Title:  Inhibitors of HIV membrane fusion

United States Patent:  6,841,657

Issued:  January 11, 2005

Inventors:  Eckert; Debra M. (Cambridge, MA); Chan; David C. (Arcadia, CA); Malashkevich; Vladimir (Belmont, MA); Carr; Peter A. (Cambridge, MA); Kim; Peter S. (Lexington, MA)

Assignee:  Whitehead Institute for Biomedical Research (Cambridge, MA)

Appl. No.:  746742

Filed:  December 21, 2000

Abstract

Inhibitors of HIV membrane fusion and a method of identifying drugs or agents which inhibit binding of the N-helix coiled-coil and the C helix of HIV gp41 envelope protein.

Description of the Invention

BACKGROUND OF THE INVENTION

Structural studies of proteins from human immunodeficiency virus type 1 (HIV-1) have been essential in the development of anti-retroviral drugs. Structure-based drug development has been most intense for reverse transcriptase inhibitors and protease inhibitors, the two classes of HIV-1 drugs in clinical use. It would also be useful to be able to carry out structure-based drug development against HIV entry.

SUMMARY OF THE INVENTION

As described herein, the cavities on the surface of the N-helix coiled-coil of HIV envelope protein gp41 subunit (e.g., HIV-1 envelope protein gp41-subunit) are targets for drugs or other agents which, by binding the coiled-coil surface, particularly the cavities, inhibit HIV entry into cells. This is useful as the basis for identifying and designing drugs or agents which inhibit entry of HIV (e.g., HIV-1, HIV-2) into cells.

Results described herein show that the coiled-coil cavity (also referred to as the hydrophobic pocket) in the gp41 core is an attractive drug target and that molecules which bind the cavity interfere with (inhibit) HIV infectivity (HIV entry into cells). Applicants have shown, for the first time, that conserved residues projecting into the hydrophobic pocket clearly play a major role in the ability of C34 to inhibit HIV-1 infection. The importance of cavity contacts (between the N-helix coiled-coil cavity and residues of the C peptide region of gp41) to gp41 function is clear. Conversely, the importance of preventing such cavity contacts in inhibiting gp41 function and, thus, inhibiting HIV-1 entry into cells, is also clear. In addition, directing drugs against the hydrophobic pocket of the central-coiled coil of gp41 targets one of the most highly conserved regions of the HIV-1 envelope proteins, which means that drugs which target the coiled-coil surface, and particularly its hydrophobic pocket, will have broad activity against diverse HIV isolates and that it will be difficult for drug-escape mutants to emerge.

A variety of methods, such as mirror-image phage display techniques (T. N. Schumacher, et al., Science, 271:1854 (1996)), combinatorial chemistry (A. Borchardt, S. D. Liberles, S. R. Biggar, G. R. Crabtree, S. L. Schreiber, Chem. Biol., 4:961 (1997); J. C. Chabala, Curr. Opin. Biotechnol., 6:632 (1995)), rational drug design and other drug screening and medicinal chemistry methods can be used to identify D-peptides, peptidomimetics and small molecules that bind the coiled-coil cavity with sufficient affinity to inhibit HIV-1 infection. The close correlation between N36/C34 stability and C34 potency, described herein, suggests that the effectiveness of such compounds will depend critically on the strength of their cavity-contacts. As described herein, candidate compounds can be tested for their ability to interfere with formation of a stable complex between C34 and N36 or their ability to disrupt binding of the two (disrupt the complex), thereby providing rapid, quantitative screens to identify and evaluate potential inhibitors of HIV-1 entry.

Alternatively, screening can be carried out to identify molecules or compounds which interfere with or disrupt binding of the N-helix coiled-coil cavity and a peptide which binds the cavity, thus providing methods of identifying molecules which are "pocket specific" binding agents or drugs. Molecules and compounds described herein (also referred to as drugs or agents) are useful to inactivate gp41 and, thus, prevent or reduce (inhibit) HIV-1 entry into cells. Without wishing to be bound by theory, it is reasonable to propose that these inhibitors bind to the pre-hairpin intermediate of gp41 and prevent its conversion to the trimeric hairpin structure of the gp41 core which corresponds to the fusion-active state of gp41. (Chan, D. C. and P. S. Kim, Cell, 93:681 (1998). Thus, the present methods are useful to identify drugs or agents which inhibit (totally or partially) formation of the fusion-active state of HIV-1 gp41 envelope protein. In the method, the ability of a candidate inhibitor (also referred to as a candidate drug), which can be any type of compound or molecule, such as a small molecule (e.g., a small organic molecule), a peptide (a D-peptide or an L-peptide), a peptidomimetic, a protein or an antibody, to bind the N-helix coiled-coil of gp41 and form a stable complex is assessed. Compounds or molecules which bind to the N-helix coiled-coil are further assessed for their ability to inhibit gp41 function (inhibit membrane fusion), such as through HIV-1 infection (viral entry) and syncytium assays, representative models of which are described and referenced herein. Those agents shown to inhibit gp41 function through such assays can be further assessed for their activity in additional in vitro assays and in appropriate animal models (e.g., Letvin, N. L., Science, 280, (5371): 1875-1880 (1998), Hirsch, V. M. and P. R. Johnson, Virus Research, 32 (2): 183-203 (1994); Reimann, K. A. et al., J. Vivol., 70 (10): 6922-6928 (1996)). Any suitable approach can be used to assess binding of candidate inhibitors to the N-helix coiled-coil and, as a result of the work described herein, to the N-helix coiled coil cavity. In one embodiment, the ability of a candidate inhibitor to bind the synthetic peptide N36 (described in Lu, M. et al., J. Biomol. Struct. Dyn. 15: 465 (1997), Chan, D. C. et al., Cell, 89, 263 (1997) and U.S. Provisional Application No. 60/043,280, entitled Core Structure of gp41 From the HIV Envelope Glycoprotein, by David C. Chan, Deborah Fass, Min Lu, James M. Berger and Peter S. Kim, filed Apr. 17, 1997) is assessed. The stability of the resulting complexes is assessed using methods described herein.

In a particular embodiment of the method of identifying compounds or molecules (drugs or agents) which bind the N-helix coiled-coil cavity, a soluble model that presents the gp41 coiled-coil cavity is used. The six helix bundle of HIV gp41 consists of an internal trimeric coiled-coil, composed of three identical N-peptides, surrounded by three C-peptides which fit into a conserved hydrophobic groove on the outside of the trimeric coiled-coil. The C-terminal end of the trimeric coiled-coil contains a large cavity into which bulky hydrophobic groups from the C-peptide pack. This hydrophobic pocket is used as the target for anti-HIV drug discovery and/or design. Unfortunately, in the absence of the C-peptide, the N-peptide is aggregated and not 100% helical. Thus, simply using an N peptide from HIV-1 gp41, such as N36, N51 (Lu, M. et al., Nature Struct. Biology, 1995) or DP-107 (Wild et al., PNAS 89:10537-10541 (1992) is unlikely to provide an effective model for the N-helix coiled-coil.

As described herein, Applicants have succeeded in producing a soluble, non-aggregating trimeric peptide model of the hydrophobic pocket of HIV gp41 and, thus, for the first time, have provided a model that properly presents this hydrophobic pocket or cavity (in a manner or configuration which forms a similar structure to the corresponding residues in the HIV gp41 structure). (The terms "pocket" and "cavity" are used interchangeably.) As described, a peptide (also referred to as a fusion protein) which includes a soluble, trimeric coiled coil portion and a portion from the N-peptide region of HIV gp41 that includes the amino acid residues which form the pocket or cavity of the N-helix coiled-coil of HIV gp41 (the pocket-comprising residues of the N-peptide) has been produced and shown to be such a soluble model, useful to identify molecules or compounds which inhibit HIV gp41 function and, thus, HIV entry into cells. The trimeric version of the coiled-coil in the peptide (also referred to as a fusion protein) can be the coiled-coil region of a protein which is not a protein of HIV (a non HIV protein, such as GCN4-pI_Q I) or a protein of HIV origin (a protein derived from HIV or having the same or a similar amino acid sequence as an HIV protein). In a specific embodiment, the soluble, non-aggregating trimeric peptide model of the large cavity, referred to as IQN17, comprises a trimeric version of the coiled-coil region of GCN4, the yeast transcription activator, and a portion of the C-terminal end of the N peptide of gp41. IQN17 contains 29 residues of GCN4-pI_Q I (formerly referred to as GCN4-pI_Q in U.S. Provisional Application No. 60/101,058) (Eckert, D. M. et al. J. Mol. Biol., 284:859-865 (1998)), including three mutations for increased solubility, and 17 residues of HIV; there is a one residue overlap between the two proteins, making the total length of the fusion protein 45 residues. The sequence of GCN4-pI_Q I is: ac-RMKQIEDKIEEI LSKQYHIENEIAR IKKLIGER (SEQ ID NO:1). The HIV Sequence is: LLQLTVWG IKQLQARIL (SEQ ID NO:20). The sequence of IQN17 is: ac-RMKQIEDKIEEIESKQKKIENEIARIKK LLQLTVWGIKQLQARIL-am (SEQ ID No:2). The HIV portion is underlined in SEQ ID No: 2; ac- represents an N-terminal acetyl group and -am represents a C-terminal amide. The sequence of the soluble, trimeric version of the coiled-coil region of GCN4 (referred to as a soluble, trimeric coiled coil of GCN4) in IQN17 is: RMKQIEDKIEEIESKQKKIENEIARIKK (SEQ ID No: 25). The superhelix parameters such as rise and pitch (Harbury, P. B. et al., Nature 371:80-83 (1994); Harbury et al., PNAS 92:8408-8412 (1995)) of the GCN4-pI_Q I coiled coil are nearly identical to the HIV gp41 N-helix coiled coil. Therefore, the resulting fusion protein molecule (IQN17) is predicted to form a long trimeric coiled coil, which presents the N-peptide hydrophobic cavity at the C terminus. IQN17 is fully helical, as determined by circular dichroism, with a molar ellipicity at 222 nm of -36,000 deg cm² dmol^-1. As determined by sedimentation equilibrium, IQN17 is close to a discrete trimeric species with a ratio of observed molecular weight to calculated molecular weight ranging from 3.00 to 3.16 times the monomer molecular weight at a concentration of 20 .mu.M. As determined by X-ray crystallography, IQN17 presents the N-peptide hydrophobic pocket in a manner that is nearly identical to the pocket in the HIV gp41 N-helix coiled coil.

The IQN17 molecule (in the natural L-handedness or enantiomeric D-handedness) can be used in screens, including high-throughput drug screens, to identify molecules that bind to the coiled-coil pocket. The IQN17 molecule, in the D-handedness, has been used as a target in mirror image phage display (Schumacher et al., Science, 271: 1854, 1996) to identify small molecules (D-peptides) which bind to the hydrophobic pocket of gp41 (in the natural L-handedness) and inhibit HIV-membrane fusion. The desired target (the N-helix of HIV gp41 which includes the hydrophobic pocket) is the exact mirror image of the naturally-occurring target. It is used to screen a library or collection of compounds or molecules which are to be assessed for their ability to bind the mirror image of the naturally-ocurring coiled-coil pocket. The mirror image of a compound or molecule found to bind the mirror image of the naturally-occurring gp41 pocket, will bind the gp41 pocket in the natural handedness. The library or collection screened can be of any type, such as a phage display library, peptide library, DNA library, RNA library, combinatorial library, collection of chemical agents or drugs, cell lysate, cell culture medium or supernatant containing products produced by cells. In the case of a phage display library, the D-target is used to screen phage coat proteins. Specific phage clones that bind to the target are identified and the mirror images of the expressed proteins are chemically synthesized with D-amino acids. By using IQN17 in mirror-image phage display, D-peptides that bind to the gp41 hydrophobic pocket have been identified. Further assessment has been carried out, as described, to demonstrate the ability of D-peptides to inhibit HIV gp41 function. D-peptides which bind the gp41 hydrophobic pocket and inhibit HIV infectivity have been identified. D-peptides which bind the hydrophobic pocket also will serve as lead molecules for drug development and/or reagents for drug discovery (where the drugs bind to the coiled-coil pocket and inhibit HIV infectivity). The IQN17 molecule, in the natural L-handedness, can be used in screens, including high-throughput screens, to identify molecules that bind to the coiled-coil pocket. IQN17 can be used to screen a collection or library of compounds or molecules which are to be assessed for their ability to bind the hydrophobic pocket. The library or collection screened can be of any type, such as a phage display library, RNA library, DNA library, peptide library, combinatorial library, collection of chemical agents or drugs, cell lysate, cell culture medium or supernatant containing products produced by cells. Compounds or molecules which bind the hydrophobic pocket also will serve as lead molecules for drug development and/or reagents for drug discovery.

Fusion proteins which are variants of IQN17 can be produced and used to screen for drugs which bind the gp41 N-helix coiled-coil pocket. Any of a wide variety of variations can be made in the GCN4-pI_Q I component of IQN17 and used in the method, provided that these changes do not alter the trimeric state of the coiled-coil. For example, the amino acid composition of the GCN4 component can be changed by the addition, substitution, modification and/or deletion of one or more amino acid residues, provided that the trimeric state of the coiled-coil is maintained. For example, the Asp residue in IQN17 (at a "f-position" of the coiled coil) can be replaced by any of the naturally-occurring amino acids. (O'Neil and DeGrado, Science 250:646 (1990)). Alternatively, this component of the fusion protein can be a trimeric version of the coiled-coil region of another protein, such as that from Moloney Murine Leukemia Virus (Fass, D. et al. Nature Struct. Biology, 3:465 (1996)), GCN4-pII (Harbury et al., Nature, 317:80, 1994) or the ABC heterotrimer (Nautiyal and Alber, Protein Science 8:84 (1999)).

Changes can also be made in the amino acid composition of the fusion protein component which is the C-terminal portion of the HIV gp41 N peptide to produce IQN17 variants. The C-terminal portion can be changed by the addition, substitution, modification and/or deletion of one or more amino acid residues. The amino acid composition of either or both components of the fusion protein can be altered, and there is no limit to the number or types of amino acid residue changes possible, provided that the trimeric state of the coiled-coil and the hydrophobic pocket of the N peptide of HIV gp41 are maintained. IQN17, IQN17 variants or any soluble model of the large cavity can be used to screen for drugs which bind the N-helix coiled-coil, especially the pocket, or for lead drug candidates or candidates for use in vaccine preparations, to be further screened using methods known to those of skill in the art, such as in a high throughput format.

Results described herein are useful to screen for inhibitors of HIV gp41 which are variants of C34 as described below. Once a variant of C34, such as a C34 variant which stably binds N36, has been identified, it can be used and further assessed as obtained or it can be modified (e.g., by altering, adding, deleting or substituting at least one amino acid residue or adding a non-amino acid substituent), if desired or needed (e.g., to enhance stability, solubility, bioavailability). Alternatively, a C34 variant can be assessed to determine if a shorter component (region of fewer amino acid residues) also is active as an inhibitor. As discussed herein, the three C34 residues Trp⁶²⁸, Trp⁶¹ and Ile⁶³⁵ that pack into the deep, conserved pocket in the N36 trimer are critical for inhibitory activity. The observation that C34 variants that have a higher affinity for the N36 coiled-coil have more potent inhibitory activity against HIV infection forms the basis for screens to identify and evaluate potential inhibitors. For example, using the "split-synthesis" technique (Chen, C. L., et al., Methods Enzymol. 267:211-219 (1996); Lam, K. S. et al., Nature, 354: 82-84, (1991)) of combinatorial peptide chemistry, a library of C34 variants is synthesized in which the three critical hydrophobic residues are randomly replaced by chemical substitutions of varying hydrophobic character. This synthesis technique results in the generation of a vast library of beads, each containing many copies of a single variant C34 peptide (i.e., a "one-bead, one-compound" type of library). To identify C34 variants which stably bind the N-helix coiled-coil, a labeled version of N36 (or a modified N-peptide) is mixed with the peptide beads under conditions (e.g., elevated temperature) that restrict binding to only those C34 variants with the highest affinity. Binding is measured by detection of the label on the N-helix peptide, using known methods. Simple modifications of the split-synthesis technique allow ready identification of the selected peptide sequence by mass spectroscopy (Youngquist, R. S. et al., J. Amer. Chem. Soc. 117, 3900-3906 (1995)). The C34 variants selected, particularly those with the highest binding affinities for N36, are tested in syncytium and infection assays for gp41 inhibitory activity. Truncated versions of these C34 variants, containing only the cavity-binding region, can also be tested for inhibitory activity. Alternatively, a library of other peptides to be assessed can be synthesized to generate a library of beads, each containing (having bound thereto) a peptide to be assessed. This library is analyzed as described above for the C34 variants and resulting hits (members with appropriate binding affinities for N36) are further analyzed for gp41 inhibitory activity. As a second example, the N36 peptide or the soluble variants described earlier, such as IQN17, GCN4-N-helix peptide can be used as a target for phage display or mirror-image phage display techniques to identify peptides that bind to the cavity.

IQN17 can also be used to raise antibodies (monoclonal and/or polyclonal) that bind to the coiled-coil cavity. IQN17 can further be used, either alone or in combination with other materials, in a vaccine, which will elicit the production of antibodies that bind to the coiled-coil in the individual to whom it is administered (the vaccinee), and thereby offer protection against infection and/or disease.

Peptides, both D-peptides and L-peptides, which fit into a deep hydrophobic pocket in the trimeric N-helix coiled-coil of HIV-1 envelope glycoprotein gp41 are also the subject of this invention. The D-peptides are the first molecules that have been shown to bind exclusively to the gp41 hydrophobic pocket. The observation that these D-peptides inhibit gp41-mediated membrane fusion processes (syncytia formation and viral infection) provides the first direct demonstration that HIV-1 infection can be inhibited by molecules that bind specifically to pocket. The validation of the gp41 hydrophobic pocket as a drug target sets the state for the development of a new class of orally bioavailable anti-HIV drugs, that work by inhibiting viral entry into cells. Such drugs would be a useful addition to the current regimen used to treat HIV-1 infection with combination therapies. D-peptides, such as the D-peptides described herein, portions, modification and variants thereof and larger molecules (e.g., polypeptides which comprise all or a portion of a D-peptide described herein, are useful to inhibit HIV membrane fusion and, thus, HIV entry into cells. D-peptides, corresponding to the D-amino acid version of phage sequences identified as described herein, are inhibitors of HIV-1 infection and syncytia formation. The C-terminal residues in these D-peptide inhibitors have the sequence pattern: CXXXXXEWXWLCAA-am (SEQ ID NO: 69). (In the phage-display library, the positions corresponding to the C residues were encoded as either C or S, the positions corresponding to the AA residues were encoded as such and the other 10 positions (indicated by X) were randomly encoded. The -am represents a C-terminal amide, added as part of the peptide synthesis procedure.) The N-terminal residues in the D-peptide inhibitors are, for example, ac-GA, ac-KKGA (SEQ ID NO: 70), or ac-KKKKGA (SEQ ID NO: 71). The ac- represents an N-terminal (acetyl group added as part of the peptide synthesis procedure. The C-terminal amide and the N-terminal acetyl group are optional components of D-peptides of this invention. Other N-terminal residues can be included, in place of or in addition to those in the previous sentence, desired (e.g., to increase solubility). For example, D-peptides of the following sequences are also the subject of this invention:

ac-XXCXXXXXEWXWLCXX-am (SEQ ID NO: 28);

ac-KKXXCXXXXXEWXWLCXX-am (SEQ ID NO: 29);

ac-KKKKKXXCXXXXXEWXWLCXX-am (SEQ ID NO: 30);

ac-XXCXXXXXEWXWLCXXX-am (SEQ ID NO: 31);

ac-KKXXCXXXXXEWXWLCXXX-am (SEQ ID NO: 32); and

ac-KKKKXXCXXXXXEWXWLCXXX-am (SEQ ID NO: 33).

The amino acid residues are represented by the single letter convention and X represents any amino acid residue (naturally occurring or non-naturally occurring) or other moiety, such as a modified amino acid residue.

Further, the ten amino acid residue "core" (the 10-mer which is flanked a each end by a cysteine residue) of the 12 amino acid residue peptide, as well as portions, modifications and variants of the 10-mers are also useful to inhibit membrane fusion and entry of IV into cells. Variants, portions and modifications of these peptides are also useful as inhibito a. As described further herein, D-peptides which comprise a consensus sequence (e.g., WXWL SEQ ID NO: 23), EWXWL (SEQ ID NO: 24), CXXXXXEWXWLC (SEQ ID NO: 63) or a portion thereof) have been shown to bind the N-helix coiled-coil and are useful to inhibit membrane fusion and entry of HIV into cells. The enantiomeric peptides (D-peptides) do not serve as efficient substrates for enzymes, such as proteases and, therefore, are more resistant to poteolytic degradation than are L-peptides; they are also less immunogenic than are L-peptides.

Specific embodiments of D-peptides of the present invention are:

(a) CDLKAKEWFWLC (SEQ ID NO: 3);

(b) CEARHREWAWLC (SEQ ID NO: 4);

(c) CELLGWEWAWLC (SEQ ID NO: 5);

(d) CLLRAPEWGWLC (SEQ ID NO: 6);

(e) CSRSQPEWEWLC (SEQ ID NO: 7);

(f) CGLGQEEWFWLC (SEQ JD NO: 8);

(g) CMRGEWEWSWLC (SEQ ID NO: 9);

(h) CPPLNKEWAWLC (SEQ ID NO: 10);

(i) CVLKAKEWFWLC (SEQ ID NO: 11);

(j) KKGACGLGQEEWFWLC (SEQ ID NO: 15);

(k) KKGACELLGWEWAWLC (SEQ ID NO: 16);

(l) KKKKGACELLGWEWAWLC (SEQ ID NO: 17);

(m) KKGACMRGEWEWSWLC (SEQ ID NO: 18);

(n) KKGACPPLNKEWAWLC (SEQ ID NO: 19);

(o) a D-peptide comprising WXWL (SEQ ID NO: 23);

(p) a D-peptide comprising EWXWL (SEQ ID NO: 24);

(q) a D-peptide comprising CXXXXXEWXWL (SEQ ID NO: 12)

(r) ac-GACEARHREWAWLCAA-am (SEQ ID NO: 34);

(r) ac-KXGACEARHREWAWLCAA-am (SEQ ID NO: 38);

(t) ac-KKKKGACEARHREWAWLCAA-am (SEQ ID NO: 43);

(u) ac-GACGLGQEEWFWLCAA-am (SEQ ID NO: 44);

(v) ac-KKGACGLGQEEWFWLCAA-am (SEQ ID NO: 64);

(w) ac-KKKKGACGLGQEEWFWLCAA-am (SEQ ID NO: 45)

(x) ac-GACDLKAKBWFWLCAA-am (SEQ ID NO: 35);

(y) ac-KKGACDLKAKBWFWLCAA-am (SEQ ID NO: 39);

(z) ac-KKKKGACDLKAKEWFWLCAA-am (SEQ ID NO: 46);

(a') ac-GACELLGWEWAWLCC-am (SEQ ID NO: 47);

(b') ac-KKGACELLGWEWAWLCAA-am (SEQ ID NO: 65);

(c') ac-KKKKGACELLGWEWAWLCAA-am (SEQ ID NO: 66);

(d') ac-GACSRSQPEWEWLCAA-am (SEQ ID NO: 36);

(e') ac-KKGACSRSQPEWEWLCAA-am (SEQ ID NO: 40);

(f') ac-KKKKGACSRSQPEWEWLCAA-am (SEQ ID NO: 48);

(g') ac-GACLLRAPEWGWLCAA-am (SEQ ID NO: 37);

(h') ac-KKGACLLRAPEWGWLCAA-am (SEQ ID NO: 41);

(i') ac-KKKKGACLLRAPEWGWLCAA-am (SEQ ID NO: 49);

(j') ac-GACMRGEWEWSWLCAA-am (SEQ ID NO: 50);

(k') ac-KKGACMRGEWEWSWLCAA-am (SEQ ID NO: 67);

(l') ac-KKKKGACMRGEWEWSWLCAA-am (SEQ ID NO: 51);

(m') ac-GACPPLNKEWAWLCAA-am (SEQ ID) NO: 52);

(n') ac-KKGACPPLNKEWAWLCAA-am (SEQ ID NO: 68);

(o') ac-KKKKGACPPLNKEWAWLCAA-am (SEQ ID NO: 53);

(p') ac-GACXXXXXEWXWLCAA-am (SEQ ID NO: 54);

(q') ac-KKGACXXXXXEWXWLCAA-am (SEQ ID NO: 55);

(r') ac-KKKKGACXXXXXEWXWLCAA-am (SEQ ID NO: 56);

(s') ac-XXCXXXXXEWXWLCXX-am (SEQ ID NO: 57);

(t') ac-KKXXCXXXXXEWXWLCXX-am (SEQ ID NO: 58);

(u') ac-KKKKXXCXXXXXEWXWLCXX-am (SEQ ID NO: 59);

(v') ac-XXCXXXXXEWXWLCXXX-am (SEQ ID NO: 60);

(w') ac-KKXXCXXXXXEWXWLCXXX-am (SEQ ID NO: 61);

(x') ac-KKKKXXCXXXXXEWXWLCXXX-am (SEQ ID NO: 62); and

(y') a variant of a sequence of(a) through (x'), wherein the variant bands the N-helix coiled-coil cavity of HIV gp41, wherein ac- at the C-terminus -am at the N-terminus are optional.

D-peptides described herein, which are ligands shown to bind the N-helix pocket, are also useful in drug screens to identify compounds or molecules (e.g., from chemical libraries, recombinantly produced products, naturally-occurring substances, culture media or supernatants) which bind the N-helix pocket and thus, are also inhibitors of HIV. For example, a competitive assay can be carried out by combining a D-peptide which binds the N-helix cavity (e.g., a D-peptide described herein); IQN17 (e.g., in the natural L-handedness), or another fusion protein which is a soluble model that presents the N-helix cavity; and a candidate inhibitor (a compound or molecule to be assessed for its ability to bind the N-helix cavity). For example, D10pep5 or D10pep1, IQN17, and a candidate inhibitor (candidate drug) can be combined using buffer conditions and peptide concentrations appropriate for binding of D10pep5 or D10pep1 to IQN17. The extent to which binding of the D-peptide occurs is determined and compared to the extent to which binding occurs under the same conditions, but in the absence of a compound or molecule (referred to as a candidate drug or candidate inhibitor) to be assessed for its ability to bind the N-helix coiled-coil cavity of HIV gp41 envelope protein (in a control). If binding of D10pep5 or D10pep1 occurs to a lesser extent in the presence of the candidate inhibitor (test sample) than in its absence (control sample), the candidate inhibitor is a ligand which binds the N-helix coiled-coil cavity and, thus, is an inhibitor. Inhibitors identified in this manner can be further assessed for their activity in viral infectivity assays and synctia formation assays, such as those described herein. Those inhibitors which show activity in such assays can be further assessed in an appropriate animal model or in humans.

Any method by which binding of the D-peptide, known to bind the N-helix cavity, can be detected can be used to assess whether the candidate inhibitor interferes with binding. For example, the D-peptide can be detectably labeled and the extent to which the label appears on the N-helix cavity (as a result of binding of the D-peptide) detected, in the presence and in the absence of the candidate inhibitor. If less label appears on the N-helix cavity of IQN17 (or other appropriate fusion protein) in the presence of the candidate inhibitor (in the test sample) than in its absence (in the control sample), then the candidate inhibitor is a ligand which binds the N-helix cavity (and interferes with binding of the D-peptide). Alternatively, the D-peptide (e.g., D10pep5 or D10pep1) and IQN17 can be labeled with a fluorophore (e.g., with EDANS; 5-(2'aminoethyl)aminonaphthalene-1-sulfonic acid) with an appropriate quencher that quenches the fluorescent signal of the fluorophore when it is in close proximity (e.g., DABCYL; 4-(4'-dimethylaminophenylazo)benzoic acid). If the candidate inhibitor binds the N-helix cavity of IQN17, fluorescence is observed, since, as a result of binding of the candidate inhibitor, the D-peptide is not brought into sufficiently close proximity to the quencher to permit it to quench the reporter signal. Alternatively, the fluorescent reporter molecule can be on the IQN17 and an appropriate quencher on the D-peptide. In either case, the position of the reporter or quencher on IQN17 must be such that when the D-peptide binds the N-helix cavity, the reporter and quencher moieties are in sufficiently close proximity to each other that quenching occurs (Tyagi, S., et al., Nature Biotechnology 16:49 (1998)).

Also the subject of this invention are drugs (compounds and molecules) which bind the N-helix coiled-coil pocket of HIV gp41 and inhibit (partially or totally) HIV entry into cells. In one embodiment, these drugs can be identified as described herein or by other methods. Drugs which bind the N-helix coiled-coil pocket of HIV gp41 are useful as therapeutic agents (to prevent HIV entry into cells or reduce the extent to which it occurs), as research tools (e.g., to study the mechanism of HIV gp41 function) and to assess the rate of viral clearance by an individual (e.g., in an animal model or an infected human).

Also the subject of this invention are compositions, useful in methods of interfering with entry of HIV into a mucosal cell; these compositions comprise an appropriate carrier or base and at least one component selected from the group consisting of:

(a) C34 peptide;

(b) DP178;

(c) T649;

(d) T1249;

(e) a derivative of (a)-(d);

(f) a D-peptide which binds to the hydrophobic pocket of HIV gp41;

(g) a derivative of (f);

(h) a combination of two or more of (a)-(g); and

(i) a molecule that inhibits HIV infectivity by binding to the N-helix coiled coil.

The compositions can comprise one such component or two or more components.

A further subject of this invention are compositions (e.g., proteins or proteinaceous materials) that can be used to elicit an immune response (e.g., antibody production) that will protect (partially or totally) against HIV infection and/or disease. Such compositions are useful as protective agents (e.g., vaccines) and to obtain antibodies (monoclonal and/or polyclonal) that are useful as research tools, diagnostic tools, drug screening reagents, and to assess viral dynamics (rates of production and clearance of virus) in animal models or infected humans.

Also the subject of this invention is a list of atomic coordinates for the X-ray crystal structure of the complex between IQN17 and D10pep1. Also the subject of this invention is a list of coordinates for the X-ray crystal structure of IQN17. These coordinates can be used (e.g., as an electronic file for computer graphics programs) to create a model of the complex which indicates how D10pep1 binds to the N-helix coiled-coil cavity and models of the N-helix coiled-coil cavity. Such models can be used, in methods known to those of skill in the art such as in computer graphics modeling, to build new models to evaluate the likelihood of binding to the N-helix coiled-coil cavity by other peptides, peptidomimetics, small molecules, drugs or other compounds. Such models can also be used to build new models for the structures of molecules (peptides, peptidomimetics, small organic molecules, drugs or other compounds) that bind the N-helix coiled-coil cavity (e.g., H. Kubinyi, Curr. Op. Drug Discov. Develop., 1:16 (1998); P. L. Wood, ibid, 1:34 (1998); J. R. Morphy, ibid, 1:59 (1998)). These models and the corresponding lists of atomic coordinates can be used to identify, evaluate, discover and design more effective and/or new D-peptides, L-peptides, peptidomimetics, other small molecules or drugs that inhibit HIV infectivity, using methods known to those of skill in the art. A further subject of this invention is a method of producing or identifying a drug which fits (packs into, binds) the N-helix coiled-coil pocket of HIV gp41 through the use of atomic coordinates of a crystal, such as a crystal of a soluble, trimeric peptide model of the HIV gp41 hydrophobic pocket described herein (e.g., IQN17 or a variant thereof), a crystal of such a model in complex with a D-peptide (e.g., IQN17 or a variant thereof in complex with a D-peptide described herein, such as D10pep1) or a crystal of the N-peptide region of HIV gp41 comprising the amino acid residues which comprise the pocket of the N-helix coiled-coil of HIV gp41. The method comprises obtaining a crystal of the soluble model, such as the empty soluble model (not in complex with a D-peptide), obtaining the atomic coordinates of the crystal (e.g., of the crystal of the empty soluble model, such as IQN17); using the atomic coordinates obtained to define the N-helix coiled-coil pocket of HIV gp41; identifying a molecule or compound which fits the N-helix coiled-coil pocket and obtaining the molecule or compound; contacting the molecule or compound with the N-helix coiled-coil pocket (e.g., by contacting it with a polypeptide which comprises the pocket (e.g., IQN17 or a variant thereof or the N-peptide) to assess (determine) the ability of the molecule or compound to fit the pocket of HIV gp41, wherein in the molecule or compound fits the pocket, it is a drug which fits the N-helix coiled-coil pocket, whereby a drug which fits the pocket is produced. The atomic coordinates of the crystal can be obtained by X-ray diffraction studies or form a computer file or Protein Data Base (PDB), such as the PDB presented herein for IQN17.

Similarly, the method can be carried out using a crystal of a soluble trimeric model in complex with a D-peptide (e.g., a D-peptide described herein, such as D10pep1) or a crystal of the N-peptide region of HIV gp41 which comprises the pocket of the N-helix coiled coil.

Drugs produces in this manner can be further assessed to conform their ability to fit into the pocket (e.g., by NMR) and can be assessed for their ability to inhibit HIV entry into cells (e.g., by a syncytia assay or infectivity assay).

Claim 1 of 10 Claims

What is claimed is:

1. A D-peptide which is a soluble, trimeric peptide model of the HIV gp41 hydrophobic pocket, wherein the D-peptide comprises SEQ ID NO: 25 and a sequence which comprises 17 amino acid residues, wherein the 17 amino acid residues comprise the sequence: LLXLTVWGXKXLQXRXX (SEQ ID NO: 42), wherein L, T, V, W, O, K, Q and R are amino acid residues represented by the single letter amino acid code and X is any D-amino acid residue.

____________________________________________
If you want to learn more about this patent, please go directly to the U.S. Patent and Trademark Office Web site to access the full patent.