NDIEALLV986LVQDRQESILV99 Round 21LQLKHKENNLM1002LQVNLEEYHLV1013LQFNLNDCNLV1024MKLKLKEDNLV1035HQLDLLEYNLG1046LRLDFSEKQLV105 Round 31LQKNLKEYNMV1062LQYNLMEDYLN1073LQMYLRGYNLV1084LPLNPKEYSLV1095MNLTLKECNLV1106LQQSLIEYNLL111

LacI is normally a tetramer and the minimum functional DNA binding species is a dimer. Thus, the peptides are displayed multivalently on the fusion protein, leading to binding to the immobilized receptor in a cooperative fashion. This cooperative binding permits the detection of binding events of quite low intrinsic affinity. The sensitivity of the assay is an advantage in that initial hits of low affinity can be identified, but the disadvantage is that the signal in the ELISA does not necessarily correlate with the intrinsic affinity of the bound peptides.

One preferred ELISA, where signal strength is better correlated with affinity, involves fusing the sequences of interest from a population of clones in frame with a gene encoding a protein, for example E. coli maltose binding protein (MBP). Once the sequences have been transferred into the monomeric fusion protein, they can be overexpressed in E. coli and used as either crude lysates or purified fusion proteins for assay by an ELISA which detects the protein bound to receptor or any convenient assay. Controls having the vector alone which expresses TGGG linker only fused to MBP, or having Gt:340-350K341R peptide fused to MBP may be used, if desired. Frozen cell stocks preferably are kept in 25% glycerol at âˆ’80Â° C. The high affinity GÎ± peptides fused to MBP preferably are analyzed by ELISA, where the resulting signal correlates to the peptide's affinity for light-activated rhodopsin. The MBP-peptide fusions are expressed and purified over an amylose affinity column and used to measure the relative affinities of peptides of interest. Those samples with an absorbance of at least two standard deviations above background may be considered to contain high affinity binding peptides. Any desired cut-off point may be used, however, depending on the assay parameters and the needs of the operator.

A suitable ELISA may be performed as follows, however those of skill in the art will be able to modify the techniques for the conditions in their assays. Serial dilutions of MBP-peptide fusion proteins are added to 96-well plates with immobilized light-activated rhodopsin previously blocked with 0.1% BSA. After 1 hour at 37Â° C., the wells are washed with phosphate buffered saline (PBS)/0.l% Tween 20, and probed with an anti-MBP antibody, followed by a goat-anti-rabbit antibody conjugated to horseradish peroxidase. Color development of the assay is allowed to proceed for 20 minutes, after which the A₄₅₀is measured on a microtiter plate reader. See Gilchrist et al., Methods Enzymol. 315:388-404, 2000, the disclosures of which are hereby incorporated by reference.

The purified fusion proteins can be further tested by measuring their ability to compete for the site of binding on the receptor using native peptide, a LacI-peptide fusion protein, or heterotrimeric G protein. Use of competitive ELISA allows one to calculate IC₅₀values for the binding of individual fusion protein to the immobilized receptor.

Peptide fusion proteins can be analyzed in a competitive ELISA format using a fusion protein co-incubation to prevent the binding of lower affinity peptide fusion proteins to the GPCR. Any convenient protein which does not interfere with peptide binding may be used, including for example, glutathione-S-transferase, green fluorescent protein or ubiquitin, however a maltose binding protein fusion protein such as MB-GÎ±_t340-35OK341R is preferred. Competitive ELISA indicates which peptide sequences have the highest affinity for light activated rhodopsin. Several different assay formats are suitable. For example, synthetic Gt:340-350K341R peptide may be used to compete with the MBP fusion proteins containing the GÎ± high affinity peptides for binding. In addition, MBP fusion proteins containing the GÎ± high affinity peptides may be used to compete with LacI-Gt:340-350K341R peptide fusion protein for binding to light-activated rhodopsin. Recombinant heterotrimeric Gt also may be tested against the high affinity peptides. The relative affinity of the variant peptides may be assessed using an ELISA format where a constant concentration of MBP-GÎ± peptide fusion proteins is competed by serial dilutions of native peptide, LacI-GÎ±t peptide fusion protein or recombinant heterotrimeric Gt. The wells advantageously may be probed with an anti-MBP antibody to measure the amount of MBP-GÎ± peptide fusion protein remaining bound. The dose-response curves may be analyzed by non-linear regression to calculate an EC₅₀.

Cloning the library into pJS142 creates a BspEI restriction site near the beginning of the random coding region of the library. Conveniently, digestion with BspEI and ScaI allows the purification of a 900 base pair DNA fragment that may be subcloned into pELM3, a vector that directs the MBP fusion protein to the cytoplasm, a reducing environment. Alternatively, the fragment can be cloned into pELM15, a vector which directs the MBP fusion protein to the periplasm, an oxidizing environment. pELM3 and pELM15 are simple modifications of the pMALc2 and pMALp2 vectors, respectively, available commercially (New England Biolabs). Digestion of pELM3 with AgeI and ScaI allows efficient cloning of the BspEI-ScaI fragment from the pJS142 library. Any suitable method may be used which is convenient to achieve the desired result. Modifications of these methods are well known by those of skill in the art of molecular biology and are contemplated for use here.

Proof that the high affinity peptides competitively bind to GPCR and interfere with or enhance its recognition of G protein can be obtained using a competitive binding assay in the presence of a heterotrimeric G protein. For example, if rhodopsin is the GPCR used in the screen, heterotrimeric G protein, transducin (Gt) may be used. Gt binds rhodopsin with multiple epitopes and is membrane-bound via myristoylation of the Î± subunit and farnesylation of the Î³ subunit carboxyl terminus. Poor binding by carboxyl terminal native peptide LacI constructs and/or heterotrimeric Gt indicates high affinity binding of the MBP-peptide fusion proteins. An analogous strategy of panning, peptide synthesis and binding studies may be employed for determining high affinity peptides that bind any GPCR, for example the thrombin receptors (PAR1, PAR3, PAR4), dopamine receptors (D1, D2, D3, D4, D5), vasopressin receptors (V1a, V1b, V2) and histamine receptors (H1, H2, H3), using carboxyl terminal peptide libraries for any GÎ± subunit, for example GÎ±i, GÎ±s and GÎ±q. Once peptide analogs with higher binding affinities have been elucidated, they can be exploited to inhibit GPCR-G protein interaction.

The peptides selected by this method, characterized by high affinity, specific blockade of or enhancement of a desired GPCR-mediated signaling event, may be used as therapeutic agents such as traditional pharmaceuticals or gene therapies to treat disorders which would benefit by modifying GPCR activity or used to screen additional libraries of compounds able to compete with the high affinity peptide analogs or to modulate (i.e., increase or decrease) the binding affinity of the high affinity peptide analogs or the high affinity peptide analog-fusion proteins.

Any method known in the art for selecting and synthesizing small molecule libraries for screening is contemplated for use in this invention. Small molecules to be screened are advantageously collected in the form of a combinatorial library. For example, libraries of drug-like small molecules, such as Î²-turn mimetic libraries and the like, may be purchased from for example ChemDiv, Pharmacopia or Combichem or synthesized and are described in Tietze and Lieb, Curr. Opin. Chem. Biol. 2:363-371, 1998; Carrell et al., Chem Biol. 2:171-183, 1995; U.S. Pat. Nos. 5,880,972, 6,087,186 and 6,184,223, the disclosures of which are hereby incorporated by reference.

Any of these libraries known in the art are suitable for screening, as are random libraries or individual compounds. In general, hydrophilic compounds are preferred because they are more easily soluble, more easily synthesized, and more easily compounded. Compounds having an average molecular weight of about 500 often are most useful, however, compounds outside this range, or even far outside this range also may be used. Generally, compounds having c logP scores of about 5.0 are preferred, however the methods are useful with all types of compounds. Simple filters like Lipinski's â€œrule of fiveâ€ have predictive value and may be used to improve the quality of leads discovered by this inventive strategy by using only those small molecules which are bioavailable. See Lipinski et al., Adv. Drug Delivery Rev. 23:3-25, 1997.

Combinatorial chemistry small molecule â€œlibrariesâ€ can be screened against drug targets. The idea is that diversity of chemical structures increases the chances of finding the needle in the 10²⁰⁰possible small organic molecule haystack. These collections provide an excellent source of novel, readily available leads. For example, ChemDiv uses more than 800 individual chemical cores, a unique Building Block Library, and proprietary chemistry in designing its Diversity Collections (small molecule libraries) to assemble 80,000-100,000 compounds a year. CombiLab lead library sets of 200-400 compounds also can be produced as a follow-up. In addition, ChemDiv's compounds are designed to ensure their similarity to drugs adjusted according to proprietary algorithms of â€œdrug-likeness definitionsâ€ (group similarity and advanced neural net approaches), and a variety of intelligent instruments for ADME&T (Absorption, Distribution, Metabolism, Excretion and Toxicity) properties prediction, such as partition coefficient, solubility, dissociation coefficients, and acute toxicity.

Thus, focused synthesis of new small molecule libraries can provide a variety of compounds structurally related to the initial lead compound which may be screened to choose optimal structures. Preferably, a library of compounds is selected that are predicted to be â€œdrug-likeâ€ based on properties such as pKa, log P, size, hydrogen bonding and polarity. The inventive multi-step approach which yields high affinity peptides in the first step, and small molecules in a subsequent step reduces the number of artificial hits by eliminating the lower affinity small molecules that would be selected and have to be assayed in a normal high throughput screening method. In addition, it focuses the search for molecules that can modulate the binding of a peptide the mimics the G protein rather than screening for binding to any site on the receptor. Other advantages of this technology are that it is simple to implement, amenable to many different classes of receptors, and capable of rapidly screening very large libraries of compounds.

Screening of the peptides or small molecules may be performed conveniently using receptors from any source. Generally, it is convenient to purify receptor from cells and reconstitute the receptor in lipid vesicles or to use membranes isolated from insect or mammalian cells that express or overexpress the receptor. PAR1 and rhodopsin are convenient receptors, however any suitable receptor is contemplated for use with this invention. The receptors used for screening may be purified from a natural source or purified from cells which overexpress the receptor and reconstituted in lipid vesicles. Membranes containing the receptor may be prepared from cells which natively express the receptor, for example Sf9 cells which express PAR1, or from cells which have been genetically engineered to express the receptor, for example mammalian or insect cells overexpressing PAR1. Peptides identified from screening a receptor (PAR1) expressed by three different methods are shown in Tables XI, XII, and XIII. The results indicate the methods give similar results showing a high degree of conservation, (N348; L349) being identified for all three methods of receptor expression. Initially, it is advantageous to determine the binding affinity of the peptide fusion protein or high affinity peptide against which the peptides or small molecules are screened. This allows the amount of receptor and peptide MBP peptide fusion protein or small molecule in the assay to be optimized.

Generally, it is convenient to test the libraries using a one well-one compound approach to identify compounds which compete with the peptide fusion protein or high affinity peptide for binding to the receptor. A single compound per well can be used, at about 1 Î¼M each or at any convenient concentration depending on the affinity of the receptor for the compounds and the peptide against which they are being tested. Compounds may be pooled for testing, however this approach requires deconvalution. Compounds may be pooled in groups of about 2 to about 100 compounds per well, or more, or about 10 to about 50 compounds per well at about 10 nM each or at any convenient concentration depending on the affinity of the receptor for the compounds being tested. Several different concentrations may be used if desired. Peptides desirably are screened using a pooled approach because of the layer members of peptides which are screened in the first instance. Peptides may be screened individually as well, but preferably are screened in pools of about 10⁴-10¹²peptides per well or about 10⁸-10¹⁰peptide per well or most preferably about 10⁹peptides per well.

ELISA, or any other convenient assay, such as fluorescence assays or radioimmunoassay may be used to determine (1) if one or more peptides in each well reduce the amount of binding by the high affinity peptide fusion protein or high affinity peptide, or (2) if one or more peptides in each well bind to the receptor. Compounds may be tested at a series of concentrations as well, and this generally is preferred if the affinity of the peptide or peptide fusion protein is not known. In an ELISA, wells in which the OD₄₅₀is half or less than half than that of control wells (no tested compounds) generally are considered â€œpositiveâ€ and may be further studied. Any suitable cut-off point may be used, however, depending on the assay components and the goals of the assay.

Screening against the high affinity peptide analogs can be performed using the desired GPCR immobilized onto microtiter wells, biochips, or any convenient assay surface. Binding assays performed in solution also are suitable. One, several, or thousands of candidate small molecule pharmaceutical compounds can be screened for binding to the receptor in the presence or absence of a high affinity peptide analog. The assays preferably are performed in the presence of a high affinity binding peptide to ensure that only those candidate compounds which can successfully compete for binding against the high-affinity binding peptide will be captured by the receptor. Alternatively, organic compounds or small molecules which have been identified by screening as competitively binding in the presence of a high affinity peptide analog also may be used as lead compounds in screening for further small molecule candidate compounds with even higher affinity. In either screening process, binding may be detected by any convenient method, for example by ELISA, fluorescence assays or radioimmunoassays.

By using a two-step protocol to identify compounds which block G protein signaling, high throughput screening of compounds and characterization of the selected compounds is significantly reduced in both time and cost, because only potent and strongly binding compounds are selected. The first step of identification of high affinity peptides which strongly compete with G proteins for their site of binding on G protein-coupled receptors insures this because the high affinity peptides are designed and tested for the particular desired binding specificity, ability to modify function within a cellular system and ability to modify functions in vivo.

Preferably, only the most strongly binding and effective peptide analogs or small molecules are used in the second or subsequent screening step. This two- or multi-step protocol reduces the number of false positives and identification of compounds which bind only weakly by eliminating the lower affinity small molecules that would be detected and then further studied if a conventional high throughput screening method were used. This method, therefore, is simple to implement, inexpensive, composed of only a few components, amenable to many different classes of receptors, and capable of rapidly screening large libraries of compounds. This method enables efficient identification of new classes of small organic peptidomimetic molecules that function as inhibitors or enhancers of receptor action, for example, thrombin receptor modifiers, dopamine receptor modifiers, histamine receptor modifiers, or vasopressin receptor modifiers. These identified compounds can target a single GPCR, a class of GPCR, or block or enhance a single G protein pathway activated by GPCR.

Thorough evaluation of the selected compounds (either peptides or small molecules) for use as therapeutic agents may proceed according to any known method. Properties of the compounds, such as pK_a, log P, size, hydrogen bonding and polarity are useful information. They may be readily measured or calculated, for example from 2D connection tables, if not already known prior to identification by the inventive method as a useful compound. Association/dissociation rate constants may be determined by appropriate binding experiments. Parameters such as absorption and toxicity also may be measured, as well as in vivo confirmation of biological activity. The screen may be optimized for small molecules according to methods known in the art. Additionally, it is preferable to use a software system for presentation of data that allows fast analysis of positives. See Example 36 and FIG. 2.

Pharmaceutical preparations are prepared by formulating the peptides or small molecules identified by the inventive screen according to methods well known in the art, with any suitable pharmaceutical excipient or combination of pharmaceutical excipients. Preparations may be made for administration by any route, such as intravenous, intramuscular, subcutaneous, oral, rectal, vaginal, transdermal, transmucosal, sublingual and the like, however, parenteral routes generally are preferred for peptide preparations. Any suitable vehicle may be used, for example saline or lactated Ringer's, for intravenous administration.

Dosages for treatment of GPCR-related diseases or conditions will depend on many factors such as the nature of the disorder, the GPCR involved, the route of administration, factors relating to the general physical condition and health of the patient and the judgment of the treating physician. Persons of skill in the art are well aware of these factors and consider manipulation of dosage to obtain an optimum result to be routine. Generally, dosages for intravenous administration may vary between about 0.01 mg/kg and 1000 mg/kg, however, this range can be expanded depending on the patient's needs. Such an expanded range is considered within the scope of this invention.

Alternatively, peptides according to this invention may be provided to cells, in vivo or ex vivo, by delivery of an expression construct. Gene therapy can be performed in vivo as a direct introduction of the genetic material. The in vivo gene transfer would introduce the oligonucleotides encoding the peptides to cells at the site they are found in the body, for example to skin cells on an arm, or to lung epithelial cells following inhalation of the gene transfer vector. Alternatively, ex vivo gene transfer, the transfer of genes into viable cells that have been temporarily removed from the patient and are then returned following treatment (e.g. bone marrow cells) could also be employed.

Gene transfer vectors can be engineered to enter specific tissues or cells. Transductional targeting allows the gene transfer vectors to interact with specific cell surface receptors. Transductional targeting also can take advantage of the rate of cellular division by using gene transfer vectors that target rapidly dividing cells such as tumor cells. Transcriptional targeting recruits distinct cellular promoter and enhancer elements to influence transcription of the therapeutic gene. Transfection efficiencies are also enhanced by engineering vectors with monoclonal antibodies, carbohydrate ligands, and protein ligands that help deliver genes to specific cells.

The gene transfer vectors used to produce the high affinity peptides inside cells could be viral vectors (e.g. Retrovirus, Adenovirus, Adeno-Associated Virus, Herpes Simplex Virus, or Vaccinia Virus). As an alternative, non-viral vectors also may be used, these include such methods as injection of naked DNA, or introduction of either DNA or peptides by attachment to positively charged lipids, or cationic liposomes, electroporation or ballistic DNA injection (limited to ex-vivo applications), as well as introduction of branched peptides.

Tet-inducible retroviral vectors for the native C-terminal sequences that co-expresses GFP driven by an internal ribosomal entry site (IRES) from encephalomyocarditis virus (p-Tet-Ti-GFP) may be used. These vectors can be modified so that they encode the high affinity peptide sequences. In addition, the high affinity peptide can be driven by a sequence allowing for spatial or temporal expression. For in vitro studies, viral supernatants may be collected from a pantropic producer line such as GP-293 (Clontech) in serum-free media. Viral supernatants may be concentrated by ultracentrifugation at 4Â° C. for 2 hr at 22,000 rpm, and the pellets resuspended in 1/100 the original volume in serum-free media with a titer of at least 10⁸IU (infectious units)/ml and stored at âˆ’80Â° C.

Murine leukemia virus (MLV) derived retroviral vectors are commonly used vehicles for stable delivery of therapeutic genes into endothelial cells. For the retrovirus studies in vivo, high affinity peptides advantageously are subcloned into a replication-defective murine Moloney retrovirus vector which is Tet-inducible and co-expresses GFP driven by an internal ribosomal entry site (IRES) from encephalomyocarditis virus (pTet-GFP). These constructs may then be transiently transfected into a producer line to generate cell-free titers of 10⁶-10¹⁰IU/ml. If needed, a pantropic retroviral expression system which utilizes VSV-G, an envelope glycoprotein from the vesicular stomatitis virus (GP-293; Clontech), may be utilized to overcome low transfection efficiencies. By using this innovative cell-based gene transfer method one can obtain stable, long-term, and localized gene expression of the high affinity C-terminal peptides.

To conclusively demonstrate that the compounds identified by this method can modulate G protein signaling events implicated in disease syndromes in vivo, antagonism or enhancement of selective G protein signal transduction events may be confirmed. One method of testing the ability of compounds to compete with native G protein binding involves expressing peptides that block the receptor-G protein interface in cells bearing the receptor. Plasmid constructs that encode GPCR-binding region peptides, such as carboxyl terminal peptide sequences from the various GÎ± subunits (see Table VI) can be used to express them in cells in vivo, ex vivo or in vitro, so that the metabolic effects of selective GPCR blockade can be studied qualitatively and quantitatively. Such studies provide proof that the binding which the compounds possess is useful in vivo to modulate selective G protein signals.

Expression of the peptides is conveniently achieved using the minigene approach by methods such as those described in Examples 23 and 24, however any suitable method may be used. Minigene vectors allow the high affinity peptides to be evaluated in cellular systems prior to high throughput screening. Any desired peptide sequence may be expressed using these methods. Those of skill in the art are well aware of alternative methods for construction, transfection and expression of protein and peptide constructs comprising the high affinity peptide analogs, and such methods are contemplated for use with them.

Human embryonic kidney cells advantageously are cultured in DMEM (Gibco) with 10% fetal bovine serum (Gibco), and Pen/Strep (5000 U/ml; 5000 pg/ml/ Gibco) in an atmosphere of 95% air/5% CO₂at 37Â° C. The cells may be plated at 60-70% density the day before transfection and transiently transfected for 1.5 hours with DNA (3 Î¼g) for pcDNA 3.1 vector with the insert (pcDNA3.1-high affinity peptide) or vector alone using an Effectene kit from Qiagen. After transfection, cells are washed once before adding complete HMEC media. When required, selection for cells carrying the minigenes may be performed by adding Neomycin to the media 48 hrs after transfection. To monitor efficiency of transfection cells are transfected with the GFP plasmid (Clonetech). When necessary, transfectants may be selected using 300 Î¼g/mL geneticin (G418). The expression of the vectors in HEK transfectants can be confirmed using reverse transcription (RT)PCR and Northern blot analysis for mRNA expression, and expression of the peptides can be characterized by HPLC as described previously. See Gilchrist et al., Methods Enzymol. 344:58-69, 2002, the disclosures of which are hereby incorporated by reference.

TABLE VIExemplary Sequences of C-terminalMinigene Peptides.Peptide NameSequenceSEQ ID NO:GÎ±iMGIKNNLKDCGLF112 GÎ±iRMGNGIKCLFNDKL113 GÎ±qMGLQLNLKEYNAV114 GÎ±q**MGLQLNLKEYNTL115 GÎ±12MGLQENLKDIMLQ116 GÎ±13MGLHDNLKQLMLQ117

As discussed above, many receptors interact with and activate multiple G proteins. Using the minigene strategy to introduce the high affinity-binding carboxyl terminal peptides into cells, it is possible to inhibit specific G protein-coupled receptor interactions with individual G proteins, thus demonstrating the feasibility of specific G protein blockade in vivo with compounds identified by the inventive method. For those receptors which activate multiple G proteins each of which activates a distinct set of signaling pathways mediating a specific set of responses (for example, the thrombin receptor), one pathway can be inhibited without substantially affecting the others.

To selectively antagonize G protein signal transduction events in vivo by expressing peptides that block the receptor-G protein interface, minigene plasmid vectors were designed to express the C-terminal peptide sequence of the various GÎ± subunits following their transfection into mammalian cells. A control minigene vector also was created, encoding the carboxyl terminus of GÎ±i_1/2in random order (GÎ±iR, see Table VI). One important element necessary for the minigene approach to block intracellular signaling pathways effectively in vivo is expression of adequate amounts of the desired peptides. Therefore, expression of the minigene should be confirmed by a convenient method of detecting mRNA, protein or both. Any convenient method known in the art can be used.

To determine the cellular efficacy of the minigene approach for expressing GPCR binding peptides, and to show the specific inhibition of one G protein pathway in response to a given receptor activation signal without affecting others, compounds advantageously may be assayed in a system designed to exhibit a measurable cellular signaling endpoint. One example of such a system is the thrombin receptor, PAR1, in endothelial cells. This receptor activates multiple G proteins. Several signaling endpoints, including transcription analysis of induced PAR1 gene expression; biochemical analysis of effector molecules including [Ca²⁺], MAP kinase (â€œMAPKâ€ ) activity, adenylyl cyclase activity, and inositol phosphate accumulation; as well as functional assays such as cell proliferation and endothelial permeability are available to measure specific activation or modulation of activation of different G proteins by ligand binding at this receptor. Signaling activity may be measured by any convenient method, including: measuring inositol phosphate accumulation; measuring intracellular calcium concentration levels; measuring transendothelial electrical resistance; measuring stress fiber formation; measuring ligand binding (agonist, antagonist, inverse agonist, etc.); measuring receptor expression; measuring receptor desensitization; measuring kinase activity; measuring phosphatase activity; measuring nuclear transcription factors; measuring cell migration (chemotaxis); measuring superoxide formation; measuring nitric oxide formation; measuring cell degranulation; measuring GIRK activity; measuring actin polymerization; measuring vasoconstriction; measuring cell permeability; measuring apoptosis; measuring cell differentiation; measuring membrane association of a protein that translocates upon GPCR activation, such as protein kinase C; measuring cytosolic accumulation of a protein that translocates upon GPCR activation, such as protein kinase C; measuring cytosolic accumulation of a protein that translocates upon GPCR activation, such as src; and measuring nuclear association of a protein that translocates upon GPCR activation, such as Ran. The functional effects of GÎ± C-terminal minigenes in the mechanism of thrombin-induced cell retraction, as measured by the change in transendothelial electrical resistance (TEER) also can be used to measure G protein inhibition.

For example, thrombin-mediated PAR1 gene induction was inhibited in human microvascular endothelial cells (HMEC) expressing the GÎ±i minigene construct. Expression of the GÎ±q minigene construct, however, affected thrombin-mediated inositol phosphate accumulation. Expression of GÎ±q also specifically decreased both thrombin-induced intracellular [Ca⁺⁺]_irise and thrombin-induced MAPK activity.

Thrombin activation of the GÎ±i mechanism in HMEC decreases cAMP levels increased in response to isoproterenol (which acts through GÎ±s). Assay for cAMP level increases in response to isoproterenol alone may be compared to increases after thrombin pre-incubation in cells expressing GÎ±i to show that expression of the GPCR binding peptide blocks GÎ±i signaling.

Recent work by Gohla et al., J. Biol. Chem. 274:17901-17907, 1999, elegantly demonstrated that thrombin receptors induce stress fiber accumulation via GÎ±12 in an EGF receptor-independent manner. Stress fiber formation appears to be Rho dependent. Both G12 and G13 have been implicated in the Rho signaling pathway. Therefore, expression of GÎ±12 and GÎ±13 GPCR-binding peptides in HMEC were used to determine whether these peptides could block the appearance of stress fibers in response to thrombin.

The extracellular signal-regulated kinase (ERK) subfamily of mitogen-activated protein kinases (MAPKs) regulates numerous cell signaling events involved in proliferation and differentiation. This forms the basis of another assay which can determine whether GPCR binding peptides can affect a specific G protein mediated pathway. Transfection of HMEC cells with minigenes encoding GPCR binding peptides along with HA-MAPK followed by immunoprecipitation of the HA-MAPK permits measurement of the effects only on cells expressing GPCR binding peptides.

Many studies have shown that the M₂muscarinic receptor (mAChR) couples exclusively to the Gi/GO family. See Dell'Acqua et al., J. Biol. Chem. 268:5676-5685, 1993; Lai et al., J. Pharm. Exp. Ther. 258:938-944, 1991; Offermanns et al., Mol. Pharm. 45:890-898, 1994; Thomas et al., J. Pharm. Exp. Ther. 271:1042-1050, 1994. The M₂mAChR can efficiently couple to mutant GÎ±q** in which the last five amino acids are substituted with the corresponding residues from GÎ±i or GÎ±O, suggesting that this receptor contains domains that are specifically recognized by the carboxyl terminus of GÎ±i/O subunits. See Liu et al., Proc. Natl. Acad. Sci. USA 92:11642-11646, 1995.

To test inhibition of G protein-coupled receptor-mediated cellular responses by carboxyl terminal GÎ± peptides expressed using minigene constructs, prototypical directly GÎ²Î³ activated channels (GIRK channels) regulated by a pertussis toxin-sensitive M₂mAChR was chosen as the model. In this model, the importance of the GÎ± carboxyl terminus and the downstream effector system have been well established. See Krapivinsky et al., J. Biol. Chem. 270:29059-29062, 1995; Krapivinsky et al., J. Biol. Chem. 273:16946-16952, 1998; Sowell et al., Proc. Natl. Acad. Sci. USA 94:7921-7926, 1997. Inhibition of M₂mAChR activation of inwardly rectifying potassium currents can be tested to demonstrate inhibition of a downstream functional response following agonist stimulation of GPCR on cells transiently transfected with a GÎ± carboxyl terminal peptide minigene or treated with a pharmaceutical compound identified by screening against high affinity GÎ± peptides.

GIRK channels modulate electrical activity in many excitable cells. See Breitwiese et al., J. Membr. Biol. 152:1-11, 1996; Jan et al., Curr. Opin. Cell Biol. 9:155-160, 1997; Wickman et al., Curr. Opin. Neurobiol. 5:278-285, 1995. Because the channel opens as a consequence of a direct interaction with GÎ²Î³, whole cell patch clamp recording of I_KAChcan be used to demonstrate inhibition of a downstream functional response following agonist stimulation of GPCR on cells transiently transfected with a GÎ± carboxyl terminal peptide minigene or treated with a pharmaceutical compound identified by screening against high affinity GÎ± peptides. Superfusion of cells expressing GIRK1/GIRK4 with their ligand, acetylcholine (ACh), activates inwardly rectifying potassium currents.

Using well-established receptor models accepted to be indicative of in vivo cellular results, this type of data can show that the individual G proteins activated via a given GPCR have specific roles in mediating cellular events and can be modulated in a specific fashion by ligands mimicking GPCR binding regions of individual GÎ± subunits. In particular, for receptors such as the thrombin receptor, which activate multiple G proteins, each of which activates a distinct set of signaling pathways mediating a specific set of responses, it is possible using the inventive methods to block one pathway while leaving all the others functional. The high affinity peptide analogs identified in vitro by consecutive affinity purification and competitive binding are capable of specifically inhibiting the downstream consequences of G protein signaling.

The assays described above clearly establish the ability of compounds identified by in vitro competitive binding studies to modulate a particular GPCR-G protein interaction selectively, even when the GPCR regulates multiple G proteins within the cell. Moreover, the peptides compete very effectively with the native sequence. In addition, the minigene approach described above and exemplified in the examples below allows a systematic test of the roles of other G proteins such as GÎ±12 and GÎ±13, which may be involved in the mechanism of increase of endothelial permeability, and clearly demonstrates the viability of this approach to select and identify GÎ± subunit modulating compounds. The peptides therefore are suitable for use in treatment of any disorder or syndrome characterized by G protein signaling excess.

In another aspect, the invention relates to methods to identify the G proteins with which a specific orphan receptor is coupled, using the materials provided by the invention. For example, the described methods can be used to test any GPCR with a battery of GÎ± subunit peptides to determine which species of G protein(s) mediates the effects of the receptor. The methods described in Examples 12-14 are suitable. Those of skill in the art are capable of designing other assays, or variations and modifications using these assays as guides.

Rhodopsin can be measured spectrophotometrically in many of its conformational states. The high affinity, biologically active rhodopsin state can be easily differentiated from its precursor, MI, by the â€œextraâ€ MII assay. See Example 35. This assay relies on the observation that under conditions of high pH and low temperature, MII is stabilized in the presence of Gt and can be spectrophotometrically measured. The ability of the C-terminal peptide to stabilize Meta II in the same manner as the heterotrimeric Gt, provides the tools to investigate the structural basis of the interaction of G proteins with the agonist binding sites of activated receptors.

The screening platform according to one embodiment of the invention can identify small molecules that increase the binding of the high affinity peptides which mimic G protein or stabilize the active conformation of the GPCR. These small molecules have an appropriate dose curve, and have an EC₅₀in the low Î¼M range. Samples of urea-washed rod outer segments typically have little or no (<5%) stabilization of MII unless G protein is added, however, the small molecules identified by the invention in one example screen stabilize the active (signaling) conformation of rhodopsin (MII). Addition of the small molecule alone results in 70% stabilization. The EC₅₀for stabilization of MII also appears to be in the low Î¼M range. Further, these same small molecules were added to human embryonic kidney (HEK) cells with measurement of their calcium response to a second GPCR do not enhance signaling of an unrelated GPCR and do not appear to cause an acute toxic response.

Drug discovery has evolved from an essentially random screening of products, into a process that includes the rational and combinatorial design of large numbers of synthetic molecules as potential bioactive agents, such as agonists, antagonists and inverse agonists, as well as the structural characterization of their biological targets, which may be polypeptides, proteins, or nucleic acids. Several approaches to facilitating the understanding of the structure of the therapeutic targets have been developed. These include sequencing of proteins and nucleic acids (Findlay et al., Protein Sequencing: A Practical Approach, IRL Press, Oxford, 1989; Adams et al., In Automated DNA sequencing and Analysis, Academic Press, San Diego, 1994), elucidation of secondary and tertiary structures via NMR (Jefson, Ann. Rep. Med. Chem. 23:275, 1998; Erikson and Fesik, Ann. Rep. Med. Chem. 27:271-289, 1992), X-ray crystallography (Erikson and Fesik, Ann. Rep. Med. Chem. 27:271-289, 1992) and computer algorithms for predicting protein folding (Copeland, Methods of Protein Analysis: A Practical Guide to Laboratory Protocols, Chapman and Hall, New York, 1994; Creighton, Protein Folding, W.H. Freeman and Co., 1992). Experiments such as ELISA (Kemeny and Challacombe, ELISA and other Solid Phase Immunoassays: Theoretical and Practical Aspects; Wiley, New York, 1988) and radioligand binding assays (Berson and Yalow, Clin. Chim. Acta, 22:51-60, 1968; Chard, An Introduction to Radioimmunoassay and Related Techniques, Elseveier press, Amsterdam/New York, 1982), surface-plasmon resonance (Karlsson et al., Anal. Biochem. 300:132-138, 2002), and scintillation proximity assays (Kariv et al., J. Biomol. Screen. 4:27-32, 1999) also can be used to understand the nature of the receptor-G protein interaction.

Peptides that block the protein-protein interactions of interest do so by binding to the surface of one of the interacting proteins and mimicking the interactions of the complete protein with the receptor. One can study the conformation of the active receptor-bound peptides when they are exchanging with the bound form, using transferred-NOESY NMR methods or X-ray diffraction if the peptides are more tightly bound. The bound peptide conformations can provide useful templates for the design of non-peptide small molecule drug leads which block the protein-protein interactions of interest. The binding sites of the peptides on the receptors also can be investigated using photochemical crosslinking, by substitution of peptide residues with photoactivatable amino acid analogs, crosslinking of the peptide to the receptor binding sites, cleavage of the receptor into peptide fragments and mass spectrometry analysis of the location of the binding sites. Combining structural data from a variety of experiments allows the development of models of the interacting protein surfaces using computer graphics and guides the design of novel non-peptide molecules to modulate the interactions.

In X-ray diffraction crystallography, a crystalline form of the molecule under study is exposed to a beam of X-rays and the intensity of detracted radiation is measured at a variety of angles from the angle of incidence. The beam of X-rays is diffracted into a plurality of diffraction â€œreflections,â€ with each reflection representing a reciprocal lattice vector. From the diffraction intensities of the reflections, the magnitudes of a series of numbers, known as â€œstructure factors,â€ are determined. The structure factors in general are complex numbers, having a magnitude and a phase in the complex plane, and are defined by the electron distribution within the unit cell of the crystal.

Crystals can be formed of receptor or portions of receptor bound to peptides that stabilize a particular conformation of interest. The methods of this invention, which identify peptides using combinatorial techniques that scan the complete set of possible amino acid sequences to find those that bind specifically to a particular receptor with high affinity, can identify peptides that bind to particular conformations of a GPCR. These peptides can be bound (and co-crystallized) with the receptor for structural determination studies by NMR or crystallography. Co-crystallization in this manner may be performed according to any method known in the art, for example the methods of Kimple et al., Nature 416:878-881, 2002, the disclosures of which are hereby incorporated by reference.

Therefore, in another embodiment, assays for identifying peptides that bind to a particular conformer of a GPCR are performed according to the methods described above for selection of the high affinity peptide analogs that bind activated rhodopsin. Once the high affinity peptides have been identified, they can be used in peptidomimetic studies. Compounds that mimic the conformation and desirable features of a particular peptide, e.g., an oligopeptide, but that avoid undesirable features, e.g., flexibility (loss of conformation) and metabolic degradation, are known as â€œpeptidomimetics.â€ Peptidomimetics that have physical conformations that mimic the three dimensional structure of the high affinity peptide analogs, that have surface active groups that allow binding to the receptor, or that have physical conformations that mimic the three dimensional structure of the high affinity peptide analogs can be used to make pharmaceutical compositions. Drugs with the ability to mimic the function of the high affinity peptide analogs that bind to the designated receptors can be identified using rational drug design according to this invention. The compounds preferably include the surface active functional groups of the high affinity peptide analogs, or substantially similar groups, in the same or substantially similar orientation, so that the compounds possess the same or similar biological activity. The surface-active functional groups in the high affinity peptide analogs possess a certain orientation when the receptor is present, in part due to their secondary or tertiary structure. Rational drug design involves both the identification and chemical modification of suitable compounds that mimic the function of the parent molecules.

The physical conformation of the peptidomimetics are determined, in part, by their primary, secondary and tertiary structure. The primary structure of a peptide is defined by the number and precise sequence of amino acids in the high affinity peptide analogs. The secondary structure is defined by the extent to which the polypeptide chains possess any helical or other stable structure. The tertiary structure is defined by the tendency for the polypeptides to undergo extensive coiling or folding to produce a complex, somewhat rigid three-dimensional structure.

Computer modeling technology allows scientists to visualize the three-dimensional atomic structure of a selected molecule and derive information that allows the rational design of new compounds that will mimic the molecule or which will interact with the molecule. The three-dimensional structure can be determined based on data from X-ray crystallographic analyses and/or NMR imaging of the selected molecule, or from ab initio techniques based solely or in part on the primary structure, as described, for example, in U.S. Pat. No. 5,612,895. The computer graphics systems enable one to predict how a new compound will link to the target molecule and allow experimental manipulation of the structures of the compound and target molecule to perfect binding specificity.

Many databases and computer software programs are available for use in drug design. For example, see Ghoshal et al., Pol. J. Pharmacol. 48(4):359-377, 1996; Wendoloski et al., Pharmacol. Ther. 60(2):169-183, 1993; and Huang et al., J. Comput. Aided Mol. Des. 11:21-78, 1997. Databases including constrained, metabolically stable non-peptide moeties may be used to search for and to suggest suitable analogs of the high affinity peptides identified in the screen. Searches can be performed using a three dimensional database for non-peptide (organic) structures (e.g., non-peptide analogs, and/or dipeptide analogs) having three dimensional similarity to the known structure of the active regions of these molecules. See, for example, Allen, Acta Crystallogr. B. 58:380-388, 2002.

Alternatively, three dimensional structures generated by other means such as molecular mechanics can be consulted. In addition, search algorithms for three dimensional database comparisons are available in the literature. Rufino et al., J. Comput. Aided Mol. Des. 8:5-27, 1994. Commercial software for such searches is also available from vendors such as Accelrys Inc. (9685 Scranton Road, San Diego, Calif. 92121-3752). The searching is done in a systematic fashion by simulating or synthesizing analogs having a substitute moiety at every residue level. Preferably, care is taken that replacement of portions of the backbone does not disturb the tertiary structure and that the side chain substitutions are compatible to retain the high affinity peptide/receptor interactions.

Using information regarding the bond angles and spatial geometry of the critical amino acids, one can use computer programs as described herein to develop peptidomimetics. Thermal protein unfolding, or thermal â€œshiftâ€ assays have been used to determine whether a given ligand binds to a target receptor protein. In a physical thermal shift assay, a change in a biophysical parameter of a protein is monitored as a function of increasing temperature. For example, in calorimetric studies, the physical parameter measured is the change in heat capacity as a protein undergoes temperature-induced unfolding transitions. Differential scanning calorimetry may be used to measure the affinity of a ligand for a G protein coupled receptor. Grauschopf et al., Biochemistry 39:8878:87, 2000; Brandts et al., Biochemistry 29:6927-40, 1990. Thus, using methods common to those skilled in the arts, the high affinity peptides may be assayed for their ability to modulate thermal shift of the receptor.

Because of the difficulty in obtaining high-resolution crystallographic structures from GPCRs, a variety of biophysical methods have been applied to characterize the interactions between the G protein, the receptor and the ligand. These include fluorescence resonance energy transfer (FRET) experiments performed with fluorescence-labeled peptide analogs (Bettio et al., Biopolymers 60:420-37, 2001), bioluminescence resonance energy transfer (BRET) experiments (Ayoub et al., J. Biol. Chem. 277:21522-8, 2002), photoaffinity labeling (Turek et al., J. Biol. Chem 277:16791-16797, 2002), fluorescence spectroscopy (Ghanouni et al. J. Biol. Chem. 276:24433-24436, 2001), site-direct spin labeling (Hubbell et al., Nat. Struct. Biol. 7:735-739, 2000), Fourier transform infrared difference spectroscopy (Vogel et al., Biochemistry 35:11149-11159,1996), and intrinsic tryptophan fluorescence spectroscopy. Farrens et al., J. Biol. Chem. 270:5073-5076, 1995.

The following non-limiting examples are provided to illustrate certain aspects of this invention.

EXAMPLES Example 1 Construction of a Peptide Library

Construction of a biased peptide library has been described previously. Martin et al., J. Biol. Chem. 271:361-366, 1996; Schatz et al., Meth. Enzymol. 267:171-191, 1996. The vector used for library construction was pJS142 (see FIG. 3). This vector had a linker sequence between the LacI and the biased undecamer peptide coding sequence, as well as restriction sites for cloning the library oligonucleotide. The oligonucleotide synthesized to encode the mutagenesis library was synthesized with 70% of the correct base and 10% of each of the other bases at each position. This mutagenesis rate leads to a biased library such that there is approximately a 50% chance that any of the 11 codons will be the appropriate (native) amino acid and approximately a 50% chance that it will be another amino acid. In addition, a linker of four random NNK (where N denotes A, C, G or T and K denotes G or T) codons were synthesized at the 5â€² end of the sequence to make a total of 15 randomized codons. Using this method, a library with greater than 10⁹independent clones per microgram of vector used in the ligation was constructed based on the carboxyl terminal sequence of GÎ±t (IKENLKDCGLF; SEQ ID NO:15). The nucleic acid used for creating this library was: 5â€²-GAGGTGGTNNKNNKNNKNNKattcaaggagaacctgaaggactgcggcctcttcTAACT AAGTAAAGC-3â€², wherein N=A/C/G/T and K=G/T; SEQ ID NO:118).

Example 2 Sequences for the Creation of GÎ± Subunit Peptide Libraries

Libraries were created using the methods of Example 1 and the sequences listed below in Table VII.

TABLE VIIC-Terminal GÎ±â€ƒSubunit Peptide Library Constructs.GÎ±SEQSub-IDunitRELinkerPeptide Coding RegionStopRENO:Gs5-GAGGTGGTNNKNNKNNKNNKattcgtgaaaacttaaaagattgtggtcgtttcTAACTAAGTAAAGC-3â€²14 G115-GAGGTGGTNNKNNKNNKNNKctgcagctgaacctgaaggagtacaatctggtcTAACTAAGTAAAGC-3â€²119 G125-GAGGTGGTNNKNNKNNKNNKctgcaggagaacctgaaggacatcatgctgcagTAACTAAGTAAAGC-3â€²120 G135-GAGGTGGTNNKNNKNNKNNKctgcatgacaacctcaagcagcttatgctacagTAACTAAGTAAAGC-3â€²121 G155-GAGGTGGTNNKNNKNNKNNKctcgcccggtacctggacgagattaatctgctgTAACTAAGTAAAGC-3â€²122 Gz5-GAGGTGGTNNKNNKNNKNNKatacagaacaatctcaagtacattggcctttgcTAACTAAGTAAAGC-3â€²123

Example 3 Isolation of Membranes from Insect Cells Expressing Thrombin Receptor

Sf9 cells (2Ã—10⁸cells) were cultured with 200 ml of Grace's insect cell culture medium (Life Technologies, Inc., Grand Island, N.Y.) containing 0.1% Pluronic F-68 (Life Technologies, Inc., Grand Island, N.Y.)), 10% fetal calf serum, and 20 Î¼g/ml gentamicin in a 1-liter spinner flask at 27Â° C. for 25 hours. Sf9 cells were infected with the ThR/pBluebac recombinant virus at a multiplicity of infection of 3-5, and cultured at 27Â° C. for 4 days. The cells were harvested, washed with phosphate buffered saline, and then resuspended in 10 mM Tris-HCl, pH 7.4. Cells were then homogenized with a hand-held homogenizer set at low speed for 20 seconds. The broken cells then were sedimented at 17,000Ã—g for 15 minutes. The supernatant was discarded, and the pellet resuspended in a buffer consisting of 50 mM Tris-HCl, pH 7.4 and 10% glycerol. Concentration of receptor in the membrane preparation ranged from 1-10,000 pmole/mg. For screening, a final concentration of 200 Î¼g/ml was used. The thrombin receptors were tested for their ability to bind to the native Gq-C terminal peptide using a MBP-Gq fusion protein.

Example 4 Isolation of Membranes from Mammalian Cells Overexpressing Thrombin Receptor

PAR1 receptor cDNA (2.1 kb insert) was obtained by polymerase chain reaction and cloned into the mammalian expression vector pBJ5. The resulting plasmid was transfected into Chinese hamster ovary cells by the calcium phosphate coprecipitation method. The PAR1-transfected cells were grown with Dulbecco's modified Eagle's medium containing 10% fetal calf serum, 100 units/mL penicillin and 100 Î¼g/mL streptomycin. The cells were detached using PBS with 5 mM EDTA and washed twice in PBS. The pellet was either used immediately for membrane preparation or stored frozen at âˆ’20Â° C. Pellets were homogenized in 20 mM Tris-HCl, pH 7.5, with 5 mM EDTA and 0.5 mM PMSF, using a Dounce homogenizer (10 strokes) and sonicated for 10 seconds. Nuclear debris and intact cells were removed by centrifugation at 3000 rpm for 10 minutes. The supernatant was sedimented at 12,000Ã—g for 30 minutes and the resulting pellet suspended in 25 mM Tris-HCl, pH 7.5, 25 mM MgCl₂, 10% sucrose, 0.5 mM PMSF, 50 Î¼g/mL antipain, 1 Î¼g/mL aprotinin, 40 Î¼g/mL bestatin, 100 Î¼g/mL chymostatin, 0.5 Î¼g/mL leupeptin and 0.7 Î¼g/mL pepstatin. The membranes were aliquoted and frozen at âˆ’80Â° C.

Example 5 Preparation of Rod Outer Segments

Bovine rod outer segments (rhodopsin-containing membranes) were prepared from fresh retinas under dim red light as described by Arsharky et al., J. Biol. Chem. 269:19882-19887, 1994. The retinas were placed in a beaker for dissection filled with 200 mL of 30% (w/v) sucrose in isolation buffer (90 mM KCl, 30 mM NaCl, 2 mM MgCl₂, 0.1 mM EDTA, 1 mM DTT, 50 Î¼M phenylmethylsulfonyl fluoride, 10 mM MOPS, pH 7.5) on ice with constant moderate stirring of the solution during dissection. Following dissection, the retina solution was left in the dark for one hour on ice. The retina-sucrose solution was distributed into eight 50 mL tubes and sedimented at 3000Ã—g for four minutes at 4Â° C. The supernatant was decanted into eight fresh centrifuge tubes and placed on ice. The tubes were filled to 1.5 cm below top with isolation buffer, then sedimented at 17,000Ã—g for 20 minutes (â€œspin 1â€ ).

The pellets were resuspended in a small volume of 30% sucrose and consolidated from eight tubes into four tubes. The tubes were filled to 1.5 cm below top with 30% sucrose, sedimented at 5000Ã—g for four minutes at 4Â° C., and the supernatant decanted into four clear tubes. These tubes were filled to 1.5 cm below top with isolation buffer and sedimented at 17,000Ã—g for 20 minutes at 4Â° C. (â€œspin 2â€ ).

A stepwise sucrose gradient was prepared in six gradient tubes using the solutions in Table VIII, below, with a sequence from top to bottom of #2, #3, #4.

TABLE VIIISucrose Gradient Solutions.Solution#2 (0.84 M)#3 (1.0 M)#4 (1.14 M)42% Sucrose51.30g61.05g69.75g1.0 M MOPS750Î¼L750Î¼L750Î¼L2.0 M KCl2250Î¼L2250Î¼L2250Î¼L3.0 M NaCl750Î¼L750Î¼L750Î¼L2.0 M MgCl₂75Î¼L75Î¼L75Î¼LTotal Weight83.25g84.75g86.25g

The pellets from â€œspin 1â€ and â€œspin 2â€ were resuspended in isolation buffer using 1 mL 26% sucrose buffer per tube. After making a slurry, each tube was homogenized with a 1 mL pipette and the tubes consolidated. The pellet solution was carefully laid onto the sucrose gradients and was not allowed to invade the gradient layers. The gradient tubes were subjected to 24,000Ã—g for 30 minutes at 4Â° C. in a swinging bucket rotor, after which the orange layer containing the membranes was collected carefully, to avoid disturbing the pellet or the dark solution near the pellet. The membranes were distributed into six 50 mL tubes and placed on ice. The tubes then were filled to 1.5 cm below top with isolation buffer and sedimented at 17,000Ã—g for 20 minutes at 4Â° C. The supernatant was discarded and the pellets resuspended in 1 mL isolation buffer containing 5 Î¼g/mL pepstatin and 10 Î¼g/mL E-64. This suspension was stored in a foil-wrapped 15 mL conical tube at âˆ’80Â° C. until needed, then thawed, homogenized in EDTA buffer (10 mM Tris, pH 7.5, 1 mM EDTA 1 mM DTT) and sedimented at 30,000Ã—g for 30 minutes. The supernatants were discarded and the pellets resuspended and sedimented again as described above. The pellets then were resuspended in urea buffer (10 mM Tris, pH 7.5, 1 mM EDTA, 1 mM DTT, 7 M urea), homogenized and sedimented at 45,000 kg for 40 minutes. These pellets were resuspended and homogenized in Buffer A (200 mM NaCl, 10 mM MOPS, pH 7.5, 2 mM MgCl₂, 1 mM DTT, 100 Î¼M PMSF), then sedimented at 30,000Ã—g for 30 minutes. The pellets each were resuspended and homogenized by pipetting in 1 mL buffer A and stored at âˆ’80Â° C. in 100 Î¼L aliquots in foil-covered tubes for use in assays. For screening, the receptor was added to wells at 10 Î¼g/ml. Binding assays were performed as in Example 18.

Example 6 Purification of PAR1 Thrombin Receptor from Insect Cells and Reconstitution of Receptors into Lipid Vesicles

Sf9 cells (2Ã—10⁸cells) were cultured in Grace's insect cell culture medium (Life Technologies, Inc., Grand Island, N.Y.) containing 0.1% Pluronic F-68 (Life Technologies), 10% fetal calf serum and 20 Î¼g/mL gentamicin in a 1 L spinner flask at 27Â° C. for 25 hours. The cells were infected with ThR/pBluebac (recombinant virus) at a multiplicity of infection of 3-5 and cultured at 27Â° C. for four days. The cells were harvested, washed with phosphate buffered saline containing 2.7 mM EDTA and stored at âˆ’70Â° C. until used. The cells were resuspended in lysis buffer (2.5 mM Tris-HCl, pH 7.2, 7.5 mM NaCl, 10 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 10 mg/mL leupeptin, 10 mg/mL aprotinin, 50 mM NaF) and washed. All subsequent steps were done on ice with cold buffers and centrifuge rotors at or below 4Â° C. The cells were homogenized for one minute at maximum speed and sedimented for 45 minutes at 30,000Ã—g. The pellet was resuspended in lysis buffer and the homogenation/washing step repeated three times. The resulting pellet was resuspended in 30 mL solubilization buffer (20 mM Tris-HCl, pH 7.4, 15 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 10 mg/mL leupeptin, 10 mg/mL aprotinin, 50 mM NaF, 0.1% (w/v) digitonin, 0.1% (w/v) Na deoxychoate) and then homogenized for one minute. The suspension was stirred for 90 minutes at 4Â° C. and then sedimented for 60 minutes at 30,000Ã—g. The supernatant was loaded onto an anti-PAR1 monoclonal antibody column equilibrated with solubilization buffer containing 0.2% digitonin. After application of the supernatant, the column was washed with 10 column volumes of 10 mM Tris-HCl buffer, pH 7.4, containing 0.2% (w/v) Na dodecyl maltoside. The receptor was eluted using 10 mM triethylamine, pH 11.8. The eluted fractions were neutralized immediately using 1 M HEPES, pH 6.4. The pooled fractions were dialyzed against 50 mM HEPES buffer, pH 7.4, containing 50% (v/v) glycerol, 0.1 M NaCl and 0.2% (w/v) Na dodecyl maltoside. Aliquots were stored at âˆ’80Â° C.

For preparation of lipid vesicles, 200 Î¼L phosphatidylserine (50 mg/mL in CHCl₃; Matreya) was dried in a rotary evaporator for 30 minutes or using a stream of dry N₂. After addition of 200 Î¼L buffer A (50 mM HEPES, 100 mM NaCl, 0.2% (w/v) Na dodecylmaltoside), the tube was sealed under an N₂atmosphere and sonicated in a bath sonicator for 30 minutes. Reconstitution of receptors into lipid vesicles was performed the same day, using purified receptor prepared as in Example 5. Purified receptor stocks (200 Î¼g/mL) were thawed on ice and 50 Î¼L was incubated for 20 minutes at 4Â° C. with the appropriate agonist peptide (100 nM final concentration). In the case of thrombin receptor, the agonist is thrombin receptor agonist peptide (CalbioChem). After addition of 80 Î¼L sonicated lipids and 50 Î¼L buffer A, the samples were mixed using a vortex machine and placed on ice for 10 minutes. The samples then were loaded onto a 1 mL Extracti-gelâ„¢ column which had been washed with 0.2% BSA and pre-equilibrated with 5 mL Buffer A without Na dodecylmaltoside. The reconstituted vesicles were eluted from the column with 2.5 mL HEK buffer.

Samples (100-200 Î¼L) were collected for purity analysis by SDS-PAGE. The concentration for each batch generally was about 10-1000 Î¼g/ml. For use, receptor was placed in microtiter plates at about 1-100 Î¼g/ml. The purified, reconstituted thrombin receptors were tested for their ability to bind to the native Gq-C terminal peptide using a MBP-Gq fusion protein. As a control, empty vesicles also were tested for their ability to bind to the native Gq-C terminal peptide using a MBP-Gq fusion protein.

Example 7 Identification of GPCR-Binding High Affinity Peptide Analogs (Panning)

Electrocompetent cells were produced as follows. A single colony of ARI814 bacteria was grown overnight at 37Â° C. in 5 ml sterile SOP (20 g/L Bacto-tryptone; 10 g/L Bacto-yeast extract; 5 g/L NaCl; 2.5 g/L anhydrous K₂HPO₄; 1 g/L Mg₂SO₄.7H₂O). One milliliter of this overnight growth was added to 500 ml SOP and the bacteria were allowed to grow until the OD₆₀₀read 0.6-0.8. All further washing steps were done in the cold. The cells were placed in an ice-water bath for at least 15 minutes, then subjected to centrifugation at 4000Ã—g for 15 minutes at 4Â° C. followed by resuspension in 500 ml 10% glycerol. After sitting on ice for 30 minutes, the cells were washed twice more in 500 ml and 20 ml 10% glycerol with sedimentation as above, and finally sedimented at 5000Ã—g for 10 minutes at 4Â° C. and resuspended in 1 mL 10% glycerol. Cells were quick-frozen using dry ice and isopropanol in 100 Î¼L aliquots for later use.

To transfect, aliquots (40 Î¼L) of thawed ARI814 cells were placed into each of three chilled microcentrifuge tubes. A peptide display library based on the undecamer carboxyl terminal peptide of GÎ±t (SEQ ID NO:15) was prepared according to Example 1. Two microliters of library plasmid were added to the tubes and mixed. For the first round of â€œpanning,â€ 200 Î¼l of the plasmid library was added. For subsequent rounds, three sets of transfections were performed (adherent plasmids from wells containing receptor (+); adherent plasmids from wells containing no receptor (âˆ’); and the PRE sample which was not incubated). See below.

In each round of panning, less library was used (round 2:100 Î¼l; round 3:50 Î¼l; round 4:10 Î¼l). After the panning was completed, the DNA for the LacI fusion protein was eluted. This DNA (50 Î¼l) was used to transfect E. Coli cells by electroporation, using cold, sterile 0.1 cm electrode gap cuvettes. The cuvettes were pulsed one time using a BioRad E. coli Pulsar set to 1.8 kV, 25 Î¼F capacity, time constant 4-5 mseconds, with the Pulser Controller unit at 200 mÎ©. Immediately, 1 mL of SOC was added and the mixture transferred to a labeled 17Ã—100 mm polystyrene tube. The tube was shaken for one hour at 37Â° C. Aliquots were taken from each set to plate 100 Î¼L undiluted to 10^âˆ’6dilution samples on LB-Amp plates. Counts of the PRE plates indicated library diversity, while comparison of the (+) and (âˆ’) plates indicated whether specific clones were being enriched by the panning procedure.

The remaining Ëœ900 Î¼L in the + receptor tube was added to a 1 L flask containing 200 mL LB-AMP media, prewarmed to 37Â° C., and grown at 37Â° C., shaking until OD₆₀₀=0.5. The tube of cells then were placed in an ice water bath for at least 10 minutes, and kept chilled at or below 4Â° C. during the subsequent washing steps. The cells were sedimented at 5000Ã—g for six minutes, resuspended in 100 mL WTEK buffer, sedimented at 5000Ã—g for six minutes, resuspended in 50 mL TEK buffer, resedimented at 5000Ã—g for six minutes and resuspended in 4 mL HEK buffer. The cells were divided into the cryovials and stored at âˆ’70Â° C. One tube was used for the next round of panning and the other saved as a backup.

The panning process is illustrated in FIG. 1. For screening of the library by â€œpanning,â€ rhodopsin receptors prepared according to Example 5 were immobilized directly on Immulon 4 (Dynatech) microtiter wells (0.1-1 Î¼g of protein per well) in cold 35 mM HEPES, pH 7.5, containing 0.1 mM EDTA, 50 mM KCl and 1 mM dithiothreitol (HEK/DTT). After shaking for one hour at 4Â° C., unbound membrane fragments were washed away with HEK/DTT. The wells were blocked with 100 Î¼l 2% BSA in HEKL (35 mM HEPES; 0.1 mM EDTA; 50 mM KCl; 0.2 M Î±-lactose; pH 7.5, with 1 mM DTT). For rounds 1 and 2, BSA was used for blocking; in later rounds 1% nonfat dry milk was used. For the first round of panning, about 24 wells of a 96-well plate were used. In subsequent rounds, 8 wells with receptor and 8 wells without receptor were prepared.

The Gt library was thawed (2 mL aliquot) and mixed with 6 mL lysis buffer on ice. Lysis buffer contains 4.25 mL HE (25 mM HEPES: 0.1 mM EDTA; pH 7.5); 1 mL 50% glycerol; 750 Î¼L 10 mg/mL protease-free BSA in HE; 10 Î¼L 0.5 M DTT; and 6.25 Î¼L 0.2 M PMSF. Freshly prepared lysozyme solution (150 Î¼L 10 mg/mL lysozyme in cold HE) was added and the tube was gently inverted several times and incubated on ice for no more than two minutes. The extent of lysis is evidenced by an increase in viscosity that can be observed by noting the slow migration of bubbles to the top of the tube after mixing. Lysis was terminated by mixing in 2 mL 20% lactose and 250 Î¼L 2M KCl. The tube was centrifuged immediately at 13,000Ã—g for 15 minutes at 4Â° C. and the supernatant transferred to a new tube. A small aliquot of 0.1% (the PRE sample) was saved in a separate, labeled tube. The blocked rhodopsin receptor-coated plate was rinsed four times with HEKL/1% BSA and exposed to room light for less than five minutes on ice to activate the rhodopsin for light-activated rhodopsin (Table IX), or left in the dark for dark-adapted (inactive) rhodopsin (Table X). Immediately thereafter, the crude bacterial lysate from the peptide library (200 Î¼L) was added to each well and allowed to shake gently for one hour at 4Â° C. For round 2, this same procedure was followed. In round 3, the amount of lysate used was reduced to 100 Î¼L. In subsequent rounds, the lysate was diluted 1:10 in HEKL/BSA. In all rounds, 5-10 Î¼L 200 Î¼M native peptide was added to the wells to chase off peptides that were bound with lower affinity.

After incubation with the bacterial lysate, the wells were washed four times into cold HEKL/1% BSA. Sonicated salmon sperm DNA (200 Î¼L 0.1 mg/mL in HEKL/1% BSA was added to each well and shaken gently for 30 minutes at 4Â° C. The plates were washed four times with cold HEKL and twice with cold HEK, then eluted by adding 50 Î¼L/well 1 mM IPTG/0.2 M KCl in HE with vigorous shaking at room temperature for 30 minutes. The eluants from each group of wells (+ or âˆ’ receptor) were combined in one or more microcentrifuge tubes as necessary. The volume of the PRE sample which had been saved previously was brought up to match the volume of the eluant samples and precipitated in parallel with them. To precipitate, 1/10 volume of 5M NaCl was mixed with each of the samples, then 1 Î¼L 20 mg/mL glycogen was mixed with the samples. An equal volume of RT isopropanol was then added and mixed thoroughly. The samples were subjected to centrifugation at 13,000Ã—g for 15 minutes and the supernatant aspirated. The pellet was washed with 500 Î¼L cold 80% ethanol and again subjected to centrifugation at 13,000Ã—g for 10 minutes. The pellets of plasmid DNA were resuspended in sterile, double-distilled water, 200 Î¼L for the PRE sample and 4 Î¼L for the + or âˆ’ receptor samples and stored at âˆ’20Â° C.

Both light-activated rhodopsin and dark-adapted rhodopsin were used to screen the library in this manner. See Tables IX and X, below. Six of the sequences obtained using light-activated rhodopsin were 100-1000 times more potent than the native sequence at binding rhodopsin and are listed in Table IX. When the GÎ±t library was used to pan light-activated rhodopsin, residues L344, C347 and G348 were invariant. Also, in each of the highest affinity sequences, the basic residue at position 341 (R341) was changed to a neutral residue. When the GÎ±t library was used to pan dark-adapted odopsin, the L344, C347 and G348 residues were no longer variant (L344 present in 62.5% of sequences, C347 present in 25% of sequences, G348 present in 75% of sequences) and the residue at position 341 was usually unchanged. Thus, the conformation of the receptor in its inactive, dark-adapted state allows it to bind to a different set of peptide analogs than the light-activated receptor. In addition, it appears that in the light-activated receptor, it is the last seven amino acids of the peptide which are most important (344-350) while the first six amino acids (340-345) are more important for dark-adapted rhodopsin binding.

TABLE IXLight-Activated Rhodopsin HighAffinity Sequences.Clone No.SEQ ID NO:SequenceLibrary Sequence124IRENLKDCGLF 8125LLENLRDCGMF 9126IQGVLKDCGLL 10127ICENLKECGLF 18128MLENLKDCGLF 23129VLEDLKSCGLF 24130MLKNLKDCGMF 3131LLDNIKDCGLF 4132ILTKLTDCGLF 6133LRESLKQCGLF 11134IHASLRDCGLF 13135IRGSLKDCGLF 14136IFLNLKDCGLF 15/28137IRENLEDCGLF 16138IIDNLKDCGLF 17139MRESLKDCGLF 19140IRETLKDCGLL 26141ILADVIDCGLF 27142MCESLKECGLF

TABLE XDark-Adapted Rhodopsin High Affinity Sequences.Clone No.SEQ ID NO:SequenceLibrary Sequence124IRENLKDCGLF 2143IREKWKDLALF 3144VRDNLKNCFLF 7145IGEQIEDCGPF 17146IRNNLKRYGMF 21147IRENLKDLGLV 26148IRENFKYLGLW 33/37149SLEILKDWGLF 41150IRGTLKGWGLF

Example 8 Screens of PAR1 with a Gq Peptide Library

The methods of Example 7 were used to screen different sources of PAR1 receptor using the Gq library. Purified PAR1, reconstituted in lipid vesicles (Example 6), membranes prepared from Sf9 insect cells expressing PAR1 (Example 3) and membranes prepared from mammalian cells overpressing PAR1 were used. The results of the screens are presented in Tables XI, XII and XIII, respectively. The peptide used as the competitor for all three screens was LQLNLKEYNLV (SEQ ID NO:2). The 4-residue linker sequences are random and are optionally present at the amino terminus of the binding peptide. These results show that the identified high affinity peptides are similar for all three sources of screened PAR1. When a Gq-biased library is used to pan PAR1, the positions that appear to be critical for receptor recognition, and thus are invariant, are N348, L349 and V350.

TABLE XIReconstituted Purified Recombinant PAR1Receptor; Screening Results.SEQ IDCloneLinkerSEQ ID NO:LQLNLKEYNLVNO:2R2-16*SWV151LQFNLNDCNLV102 R2-17FVNC152LQRNKKQYNLG160 R2-18EVRR153MKLKLKEDNLV103 R2-20*RVQ154HQLDLLEYNLG104 R2-21RLTR155LQLRYKCYNLV161 R3-37SR*K156LQQSLIEYNLL111 R3-38MTHS157VHVKLKEYNLV162 R3-44SGPQ158LQLNVKEYNLV163 R3-46ML*N159LRIYLKGYNLV164

TABLE XIIPAR1 Receptor Sf9 Insect Cell Membranes;Screening Results.SEQ IDCloneLinkerSEQ ID NO:LQLNLKEYNLVNO:2S1-13S*IR165MKLNVSESNLV94 S1-18RWIV166LQLNLKVYNLV175 S1-23G*GH167LELNLKVYNLF176 S2-26RSEV168LQLKHKENNLM100 S2-30CEPG169LHLNMAEVSLV177 S2-36HQMA170LQVNLEEYHLV101 S3-6â€ƒVPSP171LQKNLKEYNMV106 S3-8â€ƒQMPN172LQMYLRGYNLV108 S3-10MWPS173LKRYLKESNLV178 S3-15C*VE174MNLTLKECNLV110

TABLE XIIIMammalian (CHO) Cells Overexpressing PAR1;Screening Results.SEQ IDCloneLinkerSEQ ID NO:LQLNLKEYNLVNO:2C4-5â€ƒPRQL179LQLKRGEYILV183 C4-19VRPS3LQLNRNEYYLV4 C5-10SRHT11LRLNGKELNLV12 C5-12FFWV180CSLKLKAYNLV184 C4-16QRDT181LQMNHNEYNLV185 C7-3â€ƒNERN182PQLNLNAYNLV186 C7-10LPQM9QRLNVGEYNLV10 C7-13LSTN7LHLNLKEYNLV8 C7-14LSRS5LQQKLKEYSLV6

Example 9 Identification of GPCR-Binding High Affinity Peptide Analogs (Panning)

The methods of Example 7 were repeated using recombinant reconstituted Î²₂adrenergic receptor panned with the Gs Library. Results of the panning screens and ELISA binding affinity of the selected peptides are shown in Table XIV, below.

TABLE XIVÎ²2-Adrenergic Receptor screenedwith Gs library.SEQ ID NOCompetitorQRMHLRQYELL13ELISA AG1QGMQLRRFKLR187.435 AG20RWLHWQYRGRG188.431 AG19PRPRLLRFKIP189.361 AG2QGEHLRQLQLQ190.330 AG4QRLRLGPDELF191.291 BAR1QRIHRRPFKFF192.218 AG3QRMPLRLFEFL193.217 BAR2QRVHLRQDELL194.197 AG11DRMHLWRFGLL195.192 AG9QRMPLRQYELL196.190 BAR3QWMDLRQHELL197.185 AG18QRMNLGPCGLL198.155 BAR20NCMKFRSCGLF199.079 AG13QRLHLRGYEFL200.054 BAR11HRRHIGPFALL201.048 BAR8ERLHRRLFQLH202.047 BAR40PCIQLGQYESF203.028 BAR31QRLRLRKYRLF204.026

Example 10 Identification of GPCR-Binding High Affinity Peptide Analogs (Panning)

The methods of Example 7 repeated using rhodopsin screening with a Gt library. Results of the panning screens and ELISA binding affinity of the selected peptides are shown in Table XV, below. To identify the rank order of binding, the lysates were analyzed using ELISA methods in which the secondary antibody was conjugated to HRP. Following addition of the substrate, the microplate was read using a spectrophotometer. The binding is the OD₄₅₀for wells with receptorâ€”OD₄₅₀for wells in which no receptor (control wells with empty lipid vesicles).

TABLE XVRhodopsin screened with Gt library.SEQ ID NO:CompetitorIRENLKDCGLF124ELISA L33IVEILEDCGLF2051.007 L4MLDNLKACGLF206.908 L3ILENLKDCGLF207.839 L14LRENLKDCGLL208.833 L38LLDILKDCGLF209.823 L15VRDILKDCGLF210.621 L34ILESLNECGLF211.603 L17ILQNLKDCGLE212.600 L7MLDNLKDCGLF213.525 L10IHDRLKDCGLF214.506 L20IRGSLKDCGLF135.423 L6ICENLKDCGLF215.342 L8IVKNLEDCGLF216.257 L13ISKNLRDCGLL217.187 L10IRDNLKDCGLF218.162

Example 11 Additional Peptide Analogs

Chinese hamster ovary-expressed PAR1 was screened against the Gt, G12 and G13 libraries, using the competitor peptide indicated in Table XVI below. Additional peptide analogs were identified using the Gt, G12 or G13 library as indicated and IRENLKDCGLF (SEQ ID NO:124), LQENLKDIMLQ (SEQ ID NO:38) or LQDNLKQLMLQ (SEQ ID NO:233), respectively as competitor with screening for high affinity binding to PAR1 receptor obtained from Chinese hamster ovary cells as described in Example 1, indicated in Table XVII, below.

TABLE XVIPeptides Identified with CHO EXPRESSED PAR1.Gt libraryG12 libraryG13 library(IRENLKDCGLF;(LQENLKDIMLQ;(LQDNLKQLMLQ;SEQ ID NO:124)SEQ ID NO:38)SEQ ID NO:233)IREFLTDCGLF 219LQENLKEMMLQ 225LQDNLRHLMLQ 234 IRLDLKDVSLF 220LEENLKYRMLD 226LQDKINHLMLQ 235 ICERLNDCGLC 221LQEDLKGMTLQ 227LQANRKLGMLQ 236 PRDNTKVRGLF 222LQETMKDQSLQ 228LIVKVKQLIWQ 237 FWGNLQDSGLF 223PQVNLKSIMRQ 229MRAKLNNLMLE 238 RRGNGKDCRHF 224 WQHKLSEVMLQ 230LQDNLRHLIQ 239 LKEHLMERMLQ 231LQDNRNQLLF 240 LLGMLEPLMEQ 232

TABLE XVIIPAR1 Binding Peptides Screened using a G11 Library(LQLNLKEYNLV; SEQ ID NO: 2)SF9CHO EXPRESSEDSEQ ID NO:Recomb/ReconstSEQ ID NO:EXPRESSEDSEQ ID NO:LQLNVKEYNLV163LQLNVKEYNLV163LQLNLKVYNLV175 LQLNRKNYNLV241LQLRVKEYKRG244LQLKHKENNLM100 LQLRYKCYNLV161LQLRYKCYNLV161LQKNLKEYNMV106 LQLDLKESNMV242LQIYLKGYNLV245LQVNLEEYHLV101 LQLNLKKYNRV243LQFNLNDCNLV102LFLNLKEYSLV257 LQLRVKEYKRG244LQRNKKQYNLG160LELNLKVYNLV258 LQRNKKQYNLG160LQRNKNQYNLG254LPLNPKEYSLV109 LQIYLKGYNLV245LQQSLIEYNLL111LPLNLIDFSLM259 LQFNLNDCNLV102LRLDFSEKQLV105LPRNLKEYDLG260 LQYNLKESFVV246LYLDLKEYCLF255LRLNDIEALLV261 LQQSLIEYNLL111HQLDLLEYNLG104LVLNRIEYNLL262 LQRDHVEYKLF247VQVKLKEYNLV251LHLNMAEVSLV177 LVIKPKEFNLV248MKLKLKEDNLV103MNLTLKECNLV110 IQLNLKNYNIV249SAKELDQYNLG256MKLNVSESNLV94 HQLDLLEYNLG104VHVKLKEYNLV162LKRYLKESNLV178 MQLNLKEYNLV250LKRKLKESNMG263 VQVKLKEYNLV251LKRKVKEYNLG264 QLLNQYVYNLV252LELNLKVYNLF176 MKLKLKEDNLV103LQMYLRGYNLV108 WRLSLKVYNLV253 LQLKRGEYILV183 LQLNRNEYYLV4 LRLNGKELNLV12 CSLKLKAYNLV184 LQMNHNEYNLV185 PQLNLNAYNLV186 QRLNVGEYNLV10 LHLNLKEYNLV8 LQQKLKEYSLV6

Example 12 Preparation of LacI Lysates

In the last round of panning, several clones were selected from the (+) receptor plates and grown up overnight in LB-Amp media. Three hundred microliters of the overnight culture was diluted in 3 mL LB-Amp media for â€œELISA lysate culture.â€ Another 30 Î¼L was added to an equal volume of 50% glycerol was stored in labeled microcentrifuge tubes at âˆ’70Â° C. The remaining 4.5 mL was used to make DNA using a standard miniprep protocol (Qiagen Spinprepâ„¢ kits) and sequenced using a 19 base pair reverse primer which is homologous to the vector at a site 56 basepairs downstream from the TAA stop codon that terminates the random region of the library (GAAAATCTTCTCTCATCCG; SEQ ID NO:265). The DNA was stored at âˆ’20Â° C. The ELISA lysate culture was allowed to shake for one hour at 37Â° C. Expression was induced by adding 33 Î¼L 20% arabinose (0.2% final concentration) with shaking at 37Â° C. for 2-3 hours. The culture then was subjected to sedimentation at 4000Ã—g for five minutes, the pellet resuspended in 3 mL cold WTEK buffer, resedimented at 4000Ã—g for five minutes and the pellet resuspended in 1 mL cold TEK buffer. After transfer to 1.5 mL microcentrifuge tubes, the pellet was sedimented at 13,000Ã—g for two minutes and the supernatant aspirated. The cell pellet was resuspended in 1 mL lysis buffer (42 mL HE, 5 mL 50% glycerol, 3 mL 10 mg/mL BSA in HE, 750 Î¼L 10 mg/mL lysozyme in HE and 62.5 Î¼L 0.2 M PMSF) and incubated on ice for one hour. One hundred ten microliters 2M KCl was added to the lysis mixture and inverted to mix, then sedimented at 13,000Ã—g for 15 minutes at 4Â° C. The clear crude lysate (about 0.9 mL supernatant) was transferred to a new tube and stored at âˆ’70Â° C.

Example 13 PAR1 Receptor-Specific Binding of LacI-Peptide Fusion Proteins

The binding properties of the peptide encoded by individual clones were assayed as follows. Purified PAR1 receptor prepared from Sf9 insect cells (1-10,000 pg/mL in 50 mM Tris-HCl, pH 7.4, 10% glycerol) was reconstituted in lipid vesicles according to Example 6. A serial dilution of the membranes containing receptor ranging from 0.2 to 20,000 Î¼g/mL (+/âˆ’ receptor) was added to wells on a microtiter plate and shaken gently for one hour at 4Â° C. After washing, a 1:1 to 1:10,000 serial dilution of a LacI-Gq lysate prepared from the LacI-Gq clone according to the methods described in Example 12 was added to the wells, the plate was shaken gently for one hour at 4Â° C., and washed. Anti-LacI antibodies (Stratagene) were added (1:1000) and the plate shaken gently for one hour at 4Â° C. After washing, HRP-conjugated goat anti-rabbit antibodies (Kierkegaard and Perry Laboratories) were added (1:2500) and the plate shaken gently for one hour at 4Â° C. The plate was washed, color was developed using horseradish peroxidase, and then read in an ELISA reader at OD₄₅₀. The general methodology for the ELISA is illustrated in FIG. 4. The results, see FIG. 5, show that the LacI-Gq fusion protein binds thrombin receptor in a concentration-dependent manner. The ability of the LacI-Gq fusion protein to bind the empty vesicles was significantly less than vesicles reconstituted with thrombin receptor.

Example 14 Screening in the Presence of a High Affinity Peptide

To identify peptides having even higher affinity to light-activated rhodopsin than those identified by the panning procedure described in Example 7, a high affinity peptide was included in the library incubations in rounds three and four. Peptide 8 (LLENLRDCGMF; SEQ ID NO:125) had been identified in the first screening as a peptide exhibiting binding to light-activated rhodopsin 1000-fold higher than the native sequence. Screening of the GÎ±t library was performed as in Example 7, except that 10 Î¼L 100 Î¼M (100 nM final concentration) peptide 8 was included in the wells in rounds three and four. This screen revealed several clones that both bind rhodopsin with very high affinity and stabilize it in its active form, metarhodopsin II. See Table XVIII, below. Comparing Tables IX and XVIII, it is clear using peptide 8 in the screen resulted in a change at position 341 to a neutral residue. Residues L344, C347 and G348 remained stable whether peptide 8 was included in the screen or not. Use of peptide 8 resulted in a higher incidence of isoleucine at position 340 (17% with native peptide versus 71% with peptide 8) and a lower incidence of glutamine at position 342 (67% with native peptide versus 29% with peptide 8) This type of information not only contributes to the discovery of highly potent analog peptides for use as drugs or drug screening compounds, but also furthers the understanding of the structural framework which underlies the sites of contact between GÎ± and receptor.

Binding assays performed on some of the clones identified in this way are shown in FIG. 6. All peptides identified using peptide 8 in the screening process bound with equal or greater affinity to light-activated rhodopsin as did peptide 8. Compare the first bar (HAP=peptide 8) with the remaining bars.

TABLE XVIIIExemplary Light-Activated Rhodopsin HighAffinity Sequences Identified in Screenswith Addition of Peptide 8.Clone No.SEQ ID NO:SequenceLibrary Sequence124IRENLKDCGLF Peptide 8125LLENLRDCGMF 3266ILENLKDCGLL 7213MLDNLKDCGLF 8216IVKNLEDCGLF 10218IRDNLKDCGLF 13217ISKNLRDCGLL 17212ILQNLKDCGLF 19206MLDNLKACGLF

Example 15 Subcloning into MBP Vectors and Preparation of MBP Crude Lysates

pELM3 was digested at room temperature with AgeI (New England Biolabs) and the cut vector was separated from uncut vector on a 0.7% agarose gel. DNA was purified (Qiagen Extract-a-gel kit) and digested with ScaI (New England Biolabs). The 5.6 kb MBP vector fragment was separated on a 1% agarose gel and purified as above. During the final affinity purification round of the peptide library, a 20 mL portion of the 200 mL amplification culture was set aside before harvesting the cells. This 20 mL portion was allowed to grow to saturation, usually overnight, and DNA was prepared from the cells (Qiagen midi-prep kit). The pJS142 plasmid DNA was digested with BspEI and ScaI. The 0.9 kb peptide-encoding fragment was separated from the 3.1 and 1.7 kb vector fragments on a 1% agarose gel and purified.

Different ratios of the 5.6 kb MBP vector fragment and the peptide-encoding 0.9 kb fragment (1:2, 1:1, 2.5:1, 5:1, 10:1) were ligated in ligase buffer containing 0.4 mM ATP at 14Â° C. overnight with T4 DNA ligase. The ligation was terminated by increasing the temperature to 65Â° C. for ten minutes. To lower the background, the ligation mixture was digested with XbaI before isopropanol precipitation using 1 Î¼L glycogen as a carrier. After one wash with 80% ethanol, the pellet was resuspended in 20 Î¼L double-distilled water. ARI814 cells were transformed as described in Example 7 using 1 Î¼L of the precipitated XbaI digested ligation mix. After allowing the cells to shake for one hour at 37Â° C. in 1 mL SOC, 100 Î¼L of the suspension was spread on LB-Amp Plates. Crude lysates were prepared as described for LacI lysates in Example 12.

Example 16 MBP-Peptide Fusion Protein Purification

An overnight culture (1 mL) of a single MBP-peptide fusion protein clone was inoculated into 200 mL LB-AMP media. The culture was shaken at 37Â° C. until OD₆₀₀=0.5. Protein expression was induced by addition of 150 Î¼L 1 M IPTG (final concentration 0.3 mM), with continued shaking at 37Â° C. for two hours. The culture then was sedimented at 5000Ã—g for 20 minutes and resuspended in 5 mM column buffer (10 mM Tris, pH 7.4; 200 mM NaCl; 1 mM EDTA; 1 mM DTT) and 16.25 Î¼L 0.2 M PMSF was added. The resuspended cell pellet was then stored at âˆ’70Â° C. The stored pellet was thawed in cold water and placed in an ice bath. The pellet was sonicated in short pulses of less than 15 seconds with a Fisher Scientific 55 Sonic Dismembrator (40% constant time, output 5, repeating five times with a total one minute duration). The sonicated pellet was subjected to centrifugation at 9000Ã—g for 30 minutes, after which the supernatant was saved and diluted to 100 mL using column buffer. Usually, the protein concentration was approximately 2.5 mg/mL. A column was prepared by pouring 7.5 ml amylose resin in a BioRad disposable column and washing with eight volumes of column buffer. The diluted crude extract was loaded by gravity flow at about 1 mL/min and the column was washed again with eight volumes of column buffer. The fusion protein was eluted with 10 mL 10 mM maltose in column buffer and concentrated using Amicon centriplus 30â„¢ columns, then aliquoted and stored at âˆ’70Â° C.

Example 17 Method for Screening Library Crude Lysates by ELISA

Microtiter wells were coated with 0.1-1.0 Î¼g/well rhodopsin receptor in a final volume of 100 Î¼L HEK containing 1 mM DTT with shaking at 4Â° C. for one hour. The wells then were blocked with bovine serum albumin (BSA) by adding 100 Î¼L 2% BSA in HEK with 1 mM DTT to the wells and continuing shaking at 4Â° C. for at least 30 minutes, then washed four times with HEK containing 1 mM DTT. Crude lysates were diluted 1:50 in HEK containing 1 mM DTT and added to the coated wells (100 Î¼L per well). The plates were shaken at 4Â° C. for one hour, washed four times with PBS/0.05% Tweenâ„¢20 1 mM maltose and then probed with 100 Î¼L 1:1000 rabbit anti-MBP antibodies (New England BioLabs) in PBS containing 0.05% Tweenâ„¢ 20 and 1 mM maltose, with shaking for 30 minutes at 4Â° C. After another wash, the wells were probed with 100 Î¼L 1:7500 goat anti-rabbit secondary antibodies conjugated to horseradish peroxidase in PBS containing 1% BSA and 1 mM maltose with shaking for 30 minutes at 4Â° C. The plate was washed four times with PBS containing 0.05% Tweenâ„¢ 20 and 1 mM maltose. Horseradish peroxidase substrate (Bio-Fx; 100 Î¼L) was added and the color developed for 20-30 minutes. The reaction was stopped by addition of 100 Î¼L 2N sulfuric acid and the plate read at OD₄₅₀. If the color reaction occurred too quickly (less than 10 minutes) or if the background in negative control wells was too high (greater than 0.2) the assay was repeated using 1:100 or 1:200 dilutions of the crude lysates.

Example 18 Binding Assay of High Affinity Rhodopsin Binding Peptide Fusion Proteins

The entire population of peptide-coding sequences identified in round 4 of panning (see Example 7) was transferred from pJS142 to pELM3 (New England Biolabs). This plasmid is a pMal-c2 derivative with a modified polylinker, inducible by isopropyl Î²-thiogalacto-pyranoside and containing the E. coli malE gene with a deleted leader sequence and leads to cytoplasmic expression of MBP fusion proteins. The MBP-carboxyl terminal peptide analog fusion proteins were expressed in E. coli.

For the assay, in the dark, 1 Î¼g/well of ROS membranes (rhodopsin) as described in Example 5 was directly immobilized on microtiter wells in cold HEK/DTT for one hour at 4Â° C. The wells were rinsed, blocked with 1% BSA in HEK/DTT for one hour at 4Â° C. and rinsed again. Bound rhodopsin was activated by exposure to light for 5 minutes on ice before addition of the MBP fusion proteins (crude bacterial lysates were diluted 1:50 in HEK with 1 Î¼M dithiothreitol; purified proteins were used at 0.2-120 nM). The MBP-GÎ±t340-350K341R (pELM17) fusion protein and MBP with linker sequence only (pELM6) were present in control wells at 50 nM final concentration. After 30 minutes, wells were washed and rabbit anti-MBP antibody (New England Biolabs) was added. The anti-MBP antibody was used at a 1:1000 dilution for crude lysates and a 1:3000 dilution for purified proteins. After 30 minutes, wells were rewashed and goat anti-rabbit antibody conjugated to horseradish peroxidase (1:7500 dilution for crude lysates; 1:10,000 dilution for purified proteins; Kierkagaard & Perry Laboratories) was added. After 30 minutes, the plate was washed four times with PBS containing 0.05% Tweenâ„¢20. Horseradish peroxidase substrate (100 Î¼l) was added and color was allowed to develop for about 20 minutes. The reaction was stopped by addition of 100 Î¼l 2N sulfuric acid. The results are presented in FIG. 7. Values indicate absorbance at OD₄₅₀. The positive control for the assay was pELM 17, which encodes the MBP fusion protein GÎ±_t340-350K341R. pELM6, which expresses MBP protein fused to a linker sequence only, served as the negative control. â€œNo lysateâ€ control wells were included to reflect any intrinsic, non-specific binding within the assay. See FIG. 7.

The IC₅₀values of the high affinity MBP fusion proteins ranged from 3.8 to 42 nM, up to 3 orders of magnitude more potent than the 6 Î¼M IC₅₀of MBP-GÎ±t340-350K341R. In all the highest affinity sequences, position 341, which is a positively charged residue in the native sequence, was changed to a neutral residue. Leu344, Cys347, and Gly348 were found to be invariant and hydrophobic residues were always located at positions 340, 349, and 350, indicating the critical nature of these residues.

Example 19 Binding of High Affinity Peptide Fusion Proteins to Rhodopsin Can Be Competitively Inhibited by Heterotrimeric Gt

When light-activated rhodopsin was screened for peptides based on the C-terminus of Gt, a large number of high-affinity sequences were obtained. Binding of MBP fusion proteins containing the high affinity peptide from the library (sequences from clones 8, 9, 10, 18, 23, 24, as well as pELM17 which encodes the wild-type peptide sequence, and pELM6 which contains no peptide; MBP-8, MBP-9, MBP-10, MBP-18, MBP-23, MBP-24, MBP-pELM17) were assessed for their ability to bind rhodopsin (0.5 Î¼g rhodopsin/well) in the presence or absence of heterotrimeric Gt.

Lysate (50 Î¼l) from each clone was added and incubated in the light. After 45 minutes, 1 Î¼M heterotrimeric Gt was added and the solution incubated for 30 minutes. Anti-MBP antibody was added, followed by goat anti-rabbit alkaline phosphatase conjugated antibody and substrate. The color was allowed to develop. Absorbance data are presented in FIG. 8.

Most peptide sequences obtained were highly homologous to the native GÎ±t C-terminal sequence. Several of these sequences are of very high affinity (>1000-fold higher than the parent peptide) and are potent and specific antagonists of receptor-mediated G protein activation. The high-affinity peptide fusion proteins were tested for binding to light-activated rhodopsin and for their ability to stabilize the MII conformation (Table XXI).

The screen used MBP-8 because this peptide bound to rhodopsin with high affinity and stabilized MII. MBP-18 and MBP-24, which both showed even higher binding affinities than did MBP-8 to rhodopsin, were not used because the affinity was so high that the small molecules might not have been able to competitively inhibit their binding. Of course, the screen may be repeated using another peptide as is convenient, for example peptides that are of higher affinity to find even more potent small molecules.

TABLE XIXAbsorbance at OD₄₅₀in a Panning ELISA andEC50 values for MII binding and MIIStabilization for Selected MBP-High AffinityPeptide Fusion Proteins.SEQMIIMIIIDELISAbindingstabilizationNO:OD₄₅₀EC₅₀EC₅₀GtIKENLKDCGLF15.016000nM>100Î¼M 9LQQVLKDCGLL267.3510nM1.05Î¼M 10ICENLKDCGLF215.3642nM5.40Î¼M 8LLENLRDCGMI268.547.8nM0.94Î¼M 18MLENLKDCGLF128.583.8nM1.24Î¼M 24MLKNLKDCGMF130.616.6nM0.49Î¼M 23VLEDLKSCGLF129.6620nM3.50Î¼M

Heterotrimeric Gt competitively inhibited high affinity peptide fusion protein binding to light-activated rhodosin. See FIG. 8. The heterotrimeric Gt contains multiple determinants of rhodopsin binding and is membrane bound via myristoylation of the Î± subunit and farnesylation of the Î³ subunit carboxyl terminus. Thus, the selected peptide sequences from the combinatorial library bind to the receptor with very high affinity.

Example 20 Binding of MBP Clones to PAR1

To identify high affinity peptides that bind PAR1, membranes prepared from mammalian cells (Chinese hamster ovary) overexpressing PAR1 were panned with the G11 peptide library. ELISA binding affinity results of selected clones are shown in FIG. 9 for their binding to membranes prepared from SF9 cells expressing either PAR1 or the Gq-coupled muscarinic M1 receptor. To quantitate the binding, purified MBP clones were analyzed using ELISA methods in which the secondary antibody was conjugated to HRP. The binding for the control MBP-Gq fusion protein is shown. See FIG. 9. The data are the average of two separate experiments done in duplicate. MBP clones PAR-13 and PAR-34 both show both high affinity binding for PAR1 as well as specificity. MBP clones PAR-23 and PAR-33 appear to be both of low affinity and low specificity. See Table XIII for the sequences.

Example 21 Binding Specificity of LacI-Peptide Fusion Proteins

PAR1 binding clones of LacI-peptide fusion protein selected from the G11 Library were diluted 1:100 in HEK/DTT and tested for dose-responsive binding to Sf9 insect cell membranes from cells expressing no receptor, the M1 receptor (which couples to Gi) or PAR1 receptor, prepared according to Example 3. Increasing amounts of membrane as indicated in FIG. 10 were coated in microtiter wells, incubated and rinsed. LacI-peptide fusion protein lysates were added, incubated and rinsed, and the receptor-bound LacI-peptide fusion protein was measured as described above using a LacI antibody. Results for a single, representative clone are presented in FIG. 10, and demonstrate the specificity of the selected peptides for PAR1.

Example 22 Binding of Native GÎ±q-Maltose Binding Protein-Peptide Fusion Protein to PAR1

Microtiter wells were coated with purified, reconstituted PAR1 in the presence of 100 nmoles thrombin receptor activating peptide, as described above in Example 6. Purified maltose binding protein-GÎ±q (MBP-GÎ±q) was added at the concentrations indicated in FIG. 11 and incubated one hour on a shaker at 4Â° C. The wells were rinsed and then probed with a rabbit anti-maltose binding protein antibody, followed by alkaline phosphatase conjugated secondary antibodies, as described above. Substrate was added and the color was allowed to develop about 20 minutes. Absorbance at 405 nm was measured and dose-response curves were calculated using GraphPad Prism (version 2.0). See results in FIG. 11. The calculated IC₅₀of GÎ±q binding to activated PAR1 was 214 nM.

Example 23 Design of Oligonucleotides for GÎ± Peptide Minigene Constructs

cDNA encoding the last 11 amino acids of GÎ± subunits was synthesized (Great American Gene Company) with newly engineered 5â€²- and 3â€²-ends. The 5â€²-end contained a BamHI restriction enzyme site followed by the human ribosome-binding consensus sequence (5â€²-GCCGCCACC-3â€²; SEQ ID NO:269), a methionine codon (ATG) for translation initiation, and a glycine codon (GGA) to protect the ribosome binding site during translation and the nascent peptide against proteolytic degradation. A HindIII restriction enzyme site was synthesized at the 3â€² end immediately following the translational stop codon (TGA). Thus, the full-length 56 bp oligonucleotide for the GiÎ±_1/2carboxyl terminal sequence was 5â€²-gatccgccgccaccatgggaatcaagaacaacctgaaggactgcggcctcttctgaa-3â€² (SEQ ID NO:270) and the complimentary strand was 5â€²-agctttcagaagaggccgcagtccttcaggttgttcttgattcccatggtggcggcg-3â€² (SEQ ID NO:271). See FIG. 12. As a control, oligonucleotides encoding the GÎ±i_1/2carboxyl terminus in random order (GÎ±iR) with newly engineered 5â€²- and 3â€²-ends also were synthesized.

The DNA was diluted in sterile ddH₂O to form a stock concentration at 100 Î¼M. Complimentary DNA was annealed in 1Ã— NEBuffer 3 (50 mM Tris-HCl, 10 mM MgCl₂, 100 mM NaCl, 1 mM DTT; New England Biolabs) at 85Â° C. for 10 minutes then allowed to cool slowly to room temperature. The DNA then was subjected to 4% agarose gel electrophoresis and the annealed band was excised. DNA was purified from the band using a kit, according to the manufacture's protocol (GeneClean II Kit, Bio101). After digestion with each restriction enzyme, the pcDNA 3.1(âˆ’) plasmid vector was subjected to 0.8% agarose gel electrophoresis, the appropriate band cut out, and the DNA purified as above (GeneClean II Kit, Bio101). The annealed/cleaned cDNA was ligated for 1 hour at room temperature into the cut/cleaned pcDNA 3.1 Î¼lasmid vector (Invitrogen) previously cut with BamHI and HindIII.

For the ligation reaction, several different ratios of insert to vector cDNA (ranging from 25 Î¼M:25 pM to 250 pM:25 pM annealed cDNA) were plated. Following the ligation reaction, the samples were heated to 65Â° C. for 5 min to deactivate the T4 DNA ligase. The ligation mixture (1 Î¼l) was electroporated into 50 Î¼l competent cells as described in Example 7 and the cells immediately placed into 1 ml of SOC (Gibco). After 1 hour shaking at 37Â° C., 100 Î¼l of the electroporated cells containing the minigene plasmid DNA was spread on LB/Amp plates and incubated at 37Â° C. for 12-16 hours. To verify that insert was present, colonies were grown overnight in LB/Amp and their plasmid DNA purified (Qiagen SpinKit). The plasmid DNA was digested with Ncol (New England Biolabs, Inc.) for 1 hour at 37Â° C. and subjected to 1.5% (3:1) agarose gel electrophoresis. Vector alone produced 3 bands. When the 56 bp annealed oligonucleotide insert is present, there is a new NcoI site resulting in a shift in the band pattern such that the digest pattern goes from three bands (3345 bp, 1352 bp, 735 bp) to four bands (3345 bp, 1011 bp, 735 bp, 380 bp). See FIG. 13. DNA with the correct electrophoresis pattern was sequenced to confirm the appropriate sequence. This method may be used to insert any high affinity peptide to create a minigene constant.

Example 24 Expression of Peptides from Minigene Constructs

Expression of the GPCR binding peptides was achieved using constructs which included minigene inserts corresponding to the carboxyl terminal sequences of various G protein Î± subunits (GÎ±i, GÎ±o, GÎ±s, GÎ±q, GÎ±11, GÎ±12, GÎ±13, GÎ±14), as well as a control minigene containing the GÎ±i sequence in random order (GÎ±iR). The minigene insert DNAs were made by synthesizing short complimentary oligonucleotides corresponding to the peptide sequences from the carboxyl terminus of each GÎ± with BamHI and HindIII restriction sites at the 5â€² and 3â€² ends, respectively. Complementary oligonucleotides were annealed and ligated into the mammalian expression vector pcDNBA3.1 according to the methods of Gilchrist et al., J. Biol. Chem. 274:6610-6616, 1999, the disclosures of which are hereby incorporated by reference.

Human embryonic kidney (HEK) 293 cells were transfected using a standard calcium phosphate procedure according to the methods of Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York, vol. 1-3 (1989), the disclosures of which are hereby incorporated by reference. To confirm the transcription of minigene constructs in transfected cells, total RNA was isolated from the cells 48 hours post transfection with pcDNA-GÎ±i or pcDNA-GÎ±iR using methods known in the art. Reverse transcriptase PCR was used to make cDNA and PCR analysis was performed using the cDNA as template with primers specific for the relevant GÎ± carboxyl terminal peptide insert (forward: 5â€²-ATCCGCCGCCACCATGGGA (SEQ ID NO:272); reverse: 5â€²-GCGAAAGGAGCGGGGCGCTA (SEQ ID NO:273)). These primers for the GÎ± minigenes amplify a 434 bp fragment only if the inserted peptide-encoding oligonucleotides are present; no band is observed in cells transfected with the empty pcDNA3.1 vector. The PCR products were separated on 1.5% agarose gels. The presence of a single 434 bp band indicated that GÎ± carboxyl terminus peptide minigene RNA had been transcribed. See FIG. 14. Control experiments were done using a T7 forward primer with the vector reverse primer to verify the presence of the pcDNA3.1 vector, and G3DPH primers (Clonetech) to approximate the amount of total RNA.

To verify that the peptide was being produced in the transfected cells, the cells were lysed and homogenized 48 hours post transfection according to known methods. Cytosolic extracts were analyzed by gradient reversed phase HPLC as follows: 100 Î¼L of cytosolic fraction extract was loaded onto a C4 column (Vydac) equilibrated with 0.1% TFA in ddH₂O. The peptide was eluted using 0.1% TFA in an acetonitrile gradient which increased from 0-60% over 45 minutes. Peaks were collected, lyophilized, and analyzed using ion mass spray analysis (University of Illinois-Urbana Champagne). Mass spectrometry analysis for peak 1 from GÎ±_1/2peptide vector (pcDNA-GÎ±i) transfected cells, and from cells transfected with pcDNA-GÎ±iR indicated that a 1450 Dalton peptide (the expected molecular weight for both 13 amino acid peptide sequences) was present in each cytosolic extract. The minigene-encoded peptides were the major peptides found in the cytosol, strongly indicating that the vectors produced the appropriate peptide sequences in large amounts.

Example 25 Interfacial G Protein Peptide Inhibition of Thrombin-Mediated Inositol Phosphate Accumulation

HMEC were seeded onto 6-well plates 24 hours before transfection at 1Ã—10⁵well. Cells were transiently transfected with pcDNA3.1, pcDNA-GÎ±i, pcDNA-GÎ±iR, or pcDNA-Gq as described in Example 24. After 24 hours, cells were incubated in 2 mL culture medium containing 4 Î¼Ci/mL [³H]-myoinositol to obtain steady-state labeling of cellular inositol lipids. Transiently transfected cells were assayed for inositol phosphate (IP) accumulation 48 hours after transfection. Two hours prior to stimulation with Î±-thrombin, cells were washed, and medium replaced with medium containing 5 mM LiCl. Cells were stimulated with 10 nM Î±-thrombin for 10 minutes. Inositol phosphate (IP) formation was stopped by aspiration of the medium and addition of ice-cold methanol (final concentration 5%).

Perchloric acid-lysed cells were centrifuged at 2500 rpm, 4Â° C. for 5 min. The supernatant containing IP was eluted through a Poly-Prep chromatography column (Bio-Rad) containing 1.6 ml anion exchange resin (DOWEX AG1-X8, formate form, 200-400 mesh). The perchloric acid-precipitated pellets (containing phosphatidylinositols and lipids) were resuspended in 1 ml chloroform-methanol-10 M HCl (200:100:1, v/v/v). These suspensions were mixed with 350 Î¼L HCl and 350 Î¼L chloroform and sedimented for 5 min at 2500 rpm to separate the phases. The lower, hydrophobic phase was recovered and dried in counting vials to determine the amount of radioactivity in total phosphatidylinositols. The relative amount of [³H]-IP generated was calculated as follows: ([³H]-IP (cpm)/[³H]-IP (cpm)+[³H]-inositol (cpm)). Each value was normalized using the basal value (no thrombin stimulation) obtained in pcDNA transfected cells. See FIG. 15. The results presented are the normalized meanÂ±SEM of at least 3 independent experiments performed in triplicate. The ** symbol indicates p<0.005. Results indicate that addition of thrombin increased IP production in control cells (pcDNA, pcDNA-GiR). Cells transfected with PcDNA-Gq had no thrombin-mediated IP production increase, while cells transfected with pcDNA-Gi had a normal response. This indicates that transfection of the Gq C-terminal minigene vector into HMEC with subsequent expression of the Gq C-terminal peptide can inhibit thrombin-mediated increases in IP.

Example 26 Interfacial G Protein Peptide Inhibition of Thrombin-Induced P1 Hydrolysis and Intracellular Ca⁺⁺ Rise

To determine whether expression of the GÎ±q C-terminal minigene vector could affect intracellular [Ca⁺⁺]_ilevels, HMEC were transfected with empty vector (pcDNA) or with pcDNA-GÎ±i, pcDNA-GÎ±q, or pcDNA-GÎ±iR minigene DNA (1 Î¼g), which encode high affinity peptides identified for their ability to bind the receptors. Transfected cells were seeded at a low confluency on coverslips in a 24-well plate 48 hours post transfection. The cells were allowed to adhere for two hours. The medium was aspirated and each coverslip was incubated with 10 Î¼M Oregon Green 488 BAPTA-1 acetoxymethyl ester (a calcium-sensitive dye) and 0.1% (v/v) Pluronic F-127 and allowed to incubate for 20-30 minutes at 37Â° C., then rinsed twice with wash buffer. Basal conditions were established before addition of thrombin (Ëœ70 nM) in Ca⁺⁺ buffer. Recordings were made every 10 seconds and continued for 170 seconds after stimulation with thrombin. Images were quantitated using NIH Image. Data from at least 70 individually recorded cells were used to calculate the changes in fluorescence (y-axis). See FIG. 16A, which presents fluorescence in ([Ca⁺⁺]_ilevel) increase 30 seconds after thrombin addition. Each bar in FIG. 16A represents the mean ((F_Sâˆ’F_B/F_Bâˆ’1)Â±SEM of over 70 individually recorded cells. The ** symbol indicates p<0.005. FIG. 16B shows the kinetics of [Ca⁺⁺]_i; fluorescence changes after cell stimulation with thrombin. Data presented are the mean ((F_Sâˆ’F_B/F_Bâˆ’1)Â±SEM at each recording point for cells transfected with CDNA for the empty vector (pcDNA) or the Gq C-terminal minigene vector (pcDNA-GÎ±q). The arrow indicates the time thrombin was added. Each time point represents over 100 individually recorded cells.

As shown in FIG. 16, following cell activation by addition of thrombin there was a transient increase in intracellular [Ca⁺⁺]_ilevels. Thirty seconds after the addition of thrombin, cells transfected with pcDNA-GÎ±q had a calcium response that was 44% decreased as compared to cells transfected with pcDNA (FIG. 16A). pcDNA-GÎ±q transfected cells had a 45% decrease compared to those transfected with pcDNA when all time points measured after thrombin stimulation are averaged (FIG. 16B). This decrease appears to be specific as cells transfected with pcDNA-GÎ±i or pcDNA-GÎ±iR did not have any effect on thrombin stimulated [Ca⁺⁺]_ilevels. Thus, cells expressing the GÎ±q C-terminal peptide appear to be inhibited in their ability to stimulate [Ca⁺⁺]_ilevels following activation with thrombin, indicating a specific block of this downstream mediator by expression of GÎ±q.

pcDNA, pcDNA-GiR, pcDNA-Gi, pcDNA-Gq, or pcDNA-Gs minigene constructs were transfected into HMEC and used to assay inositol phosphate (IP) accumulation. After 24 hours, cells were reseeded onto 24-well plates and labeled with [³H]-myoinositol (2 Î¼Ci/ml). After 48 hours, cells were rinsed, and incubated with or without thrombin (10 nM) for 10 minutes. Total IP accumulation was assayed as described above using Dowexâ„¢ columns to separate [³H]IP. The relative amount of [³H]IP generated was calculated as follows: ([³H]IP (cpm)/[³H]IP (cpm)+[³H]inositol (cpm)). Each value was normalized by the basal value (no thrombin stimulation) obtained in pcDNA transfected cells. See FIG. 17. The results presented are the normalized meanÂ±SEM of at least three independent experiments performed in triplicate. The ** symbol indicates p<0.005.

Example 27 Prevention of Thrombin-Induced MAPK Activity by GPCR-binding C-terminal Peptides

Hemagglutanin (HA)-MAPK (1Ã—10⁵/mL was co-transfected into HMEC with the pcDNA, pcDNA-GÎ±i, pcDNA-GÎ±q or pcDNA-GÎ±iR minigene constructs using the methods described in Example 24. After 30 hours, cells were serum-starved for 18 hours and then treated with 10 nM thrombin for 20 minutes. Cells were then lysed with RIPA buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 2 mM EDTA, 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS, 10% glycerol, 10 Î¼g/mL aprotinin and 10 Î¼g/mL leupeptin) and HA-MAPK protein immunoprecipitated using 12CA5 antibody (Roche Molecular Biochemicals; Indianapolis, Ind.) (one hour, 4Â° C.) and Protein A sepharose beads (three hours, 4Â° C.). Immune complexes were washed three times in RIPA buffer. Kinase activity in the immunoprecipitates was measured using maltose binding protein (MBP) substrate and a kinase assay kit (Upstate Biotechnology, Inc., Lake Placid, N.Y.). MAPK activity (nmol/min/mg) was obtained for each, and the relative increase of MAPK activity (thrombin-mediated fold increase) was calculated as follows: (stimulated activity (nmol/min/mg)âˆ’basal activity (nmol/min/mg))/basal activity (nmol/min/mg). Results are presented as the meanÂ±SEM of at least three independent experiments in FIG. 18. A * symbol indicates p<0.05.

Addition of 10 nmol thrombin resulted in a 3.66 fold increase in HA-MAPK activity in cells transfected with the pcDNA control vector. Similarly, cells transfected with pcDNA-GiR had an essentially equivalent increase in thrombin mediated MAPK activity with (4.46 fold increase). However, endothelial cells transfected with a minigene construct encoding the GÎ±i, GÎ±q, GÎ±12 or GÎ±13 GPCR binding peptides showed a significant decrease in thrombin-mediated HA-MAPK activity (59%, 57%, 50% and 77%, respectively) compared to cells transfected with pcDNA.

Example 28 Reduction of Thrombin-Induced Transendothelial Electrical Resistance

Transendothelial electrical resistance (TEER) was measured by passing an alternating current (50 Î¼A; 2 pulses every minute) across monolayers of HMEC expressing GÎ±q, GÎ±i, GÎ±iR or no minigene construct. Basal TEER did not change significantly with minigene transfection. Upon addition of 10 nM thrombin, however, there was a decrease in the TEER of cells expressing the GÎ±q minigene compared to non-transfected cells in the presence of 10 nM thrombin. See FIG. 19 (representative of multiple experiments). The decrease in transendothelial electrical resistance in response to thrombin was significantly reduced in endothelial cells transfected with the minigene for the carboxyl terminus of GÎ±q, while there was no effect in cells transfected with GÎ±i, GÎ±iR, or empty vector. These results suggested that GÎ±q is partially responsible for the effects of thrombin on endothelial cell shape changes.

Example 29 Inhibition of Thrombin-Mediated Stress Fiber Formation

HMEC cells were transfected with pcDNA, pcDNA-GÎ±12 or pcDNA-GÎ±13 minigene constructs 1 Î¼g each/100 mm dish. As a marker for transfected cells, the pGreenLantern-1 plasmid, containing the gene for green fluorescent protein (GFP) was co-transfected together with minigene constructs. After 48 hours, cells were serum starved for 18 hours and treated with 10 nM thrombin for 20 minutes. After exposure to thrombin, the cells were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 and stained for F-actin with 1 mM rhodamine-phalloidin for 30 minutes. Cells were extensively washed, mounted using Vectashieldâ„¢ antifade mounting medium (Vector Laboratories, Inc.). Cells were observed with an inverted microscope (Diaphot 200, Nikon, Inc.) equipped for both differential interference contrast microscopy and epifluorescence observation using a 60x oil-immersion objective. Fluorescence and DIC images were recorded for each cell field with a cooled, integrating CCD array camera (Imagepoint, Photometrix, Ltd.) connected to the microscope. See FIG. 20 for fluorescence images showing inhibition of thrombin-mediated stress fiber formation by GÎ±12 and GÎ±13 peptides.

Serum-starved cells transfected with pcDNA exhibited a thin cortical F-actin rim at their margins, and contained few stress fibers (FIG. 20, panel A). Those present were inconspicious and in apparently random orientation. For HMEC transfected with pcDNA after a 20-minute exposure to thrombin actin had reorganized into prominent stress fibers, typically arranged in a parallel pattern along the longitudinal axis of the cell (FIG. 20, panel B). A very different pattern was observed for cells transfected with pcDNA-GÎ±12 (FIG. 20, panel C) or pcDNA-GÎ±13 (FIG. 20, panel D) minigenes after exposure to thrombin. In both pcDNA-GÎ±12 and pcDNA-GÎ±13 transfected cells, thrombin stimulation did not result in the appearance of stress fibers. In cells transfected with pcDNA-GÎ±13, the peripheral actin rim appears thicker and more linear, providing a clear outline of cell-cell junctions. Thus, in agreement with earlier reports, thrombin induced rapid stress fiber formation in endothelial cells. Transfection of either pcDNA-GÎ±12 or pcDNA-GÎ±13 minigenes resulted in cells that no longer showed thrombin-induced stress fiber formation. Given that stress fiber formation is dependent on the small GTPase Rho, these results concur with other findings that GÎ±12 and GÎ±13 are intimately linked to Rho signaling and demonstrates the ability of GPCR binding peptides to specifically block this G protein pathway when expressed intracellularly.

Example 30 Inhibition of G Protein Activity by GPCR Binding Peptides in Single Intact Cells

Human embryonic kidney (HEK) 293 cells, which stably express the M₂mACR (Ëœ400 fmol receptor/mg protein), were grown in DMEM (Gibco) supplemented with 10% fetal bovine serum (Gibco), streptomycin/penicillin (100 U each; Gibco) and G418 (500 mg/L; Gibco). Cells were grown under 10% CO₂at 37Â° C. In all transfections for electrophysiological studies, the CD8 reporter gene system was used to visualize transfected cells using Dynabeadsâ„¢ coated with anti-CD8-antibodies (Dynal). The following amounts of cDNA were used to transfect the cells: pC1-GIRK1 (rat)â€”1 Î¼g; nH3-CD* (human)â€”1 Î¼g; pcDNA3.1, pcDNA-GÎ±i, pcDNA-GÎ±iR, pcDNA-GÎ±q, or pcDNA-GÎ±sâ€”4 Î¼g. Thus, typically the total amount of cDNA used for transfecting one 10 cm disk was 7 Î¼g. The cDNAs for GIRK1 and GIRK4 were gifts from F. Lesage and M. Lazdunski (Nice, France). A standard calcium phosphate procedure was used for transient transfection of HEK cells according to the methods of Schenborn et al., Meth. Mol. Biol. 130:135-145, 2000. All assays were performed 48-72 hours post transfection.

Whole cell currents were recorded from stably M₂mAChR-expressing HEK 293 cells that had been transiently transfected with cDNA for GIRK1, GIRK4 and either pcDNA-GÎ±i, pcDNA-GÎ±s, or pcDNA-GÎ±q. For the measurement of inwardly rectifying K⁺ currents, whole cell currents were recorded using an extracellular solution contained 120 mM NaCl; 20 mM KCl; 2 mM CaCl₂; 1 mM MgCl₂; and 10 mM HEPES-NaOH, pH 7.4. The solution for filling the patch pipettes was composed of 100 mM potassium glutamate; 40 mM KCl; 5 mM MgATP; 10 mM HEPES-KOH, pH 7.4; 5 mM NaCl; 2 mM EGTA; 1 mM MgCl₂; and 0.01 mM GTP. Membrane currents were recorded under voltage clamp, using conventional whole cell patch techniques. See Bunemann et al., J. Physiol. 489:701-777, 1995 and Bunemann et al., J. Physiol. 482:81-89, 1995, the disclosures of which are hereby incorporated by reference. To minimize variations due to different transfections or culture conditions, control experiments (transfection with pcDNA-GÎ±iR) were done in parallel. Patch-pipettes were fabricated from borosilicate glass capillaries, (GF-150-10, Warner Instrument Corp.) using a horizontal puller (P-95 Fleming & Poulsen). The DC resistance of the filled pipettes ranged from 3-6 MÎ©.

Membrane currents were recorded using a patch-clamp amplifier (Axopatch 200, Axon Instruments). Signals are analog-filtered using a lowpass Bessel filter (1-3 kHz corner frequency). Data were digitally stored using an IBM compatible PC equipped with a hardware/software package (ISO2 by MFK, Frankfurt/Main, Germany) for voltage control, data acquisition and data evaluation. To measure K⁺ currents in the inward direction, the potassium equilibrium potential was set to about âˆ’50 mV and the holding potential was âˆ’90 mV. Agonist-induced currents were evoked by application of acetylcholine (ACh; 1 Î¼M) using a solenoid operated superfusion device which allowed for solution exchange within 300 mseconds. Linear voltage ramps (from âˆ’120 mV to +60 mV within 500 mseconds) were applied every 10 seconds. By subtracting non-agonist dependent currents, the current voltage properties of the agonist induced currents could be resolved. To exclude experiments in which currents were recorded from cells that may not have expressed the functional channel, only those cells that exhibited a basal non-agonist dependent Ba⁺⁺ (200 Î¼M) sensitive inwardly rectifying current were used for analysis. For analysis of the data, the maximal current density (peak amplitude) of ACh-induced inwardly rectifying K⁺ currents was measured at âˆ’80 mV and compared.

Superfusion of HEK 293 cells transiently transfected with GIRK1/GIRK4 and either pcDNA-Gi or pcDNA-GiR DNA with 1 Î¼M ACh revealed that cells transfected with pcDNA-GÎ±i DNA have a dramatically impaired response to the M₂mAChR agonist. See FIG. 21, which summarizes data showing the maximum amplitude of ACh evoked currents for the different transfection conditions (cells transfected with GIRK1/GIRK4 and pcDNA-Gi or cells transfected with GIRK1/GIRK4 and pcDNA-GiR). The pcDNA-Gi minigene vector results in high intracellular expression of the GÎ±i peptide, leading to diminished ability of the receptor to signal the heterotrimenic GÎ±i.

The maximum current evoked by ACh was 3.7+/âˆ’1.5 pA/pF (n=14) in cells transfected with pcDNA-Gi, compared to 24.1+/âˆ’8.8 pA/pF (n=11) in cells transfected with pcDNA-GiR. This indicates that the GÎ±i minigene construct completely blocked the agonist mediated M₂mAChR GIRK1/GIRK4 response while the control minigene construct (pcDNA-GiR) had no effect. Compare FIG. 21A to FIGS. 21B and 21C. Cells transfected with minigene constructs encoding GÎ± carboxyl termini for GÎ±q or GÎ±s pcDNA-GÎ±q or pCDNA-GÎ±s were not significantly different than those of cells transfected with the control vectors. These findings confirm the specificity of the inhibition of M₂mAChR-activated G protein-coupled inwardly rectifying K⁺ current responses by expression of the GÎ±i minigene.

Example 31 Selective G Protein Signaling Inhibition in Human Microvascular Endothelial Cells

Different measures of G-protein signaling final actions were assayed in human microvascular endothelial cells (HMEC) which natively express the thrombin receptor, PAR1. The cells were seeded onto 6-well plates at 1Ã—10⁵cells/well and transiently transfected after 24 hours with minigene constructs containing GÎ± carboxyl terminal peptides (pcDNA, pcDNA-GÎ±i, or pcDNA-GÎ±iR; 1 Î¼g per well) using Effectene (Qiagen) according to the manufacturer's protocol. After 24 hours, the cells were labeled with 3 Î¼Ci/ml [³H]-adenine for 30 minutes at 37Â° C. After another 24 hours, the cells were washed with serum-free medium containing 1 mM isobutyl-methyl xanthine. To stimulate cAMP accumulation, cells were treated with 1 Î¼M isoproterenol for 30 minutes at 37Â° C. To see the inhibitory effects of thrombin on cAMP accumulation, cells were pretreated with thrombin (50 nM) for 15 minutes prior to addition of isoproterenol. The reactions were terminated by aspiration of media followed by addition of ice-cold 5% trichloroacetic acid.

Results are provided in FIG. 22 as (cAMP/cAMP+ATP) X 1000. Three separate experiments were done in duplicate. The ** symbol indicates p<0.005. Basal cAMP levels were essentially equivalent for all conditions tested. Endothelial cells stimulated with isoproterenol to activate Î²-adrenergic receptors increase their cAMP levels through the Gs pathway. Cells transfected with pcDNA, pcDNA-GÎ±i, or pcDNA-GÎ±iR showed little difference with 82-, 64-, and 77-fold increases in isoproterenol-mediated cAMP accumulation, respectively. When the endothelial cells were pre-incubated with thrombin prior to addition of isoproterenol, a decrease in cAMP levels was observed due to thrombin activation of the Gi pathway. Endothelial cells transfected with pcDNA and pre-incubated with thrombin showed a 39% decrease in cAMP level over cells stimulated with only isoproterenol. Similarly, cells transfected with pcDNA-GÎ±iR and pre-incubated with thrombin showed had a 43% decrease over cells stimulated with only isoproterenol. However, cells transfected with pcDNA-GÎ±i and pre-incubated with thrombin had only a 0.1% decrease in cAMP levels as compared to cells stimulated with only isoproterenol. Thus, cells expressing the GÎ±i C-terminal peptide appear to be unable to inhibit adenyl cyclase following activation with thrombin, indicating that thrombin-mediated Gi signaling was specifically blocked by expression of the pcDNA-GÎ±i minigene.

Example 32 Screening Method to Identify Inverse Agonists

Urea-washed rod outer segment membrane fragments containing rhodopsin receptor are immobilized onto microtiter wells and blocked as described in Example 7. The receptor is light-activated. Labeled native GÎ±t carboxyl terminal peptide is added to each well and allowed to shake gently for one hour at 4Â° C. The wells are washed to remove unbound peptide. Crude bacterial lysates (labeled) from a GÎ±t carboxyl terminal peptide prepared according to the methods described in Example 7 (200 Î¼L) are added to each well and incubated with shaking for one hour at 4Â° C.

The wells then are washed to remove unbound label. The supernatants or well-bound labels are quantitated by ELISA to detect dissociation of labeled native peptide from the receptor after incubation with library peptides compared to control wells incubated in the absence of library peptides.

Example 33 Small Molecule Library Screening Method

Small molecule libraries are screened for inhibition of GPCR-mediated G protein signaling as follows. PAR1 thrombin receptor prepared from insect cells according to Example 3 are immobilized onto microtiter wells, blocked and washed. A small molecule library purchased from Chem Div (San Diego, Calif.) are added simultaneously with MBP-peptide fusion protein (0.1-1000 nM) in a 96- or 384-well plate and allowed to shake for one hour at 4Â° C. Initial screens are performed with the small molecules at about 5-5000 nM. The wells are washed four times in cold PBS containing 0.05% Tween 20â„¢ and 1 mM maltose. The amount of maltose binding protein adhering to the wells is quantitated with anti-MBP antibodies as described in Example 17, versus control wells incubated without library compounds.

Example 34 Identification of Very High Affinity Activated Rhodopsin-Binding Gt-Based Peptides

A combinatorial peptide library based on the C-terminal sequence of Gt was constructed by introducing all possible mutations at each position, but with an overall bias toward the GÎ±t sequence with a K341R change and panned for high-affinity binding. See Martin et al., J. Biol. Chem. 271:361-366, 1996; Gilchrist et al., Methods Enzymol. 315:388-404, 2000 the disclosures of which are hereby incorporated by reference, and Examples 7 and 17 for methods used. Specific residues within the C-terminal sequence were highly conserved. Perhaps more interesting is not only the selection against the native amino acid at a given position (R341, the second residue in the peptide shown below) but the apparent selection for a specific amino acid at that location (leucine). See Table XX. Table XXI shows amino acid sequences obtained from screening dark-adapted bovine rhodopsin with the same combinatorial peptide library based on the C-terminal sequence of Gt. As observed with Gt, specific residues within the carboxyl terminal sequence were conserved and specific residues were selected against. Notably, at identical positions there are extreme differences between the selection of light-activated and dark-adapted rhodopsin (i.e., position C347) indicating that upon activation the receptor undergoes a conformational change unmasking new sites which the G protein can interact.

TABLE XXAlignment of the Highest-Affinity Amino AcidSequences Screened based on the C-TerminalSequence of Gt with Light-Activated Rhodopsin. [Image Omitted]

TABLE XXIAlignment of Amino Acid Sequences Screened basedon the C-Terminal Sequence of Gt with Dark-AdaptedRhodopsin. [Image Omitted]

In all the high affinity sequences selected for binding to the light adapted rhodopsin, position 341, which normally is a positively charged residue was changed to a neutral one. There is an obvious selection for a specific amino acid change from R to L. Peptides synthesized with this single change were assayed for high affinity binding, and the results are shown in Table XX. There was not a selection for a neutral amino acid at position 341 in peptides selected for binding to dark-adapted rhodopsin. See Table XXI. Arg is found 75% of the time. For peptides selected for light-activated rhodopsin, Leu344, Cys347, and Gly348 were found to be invariant, and hydrophobic residues were always located at positions 340, 349 and 350, indicating the critical nature of these residues. This differs considerably from the peptides selected for binding to dark-adapted rhodopsin, which did not show any invariant positions. Most striking is the apparent selection against the Cys347 position. Cys347 and Gly 348 both are part of a type IIâ€² Î²-turn which is required for MII stabilization. This suggests that a site on rhodopsin which is required to bind the Cys 347 of Gt is unmasked only after the receptor has received a photon of light and formed MII. See Gilchrist et al., Methods Enzymol. 15:388, 2000. Other works indicate that the critical nature of Cys347 for binding light-activated rhodopsin is due to its hydrophobicity. Aris et al., J. Biol. Chem. 276:2333, 2001. Replacement of Cys 347 with a hydrophobic amino acid (Cys347Met, Cys347Val and Cys347Abu)(Abu=2-aminobutyric acid) stabilizes MII to the same extent and with similar potency as the parent peptide. The apparent selection of Lys at 347 in the dark-adapted rhodopsin peptides clearly indicates that binding of the GÎ±t peptide to dark-adapted rhodopsin is very different from light-activated rhodopsin. These results show that the site on rhodopsin recognized by the C-terminal tail of Gt differs depending on whether the receptor is dark-adapted or light-activated. This implies that high affinity peptides selected for binding to light-activated rhodopsin only bind the activated state of the receptor and not dark-adapted receptor.

Example 35 Assays for Determining Peptide, Peptide-Fusion Protein or Small Molecule Affinities for Metarhodopsin II

For the â€œextra MIIâ€ assay, EDTA-washed rhodopsin (Example 5; 5 Î¼M) is incubated in a 50 mM HEPES buffer, pH 8.2, with 100 mM NaCl, 1 mM MgCl₂, and 1 mM DTT at 5.3Â° C., in the absence or presence of varying concentrations of Gt340-350 analogs or Gt. The sample is maintained at 5.4Â° C. using a water-jacketed and thermostated circulator cuvette holder in an SLM Aminco DW2000 spectrophotometer at 390 and 440 nm. A flash of light bleaching 10% of the rhodopsin is presented and after a 1 min incubation, a second spectrum is measured and the difference in spectrum calculated. â€œExtraâ€ MII is calculated as the difference between the absorbance at 390 and 440. Dose response curves of MII stabilization by Î±t340350 (Î»), mutant Î±340-350K341L(Î½), and heterotrimeric Gt (v) were analyzed by non-linear regression using the program GraphPad PRISM and are shown in FIG. 23.

For the MII decay assay, the absorbance spectra of EDTA-washed ROS (10 Î¼M) is measured in an SLM Aminco DW2000 spectrophotometer in 10 mM K₂PO₄, pH 6.5 containing 0.1 M KCl, 0.1 mM EDTA, 1 mM DTT, in the presence of peptide, fusion protein expressing high affinity peptide or small molecule. The spectra are measured in the dark, then completely bleached in room light. The spectra of the bleached sample is measured at intervals of 30 minutes over a 6 hour period.

Example 36 Analysis of Data From Small Molecule Library Screen

Competition ELISA assays were employed to screen a small molecule library (a 10,000 compound library representative of ChemDiv's Diverse Collection of drug-like molecules) for compounds that bind activated rhodopsin and increase/decrease the binding of MBP-8 high affinity peptide fusion protein. MBP 8 was selected based on its mid-range affinity. The screen may be repeated using an MBP which displays higher affinity and ability to stabilize MII (i.e., MBP 18; Table XIX). These types of screens may be used with libraries of any size, therefore it is possible to increase the size of the compound library by 10 fold or greater and continue screening for small molecule hits in a similar manner.

A software program that displays results of screening as a colorometric readout with a unique color coding that represents the amount of inhibition or stimulation of bound light-activated rhodopsin-bound peptide fusion protein is advantageous and preferred. Two representative 96-well plates in which light-activated-rhodopsin-bound MBP-8 high-affinity peptide fusion proteins were assayed for competitive binding by 80 different compounds. Experiments were done in duplicate, and the results of the two separate plates averaged.

Dose response curves of MII (FIG. 24) indicate that both PL_â€”0302 R3.C4 (Ïƒ), and heterotrimeric Gt (Î½) stabilize the active form of rhodopsin. EDTA-washed rhodopsin (5 Î¼M) was incubated in a 50 mM HEPES buffer, pH 8.2, with 100 mM NaCl, 1 mM MgCl₂, and 1 mM DTT at 5.3Â° C., and â€œextraâ€ MII was measured. For compounds that enhanced MBP-8 binding over 25% using the color coded readout, dose studies were performed to generate EC₅₀curves. Table XXII below provides the calculated EC₅₀for metarhodopsin II stabilization of each compound.

TABLE XXIIEC₅₀values for selected small molecules on thebinding of MBP-8 to MII.Small moleculeMWBinding of MBP-8Name(daltons)EC₅₀(Î¼M)PL_0568 R1.C5291.20.96PL_0551 R8.C1328.50.95PL_0894 R3.C7424.910.1PL_0302 R3.C4290.2711.9PL_1012 R2.C1433.55.12

Example 37 Very High Affinity Agonists for Rhodopsin Have No Effect on PAR-1-Stimulated Ca⁺⁺ Transients

Small molecules PL_â€”0568 R1.C5, PL_â€”0551 R8.C1, PL-0894 R3.C7, PL_â€”0302 R3.C4, and PL_â€”1012 R2.C1 were tested for their effect on the ability of an unrelated receptor (PAR1) to activate Ca⁺⁺ signaling. Human embryonic kidney cells were cultured in a 96-well format and allowed to adhere for 2 hours. The medium was aspirated and the plate incubated at 37Â° C. for 30 minutes in 0.5 mL loading buffer (20 mM HEPES (pH 7.4), 130 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgSO₄, 0.83 mM Na₂HPO₄, 0.17 mM NaH₂PO₄, 1 mg/ml BSA, 25 mM mannose) containing 0.1% (v/v) Pluronic F127 and 10 Î¼M Oregon Green Bapta-1 acetoxymethyl ester. The small molecules were added to the appropriate wells after 30 minutes and the cells incubated at 37Â° C. for another 30 minutes. The 96-well plate was tested for calcium concentration using a Flexstationâ„¢ system. Basal conditions were established before addition of thrombin (Â±70 nM). Recordings were made every 5 seconds and continued for >100 seconds after stimulation with thrombin. See FIG. 25.

Example 38 Modulation of MBP-8 Binding to Rhodopsin by Small Molecules

To test the effects of the small molecules in cells, light response experiments were carried out on isolated rods from the dark-adapted retina of a salamander. Single rods were isolated by shredding a small piece of retina. Photoreceptors were mechanically isolated from the dark-adapted retinas and placed in a gravity-fed superfusion chamber on the stage of an inverted microscope. Membrane currents were recorded with a suction electrode as described by Baylor et al., (Baylor et al., J. Physiol. (Lond.). 288:589-611, 1979; Baylor et al. J. Physiol. (Lond.). 288:589-611, 1979) in Ringer solution containing 120 mM NaCl, 2.0 mM KCl, 2 mM NaHCO₃, 1.6 mM MgCl₂, 1.0 mM CaCl₂, 10 mM glucose, and 3 mM HEPES, pH 7.6, as described by Rieke and Baylor, Biophys. J. 71:2553-2572, 1996. Membrane current collected by the suction electrode was amplified, low-pass filtered at 20 Hz (3 dB point; 8-pole Bessel low-pass), digitized at 100 Hz and stored on a computer for subsequent analysis. Light responses were elicited by 10-msecond flashes of 50-500 nm light. The flash strength was controlled with calibrated neutral density filters. The cell was positioned in the suction electrode to collect as much dark current as possible. Solution changes (by which addition of the small molecule was effected) were achieved with a series of electronically controlled pinch valves (Biochem Valves, Boonton, N.J.) the outlets of which were connected to a common perfusion pipe about 100 Î¼m in diameter. Solution changes with this system were completed in 200-300 mseconds. Solutions were driven by positive pressure through a pair of glass pipes with openings about 50 Î¼m in diameter. The pipes were mounted on a piezoelectric translation stage (Burleigh Instruments, Fishers, N.Y.). Solution changes at the cut end of the outer segment were completed in less than 10 mseconds with this system.

Light stimuli were delivered from a dual beam optical bench. Monochromatic lights were obtained by passing the light from a tungsten-halogen bulb through interference filters with 10 nm nominal bandwidths. Wavelength (520 nm for rods and 440 nm or 620 nm for cones) and intensity of the stimulating light were set with calibrated narrow band interference and neutral density filters, respectively. Salamander L, S, and ultraviolet-sensitive cones have peak sensitivities at 600 nm, 430 nm, and 360 nm (Makino and Dodd, J. Gen. Physiol. 108:27-34, 1996) and are readily identified by the relative amplitudes of their responses to 620 nm, 440 nm, and 380 nm lights. After identification, S cones were stimulated with 440 nm light and L cones were stimulated with 620 nm. Ultraviolet-sensitive cones were not studied. Light intensities were controlled with a set of calibrated neutral density filters, and light flashes were produced by an electronically controlled shutter in the light path.

The results are presented in FIGS. 26 and 27, which show the light response of an isolated rod from the dark-adapted retina of a salamander in the presence or absence of 5 Î¼M compound PL_â€”0302R3C4, respectively. Panel A of each figure shows the membrane current (response) plotted against time for the light responses as a result of increasing light flashes. In panel B, the peak responses have been normalized so that the current at the highest light flash is 1.0. The circles correspond to the peak response for each light flash. In FIG. 27, panel B, the bissecting lines indicate that in the presence of PL_â€”0302R3C4 a lower intensity is required to get the same change of current. The results indicate that compound PL_â€”0302R3C4 increases the peak response (as measured by a change in current) 20%-50%, depending on the intensity of the light flash and thus the amount of rhodopsin receptors activated. The results are representation of these separate experiments. The results suggest taht the compound PL_â€”302R3C4 can serve as an allosteric agonist and increase the signaling activity of the receptor in cells.

Small molecules also were tested for their ability to enhance the binding of MBP-8. PELM6 (the MBP control) and MBP-8 were plated on 96 well plates that contained EDTA-washed rhodopsin. A small molecule compound library was added, and the amount of pELM6 or MBP-8 that remained bound was measured. Standard methods were used.

FIGS. 28-32 show MBP-8 binding curves for the depicted small molecules' ability to enhance binding of the high affinity peptide fusion protein, MBP-8 to EDTA-washed rhodopsin.