WO1991009625A1 - Monoclonal antibodies which neutralize hiv-1 infection and their anti-idiotypes - Google Patents

Abstract

Monoclonal antibodies are revealed which bind to the gp120 protein on the envelope of HIV-1, both in the PND region and in the CD4-binding region. These antibodies neutralize HIV-1. They inhibit the infection of T cells, and also inhibit syncytium formation. Further, the antibodies are group-specific and neutralize different strains and isolates of HIV-1. These antibodies have a variety of uses, including the treatment and prevention of AIDS and ARC. The antibodies are used in immunotoxins which are specifically toxic to HIV-1 infected T cells. In addition, anti-idiotypic antibodies to these HIV neutralizing antibodies, or derivative products such as chimeric antibodies and human antibody fragments, can be used to immunize against HIV-1.

Description

MONOCLONAL ANΗBODIES WHICH NEUTRALIZE HIV-1 INFECTION AND THEIR ANTI-IDIOTYPES

Background of the Invention

1. Neutralizing and Anti-CD4 Binding Site Antibodies Acquired Immune Deficiency Syndrome, generally known by its acronym AIDS, is probably the most serious health threat confronting society. It could reach epidemic proportions in the general population before the end of this century. The disease runs a painful and debilitating course and results in the death of its victim. In fact, from diagnosis onward, the average life span of an AIDS victim is only a few years.

AIDS is caused by a virus which has at various times been called human T-cell lymphotropic virus type III (HTLV-III), or lymphoadenopathy-associated virus (LAV). The virus is currently known as human immunodeficiency virus I (HIV-1).

HIV-1 also causes a somewhat less serious immunodeficiency syndrome known as AIDS related complex (ARC). ARC will often precede the onset of AIDS. There are currently many more ARC cases than there are AIDS cases. As

SUBSTITUTE SHEET the number of cases continues to increase, ARC will, in itself, become an extremely costly and serious health problem.

AIDS results because infection with HTV-1 depletes T helper/inducer lymphocytes (hereinafter referred to as "T cells"). T cells are essential because they control the production of antibodies by the B cells, the maturation of cytotoxic T lymphocytes (killer T cells), the maturation and activity of macrophages and natural killer cells, and, directly and indirectly, numerous other regulator and effector functions of the immune system. Therefore, HIV-1 infection severely compromises the immune response, leaving the victim unable to defend against secondary opportunistic infections. It is often the secondary infections which debilitate the victim and cause death.

In addition to their susceptibility to secondary infections, AIDS victims frequently develop otherwise rare conditions. A large number develop a rare form of skin cancer known as Kaposi's sarcoma.

Infection of a T cell with HTV-1 follows from interaction between an epitope borne by HIV-1 and a receptor site which is located on the T cell surface, known as the CD4 antigen. The epitope on HIV-1 is borne by the envelope glycoprotein gpl20

(molecular weight 120,000 daltons). The glycoprotein gpl20 is produced when a precursor glycoprotein gpl60 is cleaved apart into gp41 (molecular weight 41,000 daltons) and gpl20.

HIV-1 is a retrovirus. After the virus has entered the cell, a viral enzyme called reverse transcriptase transcribes the viral genomic RNA into DNA in the host cell nucleus. The newly synthesized DNA is incorporated into the host cell genome under a variety of activation conditions, and the infected T cell begins to transcribe the new DNA to make copies of messenger RNA and genomic RNA. The viral genomic RNA's are packed with core proteins, reverse transcriptase, and certain other proteins. They are then enveloped by parts of the cellular membrane and budded off from the cell as newly synthesized virions. These new virions can enter and infect other T cells.

There are two known mechanisms by which HIV-1 is transmitted to other T cells. The first occurs when the free virus binds to the CD4 antigen on the T cells. The second is through direct, cell-to-cell transmission of the virus.

Direct, cell-to-cell transmission occurs when an infected cell, which expresses the viral gp 120 on its surface, binds with the CD4 antigen of an uninfected cell or cells. As a result, the cells fuse and virions can pass to the uninfected cell(s).

Direct cell-to-cell contact and the resulting fusion are a significant source of cellular infection, and may be a major mechanism of T cell destruction in HIV-1 infected individuals.

Infected and uninfected cells often fuse in large groups, thereby

forming multi-nucleated aggregates known as syncytia. The cell fusion causes the death of cells in the syncytia. See Lifson et al.

"Induction of CD4-Dependent Cell Fusion by the HTL-III/LAV

Envelope Glycoprotein", Nature 323:725-27 (1986).

The majority of cell death is believed to take place in syncytia. Concentrations of free virus in the bloodstream of infected individuals are typically very low, suggesting that significant infection does not occur as a result of free virus in the bloodstream. Because of the low relative numbers of infected T cells, it also seems unlikely that significant cell infection can occur from discrete fusion of individual infected and uninfected cells. In one study it was found that the proportion of infected T cells in infected individuals is usually only one out of every 10,000 to

100,000 white blood cells. Nevertheless it was reported that the number of CD4 positive cells (i.e., T cells) gradually decreased.

The HIV-1 titers of neutralizing antibodies in the serum of infected individuals is usually so low as to be insufficient to neutralize the HIV-1 infection. Thus, monoclonal antibodies which neutralize HIV-1 would be particularly useful for treatment.

T Monoclonal antibodies are produced by hybridoma cells.

Hybridomas are cells which have all been cloned from a single fused cell. All the clones are identical to the parent. Accordingly, all the hybridomas of the same clone produce antibodies of the same idiotype which bind to the same epitope of the antigen.

A method of making monoclonal antibodies was first described by Kohler and Milstein. See Kohler and Milstein,

Nature 256:495-97 (1975); Kohler et al.. Eur. J. Immunol.. 6:511-19

(1976). A host animal, usually a mouse, is immunized with an antigen and then sacrificed. Lymphocytes containing B-cells are then removed, usually from the spleen or other lymphoid tissues.

The removed lymphocytes are fused with myeloma cells to form hybridomas. The hybridomas which produce antibody against the designated epitopes of the immunizing antigen are cloned and screened. These hybridomas are then used to manufacture the desired monoclonal antibodies.

In summary, a monoclonal antibody that inhibits infection of susceptible cells by many strains of HIV-1, either by preventing attachment of free virions or by inhibiting direct cell-to-cell transmission of virus through syncytium formation, has great potential therapeutic value. Such an antibody could be useful in treating patients with AIDS or ARC, or could be used to prevent AIDS in asymptomatic healthy HIV-1 infected individuals, or in individuals in high-risk groups for AIDS exposure and infection.

Such an antibody could target the CD4 binding site of the virus or another neutralization site on the virus. As an alternative to administering neutralizing monoclonal antibodies, however, one could use a vaccine derived from such monoclonal antibodies.

2. Anti-idiotypes

It is fundamental that the introduction of a foreign antigen to an animal causes an immune response, including the production of antibodies (Abl) specific for that antigen. It is also well recognized that anti-idiotype antibodies (Ab2) to the paratope of Abl, produced by immunizing with Abl, bear the internal image of the antigenic determinant which Abl recognizes. As a result, immunization with Ab2 can elicit in vivo the production of another set of antibodies (Ab3) which, like Abl, bind to the antigen which was originally introduced. If the antigen originally introduced is a pathogen such as HIV-1, immunization with Ab2 will actively protect against that pathogen.

The Ab2 are, therefore, useful as vaccines, because they induce production of endogenous Ab3. If the Abl originally administered were specific to the CD4 binding site of HIV-1, or were Abl which otherwise neutralized HIV-1, then the resulting

SUBSTITUTE SHEET Ab3 will neutralize HIV-1 (or prevent it from binding to the CD4 receptor). Vaccination (or "active immunization") with anti- idiotypes of these sorts appears to be a promising means to stop

the spread of AIDS. Summary of the Invention

The monoclonal antibodies (mAbs) of the invention bind to the CD4 binding region of HIV-1 or to a neutralizing epitope in the principal neutralizing determinant region on gpl20. They inhibit HIV-1 infection of T cells by free virions, and they also inhibit syncytium formation. The monoclonal antibodies of the invention are group specific and can neutralize and cross-protect

against different strains and different isolates of HIV-1.

The mAbs of the invention can be used for treatment of

AIDS and ARC and for passive immunization against HIV-1 infection. In these procedures, the antibodies can be used as

whole antibodies or as antibody fragments or they can be conjugated to cytotoxic or antiviral agents, or to microcarriers which contain such agents, in order to target the delivery of these agents to infected cells. The targeted delivery of therapeutic agents can also be achieved with bispecific antibodies derived from the anti-HIV-1 antibodies of this invention which have been

provided with a second specificity for the agent to be delivered to the target. Polyclonal or monoclonal anti-idiotype antibodies against the paratope of the antibodies of the invention can be used to stimulate a neutralizing immune response against HIV-1.

The mAbs and anti-idiotypes of this invention can be used in vivo as antibodies derived wholly from mice or other animals.

Alternatively, especially for therapeutic use, the mAbs and anti- idiotypes can be made in the form of whole human antibodies, animal/human chimeric antibodies, single chain antibodies, or antobody fragments. For the chimeric antibodies, the constant region is human-derived, and the variable region (or only the antigen binding region) is animal-derived.

The mAbs and anti-idiotypes of this invention are produced by continuous, stable antibody-producing cell lines. These cell lines can be produced by hybridoma techniques and by genetic engineering techniques.

This invention also pertains to peptides which correspond to epitopic segments of gpl20 recognized by the antibodies of this invention. The peptides can be used in vaccine compositions for generating a cross-protective, neutralizing immune response against HIV-1. They can also be used to detect neutralizing antibodies against HIV-1 in a biological fluid. Brief Description of the Drawings

Figure 1 is a plot showing the relative effectiveness of four of the mAbs of the invention (BAT085, BAT123, BAT267,

BAT509) in neutralizing HIV-1 infection of H9 cells, as compared with another anti-gpl20 mAb (BAT496) and an irrelevant murine mAb to human chorionic gonadotropin (α-HcG). The percentage of infected cells was determined nine days after infection.

Figure 2 is a plot showing the relative effectiveness of the four of the mAbs of the invention in neutralizing HIV-1 infection of H9 cells, as compared with the irrelevant mAb (α-HcG). The percentage of infected cells was determined thirteen days after infection.

Figure 3 is a plot of purification of the immunoconjugate.

Immunoconjugate was eluted with a NaCl gradient ( — ) and absorbance at 280 nm was recorded ( ). The immunoconjugate elutes as a single peak at 110 mM NaCl.

Figure 4 is a flow cytometric analysis plot showing the relative binding activities of the immunoconjugates BAT123-PAP-S and G3.519-PAP-S to HTLV-III_B infected H9 cells. Infected H9 cells were treated without immunoconjugates (a), with

BAT123-PAP-S at 50 ng/ml (b), at 1 μg/ml (c), and G3.519-PAP-S

at 5 Mg/ml (d). Figure 5 is a plot showing the cytotoxic effects of

G3.519-PAP-S when presented to H9 cells uninfected (filled circle) or infected with HTLV-III_B (open circle), HTLV-III_MN (open triangle), or HTLV-III_RP (open square). The immunoconjugate killed H9 cells infected with HTLV-III_j^, more effectively than it killed H9 cells infected with HTLV-m_B and HTLV-IIV

Figure 6 is a plot showing the cytotoxic effects of

BAT123-PAP-S when presented to H9 cells uninfected (filled circle) or infected with HTLV-III_B (open circle), HTLV-III_MN (open triangle), or HTLV-ITV (open square). BAT123-PAP-S was most effective at killing H9 cells infected with HTLV-III_B, moderately effective at killing H9 cells infected with HTLV-III_MN, and relatively ineffective at killing H9 cells infected with HTLV-IIV Figure 7 is a chart showing the specificity of the cytotoxic effects of immunoconjugates. The addition of specific monoclonal antibody to the target cells prior to incubation with the immunoconjugates efficiently and specifically blocks the cytotoxicity of the immunoconjugates in a dose-dependent fashion. BAT123-PAP-S is the immunoconjugate in Panel A, and

G3.519-PAP-S is the immunoconjugate in Panel B, both represented by the open columns. An irrelevant monoclonal antibody, which does not inhibit the cytotoxicity of the immunoconjugate, is represented by the hatched column.

Figure 8A shows the results of competition assays of the binding between BAT123 and AB19-4 by synthetic peptides corresponding to the BAT123 binding regions in HTLV-III_B,

HTLV-III_MN and HTLV-III_RJ. (T64-63-6 is an irrelevant peptide used as control).

Figure 8B shows the results of competition assays of the binding between BAT123 and AB 19-31 by synthetic peptides corresponding to the BAT123 binding regions in HTLV-III_B,

HTLV-III_MN and HTLV-III_RF (T64-63-6 is an irrelevant peptide used as control).

Figure 9 shows specific gpl20 binding of Ab3s generated in rabbits immunized respectively with AB19-4 and AB19-31. Figure 10 shows the reactivity of AB19-4-HRP conjugate with solid-phase antibodies in ELISA. Wells of Immunlon 2 plates were coated with 100 μl of BAT123 (filled circle), CAGl-51-4

(filled triangle), G3.519 (open circle), and purified polyclonal human IgG (open triangle) (10 μg/ml). AB19-4-HRP of varying concentrations was tested for binding. Results are expressed as mean ± SD, n = 3.

Figure 11 shows the inhibition of the binding between

SUBSTITUTE SHEET AB19-4-HRP conjugate and solid-phase BAT123 by gpl20.

Microtest plates were coated with 100 μl of BAT123 (10 μg/ml) for ELISA as described in Fung et al., J. Immunol. 145:2199-2206

(1990). The inhibition by varying concentrations of purified HTLV-πi_B gpl20 (filled triangle), unconjugated AB19-4 (filled circle), and G3.519 (open triangle) is expressed as the percent decrease of OD of the test well compared to that of the control without inhibitors. Results are expressed as mean ± SD, n = 3.

Figure 12 shows inhibition of the binding between AB19-4- HRP conjugate and solid-phase BAT123 by the synthetic epitope peptides. The amino acid sequences of R15K (open circle), R15N (filled circle), and S15Q (open triangle), defining the corresponding peptidic segments in the gpl20 of HTLV-III_B, HTLV-HI_MN, and HTLV-IH_RP respectively were shown in Table I, Fung et al., J. Immunol. 145:2199-2206 (1990). Peptide T19V defining a distinct segment in the C2 region (amino acid residue #254-275) of HTLV-III_B gpl20 was used as control (filled triangle). The ELISA procedure is described in Fung et al., 1 Immunol. 145:2199-2206 (1990). Results are expressed as mean ± SD, n = 3.

Figures 13A and 13B, respectively, show reactivity of goat

and sheep antisera to HTLV-III_B gpl20 with solid-phase AB19-4 in ELISA. Immune (filled symbols) and non-immune (open symbols) sera from goat (13A) and sheep (13B) were examined for binding to AB19-4 (filled circle, open circle) and protein A

purified normal mouse IgG (filled triangle, open triangle), as described in Fung et al.. J. Immunol. 145:2199-2206 (1990). Each value is the mean of duplicate determinations. The vertical bars indicate the range of OD values.

Figure 14 shows reactivity of gpl20-affinity purified Ab3

(generated from AB19-4) with solid-phase Ag in ELISA. Microtest plates were coated with HTLV-III_B gpl20 (filled circle) and KLH (open circle) at 0.5 μg/ml, and peptides R15K (filled triangle), R15N (filled square), S15Q (open triangle) and T19V (a peptide not homologous to the PND, represented by the open squares) at 2 μg/ml. Gpl20-affinity purified Ab3 of varying concentrations were tested for binding to the Ag as detected by

HRP-conjugated goat anti-rabbit IgG and HRP-conjugated goat anti-mouse IgG, respectively. Results are expressed as mean ±

SD, n = 3.

Figures 15A and 15B show flow cytometric analysis of the binding of gpl20-affinity purified Ab3 (Fig. 16A) and of BAT123

(Fig. 16B) to HIV-1-infected H9 cells. Uninfected H9 cells, — ;

H9 cells infected with HTLV-III_B, ; or HTLV-III_MN, ^•-. Figure 16 shows neutralization of HIV-1 by gpl20-affinity purified Ab3. The inhibition of infection of CEM-SS cells by

HTLV-III_B (filled circle, open circle) and HTLV-III_MN (filled triangle, open triangle) with gpl20-affiιιity purified Ab3 (filled symbols) and with sham purified pre-immune rabbit serum substances (open symbols) was determined by a syncytium-forming microassay as described in Fung et al., J. Immunol. 145:2199-2206

(1990). Vn is the mean number of syncytia in triplicate test wells, and Vo the mean number of syncytia in the triplicate control wells. The vertical bars represent SD.

Figure 17 shows neutralization of HIV-1 by BAT123. The inhibition of infection of CEM-SS cells by HTLV-HI_B (filled circle, open circle) and HTLV-III_MN (filled triangle, open triangle) with

BAT123 (filled symbols) and with negative control BAT496 (open symbols) was assayed as described in Fig. 16.

Figure 18 shows the binding of AB20-4 to solid-phase

G3.519 (filled circles), and the binding of an irrelevant murine

IgGl to solid-phase G3.519 (open circles), as measured using a sandwich ELISA where G3.519-HRP is the capture/detecting antibody.

Figure 19 shows the inhibition of binding of G3.519-HRP solid-phase HIV-1 gpl20 by AB20-4 (filled circles), G3.519 (open circles) and an irrelevant murine IgGl (open triangles).

Figure 20 shows the inhibition of binding of AB20-4-HRP to solid-phase G3.519 by peptide T35S (having the sequence of the

gpl20 region to which G3.519 binds, denoted by the filled circles) and an irrelevant peptide G19C (open circles).

Detailed Description of the Invention

A. Summary of Procedures Used

The monoclonal antibodies of the invention bind to the viral envelope glycoprotein gpl20. In the processing of HIV-1 specific envelope protein in infected T cells, gp41 is a transmembrane protein and is largely not exposed. In contrast, gpl20 is an external envelope protein which is extracellular. Thus, in infected T cells the gpl20 protein offers binding epitopes for the monoclonal antibodies of the invention. More specifically, the mAbs of the invention include mAbs

which bind to the CD4 binding region of HIV-1 (such as G3.519), or to a neutralizing epitope on the principal neutralizing determinant of gpl20 (such as BAT123 and BAT267). The monoclonal antibodies of the invention were found to be effective in inhibiting infectivity and in inhibiting syncytium formation. This indicates that they will likely be very effective for in vivo

neutralization, as the majority of cell death is believed to occur in syncytium. Importantly, the antibodies can neutralize different strains and different isolates of HIV-1 (i.e. the antibodies are group specific). The neutralizing antibodies also inhibit syncytium formation by various strains of HIV-1 which have a substantial degree of heterogeneity in the amino acid sequence of gpl20.

The neutralizing antibodies of this invention can have high potency in neutralizing infectivity. For example, the mAbs against the principal neutralizing determinant ("PND") can inhibit, with an

IC_<- of less than 10 ng/ml, the infection of susceptible human T cells lines by HIV-1_B at 20 times TCID in a nine day assay.

The PND is the peptide segment on gpl20 from amino acid residue numbers 296 to 331, as determined from the gpl20 sequence of the HTLV-III_B, or sequences of the corresponding regions from other HIV-1 strains. See Devash, Y., Proc. Nat'l Acad. Sci. USA 87:3445-3449 (1990). The PND peptide segment is in the relatively variable region, V3, of gpl20. However, recent studies indicate that there are conserved features in the PND segment. The amino acid sequences of PND segments in field HIV-1 isolates from patients are closely related. See LaRosa, G.J. et. al, Science 249:932-935 (1990). Antibodies which target the

PND are generally effective in neutralizing HIV-1 infection.

Abl which target the PND and neutralize HIV-1 include BAT123 and BAT267.

Suitable monoclonal Abl, including BAT123 and BAT267,

are produced by first immunizing an animal, preferably a mouse,

with a suitable antigen, which in this case is inactivated HIV-1. The antigen can be in whole form, e.g., whole HIV-1 virions, or cells infected with a virus and expressing the virus or its immunogenic domains can also be used. Specific viral proteins, such as the envelope glycoproteins, may be purified from the lysates of infected cells or viruses. The immunogenic domains of HIV-1 on gpl20, or synthetic or recombinant peptides which have the same or an immunologically equivalent sequence to these immunogenic domains, can also be used. These synthetic or recombinant peptides for use in immunization can be synthesized by conventional techniques, such as with the RaMPS system (DuPont

DeNemours & Co.), which applies Fmoc chemistry. Alternatively, recombinant peptides containing these peptides may be biosynthesized by expressing in is. coli or eukaryotic cells the gene segments containing the appropriate coding sequences. When using a synthetic peptide segment as an immunogen, it is usually more effective to conjugate it to a protein carrier, for example, HBsAg, hepatitis B virus core antigen, ovalbumin, bovine serum albumin, or preferably keyhole lympethemocyanin ("KLH").

If the peptidic segment lacks a lysine residue or if the lysine residue is in the middle part of the segment, it is desirable to add a lysine residue at the C-terminal end. Because the N-terminus already has an α-amino group, the modified synthetic peptide will have two available amino groups for linking.

Multiple molecules of peptides can be conjugated to each molecule of the carrier to make the immunogen. With KLH, a preferred molar ratio for peptide/KLH is 10. The conjugation can be done with well established methods using glutaraldehyde or bis

(sulfosuccinimidyl) suberate or preferably disulfosuccinimidyl tartrate as the cross-linkers.

One preferred immunization protocol for preparing the Abl monoclonal antibodies is to inject into each mouse 50 μg of the conjugate of KLH and the aforementioned recombinant or synthetic peptides in Freund's complete adjuvant. Two and four weeks later, the same amount of antigen is given subcutaneously in Freund's incomplete adjuvant. After about six weeks, the fourth antigen injection is given intraperitoneally in saline. Mice are sacrificed 4 days after the last injection and the spleens (or sometimes the lymph nodes) are removed for preparing single cell suspensions for fusion with myeloma cells. Lymphocytes from the spleens (or lymph nodes) which have been removed from the mice can be fused with myeloma cells to prepare hybridomas secreting the Abl monoclonal antibodies.

The fusion procedure with polyethylene glycol and other various procedures concerning the cloning and the culturing of hybridomas have been well established. One preferred protocol is the well- known one described by Hudson, L. and Hay, F.C. (Practical Immunology, 2nd edition, pp. 303-313, 1980, Blackwell Publishing Co., Boston), in which the lymphocytes are fused with non- secreting mouse myeloma cells, such as NS-1 or Sp2/0 cells, using polyethylene glycol. The fusion reagent used to make BAT123 was polyethylene glycol mixed with dimethyl sulfoxide (DMSO) in calcium magnesium-free phosphate buffered saline (PBS).

Reagents other than those discussed can be used for the chemical fusion. Another alternative is to use electrical fusion rather than chemical fusion to form hybridomas. This technique is well-established. Instead of fusion one can also transform a B-cell to make it immortal using, for example, an Epstein Barr Virus or a tranforming gene. (For a method of transforming a B-cell, See "Continuously Proliferating Human Cell Lines Syn¬ thesizing Antibody of Predetermined Specificity," Zurawski, V.R.

et al, in Monoclonal Antibodies, ed. by Kennett R.H. et al, Plenum Press, N.Y. 1980, pp 19-33.)

The screening of hybridomas for monoclonal antibodies reactive with the immunogen can be performed with an enzyme

linked immunosorbent assay (ELISA). A synthetic or recombinant peptide having the same sequence as a portion of the immunogen is used as the solid-phase antigen. A preferred solid phase antigen is the conjugate of such a synthetic or recombinant peptide with a carrier protein different from that used with the immunogen. An appropriate carrier protein can be bovine serum albumin or ovalbumin, provided they were not used as carriers in the immunization.

Clones of hybridomas which showed highest reactivities with the PND of gpl20 were selected for further screening by an immunofluorescence assay. The immunofluorescence assay was run to determine which of the ELISA positive monoclonal antibodies would bind specifically to intact, live infected T cells, but not to uninfected T cells. This was determined using immunofluorescence flow cytometric analysis of staining of HTLV- III_B-infected H9 cells. The clones which showed the highest reactivities with the CD4 region of gpl20 were screened using a p24 assay of HTLV-III_B-infected H9 cells.

After the Abl are isolated and characterized, they are used to immunize mice to create the Ab2. A similar protocol to that described for the Abl immunization can be used to create the

Ab2. A particularly preferred protocol for this immunization, used in producing AB19-4 and AB 19-31 fromBAT123, and in producing AB20-4 and the other anti-idiotypes in the AB20 series from

G3.519, is to first conjugate Abl to KLH using glutaraldehyde as described by Maloney et al, Hybridoma 4:191 (1985). Mice are then immunized intraperitoneally with 100 μg of the Abl-KLH conjugate at one month intervals for three months. Three days after the final immunization, the mice are killed, and the spleen cells are isolated and fused with Sp2/0 myeloma cells to create the

Ab2, using the fusion techniques set forth above. See Fung et al,

J. Immunol 145:2199-2206 (1990).

After isolation of monoclonal antibodies specific to gpl20 and to intact, live infected cells, the effectiveness of the antibodies in neutralization of HIV-1 was tested. Monoclonal antibodies from each clone which was immunofluorescence positive were isolated. A comparison was made of the number of cells infected by HIV-1 in the presence or absence of the monoclonal antibodies. Different titers of each antibody were used in order to compare their potency. For the mAbs against the PND, the neutralization was assayed with H9 cells and monitored by an immunofluorescence technique. For the mAbs against the CD4 binding region, the neutralization was determined with an infectivity assay using H9 cells or CEM-SS cells.

The second functional test of neutralization is by syncytium inhibition. In the syncytium inhibition assay infected T cells were added to a well seeded with HeLa cells which had been artifically transfected with CD4 genes and expressed the CD4 antigen on their surface. The CD4 antigen on the cell surface fuses with infected T cells to form multi-nucleated giant cells. It was determined which concentrations of mAbs to the PND, and which mAbs to the anti-CD4 binding region (International Application No. PCT/US90/02261), would inhibit giant cell formation. The antibodies are tested in these assays to determine their ability to neutralize different viral strains and isolates. In conclusion, a variety of ways of preparing and testing the products of the invention are possible and are within the scope of the invention. The advantages and uses for the products of the invention will now be discussed. B. Advantages and Uses As noted above, one advantage of monoclonal antibodies is their specificity. This specificity is highly pertinent to use of the monoclonal antibodies of the invention in therapy, because it means that lower dosages can be used.

The prophylactic and therapeutic uses for the monoclonal antibodies of the invention include both jn vivo immunotherapy

and extracorporeal immunotherapy. Direct in vivo treatment with the monoclonal antibodies of the invention involves administering them internally, preferably via intravenous injection. They can be administered subcutaneously or intramuscularly.

If treatment of infected cells in the brain is needed, it may be possible to couple the monoclonal antibody to an agent, such as certain lipophilic substances, which allows it to pass through the blood-brain barrier. The antibodies of this invention can neutralize different strains and isolates of HIV-1 in the patient population.

In extracorporeal therapy, blood leukocyctes are removed from the patient and treated with neutralizing antibody. The monoclonal antibody is then added to the leukocytes. The leukocytes can also be stimulated with immunopotentiating drugs, for example interleukin-2. The leukocytes are then returned to the patient. The mouse-derived monoclonal antibodies of the invention can be used for both direct in vivo and extracorporeal

immunotherapy. However, when mouse-derived monoclonal antibodies are used in humans as therapeutic agents, the patient may produce endogenous antibodies against the Fc region. In fact, production of such antibodies against the Fc region could reduce the immunogenic effect of the administered antibodies before they can function to neutralize HIV-1.

Thus, the preferred antibodies of the invention have human constant regions. These preferred antibodies include whole human antibodies, chimeric antibodies wherein the variable region is of murine origin and the constant region is of human origin, and antibodies wherein only the complementarity determining regions

(CDR) are of murine origin and the remainder of the variable regions, and the entire constant regions, are of human origin. Alternatively, antibody fragements, such as V_H, V_L, F-_* Fd, Fab and F(ab')₂, none of which have complete constant regions, can be used.

Chimeric antibodies can be produced by transfecting non- producing mouse myeloma cells with the hybrid genomic DNA, or cDNA. See V.T. Oi et al., Bio Techniques 4(4^:214-221 (1986); L.K. Sun et al.,"Chimeric Antibodies with 17-1A-Derived Variable and Human Constant Regions". Hybridoma 5 (1986). The hybrid genomic DNA or cDNA will contain the human constant regions and the mouse variable region. If one is making an antibody in which only the CDRs are of mouse origin, the hybrid genomic

DNA or cDNA will contain human constant regions, mouse CDR regions, and the remainder of the variable regions will be human.

Human antibodies can be produced by using human expression libraries (e.g., Stratagene Corp., La Jolla, California) to produce fragments of human antibodies (V_H, V_L, F_v, Fd, Fab, or

F(ab')₂). One can use the fragments to construct whole human antibodies using techniques similar to those for producing chimeric antibodies. One can also create single peptide chain antibodies. In such antibodies, the heavy and light chain F_v regions are connected. See Huston, J.S. et al., Proc. Natl. Acad. Sci. USA

85:5879-5883 (1988).

Another alternative form of monoclonal antibody suitable for use in therapy are derivative antibodies which draw cytotoxic cells such as macrophages or cytotoxic T cells toward the targeted

HIV-1-expressing T cells. These derivative antibodies include bi- specific antibodies having a specificity for a receptor of a cytotoxic cell and a specificity for the infected cells. Such hybrid bi-specific antibodies can include two different Fab moieties, one Fab moiety having antigen specificity for the targeted infected cells, and the other Fab moiety having antigen specificity for a surface antigen of a cytotoxic cell, such as CD3 or CD8. The bi-specific antibodies of the invention can be a single antibody having two specificities, or a heteroaggregate of two or more antibodies or antibody fragments. See. e.g._τ C. Reading, U.S. Patent Nos. 4,474,893 and 4,714,681; Segal et al, U.S. Patent No. 4,676,980.

Certain factors, such as granulocyte monocyte-colony stimulation factor (GM-CSF) or monocyte-colony stimulation factor (M-CSF), are known to induce the proliferation of leukocytes, including those mediating antibody dependent cellular cytotoxicity ("ADCC"). In in vitro experiments, GM-CSF and M-

CSF have been shown to augment the ADCC activity on tumor cells mediated by monoclonal antibodies specific for surface antigens expressed on the tumor cells. The therapeutic effect of specific monoclonal antibodies of the invention, conjugates, or polyclonal antibodies in suppressing the immune response could perhaps be enhanced by combining them with factors that augment ADCC activities.

Immunotherapy for patients with AIDS or ARC is appropriate with the mAbs and related products of the invention. A variant of immunotherapy is protection through passive immunization. The antibodies of this invention are particularly

suitable for passive immunization because they can cross-protect against HIV-1 of different strains and isolates.

In the passive immunization procedure, patients who are asymptomatic (not yet showing symptoms of AIDS or ARC), or who are seronegative but in a high risk group, are treated to inhibit infection. The targets include fetuses born in or babies born to HIV-1-carrier mothers and health professionals working with AIDS patients, or blood products. The agent for treatment, again, can be the mAbs of the invention, or the human or humanized antibodies, or the fragments, discussed above. Much attention in the effort to stop AIDS has focused on the search for a vaccine. In one type of proposed vaccine the immunizing agent is a portion of HIV-1 which itself is non-infective but which nonetheless induces antibody production.

Monoclonal antibodies which neutralize HIV-1 can help in the search for such a vaccine. They can be used to help locate, identify, and study the "neutralizing" epitopes on HIV-1 which bind the monoclonal antibodies. These epitopes are likely to be the non-infective but nonetheless immunogenic portion of the molecule. Study of these epitopes allows synthesis of a non-pathogenic immunogen with a structure which is the same or

immunologically equivalent to the epitope. For example, the immunogen can be a peptide which comprises an amino acid sequence that is the same or similar to the epitope bound by an anti-HIV-1 antibody which neutralizes HIV-1.

It has now been discovered that two of the anti-PND mAbs of this invention recognize epitopes located in a region of gpl20 having the following amino acid sequence:

RPNNNTRKRIRIQRGPGRAFVTIGK

This segment represents a 25 amino acid residue long segment of gpl20 (residue # 294 to residue # 318). One antibody (BAT267) reacts with a peptide having the sequence RPNNNTRKRIRIQRG (peptide a) and the other antibody (BAT123) reacts with a peptide having the sequence RIQRGPGRAFVTIGK (peptide b). Other strains of HIV have regions corresponding to this segment.

These two 15 amino acid residue long peptides represent two adjacent, overlapping segments of gpl20 of HIV-1B: peptide "a" represents the segment of residue #294 to residue #308 and peptide "b" of residue #304 to #318. BAT267 reacts with peptide "a" and not peptide "b", which shares five amino acids RIQRG, or another 15 amino acid long peptide, which represents a segment of gpl20 (residues #284 to #298) adjacent to peptide "a" and shares five amino acids RPNNN. These results suggest that

BAT267 recogmzes an epitope either borne entirely by all or a

part of the middle five amino acid residues TRKRI or formed by all or a part of these five amino acids with some of the flanking amino acid residues. Based on similar results, BAT123 seems to react with an epitope either borne entirely by all or a part of

PGRAF or formed by the combination of all of a part of PGRAF with some of the flanking amino acid residues.

The CD4 receptor binding region on gpl20 includes a s e g m e n t h aving t h e ami n o a c i d s e qu e n c e

IINMWQKVGKAMYAP. This sequence corresponds to residues

423 to 437 of gpl20, and is bound by the anti-CD4 binding region antibody G3.519.

The peptidic immunogens of this invention can comprise the above-identified amino acid sequences or immunochemical and immunological equivalents thereof. These equivalents include, for example, any of the actual epitope portions of any of these sequences, corresponding peptidic regions from various HIV-1 strains and peptides generated by various changes such as insertions, deletions and substitutions of amino acids.

The peptides of this invention can be coupled together to form larger, multivalent oligopeptides. The peptides may be prepared by chemical synthesis. Alternatively, they may be prepared by recombinant DNA technology where DNA sequences

encoding the peptides are synthesized or isolated from HIV-1 DNA and expressed in an appropriate expression system.

The peptides may be used in immunoassays to identify neutralizing antibody or to screen for the presence of neutralizing antibody in serum. The peptides may also be used individually or in combination to elicit a immune response against HIV-1. For this purpose, the peptides may be formulated in vaccine compositions, generally for administration at concentrations in the range of 1 μg to 20 mg/kg of host. Physiologically acceptable vehicles such as water, saline, or phosphate buffered saline can be used in the formulations. Adjuvants, such as aluminum hydroxide gel, can also be employed. The route of administration can be intramuscular, intraperitoneal, subcutaneous, or intravenous. The compositions can be given one time or mutiple times, usually at one to four week intervals.

In preferred embodiments of the vaccine composition, the peptides are coupled to a carrier protein such as a foreign keyhole limpet hemocyanin. This can enhance the immunogenicity of the haptenic peptides. Another type of vaccine employs anti-idiotype antibodies

(Ab2), and induces a HIV-neutralizing antibody (Ab3) in an organism. These Ab2 bear the same conformation on their paratope as the antigen (HIV-1) which initially stimulated antibody production. See J.L. Marx, "Making Antibodies Without

Antigens". Science 288:162-65 (1986).

Thus, parotope-specific anti-idiotypic antibodies with partially the same structure as the PND on HIV-1 can be made by immunizing an animal with the monoclonal antibody to the PND of HIV-1. Similarly, parotope-specific anti-idiotypic antibodies with partially the same structure as the CD4 binding region on

HIV-1 can be made by immunizing an animal with the monoclonal antibody to the CD4 binding region of HIV-1.

Advantageously, because these anti-idiotype antibodies consist of protein and do not carry any viral nucleic acid, they would be of much less concern for pathogenicity than the killed or inactivated virus. As noted above, the same sorts of derivative antibodies and fragments which are less immunogenic than murine mAbs, e., human, chimeric mouse/human, single chain, and the human antibody fragments V_H, V_L, F^ Fd, Fab and F(ab')₂, are preferable for the anti-idiotype antibodies of the invention.

Alternatively, single chain polypeptides containing the antigen combining region (paratope) of the anti-idiotypic antibody can be used. Such polypeptides can be produced by genetic

engineering techniques. See e.g., Skerra, A. and Pluckthun, A.

SUBSTITUTE SHEET Science 240:1038-1041 (1988); Better, M. et al. Science

240:1041-1043 (1988).

The anti-idiotypes of the invention can be used for active immunization, and are preferably administered together with appropriate adjuvants, such as threonyl muramyl dipeptide. In addition, the anti-idiotypes can also be used as boosters, to enhance the immunogenicity of another type of vaccine. The other vaccine could be a protein subunit vaccine, such as the peptidic immunogens of the invention described above, or killed or inactivated HIV-1. The anti-idiotypes of the invention would enhance the anti-HIV-1 immune response, and thus enhance the immonogenicity of the other vaccine. The other vaccine would be admimstered simultaneously or shortly after administering the anti- idiotype. In addition to the above-noted uses for the anti-idiotypes of the invention, the mAbs of the invention can also be used to aid in the delivery of cytotoxic or antiviral agents, by incorporating them into, for example, microcarriers or liposomes. Exemplary cytotoxic agents include pokeweed antiviral protein from seeds (PAP-S), cytotoxic steriods, gelonin, abrin, ricin and phospholipases. Examples of antiviral agents are interferon, azidothymidine and ribovirin. Once again, it should be noted that human, chimeric mouse/human, single chain, and human antibody fragments, as well as bispecific monoclonal antibodies, are also suited to aid in drug delivery.

Conjugates of antibodies and toxins have been actively explored as therapeutic agents for destroying tumor cells in cancer patients (L.E. Spitler, Immunotoxin. ed. A.E. Frankel, Kluwer

Academic Publishers, Boston, pp 493-514 (1988)). These immunoconjugates are known to exhibit relatively long half-lives in vivo and their side effects in most patients are clinically tolerable (L.E. Spitler, Immunotoxin. ed. A.E. Frankel, Kluwer

Academic Publishers, Boston, pp 493-515 (1988)). Therefore, a rational approach for treating AIDS patients and HIV-infected individuals is to target the HIV-infected and producing cells by using conjugates of toxins (e.g.. PAP-S) and monoclonal antibodies of the invention.

The details for the procedure by which the monoclonal antibodies and immunoconjugates of the invention are made will now be described.

Example I: Preparation of the Hybridomas and Monoclonal Antibodies a) Preparation of Virus for mAbs Against the PND

In order to maintain a supply of inactivated HIV-1, a virus stock was prepared as follows. The H9 clones of the human T cell line (which is described by M. Robert-Guroff et al. in Nature 316:72-74, supra) were maintained in culture. These H9 cells were infected with HIV-1 (HTLV πi_B), which was a gift from Dr. R.

Ting, Biotech Research Laboratory, Rockville, Maryland. Maintaining the infected H9 cells in culture permits the cells to reproduce and to continuously synthesize a supply of

HIV-1. The H9 cells were cultured in a growth medium of 20%

FBS (heat-inactivated) RPMI 1640, supplemented with 5mM

L-glutamine, 5mM HEPES, 50 units/ml penicillin and 50 mg/ml streptomycin.

Purified HIV-1 was obtained by first centrifuging the cell culture at 1000 g for ten minutes to remove the cells and debris. The supernatant was then centrifuged at 90,000 g for one hour. The virus pellet was resuspended in minimal volume of phosphate buffered saline pH 7.4 and loaded into a centrifuge tube with a preformed sucrose gradient (20%-60%). The sample was then centrifuged at 100,000 g for sixteen hours. The virus was collected at the gradient of 38%. The virus was then aliquoted and frozen at -80°C after the protein content was measured. b^ Immunization Procedure for mAbs Against the PND

Male Balb/c mice were used for the immunization. Each mouse received 100 micrograms of inactivated HIV-1. The inactivation of the virus was performed according to FDA approved protocol, by UV irradiation and addition of a detergent,

Nonidet P-40 (0.1%). A volume of suspension containing 100 micrograms of virus per mouse was suspended in 200 microliters phosphate buffered saline (PBS), and emulsified with equal volumes of Freund's complete adjuvant.

Each mouse was immunized subcutaneously with 100 micrograms of the emulsified virus. The mice were injected at sites with high concentrations of lymph nodes, for example, the underside of the intersection of the limbs and the trunk. One month later the mice received subcutaneous booster injections at the same sites with the same quantity of virus. The boosters were prepared essentially in the same manner as was the first injection, except that for the boosters the emulsification was done in Freund's incomplete adjuvant.

One month later, each mouse was reimmunized subcutaneously with 100 micrograms of virus suspended in PBS.

Each mouse was injected subcutaneously at the intersection of each limb with the trunk, and intraperitoneally. Three days after the last injection, the mice were sacrificed and their spleens were removed. The spleen cells were then fused with myeloma cells by the following procedure. c) Fusion

Suspensions containing a five-to-one ratio of spleen cells to myeloma cells were prepared. The myeloma cells chosen were NS-1. The NS-1 cells were conditioned to have a doubling time about every seventeen hours. They were used for fusion when in the log phase. The NS-1 cells were subcultured in bacteriological plates (100 mm) at a concentration of 6 x 10⁴ cells/ml in 10 ml of Dulbecco's Modified Eagle's Medium (DMEM) containing 5% Fetal Bovine Serum (FBS), 100 units/ml of penicillin and 100 micrograms/ml of streptomycin. The medium was changed every three days. Alternatively, the cells were subcultured at 1.54 x 10⁵ cells/ml in 10 ml of the same medium, and the medium was changed every two days.

The spleen cells were prepared by placing the spleen on a bacteriological plate (100 mm) and injecting 20 ml of calcium magnesium free PBS (CMF-PBS) into both ends of the spleen to flush out the spleen cells. The flushed spleen cells were then transferred to a 50 ml centrifuge tube.

The spleen cells were centrifuged at 400 g for five minutes, and then suspended in 5 ml of 0.83% NH₄C1 (0.155 M) for ten minutes at room temperature to lyse the erythrocytes. 5 ml of

CMF-PBS was added to the tube to stop the lysis. The cells were then pelleted, and resuspended in 10 ml of CMF-PBS.

The concentration of lymphocytes was determined by adding 40 microliters of cell suspension to 10 ml of saline together

TM with 3 drops of Zapoglobin . The number of lymphocytes was counted with a hemacytometer and from this value the concentration of cells was determined. The concentration was then multiplied by the dilution factor of 250 to yield the actual concentration of cells in the suspension.

The Sp2/0 cells were transferred from five of the bacteriological plates (100 mm) to a 50 ml centrifuge tube. The cell concentration was determined using the counting technique described above. 5 x 10⁷ of the Sp2/0 cells were then suspended in 10 ml of CMF-PBS and mixed with 2.5 X 10⁸ spleen cells in a 50 ml centrifuge tube. The cells were spun down and washed once with 10 ml of

CMF-PBS. The supernatant was aspirated as much as possible with a glass Pasteur pipette. The tube was gently tapped to free the cell pellet.

Prior to preparing the cells, a fusion mixture had been prepared as follows. 5 g of polyethylene glycol 1450 (purchased from Kodak) had been mixed with 5 ml of CMF-PBS and 0.5 ml

of DMSO. This mixture had then been warmed to 56°C to melt

SUBSTITUTE SHEET it, titrated to a final pH of 7.0, and filtered through a 0.22 micron

Millipore filter in order to sterilize it. 1.0 ml aliquots had been added to Cryotubes, and these had been stored at -70°C.

To prepare the fusion mixture for use, one of the aliquots in the Cryotubes was melted by heating it to 37°C. Separately, a tube containing 1.0 ml of DMEM (without serum) was heated to

37°C.

The 1.0 ml aliquot of polyethylene glycol fusion mixture was added to the cell suspension and the suspension was mixed well. Forty-five seconds after the polyethylene glcyol fusion mixture had been added, 2.0 ml of the pre-heated DMEM (without serum) was added dropwise with mixing. The remaining 8 ml of the pre-heated DMEM (without serum) was then added. The cells were left at room temperature for 10 minutes. 2.0 ml of FBS was added to the suspension and the suspensions were mixed well. The combination of the FBS and the DMEM can help prevent adherence of cells to the test tube walls. The suspensions were then centrifuged at 400 g for four minutes. After having been spun down, the cells were suspended in

116 ml of a modified medium, supplemented with 5% FBS, 100

units/ml of pencillin, 100 micrograms/ml of streptomycin, and Littlefield's hypoxanthine, aminopterin and thymidine (HAT).

The concentration of the cell suspension was adjusted to

3.3 X 10⁵ of the spleen cells per 200 microliters of suspension. 200 microliter aliquots of suspension were then distributed to each well of a 96 well microtiter plate. After seventeen such plates were prepared, the plates were transferred to an incubator and maintained at 37°C in 5% C0₂.

The cells were grown for seven days in the plates, then the growth medium was withdrawn and new medium was added. Four days after that, the medium was again changed. d) Preparation of mAbs Against the CD4-Binding Domain

Rather than using the inactivated virus in the immunization, the envelope glycoprotein, gpl20, was used in immunizing to make the mAbs against the CD4-binding domain. The gpl20 was prepared from H9/HTLV-IIIB cell extracts. The immunization, virus preparation, fusion and screening for preparation of these mABs is described in published International Application PCT/US90/02261. e) ELISA Procedure for Preparing the mAbs Against the PND

Four days after the medium was changed, an enzyme linked

immunosorbent assay (ELISA) was performed on the antibodies in the 96 well microtiter plates to determine which would bind the gpl20 protein of HIV-1. The ELISA was carried out as follows.

Purified gpl20 protein was prepared as described in W.G. Robey, "Prospect for Prevention of Human Immunodeficienty Virus Infection: Purified 120-kD Envelope Glycoprotein Induces Neutralizing Antibody", Proc. Natl. Acad. Sci. USA J&7023-27

(1986). 50 microliters of a gpl20 suspension (at a concentration of 0.1 to 1.0 micrograms/ml) was added to wells of 96-well Immulon I plates with a twelve-channel pipette. The plates were covered and incubated for eighteen hours at 4°C, in order to allow the protein to bind to the plate.

The liquid contents of the plates were then emptied, and

200 microliters of 0.1 M NH₄C1 was added to each well in order to saturate any remaining binding sites on the plates. The NH₄C1 solution was left in the wells for thirty minutes at room temperature.

The NH₄C1 solution was then removed and the wells were washed three times with PBS and 0.05% Tween 20. Some of the PBS/0.05% Tween 20 solution was left in the wells until the antibody suspension described below was added. 50 microliters of the cell fusion supernatant from each well of the seventeen 96 well plates was added to each of the wells on

the Immulon I plates, and incubated for one hour. Following incubation, the plates were rinsed three times with PBS/0.05%

Tween 20 in order to remove any unbound antibody.

The cell fusion supernatant will contain the antibody which

is produced by the various hybridomas in the 96 well plates. The antibody which is specific to gpl20 will bind thereto. Inasmuch as the gpl20 is bound to the Immunlon I plate, the antibody specific to gpl20 will also become bound to the plate.

The next stage is to add the marker which will indicate the amount of bound antibody in each well. The marker chosen was horseradish peroxidase. This marker was conjugated with goat anti-mouse IgG to yield peroxidase-conjugated goat anti-mouse

IgG. The goat anti-mouse IgG will bind to any mouse monoclonal antibody which is bound to the palte. The peroxidase marker can then be activated to indicate the quantity of bound antibody by an exzyme reaction.

For detection of the bound antibodies, 100 microliters of the peroxidase-conjugated goat anti-mouse IgG diluted at 1:1000 in PBS/0.05% Tween 20 and 1% BSA was added to each well.

The plates were incubated for one hour at room temperature. Thereafter, the plates were washed three times with PBS/0.05%

Tween 20 to remove any unbound goat anti-mouse IgG conjugate.

The next step is to activate the peroxidase marker which is conjugated to the goat anti-mouse IgG. This is done by adding

200 microliters of 3,3,5,5'-tetramethyl benzidine substrate solution to each well, and incubating at room temperature for 30 minutes.

The color reaction is stopped by adding 50 microliters of 2.0 M H₂S0₄.

The intensity of color was determined with an ELISA reader at 450 nm. The amount of antibody specific to gpl20 is proportional to the intensity of the color.

It was found that there were approximately 200 wells in the 96 well microtiter plates which produced antibodies which bound to gpl20 to at least some extent. Of these 200 wells the 39 which produced antibody showing the highest color intensity were selected for another screening step. f) Immunofluorescence Assay Using Live T-Cells for mAbs Against the PND

An immunofluorescence assay was performed to determine whether any of the antibodies which were reactive with gpl20 in the ELISA would bind specifically to live HIV-1 infected H9 cells.

The H9 cell line is permissive to persistent infection by HIV-1. This cell line was obtained from the American Type Culture

Collection in Rockville, Maryland. Antibody which binds to

infected cells, but not uninfected cells, is probably selective to a domain of the HIV-1 envelope protein on the extracellular side of the cell membrane. The immunofluorescence assay helps to select those gpl20 reactive antibodies which have a high potential to recognize the neutralization epitopes on the HIV-1 virion, and to inhibit syncytium formation by infected T-cells. Cultures of infected H9 cells were maintained as described above under the heading (a), "Preparation of Virus". The procedure by which the assay was performed is described below.

(i) Assay Procedure

50 microliter aliquots of infected H9 cell suspension at a concentration of 5 x 10⁶ cells/ml was added to each of thirty-nine

1.5 ml microfuge tubes. 50 microliter aliquots of the supernatant from the 39 wells containing the ELISA positive clones was then added to each tube. The antibodies in the supernatant which react with H9 cells will bind to any H9 cells which are in the tube. The tubes were then incubated for thirty minutes at room temperature. After incubation, the tubes were spun, the supernatant was withdrawn, and the cells were washed three times with a mixture of RPMI 1640, contaimng 2% fetal calf serum and

0.1% sodium azide. The tubes were then tapped to loosen the cell pellet.

10 microliters of a labeled antibody, goat anti-mouse IgG conjugated with fluorescein isothiocyanate (FITC), was added to each test tube at a dilution of 1 to 200. This labeled antibody will bind to any monoclonal antibodies which have attached to HIV-1 infected H9 cells and provide a means for identifying these monoclonal antibodies. The tubes were again incubated for thirty minutes at room temperature. The tubes were centrifuged, and the cells were washed with the same medium as before.

The cells were then resuspended in PBS, placed onto individual slides and cover-slipped. The cells were viewed with a fluorescence microscope.

To determine which of the thirty-nine selected wells contained antibodies which bound to uninfected H9 cells, an essentially identical procedure as described above was performed, using infected H9 cells instead. (ii) Results

Seven of the thirty-nine wells tested contained clones which produced monoclonal antibodies binding to live infected H9 cells but not to uninfected H9 cells. That is, when using antibodies from these seven wells the infected cells fluoresced, but the uninfected cells did not.

Cells and antibodies from the seven wells which contained

immunofluorescence positive clones were collected. g) ELISA and Immunofluorescence Staining in Preparing mAbs Against the CD4-Binding Region

Essentially the same ELISA and immunofluorescence

staining procedure was used in preparing the mAbs against the CD4-binding region. The ELISA was used to determine binding to g l20 and the immunofluorescence staining was used to determine binding to HIV-1 infected H9 cells. These procedures are more fully described in International Application No.

PCT/US90/02261. h) Single Cell Cloning for mAbs Against the PND

Cell suspensions from each of the thirty-nine ELISA positive wells were expanded in the wells of a twenty-four well plate. After five days of growth in the twenty-four well plate, the cell suspension from the seven wells tested immunoreactive to infected H9 cells which were diluted to thirty, fifty and one hundred cells per ml. 0.1 ml of the diluted cell suspensions

(containing an average of three, five and ten clones, respectively) was placed into the wells of a ninety-six well plate. The wells had previously been coated with histone. After each cell grew up to become a colony, the cells were checked under a microscope. The cells of each colony did not

move about and form satellite colonies. The single-cell clone from each of the seven clonings showing strongest reactivities in ELISA and immunofluorescence was chosen and expanded in culture. i) Sodium Dodecyl-Sulfate Polyacrylamide Gel Electrophoresis ( SDS-PAGE^) and Western Blot Procedure for the anti-PND mAbs In Western blot analysis, the virus is solubilized into its component proteins. The clones which produce monoclonal antibodies binding to the exterior envelope protein of HIV-1

(gpl20) are the ones which are desired. The procedure is described below. 30 micrograms of HIV-1 was solubilized by heating it in a sample buffer (which contained 2% SDS and 5% beta-mercaptoethanol) at 100°C for five minutes. It was then loaded onto a 12% slab poly-acrylamide gels 1.5 mm thick. The gel was run at constant voltage of 35 mV for 8 hours at room temperature. The procedure was described in "Procedure for

Preparation of Gels for Western Blot Detection of HTLV-III Antibodies", published by Biotech Research Laboratories, Inc., Rockville, Maryland. The protein bands were transferred onto nitrocellulose paper by setting the power at 30 volts (about 0.1 A) and πinning for 16 hours at room temperature. The next morning, the voltage was increased to 60 volts (about 0.2A) and the transfer was run for 1-2 hours to maximize the transfer of gpl20 and gpl60. The transfer buffer contained 24 g of Tris base, 57.6 g of glycine and 800 ml of methanol. Water was added to make the solution up to 4 liters.

The nitrocellulose sheets were then rinsed with PBS/0.05%

Tween 20 and placed in a tray containing Blotto buffer. The tray was gently shaken for two hours at room temperature. Blotto buffer consists of 50 g of non-fat dry milk, 1.0 g of antifoam A

(optional), 0.1 g of merthiolate, and sufficient PBS to make a final volume of 1.0 liter. The buffer pH was adjusted to 7.0.

The nitrocellulose sheets where then rinsed in PBS/0.05% Tween 20 and dried on a paper towel between weighted plexiglass plates. The nitrocellulose sheets were then cut into strips 0.5 cm

wide, each of which was numbered consecutively. The strips can either be used immediately or stored dry and in the dark for up to one month. The strips which carry the gpl20 band were to be used in the next stage.

The gpl20 nitrocellulose strips were then prepared to allow

binding of monoclonal antibody to the protein bands. Forty of these strips were individually placed into an assigned slot of a slot tray and pre-soaked for twenty minutes in PBS/0.3% Tween 20.

TM The pre-soak solution was aspirated into a Clorox containing trap. The strip wells was then rinsed once with PBS/0.05% Tween

20, the tray was shaken several times, and the solution was aspirated off.

The positive control was made of 2.0 ml of Blotto buffer/4% goat serum (which is made by mixing 100 ml of Blotto buffer and 4 ml of heat inactivated normal goat serum) added to one strip after which 10 microliters of heat inactivated AIDS patient serum was added to the well. 2.0 ml of supernatant was withdrawn from each of the thirty-nine wells in the microtiter plates which contained ELISA positive clones. Mixtures were made which consisted of 2.0 ml of supernatant, 5% non-fat dry milk, 50 microliters of 1 M HEPES (pH 8.0), and merthiolate.

An aliquot of suspension was added into each strip well

TM containing a strip. The mixture was then aspirated into a Clorox containing trap. The strips wells were rinsed once with PBS/0.05% Tween 20, rocked several times by hand, and aspirated with wash buffer. The strips were then washed three times with

PBS/0.05% Tween 20, allowing five minutes for each rinse.

The strips were then reacted with the staining reagents, which permit visualization of specific antibody binding to gpl20. The reagent chosen was horseradish-peroxidase. This reagent exhibits color when contacted by a working substrate which consists of 10 ml of PBS, pH 7.4, 2.0 ml of substrate stock, and 4.0

microliters of 30% H₂0₂. Substrate stock is made by dissolving 0.3 g of 4-chloro-l-napthol in 100 ml of anhydrous methanol.

2.0 ml of Blotto/4% goat serum, containing 1:100 biotinylated goat anti-mouse IgG, was then added to each strip well containing murine mAb, and biotinylated goat anti-human IgG to the control strip reacted with AIDS patient serum. The trays were incubated at room temperature for thirty minutes on a rocking platform. The goat anti-mouse IgG conjugate will, of course, bind to any monoclonal antibody which has bound to the gpl20 on a strip. The strip wells were then rinsed once with PBS/0.05%

Tween 20, and shaken by hand several times to remove excess goat anti-mouse IgG conjugate. The wash buffer was discarded.

The strip wells were then washed three times with PBS/0.05%

Tween 20. Each washing lasted for five minutes. 2.0 ml of Blotto/4% goat serum containing 1:1000 horseradish-peroxidase-avidin D conjugate was added to each strip well. The avidin in this conjugate binds to the biotin in the goat anti-mouse IgG conjugate. Therefore the horseradish-peroxidase marker becomes linked to goat anti-mouse IgG and thereby marks any bound antibody. Following addition of the conjugate, the trays were incubated for thirty minutes at room temperature on a rocking platform. Each strip well was washed three times with PBS/0.05%

Tween 20, five minutes per wash, then once with PBS. 2.0 ml of the working enzyme substrate was added to each well, and the trays were incubated at room temperature until color developed. The working substrate solution contained 0.05% 4-chloro-

1-naphthol and 0.01% H₂0₂ in phosphate buffer saline at pH 7.4.

From the Western blot analysis, using antibody from the thirty-nine ELISA positive wells.only antibody from six of these thirty-nine wells was found to react with gpl20. All six of these wells were among the seven wells which had been found immunofluorescence positive in the immunofluorescence assay. Thus, only one of the seven immunofluorescence positive clones was not also positive in Western blot analysis. j) Reactivities of mAbs against the CD4-Binding Region The binding of the mAbs against the CD4 binding region was examined for reactivity with several different strains of HTV-1, and for the amino acid sequence to which they bound, by the methods described in International Application No.

PCT/US90/02261. h) Production and Purification of Monoclonal Antibodies

Against the PND

To produce large quantities of mAbs against the PND, the following procedure was performed. The seven immunofluorescence positive clones, which have situated in the wells in the second twenty-four well plate, were grown up in a 100 mm tissue culture plate. The expanded culture of the selected seven single-cell clones were then separately injected into the peritoneal cavity of pristane treated mice, using five million cells per mouse. After seven days the ascites fluid of each mouse was collected and frozen.

The monoclonal antibodies in the ascites fluid were purified as follows. The frozen ascites fluid was thawed and filtered through a nylon cloth to remove viscous material. Sufficient phenylmethyl sulfonyl fluoride was added to the ascites fluid so that there was a final concentration of 0.1 mM. 0.05 ml of 1.2M acetate buffer (pH 4.0) was added for every milliliter of ascites fluid. The final concentration of the acetate buffer was 60 mM. The pH was adjusted to 4.5.

For every milliliter of treated ascites fluid, 25 microliters of caprylic acid (MW of 144.21, density of 0.91) was added dropwise with vigorous stirring. The suspension was kept at room tempera- ure and stirred continuously for 30 more minutes. The suspension was then centrifued at 15,000 g for ten minutes in order to remove the precipitate. The supernatant, which contains IgG, was neutralized by adding a volume of 1 M HEPES buffer (pH 8.0) equal to one-tenth the volume of the supernatant. The IgG was then precipitated with 50% (NH₄)₂S0₄.

The precipitate was then dissolved in HEPES saline buffer.

This solution was dialysed overnight against HEPES saline buffer in order to remove (NH₄)₂S0₄ from the IgG. The HEPES saline buffer was changed twice during the dialysis. After dialysis, the

HEPES buffer saline contains purified dissolved IgG. The purified

IgG was used in the infectivity assays and the syncytium formation assays which follow. Example II: Verifying the Efficacy of the mAbs Against the PND a) Neutralization Assay

An assay was performed to determine the effectiveness of the monoclonal antibodies of the invention in inhibiting infection of T-cells by HIV-1 virion. A comparison was made of the number of cells infected when HIV-1 alone was added to a cell culture, with the number infected when HIV-1 and the monoclonal antibodies of the invention were added. H9 cells were selected for the neutralization assay. i) Preparing the Virus. Antibody and Cells H9 cells were prepared by washing a cell culture with H9 growth medium. The H9 growth medium contained 20% FBS

(heat inactivated) in RPMI 1640, 5 mM of L-glutamine, 50

SUBSTITUTE SHEET units/ml of penicillan, 50 mg/ml of streptomycin, and 5 mM of

HEPES. The cells were then resuspended to a final concentration of 2 x 10⁶ cells/ml. The suspension was then incubated with 2 micrograms/ml of polybrene in a water bath at 37°C for twenty minutes.

After incubation, the cells were spun down at 700 g for seven minutes. The supernatant was then discarded, and the cells were resuspended in H9 growth medium and washed again to remove the polybrene. The cells were then resuspended to 2 x 10⁶ cells/ml in growth medium.

Six of the seven immunofluorescence positive clones were chosen for use in the neutralization assay. The antibodies from the purified ascites (as described above) were sterilized by passing them through a 0.22 micron Millipore filter. The solution was then diluted in the H9 growth medium to yield different final concentrations of 100, 10, 1, 0.1, and 0.01 micrograms/ml.

Virus at 20 TCID₅₀, or twenty times the TCID₅₀ value, was used in the infection of H9 cells. The TCID₅₀ value of the virus preparation was determined in previous infectivity assays under the same experimental conditions. It is defined as the virus titer at which 50% of the experimental wells are infected. 20 TCID_™ was equivalent to roughly a 4.72 x 10⁵ dilution of the viral stock. In the infectivity assays, 30 microliters of virus suspension, and 30 microliters of each of the antibody solutions, were mixed in the wells of a microtiter plate at 4°C for one hour. Each well was done in duplicate. The plate was then warmed in an incubator at 37° C and 5% C0₂ for thirty minutes. 30 microliters of the polybrene treated H9 cell suspensions was then added to each well.

The microtiter plates were then incubated for one hour at

37° C in an incubator. 110 microliters of the growth medium was added to each well, bringing the total volume to 200 microliters.

The plates were incubated for three days, and new growth medium was replaced every three days. Cells were collected on the third, sixth, ninth and thirteenth day.

The identical procedure described above was also performed using murine monoclonal antibody to human chorionic gonadotropin (anti-HcG) rather than one of the anti-HIV-1 antibodies of the invention. The cells treated with the anti-HCG antibody served as a negative control. ii) Immunofluorescence Assay of Infected Cells 100 microliter aliquots of the cell suspensions collected on days 9 and 13 were washed with 3 ml of PBS. The cell suspension was centriguted at 700 g for seven minutes and was washed again in PBS. The cells were finally resuspended in 50 microliters of

PBS and 10 microliters of suspension was dotted onto a glass slide.

This suspensions were air dried and then fixed with 1:1 acetone/methanol for ten minutes, air dried and stored at -20°C before assay.

In the assay, the fixed cells were rehydrated in PBS for twenty minutes and then incubated with 5% normal goat serum in

PBS for another thirty minutes. After dripping away the excess normal goat serum, the cells were incubated at room temperature for one hour with anti-p24 monoclonal antibody (at a dilution of

1:100) containing 2% normal goat serum. This antibody binds specifically to the p24 core protein of HIV-1. The slides were kept in the humidifier to avoid drying. After the incubation, the slides were rinsed for three times in PBS for a total of 30 minutes. Then fluorescein conjugated goat anti-mouse IgG (F(ab')₂) fragment was added at a dilution of 1:20. The slides were incubated for one hour at room temperature. The slides were then rinsed in three changes of PBS for thirty minutes and counterstained with 0.5%

Evans blue for five minutes, washed and mounted in Fluoromount G. The cells were then observed under a fluorescence microscope.

The number of infected cells were counted at the magnification of 400x. Four data points were collected from each slide by random sampling over the field. ϋi) Results

The results are depicted graphically in Figs. 1 and 2, where

the percentage of immunofluorescence cells is plotted against the concentration of antibody in suspension. The results in Fig. 1 are from cells collected on day 9. In Fig. 2 the cells were collected on day 13.

Turning to Figs. 1 and 2, it can be seen that four of the six antibodies tested (designated as BAT123, 267, 509, and 085) were effective in inhibiting infection. In particular, BAT123 showed almost complete inhibition of infection on day 9. This results is to be contrasted with the negative control anti-HcG antibody, which exhibited virtually no inhibition. Nearly 100% of the cells treated with anti-HcG were immunofluorescent, irrespective of the concentration of antibody. The similar result was obtained with

monoclonal antibody BAT496 which is reactive with gpl20 but shows no neutralization activity. For this reason, BAT496 was not assayed on day 13 and does not appear in Fig. 2.

It should be noted that another antibody, BAT401, was tested for neutralization. However, the results do not appear in

Figs. 1 and 2 because it was found less effective in syncytium formation inhibition. A comparison of Figs. 1 and 2 shows that as time goes on, more of the cells in the suspension become infected. This result is expected. The amount of antibody in suspension available to neutralize the virus is decreasing due to change in medium and probably degradation or internalization. However, the infected H9 cells continually produce more virus, and this virus eventually infects all the cells.

The plots in Figs. 1 and 2 show that with a decreasing concentration of antibody, a greater number of cells are infected. This indicates that the neutralizing effect of the antibodies is dosage dependent. The IC₅₀ value of each monoclonal antibody, which is the dosage at which 50% of the cells are infected, was calculated. The results as taken on day 9 appear below in Table

I.

It can be seen that the monoclonal antibodies which are most effective at inhibition (BAT123, 267 and 509), do so in nanogram quantities. This indicates that these monoclonal antibodies may also be very effective in minute quanitities for jn vivo AIDS therapy. b) Inhibition of Syncytium Formation by the anti-PND mAbs

Another test for the monoclonal antibodies of the invention was to determine whether they inhibited syncytium formation.

Inhibition of syncytium formation would enhance the therapeutic value of the antibodies, inasmuch as the majority of cell infection and cell death in vivo is believed to occur via syncytium.

The syncytium assay was based on the assumption that the exterior envelope protein of the virus in infected H9 cells binds to the CD4 antigen which is carried by T cells. In the assay, infected H9 cells are added to a well containing CD4 DNA transfected HeLa cells. HeLa cells are used because they adhere, in a monolayer, to the bottom of the well. These transfected HeLa cells express abundantly CD4 antigen on their cell surface. Thus, they have the ability to fuse with infected H9 cells. Therefore, if syncytium formation occurs, aggregates of HeLa and H9 cells will be bound to the well. These multi-nucleated giant cells can readily be observed and counted.

The protocol for the syncytium formation assay is set forth below. (i) Procedure for Syncytium Formation Assay

HeLa-CD4⁺ cells (which express the CD4 antigen on the surface) were grown in a HeLa-T4 growth medium, which

contained 5% FBS (heat inactivated) in DMEM, 5mM L-glutamine, 50 units/ml of penicillin, 50 mg/ml of streptomycin, and 5 mM of HEPES. The cells were harvested by trypsinization to remove the cells from the flask, and washed. The cells were then seeded onto a 96 wells microtiter plate at a density of 10,000 cells per well. The plates were incubated at 37°C for thirty-six hours until 90% confluency was reached.

Both infected and uninfected H9 cells were then prepared.

For preparing these cells, the cell suspension was first washed twice with H9 growth medium (20% FBS in RPMI 1640, 5 mM of

L-glutamine, 50 units/ml of penicillin, 50 mg/ml of streptomycin and 5 mM of HEPES.) The cells were then resuspended in

HeLa-T₄ at a concentration of 0.4 million/ml.

The antibodies were prepared by first performing a sterile filtration on the seven antibody solutions which had been used in the neutralization assay. Six of these solutions contained antibodies of the invention, and the seventh contained the anti-HcG. Each solution was then diluted to make two final

concentration of 1.0 and 10 micrograms/ml.

SUBSTITUTE SHEET 50 microliters of each antibody solution and 50 microliters of infected H9 cell suspension was added to the various wells of the microtiter plate. The microtiter plate wells had previously been coated with the HeLa-CD4⁺ cells. In another HeLa- CD4⁺-coated well, infected H9 cell suspension was added without the addition of antibody. This well was to serve as a positive control. In yet another coated well, uninfected H9 cell suspension was added. This well was to serve as a negative control. The experiments were done in triplicate. The plates were then incubated for eighteen hours at 37° C and 5% C0₂. The plates were washed gently twice with DMEM in order to remove unattached H9 cells. The DMEM was removed and the cells were fixed by adding 200 microliters of methanol per well for seven minutes. After removing the methanol, the cells were air dried, and then stained with 100 microliters of 1.4% methylene blue for ten minutes. The cells were rinsed with distilled water three times.

After staining, the cells were then observed under an inverted microscope (at a magnification of 100 times), and the number of syncytia per field was determined. Aggregates of cells were considered to be a syncytium if more than five nuclei were

present. Each well was counted three times randomly.

SUBSTITUTE SHEET (ii) Results The negative control well showed no syncytium formation. The results for the remainder of the wells appear below in Table

II, expressed as a mean ± standard deviation. TABLE II

Inhibition of Syncytium Formation Between HIV-infected H9 Cells and HeLa-CD4⁺ Cells Antibody*

& Concentration Number of S nc tia Inhibition

* The 1.0 microgram/ml and the 10 microgram/ml solutions of antibody are designated "1" and "10" respectively.

** Not significantly different from negative control.

It can be seen from Table II that screening by the above-described methods is needed to identify the antibodies best- suited for therapeutic use. The same antibodies which lowered

infectivity of free HIV-1 virions (as shown in Figs. 1 and 2) also were effective in inhibiting syncytium formation. BAT123, 267 and

509 were particularly effective in both applications. BAT496 was almost ineffective in both applications as was, of course, the negative control anti-HcG. Although BAT085 was effective in neutralization, it was not among the most effective in syncytium inhibition.

BAT401 was not very effective at syncytium inhibition, although it was effective in the neutralization assay. This result indicates that antibodies which are effective in inhibiting HIV-1 infection are not necessarily effective in inhibiting syncytia formation. Accordingly, the hybridoma producing BAT123, which was most effective at inhibiting both infectivity by the HIV-1 virions and syncytium formation, was deposited at the ATCC in Rockville, Maryland, under Accession number HB 10315. The Table II results demonstrate that, similar to neutralization as shown in Table I, syncytium inhibition is also dosage-dependent. The solutions with 10 microgram/ml of antibody were generally more effective in inhibition than the 1 microgram/ml solutions. Example III: Neutralization of Different Strains and Isolates of

HIV-1 bv the anti-PND mAbs

Several antibodies were found to inhibit the infectivity of free HIV-1 virions and the syncytium formation between HeLa-CD4+ cells and H9 cells infected by HIV-1_B. Since genomic analyses indicate that the virus mutates significantly both in vivo and in vitro (Alizon, M., Wain-Hobson, S., Montagnier, L.

and Sonigo, P. (1986) £gU 46:63-74; Starcich, B.R., Hahn, B.H., Shaw, G.M., McNeely, P.D., Modrow, S., Wolf, H., Parks, E.S.,

Parks, W.P., Josephs, S.F., Gallo, R.C. and Wong-Staal, F. (1986)

Cell 45:637-648)_τ the application of these neutralizing monoclonal antibodies as agents for therapy and protection relies heavily on whether they are group-specific and protect HIV-1 infection caused by a large proportion of strains of the virus in the population. It is important to know whether BAT123 and the other neutralizing monoclonal antibodies of the invention recognize distinct neutralization epitopes in the viral envelope protein gpl20 which have conserved amino acid sequences among different strains and isolates of HIV-1. In order to understand these characteristics of the antibodies, we studied whether these antibodies can inhibit the syncytium formation by other strains of

HIV-1 with a substantial degree of heterogeneity in the amino acid sequence of gpl20. The strains selected for the study were the RF, AL, MN, Z84 and Z34 strains. See Starcich £t .al., supra.

The neutralization antibody BAT123 was chosen for use in the study because it was shown to elicit highest potency in the neutralization of the virus. In order to evaluate the effectiveness of the neutralizing antibodies on different HIV-1 variants existing in the infected population, blood specimens were randomly collected from infected individuals in different parts of the United States (Houston, Texas; Los Angeles, California; Boston,

Massachusetts). These individuals also had different disease states. The effect of BAT123 on the viral infection was examined in the lymphocyte preparations by co-culture experiments. a) Syncytium formation assay The syncytium formation assay was performed as described in Example II. b) Co-culture assay

The procedure used is similar to that described earlier. 30 ml of heparinized blood from each patient was freshly collected and processed for mononuclear leukocytes by density-gradient centrifugation. Briefly, the whole blood was diluted with equal volume of phosphate-buffered saline (PBS). 25 ml of the diluted blood was laid over 10 ml of Ficoll-Paque (Pharmacia) and centrifuged at 1500 x g for 30 minutes. After centrifugation, the interphase containing mononuclear leukocytes was removed and washed twice in PBS. The mononuclear leukocytes were then

cultured at 0.5 - 1 X 10⁶/ml in the RPMI 1640 medium supplemented with 15% heat-activated fetal bovine serum, 2mM

L-glutamine, 10% interleukin-2 (Cellular Products), 25 neutralizing units/ml sheep anti-human alpha interferon (Interferon Science),

100 units/ml penicillin, 100 ug/ml streptomycin and 2 ug/ml Polybrene.

Equal volumes of phytohaemagglutinin (PHA)-stimulated mononuclear leukocytes from normal donor blood was mixed with the patient culture. The mononuclear leukocytes from the normal donor blood was stimulated for one day early with 2 ug/ml PHA-P (Sigma). They were washed twice in PBS to remove the lectin.

BAT123 was added to the test culture at the final concentration of

10 ug/ml. The total volume of the culture was 10 ml. Five ml of the cell culture was removed at 3-4 day intervals, and then centrifuged at 1,500 x g for 15 minutes to remove the cells and debris. The supernatants were collected and assayed for reverse transcriptase activities after precipitation of the virus using 10% polyethylene glycol (PEG) (Gupta, P., Galachandran, R., Grovit,

K., Webster, D. and Rinaldi, C. Jr. (1987) J. Clin. Microbiology

2i:1122-1125). c) Reverse transcriptase assay for anti-PND mAbs

The procedure for the measurement of reverse transcriptase activity is one described by Barre-Sinoussi, F., Chermann, J.C., Rey, F. Nugeyre, M.T., Charmaret, S., Gruest, J., Daugnet, C.

Axler-Blin, C, Vezinet-Brun, F., Ronziou, C, (1984) Science

220:86-87. Briefly, the PEG-precipitated virus was solubilized for

20 minutes in 100 μl of Tris-buffered saline (pH 8.2) containing 0.1% Triton X-100, 2mM dithiothreitol, 0.2mM leupeptin and 50 mM of amino-n-caproic acid. In the assay, 100 μl of the substrate solution in 50 mM Tris-HCl pH 8.2 containing 8mM MgCl₂, 20 uCi ³H-thymidine triphosphate (2mCi/ml), 0.05 units of

template-primer poly(rA).p(dT)._|2_ι o ^was added to 25 μl of the solubilized virus. No template-primer was added to the corresponding control, but substituted with distilled water instead. The reaction mixtures were incubated at 37°C for one hour and the reaction was terminated by addition of 5% cold trichloracetic acid and finally filtered over Whatman GF/C filters which were washed thoroughly and counted for radioactivity using a scintillation counter. The specific reverse transcriptase activities were calculated as the difference in radioactivity when the template-primer was added. d) Results & Discussion We studied the anti-PND mAbs with regard to their group- specificity to the virus and their cross-protection to six different

HIV-1 strains (HIV-1_B, HIV-IRF, HIV-1^, HIV-1_MN, HIV-l^,

SUBSTITUTE SHEET and HIV-l^ In syncytium formation assay between

HeLa-CD4+ cells and H9 cells chronically infected with these strains of HIV-1 respectively, BAT123 at 25 μg/ml inhibited

syncytium formation by almost 80%. It also reduced the syncytium

formation of H9 cells infected with HIV-1_MN, HIV-1^,

HlV-l_β and HIV-l^-o^ by approximately 50%, and HIV-l^ by

23%. (See Table III).

TABLE III

CROSS-PROTECTION OF SYNCYTIUM FORMATION BY H9 CELLS INFECTED WITH DIFFERENT HIV-1 STRAINS

Infected H9 With

Cells Antibody

H9 uninfected (control)

H9-HIV-1B 2.33 ± 0.51*

HIV - 1 2.08 ± 0.38

HIV - 1VT 7.08 ± 0.66

HIV - lrt; 1.91 ± 0.55 HIV - 1** 12.41 ± 1.46

HIV - lZfr34; 1.58 ± 0.14

* Expressed as number of syncytia per microscopical field (x ± S.E., n = 11 or 12), p 0.05, paired student's t test.

The co-culture experiments used lymphocytes isolated from the peripheral blood of patient clinically diagnosed as positive but in an asymptomatic state for AIDS or ARC. Out of 32 patient blood specimen tested, the virus had been isolated from 18 samples as measured for reverse transcriptase activities. When

10 μg/ml BAT123 was added in the culture medium throughout the experiments, the viral replication was inhibited in all of the 18 virus-positive cultures. The degree of inhibition ranged from 43.7 to 100%. Among the 18 samples, 8 samples were effectively

inhibited by over 90%. The results from our jn vitro experiments suggest that the neutralizing monoclonal antibody BAT123 can cross-protect different diverse strains of HIV-1 in the syncytium formation assays and inhibit viral infection in patient blood specimen.

Example IV: Determining The Peptidic Segments of gpl20 Reactive With anti-PND Monoclonal Antibodies

In order to map the epitopes on gpl20 of HTV-1 that are recognized by the monoclonal antibodies, Western blot assays were used to determine the reactivities of some of the monoclonals. The strips were obtained from Dr. Steve Petteway, Medical

Products Department, DuPont de Nemours and Company, Wilmington, Delaware. The synthetic peptides on the strips are 8-20 amino acid residue long. These peptides represent overlapping peptidic segments across the entire length of gpl20 of HTV-1B strain. Several tens of peptide solutions had been adsorbed on individual strips in equally spaced regions. The strips were provided in a dry form.

The immunoblotting procedure using the nitrocellulose strips is the same as the Western blot procedure used to determine whether the monoclonal antibodies react with gpl20 described in the preceding section.

Three of the monoclonal antibodies, BAT123, BAT267, and BAT085, showed very clear and specific reactivities with particular peptides in the Western blot assay. The reactivities for these three mAbs were as follows:

BAT267 RPNNNTRKRIRIQRG (residue #298-312)

BAT123 RIQRGPGRAFVTIGK (residue #308-322) BAT085 VQKEYAFFYKLDIIP (residue #169-183)

The 15 amino acid long peptides reactive with BAT267 and

BAT123 overlap by 5 amino acids. However, the antibodies react with just one of them and do not react with the other to any measurable extent. The antibodies do not react with peptides overlapping at the other ends either, i.e. BAT267 does not react with LNQSVRINCTRPNNN and BAT123 does not react with

VΗGKIGNMRQAHCN. These results suggest that the antibodies react with an epitope borne by either all or a part of the middle five amino acids or a combination of these amino acids with some of the flanking amino acids. Similar findings have been made for BAT085 and similar conclusions may be made for it. Example V: Immunotoxins with anti-CD4 Binding Region mAbs a) Methods

(i) Antibodies

Murine monoclonal antibody G3.519 (IgGl) binds to a conserved region in or near the CD4 receptor binding site on gpl20 and is known to have neutralizing activities against diverse strains of HTV-1. See Published International Application

PCT/US90/02261. BAT123 (IgGl) recognizes a relatively variable peptidic segment in gpl20 and exhibits effective neutralizing activity against HTLV-III_B strain. The details of the neutralizing activity of BAT123 is described above. BAT123 and G3.519 were used in the immunoconjugate study.

(ii) Purification of PAP-S

PAP-S was purified from seeds of Phytolacca americana (pokeweed) using a method of Barbieri et al. (L. Barbieri, et al..

Biochem J.. 203: 55-59 (1982)). Briefly, pokeweed seeds (100 g) were homogenized in 500 ml of 5 mM phosphate buffer (pH 6.0).

Insoluble materials and lipid were removed after centrifugation at

10,000 x g for 1 hour. Supernatant was applied to a CM-Sepharose Fast Flow (Pharmacia, Piscataway, NJ) column equilibrated with 5 mM phosphate buffer. After washing, bound proteins were eluted with a NaCl gradient (0-0.3M). The peak containing activity of ribosome inactivation was pooled and dialyzed against PBS using a membrane with a molecular weight

cut-off of 10,000 daltons.

(iii) Antibody and toxin conjugation Monoclonal antibody (10 mg) was reacted with a heterobifunctional crosslinking reagent,

N-succinimidyl-3-(2-pyridyldithio) propionate (Pharmacia) at 1:3 molar ratio as described by Carlsson et al. (J. Carlsson, et al..

Biochem J.. 173:723-737 (1978)). Pokeweed antiviral protein (6 mg) was reacted with 2-iminothiolane at 1:3 molar ratio as described previously (J.M. Lambert, et al. Biochemistry. 17:

406-416 (1978)). Excess of chemicals were removed by gel filtration using a 10PD columm (Bio-Rad, Richmond, CA).

Chemically modified antibody and PAP-S were combined and incubated at room temperature for 2 hours or at 4°C overnight.

Uncoupled PAP-S was removed by gel filtration on a Sephacryl

S-200 (HR) column equilibrated with PBS. The fractions containing antibody and conjugate were pooled, concentrated to 10 ml and then dialyzed against 5 mM phosphate buffer, pH 6.0. Two ml samples containing immunoconjugate and uncoupled antibody were injected into a cation exchange column (Mono S,

Pharmacia) equilibrated with 5 mM sodium phosphate buffer, pH 6.0. Immunoconjugate was eluted with a NaCl gradient and absorbance at 280 nm was recorded. The immunoconjugate was eluted as a single peak at 110 mM NaCl. The composition of the immunoconjugate was analyzed by a 7.5% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) under a non-reducing condition (Fast System, Pharmacia) along with molecular weight markers (Bio-Rad).

(iv) Cytoxicity assay of immunoconjugates

The ability of the immunoconjugate to kill HIV-infected cells was assayed by the inhibition of (³H)-thymidine incorporation of infected cells. H9 cells either uninfected or chronically infected by HIV-1 strains (HTLV-ffl_B, HTLV-HI_RF, and HTLV-IH_MN) were maintained in log phase in RPMI1640 supplemented with 15% heat-inactivated fetal bovine serum, 100 U/ml of penicillin and 100 μg/ml of streptomycin. One hundred and eighty μl of 5 x 10⁴

cells/ml were dispensed into each well of a 96-well microtiter plate. Twenty μl of purified immunoconjugate of 5-fold serial dilutions (100 ug/ml - 160 ng/ml) were added into the wells in triplicates. In some assays in which the specificity of killing by the immunoconjugates was examined, infected H9 cells were incubated with mAb at 37°C for 30 minutes prior to addition of the

corresponding immunoconjugates. The controls were an irrelevant immunoconjugate of PAP-S or a mixture of the unconjugated antibody and PAP-S at equivalent concentrations under the identical conditions. The cell cultures were kept at 37°C for 24

hours and pulsed with 1 μCi per well of (³H)-thymidine for another 4 hours. Cells were harvested on glass fiber filters using a Skatron cell harvester and the (³H)-thymidine retained on the dry filter was measured by scintillation counting. The inhibition of cellular thymidine incorporation was calculated by comparing the radioactivity of test cultures to that of the control. (v) Flow cytometry

The binding activity of the immunoconjugates to H9 cells infected by HTLV-IIIB was studied by flow cytometry. Briefly, H9 cells in the log phase uninfected or infected by HIV-1 were washed twice in PBS containing 1% bovine serum albumin at 4°C. The cells were resuspended at 1 x 10⁷/ml in the same buffer. Fifty μl of the cell suspension were incubated with 50 μl of diluted immunoconjugates (10-0.1 μg/ml) at 4°C for 30 minutes. The cells were then washed with 3 ml of the buffer. Then 50 μl of goat anti-mouse F(ab')₂ conjugated with fluorescein isothiocyanate (1: 10 dilution) was added, and the mixture was incubated for another 30 minutes at 4°C. The cells were then washed with 3 ml of the buffer and fixed with 0.5 ml 1% paraformaldehyde and assayed by flow cytometry.

(b) Results

(i) Purification and characterization of immunoconjugates

Immunoconjugate was isolated from uncoupled monoclonal antibody using a cation exchange column (Mono S). Since the isoelectric points of both mAbs BAT123 and G3.519 are approximately 6.8, unmodified mAb did not bind to the column during the sample loading in 5 mM phosphate buffer, pH 6.0. The immunoconjugate was eluted from the Mono S column as a single peak at 110 mM NaCl concentration (Fig.3). However, this peak, when analyzed by 7.5% SDS-PAGE under the non-reducing condition, resolved into two protein bands. The higher molecular weight band represented the conjugate containing two molecules of PAP-S per molecule of antibody and the lower band a conjugate with one molecule of PAP-S per antibody molecule.

Densitometric analysis of Coomassie blue stained gels indicated that the higher molecular weight conjugate accounted for about 25% of the total immunoconjugates (data not shown). The binding activity of the immunoconjugates determined by ELISA using synthetic oligopeptides comprising the binding epitopes of the respective mAbs revealed no impairment of antibody activity after conjugation (data not shown). Both mAbs BAT123 and

SUBSTITUTE SHEET G3.519 are specific to gpl20 but recognize distinct epitopes with different binding constants (2.9 x 10¹⁰ M^"1 and 6.9 x 10⁸ M^'1 respectively). As shown in Figure 4, BAT123-PAP-S bound more

than 90% of HTLV-III_B-infected H9 cells at 0.05 μg/ml, while G3.519-PAP-S required 5 ug/ml to show the same degree of binding.

(in Cytotoxicity of G3.519-PAP-S on HIV-1 infected cells

G3.519-PAP-S binds to the CD4-binding region of gpl20 in which the amino acid sequence is conserved. H9 cells infected separately with three diverse strains of HIV-1 were all sensitive to

G3.519-PAP-S treatment (Figure 5). The immunoconjugate killed

H9 cells infected with HTLV-IIIMN more effectively (IC_5Q=1.4 x 10^"10 M) than it killed H9 cells infected with HTLV-IIIB (IC_5Q

= 1.7 x 10^"9 M) and HTLV-IIIRF (IC_5Q = 1.2 x 10^"9 M). In contrast to the differing cytotoxic activity exhibited by

BAT123-PAP-S against H9 cells infected with HTLV-IIIB and

HTLV-IIIRF, H9 cells infected with those strains were equally sensitive to G3.519-PAP-S (Figure 5). G3.519-PAP-S was not o cytotoxic to uninfected H9 cells up to a concentration of 5.3 x 10 M.

(iii) Cvtotoxicity of BAT123-PAP-S on HIV-1-infected cells

The cytotoxic activity of the BAT123-PAP-S immunoconjugate was evaluated using H9 cells infected with the same diverse strains of HIV tested with G3.519-PAP-S. Incubation of H9 cells infected by these diverse strains of HIV-1 with

BAT123-PAP-S showed various degrees of specific killing (Figure 6). H9 cells infected with HTLV-III_B were most susceptible to BAT123-PAP-S treatment with an IC_5Q of 5.2 x 10^"11 M. At 2.6 x 10^"9 M BAT123-PAP-S inhibited DNA synthesis by more than

90%. BAT123-PAP-S also killed H9 cells infected with

HTLV-III_MN (IC₅₀ = 4.9 x 10^"9 M). However, BAT123-PAP-S was relatively ineffective in killing H9 cells infected with the

HTLV-III_RP strain (IC_5Q = 3.9 x 10^"8 M). BAT123-PAP-S did not inhibit DNA synthesis of uninfected H9 cells up to a concentration of 5.3 X 10^"8 M.

(iv) Specific cytotoxicity of immunoconjugates The specificity of immunoconjugates against target cells was studied by antibody blocking experiments. As shown in Figure 7, addition of specific mAb prior to the incubation of target cells with the immunoconjugates efficiently blocked the cytotoxicity of immunoconjugates in a dose-dependent fashion. BAT123-PAP-S at 60 ng/ml inhibited 90% of DNA synthesis of H9 cells infected with HTLV-iπ_B strain. Addition of 4 μg/ml of BAT123 abolished approximately 60% of its cytotoxicity. Similarly, G3.519-PAP-S at 0.6 μg/ml inhibited 55% of DNA synthesis of H9 cells infected with HTLV-III_B. In the presence of 50 fold excess of mAb

G3.519, the immunoconjugate inhibited only 13% of DNA

synthesis of target cells. Irrelevant monoclonal antibody did not inhibit the cytotoxicity of immunoconjugates even at 200 fold excess of the free antibody. In addition, irrelevant mAb containing

PAP-S did not show any cytotoxicity against H9 cells infected with

HTLV-III_B (data not shown).

(c) Discussion Two immunoconjugates were synthesized by chemically coupling different murine neutralizing monoclonal antibodies which recognize gpl20 of HIV-1 to PAP-S through a disulfide

bond linkage. These immunoconjugates, designated

BAT123-PAP-S and G3.519-PAP-S, showed specific cytotoxicity against human T cells infected with HIV-1. Epitope mapping studies using synthetic polypeptides revealed that mAb BAT123 recongizes the relatively variable region (amino acid 308-322) of gpl20 and mAb G3.519 recognizes a relatively conserved region

(amino acid 423-437) of gpl20. When tested against H9 cells infected with HTLV-IIIB,

BAT123-PAP-S showed cytotoxic effectiveness at lower

concentrations than G3.519-PAP-S (IC_5Q 5.2 x 10^"11 M vs 1.7 x ^10-9 M). The increased effectiveness of BAT123-PAP-S with HTLV-III_B infected H9 cells may result from the higher affinity of BAT123 antibody and/or the location of the epitope on gpl20, since both immunoconjugates contained PAP-S and the same target cells were used in the studies.

Although BAT123 is directed against the variable region in gpl20 of HTLV-πi_B, BAT123-PAP-S still was effective in killing H9 infected with HTLV-III_MN while H9 cells infected with HTLV-III_RP were not effectively killed. However, in neutralization studies, BAT123 showed a similar degree of ineffectiveness against both HTLV-III_RP and HTLV-iπ_MN. The recent data with

BAT123-PAP-S showing its ability to kill H9 cells infected with both HTLV-III_B and HTLV-III_MN suggests a broader application of this antibody for use as an immunoconjugate. Epitope mapping of G3.519 revealed that the binding site of the antibody resides on the CD4 binding region of gpl20 which is conserved among diverse strains of HIV. G3.519-PAP-S specifically killed H9 cells infected with three diverse strains. It is interesting to note that even though mAb G3.519 was generated against gpl20 of the HTLV-III_B strain, H9 cells infected with

HTLV-πi_MN strain were more sensitive to G3.519-PAP-S than H9

cells infected with HTLV-IH_B (Figure 5, IC_5Q = 1.4 x 10^"10 M for HTLV*III_MN vs. 1.7 x 10-9 _{M for} HTLV-III_B).

These in vitro studies clearly show that immunoconjugates directed to the CD4 binding region of gpl20 are effective in killing

human T cells infected with diverse strains of HIV-1. These immunoconjugates may have a longer half-life in vivo as compared to the sCD4-toxin conjugates (which are known to be cleared rapidly in vivo (D J. Capon, et al.. Nature. 337: 525-531 (1989)), and may represent a better approach for specific killing of HIV infected cells. Example VI: Anti-idiotypic antibody to BAT123 and G3.519

Antibodies can be induced by the individually unique idiotypic determinants (idiotopes) of the first antibodies (Ab-1) that are induced by specific antigens. A subset of these anti-idiotypic antibodies or anti-id (Ab-2) recognizes the antigen-combining site (paratope) of Ab-1 and thus bear the internal image of the

nominal antigen. These paratope-specific anti-id, whose binding to Ab-1 can be inhibited by the antigen, are referred as Ab-2β. Anti-id recognizing other idiotypic determinants are classified as Ab2α and Ab-2γ: those whose binding to Ab-1 can be inhibited by the antigen are Ab-2γ and those whose binding to Ab-1 cannot be inhibited by the antigen are Ab2α (Jerne, N.K. _εiϋl. EMBO J.

1:243, 1982). In several experimental systems involving various antigens, Ab-2β have been shown to induce in animals antibodies

(Ab-3) that are reactive with the antigen in a way similar to that

Ab-1 reacts with the antigen (Kennedy, R.C. et al. Sci. Am.255:48,

1986). The HIV-neutralizing anti-PND Mab, BAT123, which binds to a peptide segment of amino acid residue #308-322 of HIV-1 gpl20, was selected to generate anti-idiotypic (anti-id) Mabs.

BAT123 was shown to neutralize the infectivity of HTLV-IIIB virions and to block syncytium formation between HTLV-IIIB-infected T cells and uninfected CD4⁺ HeLa cells (see above). In immunofluorescene flow cytometric analysis, BAT123 was shown to bind specifically to HTLV-III_B and HTLV-III_MN-infected T cell line (H9). Recent studies suggest that HTLV-iπ_B and HIV-III_MN may be among the prevalent HTV-1 strains in the HIV-infected populations in the U.S.A. and in

Europe (Goudsmit .el jil. Abstract in Annual Meeting of the Laboratory of Tumor Cell Biology. National Cancer Institute. U.S.A.). The development of neutralization mAbs against HTLV-iπ_B and _MN strains and of their anti-id for therapeutic and prophylactic applications seems to be an attractive approach.

The anti-CD4 binding region mAb G3.519 was also selected to generate anti-ids. G3.519 exhibited significant inhibition of binding to HTLV-III_B and HTLV-III_RP to CD4+ C8166 cells. The

G3.519-PAP-S immunoconjugate was cytotoxic to H9 cells infected with HIV-1 strains HTLV-III_B, HTLV-III_MN, and HTLV-III_RF.

(a) Generation of hybridomas secreting antibodies specific for idiotypes of Mab BAT123.

(i) Generation and screening of anti-Id mAb

BAT123 was conjugated to keyhole limpet hemocyanin

(KLH) using glutaraldehyde as described by Maloney, D. G., M.

S. Kaminsky, D. Burowski, J. Haimovich, and R. Levy, Hybridoma 4:191 (1985). Five male BALB/c mice (Harlan Sprague Dawley

Inc.) were immunized i.p. with 100 μg of BAT123-KLH conjugate in CFA at 1-month intervals for 3 months. Three days after the final immunization, the mice were sacrificed, and spleen cells isolated and fused with Sp2/0 myeloma cells as described by Fung, M.C. et al, Bio /Technology 5:940 (1987). After selection using

HAT medium, culture supernatants were tested for reactivity with BAT123 by ELISA, in which anti-BAT123 antibodies were bound by solid-phase BAT123, and detected by BAT123-HRP conjugate, which was prepared by the method of Wilson and Nakane, "Recent developments in the periodate method of conjugating horseradish peroxidase (HRPO) to antibodies." Immunofluorescence and Related Staining Techniques. W. Knapp ed. Elsevier/North- Holland, Amsterdam and New York, p. 215. (1978). Briefly, wells of 96-well ELISA plates (Immunlon 2,

Dynatech Laboratories, Alexandria, VA) were coated with 50 μl of BAT123 (2 μg/ml in PBS, pH 7.5) for overnight at room temperature, and treated with BLOTTO containing 5% non-fat dry milk in PBS. Fifty μl of hybridoma culture supernatants was added to the antibody-coated wells and incubated for 1 hour at room temperature. After washing with PBS contaimng 0.05%

Tween 20 (PBST), 100 μl of BAT123-HRP conjugate (dilution

1:1000, in PBST with 1% BSA), was added to each well for incubation at room temperature for 1 h. The plates were then washed, 200 μl peroxidase substrate solution contaimng 0.1% 3', 3', 5', '5 - tetramethyl benzidine (Sigma) and 0.003% H₂0₂ (Sigma) was added to each well for color development for 30 minutes. The reaction was stopped by addition of 50 μl of 2 M H₂S0₄, and the OD measured by a BioTek ELISA reader at 450 nm.

Reactive culture supernatants with OD greater than 0.2 were further tested for the ability to inhibit the binding of ¹²⁵I- labeled gpl20 to immobilized BAT123. The gpl20 was purified from HTLV-iπ_B-infected H9 cell lysates as described by Sun et al, /. Virol 63:3579 (1989), and radioiodinated by the method described by Bolton, A. E. et al Biochem. J. 133:529 (1973).

In this assay, wells of 96-well PVC plates (Falcon, Becton Dickinson, Lincoln Park, NJ) were coated with 100 μl of BAT123

(4 μg/ml in PBS, pH 7.5) for overnight, treated with BLOTTO for

1 h at room temperature, and washed with PBST. Forty μl of ^12SI

- gpl20 (approximately 30,000 cpm) and 60 μl of hybridoma supernatant were added to each well. After the plate was incubated for 1 h at room temperature and washed, the wells were air-dried and the radioactivity in each well measured by a Packard gamma counter. Microcultures showing strong inhibition of the binding of ¹²⁵I - gpl20 to BAT123 were single-cell cloned by a limiting dilution method. The clone whose antibody showed the strongest activity in inhibition, designated AB19-4 and found to be

IgGl, K, was selected for further characterization. The antibody

(designated AB 19-31) of another clone also showed strong activity in inhibition. The positivity in the above two screening assays indicates that

the mAb was Ab2β or Ab2γ. To confirm the anti-Id nature of this mAb, its identical binding to BAT123 and its chimeric form

CAGl-51-4 (the V regions of CAGI-51-4 are identical to those of

BAT123), AB19-4 was tested by ELISA. The mAb to be tested was conjugated to HRP, and the procedures for ELISA were the

same as described above. The mAb G3.519, also of l,k subclass

as BAT123, was used as a control for reactivity with C rather than with V regions.

As shown in Fig. 11, the AB19-4-HRP conjugate bound specifically to BAT123 but not to murine G3.519. Also, AB19-4 reacted with CAGI-51-4 and with BAT123 (Fig. 11). To aid in characterizing the binding region of AB19-4 and

AB 19-31, competition assays of the binding between BAT123 and

AB19-4, and BAT123 and AB 19-31, by synthetic peptides corresponding to the BAT123 binding regions in HTLV-III_B,

HTLV-III_MN and HTLV-III_RP were run. The results are shown in Figures 8A and 8B. T64-63-6 is an irrelevant peptide used as control.

To demonstrate that the anti-idiotypes AB19-4 and AB19-

31 generated suitable Ab3 in vivo, the Ab3 generated in rabbits immunized with AB19-4 and AB 19-31 were assayed for binding to gpl20, and the control proteins BSA and KLH. The results are shown in Figure 9.

The effectiveness of the rabbit Ab3 in neutralizing two strains of HIV-1 (HTLV-III_B, HTLV-IH_MN) was also demostrated using a syncytial formation assay. The results are shown in Figure 16. (ii) Inhibition of Id-anti-Id reactions by gp!20 and synthetic peptides defining the binding epitope of BAT123.

In a competition ELISA, the binding of Ab2-HRP conjugate to immobilized BAT123 was tested in the presence of purified

HTLV-III_B gpl20 and the synthetic epitope peptides of BAT123, derived from the gpl20 of HTLV-III_B, HTLV-III_MN and HTLV-

III_RP: R15K, R15N, and S15Q respectively. Their amino acid

sequences were obtained from Myers et al. eds. Human retroviruses and AIDS. Los Alamos National Laboratories. Los

Alamos, NM. (1989). The oligopeptides were synthesized using a DuPont RaMPS peptide synthesis system (Wilmington, DE). In the ELISA, wells of Immunlon 2 plate were coated with 100 μl of BAT123 (10 μg/ml). Fifty μl of serial dilutions of gpl20 (20 μg/ml to 20 ng/ml) or peptides (200 μg/ml to 80 ng/ml) and 50 μl of diluted Ab2-HRP conjugate were added to the wells. KLH and an irrelevant HTLV-III_B gpl20-derived peptide T19V in the C2 region (amino acid residue #254-275) were used as negative controls. The working dilution of the Ab2-HRP conjugate was determined to be within the linear range of its binding to BAT123.

Subsequent steps of the ELISA were the same as described above. The degree of inhibition by gpl20 and the peptides was expressed

as the percent decrease in OD as compared to the control without inhibitors.

HTLV-III_B-gpl20 inhibited the binding between AB19-4-

HRP conjugate and BAT123 in EUSA (Fig.11). G3.519 did not inhibit the binding. These results indicate that AB19-4 is either an Ab2ø or an Ab2γ.

To analyze whether AB19-4 is an Ab2β or Ab2γ, it is necessary to determine whether the idiotope on BAT123 recognized by AB19-4 coincides with the complementarity determining region (CDR) or Ag-binding site of BAT123. For these analyses, small synthetic epitope peptides of gpl20 (15 amino acid residues) reactive with BAT123 were employed. See Table

IV below.

TABLE TV

Amino acid sequences of the V3 neutralization domain in HTLV- III_B gpl20 recognized by BAT123 and its corresponding regions in HTLV-III_MN and HTLV-III_RF. Peptide HIV Isolate Amino Acid Sequence*

R15K IIIB R I Q R G P G R A F V T I G K

R15N MN R I H - Y - T K N

S15Q RF S I T K V I Y A T - Q

'Homologous sequences are represented by hyphens, and deletions by spaces.

If AB19-4 is an Ab2γ, it should be only the intact gpl20 but not the peptides that can inhibit its binding to BAT123 via steric

hindrance; on the other hand, if AB19-4 is an Ab2β, the peptides should also inhibit the binding because they bind to the CDR. In

ELISA, peptide R15K of HTLV-III_B and R15N of HTLV-III_MN were found to compete the binding effectively with ID₅₀ 0.35

μg/ml and 18 μg/ml respectively (Fig. 12). Peptide S15Q of HTLV-III_RP showed slight inhibition at high concentrations ( > 100 μg/ml) and the irrelevant peptide T19V had no effect on the binding (Fig. 12). The relative inhibitory activities of these epitope peptides in the assays are consistent with the binding affinities of BAT123 to these peptides in ELISA. These results indicate that AB19-4 constitute a good Ab2β candidate. This tentative conclusion was substantiated by the functional studies described below.

(iii) Reactivity of Ab2 with goat and sheep anti- HTLV-III_B gp!20 antisera. Wells of Immunlon 2 plates were coated with 100 μl of Ab2

(10 μg/ml in PBS) for overnight at room temperature. After the wells were treated with BLOTTO and rinsed with PBST, 100 μl of goat or sheep antiserum to HTLV-III_B gpl20 (diluted 1:100 to 1:1600) in PBST containing 1% BSA and 10% normal mouse serum, was added. The binding was detected by HRP-conjugated rabbit anti-goat IgG (Fisher Scientific Co., Springfield, NJ), which

also reacted well with sheep IgG. Control wells were coated with protein A-purified normal mouse IgG. Both normal goat and sheep sera were also tested.

Figures 13A and 13B show that both the goat and sheep antisera bound to solid-phase AB19-4 but not to normal mouse

IgG, indicating that the reactive antibodies contained therein bound to the V region of AB19-4. These results suggest that the idiotope on murine BAT123 recognized by AB19-4 is shared by the Abl generated against the antigenic epitope on gpl20 in other animal species. This further indicates that AB19-4 is an Ab2β.

(iv) Immunization of rabbits with Ab2 and production of Ab3.

Three adult NZW rabbits (Jackson Laboratory, Bar Harbor, ME) were given three bi-weekly s.c. injections of 100 μg of Ab2 emulsified with IFA (1:1) (Difco Laboratories, Detroit, MI). Pre- immune sera were collected for use as controls. Sera were obtained prior to each subsequent injection and monitored for anti-gpl20 reactivity by ELISA. The sera used for characterization of Ab3 were collected one month after the final immunization.

Ab3 in antisera showing anti-gpl20 reactivity was affinity- purified by AffiGel-10 (BioRad, Richmond, CA) coupled with purified HTLV-III_B gp 120. Five ml of diluted antiserum (1:1 with

PBS) was loaded on a 3-ml gel column, which was then washed with 30 ml PBS. The bound proteins were eluted with 0.5 M

acetic acid, and this eluate was immediately adjusted to neutral pH with 1 N NaOH and dialyzed against PBS for overnight at 4°C.

The protein in the eluate was quantitated by the BCA protein assay (Pierce Chemical Co., Rockford, IL) and its purity examined by SDS-PAGE under reducing and non-reducing conditions. Pre- immune sera were sham purified by the same procedure and the eluate used as control.

After the third immunization, antisera from all three rabbits showed reactivity with gpl20 in ELISA with titers of 5,000 -

10,000. Representative results of the antiserum from one of the three rabbits showing highest reactivity with gpl20 are shown in

Fig. 9.

When the rabbit antisera, which were diluted 100 fold or more to minimize possible nonspecific inhibitory activity, were tested for neutralizing activity in HIV-1-induced syncytium formation assays, no significant neutralizing activity was observed.

To investigate quantitatively the BAT123-like Ab3 in the rabbit antisera, gpl20-reactive substances in the antiserum from the same rabbit as shown in Fig. 9 were affinity-purified by AffiGel-10 conjugated with HTLV-III_B gpl20. SDS-PAGE analyses under both reducing and non-reducing conditions revealed that the eluate

contained over 85% rabbit IgG. Pre-immune serum from the same rabbit was also sham purified and the eluate used as control. (v) Reactivities of Ab3 and BAT123 with gp!20 and epitope peptides.

Wells of Immunlon 2 plates were coated with 100 μl of gpl20 (1 μg/ml) or peptides R15K, R15Q, or S15Q (2 μg/ml) for overnight at room temperature. Serially diluted Ab3 or BAT123

(4 μg/ml to 1 ng/ml) was added and their binding detected by

HRP-conjugated goat anti-rabbit IgG or HRP-conjugated goat anti-mouse IgG (Fisher Scientific, Springfield, NJ), respectively.

KLH and peptide T19V were also tested as controls. In testing the purified Ab3 for reactivity with various Ag in

ELISA, it was found to react specifically with HTLV-III_B gpl20 and with peptides R15K of HTLV-III_B and R15N of HTLV-III_MN, but not with KLH, peptide S15Q of HTLV-III_RP and the irrelevant peptide T19V (Fig. 14). The sham purified pre-immune rabbit serum substances did not react with gpl20 or with any of the peptides (data not shown). In parallel assays, BAT123 (Abl) reacted with gpl20, peptides R15K and R15N (data not shown). Contrary to BAT123, the Ab3 reacted more strongly with peptide R15N than with peptide R15K; both of them did not react with peptide S15Q.

The binding of the purified rabbit Ab3 and BAT123 to H9 cells infected by HTLV-III_B, HTLV-III_MN, or HTLV-IH_RF was analyzed by flow cytometric methods using a Coulter EPIC cell analyzer as described by Kim, Y. W. et al., /. Immunol. 144:1257

(1990). Sham gpl20-purified pre-immune serum substances and an irrelevant mAb anti-hCG were used as controls. In the experiments examining the effect of Ab2 on the binding of BAT123 to HTLV-III_B-infected H9 cells, the cells were incubated with 100 ng/ml BAT123 in the presence of Ab2 (1 μg/ml or 10 μg/ml).

In the flow cytometric assay as shown in Figure 15A, the

Ab3 at 10 μg/ml stained H9 cells infected with HTLV-III_B or HTLV-III_MN, by 54.2% and 35.1% respectively. In the same assay,

BAT123 stained 81.6% of H9 cells infected by HTLV-III_B and

44.2% H9 cells infected by HTLV-III_MN (Fig. 15B). Neither the

Ab3 nor BAT123 bound to H9 cells infected with HTLV-III_RP

The sham purified pre-immune serum substances as well as the irrelevant mAb anti-hCG did not bind to the uninfected or HIV-1- infected H9 cells.

(vi) HIV-1 neutralization assay.

The syncytium-forming microassays using CEM-SS cells as targets were performed as described by Nara. P. L. et al., AIDS Res. Hum. Retroviruses. 3:283 (1987); Sun, N. C. et al., /. Virol.

63:3579 (1989). Briefly, 50 μl of diluted Ab3 or BAT123 was mixed with 50 μl viral culture supernatant containing 200 syncytium-forming units (SFUs) of HTLV-III_B, HTLV-III_MN, or

HTLV-IH_RP, and incubated for 1 h at room temperature. The mixtures were added into microculture wells containing 5 x 10⁴

DEAE-dextran-treated CEM-SS cells, and the cell cultures were maintained in 5% C0₂ at 37°C for 3 to 4 days. The syncytia were enumerated under an inverted microscope. The neutralizing activity was expressed as ID₅₀, defined as the concentration required to achieve 50% inhibition of the infection (i.e. Vn/Vo =

50%), where Vn is the SFUs in the test wells and Vo the SFUs in the control without test antibodies. Sham purified pre-immune rabbit serum substances and a non-HIV-neutralizing mAb BAT496 (Fung, M. S. et al., Bio/Technology. 5:940 (1987)) were used as controls.

In the syncytium-forming microassay for HIV infectivity, in which 5 x 10⁴ CEM-SS cells were infected with a viral inoculum of

100-200 SFUs, the Ab3 was found to neutralize HTLV-III_B and HTLV-III_MN with ID^ 0.7 and 0.33 μg/ml respectively (Fig. 18), whereas BAT123 neutralized only HTLV-IH_B with ID₅₀ 0.11 μg/ml, but not HTLV-III_MN (Fig. 19). Both the Ab3 and BAT123 did not neutralize HTLV-III_RP at concentrations as high as 2 μg/ml (data not shown). The sham purified pre-immune serum substances as well as the control mAb BAT496 showed no effect on the infectivity of the viruses in the assays.

(vii) Results

The results indicate that AB19-4 is a paratope-specific

anti-id (Ab-2β) of BAT123 (Ab-1). The Ab-3 produced in Ab-2-immunized rabbits elicited an Ab-1 like Ab-3 response which was specific to gpl20 and to its specific epitope peptide and conferred HIV-neutralizing activities. Further, there was broadening of neutralizing activities of Ab-3 for HIV strains including both MN and IIIB as compared to Ab-1. The finding that Ab-3 (AB19-4) exhibit broader reactivity to HIV is important, because it shows that via some as yet undefined mechanisms of jn vivo immunomodulation, a type-specific anti-HIV humoral immunity can be transformed, into a broader anti-HIV reactivity. In view of the genomic heterogeneity of HIV, it is obvious that the paratope-specific anti-id's generated from broadly reacting HIV-neutralizing antibodies may have wider application.

(b) Generation of hybridomas secreting antibodies specific for idiotypes of mAb G3.519. (i) Generation and screening of anti-Id mAb.

Anti-idiotypes to mAb G3.519 were generated and screened

by essentially the same process as the anti-idiotype AB19-4 was made. The resulting anti-idiotypes were designated the AB20 series, and AB20-4 was deemed the one with the most suitable properties for further studies.

(ii) Properties of AB20-4

To verify that AB20-4 bound to G3.519, a double-capture ELISA was run with G3.519 as the solid phase antigen, and

G3.519-HRP as the marker. The results are shown in Fig. 18, where it can be seen that the amount of AB20-4 bound increased in a concentration-dependent fashion, and that an irrelevant peptide (designated by open circles) did not bind to AB20-4. To demonstrate that AB20-4 competes with gpl20 for binding to G3.519, a competition assay, using solid-phase gpl20 and G3.519-HRP as the marker, was performed. The results are shown in Fig. 19, where it can be seen that increasing concentrations of AB20-4 or G3.519 inhibited binding of the G3.519-HRP to solid-phase gpl20, and that an irrelevant murine

IgGl did not.

The peptide T35S has the same sequence as the gpl20 CD4-binding site of HTLV-IH_B (amino acid residue numbers 413- 447, amino acid sequence TTTLPCRIKQIINMWQKVGKAMYAPPISGQIRCSS). To demonstrate that T35S inhibited binding between G3.519 and AB20-4, an assay was run with G3.519 as the solid-phase antigen, and AB20-4-HRP as the marker, where T35S was used as an inhibitor. The results are shown in Fig. 20, where it can be seen that T35S does inhibit binding between G3.519 and AB20-4-HRP,

whereas an irrelevant peptide (G19C) does not. These results indicate that AB20-4 is a suitable anti-idiotype for use as a vaccine against HIV-1. The hybridoma cell line producing AB20-4 was placed on deposit at the American Type

Culture Collection (ATCC), Rockville, Maryland under provisions of the Budapest Treaty, under ATCC Accession No. . Equivalents

It should be understood that the terms, expressions, and descriptions herein are for clarification only and not limitation, and that the scope of protection is limited only by the claims which follows.

Claims

1. An antibody, or a fragment or polypeptide containing the complementarity determining region thereof, or an immunologically equivalent polypeptide, which generates the production of HIV-neutralizing antibodies in host organism.

2. An anti-idiotypic, paratope-specific antibody, or a fragment or polypeptide containing the complementarity determining region thereof, or an immunologically equivalent polypeptide, specific for the idiotope of an antibody which neutralizes HIV.

3. An anti-idiotypic antibody, fragment or polypeptide of Claim 2, which generates the production of

HIV-neutralizing antibodies in host organism.

4. An anti-idiotypic antibody, fragment or polypeptide of Claim 2, wherein the antibody which neutralizes HIV binds to gpl20 of HIV-1.

5. An anti-idiotypic antibody, fragment or polypeptide of Claim 4, wherein the antibody which neutralizes HIV binds to the region of gpl20 of HIV having the amino acid sequence RPNNNTRKRIRIQRGPGRAFVΗGK or a portion thereof.

6. An anti-idiotypic antibody, fragment or polypeptide of

Claim 5, wherein the antibody which neutralizes HIV is the monoclonal antibody BAT123.

7. An anti-idiotypic antibody, fragment or polypeptide of Claim 4, wherein the antibody which neutralizes HIV and binds to the region of gpl20 of HIV having the amino acid sequence

ΉTLPCRIKQIINMWQKVGKAMYAPPISGQIRCSS or a portion thereof.

8. An anti-idiotypic antibody, fragment or polypeptide of

Claim 7, wherein the antibody which neutralizes HIV is the monoclonal antibody G3.519.

9. An anti-idiotypic antibody of Claim 2 which is a human antibody, a chimeric animal/human antibody, a single chain antibody, or a human antibody fragment.

10. An anti-idiotypic antibody of Claim 9 which is a chimeric murine/human antibody.

11. An anti-idiotypic, paratope-specific monoclonal antibody specific for the idiotype of an gpl20-specific antibody which neutralizes different strains and different isolates of HIV-1.

12. A method of immunizing an organism against HIV, comprising administering an antigenic amount of an antibody, or a fragment or polypeptide containing the complementarity determining region thereof, which generates the production of HIV-neutralizing antibodies in the organism.

13. A method of immunizing an organism against HIV, comprising administering an antigenic amount of an anti-idiotypic, paratope-specific antibody, or a fragment or polypeptide containing the complementarity determining region thereof, or an immunologically equivalent polypeptide, specific for the idiotype of an antibody which neutralizes HIV.

14. A method of Claim 13, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for an the paratope of an antibody which binds to gpl20 of HIV-1.

15. A method of Claim 14, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for an the paratope of an antibody which binds to the region of gpl20 of HIV having the amino acid sequence

RPNNNTRKRIRIQRGPGRAFVΠGK or a portion thereof.

16. A method of Claim 15, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for the paratope of the monoclonal antibody BAT123.

17. A method of Claim 14, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for an the paratope of an antibody which binds to the region of gpl20 of HIV having the amino acid sequence

TITLPCRIKQIINMWQKVGKAMYAPPISGQIRCSS or a portion thereof.

18. A method of Claim 17, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for the paratope of the monoclonal antibody G3.519.

19. A method of Claim 13, wherein the anti-idiotypic antibody is a human antibody, a chimeric animal/human antibody, a single chain antibody, or a human antibody fragment.

20. A method of Claim 19, wherein the anti-idiotypic antibody is a chimeric murine/human antibody.

21. A method of immunizing an organism against HIV,

comprising administering an antigenic amount of an anti-idiotypic, paratope-specific monoclonal antibody specific for the idiotype of an gpl20-specific antibody which neutralizes different strains and different isolates of HIV-1.

22. A method of immunizing an organism against a particular pathogen, comprising admininstering an anti-idiotypic paratope-specific monoclonal antibody specific for the particular pathogen, and administering another vaccine which induces protective antibodies against said pathogen.

23. The method of claim 22 wherein the pathogen is HIV-1, and the monoclonal antibody is AB19-4 or AB20-4.

24. A cell which produces an antibody, or a fragment or polypeptide containing the complementarity determining region thereof, which generates the production of HIV-neutralizing antibodies in a host organism.

25. A cell which produces an anti-idiotypic, paratope-specific antibody, or a fragment or polypeptide containing the complementarity determining region thereof, or an immunologically equivalent polypeptide, specific for the idiotype of an antibody which neutralizes HIV.

26. A cell of Claim 25, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for the paratope of an antibody binds to gρl20 of HIV-1.

27. A cell of Claim 26, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for the paratope of an antibody which binds to the region of gpl20 of HIV having

the amino acid sequence RPNNNTRKRIRIQRGPGRAFVTIGK or a portion thereof.

28. A cell of Claim 27, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for the paratope of the monoclonal antibody BAT123.

29. A cell of Claim 26, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for the paratope of an antibody which binds to the region of gpl20 of HIV having the amino acid sequence TITLPCRIKQIINMWQKVGKAMYAPPISGQIRCSS or a portion thereof.

30. A cell of Claim 29, wherein the anti-idiotypic antibody, fragment or polypeptide is specific for the paratope of the monoclonal antibody AB20-4.

31. A cell of Claim 26, which is a transfected myeloma which produces a chimeric animal/human antibody.

32. A cell of Claim 31, wherein the chimeric antibody is a murine/human antibody.

33. A hybridoma cell line which produces an anti-idiotypic, paratope-specific monoclonal antibody specific for the idiotype of an gpl20-specific antibody which neutralizes different strains and different isolates of HIV-1.

34. The hybridoma cell line AB19-4, ATCC Accession No. HB

10315.

35. The hybridoma cell line AB20-4, ATCC Accession No.

36. A method of using an anti-idiotype antibody to enhance the immunogenicity of another vaccine, comprising administering the anti-idiotype at about the same time as the other vaccine.

37. The method of claim 36 wherein the other vaccine is a protein subunit vaccine, or a vaccine based on killed or inactivated virus.

38. The method of claim 37 wherein the anti-idiotype is specific for the idiotype of a gp-120 specific antibody which neutralizes different strains and isolates of HTV-1.