WO1990004978A1

WO1990004978A1 - Method of directing immune response to a viral binding site

Info

Publication number: WO1990004978A1
Application number: PCT/US1988/003886
Authority: WO
Inventors: Tran C. Chanh; Ronald C. Kennedy; Vernon C. Maino; Noel Warner
Original assignee: Becton Dickinson and Co
Current assignee: Becton Dickinson and Co
Priority date: 1988-11-01
Filing date: 1988-11-01
Publication date: 1990-05-17
Anticipated expiration: 1991-05-01

Abstract

Immunogenic compositions comprising a plurality of monoclonal antibodies capable of (1) blocking binding of a virus to its receptor and (2) inducing formation of an anti-idiotypic antibody capable of binding to the virus are provided, along with techniques for producing such compositions.

Description

METHOD OF DIRECTING IMMUNE RESPONSE

TO A VIRAL BINDING SITE

FIELD OF THE INVENTION

This invention is related to the use of antibodies as immunogens and is particularly directed to the use of anti-CD4 antibodies as immunogens to direct an immune response against the HIV virus.

BACKGROUND OF THE INVENTION

Acquired Immune Deficiency Syndrome (AIDS) is a devastating disease resulting from infection of critical cellular components necessary for maintenance of the immune system. The etiologic agent is a human T lymphotrophic retrovirus previously known as LAV, HTLV- III, or ARV, and now called human immunodeficiency virus (HIV-1) by consensus. This and other HIV isolates (such as HIV-2), infect, replicate in, and are) cytopathic for T cells with the helper/inducer phenotype that is defined by the presence of the CD4 antigen. Infection with the AIDS virus can result in a progressive depletion of CD4⁺ T cells that eventually leaves the patient susceptible to opportunistic infections and/or malignancies that are the clinical manifestations of AIDS.

Considerable scientific evidence now exists showing that the CD4 molecule on the surface of -the T helper/inducer cells functions as a receptor for HIV. It has recently been shown that the HIV interaction with CDM occurs via an envelope glycoprotein having a molecular weight of 110,000-120,000 known as gp120 (also referred to in some publications as gp110 because of its variable molecular weight). The gp120 molecule is the larger subunit of the gplβO envelope protein of HIV which, upon maturation of the virus, is processed into an exterior 120 kd subunit (gp120) and a smaller transmembrane 41 kd subunit (gp41).

The CD4 antigen on T lymphocytes has been characterized as being a 60,000 Dalton protein expressed on approximately 65% of human peripheral blood T lymphocytes. The antigen is often defined by its ability to react with monoclonal antibodies, specifically the Leu3a monoclonal antibody manufactured by Becton- Dickinson. Antibodies reactive with the same protein molecule (but not necessarily the same determinant on this molecule) are 0KT4 and 0KT4a, manufactured by Ortho, and anti-T4, manufactured by Coulter, Inc.

In vitro studies have demonstrated that infection of T cells can be effectively blocked by monoclonal antibodies such as anti-Leu3a and OKT4a reactive with the CD4 molecule. By binding to CD4, an antibody would block binding by gp120 of HIV and consequent cell infection. Later studies characterized a number of anti-CD4 antibodies capable of blocking virus infectivity to CD4⁺ lymphocytes. Such data demonstrated that only antibodies that reacted with a specific region of the CD4 molecule were able to block virus infectivity. The prototype monoclonal antibody which defines this region is the anti-Leu3a anti-CD4 antibody.

However, production from anti-CD4 antibodies of an effective neutralizing anti-idiotypic antibody against HIV has proven difficult. A study of four candidate anti-CD4 monoclonal antibodies that were potent inhibitors of virus binding (0KT4a, 0KT4d, 0KT4f, and Leu3a) attempted to use these individual monoclonal antibodies to produce anti-idiotypic sera in rabbits that would bind to HIV-1. However, the anti-idiotypic sera raised against each of the four candidate CD4 monoclonal antibodies did not react with virus nor in- hibit virus binding to CD4⁺ T cells. Although one monoclonal anti-idiotypic antibody, identified as HF1.7, was generated against anti-Leu3a and was shown to bind HIV-1 in another study, even this anti-id monoclonal antibody was only partially capable of neutralizing HIV infection of human T cells in vitro. Accordingly, there remains a need for antibody preparations capable of evoking an immune response to produce antibodies capable of neutralizing HIV infections. A method of directing an immune response to a viral binding site would be potentially used for numerous other viral infections in addition to HIV infection.

RELEVANT LITERATURE

Dalgleish et al., Nature (1984) 312 : 763, describes anti-idiotypic technology relating to HIV.

McDougal et al., J. Immunol. (1986) 137:2937-2944 describes the binding of HIV to the CD4 molecule and the production of rabbit-anti-idiotypic sera raised against anti-CD4 monoclonal antibodies. Chanh et al., Proc. Natl. Acad. Sci. USA (1987) 84: 3891-3895 describes a monoclonal anti-idiotypic antibody capable of mimicking the CD4 receptor and binding HIV. Fingeroth et al., Proc. Natl. Acad. Sci. USA (1984) 81: 4510 describes the receptor for Epstein-Barr virus (EBV).

SUMMARY OF THE INVENTION

The present invention provides a method of directing an immune response to a specific viral binding site. The method provides a collection of antibodies that can be used as an immunogen, for example to produce a diagnostic antiserum or as a vaccine. For example, binding of HIV particles to human cells can be prevented by inducing an immune response in human cells to a plurality of gp120-internal-image anti-CD4 antibodies of different specificities. As a result of the immune response, antibodies against the gp120-internal- image anti-CD4 antibodies will be produced that are capable of binding the gp120 antigen and neutralizing HIV infection. By providing a plurality of different specificities in the gp120-internal-image anti-CD4 antibodies presented to the human B cells, a broad immune response occurs which is better able to neutralize polymorphic HIV viruses. The plurality of gp120- internal-image anti-CD4 antibodies could alternately be used to produce a diagnostic antiserum capable of detecting the presence of HIV. Similar antibody collections can be prepared to other viruses if the binding site on the virus and the receptor on the host cell are known.

DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS

The present invention provides a method of preparing a collection of antibodies that can be used as a vaccine against a viral infection, such as acquired immune deficiency syndrome (AIDS), or can be used to prepare an antiserum used to diagnose the presence of infective viral particles. In contrast to classical vaccine antigen approaches for directing immune responses, the immunogenic preparation described in this disclosure is not based on the virus itself or on virus- derived materials. Nor need any viral substance be manufactured or administered to human subjects. Instead the composition is generated by a selection procedure applied to antibodies prepared using the viral receptor as an immunogen. Specifically, the composition contains a collection .of monoclonal antibodies of different specificities that are reactive with the viral receptor, such as HIV-reactive epitope(s) on the CD4 antigen molecule found on the cell surface of T helper lymphocytes. By following the specified selection procedure, which in preferred embodiments is broken down into three steps, a collection of antibodies is obtained that is capable of eliciting an immune re- sponse to provide an antiserum or cell mediated response that reacts with the viral binding site; i.e., the immune response produces receptor-like anti-idio- typic antibodies. The collection of antibodies is therefore useful either as a vaccine or as an agent for inducing an antiserum used in diagnosis or for other purposes.

The three selection procedures are carried out on monoclonal antibodies (referred to as the source collection of antibodies) produced using the receptor for the virus as an immunogen. These monoclonal antibodies can be prepared by any suitable technique, such as innoculating an animal with a cell having a virus- specific receptor on its surface, carrying out an in vitro or in vivo immunization using a purified viral receptor, or any other technique for inducing antibody formation to the receptor molecule. If the source collection is prepared using a crude receptor preparation (such as a whole-cell or a cell-membrane preparation) as the immunogen, an initial screening is usually necessary to identify antibodies reactive with the receptor, as opposed to antibodies induced by other immunogens. This is readily accomplished if antibodies that are known to bind specifically to the receptor are known, such as the Leu 3a antibody discussed above. If the receptor is not known, it should be identified (characterization by binding to specific antibodies is sufficient for this purpose) in order to identify the source collection of antibodies. Examples of known receptors for viruses are the CD4 receptor for HIV discussed above and the 03d receptor (also known as CR2 or CD21) for Epstein-Barr virus (EBV). The method can be applied to these or other known receptors or can be applied to currently unknown virus receptors as they are discovered.

Since the receptor molecule is usually much larger than its binding site with the virus particle, only a small fraction of the monoclonal antibodies produced using the receptor as an immunogen will react with determinants at the binding site. The first key step of the present invention is to select monoclonal antibodies from the source collection that react with the binding site on the receptor that interacts with the virus particle. Such antibodies can be detected by their ability to interfere with the binding of viral particles (or the viral coat protein or other protein known to interact with the receptor) with the receptor. The resulting antibodies are referred to as virus-blocking antibodies and represent a first collection on which later selection techniques can be carried out.

The second selection is to obtain from the virus-blocking antibodies those antibodies that are capable of inducing an immune response that cross- reacts with the virus. In other words, the first anti- body (Ab1) induces production of a second antibody

(Ab2), also known as the anti-idiotypic antibody (anti- id), which reacts with the virus. Since Ab2 reacts both with Ab1 and the virus, Ab1 and the virus have similar binding sites, and Ab1 is referred to as having a viral internal image.

If an Ab1 collection is eventually to be used as a vaccine, it is preferred to use the host organism or a closely related species to raise Ab2, since different organisms can respond differently to the same immunogen (here Ab1). Thus, if Ab1 is being tested for possible use in a human vaccine, Ab1 should either be innoculated into a human (or a non-human primate, for ethical reasons) or an in vitro human immune response should be elicited in order to provide the best possible guidance in selecting immunogenically active antibodies. However, if an Ab1 collection is to be used to produce an antiserum that will be used in a diagnostic procedure, any active immune system can be used to generate Ab2, and even when Ab1 will be used In a vaccine, the system used to generate Ab2 is not required to be the host system. The binding of Ab2 to the viral particles (or the purified surface protein or other component known to react with the receptor) can be carried using any technique that determines the binding of an antibody to an antigen. Examples include competitive reactions in which Ab2 interferes with the binding between a purified receptor and a purified viral protein, a host cell and a whole virus, or a combination of purified components and cells or viruses.

After virus-blocking antibodies have been selected for viral internal images, it is still necessary to carry out an additional selection technique, since use of a single virus-internal-image virus- blocking monoclonal antibody may be ineffective in neutralizing viral activity because of its high specificity and resulting narrow range of binding. The final product composition of the present invention is a mixture of antibodies selected by the two procedures set forth above which have different specificities. Such antibodies, bind to slightly different locations on the receptor and induce formation of antibodies that interact with slightly different locations at or near the binding site on the virus particle. By providing a mixture of 2-100, preferably 2-25, and more preferably 2-5 monoclonal antibodies having the other binding properties indicated above but with slightly different specificities, a broader spectrum vaccine or an anti- genic mixture capable of providing a broader diagnostic capability is obtained.

in order to better define individual aspects of the invention, the preparation of antibodies and selection procedure is described in connection with HIV and its receptor CD4. However, it should be recognized that the detailed description which follows can be applied to the preparation of an antibody collection for use as a vaccine or diagnostic antigen when the receptor that interacts with the virus is known. In the HIV-CD4 system, Ab1 antibodies are prepared against the CD4 antigen. Either purified CD4 antigen, a crude preparation of the antigen, or cells known to present CD4 antigen on their surface can be used as the immunizing agent since the selection process will eliminate antibodies of other specificities. However, the amount of screening that must be done can be minimized by using a purified CD4 preparation as the immunogen. Anti-CD4 monoclonal antibodies are obtained using standard techniques.

However, mere binding to CD4 is not sufficient since many determinants on the CD4 molecule are distant from the binding site of CD4 with the gp120 antigen of HIV. Accordingly, an anti-CD4 monoclonal antibody of the invention must be capable of blocking HIV binding to CD4. However, such blocking of HIV binding to CD4 is itself insufficient to define an antibody of the invention, since, as indicated in the background section, monoclonal antibodies capable of blocking HIV binding have previously been known which were incapable of inducing an immune response to provide the desired protective effect. Accordingly, HIV-blocking anti-CD4 monoclonal antibodies must then be subjected to a selection procedure in which the HIV-blocking anti-CD4 monoclonal antibodies (Ab1) are used to induce anti- idiotype antibodies (Ab2). The ability of the resulting-anti-id antibodies to bind with the gp120 envelope protein from HIV is then determined. If the HIV- blocking anti-CD4 monoclonal antibody induces an anti- idiotypic antibody reactive with the gp120 envelope protein, it is considered to have an internal image of the gp120 envelope protein and is a candidate for use in preparing the antibody collection of the invention. Such antibodies are referred to as gp120-internal-image anti-CD4 antibodies. However, a further selection remains to be made before a collection of antibodies suitable for practice of the present invention is designated. The collection must contain a plurality of gp120-internal- image anti-CD4 antibodies with different specificities. By providing such a collection, an immune response is generated in the challenged host that is better able to neutralize an HIV infection than would be possible by challenge with a single gp120-internal- image anti-CD4 monoclonal antibody. Similarly, if an antiserum is being prepared for a diagnostic test, a broad response range will be obtained. The Ab1 collection of Gp120-internal-image anti-CD4 monoclonal antibodies will also be useful as an HIV mimic for a control sample that will not require use of active virus particles.

In part, the present invention arises as a result of extension of the original characterization of CD4 epitopes using a new and large panel of monoclonal anti-CD4 antibodies . generated in the laboratories of the inventors. Initial characterization of these antibodies is shown in a table in the examples which

follow.

A superior collection of Ab1 monoclonal anti- bodies to be used for active immunization protocols, to generate antisera, or to mimic viral particles can be selected on the basis of the following criteria: (1) antibodies that demonstrate the most effective blocking of HIV infection as determined by syncytia formation, fluorescence binding assays, and blocking of reverse transcriptase activity during HIV infection of susceptible human cell lines and (2) antibodies within group (1) antibodies capable of eliciting strong HIV-reactive, neutralizing anti-idiotypic responses, particular in human systems if the Ab1 collection is to be used as a vaccine. Neutralizing can be measured in vitro by measuring reduction of either viral reverse transcriptase activity or syncytia formation; analyses for both conditions are well known in the art. A gp120-internal- image anti-CD4 preparation which induces a neutralizing anti-idiotypic response should inactivate divergent HIV isolates.

In a related screening test, it is possible to use combinations of antibodies from group (1) that demonstrate simian immunodeficiency virus (SIV) blocking activity on monkey peripheral blood lymphocytes. It has been shown that SIV, a virus very similar to HIV- II, causes a similar disease in Rhesus monkeys. SIV is capable of binding to the CD4 glycoprotein on human cells, and this interaction is also blocked by anti- Leu3a monoclonal antibodies. It therefore appears that SIV uses the same determinant (s) as a receptor on monkey CD4. The recent demonstration by the inventors that some anti-CD4 monoclonal antibodies which demonstrate effective blocking of HIV binding on human T cells but not blocking of SIV binding to these cells suggests that SIV binding can be used to determine fine specificity of CD4 epitopes defined by monoclonal antibodies involved in HIV (and gp120) binding.

The individual assays used in selection steps in preparing the indicated collection of antibodies are not limited to particular types of assay but can be any assay which demonstrates the required property (such as blocking of virus binding). Many such assays are now known and numerous others will doubtless be rapidly developed because of the continuous research and development being carried out in the HIV field and other areas of virus research. Several procedures are generally described below and references are given in order to provide guidance in achieving the proper selection of antibodies.

Anti-CD4 monoclonal antibodies can be produced by standard techniques. The anti-CD4 monoclonal antibodies can be obtained from any species or can be a chimeric antibody of the type described in Morrison et al., Proc. Natl. Acad. Sci. USA (1984) 81 :6851-6855. For example, mice can be immunized intraperitoneally or intravenously with CD4⁺ cells. Examples of CD4⁺ cells include the myelomonocytic cell line U937, the acute lymphocytic leukemia (ALL) cell line HPB-ALL, and the CD4-transfected (EL4) mouse cell line 119-16S2. A typical immunization consists of a series of inoculations over several days prior to isolation and fusion of splenocytes. Numerous fusion partners can be used, such as the mouse fusion partner SP2/o/Agl4. Viable fused cells, prepared by polyethylene glycol-mediated fusion or other fusion techniques, are initially screened for reactivity with CD4⁺ cells or purified preparations of CD4 antigen. Monoclonal hybridomas are prepared from anti-CD4 producers, and their antibodies purified by standard techniques, such as affinity chromatography.

These or any other anti-CD4 antibodies are then subjected to a procedure which selects for blocking of HIV binding to the CD4 receptor. One such method involves the inhibition of binding of labeled HIV particle to CD4⁺ cells as measured by flow cytometry. Alternatively, antibodies which block the binding of purified or recombinant gp120 to CD4⁺ cells can be selected as virus-blocking antibodies.

Once HIV-blocking anti-CD4 monoclonal antibodies have been selected, the resulting collection is again treated to select for gp120-internal-images-.

This selection will typically be carried out by generating anti-idiotypic antibodies in vivo or in vitro followed by determining the ability of the thus-produced anti-id antibodies to react with HIV or gp120. A typical immunization would be carried out as described above. Anti-id antibodies would then be incubated in vitro with HIV to determine if the anti-id antibody is capable of binding to and/or neutralizing the virus. Procedures can also be carried out to determine if the resulting anti-id recognizes the gp120 envelope protein, such as the well known Western blot analysis.

For example, in a neutralization assay, the treated HIV is added to CD4⁺ cells. At intervals, aliquots of culture fluids are removed and reverse transcriptase activity determined by standard methods. Comparison of reverse transcriptase activity to controls will allow determination of whether infection of cells in in- hibited.

Once gp120-internal-image anti-CD4 antibodies have been selected, they are combined to produce an effective immunogenic preparation of the invention.

Since the specificities of the individual components of the collection are intended to differ and since specificities are determined by the amino acid sequence of the variable region of the antibodies, one technique for selecting different specificities is to select antibodies having a different variable region amino acid sequence. However, any characteristic that indicates a difference in specificity can be used to select the collection of antibodies. For example, epitope mapping and/or differences in reactivity with CD4 antigens can also be used to select a plurality of the desired antibodies.

For example, chimeric human/mouse CD4 molecules can be generated and used to detect difference in binding of the anti-CD4 antibodies. CD4 and L3T4 cDNA gene segments can be cloned and transformed simultaneously into a bacterial host where homologous recombination events can occur. Using this system, a series of CD4/L3T4 and L3T4/CD4 recombinants have been generated. Each of these recombinants have a short human segment in a mouse receptor molecule, the human segment appearing at different locations in the different molecules.

These recombinant molecules can be used to measure both anti-CD4 reactivity and HIV-I binding. In fact, once the recombinant molecules have been used to identify Ab1 antibodies having different specificities, the recombinant molecules themselves can be used as immunogens to produce dupl i cat e Ab 1 ant i bodi es of ident i cal s peci f i c i ti es when injected into mice, since only the human segment of the recombinant molecule will be immunogenic.

An additional embodiment of this invention is the therapeutic use of the proposed collection of anti- bodies in the treatment of AIDS. Since the response to this preparation would be an active immune response against the HIV binding site, the Ab1 collection can be used for therapy in AIDS as an active immunization to induce the anti-id response. However, patients who already have the disease and would therefore be the group considered for therapy have by definition a compromised immune system and therefore may not be able to mount an effective anti-id response against an Ab1 panel. This patient group is characteristically unable to make good immune responses against neo-antigens. However, their ability to mount a good response against previously sensitized antigens as carriers is not severely impaired. Therefore, for therapeutic applications of the described vaccine, the Ab1 panel will be conjugated by standard chemical means to immunogens (preferably protein in nature) to which the patient has been previously sensitized. Such immunogens are generally known as carrieradjuvants. Examples of carrier adjuvants include diptheria, pertussis and tetanus toxoid (DPT). Techniques for attaching immunogens to carrier proteins are well known and are described in detail in the scientific literature. See, for example, Burdon and Knippenberg, eds., Laboratory Techniques in Biochemistry and Molecular Biology, Elsevier, pp. 221-241, for a review of macromolecular conjugation. This mixture will be administered in an acceptable adjuvant form, preferably as an alum precipitate. The invention now being generally described, the same will be better understood by reference to the following examples which are provided for purposes of illustration only and are not to be considered limiting of the invention unless so specified.

EXAMPLES

Generation and Characterization of

CD4 Specific Monoclonal Antibodies

Female Balb/a mice were immunized intraperitoneally (I/P) or intravenously (I/V) with CD4⁺ cells. These CD4⁺ cells included PHA or ConA activated (CD8 depleted) peripheral blood lymphocytes, the myelomono- cytic cell line U937, the acute lymphocytic leukemia (ALL) cell line HPB-ALL, and the CD4 transfected (EL4) mouse cell line, 119-16S2. Generally, the immunization scheme consisted of 2-5 I/P inoculations of 10' cells, followed by an I/V dose of 5x10° cells three days prior to fusion. Immune splenocytes were fused to the non- secreting mouse fusion partner SP2/o/Ag14 using a modified polyethylene glycol mediated-fusion protocol. The primary screen, usually at 7-10 days post fusion, was performed on a Pandex Screen Machine, using a method Particle Concentration Fluorescence Immunoassay (PCFIA) adapted for viable cells. Hybrid supernates were assayed against CD4⁺ and CD4^" cell lines (HPB-ALL and LB respectively). Antibody-containing supernates from the primary screens which reacted in an appropriate fashion were then subjected to single-color FACS analysis on peripheral blood mononuclear cells. CD4 specific antibodies could be recognized by their characteristic staining pattern and the percentage of cells stained. These antibodies were confirmed as being CD4 specific by (1) two-color FACS analysis against standard CD4 reagents (anti-Leu3a, anti-Leu3b) and (2) immunoprecipitation of the 55-kD CD4 antigen from io dinated HPB-ALL cell lysates. All hybridomas derived in this way were subjected to at least two rounds of cloning by limiting dilution, and representative clones were injected I/P into pristane primed Balb/c mice for ascites production. Purified antibody preparations were obtained from these ascites either by Protein A or by goat anti-mouse Ig affinity chromatography.

Epitope Analysis

CD4 specific antibodies were purified as outlined above, and FITC conjugates of each prepared.

Titer points for each FITC-conjugated CD4 antibody were determined by FACS analysis. Blocking studies were then performed by incubating 10 peripheral blood lymphocytes with an excess (3-5 μg) of unlabeled antibody followed by addition of a given FITC conjugate at its titer point (usually 0.1-1 μg per 10° cells). A checkerboard analysis using a panel of unlabeled vs. labeled CD4 antibodies demonstrated that a number of distinct epitopes were being recognized (see Table 1). Studies involving the binding of these CD4 antibodies to peripheral blood lymphocytes of cynomologus monkeys quickly demonstrated that antibodies which cluster together in the blocking assays clearly are different. In Table 1 it can be seen that the antibodies anti-Leu 3a, L77, L93, L34 and L80 appear identical; however, L34 and L80 did not stain monkey CD4⁺ cells indicating that L34 and L80 are recognizing different epitopes to those recognized by Leu 3a, L77 and L93.

Two apparently unique CD4 epitopes are represented by the antibodies L83 and L88. L83 is unusual in that it does not block the binding of any other CD4 antibody conjugate, and blocking of L83 FITC conjugate can only be achieved by incubating cells with cold unlabeled L83. The reverse is true of L88, in that only L83 fails to block L88 FITC, and unlabeled L88 blocks all anti-CD4 FITC conjugates except L83. Conformation

al changes in the CD4 molecule can be induced by binding of antibody to a specific epitope on the CD4 molecule. Cynomologus monkey cells do not stain positively with FITC conjugates of L34. L80, L69, or L88, i.e., the epitopes recognized by these antibodies are not expressed. However, monkey cells will stain positively with these FITC conjugates after pretreatment with cold L83, indicating that L83 binding results in a refolding of the protein molecule with subsequent expression of previously hidden epitopes.

Generation of Chimeric Human/Mouse CD4 Molecules Using Bacterial in vivo Recombination

Briefly, the CD4 and L3T4 cDNA (Maddon et al., Cell (1986) 42:93; Littman et al., Proc. Natl. Acad. Sci. USA (1983) 80:3968) gene segments are cloned into a separate trpB defective plasraid. When these plasmids are transformed simultaneously into a trpB negative bacterial host, two homologous recombination events canoccur. The first detectable event is in the trpB region, generating a functional trpB gene and allowing growth on selection media. The second event is in the only other homologous region of the two plasmids, in the CD4 and L3T4 gene segments. The second event is detected by altered size mobility of plasmid DNA in agarose gels and by antibiotic resistance patterns of plasmid containing bacterial colonies. With this system, a series of nested CD4/L3T4 and L3T4/CD4 recombinants can be generated. The exact site of recombination can be mapped easily using restriction enzyme analyses followed by dideoxy sequencing of the appropriate DNA fragment. Transfection of the recombinant cDNA into L3T4 negative murine cells will provide cells expressing the chimeric protein which can be tested for the presence of both anti-CD4 reactivity and HIV-1 binding. We already have generated several CD4/L3T4 recombinant molecules using, this system. HIV-binding Assays

One of the parameters to evaluate anti-CD4 antibodies, to determine which idiotypes (V region sequence) should be used in an immunogenic Ab1 vaccine preparation, is to measure the capacity of anti-CD4 antibodies to block HIV binding to its receptor.

A method used to define those antibodies that block HIV binding is to measure the inhibition of fluorochrome-conjugated (FITC or phycoerythrin) HIV binding to CD4⁺ cells by flow cytometry. This analysis has been conducted using flow cytometry with fluorescent conjugates of isolated heat inactivated HIV as described by Achour et al., Ann. Inst. Pasteur (1986)

137:291. This assay has been shown to be specific and enables the identification of HIV susceptible cells. It can also be used to characterize further the viral receptors from these cells. In a manner similar to the procedure described for analysis of gp120 binding (below), the panel of anti-CD4 monoclonal antibodies described can be evaluated for their ability to block binding of HIV. The antibodies which demonstrate the most effective HIV blocking activity will be further evaluated for their ability to block other parameters of HIV infection, including syncytia formation and reverse transcriptase activity.

Related methods for assessing which anti-CD4 antibodies block HIV binding can be used. Since HIV binds to the CD4 molecule via the gp120 envelope protein, anti-CD4 antibodies can be assessed by their ability to block binding of purified or recombinant gp120 to CD4⁺ cells. Binding of gp120 to CD4⁺ cells can be measured with either ^{1 25}I labeled gp120 in a direct binding assay or bound gp120 will be detected using a ¹²⁵I labeled anti-gp120 antibody and binding quantitated using a gamma counter. ¹²⁵l-BSA or irrelevant immunoglobulin can be employed as a non-specific control and specific binding will be assessed. For these studies, CD4⁺ cell lines, including CEM, HPB-ALL, and HUT-78 will be evaluated for gp120 binding. Anti- CD4 antibodies can be titrated against a concentration of gp120 which produces 50% maximal binding.

Titration curves generated from the resulting data will be used to compare relative efficacy of the anti-CD4 antibodies to block specific gp120 binding. These studies will employ similar cell and virus concentrations described by McDougal et al., J. Immunol. (1986) 137:2937-2944.

Incubation of gp120 with CD4⁺ cells will be performed at 37°C for 30 minutes. In some experiments immunoprecipitation will be performed following cross- linking of the labeled gp120/CD4 complex with dithiobis(succinimidyl)propionate, a chemical cross-linker that can be cleaved under reducing conditions. The complexes are evaluated by polyacrylamide gel electro- phoresis (PAGE).

This analysis will allow direct assessment of the capacity of various anti-CD4 monoclonal antibodies to block gp120 binding. Specific blocking of the binding site by anti-CD4 monoclonal antibodies should result in a reduction in the amount of identified gp120. The anti-CD4 monoclonal antibodies also will be evaluated for their ability to precipitate the cross-linked complexes. Those antibodies capable of precipitating the cross-linked complexes can be mapped outside the gp120 binding region of CD4 since the binding of gp120 would have no effect on the accessibility of their epitopes.

Binding studies with gp120 also can be performed with fluorescent-conjugated preparations of this protein using either fluorescein (FITC) or phycoerythrin (PE) derivatives. Binding of gp120 to CD4⁺ cells and inhibition of binding by anti-CD4 antibodies will be evaluated using flow cytometry (FACS). Preparation of fluorescent conjugates will be performed by standard methods. Blocking studies can also be performed in the reverse direction; preparations of purified gp120 can be evaluated for inhibition of fluorescent-conjugated anti-CD4 antibodies. These studies will be conducted only with those antibodies that demonstrate inhibition of gp120 binding since the amount of available gp120 is likely to be limiting.

Using one of the techniques described above, blocking of fluorochrome-conjugated HIV and analysis by flow cytometry, the ability of various antibodies to block HIV binding was determined. As shown in the table below, eight different monoclonal antibodies were prepared that reacted with the CD4 antigen. However, only five of these antibodies were capable of blocking HIV binding, and one of these produced only weak blocking. The Figure shows in graphical fashion the blocking of HIV binding by the Leu-3a antibody and the lack of such blocking by the L-71 antibody.

Enzyme-linked Immunosorbent Assays

The HTLV-III ELISA (Electro-Nucleonics, Inc., Silver Spring, MD) and the LAV EIA (Genetic Systems, Seattle, WA) are carried out according to the manufacturers' specifications. Horse radish peroxidase-goat anti-mouse IgG antibodies (Vector Laboratories, Burlingame, CA) can be substituted for the goat anti-human IgG enzyme conjugate. The ELISA using psoralen and UV- inactivated HIV-1 can be done as described previously (16). Briefly, culture supernate from 7 day HIV (HTLV- IIIB) infected A3.01 cells are concentrated 10-fold and inactivated by exposure (three times) to psoralen (Calbiochem, San Diego, CA) at a final concentration of 10 μg/ml and UV light (360 nm) for 15 minutes. The inactivated HIV is passed over a column of Sepharose 4B. The void volume is collected and concentrated for use. The ELISA is performed using microtiter plates coated with 50 μl/well containing 50 μg/ml of antigens whose concentration is estimated by using an extinction coefficient of 14 for a 1 % solution at 280 nm .

Neutralization of HIV Infection in vitro

This assay is performed as previously described (16). Briefly, 1000 or 100 TCID₅₀ of HIV in100 μl is incubated with 100 μl of antibody or control protein for one hour at 37°C. The concentrations of monoclonal antibodies is adjusted to yield a final concentration of 0.5 mg/ml. After incubation, the treated HIV is added to 10° A3.01 cells and is incu- bated at 37°C for 2 hours in the presence of Polybrene at 10 μg/ml. The cells are then washed and resuspended in RPMI 1640 10% FCS at a density of 10⁶/ml. At indicated time intervals, aliquots of culture fluids are removed and reverse transcriptase (RT) activity is determined as described in Chanh et al., EMBO J. (1986) 5:3065. Cell-free HIV is harvested from chronically infected A3.01 cell culture and titrated on uninfected A3- 01 cells. The titer is expressed as 50% tissue culture infective dose (TCID₅₁). Purification of Antl-CD4 Monoclonal Antibodies

The collection of selected anti-CD4 antibodies which will be used as an immunogen is purifed from mouse ascites using High Performance Liquid Chromatography on a Waters 600 system. Monoclonal antibodies in either serum or ascites are diluted 1:10 with buffer A (25 mM MES, pH 5.4), filtered through a 1.45 μm filter and loaded on a Bakerbond ABx mixed-mode column (10 mm x 250 mm) at 2.5 ml/min. The column is then washed with buffer A until the A₂₈₀ is less than 0.010. A linear gradient is developed from 100% buffer A to 75% buffer A, 25% buffer B (500 mM Ammonium Sulfate, 20 mM Sodium Acetate, pH 7.0) over 40 minutes. The column is then taken to 100% buffer B over 10 minutes, washed with 10 column volumes water, and re-equilibriated with buffer A. Due to slight variations among different monoclonal antibodies, specific elution conditions are developed for each monoclonal antibody prior to scale- up for preparative purification. Typical yields are approximately 100 mg purified IgG per run. This system also will purify monoclonal antibodies under pyrogenfree conditions following GMP procedures.

The central question asked of these antibodies is whether they induce a particular type of immune response. Hence for these studies, purity of the materi- al is not a key criterion. The HPLC procedure described has yielded immunoglobulin preparations of at least 80% purity, adequate for this study; however, for vaccination of chimpanzees, strict GMP procedures can be used and product quality equivalent to that required for human use can be manufactured. The purification

procedures are set forth below.

Clarification of Ascites or Serum

The containers of ascites or serum are removed from the freezer and thawed in the refrigerator. The thawed material is centrifuged at 10,000 rpm at 2-10°C for 30-60 minutes to remove additional fibrin clots that have formed. Alternately, the material can be ultracentrifuged, if necessary, at 30,000 rpm at 2-10°C for 30-90 minutes to aid in removing lipid. The clarified material is aseptically withdrawn from the centrifuge tubes, leaving behind any additional fibrin clots, and refrigerated at 2-10°C.

Precipitation of Monoclonal Antibody

An ammonium sulfate solution that is saturated at 4°C is prepared by mixing ammonium sulfate and sterile water for injection. The pH of the ammonium sulfate is adjusted to 7.2 with ammonium hydroxide. A calculated volume of the solution is added drop-wise to the ascites or serum to give a final solution that is 45% (v/v) ammonium sulfate solution, stirred and incubated overnight at 2-10°C. The suspension of precipitated monoclonal antibody is then centrifuged at 9,000- 20,000 rpm for 30 minutes at 2-10°C. The supernatant is carefully poured off and discarded. The precipitate is dissolved in TRIS Buffer and precipitated once again by the addition of saturated ammonium sulfate as described above. The precipitate is again dissolved in TRIS Buffer.

Desalting

Following ammonium sulfate precipitation, the monoclonal antibody is desalted chromatographically by passage over a column of Cellufine GH25, equilibrated with TRIS Buffer. The monoclonal antibody elutes in the excluded volume of the column, free of ammonium sulfate. In some cases, the product is frozen and stored at this point in the processing. Ion-exchange Chromatography

The desalted monoclonal antibody is further purified by ion-exchange chromatography on DEAE-Fractogel. Selective elution of monoclonal-antibody enriched fractions is achieved by adjusting the relative % of sodium chloride in the running buffer. Further purification, if necessary, is achieved by a second ion-exchange chromatography step on a Mono-Q chromatography column. Again, selective elution of monoclonal-anti- body enriched fractions is achieved by adjusting the relative % of Buffers A & E. In some cases, the product may be frozen and stored either following ionexchange chromatography and/or Mono-Q chromatography. Re-precipitation by Ammonium Sulfate

The antibody is reprecipitated wiih a 50% (v/v) ammonium sulfate solution. This solution is added drop-wise to the antibody, stirred and incubated over night at 2-10°C. The suspension of precipitated monoclonal antibody is then centrifuged at 9,000-10,000 rpm for 30 minutes at 2-10°C. The supernatant is carefully poured off and discarded. The precipitate is dissolved in monoclonal antibody diluent. Initial Filtration

The product is initially filtered after bufferexchange and before the adjustment of concentration. The antibody solution is filtered into a non-pyrogenic glass bottle. A sterile, non-pyrogenic, low protein binding, non-fiber releasing, 0-2 micron membrane filter is used to filter the solution.

Samples are taken at the bulk stage before bottling to determine immunoglobulin concentration, pyrogenicity , identify, purity and activity. In some cases, the product may be frozen in bulk after initial filtration. If frozen at this point, the product is thawed at temperatures between 2-10°C All steps are performed under aseptic conditions inside the clean room after the inital filtration step.

Adjustment of Concentration

The concentration of monoclonal antibody is determined by analytical High Pressure Liquid Chromatography (HPLC). A calculated amount of monoclonal antibody diluent is added to adjust the concentration to the product specifications as required. The adjusted solution is then ready for final filtration.

Final Filtration

A sterile, pyrogen-free, low protein binding, non-fiber releasing, 0.2-micron membrane filter is used to finally filter the antibody solution into a sterile, non-pyrogenic glass bottle, under aseptic conditions.

Labeling and Storage of the Bulk

Container Prior to Aseptic Filling

The bulk container of investigational monoclonal antibody product is labeled "In-Process" and lists the product name, lot number, quantity, date produced, and storage conditions. The bulk container is used for the aseptic filling of vials.

Aseptic Filling

The filling of vials is conducted under aseptic conditions since the monoclonal antibody product must be sterile and pyrogen-free at the time of clinical use. Since the product does not normally contain any added preservative system, contamination by microorganisms or particulates must be prevented by filling under strict aseptic conditions. The use of laminar flow air that is HEPA filtered, complete clean room apparatus, including gloves, sterilization of filling machines and ancillary equipment, and other aseptic conditions is maintained during all filling operations. All publications and patent applications mentioned in this specification are indicative of the level of skill of those skilled in the art to which this invention pertains. All publications and patent applications are herein incorporated by reference to the same extent as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference.

The invention now being fully described, it will be apparent to one of ordinary skill in the art that many changes and modifications can be made thereto without departing from the spirit or scope of the appended claims.

Claims

WHAT IS CLAIMED IS:

1. A method of preparing a binding-site-specific immunogenic composition, comprising:

selecting, from a source collection of monoclonal antibodies prepared using a viral receptor as an immunogen, Ab1 monoclonal antibodies capable of blocking binding of a virus to said receptor to provide a first collection of antibodies;

selecting, from said first collection of antibodies, antibodies capable of inducing an immune response to provide anti-idiotype Ab2 antibodies capable of binding to said virus to provide a second collection of antibodies; and

selecting, from said second collection of antibodies, a plurality of antibodies of different specificities.

2. A method of Claim 1, wherein said virus is HIV.

3. The method of Claim 2, wherein said receptor is CD4.

4. A compos it ion compris ing a plurality of vi rus- internal-image virus-blocking monoclonal antibodies of different specificities.

5. The composition of Claim 4, wherein said virus is HIV.

6. A composition comprising a plurality of virus- internal-image virus-blocking antibodies of different specificities prepared by the method of Claim 1.

7. The composition of Claim 4, wherein said antibodies are conjugated to an immunogenic adjuvant.

8. The composition of Claim 7, wherein said adjuvant is diptheria, pertussis, or tetanus toxoid.