WO2011050222A2 - Therapeutic compositions that disrupt the hydrogen bonded ring structure in gp41 and methods for treating hiv - Google Patents

Abstract

The Invention relates to therapeutic compositions and methods for treating HIV and other viral diseases and to vaccines for preventing HIV and other viral diseases.

Description

IN THE PATENT COOPERATION TREATY

Title: Therapeutic compositions that disrupt the hydrogen bonded ring structure in gp41 and methods for treating HIV

Statement of support

[001] This invention was made with support of the Bill and Melinda Gates Foundation and the University of California, Santa Cruz start-up fund.

Relationship to other applications

[002] To the extent legally relevant and allowed by national law, this application claims priority to and the benefit of US Provisional application No. 61 ,253,858 filed 22 October 2009 and titled "Therapeutic compositions that disrupt the hydrogen bonded ring structure in gp41 and methods for treating ΗΓν" . This application is also related to, by virtue of one or more common inventors, US provisional application 61/195, 1 12 filed 4^th October 2008 and International application No. PCT/US09/59583 filed 5^th October 2009.

Sequence listing

[003] The information recorded in electronic form (if any) submitted (under Rule 13ier if appropriate) with this application is identical to the sequence listing as contained in the application as filed.

Field of the invention

[004] The Invention relates to therapeutic compositions and methods for treating HIV and other viral diseases and to vaccines for preventing HIV and other viral diseases. Specifically the invention relates to thera- peutics applications against HIV infections and to methods for creation, screening and identification of viral epitopes of HIV that have therapeutic value.

Background

[005] A major goal in HIV vaccine research is the identification of antigens able to elicit the production of broadly neutralizing antibodies (bNAbs) effective against primary isolates of HIV. Despite more than 20 years of effort, antigens able to elicit broadly neutralizing activity have yet to be described (Burton et al., 2004; Fauci et al., 2008). The applicant has investigated the molecular features of the HIV- 1 envelope glycoproteins, gpl 20 and gp41 , that confer sensitivity and resistance of viruses to neutralization by broadly neutralizing antibodies.

Brief description of the invention

[006] Disclosed are to therapeutic compositions and methods for treating HIV and other viral diseases and to vaccines for preventing HIV and other viral diseases. These therapeutic compositions contain species and compositions that have been identified by a novel method for identifying mutations in envelope proteins, which mutations provide enhanced sensitivity to neutralization of an virus by anti-viral antisera; in particular neutralization of an HIV virus by anti-HIV antisera. This novel method is disclosed in US provisional application

61/195, 1 12 filed 4* October 2008 and also in related International application No. PCT/US09/59583 filed 5* October 2009.

[007] Identification of the determinants of sensitivity and resistance to broadly neutralizing antibodies is a high priority for HIV research. Analysis of the swarm of closely related envelope protein variants in HIV infected individuals revealed a mutation that markedly affected sensitivity to neutralization by antibodies and antiviral entry inhibitors targeting both gp41 and gpl 20. This mutation mapped to the C34 helix of gp41 and disrupted an overlooked structural feature consisting of a ring of hydrogen bonds in the gp41 trimer. This mutation affects the assembly of the 6 helix bundle required for virus fusion, and alters the conformational equilibrium so as to favor the pre-hairpin intermediate conformation required for the binding of the HIV-1 env membrane proximal external region (MPER) specific neutralizing antibodies, 2F5 and 4E10, and the antiviral drug, Fuzeon. Targeting cooperative interactions that stabilize conformational transitions provides new approach to the design of vaccine antigens and antiviral compounds. Methods for measuring the integrity of the pre-hairpin intermediate conformation include those described by Yang Xu et al., "Development of a FRET Assay for Monitoring of HIV gp41 Core Disruption" J. Org. Chem. 2007, 72, 6700-6707.

[008] The invention encompasses the use of various compounds used for therapeutic purposes. These compounds may be contacted with a virus, such as an HIV virus, and interact with and/or bind to one or more regions on the viral envelope (env) protein or other viral protein or glycoprotein. This interaction thereby (i) exposes one or more previously unexposed epitopes which epitope can bind specifically with a neutralizing antibody and/or (ii) limits, inhibits or prevents fusion of a viral membrane with a cell membrane, thereby inhibiting infection of a cell by a virus. Such compounds may be therapeutic compositions, drugs, small molecules or antibodies.

[009] Also disclosed are therapeutic methods that employ compositions such as drugs and small molecules or antibodies that interact with specific antigens or epitopes or regions of the glycoproteins or polypeptides described, thereby (i) exposing a previously unexposed epitope which epitope can bind specifically with a neutralizing antibody and/or (ii) limiting, inhibiting or preventing fusion of a viral membrane with a cell membrane, thereby inhibiting infection of a call by a virus. Also disclosed are the therapeutic compositions, drugs, small molecules or antibodies used in the above method.

[0010] Also disclosed are generic and specific sequences, mutations, antigens and epitopes that may be used therapeutically for the treatment and/or prevention of viral infection such as HIV infection, and vectors, pseudovi- ruses and other constructs that comprise specific polynucleotide sequences and mutations that encode antigens and epitopes of the invention.

[001 1] Also disclosed are therapeutic methods that comprise delivering a vaccine to a subject wherein the vaccine may comprise one or more antibodies or antigens or epitopes of the invention, or polynucleotide sequences or vectors encoding antigens and epitopes of the invention.

[0012] Also disclosed are therapeutic methods and therapeutic compositions comprising drugs such as small molecules that target a specific antigens or epitopes of the invention, thereby limiting, inhibition or preventing fusion of a viral membrane with a cell membrane, thereby inhibiting infection of a call by a virus.

[0013] One particular embodiment is a method for inhibiting the fusion of an HIV virus to a host cell, the method comprising exposing the HIV virus to a drug compound that disrupts the hydrogen-bonded ring structure between the N36 and C34 helices of gp41.

[0014] Another embodiment is a method for increasing the immunogenicity of HIV envelope proteins the method comprising exposing the HIV virus to a drug compound that disrupts the hydrogen bonded ring structure between the N36 and C34 helices of gp41. Methods for measuring the integrity of the hydrogen-bonded ring structure are known in the art and include those described by Yang Xu et al., "Development of a FRET Assay for Monitoring of HIV gp41 Core Disruption" J. Org. Chem. 2007, 72, 6700-6707. The drugs used in these methods may be, for example, antibodies, small molecules or peptidomimetics, including, for example, Fuzeon, 4E10, 2F5.Q665R, Q655K and Q655E. In some of the methods the drug compound interacts with the gpl20 fragment of the HIV envelope protein. In some of the methods, the drug compound is an inhibitor of HIV fusion binding and becomes a more effective inhibitor in the presence of a molecule that disrupts the disulfide bonded ring structure of gp41.

[0015] An important element of the invention is the mechanism by which the drug works to prevent viral fusion and/or to expose previously hidden epitopes. In various methods the drug compound disrupts the hydrogen bonded ring structure between the N36 and C34 helices of gp41 and thereby exposes neutralizing epitopes which are recognized by endogenous or exogenous antibodies which then are able to neutralize the virus.

[0016] The invention encompasses methods for screening for a drug that prevents or attenuates intracellular membrane fusion, the method comprising exposing the multimeric coiled coil bundle of the activated fusion complex to a drug candidate, wherein disruption of one or more hydrogen bonds of the fusion complex is associated with prevention or attenuation of intracellular membrane fusion. The fusion complex may comprise a cellular hairpin membrane fusion protein. The cellular hairpin membrane fusion protein may be a cellular SNARE protein. The multimeric coiled coil bundle may be a 4 helix bundle. The intracellular membrane fusion may be associated with secretion of a hormone, cytokine or neurotransmitter.

[0017] The invention also includes a method of treating, attenuating or preventing HIV infection, the method comprising administering to a patient a drug compound which disrupts one or more intra-molecular or inter- molecular hydrogen bonds of the hydrogen-bonded ring structure of gp41 trimer, wherein disruption of the hydrogen-bonded ring structure is associated with attenuation or prevention of HIV infection.

[0018] The invention also encompasses a synthetic helical peptide wherein the peptide sequence binds specifically to at least a fragment of the N-36 helix of gp41 , wherein the fragment includes the residue Q655 and wherein binding of the synthetic helical peptide to the N-36 helix disrupts hydrogen bonded ring structure between the N36 and C34 helices of gp41.

[0019] The invention also encompasses a peptidomimetic drug that binds specifically to helical sequences adjacent to the Q655 or Q553 residues of gp41 , and disrupts of prevents the formation of a hydrogen bonded ring structure involving Q655 from the C34 helix and Q533 from the N34 helix. The peptide or peptidomimetic binds to or interacts with the N36 or C34 helices of gp41 and thereby disrupts one or more intra-molecular or inter- molecular hydrogen bonds of the hydrogen-bonded ring structure of gp41 trimer, wherein disruption of the hydrogen-bonded ring structure is associated with attenuation or prevention of HIV infection. The peptide or peptidomimetic may disrupt one or more of the intra-molecular or inter-molecular hydrogen bonds stabilizing the multimeric coiled coil bundle in the activated fusion complex required for the fusion and release of synaptic vesicles. It may also disrupt one or more of the intra-molecular or inter-molecular hydrogen bonds stabilizing the multimeric coiled coil bundle structurally homologous to the N36 or C34 helix of HIV in the activated fusion complex required for the fusion and release of vesicles or granules containing pro-inflammatory proteins, cytokines, hormones, or vasoactive substances.

[0020] The invention also encompasses a method of attenuating or preventing HIV-l infection comprising administering to a patient an effective amount of an agent which disrupts one or more intra-molecular or inter- molecular hydrogen bonds of the hydrogen-bonded ring structure of gp41 trimer, wherein disruption of the hydrogen-bonded ring structure makes HIV-l susceptible to neutralization by the patient's antibodies which thereby attenuate or prevent HIV infection. The agent may comprise a peptide or peptidomimetic compound.

[0021] The invention also encompasses a peptide having a formula selected from the group consisting of:

X-YTSLIHSLIEESQNQ[*]EKNEQELLELDKWASLWNWF-Z

X-YTNTIYTLLEESQNQ[*]EKNEQELLELDKWASLWNWF-Z

X-YTGIIYNLLEESQNQ[*]EKNEQELLELDKWANLWNWF-Z

X-YTSLIYSLLEKSQIQ[*]EKNEQELLELDKWASLWNWF-Z

X-LEANISKSLEQAQIQ[*]EKNMYELQKLNSWDEFGNWF-Z and

X-LEANISQSLEQAQIQ[*]EKNMYELQKLNSWDVFTNWL-Z

[0022] For the above listed peptides, amino acid residues are presented by the single-letter code; X comprises an amino group, an acetyl group, a 9-fluorenylmethoxy-carbonyl group, a hydrophobic group, or a macro- molecule carrier group; Z comprises a carboxyl group, or an amido group, or a hydrophobic group, or a macromo- lecular carrier group; and [*] represents any amino acid other than Q or N. In some embodiments [*] represents R, K, S, or E.

[0023] Another discovery of potential relevance to the understanding of the determinants of neutralization sensitivity and resistance of HIV-l is disclosed in J. Virol, doi: 10.1 128/JVI.00790- 10, 'Mutation at a single position in the V2 domain of the HIV-l envelope protein confers neutralization sensitivity to a highly neutralization resistant virus' by Sara M. O'Rourke, Becky Schweighardt, Pham Phung, Dora P.A.J. Fonseca, Karianne Terry, Terri Wrin, Faruk Sinangil, and Phillip W. Berman, hereby incorporated by reference. In this work the authors made use of the swarm of closely related envelope protein variants (quasispecies) from an extremely neutralization resistant clinical isolate in order to identify mutations that conferred neutralization sensitivity to antibodies in serum from HIV- l infected individuals. The authors describe a virus with a rare mutation at position 179 in the V2 domain of gpl 20, where replacement of aspartic acid (D) by asparagine (N) converts a virus that is highly resistant to neutralization by multiple polyclonal and monoclonal antibodies, as well as antiviral entry inhibitors, to one that is sensitive to neutralization. Although the V2 domain sequence is highly variable, D at position 179 is highly conserved in HIV-l and SIV and is located within the LDI/V recognition motif of the recently described alpha-4B7 receptor binding site. Our results suggest that the D179N mutation induces a conformational change that exposes epitopes in both the gp l 20 and gp41 portions of the envelope protein such as the CD4 binding site and the MPER that are normally concealed by conformational masking. These results suggest that D179 plays a central role in maintaining the conformation and infectivity of HIV- 1 as well as mediating binding to alpha-4-beta-7 (ά4β7).

Brief description of the figures

[0024] Figure 1. Mutation of neutralization resistant clone 022 from subject 108060 (A) Amino acids from neutralization-resistant clone 022 are shown as open rectangles. Amino acids from neutralization-sensitive clone 024 were inserted by in vitro mutagenesis, and are shown as shaded rectangles. (B) Schematic showing the position of the Q655R mutation in relation to the entry inhibitor Fuseon (or T-20 peptide), the MPER, and peptides recognized by the broadly neutralizing monoclonal antibodies 2F5 and 4E10. The locations of gp41 structural elements are shown as follows: shaded boxes for the hydrophobic fusion domain (FD), the transmembrane domain (TMD), and the MPER. Sequences defining the C34 and N36 helices are shown as open rectangles.

[0025] Figure 2. Conformational transitions involved formation of the fusion active 6 helix bundle in gp41 . Binding of the HIV envelope protein gp l 20 by CD4 and either of the CXCR4 or CCR5 chemokine receptors (agonists) triggers the formation of the pre-hairpin intermediate structure. This transition is inhibited by antagonists such as CD4 blocking antibodies found in HIV+ sera and the CD4 blocking MAb, bl 2. The transition from the pre-hairpin intermediate to the fusion active 6 helix bundle structure is facilitated by cooperative interactions between the N-36 and C34 helices and the hydrogen bonded ring structure involving Q655. This transition is antagonized by bNAbs in HTV+ sera, MAbs such as 2F5 and 4E10, the antiviral drug, Fuzeon, and mutations such as Q655R, Q655K, Q655S, and Q655E that destabilize the hydrogen bonded ring structure of the 6 helix bundle.

[0026] Fig. SI shows the method of swarm analysis

[0027] Fig. S2 shows the results of the infectivity screen used to identify env clones used in the neutralization assay

[0028] Fig. S3 shows the location of amino acid differences between sensitive and resistant clones

[0029] Fig. S4 shows graphs of neutralization for various clones

General Representations Concerning the Disclosure

[0030] The embodiments disclosed in this specification are exemplary and do not limit the invention. Other embodiments can be utilized and changes can be made. As used in this specification, the singular forms "a", "an", and "the" include plural reference unless the context clearly dictates otherwise. Thus, for example, a reference to "a part" includes a plurality of such parts, and so forth. The term "comprises" and grammatical equivalents thereof are used in this specification to mean that, in addition to the features specifically identified, other features are optionally present. Where reference is made in this specification to a method comprising two or more defined steps, the defined steps can be carried out in any order or simultaneously (except where the context excludes that possibility), and the method can optionally include one or more other steps which are carried out before any of the defined steps, between two of the defined steps, or after all the defined steps (except where the context excludes that possibility). Where reference is made herein to "first" and "second" features, this is generally done for identification purposes; unless the context requires otherwise, the first and second features can be the same or different, and reference to a first feature does not mean that a second feature is necessarily present (though it may be present). Where reference is made herein to "a" or "an" feature, this includes the possibility that there are two or more such features. This specification incorporates by reference all documents referred to herein and all documents filed concurrently with this specification or filed previously in connection with this application, including but not limited to such documents which are open to public inspection with this specification.

Definitions

[0031] The terms "amino acid" and "amino acid sequence" refer to an oligopeptide, peptide, polypeptide, or protein sequence, or a fragment of any of these.

"Amplification" relates to the production of additional copies of a nucleic acid sequence e.g., using polymerase chain reaction (PCR).

The term "antibody" refers to intact immunoglobulin molecules as well as to fragments thereof, such as Fab, F(ab')2. and Fv fragments, which are capable of binding an epitopic determinant.

The term "similarity" refers to a degree of complementarily. There may be partial similarity or complete similarity. The word "identity" may substitute for the word "similarity." A partially complementary sequence that at least partially inhibits an identical sequence from hybridizing to a target nucleic acid is referred to as "substantially similar."

The phrase "percent identity" as applied to polynucleotide or polypeptide sequences refers to the percentage of residue matches between at least two sequences aligned using a standardized algorithm such as any of the BLAST suite of programs (e.g., blast, blastp, blastx, nucleotide blast and protein blast) using, for example, default parameters. BLAST tools are very commonly used and are available on the NCBI web site.

A "variant" of a particular polypeptide sequence is defined as a polypeptide sequence having at least 40% sequence identity to the particular polypeptide sequence over a certain length of one of the polypeptide sequences using blastp with the "BLAST 2 Sequences" tool set at default parameters. Such a pair of polypeptides may show, for example, at least 50%, at least 60%, at least 70%, at least 80%, at least 86%, at least 90%, at least 95%, or at least 98% or greater sequence identity over a certain defined length of one of the polypeptides.

Detailed description of the invention

[0032] Described is a new way to identify mutations that effect sensitivity and resistance to virus neutralization by anti-HIV antisera. Some of these mutations occur in previously undescribed sites critical for preservation of the structure and function of HIV. One of these sites appears to affect a previously overlooked hydrogen bonded ring structure in the trimeric form of the HIV envelope protein, gp41 . This novel structure is formed by oligomeric interactions between the C34 and N-36 helices of gp41 and is located close to the C-terminus of the domains that undergo massive rearrangement to form the 6 helix bundle required for virus entry and fusion. Disruption of this structure by naturally occurring or experimental mutations renders the virus much more sensitive to neutralization by antibodies. Disclosed is the development of and the use of small molecule drugs that target this site which interfere with virus fusion in such a way as to prevent, or lower the efficiency of fusion and therefore virus infection.

[0033] A mutation mapped and herein disclosed is located in the middle of a sequence that forms the basis of a commercially marketed HIV antiviral drug, Fuzeon. The structure identified allows for the rational design of new compounds targeting the same area as Fuzeon, but which work by a different mechanism.

[0034] The molecular structures responsible for HIV fusion have been conserved through evolution, and homologous structures are present in other viruses, such as influenza, and vesicle proteins required for the export and secretion of a number of important molecules (e.g. hormones, cytokines, an neurotransmitters). Targeting weak, hydrogen bonded interactions of the type that we have identified may provide a new approach to the development of small molecule therapeutics that disrupt such structures.

[0035] Disclosed is a new method ("swarm analysis") used to identify mutations that confer sensitivity and resistance to neutralization by bNAbs (broadly neutralizing antibodies) in polyclonal HIV+ sera with broad neutralizing activity. The method takes advantage of the swarm of closely related virus variants that occur in each HIV-infected individual to establish panels of envelope proteins that differ from each other by a limited number of mutations causing amino acid substitutions ( 1 -3%). By studying the effect of these mutations in swarms of viruses from the same individual, we can identify specific amino acids that affect sensitivity and resistance to neutralization by HIV+ sera. We have used this method to identify a novel structural element in the gp41 fragment of the HIV envelope glycoprotein that appears to stabilize the oligomeric 6 helix bundle in the HIV- 1 fusion apparatus. This oligomeric 6 helix structure is important in promoting fusion of the viral membrane to membrane of the host cell being infected. Mutations that affect this structure confer sensitivity or resistance to virus neutralization, i.e., they make the virus more or less sensitive to neutralizing Abs such as broadly neutralizing antibodies.

[0036] The studies described made use of a large collection of clinical specimens from new and recent HIV infection collected in the course of a Phase 3 clinical trial (VAX004) of a candidate HIV-1 vaccine, AIDS VAX B/B (Flynn NM, Formal DN, Harro CD, Judson FN, Mayer H, Para MF; "Placebo-controlled phase 3 trial of a recombinant glycoprotein 120 vaccine to prevent HIV-1 infection." The Journal of infectious diseases

2005; 191 :654-65). This collection of specimens is unique in that they were obtained within six months of infection and are representative of viruses currently circulating in North America. Transmission of HrV-1 involves a genetic bottleneck where, out of the myriad of genetic variants in each HIV infected donor, only a single homogeneous variant of HIV- 1 successfully replicates in the recipient. This variant replicates to very high titers in the first days and weeks after HIV- 1 infection and eventually starts to mutate in response to error-prone reverse transcription to generate a swarm of closely related variants (Richrnan et al., 2003; Wei et al., 2003). The swarm of viruses further diversifies in response to selective pressures imposed by both cellular and humoral antiviral immune responses and in response to drug therapy. Virus variation, driven by the relentless error-prone reverse transcription and selection by the immune system, occurs throughout the course of HIV infection, and is perhaps the greatest challenge in the development of vaccine and therapeutic products. The applicants reasoned that by studying viruses from early infections, sequence variation would be limited compared to sequences collected at later times. The analysis described is made possible by high throughput, automated methods for virus infectivity and neutralization assays as well as systems for the construction and analysis of pseudotype viruses

(Schweighardt et al., 2007, J Acquir Immune Defic Syndr 46: 1 - 1 1 and Whitcomb et al.,2007, Antimicrob Agents Chemother 51 :566-75) with defined amino acid sequences. This technology allows for the accurate and efficient analysis of thousands of individual envelope glycoproteins for sensitivity/resistance to neutralization by panels of HIV+ sera. These analyses provide particular insight into the strategies employed by HIV to evade the immune response and can guide the development of a new generation of HIV vaccine antigens, one or more of which are described herein.

Experimental methods and results

[0037] Cryopreserved plasma was obtained from 28 randomly selected individuals who became infected with HIV during the course of the VAX004 clinical trial. The specimens were all collected from the first post- diagnosis blood draw, with a mean estimated time post infection of 109 +/- 58 days. Populations of gpl60 genes were amplified from each patient plasma by RT-PCR and ligated into a plasmid expression vector to create libraries of envelope genes (Schweighardt et al., 2007). A diagram that describes the swarm analysis strategy is provided in Figure S I . The plasmid libraries from each individual were then used to create pseudoviruses for neutralization assays. Because HIV infection is known to result in a high frequency of defective envelope genes, it was necessary to screen individual clones for infectivity prior to performing virus neutralization assays. For this purpose 24-48 individual colonies were selected from each library, and the plasmids from each used to construct pseudotype viruses for initial screening in infectivity and receptor tropism assays. Data from these infectivity studies on a cell line (CCR5/CD4 U89) expressing CD4 and CCR5 are provided in the supplemental information (Figure S2). Based on the results of this assay, sets of 10 pseudotype viruses with robust infectivity were selected from each individual for use in a pseudotype virus neutralization assay. These 280 pseudotype viruses were then tested for sensitivity/resistance to neutralization by a panel of four standard ΗΓ7+ sera (Z23, Z1679, Z1684, and N 16) known from previous studies to possess bNAbs. The results of these studies provided insights into both virus variation and variation in the specificity of bNAbs in different HrV+ sera. Overall three different neutralization phenotypes were observed in the viruses. We found that one individual ( 1/28) possessed viruses that were extremely resistant to neutralization, such that none of the 10 clones were sensitive to neutralization by any of the HIV+ sera. Conversely we found that some individuals (3/28) possessed viruses that were extremely sensitive to neutralization, such that almost all of the clones were sensitive to neutralization by all four HIV+ sera. However, in the majority of the individuals (24/28), we found a mixture of neutralization sensitive and resistant clones.

[0038] When the activities of the four HIV+ sera were compared, differences in the apparent potency and specificity of the bNAbs were observed. For example in some cases (e.g. 108059) only one of the four sera was able to neutralize the clones from a particular individual (Table 1 A). This result suggested that serum Z23 possessed at least one population of neutralizing antibodies that was missing or under-represented in the antibodies from the other HIV+ sera. One particularly interesting pattern of neutralization was found in subject 108060 (Table I B) where all four ΗΓ + sera neutralized three of the ten clones. These results raised the possibility of a mutational difference between clones that affected a population of neutralizing antibodies common to all four ΗΓν+ sera. Because we expected sequence variation between clones from the same individual to be minimal, we reasoned that comparison of the sequences between the neutralization sensitive and resistant variants would allow us to identify the mutation that conferred neutralization sensitivity.

[0039] Further examination of the dataset revealed that 7/28 individuals exhibited a similar pattern of neutralization sensitivity, where at least one clone was sensitive to neutralization by all four ΗΓ7+ sera and at least one clone was resistant to all four HIV+ sera. Based on this observation, we selected pairs of viruses (one neutralization sensitive, and the other neutralization resistant) from seven of the 28 individuals with the largest differences in neutralization titers for further analysis.

[0040] We next sequenced the envelope glycoproteins from each neutralization sensitive/resistant pair and compared the sequences. In some cases we found that sequence variation was minimal between the two clones from the same individual, whereas in other cases there were a large number of amino acid differences between neutralization sensitive and resistant clones (Figure S3). In one case (subject 108048), there were only two amino acid differences between the neutralization sensitive and resistant clones. In contrast, other viruses (e.g. 108068) showed a large number of amino acid differences (48) between neutralization sensitive and resistant viruses. Pairs with limited sequence variation allowed for the possibility of in vitro mutagenesis to localize the amino acids responsible for conferring sensitivity or resistance to neutralization by HIV+ sera. To explore this possibility, we initially selected the viruses from subject 108060 for further analysis.

[0041 ] Identification of a mutation in gp l 60 from subject 108060 that confers sensitivity to neutralization by HIV+ sera. It can be seen (Table 1 A) that three of the ten clones from subject 108060 (clones 002, 018, and 024) were sensitive to neutralization by all four HIV+ sera, and of the remaining seven clones, most were resistant to neutralization by HIV+ sera Z1679, Z 1684, and N 16, but somewhat sensitive to HIV+ sera from Z23. Based on the fact that there was at least a 10-fold difference in neutralization sensitivity with all four HIV+ sera, clones 022 and 024 were selected for further study. When the gpl 60 sequences of the neutralization resistant variant (clone 022 wtR) and a neutralization sensitive variant (clone 024 wtS) were compared (Figure S3), it was found that they differed at only seven positions. Two of the amino acid differences were in gpl20, two amino acid differences were in the gp41 ectodomain, and the remaining three differences were in the cytoplasmic tail of gp41. To determine which amino acids were responsible for the difference in sensitivity to neutralization between clone 022 and clone 024, a series of mutant envelope proteins were constructed and used to create pseudovirions where polymorphisms from the neutralization sensitive variant (clone 024) were introduced into the neutralization resistant (clone 022) background (Figure 1 A).

[0042] We found (Table 2A) that the replacement of asparagine (N) for serine (S) at position 323 (N323S) in the V3 domain of gp l20 had no effect on sensitivity to neutralization. Similarly, the substitution of N for glycine (G) at position 530 in the C5 domain (N530G) of gpl 20 had no effect. Replacement of lysine (K) at position 634 of the second heptad repeat domain (C34 helix) of gp41 with glutamic acid (E) in the mutant K634E also failed to show a significant difference in neutralization sensitivity.

[0043] However, the replacement of glutamine (Q) for arginine (R) at position 655 (Q655R) resulted in a remarkable increase (>30 fold) in neutralization sensitivity by all four of the HIV+ sera.

[0044] Mutations in the cytoplasmic tail region (832/833 and 827/832/833) were also examined and had no significant effect. The primary data used to calculate 50% neutralization titers with HIV+ serum Z23 are presented in supplemental information (Figure S4). It can be seen that the neutralization curves were well behaved for all of the mutants.

[0045] Localization of residue 655 on linear sequence and 3-D structure of gp41. To better understand the impact of this mutation on the structure and function of the 108060 envelope glycoprotein, we located residue 655 on the linear sequence and 3-dimensional structure of gp41. Examination of the linear sequence (Figure I B) revealed that position 655 was located in the conserved second heptad repeat (HR2) of gp41 in a region also known as the C34 helix. This part of the molecule is known to play an integral role in virus fusion and indeed forms an essential component of the 6 helix bundle in the trimeric structure of gp41 that mediates fusion of the viral membranes with cellular membranes in the course of HIV infection. Position 655 is also located within in the T20 peptide that serves as the basis for the antiviral drug, Fuzeon, that inhibits HIV infectivity by inhibiting virus fusion and entry: Finally the location of this mutation is only eight amino acids from a distinct structural region of gp41 , termed the Membrane Proximal External Region (MPER), that is known to contain distinct epitopes recognized by the broadly neutralizing monoclonal antibodies 2F5, 4E10 and Z13. Taken together, these results suggest that this mutation occurs in a region that is important for virus fusion, and is close to - but structurally distinct from - a region known to contain other epitopes recognized by bNAbs. Interestingly, while the Q655R mutation in the C34 helix of gp41 had a marked effect on virus neutralization, the K634E mutation also in the C34 helix had no significant effect. These results demonstrated that some amino acid substitutions in the C34 helix, but not others, can cause a significant change in sensitivity and/or resistance to neutralization by antibodies in ΗΓ + sera.

[0046] The availability of the PDB (Protein Data Bank) co-ordinates of the gp41 fusion domain allowed us to evaluate the impact of the substitution of R for Q at position 655 upon the structure and function of gp41.

Using the structure of Chan and Kim, we were able to determine that in the fusion activated form of the gp41 trimer, Q at position 655 is located two turns from the terminus of the C34 helix (Figures 2A and 2B) and is subject to both intra-molecular interactions with the N36 helix of the same monomer and inter-molecular interactions with the N-36 helix of adjacent monomers. The N36 and C34 helices within a gp41 monomer pack together in a fairly standard anti-parallel coiled-coil hair-pin structure. The 3-fold symmetric packing interface of the gp41 trimer is mediated almost exclusively by a set of parallel three-helical bundle contacts between the N36 helices of each gp41 monomer. One of the few exceptions to this is the set of contacts mediated by Q655. Although Q655 resides in the C34 helix, its side chain accepts an intramolecular hydrogen bond from Q553 of the N36 helix within the gp41 monomer, and it donates an inter-molecular hydrogen bond to the backbone carbonyl oxygen of V551 in the N36 helix of an adjacent gp41 monomer (Figure 2C). The gp41 trimeric structure is thus stabilized by a "ring" of amino acids Q655-Q553-V551 in a 3-fold symmetric repeat (Figures 2B and 3A). Hence the three copies of Q655 contribute six hydrogen bonds that specifically stabilize the trimeric structure through intramolecular as well as inter-molecular hydrogen bonding contacts. Mutations of Q655 clearly possess a significant potential to disrupt the stability of the tertiary gp41 structure as well as the quaternary structure of the gp41 trimer. Molecular modeling suggested that replacement of Q with R might impact the structure of the 6 coiled-coil bundle in two ways. First, the longer R side chain may have a steric effect that disrupts the close packing of the C34 helix with the N36 helix on the adjacent monomers (Figure 2D). A second mechanism by which this substitution at position 655 could confer neutralization sensitivity is by disruption of the intra-molecular hydrogen bond with position 553 (Figures 2D and 3B), as there is no longer a keto oxygen to act as a hydrogen bond acceptor. Hence the effect of the mutation is predicted to destabilize each of the gp41 monomers in the trimeric structure. However, the potential to form the inter-molecular hydrogen bond with V551 remains, so that if the gp41 monomer can still fold correctly, a partially stable trimer should be able to form.

[0047] Role of inter- and intra-molecular hydrogen bonds. To further investigate the role of R655 in conferring sensitivity to virus neutralization, we used in vitro mutagenesis to replace Q at position 655 with other residues predicted to affect inter- and intra- molecular interactions in the hydrogen bonded ring structure and examined their affect on neutralization sensitivity (Table 2B). Some of the replacements, such as threonine (T), failed to yield infectious viruses. We found that the conservative replacement of Q for asparagine (N) at position 655 resulted in a small but significant increase in neutralization sensitivity. Glutamine and asparagine share the same side chain amide functionality, but asparagine has one fewer side chain carbon atoms than does glutamine. Hence, the Q655N mutation is unique in that it retains the potential to form both the intra-molecular hydrogen bond and the inter-molecular hydrogen bond (Figure 3C), providing that a local distortion of the helical backbone can compensate for the shortening of the side chain by one carbon atom. This observation explains the relative insensitivity of HIV-1 to the Q655N mutation.

[0048] We next examined the replacement of Q at position 655 with lysine (K). The side chain of lysine is shorter than that of arginine and has reduced potential to interfere with the inter-helix packing structure than arginine. Modeling suggested that Q655K mutation, like the Q655R mutation, was unable to form the intra- ₍· molecular hydrogen bond with Q553, but preserved the inter-molecular hydrogen bond with V551 (Figure 3D). We found the Q655 mutation resulted in a highly neutralization sensitive phenotype. This result suggested that the destruction of the hydrogen bond was a more important factor in conferring neutralization sensitivity than the steric hindrance provided by the longer side chain of arginine. This conclusion was confirmed in the next two mutants examined (Q655S) where serine (S) replaced glutamine at position 655, and Q655E where glutamic acid (E) replaces glutamine. We found that these substitutions also resulted in a significant increase in neutralization sensitivity (Table 2B), albeit not as high as the Q655K mutation. The effect of S or E at position 655 is predicted to differ from that of the Q655R mutation in that they preserve the intra-molecular hydrogen bond, but are unable to form the inter-molecular hydrogen bonds (Figure 3E). Together these results suggest that both the inter- molecular and intra-molecular hydrogen bonds are important for stabilizing the ring structure, and that disruption of either the set of three intra-molecular hydrogen bonds or the set of three inter-molecular hydrogen bonds results in increased sensitivity to neutralization.

[0049] Monoclonal Antibody Sensitivity and Envelope Transfer - Sensitivity to neutralization by MAbs and fusion inhibitors. While the structural analysis provided insight into the functional consequences of mutations at position 655, two alternate hypotheses can account for a mechanism by which this mutation increases sensitivity to antibody-mediated neutralization. One possibility is that this mutation is located at or near an antibody binding site and that the Q655R mutation restores an epitope recognized by a population of neutralizing antibodies present in all four HIV-positive sera. Alternatively, it is possible that this mutation results in a significant conformational change that is transmitted to other parts of gp41 such as the adjacent MPER or the gp l 20/gp41 trimer complex in such a way as to increase exposure or access to antibodies at other locations on the molecule.

[0050] To explore these possibilities, antibody neutralization studies were carried out with a panel of neutralizing MAbs to epitopes in gpl 20 and gp41 as well as fusion inhibitors targeting either the gp l 20 or the gp41 portion of the HIV envelope glycoprotein. In these studies, we examined two broadly gp41 -neutralizing MAbs, 2F5 and 4E10; the broadly neutralizing b l 2 antibody able to block CD4 binding to gp l 20; and 2G 12, an antibody that binds to a carbohydrate epitope in gp l 20. In addition, we tested the antiviral entry inhibitor CD4- IgG, which binds to sequences in gp l20 and is able to neutralize lab-adapted CXCR4-dependent clinical isolates at low concentrations (0.01 to 0.1 Mg/ml), and primary clinical isolates of HIV at high concentrations (10 to 100 Mg/ml). We also examined the sensitivity of envelope mutants to enfuvirtide, a peptide virus entry inhibitor that consists of a gp41 -derived peptide that includes sequences from the C34 helix containing Q655.

[0051 ] The results of these studies are shown in Table 4, in which the sensitivities of clone 022 and clone 024 from subject 108060 to neutralizing MAbs were compared. It can be seen that the neutralization-resistant clone 022 is moderately sensitive to the 2F5 and 4E10 MAbs specific for the MPER of gp41 but resistant to neutralization by the bl 2 and 2G12 MAbs reactive with gpl 20. This virus was also sensitive to enfuvirtide and resistant to CD4-IgG. The high CD4-IgG concentration required for the neutralization of this virus is consistent with the concentration required to neutralize other primary, CCR5 -dependent viruses.

[0052] We next examined the neutralization-sensitive clone 024 that differs from the neutralization-resistant clone 022 at only seven amino acid positions. We found that this clone was 15- to 20-fold more sensitive to the MPER-specific MAbs (2F5 and 4E10) than the 022 clone. Similarly, the neutralization-sensitive clone 024 was more than 20-fold more sensitive to CD4-IgG and 3.5-fold more sensitive to neutralization by enfuvirtide (Table B). Thus, clone 024 exhibited significantly increased sensitivity to neutralization by MAbs and antiviral entry inhibitors as well as antibodies in HIV-positive sera.

[0053] We then mutated the neutralization-sensitive clone 024 so as to replace R with Q at position 655. We found that the resulting mutant ( 108060_024 R655Q) became resistant to neutralization and showed a pattern of neutralization sensitivity closely resembling that of the neutralization-resistant clone 022. Conversely, when we mutated the neutralization-resistant clone 022 to replace Q at position 655 with R, the resulting mutant

( 108060_022 Q655R), which differed from the parental neutralization-resistant clone by a single amino acid, exhibited an extraordinary increase in neutralization sensitivity (Table 3). We observed a >125-fold increase in sensitivity to CD4-IgG compared to that of the wild-type clone 022 and a 30- to 35-fold increase in sensitivity to the MPER-reactive antibodies 2F5 and 4E 10. We also noted a 17-fold increase in sensitivity to the antiviral drug enfuvirtide.

[0054] These results highlight the importance of glutamine at position 655 and suggest that epistatic mutations at other sites in clone 024 moderate sensitivity to neutralization. The results of these studies are remarkable in that they show that a single amino acid substitution in gp41 not only confers sensitivity to neutralization by MAbs and entry inhibitors directed to gp41 but also increases sensitivity to CD4-IgG, a molecule that binds to gp l 20, an entirely different protein. Thus, the Q655R mutation appears to cause a conformational change in gp41 that affects not only the binding of antibodies and entry inhibitors (2F5, 4E 10, and enfuvirtide) that bind close to the site of the mutation but also the binding of another inhibitor (CD4-IgG) that binds to a site on gpl 20 located a considerable distance from the mutation.

[0055] Transfer of the Q655R mutation to related and unrelated viruses. In order to determine whether the Q655R mutation could confer neutralization sensitivity and resistance to other viruses, this mutation was introduced into two unrelated viruses highly resistant to neutralization (from subjects 108069 and 108051 ) that normally possessed a Q at a position corresponding to 655 of the virus from subject 108060 (the 108060 virus). The results of these experiments are shown in Table 3. Interestingly, we found that the replacement of Q655 with R had little or no effect on neutralization by any of the HIV-positive sera. However, these mutations significantly increased the sensitivity to neutralization by the 2F5 and 4E10 MAbs (25- to 35-fold). These mutations also increased the sensitivities to neutralization by the entry inhibitors enfuvirtide and CD4-IgG. Thus, the mutation of Q to R at a position corresponding to 655 in the 108069 virus increased the sensitivity to enfuvirtide by more than 17-fold and increased the sensitivity to CD4-IgG by more than 20-fold. The 108069 mutant with the Q655R mutation seemed to be somewhat more sensitive to enfuvirtide and possibly CD4-IgG than the corresponding mutant of the 108051 virus.

[0056] Together, these results demonstrate that the mutation of Q to R at positions corresponding to 655 of the 108060 virus confers sensitivity to neutralizing MAbs to the MPER and antiviral compounds. targeted to the C34 helix and the MPER of gp41. However, it was interesting that these mutations failed to increase the sensitivity to bNAbs in HIV-positive sera. We do not know whether neutralizing activity in HIV-positive sera is attributable to antibodies binding to the C34 region, the MPER, or other parts of the molecule. It has been recently reported that antibodies with specificities similar to 2F5 and 4E10 are rare in HIV-positive sera, which might account for the lack of effect. Alternatively, the failure of the Q655R mutation to increase neutralization sensitivity by HIV-positive sera might be attributable to polymorphisms outside of the MPER and the C34 region that preclude the binding of otherwise bNAbs. This may well be the case since the 108069 and 108051 viruses were selected because of their resistance to neutralization by the HIV-positive sera selected for use in these studies.

[0057] Transfer of Q655R mutation to related and unrelated viruses. In order to determine whether the Q655R mutation could confer neutralization sensitivity and resistance to other viruses, this mutation was introduced into two unrelated viruses ( 108069 and 108051 ) that normally possessed a Q at a position corresponding to 655 of the 108060 virus. The results of these experiments are-shown in Table 3. Interestingly, we found that replacement of Q655 with R had little or no effect on neutralization by any of the HrV+ sera (supplementary information S4). However, these mutations significantly increased the sensitivity to neutralization by the 2F5 and 4E 10 MAbs (25- to 35-fold). These mutations also increased sensitivity to neutralization by the entry inhibitors Fuzeon and CD4-IgG. Thus the mutation of Q to R at a position corresponding to 655 in 108069 increased the sensitivity to Fuzeon by more than 17-fold and increased the sensitivity to CD4-IgG by more than 20-fold. The 108069 mutant with the Q655R mutation seemed to be somewhat more sensitive to Fuzeon and possibly CD4- IgG than the corresponding mutant in the 108051 virus. Together these results demonstrate that the mutation of Q to R at positions corresponding to 655 of 108060 confers sensitivity to neutralizing MAbs and anti-viral compounds targeted to the C34 and MPER regions of gp41 and to the CD4 binding site in gp l 20. However, this mutation is not able to confer sensitivity to neutralization by bNAbs in HIV+ sera to all viruses.

Discussion

[0058] These studies utilized a novel method for the identification and mapping of mutations that affect the sensitivity/resistance of viruses to neutralization by HIV+ sera and anti-viral entry inhibitors. This approach differs from previously described methods of mutational analysis used to study HIV in that it relies on naturally occurring mutations in the swarm of closely-related viruses that evolve during the course of HIV infection.

[0059] Identification of a mutation at position 655 in gp41 that confers sensitivity to neutralization by bNAbs In this study we identified a naturally occurring mutation (Q655R) that affects sensitivity/resistance of viruses to neutralization by bNAbs. X-ray crystallography studies showed that glutamine at position 655 is located close to the C-terminus of the C34 helix and contributes to two hydrogen bonds: one mediating an intra-molecular interaction with the N36 helix on the same monomer, and the other mediating an inter-molecular interaction with the N36 helix on an adjacent monomer. These two hydrogen bonds appear to stabilize the fusion active conformation of the 6 helix bundle in trimeric gp41 in such a way as to increase infectivity and confer resistance to neutralization. Our data suggest that naturally occurring mutations (e.g. Q655R) and experimental mutations (e.g. Q655K , Q655S or Q655E) that interfere with either the intra-molecular or inter-molecular hydrogen bonds normally provided by Q655 confer sensitivity to neutralization by interfering with the formation of the hydrogen bonded ring. In this regard, the function of this ring structure appears to be twofold: 1 ) to stabilize interactions between the backbones of adjacent N-36 helices in the core of the 6 helix bundle and 2) to stabilize the ends of the coiled-coil hairpin structures in each gp41 monomer. This latter interaction may serve a function analogous to the fibular clasp on brooch or badge.

[0060] The mechanism by which the Q655R mutation confers sensitivity and resistance to neutralization. HIV fusion is thought to be a step-wise process that begins with the binding of CD4 and a suitable chemokine receptor (CXCR4 or CCR5) to gp l 20. This triggers a conformational change resulting in the formation of the "pre-hairpin" fusion intermediate complex via rearrangement of the amphipathic helices in the external domains of gp41. The N36 helices pack in a parallel three-helical bundle. The pre-hairpin is characterized by the exposure of the N-terminal hydrophobic fusion domain and the C-terminal MPER of gp41 which are normally folded inside the gp41 trimer and not exposed to circulating antibodie. Further molecular rearrangements result in closure of the hairpin structure, resulting in anti-parallel packing of each C34 helix into the grooves on the outside of each N- helix in the gp41 trimer. Ultimately, a highly thermostable 6 helix bundle is formed, which is thought to provide the energy required to fuse viral with cellular membranes. We hypothesize that the Q655R mutation alters the conformational equilibria (Figure 4) so as to favor the pre-hairpin fusion intermediate structure where both the N terminal fusion domain and MPER are exposed. This would explain the increased sensitivity to the 2F5 and 4E10 MAbs, which recognize the exposed MPER as well as the increased sensitivity to neutralization by CD4-IgG. Previous studies have suggested that transition of the fusion intermediate to the fusion-active conformation is the rate limiting step in virus infection, and is estimated to be in the range of 15 minutes based on T-20 (Fuzeon) sensitivity. An interesting possibility is that HIV envelope glycoproteins that are "trapped" into the fusion intermediate conformation might represent superior HIV vaccine antigens, since they would expose epitopes normally hidden, and only exposed during virus fusion. The results obtained with swarm analysis are consistent with the possibility that mutations at 655 in the 108060 virus, such as Q655R and Q655K, alter the conformational equilibria to favor the gp41 trimer in the fusion intermediate conformation (Figure 4).

[0061] Sera from early infections may represent an opportunity to identify rare mutations that confer sensitivity to bNAbs. Based on the examination of sequence data in the Los Alamos HIV Sequence database, it appears that the mutation of arginine for glutamine at position 655 is extremely rare and occurs with an observed frequency of 8/1242 (0.64%). How is it then that we were able to find such a rare mutation within the first seven viruses examined? One possible explanation relates to the fact that the viruses analyzed in this study were all collected close to the time of infection, and may possess antigenic structures that are uncommon in viruses recovered from later infections due to kinetics of the development of the neutralizing antibody response. Several studies have shown that bNAbs do not occur until 6-12 months after infection. It could well be the case that viruses recovered from early infections possess a broader range of antigenic features because they are being selected primarily for infectivity rather than neutralization resistance. Once effective neutralizing antibodies are present, neutralization sensitive variants, such as Q655R, would be selected against, and rapidly disappear from plasma. The possibility that viruses from early infections may contain mutations resulting in unusual structures is consistent with a previous study where viruses recovered from the same clinical cohort as 108060 had an unexpectedly high frequency of mutations that affected the disulfide structure of gpl 20. .

[0062] Envelope proteins from early infections with rare mutations such as Q655R may represent a new source of vaccine antigens. How are mutations that occur with such low frequencies useful for HIV vaccine development? The results obtained for the Q655R mutation suggest that mutations of this type significantly alter the antigenic structure of the envelope protein in such a way as to expose important epitopes that are normally shielded from contact with the immune system. Frey et al. have hypothesized that immunization with a gp41 trimer locked into the pre-hairpin fusion intermediate conformation might be an effective way to elicit bNAbs to the MPER with activities similar to 2F5 and 4E10. We believe that the Q655R and other mutations that we have described may have "trapped" the gp41 trimer into this pre-hairpin intermediate conformation and might be effective in inducing bNAbs. The immunogenicity of such variants has not yet been explored; however, studies are underway to examine their immunogenic potential.

[0063] Virus fusion is a delicately balanced process that involves major conformational transitions triggered by ligand binding. These transitions are no doubt aided, and stabilized, by a variety of cooperative interactions. The studies described highlight a set of novel interactions mediated by hydrogen bonds that appear to facilitate fusion of viruses with cellular membranes. The 6 helix bundle structure and fusion mechanism is conserved throughout evolution and is essential for the infectivity of most enveloped viruses. A homologous 4 helix bundle plays a similar role in cellular vesicles mediating intracellular transport and secretion. It may well be that the infectivity of other enveloped viruses (e.g. influenza) as well as membrane fusion processes (e.g. intracellular transport and secretion) might also depend on stabilizing interactions from hydrogen bonded structures of the type that we have observed in gp41. Knowledge of these stabilizing interactions may be useful in understanding the details of the fusion process and may provide a new approach to the development of vaccine and therapeutic products, where alteration of these interactions may provide a functional benefit.

Materials and Methods

[0064] Sera and Plasma. Cryopreserved plasma used to clone full length envelope glycoproteins were collected in the course of a Phase 3 clinical trial of a candidate HIV vaccine (AIDSVAX B/B) sponsored by VaxGen, Inc. (S. San Francisco, CA). Deidentified specimens and data required for these studies were provided by Global Solutions for Infectious Diseases (S. San Francisco, CA). All of the viruses used in this study were obtained from patient plasma collected within six months of initial infection. HIV+ sera containing broadly neutralizing antibodies (Z23, Z1679, Z1684, and N16) were provided by Monogram Biosciences, Inc. (S. San Francisco, CA) and are known from previous studies to neutralize a variety of primary clinical isolates of HIV. The monoclonal antibodies used in these studies were obtained from two different sources. The broadly neutralizing monoclonal antibodies b l 2, 2F5, and 4E10 were obtained from the NIH AIDS Reagent Repository and Polymun A.G. (Vienna, Austria). The antiviral compound CD4-IgG was described previously and provided by GSJD (S. San Francisco, CA).

[0065] Construction of envelope gene libraries and pseudoviruses. Libraries of envelope glycoprotein were created from each subject by PCR amplification of full length envelope genes from cryopreserved plasma using the method described previously. The swarm of PCR products was cloned into a plasmid expression vector useful for the construction of pseudoviruses. The vector was specifically designed to permit the construction of pseu- dovirus libraries for use in a well-established and validated virus neutralization assay. However, instead of pooling all of the clones together and carrying out neutralization assays or drug sensitivity assays with an entire library of cloned genes from each infected individual as had been done previously, we plated out the plasmid library on agar plates and picked 24-48 clones from each individual for infectivity studies. The plasmid DNA was isolated from each clone and used to create a stock of pseudovirus particles that were then screened for infectivity, and chemokine receptor usage. After verifying infectivity and receptor usage, we then selected approximately ten CCR5-dependent pseudotype viruses with good infectivity for virus neutralization assays. The virus neutralization assays were carried out as described by Schweighardt et al.

^[0066] Sequencing and mutagenesis. Plasmids containing cloned envelope glycoproteins were sequenced using fluorescently labeled dideoxynucleotides at either Monogram Biosciences or the University of California Sequencing Facility (Berkeley, CA) using capillary electrophoresis sequencing devices (Applied Biosystems, Foster City, CA). HIV envelope glycoprotein sequences were mutagenized by a mismatched primer method using the QuikChange Mutagenesis kit (Stratagene, San Diego). All mutations were confirmed by DNA sequencing. The numbering of amino acids is made with reference to the sequence of gpl60 from clone 022 of from subject 108060. Position 655 corresponds to position 653 of the HXB2 reference strain of HIV- 1.

[0067] Virus neutralization assay. The automated virus neutralization assay described in this study has been described previously. The neutralization data reported represent IC50 values calculated from serum dilution curves. This assay employs multiple assay controls, including a positive pseudotype virus control panel and a negative pseudotype virus control panel. Assay acceptability criteria have been established to minimize interas- say variability and assure comparability of data from different experiments. The positive virus control panel includes the pseudotypes from the neutralization sensitive isolate, NL43, and the less neutralization sensitive primary isolate JRcsf. The negative virus (specificity) control consists of pseudotype viruses prepared from the envelope of the amphitropic murine leukemia virus. Previous studies (Wrinn, Montefiore, and Sinangil, manuscript in preparation) have shown that the Monogram virus neutralization assay yields comparable results to the TZM-BL pseudotype virus neutralization assay when tested on standard panels of HIV-1 isolates distributed by the NIH.

[0068] Molecular modeling. Although the complete gp41 HIV-1 glycoprotein structure is currently unavailable, a crystal structure comprising the N36 and C34 helices of gp41 (PDB accession code 1 AIK) anti-parallel helical core duplicates the essential intra-molecular as well as inter-molecular packing interactions in which the crystallographic three-fold axis corresponds to the natural gp41 trimer three-fold axis. The intra-molecular and inter-molecular hydrogen bonding contacts involving Q655 were identified in the context of the gp41 trimeric structure in the PyMOL molecular visualization software package. The potential effects of the various Q655 mutations upon both sets of packing interactions were then analyzed by in silico mutagenesis in PyMOL combined with crystallographic symmetry-constrained energy minimization molecular modeling (using the crystallographic software package Phenix to enforce the gp41 trimeric symmetry. The results of the crystal-structure-based 9] molecular modeling efforts were subsequently analyzed in PyMOL

TABLE 1. Neutralization in 108059 and 108060

A 108059 Wild Type Viruses B 108060 Wild Type Viruses

Sera / Neutralization Titers* Sera / Neutralization Titers

Clone Z1679 Z1684 N16 Z23 Clone Z1679 Z1684 N16 Z23

002 <40 <40 <40 251 022 53 58 51 1 17

005 <40 <40 <40 234 024 804 609 612 1667

008 <40 <40 <40 244 002 303 160 195 379

010 <40 <40 <40 238 003 69 57 67 151

013 <40 <40 <40 196 01 1 136 130 177 222

014 <40 <40 <40 436 012 62 57 70 241

016 44 50 49 490 013 53 50 58 158

018 <40 <40 <40 167 018 428 243 388 1378

021 <40 <40 <40 278 019 44 <40 40 145

023 <40 <40 <40 258 021 47 47 70 157

The neutralizing antibody titer (IC50) is defined as the reciprocal of the plasma dilution that produces a 50% inhibition in target cell infection. Values in bold represent neutralization titers that are at least 3 times greater than those observed against the negative control (aMLV). All clones tested were CCR5 tropic. Clone indicates gpl60 envelope genes.

TABLE 2. Neutralization of Wild Type (WT) and Mutated Clones from Subject 108060 by HIV+ sera possessing broadly neutralizing antibodies

Mutation of Clone 022 wtR from 108060 B Mutation at Position 655

Sera / Neutralization Titers* Sera / Neutralization Titers*

Clone /

Z1679 Z1684 N16 Z23 Clone mutant Z1679 Z1684 N16 Z23 Mutants

022 wtR 75 104 76 384 022 wtR 40 <20 36 281

024 wtS 728 1086 982 1926 024 wtS 1099 1193 545 4167

N323S 73 95 54 382 022 Q655R 14276 2876 2610 8422

N530G 37 42 41 308 022 Q655K 5486 8590 4276 19476

K634E 67 73 72 346 022 Q655E 564 132 366 2424

Q655R 2165 2562 4472 8290 022 Q655S 1565 472 674 2650

I827T 39 <20 113 <100 022 Q655N 148 24 57 820

832/833 104 50 63 404 022 I827T 49 <20 <20 277

827/832/833 72 53 81 279 024 R655Q 50 <20 39 372

The neutralizing antibody titer (IC50) is defined as the reciprocal of the plasma dilution that produces a 50% inhibition in target cell infection. Values in bold represent neutralization titers that are significantly above the background (Experimental Procedures). All clones tested were CCR5 tropic. Clone indicates gpl60 envelope genes. wtR and wtS indicate wild type neutralization-resistant and -sensitive clones respectively.

TABLE 3. Transfer of Q655R Mutation to Unrelated Viruses:

Sensitivity to Neutralizing Monoclonal Antibodies and Entry Inhibitors

IC50 MAbs and Fusion Inhibitors ^g/ml)

Clone Mutation CD4

2F5 4E10 bl2 2G12 Fuzeon

IgG

108060_{. .}022 wtR 3.250 5.201 >20 >20 0.068 >20

108060_{. .}022 Q655R 0.093 0.156 >20 >20 0.004 0.161

108060_{. .}024 wtS 0.151 0.333 >20 >20 0.019 0.798

108060_{. .}024 R655Q 3.434 6.546 >20 >20 0.130 >20

108069 .005 wtR 1.129 3.556 >20 >20 0.071 >20

108069_. .011 wtS 0.043 0.040 >20 >20 0.145 >20

108069. .005 Q655R* 0.052 0.044 >20 >20 0.011 1.080

108051_{. .}005 wtR >20 >20 >20 >20 0.088 >20

108051_{. .}006 wtS 1.176 1.369 >20 >20 0.008 0.231

108051. .005 Q655R* 0.343 1.314 >20 >20 0.036 5.209

*Numbering with reference to 108060 protein.

The neutralizing antibody titer (IC50) is defined as the concentration ^g/ml) of mAB or entry inhibitor that produces a 50% inhibition in target cell infection. Values in bold represent neutralization titers that are significantly above the background (Experimental Procedures). All clones tested

were CCR5 tropic. Clone indicates gpl60 envelope genes. wtR and wtS indicate wild type neutralization-resistant and -sensitive clones respectively.

Table 4. Sensitivity to neutralizing monoclonal antibodies and entry inhibitors in 108060 clones and unrelated viruses (a)

ICs_p (μ§ ηι1) of indicated MAb or fusion inhibitor

Clone Mutation ^: ■ ^■■

2F5 4E10 bl2 2G12 Enfuvitlide CD4-IgC

108060 022 wtR 3.250 5.201 >20 >20 0.068 >20

108060^"022 Q655R 0.093 0.156 >20 >20 0.004 0.161

108060~024 wtS 0.151 0.333 >20 >20 0.019 0.798

10S060_024 R6550 3.434 6.546 >20 >20 0.130 >20

108069 005 wtR 1.129 3.556 >20 >20 0.071 >20

108069~011 wtS 0.043 0.040 >20 >20 0.145 >20

108069 005 Q655R* 0.052 0.044 >20 >20 0.011 1.080

1080 1 005 wtR >20 >20 1 >20 >20 0.088 >20

108051 006 wtS 1.176 1.369 >20 >20 0.008 0.231

10805 l^"005 0655R* 0.343 1.314 >20 >20 0.036 5209

" The neutralizing antibody titer (1C₅₀) is defined as the rancoiitration (μ¾'ΊηΙ) of an MAb or entry inhibitor that produces a 50% inhibition in target cell infection. Values in bold represent neutralization titers thai are significantly above the background (see Materials and Methods). All ckmes tested were CCR5 tropic. Clones indicate gpl60 envelope proteins. wtR and wtS indicate wild-type neutralization-resistant and -sensitive clones, respectively.

* Numbering with reference to subject 10S060 protein.

Claims

1. A method for screening drug candidates to identify a drug that prevents or attenuates the ability of HIV gp41 to mediate cell fusion with a CD4+ cell, the method comprising the following steps:

a) providing a putative drug candidate

b) providing a membrane-bound trimeric HIV gp41 trimer

c) measuring the integrity of the hydrogen-bonded ring structure of the HIV gp41 trimer

d) contacting the membrane-bound trimeric HIV gp41 with the putative drug candidate

e) re-measuring the integrity of the hydrogen-bonded ring structure of the HIV gp41 trimer

f) whereby degree of integrity of the hydrogen-bonded ring structure of the HIV gp41 trimer is proportional to the ability of HIV gp41 to mediate cell fusion with a CD4+ cell, and wherein the ability of the membrane- bound trimeric HIV gp41 to mediate cell fusion is reduced by disruption of the hydrogen-bonded ring structure of the HIV gp41 trimer.

2. The method of claim 1 wherein the membrane -bound trimeric HIV gp41 is bound in the membrane of a virus, a pseudovirus, or a transfected cell.

3. The method of claim 1 wherein disruption of the hydrogen-bonded ring structure results in exposure of previously hidden epitopes which bind specifically with broadly neutralizing antibodies.

4. The method of claim 1 wherein the drug candidate is an antibody that binds specifically with an epitope of gp41 or a neutralizing antibody that targets gpl20.

5. A method for increasing the immunogenicity of HIV envelope proteins the method comprising exposing the HIV virus to a drug compound that disrupts the hydrogen bonded ring structure between the N36 and C34 helices of gp41.

6. The method of claims 1 of 5 wherein the drug compound is a small molecule.

7. The method of claims 1 of 5 wherein the drug compound is an antibody.

8. The method of claims 1 of 5 wherein the drug compound is an inhibitor of CD4 binding selected from the group consisting of: 4E10, 2F5,Q665R, Q655K and Q655E.

9. The method of claims 1 or 5 wherein the drug compound is an inhibitor of HIV fusion binding and becomes a more effective inhibitor in the presence of a molecule that disrupts the disulfide bonded ring structure of gp41.

10. The method of claims 1 or 5 wherein the drug compound disrupts the hydrogen bonded ring structure between the N36 and C34 helices of gp41 and thereby exposes neutralizing epitopes which are recognized by endogenous or exogenous antibodies which then are able to neutralize the virus.

11. A method for screening for a drug that prevents or attenuates intracellular membrane fusion, the method comprising exposing the multimeric coiled coil bundle of the activated fusion complex to a drug candidate, wherein the fusion complex comprises a cellular hairpin membrane fusion protein, wherein disruption of one or more hydrogen bonds of the fusion complex is associated with prevention or attenuation of intracellular membrane fusion.

12. The method of claim 11 wherein the intracellular membrane fusion is associated with secretion of a hormone, cytokine or neurotransmitter.

13. A method of treating, attenuating or preventing HIV infection, the method comprising administering to a patient a drug compound which disrupts one or more intra-molecular or inter-molecular hydrogen bonds of the hydrogen-bonded ring structure of gp41 trimer, wherein disruption of the hydrogen-bonded ring structure is associated with attenuation or prevention of HIV infection.

14. A synthetic helical peptide wherein the peptide sequence binds specifically to at least a fragment of the N-36 helix of gp41, wherein the fragment includes the residue Q655 and wherein binding of the synthetic helical peptide to the N-36 helix disrupts hydrogen bonded ring structure between the N36 and C34 helices of gp41.

15. A peptidomimetic drug that binds specifically to helical sequences adjacent to the Q655 or Q553 residues of gp41, and disrupts of prevents the formation of a hydrogen bonded ring structure involving Q655 from the C34 helix and Q533 from the N34 helix.

16. The peptide or peptidomimetic compound of claim 14 wherein the compound binds to or interacts with the N36 or C34 helices of gp41 and thereby disrupts one or more intra-molecular or inter-molecular hydrogen bonds of the hydrogen-bonded ring structure of gp41 trimer, wherein disruption of the hydrogen -bonded ring structure is associated with attenuation or prevention of HIV infection.

17. The peptide or peptidomimetic compound of claim 14 wherein the compound disrupts one or more of the intramolecular or inter-molecular hydrogen bonds stabilizing the multimeric coiled coil bundle in the activated fusion complex required for the fusion and release of synaptic vesicles.

18. The peptide or peptidomimetic compound of claim 14 wherein the compound disrupts one or more of the intramolecular or inter-molecular hydrogen bonds stabilizing the multimeric coiled coil bundle structurally homologous to the N36 or C34 helix of HTV in the activated fusion complex required for the fusion and release of vesicles or granules containing pro-inflammatory proteins, cytokines, hormones, or vasoactive substances.

19. A method of attenuating or preventing HIV-1 infection comprising administering to a patient an effective amount of an agent which disrupts one or more intra-molecular or inter-molecular hydrogen bonds of the hydrogen-bonded ring structure of gp41 trimer, wherein disruption of the hydrogen-bonded ring structure makes HIV-1 susceptible to neutralization by the patient's antibodies which thereby attenuate or prevent HIV infection.

20. The method of claim 19 wherein the agent comprises a peptide or peptidomimetic compound.

21. A peptide having a formula selected from the group consisting of:

X- YTSLIHSLIEES QNQ [ * ] EKNEQELLELDKWASLWNWF-Z

X- YTNTIYTLLEESQNQ [ * ] EKNEQELLELDKWASLWNWF-Z

X- YTGIIYNLLEESQNQ [ * ] EKNEQELLELDKW ANLWNWF-Z

X-YTSLIYSLLEKSQIQ[*]EKNEQELLELDKWASLWNWF-Z

X-LE ANIS KSLEQ AQIQ [ * ] EKNM YELQKLNS WDIFGNWF-Z, and

X-LE ANIS QSLEQ AQIQ [ * ] EKNM YELQKLNS WD VFTNWL-Z,

in which:

amino acid residues are presented by the single -letter code;

X comprises an amino group, an acetyl group, a 9-fluorenylmethoxy-carbonyl group, a hydrophobic group, or a macromolecule carrier group;

Z comprises a carboxyl group, an amido group, a hydrophobic group, or a macromolecular carrier group.

[*] represents any amino acid other than Q or N.

22. The peptide of claim 21 wherein [*] represents R, K, S, or E.