CA2073060A1

CA2073060A1 - Monoclonal antibody specific for non-immunodominant epitope of hiv proteins

Info

Publication number: CA2073060A1
Application number: CA002073060A
Authority: CA
Inventors: Paul J. Higgins; Charles F. Scott, Jr.
Original assignee: Individual
Current assignee: Repligen Corp
Priority date: 1990-01-16
Filing date: 1991-01-16
Publication date: 1991-07-17
Also published as: EP0511300A4; WO1991010742A1; EP0511300A1; JPH05506142A

Abstract

An antibody capable of recognizing a group common determinant and a non-immunodominant epitope of the envelope protein of HIV, wherein the binding of the antibody to the envelope protein is not blocked by serum from an HIV-infected patient.

Description

~YO 91/10742 2 ~ 7 3 ~ 6 0 PCI`/US91/00319 MONOCLONA:~J ANTIBODY SPECIFIC FOR NON~
IMMUNODOMINANT EPITOPE OF HIV PROTEINS

Backqround of the Invention This invention relates to antibodies specific for Human Immunodeficiency Virus (HIV).
HIV is the proposed causative agent of Acquired I~mune Deficiency Syndrome (AIDS). (Popovic et al., 1984, Science ~24:497). It is a pathogenic human retrovirus whose genome is capable of encoding at least six gene products. The env gene encodes a 160 kDa glycosylated protein (gpl60) that is processed by proteolysis into a 120 kD external glycoprotein (gpl20) and a 41 kD transmembrane protein tgp41). gpl20 is anchored to the virion by noncovalent interactions with gp41. gp120 and gp41 are present on the surface of both virion particles and virus-infected cells.
Different strains of HIV vary in the amino`acid sequences of proteins encoded by the viral genome, particularly in the amino acid sequence of the external envelope glycoprotein gpl20 (Starcich, 1986, Cell 45:637;
Hahn et al., 1986, Science 232:1548). Over its entire length, the gpl20 polypeptide sequence varies from one HIV variant to the next by approximately 20-25%. The extent o~ variation is not constant over the whole envelope protein. There is a pattern ~f conserved and variable regions, which suggests that the protein is divided into regions responsible for distinct functions.
A num~er of different regions have been identified; for example, the CD4 binding domain, the principal neutralizing determinant/ and cytotoxic T-cell recognition determinants.
~ he use of antibodies to target cells has previously been used in the treatment of cancer and other disorders. Zarling et al. (EPO 308 936) disclose ... ' ~BSTIT~Jl E S~IE~T : ' _._ WO 9}/10742 2 ~ 7 3 ~ ~ ) PCT/US91/00319 antibo~y heterocon~ugates specific for the gpl20 principle neutralizing domain which selectively kill HIV-infected cells. Pincus et al. (J. Immunol. (1989) 142;3070) describe antibody-toxin conjugates that also 5 recognize an immunodominant region of gpl20, and Till et ~
al. (Proc. Nat. Aca Sci., 1989, 86;1981) disclose anti- - ?
gp41-toxin conjugates.
Summary of the Invention The invention features an antibody capable of 10 recognizi~g a non-immunodominant epitope of the envelope protein of HIV, wherein the binding of the antibody to the envelope protein is not blocked by serum from an HIV-infected patient. As used herein, "antibody" refers to a whole antibody molecule, or to a fragment ox to a 15 modification of an antibody, e.g. a fragment of an antibody bay be the Fab2 fragment of the molecule, the Fabl fragment , or the heavy or light chain alone; and a modification, for example, may be a linear polypeptide molecule which includes both the heavy and light chains, as described in Huston et al., WO B8/09344 and Ladner et al., WO 88/01649. As used herein, "non-immunodominantll epitope means an amino acid ~equence within the natural confor~ation of a protein that is not significantly -immunogenic; i.e., does not e~icit an antibody response ~;
~5 in at least 75% of human patients. An antibody directed toward a non-im~unodominant region of the HIV envelope ;
protein can, according to the invention, bind that region due to the absence or low level of potentially competitive circulating antibodies. In contrast, an antibody directed toward~ an immunodominant region of the envelope protein will be partially or completely blocked from binding the targeted envelope protein region due to the presence of antibodies in the patient due to the patient's natural immune response to HIV infection. The ;
35 non-immunodominant envelope region may be`within the ~ ~

: ' SUE~STITVTE SHEET

.,` . ': :.:. , ' ,,.' ,': ' ' : ' ' '' :: ,, `` ` . ' ' ' ': :``: ' : ,` . :` ' . ' ': ' : ' ` .

~091/10742 2 0 7 3 ~ ~ ~ PCT/USg1/003~9 gpl20 or the gp41 portion of the Pnvelope protein. Non-immunodominance can be measured in vitro by binding the antibody to a target antigen in the presence of HIV-positive human serum; an antibody specific for a non-immunodominant epitope of the target antigen willdemonstrate a comparable binding efficiency to the target antigen in the presence or absence of HIV positive serum.
The binding efficiency of the antibody to the target antigen in the presence of HIV-positive serum is at least 80% of its binding efficiency in the absence of HIV-positive serum.
In preferred embodiments, the non-immunodominant epitope recognized by the antibody is group common. As used herein, "group common determinant" means an antigenic portion of a protein encoded by an H~V strain that is not specific for that strain only, but is present on at least one other HIV strain. Preferably, the antibody is capable of binding to the surface of HIV
envelope glycoprotein expressing cells; is capable of recognizing the region of the envelope protein between amino acid residues 473 and 759, inclusive, according to the numbering convention of Ratner et al., 1985, Nature 313:277; and is capable of recognizing that portion of gp120 contained within the 473 through 7S9 amino acid region, one example of such an antibody is that produced by cell line A.T.C.C. No. ~B 10321.
In other pre~erred embodiments, the antibody is covalently linked to a toxin to ~orm a conjugate, and the conjugate is capable o~ killing HIV-in~ected cells in the presence o~ human HIV~ serum; killing may occur via internalization of the antibody-toxin conjugate by HIV-infected cells. The conjugate may be made at the protein level by chemically linking the antibody and the toxin molecule or at the DNA level, by cloning the DNA sequence ' ... :~.

: .
8UE~STITUTE SHEET

.- . . : . . . , ` :. `.

W091~10742 207'~`~G ~ PCT/USgl/003t9 encoding the antibody and ligating it to a toxin-encoding DNA sequence, as described in Huston, s~pra.
As employed herein, the term "toxin" is used to include the commonly designated toxins such as poisonous lectins, ricin, abrin, modeccin, diphtheria toxin, pseudomonas exotoxin or preferably the toxic A chain portions thereof, as well as other toxic agents such as radioisotopes, cytotoxic and carcinostatic drugs.
"Toxin" may also refer to combinations of the various toxins, which can be coupled to one antibody molecule thereby accommodating variable cytotoxicity.
In another preferred embodiment of the inventiQn, the HIV-specific antibody is linked to a second antibody that is specific for an effector cell to form an antibody heteroconjugate (also known in the art as heteroaggregates or heteroantibodies). The anti-HIV
antibody of the heteroconjugate ~inds to an HIV-infected cell, i.e., the target cell to be killed, while the anti-effector antibody of the heteroconjugate binds to an effector cell such as those found within the peripheral blood lymphocyte (PBL) population, e.g., cytotoxic T
lymphocytes (also known as T cells), monocytes (in particular, macrophages), granulocytes, or large granular lymphocytes which inclùde cells with natural killer acti~ity or antibody-dependent cellular cytotoxic activity, with the result that the antibody components of the heteroconjugate bridge the effector and target cells and thus promote killing of the target cell by the cytotoxic e~fector cell.
HIV-$nfected patients can be treated by administering an amount of the antibody-toxin conjugate or an antibody heteroconjugate of the invention sufficient to kill HIV-infected cells. During active production of ViXU5, the viral envelope protein is expressed on the surface of infected cells. By SUBSTITUTE SHEEI-2 ~ 7 ~ a ~ ~ PCT/US91/00319 selectively killing the cells in which the virus is replicating using an antibody conjugate that reacts with virus-specific cell-surface antigens, the infectious cycle of the virus may be interrupted.
one advantage of an antibody or antibody-toxin conjugate sr an antibody heteroconjugate of the invention is the non-immunodominance of the HIV envelope glycoprotein epitope that the antibody is specific for.
A consequence of non-immunodominance, or failure of the human immune system to mount a detectable immune response to an epitope, is that there are few, if any, circulating antibodies to that epitope; i.e., to a gpl60 non-immunodominant epitope of the envelope protein in an non-immunodominant epitope of the envelope protein in an HIV-lS infected patient. Therefore, when an antibody-toxin conjugate or antibody heteroconjugate of the invention is used to treat an HI~-infected patient, it can selectively target HIV-infected cells without competing with circulating antibodies for the target non-immunodominant epitope.
Another advantage of certain antibodies, antibody-toxin conjugates, and antibody heteroconjugates of the invention are their ability to recognize a group common determinant of the ~IV envelope glycoprotein. Group common determinants are portions of the envelope polypeptide that are essentially invariant among di~ferent HIV strains. Therefore, an antibody capable of recognizing a group com~on determinant can recognize gpl60 from any strain o~ HIV. Treatment of an HIV-infe¢ted patient according to the in~antion will thus notbe limited to any one strain of HIV, but will inc}ude all strains to which the target determinant is common.
Other features and advantages of the invention will be apparent from the following description of the `
preferred embodiments thereof, and from the claims.
, .
~' S~lBSTlTlJl-E S~

WO91/10742 ~ 0 7 ~ Jvi PCT/US91/00319 pesçription of the Preferred Embodiment~ -we first briefly describe the drawings.
Drawinqs Fig. l is an schematic illustration of the gpl60 protein, showing the gpl20, gp41, pl21, and pENV9 regions.
Fig. 2 is a graph showing results of an ELISA in which the binding specificity of the lCl antibody to a target antigen was tested.
Figs. 3(a) - 3(d) are graphs showing results of ELISAs in which the binding speci~icity was determined ~or the lCl antibody or a control antibody in the presence of HIV positive serum.
Figs. 4(a) - 4(d) are graphs showing results of FACS analyses using the lCl antibody.
We now describe the preparation and use of antibodies of the invention.
Immunoqens Two immunogens were used to generate antibodies of the invention: gpl60 (Repligen Corp., Cambridge, MA) and the envelope protein fragment extending from amino acid residue 473 through residue 759, denoted pENV9 (Ivanoff et a}., U.S. Patent No. 4,861,707).
gp 160 was prepared for immunization by emulsification in complete Freund's adjuvant (CFA) according to standard techniques (Difco Labs, Grand Island, NY).
Production Q~ Monoclonal Antibodies Balb/cJ female mice (Jackson Labs., Bar Harbor, ME) were immunized intraperitoneally with 70~g per mouse o~ gpl60/CFA. ThQ mice were given a booster immunization of gpl60 in an emulsification in incomplete Freund's adjuvant three weeks later. Mice were ~led and the sera assayed for the presence o~ antibodies reactive with the immunogen. Mice showing a strong serological response Sl)BSTITUTE SHEET

.i. ~ .. ........ ~, - ., ~ . . . ..... . . .. . .

W09lt1074~ 2 ~ 7 ~ P~T/US9~/00319 were given a final booster immunization of response were given a final booster immunization of pENV9 in soluble form five weeks ~ollowing the first booster immunization and, 3 days later, spleen cells from these mice were fused at a ration of 5:1 with SP2/0 (A.T.C.C. No.
CRL8287, A.~.C.C. No. CRL8006) myeloma cells incapable of secreting both heavy and light immunoglobulin chains (Kearney et al., J. Immunol., 1979, 123:1548), ~y standard procedures based on the method of Kohler and -Milstein, Nature (1975) 256:495. ~ `!
Supernatants from hybridomas which appeared 10-21 days after fusion were screened for production of antibodies reactive with the pENV9 protein fragment.
Each well of a 96-well Costar flat-bottom microtiter plate was coated with pENV9 by placing a fifty microliter aliquot of a P~S solution containing the protein fragment at a final concentration of 2-10 ug/ml in each well The pENV9 solution was aspirated, the wells washed and replaced with PBS + 0.5~ BSA and incubated for 2 hours. Followi~g incubation, the wells were aspirated, washed, and 50 ul o~ hybridoma supernatant was added and incubated for 2 hours.
Following incubation, the wells were washed 3 times with PBS, and then incubated for 1 hr. with 50 ul of an appropriat~ dilution of goat anti-mouse immunoglobulin conjugated with horseradish peroxidase (HRP, Boehringer Mannheim, West Germany). The wells were wash~d again 3 times with PBS and 100 ul of 1 mM ABTS (2,2 azino-bis (3-ethyl benzth, azoline 6-sulfonic acid) in O.lM Na-Citrate, p~ 4.2, to which a 1:1000 d~lution of 30% H2O2had been added), the substrate for HRP, was added to detect bound antibody. After 30 minutes, absorbance was measured at OD410 on a Dynatech spectrophotometric aukoreader ~Virginia).
., SUE~STITUTE SHEET

W09l/10742 2 ~ r~ PCT/VS91/00319 Four hybridomas (lCl, lH9, 2D7, and 2H8) that tested positive for binding to the first immunizing antigen, gpl60, were tested for their ability to bind to the second immunizing antigen, pENV9, and to gpl60 from another different HIV strain. One hybridoma clone, designated lCl, which demonstrated reactivity with the pENV9 fragment by ELISA, produced antibody which was also capable of recognizing an envelope protein determinant ~ , that is present in more than one strain of HIV, i.e., a group common determinant, as described below.
The above four antibodies were tested in an ELISA
for bi~ding to either gpl60 from the ~IV-IIIB isolate or gpl60 from the HIV-RF isolate. The results, presented in Table I below, show that the pENV9-specific lCl antibody binds to gpl60 of both the IIIB and RF isolates, as well as to pENV9, whereas the remaining three antibodies do not recognize all three proteins.
Table I
clone gpl60~ gpl60RF pENV9 BSA
(control) lCl + ~ +
lH9 + _ 2D7 + +
2H8 ~ + O - -The isotype of the lCl clone was determined by the ELISA method to be IgG2~ using goat-anti-mouse HRP (Zymed Labs, San Francisco, CA) preparations which correspond to each of the major immunoglobulin isotypes. ~he lCl clone was subcloned and rescreened ~or the ability to bind to the Antigens described above. The lCl subclone was expanded by intraperitoneal in~ection into pristane primed Balb/c mice. Ascites fluid was recovered from the mice and the antibody was purified by Protein A affinity chromatography, as described below.
Ampli~ication and Purification of Monoclonal Antibodies 5UBSTIT~JTE SHEFT' ; ~ ' ' " ' '" "' ' ' ' '' . ' ,'; ' , . "" '` " ' ', . ' ' , ''.. ~ ,, ,`', ' ', "" ', " ` ' ,, ,, :, ,, ' WO 91/10742 2 ~ 7 3 ~3 ~ ~ Pcr/usg1/0031g g Purified lC1 antibody was prepared by injecting a hybridoma subclone that repeatedly tested positive by ELISA intraperitoneally into pristane-primed syngeneic mice. The ascites which developed were recovered two to three weeks after injection, and the antibody was purified as follows, and then dialyzed against PBS.
Ascites fluid containing IgG2a lC1 antibody was diluted five-fold in 0.1 M Tris/3 M NaCl pH 8.9, bound to a Protein-A-Sepharose affinity column equilibrated with the same buffer, and then eluted from the column with Q.15 M NaCl, 0.1 M acetic acid, Ph 3Ø Following elution, the antibody was immediately neutralized by the addition of 1 M Na2HC03.
Bindinq SPecificit~ of the lC1 AntibodY
The lC1 antibody was tested in the ELISA ~or binding with pENV9, gp120, and pl21, in order to map the epitope to which lC1 binds. pl21 ~Chang et al~ U.S.
Patent No. 4,774,175) is an 83 amino acid protein fragment spanning amino acids from approximately 566-648 within the gp41 portion o~ qpl60, and is completely contained within the pENV9 se~uence. (These regions of gpl60 are schematically illustrated in Fig. 1.) pl21 includes a major immunodominant epitope of the gp41 protein ~Chang et al., U.S. Patent No. 4,724,175 and Wang et al., 1986, ~E~. Nat. Aca. Sci. 83:6159). Fig. 2 shows the results of the ELISA assay. These results demonstrate lC1 binds specifically to both pENV9 and gpl20, but does not bind pl21. Therefore, the lCl antibody binds to a region of pENV9 that is also present 30 in gpl20, but that is not contained within the pl21 `
portion. In a control experiment, an antibody specific for gp41 (Epitope, Inc., ~eaverton, OR) was tested ~or binding to either pENV9, pl21, or gp120, showing that the anti-gp41 antibody bound to pENV9 and pl21, but did not bind to gpl20, as expected.

~U~STITlJTE SHEET

WO91/10742 PCT/US91/~319 2~7~

Bindinq of the lCl Antibodv in the ;:
Presence of Human Serum A therapeutically useful exogenous antibody specific for the HIV envelope protein must be able to bind HIV, or HIV-infected cells expressing the envelope protein, in the presence of competing circulating antibodies.
The lC1 antibody was tested for its ability to bind to target antigen in the presence of serum from an ~
lo HIV-infected patient, and compared with a control ~-antibody known to bind to an immunodominant region of ;
gp41. Figs. 3(a) - 3(d3 show the results of ELISAs in which microtiter wells were coated with the capture ~ -antigen. The lCl or control antibodies were then added in 50 ul of (1) undiluted HIV-negative serum, (2) undiluted HIV-positive serum, or (3) 0.5% BSA. After 2 hrs., the wells were washed and secondary antibody (sheep anti-mouse-H~P) which did not cross-react with human Ig was added. After 1 hr., the secondary antibody was removed, the wells were washed, ABTS was added, and the OD410 was measured after 30 min.
In ~ig. 3(a), the lC1 antibody was assayed for binding to the capture antigen, pENV9. ~he results demonstrate that lC1 bound with almost egual efficiency to pENV9 in the presence of either HIV-positive serum or HIV-negative serum (Fig. 3a). Although the lC1 antibody bo~nd with the highest efficiency to pENV9 in the presence o~ HIV negative serum at antibody concentrations above 0.01 ~g/ml, the binding o~ this antibody to pENV9 in the presQnce of HIV positive serum was about 88~ as efficient as binding to pENV9 in the presence of HIV
negative serum at an antibody concentration of 1 ~g~ml, and over 95% as efficient at 10 ~g/ml. Similar results were obtained when lCl binding was tested in the presence o~ HIV positive sera from four other patients. These SU8STITUTE SHE~ET

WO 91tlO742 2 0 7 3 ~ 5 ~ PCT/US~1/G0319 results demonstrate that, if there are pENV9-specific antibodies present in HIV positive serum, they are present in low enough titers or have a weak binding affinity or react with different pENV9 epitope so that they do not significantly interfere with the binding of the lC1 antibody to pENV9. Thus, lC1 is potentially useful as a therapeutic agent.
In Fig. 3(b), khe ability of the lCl antibody to bind the pl21 protein was tested. The lC1 antibody did not bind pl21 at all. (~he minimal reactivity observed with pl21 at lC~ concentrations above O.l~tml in 0.5~ BSA
was probably due to nonspecific binding.) Two additional control experiments were performed using a control antibody, anti-gp41 (Epitope, Inc.) and tow target antigens, pl21 and pENV9. Results present in Figs. 3~c) and 3~d) show that anti-gp41 binds to both target antigens, but binds more ef~iciently to pl21 than to pENV9, at antibody concentration~ above 0.1 ~g/ml.
The results also show that, at these same concentrations of antibody, the binding of anti-gp41 antibody is partially blocked by HIV positive serum. These results indicate that there are antibodies specific for pl21 and p~NV9 present in ~IV positive serum that are capable of partially blocking the binding o~ the anti-gp41 antibody.
Bindina of lCl An~ibody_to Cells Whether the lC1 monoclonal antibody binds to the surface of cells expressing the HIV envelope glycoprotein was determined by indirect immunofluorescence and analysis by FACs (Fluorescence Activated Cell Sorter, ~ethods in Enzymology, 1984, Parks et al., 108:197), as ~ollows.
The lCl antibody was bound to either CV1 cells ~A.T.C.C. No. CCL70) infected with a Vaccinia Virus reaombinant containing the HIV env gene, which express both gpl20 and gp41 on their surface (CV1-Env), or CV1 SUBSTITUTE SHEET
: ;~

WO91~10742 PCT/US91/00319 2 ~ 7 ~

cells infected with a Vaccinia Virus recombinant that does not contain the HIV env gene (CVl-Lac), as a negative control. Construction of a recombinant capable of expressing the full-length HIV envelope gene is described in EPO 243 029, hereby incorporated by reference. Figs. 4~a) - 4(d) show these results.
In Fig. 4ta), the lCl antibody was bound to CVl-Env cells. If Fig. 4(a) is superimposed on Fig. 4(b), which shows a FACS profile for CVl-Lac cells incubated with the lCl antibody, there is a rightward ~hift (i.e., an increase) in fluorescence intensity in the CVl-Env cells compared to the CVl-lac cells, indicating that the lCl antibody binds HIV envelope glycoprotein expressing cells significantly better than the cells not expressing the enYelope protein. This result is significant because it demonstrates that lCl binds to the target antigen in its native state, and suggests that cells expressing HIV
envelope glycoprotein may be specific targets for an immunotoxin conjugate composed of lCl linked to a toxin.
Figs. 4(c) and 4(d) are controls in which the lCl antibody was bound to uninfected CVl cells and in which buffer alone was bound to CVl-Env Cells, respectively.
(The background fluorescence that is apparent in the FACS
analyses of CVl-Env and CVl-Lac cells is probably due to alterations in the cell membrane resulting from expression of foreign viral proteins.) Pre~aration of ~ntibod~ Coniu~ates o~ Heterocon~u~ates $he antibodies may be conjugated to cytotoxic agents and used as immunotoxins (as described in, e.g., Vitetta et al., 1987, Science 238: 1098), or incorporated onto the surface of liposomes containing anti-HIV drugs or toxins to specifically target such drugs or toxins to HIV-infected cells. As employed herein, the term immunotoxin refers to a conjugate of an antibody with one or more toxins. Where co~binations of . . .

S~JB8T~ E 9HE~Ef .

.. . .. .. . .. . .. .

WO~1/10742 ~ 0 7 3 ~ ~ ~ PCT/USgl/00319 various toxins are coupled to one antibody molecule, coupling may occur by different chemical mechanisms; for example, covalent binding, affinity binding, intercalation, coordinate binding and complexation. The ~`
preferred coupling of the antibody to the toxin is, however, covalent binding either by chemical or genetic fusions.
In a pre~erred embodiment, the immunotoxin comprises an antibody reactive with a non-immunodominant, group common epitope of the HIY envelope protein linked to the exotoxin form Pseudomonas aeruqinosa. Pseudomonas exotoxin (PE) is particularly preferably to other toxins because large amounts are easily prepared, because humans do not usually have neutralizing antibodies against it, and because it does not need to be separated into subunits be~ore being conjugated. PE is an extremely active monomeric protein (~olecular weight 66kD), secreted by P. aeru~inosa which inhi~its protein synthesis in eukaryotic cells. A preferred form of`PE is the truncated molecule, designated PE40, from which the cellular binding domain has been removed (Pastan et al.
EP Publication No. 0 261 671). PE~0 can be linked to an antibody o~ the invention by chemical coupling, for exa~ple, using the heterobifunctional cross-linker SPDP `
(N-succinimidyl-3-~2-pyridyldithiol~ propionate) Sigma, St. Louis, M0) ~Pastan et al, 1986, Cell 47:641), or by genetic fusion ~Cbaudhary et al.,) 1989, Nature 339:394).
Where an antibody heteroconjugate is preferred, any suitable method of conjugation of the antibodies may be used; for example, preferred method involves cross-linking the antibodies using the cross-linker SPDP
according to the method of Karpovsky et al. (1984, J. ~`
Exp. Med. 160:1686). Following cross-linking, the heteroconjugates are separated from fee antibody by size 35 exclusion chromatography. `;;
,:

SURSTITUTE SHEET

wo 91~ln742 ~ o 7 ~ ~ ~J~ PCT/~S91/00319 An antibody-toxin conjugate or heteroconjugate of the invention can be tested for target specificity and efficiency by killing by incubating the conjugate with HIV chronically-in~ected cells that express HIV envelope glycoprotein or with uninfected cells, and pulsing the cells with 3H-Thymidine of 14C-Leucine. Toxicity can be measured by a decrease in cell division or protein synthesis in the infected cells relative to the uninfected control cells. The efficiency of cell killing can be calculated using a clonogenic assay, in which infected cells are incubated with a conjugate of the invention, plated by limiting dilution (Tarwell, 1981, J.
Immunol. 126.1614), and the number of surviving cells is compared to identically treated uninfected cells.
Use In a typical treatment employing antibodies of the invention as immunotoxins, the antibody (which binds to a protein that is expressed only in HIV-infected cells) is conjugated to a toxin (e.g. pseudomonas exotoxin~ that i5 toxic to the HIV-infected cells ~and to non-infected cells as well). By coupling the cytotoxic agent to the antibody, a high level of toxic efficacy can be achieved specifically against the targe~ cell with a ~arkedly low level of non-specific toxicity. The use of the toxic agent is possible because the antibody to which the agent is coupled will carry the agent specifically to the target (in this case, HIV-in~ected cells), thereby sparing non-infected cells from the toxin. Techniques that may be employed to con~ugate antibodies to cytotoxic agents are described in detail in ~itetta et al, su~ra, and in European Patent Applica~ion No. 279,668, published August 24, 1988.
The antibodies of the invention can be incorporated into conventional pharmaceutical formulations ~or use in treating individuals that are Sl.3BSTIT~JTE SHEET

` . . . . ., -: .. , :

WO91/10742 2 ~ ~ 3 ~ ~ Q PCT/US9l/00319 infected with HIV. In addition, such formulations may comprise pharmaceutically-acceptable carriers, diluents, salts and other materials well-known in the art.
Isotonic saline, sterile water, 10% maltose, human serum albumin, glycine or other pharmaceutically-acceptable material may be used as diluent , carriers or solvents in preparing the pharmaceutical formulations comprlsing antibodies of the invention.
~he phaxmaceutical compositions may be in a variety of dosage forms which include solid, semi-solid and liquid forms as powders, pills, tablets, liquid solutions or su~pension, suppositories, polymeric microcapsule, liposomes or injectable or infusible substances. The pharmaceutical formulations may be administered using conventional methods which include, but are not limited to, intravenous, oral, subcutaneous, intraperitoneal or intralymphatic. In addition, the antibody, i~munotoxin or heterocon~ugates of the invention may be administered in conjunction with other treatments to augment the e~fectiveness of the treatment.
Other E~bodiments Other embodiments are within the following claims.
For example, since, for the most part, monoclonal antibodies are produced in species other than humans, they are often i D unogenic to humans. In order to successfully use these monoclonal antibodies in the treatment of humans, it may be necessary to create a chimeric antibody molecule wherein the portion o~ the polypeptide involved with ligand binding ~the variable region) is derived ~rom one specie~ and the portion involved with providing structural stability and other biological functions (the constant region) is derived ;~
from a human antibody. Methods for producing chimeric `
antibodies in which the variable domain is derived from one host and the constant domain is derived from a second SUE~STITUTE SHEET

- . :, ......... ... . ~` : . . . :
~ ;: ; ? :::: ~ ` :
.... . . . ..

wo 91~10742 ~ 0 7 ~ ~ J~ PCT/US91/~319 host are disclosed by ~euberger et al. (Wo Publication No. 86/01533) and Morrison et al. (EP Publication No. o 173 494), hereby incorporated by reference.
An alternative method, in which an antibody is produced by replacing only the complementarity determining regions ~CDRs) of the variable region with the CDRs ~rom an immunoglobulin of the desired antigenic specificity, is described by Winter (GB Publication No.) 2 188 638). For example, the CDRs of a pENV9-specific, murine monoclonal antibody which recognizes a group common determinant and a nonimmunodominant domain can be grafted onto the framework of a human antibody by recombinant DNA techniques. This arrangement is ;~
particularly beneficial for use in the therapeutic applications of monoclonal antibodies.
DePosit Cell line lC1-lH5 was deposited in the American Type Culture Collection on January 10, 1990, and assigned Accession Number HB 10321.
Applicants' assignee, Repligen Corporation, represents that the A.T.C.C. is a depository affording permanence of the deposit and ready accessibility thereto by the public i~ a patent is granted. All restrictions on the availability to the public of the material so deposited will be irrevocably removed upon and granting of a patent. ~he material will be available during the pendency of the patent application to one determined by the Commissioner to be entitled thereto under 37 C.~.R.
1.15 and 35 U.S.C. 122. The deposited material will be maintained with all the care necessary to k~ap it viable and uncontaminated for a period of at least five years after the most recent request for the furnishing of a sample of the deposited microorganism, and in any case, ~or a period of at least thirty ~30) years after the date of deposit or for the enforc-able life of the patent, . i . " : .

, W~91/10742 PCT/US91/0~319 2~7~Q6~

whichever period is longer. Applicants' assignee acknowledges its duty to replace the deposit should the depository by unable to furnish a sample when requested due to the condition of the deposit. A copy of the 5 A . T ~C.C. Budapest Treaty deposit receipt will be furnished upon request.

BS~lT~ S~EE~ ~

.. .. ..... . , . .................. ... . .; .. ~ .. . .... ~ . ~ .

::: .

Claims

1. An antibody capable of recognizing non-immunodominant epitope of the envelope protein of HIV, wherein the binding of said antibody to said envelope protein is not blocked by serum from an HIV-infected patient.

2. The antibody of claim 1, wherein said epitope is group common.

3. The antibody of claim 2 wherein said antibody is capable of binding to the surface of HIV envelope glycoprotein-expressing cells.

4. The antibody of claim 3 wherein said antibody recognizes the region of the HIV envelope protein between amino acid residues 473 and 759, inclusive.

5. The antibody of claim 4 wherein said antibody recognizes a portion of gp120 within said region.

6. The antibody of claim 5, said antibody being produced by cell line having A.T.C.C. No. HB 10321.

7. The antibody of any claims 1-6 further comprising a toxin covalently linked to said antibody to form a conjugate which is capable of killing HIV-infected cells in the presence of human HIV+ serum.

8. The antibody-toxin conjugate of claim 7 wherein said conjugate is capable of being internalized by HIV-infected cells.

9. The antibody of any claim 1-6 further comprising a second antibody covalently linked to said antibody to form a heteroconjugate.

10. A method of treating HIV-infected patients, comprising administering an amount of the antibody-toxin conjugate of claim 6 or the antibody heteroconjugate of claim 8 sufficient to kill HIV-infected cells.