"HIV-RELATED ANTIGENS AND ANTIBODIES"

CROSS-REFERENCE TO RELATED APPLICATIONS

The present invention is a continuation-in- part of co-pending U.S. Patent Application, filed August 22, 1988; which in turn is a continuation-in-part of co- pending U.S. Patent Application Serial No. 135,337; filed December 21, 1987.

BACKGROUND

The present invention relates generally to immunological methods and materials and more particu- larly to peptides sharing amino acid sequence homology with proteins of Human Immunodeficiency Virus type 1 ("HIV-1") and with the human neurotrophic factor, neuro- leu in ("NLK"), to antibodies specific for and related to such peptides, and to immunotherapeutic and diagnostic procedures involving such peptides and anti¬ bodies.

Human Immunodeficiency Virus Type 1 has been designated as the viral causative agent of acquired immunodeficiency syndrome ("AIDS") and AIDS-related complex ("ARC"). Numerous review publications address the etiology of the disease state, the pathogenesis of. the viral causative agent, and the prospects for rapid development of AIDS and ARC diagnostic and/or thera¬ peutic agents. See, e.g., Fauci, Proc. Nat'l. Acad. Sci. (USA), 3, 9278-9283 (1986) and Ho et al.. New Eng. Jour. Med., 317, 278-286 (1987). _.

A variety of strategies are currently under¬ going evaluation in an attempt to develop AIDS and ARC protective and/or therapeutic vaccination procedures. In addition to the traditional approach of developing • killed and attenuated viral vaccines, substantial

research effort has been focused on recombinant' vaccinia vaccine and protein subunit vaccine stratff s, princi¬ pally based on the immune "presentation" of 5ιe core (gag) and envelope (env) glycoproteins of HIV-1. Thus, vaccinia—based strategies envision vaccination with recombinant vaccinia strains engineered to encode the HIV-1 envelope glycoprotein, gpl60, hopefully thereby presenting the glycoprotein as an immunogen to the indi¬ viduals vaccinated. In a like manner, based on demon- strations of capacity to provoke formation of antibodies capable of neutralizing ii vitro viral infectivity, immunopurified preparations of a gpl60 subunit, gpl20, have been suggested as protective vaccine components. See, Robey et al., Proc. Nat'l. Acad. Sci (USA), 83, 7023-7027 (1986); and Matthews et al., Proc. Nat'l. Acad. Sci (USA), E$3_, 9709-9713 (1986). Recombinant methods have been applied to secure relatively large scale production of various forms of HIV-1 gp!60, as well as ^' its lower molecular weight subunits, gp41 and gpl20 [see, e.g., Lasky et al., Science, 233, 209-212 (1986); Weiss et al., Nature, 324, 572-578 (1986); and Lasky et al., Cell, 5_0, 975-985 (1987)] and even "sub- subunit" polypeptides such as the fusion protein com¬ prising a 180 amino acid fragment of gpl20 known as "PB1" [see, Putney et al., Science, 234, 1392-1395 (1986)3.

Among the substantial drawbacks attending attempts to develop protective subunit vaccines based on, e.g., HIV-1 env glycoproteins or fragments thereof is the existence of significant intertypic heterogeneity in amino acid sequence and the corresponding hetero¬ geneity,of immune responses to administration of pro¬ teins derived from differing HIV-1 isolates. For example, monoclonal antibodies raised against or speci- fically immunoreactive with gpl20 derived from one HIV-1 isolate may be immunologically significant in terms of

in vitro neutralization capacity with respect to the specific isolate, but may not at all recognize gpl20 of other subtypes. See, e.g., Fung et al., Bio/Technology, j>, 940-946 (1987). Indeed, in gpl20, a pattern of alternating "variable" and "constant" regions has been observed. See, Starcich et al.. Cell, 45, 637-648 (1986) and Willey et al., P.N.A.S. (USA), 3_, 5038-5042 (1986). In the two "constant" domains of gpl20, sequence homology among various isolates is on the order of 80 percent, while in the "variable" regions conserva¬ tion is at the level of 20-30 percent. Moreover, vaccination strategies based on HIV-1 protein subunits must take into account such factors as ease of manufac¬ ture and purification in quantity. Natural isolates and recombinant products should be free of foreign DNA and contaminating foreign proteins. Finally, because little is known concerning the effects of HIV-1 proteins per se in infected hosts, substantial attention must be taken to avoid, e.g., adverse non-immunological effects which might attend administration of viral subunits.

Of interest to the background of the present invention are the disclosures of Kennedy et al.. Science, 231, 1556-1559 (1986); Chanh et al., EMBO Jour. , 5(11), 3065-3071 (1986); Kennedy et al., J. Biol. Chem., 262(12), 5769-5774 (1987); Ho et al., J. Virol., 61(6) , 2024-2028 (1987); European Patent Application No. 0 227 169, published January 7, 1987; European Patent Application No. 0 230 222, published July 29, 1987; European Patent Application No. 0 231 914, published August 12, 1987; and PCT International Publication No. WO87/02775, published May 7, 1987. Briefly summarized, these publications and published patent applications relate to immunochemical reagents projected for use in diagnosis and vaccination for AIDS and ARC and, more specifically, describe synthetic peptides modeled after, i.e., sharing amino acid sequence homology with, con-

tinuousgsequences of amino acids within the envelope glyccJl libems of HIV-1. As an example. Ho et al., J.Virolfl 61(6) , 2024-2028 (1986) describes the prepara¬ tion and testing of anti-sera to 87 "overlapping" pep- tides spanning the env glycoprotein, gpl60. Based on neutralization titers, peptides corresponding to amino acids 298-314, 458-484 and 503-532 (within the gpl20 portion of the gpl60 sequence) as well as amino acids 728-752 and 616-632 (within the gp41 portion of the sequence) were highlighted for further study. All five synthetic peptides correspond to extremely hydrophilic regions of the HIV-1 envelope and the latter three correspond to regions which are reported to be highly conserved in HTLV-III and LAV strains of HIV-1. Of particular interest to the background of the present invention are the results of recent work in the characterization and molecular cloning and express¬ ing of the human neurotrophic factor, neuroleukin. See, Gurney et al.. Science, 234, 566-574 (1986); Gurney et al.. Science, 234, 574-581 (1986); and Lee et al..

Science, 237, 1047-1051 (1987). See, also, DNA sequence GENBANK Accession No. K03515:HUM-NLK, May, 1987. Briefly summarized, these publications disclose that the deduced amino acid sequence of neuroleukin is partially homologous to the amino acid sequence of the gpl20 env glycoprotein of HIV-1 in a region which is highly "conserved" among known HIV-1 subtypes. Also disclosed is the determination that gpl20 is, per se, capable of in vitro inhibition of the action of neuroleukin on neuroleukin-dependent cultured neuron cells, suggesting that the antigen may display neurotoxic effects in an infected host by directly suppressing neuronal responses to neurotrophic factors. This determination is noted to provide valuable insights into the pathogenesis of the AIDS dementia complex.

Despite intensive research and development efforts expended on methods and materials for diagnosis and treatment of AIDS and ARC, there continues to exist a need for new and useful immunodiagnostic and- immuno- therapeutic reagents. To the extent that synthetic "sub-subunit" peptides provide the basis for such reagents, they should be readily assembled and admini¬ stered (e.g., in combination with adjuvants and "carrier" proteins), should be based on homology to "constant" rather than "variable" regions or domains in HIV-1 proteins, should be non-neurotoxic so as to avoid deleterious neurological side effects upon administra¬ tion, and should be capable of provoking immune res¬ ponses operative in neutralizing viral infectivity.

BRIEF SUMMARY

Provided by the present invention are novel immunochemical reagents including non-neurotoxic pep- tides and antibodies specifically immunoreactive there¬ with. Peptides of the invention are characterized by the absence of capacity for ^n vitro inhibition of neuroleukin action on neuroleukin-dependent neurons. The peptides are further characterized by sufficient amino acid sequence homology to residues 254 to 280 [according to the numbering system of Ratner et al., Nature, 313, 277-284 (1985)] of glycoprotein gpl20 of HIV-IIIB to allow for the capacity to stimulate forma¬ tion of antibodies of the invention which, in turn, are characterized by immunoreactivity with gpl20 and/or gpl60 isolates of HIV-IIIB and at least one other HIV-1 subtype and by the capacity for neutralization of infec¬ tivity, ii vitro, of HIV-IIIB and at least one other HIV-1 subtype [e.g., HIV-IIIRF referred to as HAT-3 and sequenced by Starcich et al.. Cell, 45, 637-648

(1986)]. Peptides of the invention are thus seen to

include amino acids in sequence which correspond to at least one antigenic determinant shared by gpl20 of HIV-IIIB and at least one other HIV-1 subtype. Preferred antibodies of the invention, including mono- clonal and polyclonal antibodies, neutralize HIV-1 infectivity (with respect to, e.g., H9 cells in culture) without substantially diminishing the capacity for association (binding) of viral particles to host cell surfaces. Antibodies of the invention also comprehend monoclonal anti-idiotypic antibodies raised, e.g., against variable, antigen-combining regions ("paratopes") of neutralizing monoclonal antibodies of the invention.

Chimeric antibodies, and fragments thereof, and especially bi-specific antibodies, are also products within the contemplation of the present invention, as are antibody-related products produced in microbial hosts, (e.g., procaryotic and eucaryotic cells in culture) which hosts are transformed or transfected with DNA sequences encoding the desired polypeptide products.

As one example, with structural information in hand concerning the idiotypic regions of antibodies of the invention, it becomes possible to employ procaryotic and eucaryotic hosts such as E. coli, yeast, insect, and mammalian cells in culture to produce useful antibody fragments (such as fab' and f(ab')2 fragments). Moreover, it is within the contemplation of the invention that chimeric antibodies (e.g. mouse/human antibodies) may be prepared using transformed mouse myeloma cells or hybridoma cells (especially heavy chain deletion mutant cells) as production hosts. Hybrid hybridoma cell producing bi-specific antibodies having diagnostic and therapeutic uses are contemplated. Presently preferred antibodies of the invention are monoclonal antibodies of the IgG 3 subtype and in particular that monoclonal antibody produced by

cell line JK112, deposited August 19, 1988 with the American Type Culture Collection, Rockville, Maryland under accession No. A.T.C.C. HB9799.

Provided by the invention are novel binary immunological compositions comprising HIV-1 viral par¬ ticles bound to antibodies of the invention, as well as ternary immunological compositions comprising antibodies of the invention bound to HIV-1 viral particles which are, in turn, bound to susceptible host cell surfaces. Presently preferred peptides of the invention include the peptide having the following sequence,

CTHGIRPWSTQLLLNGSLAE which sequence possesses total homology with the con¬ tinuous sequence of amino acid residues spanning posi- tions 254 through 274 of gpl20 of both HIV-IIIB and

HIV-IIIRF and differs from the corresponding sequence of residues of gpl20 of HIV-ARV-2 by a single residue. Also presently preferred is the peptide having the sequence, STQLLLNGSLAEEEWIRC which possesses total homology to residues 263 through 280 within gpl20 of all three of the above-mentioned HIV-1 isolates and includes a carboxy terminal cysteine residue. Peptides of the invention may optionally be provided with a variety of additional amino acid residues at their amino or carboxy terminal or at inter¬ mediate positions within the sequence in order to enhance their immunogenicity and immunochemical reagent potential. As one example, one or more tyrosine resi- dues may be provided for association with a radiolabel substance such as ι ¹²⁵ with the peptide. Similarly, one or more -reactive terminal or intermediate cysteine resi¬ dues may be provided to facilitate association of the peptide with carrier proteins commonly employed in the development of peptide vaccine _^ compositions. Moreover, polyamino acids such as poly-L-glutamic acid may be

linked to peptides of the invention to enhance anti- genicity. In general, however, care should be exercised to avoid incorporation of residues adversely affecting the immunogenicity of antigenic sites within the pep- tides.

Monoclonal antibodies presently preferred are those reactive with a peptide having the sequence:

S T Q L L L N G S L E. Vaccine compositions of the invention comprise immunologically effective amounts of one or more pep¬ tides or anti-idiotypic antibodies of the invention in combination with a diluent, adjuvant or carrier and vaccination methods of the invention involve administration of such compositions to a host susceptible to HIV-1 infection. Also within the contemplation of the invention are passive immunization compositions and vaccination methods based on admini¬ stration of antibodies of the invention.

Peptides and corresponding antibodies of the invention are useful for immunodiagnostic procedures (e.g., ELISA*s, RIA's, and the like) for the detection of HIV-1 infection and/or for monitoring progress of vaccination treatment.

In another of its aspects, the present invention provides, as novel immunochemical reagents, peptides (and antibodies related thereto) whose sequences are based on those regions of amino acids within residues 234-300 of HIV-III B, which presumptively form loops separating (i.e., intermediate between) beta-sheet forming residues within the normal secondary and tertiary conformation of that region.

Numerous aspects and advantages of the present invention will be apparent upon consideration of the following detailed description of illustrative embodi- ments thereof, reference being made to the following drawing.

BRIEF DESCRIPTION OF THE DRAWING

Figure 1 shows the predicted beta-sheet secondary structure for HLA-A A-2 domain, HCMV-H301 gene, HLA-DR B-2 domain, and HIV gpl20.

Figure 2 shematically illustrates a typical anti-parallel arrangement of the beta-sheets of a constant Ig domain. Figure 3 illustrates predicted strands of beta-sheets, B-2 through B-6, for the second conserved region of HIV-III B.

Figure 4 shows the alignment of contact residues between the HLA-A A-3 domain strands beta-4 and beta-5 and B2"-m with the homologous domains of HCMV- H301, HLA-DR, and gpl20.

DETAILED DESCRIPTION

The following Examples 1 through 8 are intended to be illustrative of practice of the present invention rather than limiting thereon. Example 1 relates to preparation of immunologically active pep¬ tides and reagents of the invention. Example 2 relates to generation of antibodies specific for peptides of the invention. Example 3 relates to screening procedures permitting characterization of immunological properties of antibodies of the invention. Example 4 relates to screening of peptides of the invention for potential neurotoxic effects as determined according to the cul¬ tured chick sensory neuron survival assay_as described in Lee et al., supra. Example 5 relates to the preparation of monoclonal antibodies. Example 6 relates to m vitro neutralization of HIV-1 using JK112 monoclonal antibody. Example 7 relates to the mapping of the amino acid sequence recognized by the JK112

monoclonal antibody. Example 8 relates to the predicted three-dimensional conformation of the antigenic epitopes of HIV-III B gpl20 and the peptides derived therefrom.

EXAMPLE 1

Table I below sets out partial amino acid sequence of glycoprotein gpl20 of HIV-IIIB, HIV-IIIRF and HIV-ARV-2 isolates in the so-called "second con¬ served region" as reported in Ratner et al., supra; Wain-Hobson, Cell, 40, 9-17 (1985); Sanchez-Pescador et al., Science, 227, 484-492 (1985); and Starcich et al., supra, using the numbering system of Ratner et al., supra. A dash indicates identity with the first-listed residue. Set out immediately following the HIV-1 sequences is the sequence of a "corresponding" sequence of amino acids in human neuroleukin. Asterisks desig¬ nate sequence homology with the HIV-1 proteins. Finally, Table I sets out the sequence of six synthetic peptides — three 19-mers and a 21-mers based on the HIV-1 sequences and two 20-mers based on NLK sequences.

2 2 2 3 4 8 5 0 9

HIV-IIIB CNNKTFNGTGPCTNVSTVQCTHGIRPWSTQLLLNGSLAEEEWIRSANFDTNAK

HIV-IIIRF

HIV-ARV-2 K 1 D—T—V-

** ** * ** **

3 4 4 4 4 4 4 9 0 1 2 3 4 4

4 0 0 0 0 0 8

NLK Q IHQGTKMIPCDFLIPVQTQHPIRKGLHHKILLANFLAQTEALMRGKSTEEARK

Peptide C19Q CNNKTFNGTGPCTNVSTVQ Peptide T19V TGPCTNVSTVQCTHGIRPV Peptide C21E CTHGIRPWSTQLLLNGSLAE Peptide S19C STQLLLNGSLAEEEWIRC Peptide M20L MIPCDFLIPVQTQHPIRKGL Peptide F20H FDCPIMQTQVPILLGKRIPH

As is readily determinable from Table I, con¬ struction of peptide C19Q is based on homology to resi¬ dues 236 through 253 of gpl20 of HIV-IIIB; peptide T19V is based on homology to residues 243 through 261; pep- tide C21E is based on homology to residues 254 through 274; and peptide S19C is based on homology to residues 263 through 280 and includes an additional carboxy ter¬ minal cysteine residue. It is also apparent that pep¬ tide M20L is based on homology to residues 401 through 421 of neuroleukin and that peptide F20H preserved the amino acid composition, but not the sequence, of peptide M20L.

The above-described peptides designated C19Q, T19V, C21E, S19C, M20L and F20H were commercially pre- pared by chemical synthesis using t-Boc protected amino acids according to standard procedures. The purity of peptides was 94-96% by RPHPLC and their composition was confirmed by amino acid analysis.

Conjugates of the peptides with a carrier protein are readily prepared by suitable means well known in the art. Suitable carriers include keyhole limpet hemocyanin (KLH) and bovine serum albumin (BSA) . Coupling of cysteine residue-containing peptides may be accomplished through use of m-maleimidobenzoic acid N-hydroxy succinimide ester (MBS) or the

N-maleimido-6-aminocaproyl ester of l-hydroxy-2-nitro-4- benzenesulfonic acid (MSAC) . Alternately, coupling to primary amino groups through glutaraldehyde may be employed. In a typical procedure for coupling to BSA, the carrier is first reacted with MBS (25 mg/ml in for¬ mamide) .at a ratio of 0.125 mg MBS per mg BSA in 50 mM sodium phosphate buffer (pH 7.5) for 30 min. at room temperature. Removal of unreacted MBS and buffer exchange is accomplished by chromatography over Sephadex G-50 equilibrated with 50 mM sodium phosphate buffer

(pH 7.0). Incorporation of MBS into the BSA was moni¬ tored by the increase in absorption of the conjugate at 280 nm. The peptides are coupled to the activated BSA by incubation at a ratio of 25 nmol peptide per nmol BSA for 3 hr. at room temperature and then unreacted peptide was removed by dialysis. The loading of the peptide on the BSA carrier is determined by amino acid analysis. Samples (100 ug) of the peptide-BSA conjugates and of the batch of BSA used for preparation of the conjugates were acid hydrolyzed (6N HCl, 110°C 24 hr.) and then analyzed using a Beckman 6300 amino acid analyzer with ninhydrin used as the detection method (Protein Structure Laboratory, University of California at Davis). The concentration of conjugate in the stock solution is calculated from the amino acid analysis. Application of these standard reaction conditions with MBS activated BSA yielded conjugates with 10 nmol T19V, 12 nmol C21E, 6 nmol S19C, and 16 nmol C19Q per nmol BSA. For coupling with MSAC, the water soluble cross-linker is dissolved in 100 mM sodium phosphate (pH 7.4) at a weight ratio of 1 mg MSAC per mg BSA and incubated 10 min. at room temperature. Removal of un¬ reacted MSAC and buffer exchange is performed by chroma- tography as before. For coupling, varying molar ratios of the peptide are incubated with the activated carrier overnight at room temperature and then unreacted peptide is removed by dialysis. Coupling with 1% glutaraldehyde in phosphate buffered saline is performed for 1 hr. at room temperature and the reaction is terminated by addi¬ tion of sodium borohydride.

The following example relates to preparation of antibodies of the invention.

EXAMPLE 2

Peptides T19V, C21E and S19C were conjugated to KLH through cysteine residues using m-maleimidoben- zoic acid-N-hydroxy-succinimide ester. The amount of peptide bound to the carrier was determined by amino acid analysis and was generally 0.2-0.4 mg peptide bound per mg of KLH. Female New Zealand White rabbits were immunized three times at two-week intervals with 500 ug of the conjugate emulsified in Freund's complete adju- vant for the primary immunization and in Freund's incom¬ plete adjuvant for the secondary immunizations. Titer and specificity of rabbit antiserum were assayed by ELISA. For the ELISA, flexible poly-vinyl chloride microtiter plates were coated overnight at 8°C with 100 ul per well of peptide at 5 ug per ml in 150 mM sodium borate buffered at pH 9.6. The wells were blocked by incubation with 3% bovine serum albumin in phosphate buffered saline for 1 hr. at room temperature, then incubated overnight with serial dilutions of the rabbit sera diluted in 10% horse serum, 20% goat serum and 0.1% Triton X-100 in phosphate buffered saline. Bound rabbit immunoglobulin was detected with a Vectastain kit (Vector Labs.) utilizing biotinylated goat antibody to rabbit immunoglobulin, avidin, and biotinylated horse radish peroxidase. Color development was achieved with o-phenylenediamine and hydrogen peroxide and the ELISA was read on a dual wavelength microplate spectrophotometer at 450 nm. - The titers in Table II below are the dilution of rabbit antiserum that gave one-half maximal color development (usually optical density of 0.5-0.6).

Each of the rabbits developed a strong anti¬ body response to its respective immunizing peptide, but not to a control peptide, M20L. The antiserum to C21E also reacted with the peptide that overlaps C21E at its N-terminus (T19V), but not with the peptide that over¬ laps at its C-terminus (S19C). The antisera to T19V and S19C showed little cross-reactivity with C21E.

The following example relates to characteriza¬ tion of antibodies according to the invention.

EXAMPLE 3

To determine if the antisera prepared accord¬ ing to Example 2, above, recognize the HIV-1 glycopro¬ tein gpl20 or its precursor, gpl60, radioimmune pre- cipitation of ³⁵ S-methionine labeled, HIV-1 infected cells was performed. All three antisera immunoprecipi- tated proteins of 120 kD and/or 160 kD from Molt-III cells infected with the HIV-IIIB isolate. These pro¬ teins were hot immunoprecipitated from uninfected Molt-III cells. In addition, preimmune sera from the rabbits did not react with gpl20 or gpl60.

The rabbit preimmune and immune sera were tested for neutralization of HIV infectivity using an assay as described in Ho et al., J.Virol. , 61(6) , 2024- 2028 (1986).

Each virus inoculum (100 ul, fifty 50% tissue culture infective dose) was preincubated with the test serum (100 ul, serial two-fold dilutions) for 1 hr. at 37°C before inoculation onto 2.0 X 10 ⁶ H9 cells in 5 ml of RPMI 1640 medium supplemented with fetal calf serum (20%) , N-2-hydroxyethylpiperazine-N'-2-ethane-sulfonic acid (10 mM) , penicillin (250. U/ml), streptomycin (250 ug/ml), and L-glutamine (2 mM) . On day 7 of culture, each culture was examined for characteristic cytopathic effects with syncytia formation and for p24 antigen in- the supernatant fluid by an immunoassay (Abbott

Laboratories, North Chicago). Neutralization was defined as >90% reduction in both syncytia formation and supernatant p24 antigen compared to control cultures, which were similarly established except that the virus inoculum was preincubated with culture medium or normal rabbit serum. HIV binding inhibition studies were per¬ formed using the protocol of McDougal et al., J.Immunol. , 137, 2937-2944 (1986). The virus stock used in the assays was prepared as follows. Supernatant fluid from HIV-IIIB-infected Molt-III cells were pre- cleared by sequential centrifugation (300 g for 7 minutes, followed by 1500 g for 20 minutes) and then concentrated 1000-fold by ultracentrifugation (90,000 g for 90 minutes) onto a cushion of 15% Renografin-60 (Squibb, Princeton, N.J.) in a 0.01 M Tris, 0.15 M NaCl, 1 mM EDTA, pH 8.0. Ten ul of this HIV preparation were pretreated with 10 ul of test sera for 30 minutes at room temperature prior to incubation with C8166 cells (5 X 10 ⁵ , 30 minutes at 37°C) . Subsequently, the cells were washed and resuspended in 25 ul of a 1:50 dilution of human anti-HIV conjugated to fluroescin. After 30 minutes at 4°C, the cells were then washed, fixed in 1% paraformaldehyde, and analyzed by flow cytometry. The results are summarized in Table III, below. The preimmune sera had no neutralizing acti¬ vity. The antiserum to peptide C21E strongly neutralized the homologous isolate, HIV-IIIB, at a titer of 1:128. More importantly, the antiserum was strongly neutralizing against the "heterologous" isolates tested, HIV-IIIRF and HIV-ARV-2. The antisera to peptides T19V and S19C had weak, but detectable neutralizing titers of 1:8 and.1:16, respectively, against HIV-IIIB, and had similar titers against HIV-IIIRF and HIV-ARV-2. Animal sera produced against the entire gpl20 envelope protein - frequently show restricted neutralization of only the HIV-1 strain from which the gpl20 protein was pre-

pared. For example, a goat antiserum to gpl20 prepared from HIV-IIIB, neutralizes HIV-IIIB at a titer of 1:32, has a very low neutralizing titer against HIV-IIIRF and does not neutralize HIV-ARV-2.

TABLE III

Reciprocal of HIV Inhibition of Neutralizin Titer HIV Binding (%

Rabbit Sera HIV-IIIB

#1 Pre-immune serum

#1 Anti-T19V 0

#2 Pre-immune serum 1

#2 Anti-C21E 2

#3 Pre-immune serum 3

#3 Anti-S19C 0

Control Sera

Human, seronegative (n=6) <4 <4 <4 0-5

Human, seropositive (n=18) 20 18 25 93-100

Human, P982 32 32 64 99

Goat , anti-gpl20 32 <4 4 94

The neutralizing titer of the antiserum to C21E is several fold higher than the mean titer of sera from HIV-1 seropositive persons. In particular, although human serum P982 is more reactive with gpl20 and gpl60 by radioimmune precipitation, the antiserum to C21E is four-fold more neutralizing against HIV-IIIB and HIV-ARV-2. The antiserum to C21E is also more neutral¬ izing than the goat antiserum to gpl20.

Most sera from HIV-1 seropositive individuals show restricted neutralization of different HIV-1 strains, as do animal antisera prepared against gpl20. Presumably, the immune response to gpl20 is directed against antigenically variable domains as opposed to regions of conserved sequence. Indeed, HIV-1 seroposi- tive human sera were largely negative in an ELISA with the T19V, C21E, and S19C peptides, and those few scoring as positive were not strongly reactive. Furthermore, sera from animals immunized with gpl20 also did not react significantly with the peptides as determined by ELISA. Chemically synthesized peptides from other regions of the envelope protein, notably the C-terminus of gpl20 and one domain of gp41 (amino acids 600-611) are almost universally recognized by sera from HIV-1 infected individuals in a peptide ELISA. Thus, when the second conserved domain of gpl20 is presented to the immune system in the context of the larger polypeptide, it appears at best minimally immunogenic.

To explore the mechanism of HIV-1 neutraliza¬ tion by the sequence-specific antisera to the second conserved domain, the sera were tested for inhibition of HIV-1 binding to CD4+ cells by using the-_method of McDougal et al. Test sera were incubated with a con¬ centrated HIV-1 preparation for 30 min. at room tempera¬ ture and then exposed to C8166 cells, a highly positive CD4 T-cell line. See, Salahuddin et al., Virology, 129, 51-64 (1983). HIV-1 bound to the T-cells was reacted

with fluorescein isothiocyanate-conjugated human anti¬ body to HIV-1 and then was quantitated by flow cytometry. As shown in Table III, neither preimmune nor immune rabbit antisera to T19V, C21E, and S19C inhibited binding of HIV-1 to the T-cell line. In particular, despite the high HIV-1 neutralizing activity of the antiserum to C21E, no inhibition of binding was observed. In contrast, and to validate the assay, human HIV-1 seropositive sera and the goat antiserum to gpl20 were very efficient at blocking HIV-1 binding to the T-cell line.

Preliminary screening studies have revealed neutralizing activity for antibodies of the invention with respect to infectivity of HIV-MN. The neutralizing antisera elicited with C21E and peptide "1-110"

(VSTVQCTHGIRPW representing HTLV-III residues 249-262) do not inhibit binding of HIV-1 to CD4+ cells, [Ho et al., Science, 239, 1021-1023 (1988)] in agreement with the mapping of the CD4 binding domain to the carboxyl- end of gpl20 [Kowalski et al.. Science, 237, 1351-1355 (1987); Lasky et al.. Cell, 50, 975-985 (1987)]. This suggests that a post-CD4 binding event required for viral penetration has been blocked. These results are consistent with those of Willey et al., J. Virol., 62, 139-147 (1988), who found that mutations of residues in gpl20 corresponding to regions within the C21E peptide sequence destroy infectivity, although the mutant gpl20 still retains binding to CD4.

The following example relates to characteriza- tion of peptides of the invention.

EXAMPLE 4.

Tests were performed to determine whether synthetic peptides having homology to gpl20 would inhibit the biological activity of neuroleukin as reported in Lee et al., supra. In initial procedures

involving peptides T19V, C21E and S19C prepared accord¬ ing to Example 1, up to 10 ug per ml (from 4-5 uM) of each of the unconjugated peptides was added to culture medium containing 4-5 biological units of recombinant mouse NLK. That concentration of NLK maintains maximum survival and growth of sensory neurons cultured from 10 day chick embryos. No inhibition of sensory neuron growth was obtained with any of the peptides.

Testing of the conjugates of the peptides to carrier proteins that had been prepared for immunization according to Example 2 revealed the T19V peptide con¬ jugate to be a potent inhibitor of neuroleukin in the culture assay.

Table IV below provides a summary of the results of screening of conjugates of peptide whose preparation is referred to in Example 1. In the Table, concentrations of stock solutions were calculated from amino acid analysis. Molecular weights employed for the calculations were: C19Q, 1985; T19V, 1970; C21E, 2208; S19C, 2074; M20L, 2306; F20H, 2306; BSA, 67,000. As noted in the Table, fifty percent inhibition of sensory neuron survival in culture medium containing 4-5 bio¬ logical units of NLK (defined as the I ₅₀ for the inhi¬ bitor) was obtained with the T19V-BSA(C) conjugate added at 5 ng per ml of culture medium (94 pM) . The C19Q peptide which overlaps the amino terminus of T19V was also an inhibitor of NLK, while the C21E peptide which overlaps the carboxyl terminus of T19V was not inhibi¬ tory. The C19Q-BSA conjugate, even though it had a slightly greater loading of peptide, was less inhibitory than the T19V-BSA conjugate and had an IC _Q of 13 ng per ml (278.pM). BSA conjugates prepared with C21E or S19C did not inhibit sensory neuron growth in NLK at up to 1 ug per ml (> 100 nM) . thus, the set of peptides define a sequence within the midsection of the T19V peptide that is a potent inhibitor of NLK.

The homologous NLK peptide, designated M20L, is also an antagonist, and even more potent than T19V. An M20L-BSA conjugate with a loading of 10 nmol M20L peptide per nmol BSA was found to have an I ₅₀ of 2 pM (two experiments). T19V-BSA conjugates with equivalent peptide loadings had an Igg of approximately 100 pM. The inhibition of NLK by M20L-BSA was sequence speci¬ fic. A BSA conjugate prepared with the control peptide, designated F20H, which had an identical amino acid com- position, but a shuffled sequence, was not inhibitory at up to 100 nM.

TABLE IV

Inhibition of NLK dependent sensory neuron survival by peptide conjugates

# - determined by amino acid analysis, mean ± SD where n is the number of amino acids in the peptide that were analyzed.

* - determined by percentage incorporation of ¹²⁵ I-labeled peptide into the conjugate. ** - As reported in Lee et al., supra.

Antibodies of the present invention include polyclonal and monoclonal antibodies of diverse mam¬ malian origins. Monoclonal antibodies derived from murine hybridoma cell sources and chimeric (e.g., mouse/human) and anti-idiotypic antibodies are currently undergoing development.

EXAMPLE 5

C21E peptide was coupled to keyhole limpet hemocyanin (KLH) with m-maleimidobenzoic acid-N-hydroxy- succinimide ester (MBS) as described in Table I and Example 1, supra. Approximately 0.2 - 0.4 mg of peptide was bound per mg of KLH. Eight 12-week old female BALB/c mice were immunized with 50 yg C21E-KLH conjugate emulsified in Freund's complete adjuvant at multiple intradermal sites (day 0). The mice were immunized a second time with C21E-KLH emulsified in Freund's incomplete adjuvant on day 14, and then were boosted on day 28 with C21E-KLH in saline delivered intraperitonealy. The spleens were harvested on day 31 for fusion with mouse SP2/0 cells.

Hybridoma fusions of immunized mouse spleen cells with SP2/0 cells (5:1 fusion ratio) followed the protocol of Galfre and Mils±ein, Methods Enzymology, 73, 1-45 (1981). Hybridomas were selected for growth in hypoxanthine-aminopterin-thymidine.

Monoclonal antibodies reactive with C21E were detected using ELISA. The C21E peptide was used to coat the wells of 96-well polyvinylchloride microtiter plates at a concentration of 5 yg/ l in 0.1 M sodium bicarbonate (pH 9.6). For the gpl60 ELISA, recombinant gpl60 coated ELISA plates were used that were purchased from MicroGeneSys (West Haven, CT) . Supernatants from hybridoma microcultures were incubated in the wells for 2 hr at 37°C. The assay was then developed with a

Vector Laboratories ABC kit using biotinylated antibody to mouse immunoglobulin and a biotinylated horseradish peroxidase, avidin complex. Color was developed with o- phenylenediamine/hydrogen peroxide and quantitated at 450 nm.

Four immunized spleens were fused separately to SP2/0 cells and 391 hybridoma lines were obtained. Of these, 55 hybridomas secreted into the culture supernatant an antibody that was reactive with both C21E and gpl60 as measured by ELISA. The 55 hybridoma cell lines and the reactivities of their respective antibodies are listed in Table V below. The secreted antibodies are of the IgG subclass and are reactive with protein A.

* ND means not determined.

TABLE V (Continued)

To determine which of those antibodies which are reactive with C21E were also reactive with HIV-1 envelope glycoprotein(s) , radioimmunoprecipitation analysis (RIPA) as described Example 3, supra, was performed on sixteen of the 55 monoclonal antibodies. The HTLV-IIIB isolate of HIV-1 was used to infect H9 T- lymphocytic cells. The infected cell cultures were fed S-35 labeled methionine and cysteine to incorporate S-35 label into viral proteins. Detergent lysates were prepared from the infected cells. For RIPA, mouse monoclonal antibodies from the culture supernatants were attached to Sepharose™ beads through goat anti-mouse immunoglobulin which previously was covalently bound to the surface of the beads. The beads were then washed and subsequently incubated with S-35 labeled, HIV- infected H9 cell lysates. After incubation for 2 hr at room temperature, the beads were then washed with RIPA buffer containing 0.5 M sodium chloride, and bound antigen was released by boiling in sample^buffer containing sodium dodecyl sulfate and B- mercaptoethanol. The released antigen was analyzed by electrophoresis on 7.5% SDS-polyacrylamide gels and

autoradiography against X-ray film to determine whether the mouse monoclonal antibodies had reacted with any HIV-1 glycoproteins.

Thirteen of the 16 monoclonal antibodies assayed, including the monoclonal antibody designated JK112, were found to immunoprecipitate radiolabeled gpl60 and gpl20 envelope glycoproteins from HIV-1 (HTLV- IIIB) infected cells. These results are presented in Table V. To further characterize the reactivity of the

JK1112 monoclonal antibody with HIV-1 envelope glycoproteins, additional isolates of HIV-1 were radiolabeled, including isolates designated RF, MN, COSTA, Z34, Z84, and AL. The RF isolated has been described in Starcich et al.. Cell, 4_5, 637-648 (1986); the MN isolate has been described in Gurgo et al., Virol. , 164, 531-536 (1988); the remaining isolates are clinical isolates of one of the co-inventors. Monoclonal antibody JK112 reacted with the gpl60 and/or gpl20 envelope glycoprotein of every isolate of HIV-1 tested.

To show that the immunoprecipitation observed was due to specific reaction of the JK112 monoclonal antibody with the C21E sequence embedded within the HIV- 1 envelope glycoprotein, the antigen binding site of the JK112 monoclonal antibody was pre-saturated with C21E peptide (1 μg per immunoprecipitation). Subsequent attempts at immunoprecipitation of gpl60 and gpl20 from HIV IIIB isolate was blocked thereby indicating a specific reactivity of the JK112 monoclonal antibody with the target C21E sequence. _ι^ __

To demonstrate the genetic stability of the JK112 hybridoma, it was subcloned by limiting dilution ^' using feeder cells prepared from un-immunized BALB/c mouse spleens. In the first subcloning, 36 of 36 subclones secreted antibody reactive with C21E. Three

subclones were chosen (JK112.6, JK112.16, and JK112.17) and each was subcloned a second time by limiting dilution. In the second subcloning, 28 of 28 JK112.6 subclones, 36 of 36 JK112.16 subclones, and 22 ^' of 22 JK112.17 subclones secreted antibody reactive with

C21E. One of each of these subclones was retained and designated JK112.6a, JK112.16a, and JK112.17a.

Ochterlony assay with antisera specific for different subclasses of mouse immunoglobulin revealed that JK112 is an IgG3 subtype antibody. Because the

JK112 monoclonal antibody binds to protein A, it can be purified from culture supernatants in one step. Supernatants were harvested from JK112 hybridoma cultures, adjusted to pH 8.0 by addition of sodium phosphate buffer to a final concentration of 20 mM, and then were passed over a column of protein A covalently bound to Sepharose CL4B™. The optical absorbance of the effluent was washed to baseline with pH 8 buffer containing 20 mM sodium phosphate and 150 mM sodium chloride. JK112 monoclonal antibody was eluted from the column using pH 4.5 buffer containing 100 mM sodium citrate. The JK112 antibody was dialyzed into phosphate-buffered saline, concentrated to 0.5 - 0.25 mg protein per ml, sterilized by filtration through a 0.22 micron filter, and stored frozen until use.

EXAMPLE 6

An HIV-1 Ln vitro neutralization assay of the JK112 monoclonal antibody was performed as follows. A titered stock of HIV-1 containing 50 tissue culture infectious doses was incubated with varying amounts (from 10 μg to 10 pg) of JK112 monoclonal antibody for 30 minutes at 37°C before dilution twenty-five fold for inoculation onto H9 cells, as per Example 3, supra.

Samples of the culture supernatant were harvested 4, 7,

10 and 14 days after infection for assay of p24 GAG antigen as an index of HIV-1 infection and lEli replication. Data are expressed as the rations the amount of p24 antigen in experimental cultures (i.e., treated with JK112) divided by the amount of p24 antigen in control cultures (i.e., virus only without JK112 antibody) and normalized to 100%. This value was then substracted from 100% to give percent neutralization. JK112 monoclonal antibody efficiently neutralized the three structurally distinct isolates of HIV-1 tested (IIIB, RF, and MN) . This data, collected 7 days after infection, is presented in Table VI, below.

HIV-1 isolate

IIIB IIIB

RF RF

MN

MN

EXAMPLE 7

Three overlapping peptides were used to map the amino acid sequence recognized by JK112 monoclonal antibody. The peptides are designated T19V, C21E, and S19C, as described in Table I, supra. In an ELISA, JK112 reacted with C21E and S19C, but notr-with T19V. Thus, JK112 reacts with the minimum sequence S-T-Q-L-L- L-N-G-S-L-A-E comprising the carboxy-terminus of the C21E peptide. As described, supra, in Example 2, the rabbit antiserum to C21E reacted with T19V and C21E

peptides, but not with S19C. Thus, the rabbit antiserum reacted with the minimum sequence C-T-H-G-I-R-P-V comprising the amino-terminus of the C21E peptide. The rabbit antiserum produced by immunizing with the C21E peptide also efficiently neutralized in vitro infection of T-lymphocytic cells by the IIIB, RF, and MN isolates of HIV-1. See Table III, supra. Thus, a minimum of two distinct epitopes are present in C21E, either of which will elicit antibodies that efficiently neutralize HIV-1 infection in the i ^~ \ vitro model.

EXAMPLE 8

As described supra, two independent lines of evidence suggest gpl20 residues 254-271 as being important for HIV-1 infectivity. Interestingly, Young Nature, 333, 215 (1988) has identified a short region of the HLA-DR beta chain which has sequence homology to gpl20 residues 261-270. That homology seemed of interest because class II molecules present on antigen presenting cells form a complex with CD4 and the T-cell antigen receptor on effector cells [Kappes et al., Ann. Rev. Biochem., 57, 991-1028 (1988)]. The possibility that the binding of HIV-1 to CD4 mimics some aspect of that association was therefore investigated.

The HLA class II molecule is a dimer consist¬ ing of an alpha and a beta chain [Kappes et al., Ann. Rev. Biochem., 57, 991-1028 (1988)]. Each chain con¬ sists of an external T-cell recognition domain (A-l and B-l, respectively) and a membrane-proximal domain (A-2 and B-2). The B-l domain of HLA-DR is highly polymor¬ phic and is important for presenting processed antigen to HLA class II-restricted T-cells. The A-2 and B-2 domains probably have a structural function and have considerable homology to immunoglobulin constant domains (the C-l Ig homology set defined by Williams et al..

Ann. Rev. I munol. , 6_, 381-405 (1988). Class II- restrictjf .-cells express CD4 and monoclonal antibodies to specifiϊ|epitopes of CD4 can block class II- restricted lymphocyte responses [Rogozinski et al., \_ Immunol. , 132, 735-739 (1984)]. The same antibodies which block CD4 function also block binding of HIV-1 [Sattentau et al.. Science, 234, 1120-1123 (1986)]. CD4 binds to class II molecules and probably helps stabilize the contact between class II molecules and the T-cell antigen receptor [Doyle et al.. Nature, 330, 256-259 (1987)]. Experiments with class I/class II chimeric molecules strongly suggest that CD4 binds to the external B-l domain of class II molecules and reinforce the view that the A-2 and B-2 domains of class I molecules simply provide a structural framework for the T-cell recognition domains [Golding et al.. Nature, 317, 425-427 (1985)].

The HLA class II molecule is structurally homologous to the HLA class I molecule. Instead of two polymorphic chains, the class I molecule has an epoly- morphic chain (with 3 domains designated A-l through A-3) which is associated with an invariant protein, B2-microglobulin (B ₂ -m). The A-l and A-2 chains form a groove or saddle which has contact residues for T-cell recognition and a binding site for foreign antigen

[Bjorkman et al.. Nature, 329, 506-512 (1987)]. The A-3 domain and B2~ provide structural support for the T-cell recognition domains, and like the membrane proximal domains of class II molecules, are also members of the C-l Ig homology set [Williams et al., Ann. Rev. Immunol. , 6 , 381-405 (1988)]. Thus, the H2 HLA-DR, A-3 HLA-A and B2~m sequences possess considerable homology and readily align.

The HLA class I A-3 domain contains two anti- parallel beta-pleated sheets. One sheet contains four beta-strands; the other contains three beta-strands, and

the two sheets are connected by a disulfide which is a hallmark of Ig domains. The association of B ₂ ~m with the class I molecule is primarily through contacts on the A-3 domain. These contacts cluster mainly over a 12 amino acid stretch (residues 231, 233, 234, 237 and 238) comprising the beta 4-strand and the loop between strands beta-4 and beta-5 [Bjorkman et al.. Nature, 329, 506-512 (1987)]. The A-3 beta-4 strand of the class I molecule contacts B2~m perpendicularly to the B2~m anti- parallel beta-sheet along a shallow groove across the inner surface of the protein.

The secondary structure of the A-3 domain is schematized in Figure 1 with the corresponding pre¬ diction of beta-strands plotted immediately underneath (Chou-Fasman algorithm. University of Wisconsin Computing Genetics Group) . In this Figure are represented predicted structures for the HLA-A A-2 (residues 203-259), the HCMV-H301 gene (residues 224- 281) _r the HLA-DR B-2 (residues 117-173) and HIV gpl20 (residues 235-295). For each sequence, a schematic representation of beta-sheet (sawtooths) is drawn above the Chou-Fasman prediction of beta-sheet regions (plus (+) indicating beta forming; and minus (-) indicating beta breaking) . Cysteines bounding the domains are indicated by asterisks. The observed and predicted beta-structure for the class I A-3 domain are in good agreement. Immediately beneath are plotted beta-strand predictions for a human cytomegalovirus (HCMV) encoded class I-like molecule (designated H301, Beck et al., Nature, 331, 269-272 (1988), the class II B-2 domain, and residues 235-295 of the carboyxl-end of the gpl20 second conserved domain. The class I and class II molecules are aligned with respect to conserved cysteine residues in the Ig homology domain and the gpl20 sequence has been aligned with the region of homology in the class II B-2 domain (residues 142 to 151). Note

that each sequence, including that of HIV-1, predicts a pattern of five alternating, anti-parallel beta- strands. The five beta-strands predicted for the gpl20 sequence (numbered 2-6 following the numbering ^' of the A-3 class I structure) fall at residues 238-242,

247-255, 260-268, 274-280, and 290-295. Thus, a "beta- sandwich" structure of anti-parallel beta-sheets is a reasonable structural prediction for this region of the gpl20 second conserved domain. For purposes of illustration. Figure 2 provides a schematic representation of a constant Ig domain showing the anti-parallel arrangement of beta- sheets. Strands of the beta-sheets are indicated by the ribbon arrows. Loops of polypeptide chain which are believed to be antigenic determinants link adjacent strands of the beta-sheets. Linear epitopes are frequently found to comprise loops of polypeptide chain exposed on the surface of proteins [Tainer et al.. Nature, 312, 127-134 (1984)]. The model of anti-parallel beta-strands for gpl20 residues 235-300 is in good agreement with the existing mutational analysis and the mapping of epitopes within that region. The epitope apparently recognized by both anti-C21E and anti-1-110, for example (CTHGIRPV), maps precisely to the proposed loop between strands beta-3 and beta-4. Similarly, the epitopes recognized by JK112 monoclonal antibody (STQLLLNGSLAE) and anti-S19C (LAEEEWIR) overlap the proposed loop between strands beta-4 and beta-5. The antiserum to T19V was poorly reactive with gpl20, perhaps because it was directed against a sequence (TGPCTNVSTVQ) located in the beta-3 strand rather than a loop region.

Figure 3 diagrammatically represents a structure for residues 234-300 of gpl20 which is consistent with the above. B-2 through B6 are the predicted strands of beta-sheet. The residues beginning

and ending the inter-strand acid sequences of the loops below, correlates projected the gpl20 region with their location in the predicted beta-sandwich structure.

TABLE VII

Epitope Residues Location in predicted Sequence in gp!20 beta-sandwich structure

KCNNK 234-238 upstream of beta-1 strand

GTGPCT 242-247 inter-strand 2-3 loop

THGIR 255-259 inter-strand 3-4 loop

LNGSLA 268-273 inter-strand 4-5 loop

RSANFTDNAK 280-289 inter-strand 5-6 loop

NQSVEI 295-300 downstream of beta-6 strand

The region of homology between gpl20 and HLA-DR aligns with the contact residues (underlined) between the HLA-A class I beta 4-strand and B ₂ -m. Figure _. 4. The contact residues are located at the end of the beta-4 strand (V ²³¹ , τ ²³³ , and R ²³⁴ ); within the inter-strand loop (G ²³⁷ and D ²³⁸ ); and at the beginning of the beta-5 strand (Q ²⁴² and W ²⁴⁴ ). The class I molecule encoded by HCMV aligns with this sequence and is believed to bind B ₂ -m [Beck et al., Nature, 331, 269- 272 (1988)], but as can be seen only two contact residues are conserved. The gpl20 and HLA-DR sequences also align with the class I sequences over the contact region. Four contact residues are conserved between HLA-DR and HLA-A, and three of these are~also present in the gpl20 sequence. The alignment between HLA-DR and gpl20 is strongest over the length of the predicted contact region in the beta-4 strand and the loop between beta-strands 4 and 5.

Mutations in the inter-strand beta-4/-5 loop destroy HIV-1 infectivity. Three mutations have been made in the predicted beta-3 strand of gpl20, one in the loop between strands beta-3 and beta-4, and three in the predicted loop between strands beta-4 and beta-5 which align with the contact residues in the HLA-A sequence. Substitution of glutamine for asparagine replaces a weak beta-sheet former with a weak beta-sheet breaker, while aspartic acid is a strong beta-sheet breaker. Mutations in strand beta-3 (N ²⁴ ° and V ²⁴ *) are infectious and cause little change in the predicted beta-structure. The linker scanning insertion of Kowalski et al.. Science, 237, 1351-1355 (1987); [Ho et al., J. Virol., 61, 2024-2028 (1987)] lengthens the predicted loop between strands beta-3 and beta-4 without altering the beta-strands themselves; however, the mutant fails to process gpl60 to gpl20. Finally, mutations in the loop between beta-4 and beta-5 (L ²⁶⁶ , N ²⁶⁹ , and G ²⁷⁰ ) all destroy infectivity; however these have little effect on the predicted beta-structure. Residue G ² ' ⁰ of gpl20 aligns with the contact residue G ²³⁷ of HLA-A, if this region is important for gpl20 association with another protein, mutations near or within the contact site might destroy infectivity by disrupting that association. This analysis indicates that the second con¬ served domain has structural features which might allow it to bind to a member of the Ig supergene family. When retroviruses bud from the infected cell, they often incorporate cellular surface antigens into their envelope. Non-random association of HIV-1 virions is seen with two cellular antigens, both of which are members .of the Ig superfamily, B2~m and HLA-DR [Hoxie et al.. Hum. Immunol., 18, 39-52 (1987)]. The B ₂ ~m detected in HIV-1 virion preparations is not associated with HLA class I molecules.

The differential reactivity of JK112 mono¬ clonal antibody witg|| l60 versus gpl20 also provides some evidence that th§l3econd conserved domain of gpl20 may be complexed with an additional protein. The presumed contact residues between gpl20 and B2~m or another member of the Ig superfamily coincide within the epitope recognized by the JK112 monoclonal antibody (STQLLLNGSLAE) as do the mutational sites which destroy HIV-1 infectivity. JK112 binds strongly to gpl60, but has little reactivity with gpl20.

The above illustrative examples are believed to establish that the preparation of peptides having sequence homology to the "second conserved region" of HIV-1 gpl20 glycoprotein can provide valuable immuno- chemical reagents. Not all such peptides are valuable and, indeed, some display a substantial potential for generating neurotoxic effects. When peptides of the invention are prepared based on homology to an antigenic determinant or epitope within the region spanning gpl20 residues 254-280 and preferably 254-274, neurotoxic potential is not observed. As is the case with illus¬ trative peptide C21E, preferred peptides of the inven¬ tion include an epitope homologous to the gpl20 sequences spanning residues 254 through 262. Assays for the diagnosis of AIDS and ARC employing peptides, peptide conjugates and antibodies of the presence invention may readily be developed accord¬ ing to well known procedures such as are referred to in PCT Application WO87/02775. In a like manner, it is within the contempla¬ tion of the invention to employ standard means to develop active and passive immunization processes using peptides, antibodies and anti-idiotypic antibodies of the invention. Thus, in addition to developing peptide conjugates for vaccination procedures as illustrated in the foregoing examples, it is within the contemplation

of the invention to provide peptides of the invention as portions of larger polypeptides such as poly-L-glutamic acid or poly-L-lysine, to subject the peptides to cationization and thereby enhance immunogenicity as in Muckerheide et al., J.Immunol., 138, 833-837 (1987), to couple peptides of the invention with antibodies known to facilitate strong serological responses to protein antigens as in Carayanniotis et al., Nature, 327, 59-61 (1987), to attach peptides to foreign helper T-cell epitopes according to the procedures of Francis et al., Nature, 330, 168-170 (1987), or to use anti-idiotypic antibodies of the invention as in Chanh et al., Proc. Nat'l. Acad. Sci. (USA), 84, 3891-3895 (1987).

While the foregoing examples relate to murine- derived hybridoma cell preparations, it is within the contemplation of the invention to generate and employ hybrid hybridomas (e.g., mouse/human) and especially human/human hybridomas prepared, for example, in a manner consistent with Borrebaeck, TIBTECH, June, 1986, pp. 147-153; Abrams et al.. Methods in Enzymology, 121, pp. 107-119 (1986); Kozbor et al.. Methods in Enzymology, 121, pp. 120-140 (1986); Suresh et al. , Methods in Enzymology, 121, pp. 210-228 (1986); and Masuho et al., Biochem. & Biophys. Res. Comm. , 135(2) , pp. 495-500 (1986). See, also, Klausner, '"Single

Chain' Antibodies Become a Reality", Bio/Technology, 4_, 1041-42 (1986), Klausner et al., "Stage Set For 'Immunological Star Wars'", Bio/Technology, 5_, 867-868 (1987) and Marx, "Antibodies Made To Order", Science, 229, 455-456 (1985).

Immunological complexes of the invention are formed ^n vitro upon contact between antibodies of the invention with HIV-1 virus and virus particles and host cells. Such complexes are expected to be formed iji vivo upon practice of vaccination procedures of the invention and expected to preclude infectivity of the virus.

Numerous modifications and variations in prac¬ tice of the invention will occur to those skilled in the art upon consideration of the foregoing illustrative examples and consequently only such limitations as appear in the appended claims should be placed thereon.