WO1989005455A1

WO1989005455A1 - Aids-related anti-t-cell antibody test

Info

Publication number: WO1989005455A1
Application number: PCT/US1988/004293
Authority: WO
Inventors: Raphael B. Stricker
Original assignee: University of California Berkeley; University of California San Diego UCSD
Current assignee: University of California Berkeley; University of California San Diego UCSD
Priority date: 1987-12-02
Filing date: 1988-12-02
Publication date: 1989-06-15
Anticipated expiration: 1990-06-02
Also published as: AU2800389A

Abstract

Methods of diagnosing the clinical stage of HIV infections is provided wherein the level of autoantibodies to histone H2B in a patient is determined. Kits for the practice of the diagnostic method are also provided.

Description

AIDS-RELATED ANTI-T-CELL ANTIBODY TEST

Technical Field

The present invention relates to diagnostic methods and kits for determining the onset of acquired immunodeficiency syndrome (AIDS) and AIDS-related condi¬ tion (ARC). More particularly, the present invention is directed to methods, such as immunoassays, and kits use¬ ful in such methods, to detect the presence of an autoantibody directed against histone H2B.

Background

Patients with AIDS and ARC are known to have abnormalities of T-cell subpopulations, including a decreased helper/inducer (CD4+) to suppressor/cytotoxic (CD8+) T-cell ratio and decreased absolute number of

T-cells with the helper/inducer phenotype. See, e.g., Fauci et al. (1984) Ann. Int. Med. 100:92-106; Ammann et al. (1983) Clin. Immunol. Immunopathol. 27;315-325. Infection of T-cells with a retrovirus, termed human immunodeficiency virus (HIV) type 1 or 2, is thought to play a role in these abnormalities. Levy et al. (1984) Science 225;840-842; Barre-Sinoussi et al. (1983) Sci¬ ence 220:868-871; Gallo et al. (1984) Science 224:500-502. However, HIV infection alone does not ade- quately explain the CD4+ T-cell abnormalities seen in

AIDS, and the nature of T-cell destruction in this dis¬ ease remains poorly characterized. Quinnan et al. (1985) Ann. Int. Med. 103:710-714. A number of autoimmune phenomena have been described in patients with AIDS and AIDS-related condi¬ tions. See, e.g., Strieker et al. (1985) N. Engl. J. Med. 3ϋ:1375-1380; Bloom et al. (1986) J.A.M.A. 25_6:491-493; Cohen et al. (1986) Ann. Int. Med. 101:175-180; Toy et al. (1985) Am. J. Hematol. 19:145-150; Lane et al. (1983) N. Engl. J. Med. 109:453-459; Dalakas et al. (1986) J.A.M.A. .26:2381; Kiprov et al. in Acquired Immune Deficiency Syndrome, pp. 299-308 (Gottlieb et al. eds. 1984); Kloster et al. (1984) Clin. Immunol. Immunopathol. 30:330-335; Pruzanski et al. (1984) AIDS Res. 1:211-220.

Serum autoantibody against a 25,000-dalton platelet protein has been found in 95% of homosexual men with immune thrombocytopenic purpura (ITP), as well as in 80% of homosexual patients with AIDS. Strieker et al. (1985), supra. The lupus anticoagulant (anti-phospholipid antibody), autoantibodies against red blood cells, and antinuclear antibodies have been detected in homosexual and hemophiliac patients with AIDS and AIDS-related conditions. Bloom et al. (1986), supra; Cohen et al. (1986), supra; Toy et al. (1985), supra: Lane et al. (1983), supra; Dalakas et al. (1986), supra. In addition, anti-lymphocyte antibodies have been reported in these patients. See, e.g., Kiprov et al. (1984), supra: Kloster et al. (1984), supra: Pruzanski et al. (1984), supra: Williams et al. (1984) J. Clin. Immunol. 4_:118-123; Tomar et al. (1985) Clin. Immunol. Immunopathol. 12:37-47; Dorsett et al. (1985) Am. J. Med. 2£:621-626; Stites et al. (1986) Clin. Immunol. Immunopathol. 38:161-177.

Kiprov et al., supra. found antibodies against T-cells in 70% of patients with AIDS. These antibodies reacted with 60% of helper/inducer T-cells but were unreactive with suppressor/cytotoxic T-cells. Dorsett et al., supra, also found selective binding of serum antibody from AIDS patients to OKT-4-positive (helper/inducer) T-cells. According to one theory, an aberrant autoimmune response in AIDS patients may be directed against the highly polymorphic HLA-DR region found on both Langerhans epithelial cells and lymphocytes. Belistro et al. (1984) N. Engl. J. Med. 110:1279-1282; Andrieu et al. (1986) AIDS Res. 1:163-174; Ziegler et al. (1986) Clin. Immunol. Immunopathol. 4JL:305-314. Evidence in favor of this hypothesis is lacking at present. Spencer et al. (1986) Lancet ii:983.

Summary of the Invention

It is an object of the present invention to provide methods and kits useful in determining the clin¬ ical stage of an HIV infection. The invention is based on the discovery of an autoantibody in the serum of AIDS and ARC patients, and the epitope of this autoantibody. The appearance of this antibody in the serum of an HIV-infected individual is associated with the onset of AIDS or ARC.

In one embodiment, the present invention is directed to a method for diagnosing the clinical stage of an HIV infection by determining the level of anti-H2B antibodies in a patient comprising: (a) providing an antibody-containing sample from a human patient infected by HIV; and (b) determining whether said sample contains antibodies to histone-H2B. Preferred methods include immunoassays, such as direct assays, competition assays, and, in particular, ΞLISAs.

In a preferred embodiment, the present inven¬ tion is directed to an immunoassay for diagnosing the clinical stage of an HIV infection by determining the level of anti-H2B antibodies in a patient comprising: (a) providing serum from a human patient infected by HIV; (b) contacting said serum with a sample of immobi¬ lized histone-H2B epitope to provide a serum-contacted sample; (c) contacting said serum-contacted sample with labeled anti-human Ig antibodies to provide label-contacted sample; and (d) determining the presence of the label on said label-contacted sample.

In another embodiment, the present invention provides a kit for an immunoassay to diagnose the clini¬ cal stage of an HIV infection comprising: (a) a solid immunoassay support having immobilized thereon histone-H2B epitopes; and (b) a container comprising labeled anti-human Ig antibody.

In still another embodiment, the present invention provides a kit for an immunoassay to diagnose the clinical stage of an HIV infection comprising: (a) a predetermined amount of histone-H2B epitope; and (b) a predetermined amount of labeled antibody to said histone-H2B epitope.

These and other embodiments of the present invention will readily occur to those of ordinary skill in the art based on the following disclosure.

Disclosure of the Invention

Disclosed herein is an AIDS-related serum autoantibody that reacts with an 18,000-dalton antigen restricted to lectin-stimulated or HIV-infected CD4+ T-cells. The antibody also suppresses proliferation of CD4+ T-cells in vitro and induces cytotoxicity of these cells in the presence of complement. It has been dis¬ covered that patients with AIDS and ARC produce this autoantibody directed against a specific antigen on stimulated or infected helper/inducer T-cells (CD4+). This autoantibody is not found in healthy heterosexual, homosexual or hemophiliac patients, but is detected in at least about 95% of patients with ARC or AIDS. Test¬ ing for the presence of this antibody, therefore, can be used to diagnose clinical AIDS or ARC. The antibody may also play a significant role in the pathogenesis of ARC and AIDS. Furthermore, it has been determined that the specific antigen to which this autoantibody is directed is the commonly occurring histone H2B, which is well known and occurs in many (if not all) eukaryotes.

According to the present invention, the autoantibody is detected by first obtaining an antibody-containing (i.e., immunoglobulin-containing) sample from a patient (e.g., blood, lymph, particularly the serum or purified immunoglobulin) that has been infected by a human immunodeficiency virus (HIV), such as HIV-1 or HIV-2. An HIV-infected patient is a patient whose blood contains anti-HIV antibodies and/or HIV polypeptides or intact virions. Then, it is determined whether the sample contains antibodies to H2B. This can be determined by any appropriate chemical or biochemical technique, the most convenient being an immunoassay where the sample is contacted with H2B. The selection . of a particular method to measure anti-H2B Ig is not critical, however, and is secondary to the discovery of the clinical significance of the presence of these autoantibodies.

Only particular epitopes found on H2B may be the targets of the autoantibody produced at the onset of clinical AIDS. Thus, it is understood that the present invention encompasses not only methods of contacting fluid samples from a patient directly with intact H2B, but also contacting the fluid sample with peptides and/or carbohydrates that mimic the H2B epitopes recog¬ nized by the autoantibody. Such peptides or carbohydrates can constitute discrete H2B fragments, or portions of a larger molecule having the identical structure as the H2B epitope region (e.g., fusion pro¬ teins, anti-idiotype antibodies, etc.). Thus, the pre¬ sent invention contemplates the use of any polypeptide, carbohydrate, glycoprotein or other entity which con¬ tains an epitope bound by the autoantibodies produced by patients with ARC or AIDS that also binds an epitope on H2B. Such cross-reactivity can be determined by a com¬ petitive immunoassay, or other techniques known in the art.

Histone H2B is well known in the art, and is available from commercial sources (e.g., Boehringer- Mannheim, Indianapolis, IN). H2B can also be purified from various human and non-human eukaryotic cells. See, e.g., Bohm et al. (1973) FEBS Lett 34:217; VanderWesthuyzen et al. (1971) FEBS Lett i£:333. The amino acid sequence of histone H2B has also been described in the literature. See, e.g., Ohe et al. (1979) J. Biochem. £5:615; Wu et al. (1984) Critical Reviews in Biochemistry 2):201. H2B can be isolated from such diverse cells as bovine (e.g., calf thymus), porcine, ovine, equine, and other mammals, as well as lower eukaryotes, such as yeast. Preferred sources of H2B are non-human sources, such as non-human mammal tis¬ sue, and lower eukaryote cells. While the amino acid sequence of H2B is highly conserved, it should be under¬ stood by those skilled in the art that it is possible that a sequence variation or difference in glycosylation may adversely affect a particular H2B's ability to bind the autoantibody. One of ordinary skill in the art, however, can readily determine the antigenicity of any selected H2B by routine screening based on the disclo¬ sure herein.

As previously stated, the preferred embodiment of the present invention employs an immunoassay to determine the presence of the autoantibody in a fluid sample from the patient. The techniques of immunoassays are well known in the art. See, e.g., Immunoassay; A Practical Guide (D. Chan & M. Perlstein, eds., 1987) (and references cited therein); Radioimmunoassay Methods (Kirkham & Hunter, eds. (1970) (and references cited therein); C.W. Parker, Radioimmunoassay of Biologically Active Compounds (1976); U.S. Patent No. 4,190,496; U.S. Patent No. 4,323,647; U.S. Patent No. 4,380,580; U.S. Patent No. 4,375,972; U.S. Patent No. 4,376,110; U.S. Patent No. 4,376,165; U.S. Patent No. 3,791,932; U.S. Patent No. 3,654,090; U.S. Patent Reissue No. 31,006; U.S. Patent No. 3,940,475; U.S. Patent No. 3,867,517; U.S. Patent No. 3,645,852; U.S. Patent No. 3,791,932, the disclosures of which are hereby expressly incorpo¬ rated herein by reference. Specific immunoassay methods for detecting particular antibodies in a sample from a patient are well known in the art. See, e.g., EPO Pub. No. 173,295; EPO Pub. No. 193,284; PCT Pub. No. WO

86/04993; EPO Pub. No. 136,798; PCT Pub. No. WO 85/04903; EPO Pub. No. 187,041; PCT Pub. No. WO 86/06414; EPO Pub. No. 202,890; PCT Pub. No. WO 86/06099; U.S. Pat. No. 4,520,113, the disclosures of which are expressly incorporated herein by reference.

There are at least hundreds of specific types of immunoassays, and at least dozens of general proto¬ cols involved in such assays, which could be employed in the present invention, all of which are known in the art. The assays can be qualitative or quantitative with respect to the autoantibody. Immunoassay protocols, according to the present invention, can be based, for example, upon known competition assays, or on known direct assays, both homogeneous and heterogeneous. In a competition assay, the analyte (i.e., the autoantibody) is caused to compete with a reagent (e.g., polyclonal or monoclonal anti-H2B antibodies), wherein the competing reagent is labeled. The amount of label detected, therefore, is inversely proportional to the amount of analyte present. In a direct assay, the analyte is detected by an antibody (i.e., anti-human immunoglobulin antibody, or anti-hlg antibody) , which itself is labeled. Thus, the amount of analyte present is directly proportional to the amount of label detected.

Examples of direct assays include enzyme-linked immunoabsorbent assays (ELISA) , sandwich-type assays, immunoblot assays, and Western-blot assays.

In the practice of the present invention it is preferred to determine the presence of the autoantibody by contacting it with anti-human immunoglobulin anti¬ body. It is understood that anti-hlg antibody includes intact antibody, or the F(ab' ^-binding fragment thereof. The anti-hlg antibody can be monoclonal or polyclonal, and can be from any of numerous animal spe¬ cies. Such antibodies are widely available commercially.

It is preferred that the anti-hlg antibody (or detector antibody) be labeled. The most common labels are radioisotopes, fluorescent materials, or enzymes capable of catalyzing detectable reactions (e.g., color change). The detector antibody can be directly labeled or indirectly labeled. By direct labeling is meant that a directly detectable label (e.g., radioisotope, fluo¬ rescent compound, or enzyme) is attached directly to the anti-hlg antibody. An indirectly labeled antibody is one which contains means for later attaching these directly detectable labels after the antibody has reacted with its epitope. The most common example of indirect labeling is the avidin-biotin system. In such a system, for example, the anti-hlg antibody would be biotinylated, and avidin is conjugated to an enzyme label. Thus, the biotinylated anti-hlg antibody is indirectly labeled. A preferred and well-known enzyme/substrate combination is horseradish peroxidase and o-phenylenediamine.

In the preferred direct assay of the present invention, H2B-epitope-containing compounds are immobi¬ lized prior to contacting with the fluid sample sus¬ pected of containing the autoantibody. A convenient method for making the H2B-epitope-containing compound immobile (i.e., insoluble) is to bind or immobilize H2B to a solid support. One of ordinary skill can select any suitable immunoassay solid support. Preferably, the solid support member is plastic (e.g., polystyrene, polypropylene, polyethylene, or polyvinyl). Such plas¬ tics readily absorb glycoproteins, thus immobilizing them on the surface. Normally, the plastic support is shaped to have a suitable receptacle or well for the fluid sample from the patient. A particularly preferred support member is a micro-immunoassay plate, or a microtiter plate. The receptacles of such plates range in volume, usually from about 250 μl to about 1 ml. Other types of solid support, such as nitrocellulose, polyacrylamide gels, and plastic and metal beads, are known in the art.

In a preferred direct assay according to the present invention, H2B is immobilized on a solid immunoassay support, contacted with a serum sample from a patient that is HIV-positive, and incubated to allow binding of any autoantibody to the antigen. The solid support is then washed with a suitable wash solution (e.g., phosphate-buffered saline, or PBS), incubated with labeled °anti-hlg antibodies, washed again, and then the presence of the label determined. In a more pre¬ ferred embodiment, the immunoassay is an ELISA using, e.g., the peroxidase or alkaline phosphatase technique, and the antigen is fixed to the wells of a microtiter plate, wherein approximately about 1 to about 10 μg of H2B per well is placed. In such an assay, it is usually desirable to block non-specific binding to the plate by also adsorbing a protein that should not react with antibodies in the test sample (e.g., gelatin) and to dilute a serum sample 1:100 in PBS.

In an alternative embodiment, H2B is immobi¬ lized on plastic microspheres, the microspheres are then contacted with patient serum, followed by contacting with anti-hlg antibodies labeled with a fluorescent com¬ pound, and then the fluorescing spheres are counted in a cell sorter by known techniques.

Another embodiment of the present invention is a standard Western blot. In this technique, peripheral blood T-cells, HUT-78 cells, or commercial preparations of H2B are treated with SDS detergent to solubilize pro¬ teins, and then electrophoresed in SDS polyacrylamide gels. The proteins are then transferred to nitrocellulose membranes and incubated with patient serum or purified immunoglobulin. Autoantibody binding to the immobilized antigen is then detected as above.

The present invention also encompasses kits suitable for the above diagnostic methods. These kits contain the appropriate reagents and are constructed by packaging the appropriate materials, including the H2B epitope-containing compounds immobilized on a solid support with labeled anti-hlg antibodies in suitable containers, along with any other reagents (e.g., wash solutions, enzyme substrate) or other materials required for the conduct of the assay. The reagents are usually premeasured for ease of use. An optional component of the kit is a set of instructions describing any of the above immunoassay methods. For example, a kit for a direct assay can comprise histone-H2B epitope immobi¬ lized on a solid immunoassay support and a container comprising labeled antibody to human immunoglobulin (Ig), as well as the other reagents mentioned above. A kit for a competition assay will contain predetermined amounts of H2B epitope and labeled anti-H2B antibody. The H2B epitope can be in a soluble form (or in solu¬ tion) for a homogeneous assay, or in immobilized for a heterogeneous assay.

EXAMPLES

Results

To evaluate antibody binding to T-cells in AIDS, serum samples were obtained from 80 subjects (Table 1). The control group of 41 patients included 10 heterosexual subjects (7 women and 3 men) with ITP, sev¬ enteen asymptomatic homosexual men and fourteen healthy hemophiliacs. None of the heterosexual controls had anti-HIV antibody in their sera, as determined by an indirect immunofluorescence assay (IFA) and immunoblotting. Levy et al. (1984), supra; Carlson et al. (1985) J.A.M.A. 25_3:3405-3408. Seventeen of the homosexual and hemophiliac controls were negative for serum anti-HIV antibody, while fourteen were HIV antibody-positive. The patient groups consisted of 27 homosexual men and 12 hemophiliacs (11 men and 1 woman) with lymphadenopathy, ITP or AIDS. Each of these patients had serum antibody against HIV. All AIDS patients met the criteria for the disease established by the Centers for Disease Control. Selik et al. (1984) Am. M. Med. 2_6ϊ493-500.

The mean helper-suppressor T-cell ratio of the 17 homosexual controls was 1.8 (range 1.2-3.5), while the mean T-cell ratio of the patients with AIDS and AIDS-related conditions was 0.3 (range 0.1-0.9), as determined using monoclonal antibodies and flow cytometry. Stites et al. (1986), supra. All of the individuals with ITP had elevated platelet-associated immunoglobulin, as demonstrated by a flow cytometric technique. Strieker et al. (1985) N. Engl. J. Med. 313:1375-1380. The six homosexual men with ITP and ten of thirteen homosexual AIDS patients who were studies also had serum antibody against a 25,000-dalton platelet protein, as previously described. Jd. This antibody was not detected in sera from any of the heterosexual or hemophiliac patients with ITP.

Using a sensitive immunoblot technique, it was found that patients with AIDS and AIDS-related condi¬ tions produce a serum antibody that binds to an antigen on normal lectin-stimulated T-cells and on non-mitogen-stimulated, HIV-infected T-cells (see Table 1). The antibody was not detected in sera from 40 of 41 controls, but it was found in sera from 37 of 39 patients with AIDS and AIDS-related conditions (Table 1). The target T-cell protein was not found on unstimulated or uninfected T-cells or other blood cells. The antigen is an 18,000-dalton protein that appears to be restricted to stimulated T-cells of the CD4+ pheno- type. It was also detected on lectin-stimulated or virus-infected cells from the human helper/inducer T-cell line, HUT-78.

The antigen is expressed on the T-cell sur¬ face, since antibody eluted from stimulated HUT-78 cells bound to the 18,000-dalton protein on immunoblots. The antibody was shown to bind via the F(ab)2 portion to autologous HIV-infected T-cells.

This finding indicates autoantibody rather than immune complex binding to the target antigen. The 18,000-dalton T-cell protein appears to be distinct from the 25,000-dalton platelet protein recognized by serum antibody from homosexual men with ITP. It also appears to be an integral T-cell protein this is distinct from HLA antigens or HIV antigens defined by available immunoblot reagents. It was also demonstrated that IgG fractions of AIDS serum containing the anti-T-cell autoantibody selectively inhibited CD4+ T-cell prolifer¬ ation in vitro. Virus-free purified IgG from AIDS patients suppressed proliferation of lectin-stimulated, uninfected CD4+ cells by 58-62%. No inhibitory effect of the IgG on proliferation of lectin-stimulated CD8+ cells was seen. Furthermore, HUT-78 cell antibody eluates from AIDS patients were cytotoxic for 60-70% of lectin-stimulated, uninfected CD4+ cells in the presence of complement. CD8+ cells were unaffected by the eluates. These findings suggest that anti-T-cell autoantibody may play an important role in the develop¬ ment of selective T-cell abnormalities in AIDS. Fur¬ thermore, the autoantibody was associated with clinical disease rather than simple exposure to HIV. Antibody against the 18,000-dalton T-cell antigen was not detected in sera from 40 to 41 heterosexual or "healthy" homosexual and hemophiliac controls, even though 14 of these controls had evidence of HIV infection (Table 1). The 17 homosexual controls had normal helper/suppressor T-cell ratios (mean, 1.8). In contrast, 95% of patients with AIDS or AIDS-related conditions were found to have antibody against the 18,000-dalton T-cell antigen (Table 1). These patients had decreased helper/suppressor ratios (mean, 0.3). Thus the presence of autoantibody against T-cells was associated with growth inhibition and cytotoxicity of lectin-stimulated CD4+ cells in vitro and with clinical disease and decreased helper/suppressor ratios in vivo.

The nature of the 18,000-dalton T-cell antigen is uncertain. It is distinct from the pl7/18 core pro¬ tein of HIV, since the latter is detected under condi¬ tions of disulfide bond reduction on immunoblots. In contrast, the T-cell antigen cannot be detected by immunoblotting after reduction of disulfide bonds. It was also found that the T-cell antigen is expressed in HUT-78 cells after infection with murine xenotropic retrovirus. Therefore expression of the antigen appears to be induced by other viral infections. Preliminary N-terminal amino acid sequence data indicates that the 18,000-dalton T-cell protein does not share homology with any previously defined polypeptide. Thus the tar¬ get T-cell protein may represent a previously uncharacterized CD4+ T-cell activation antigen that is rendered immunogenic by HIV infection, as discussed below.

What is the significance of an autoantibody directed against stimulated CD4+ T-cells in AIDS? First, the antibody does not appear to be an epiphenomenon of the disease, since it is found in homo¬ sexual men and hemophiliacs with conditions that are considered to be part of the "pre-AIDS" complex, such as lymphadenopathy and isolated ITP. Thus appearance of the autoantibody precedes the onset of AIDS. Second, induction of this autoantibody may in part explain how patients progress from HIV infection to AIDS. Autoantibody against stimulated T-cells may be induced by viral infection through interaction of the virus with cell surface components such as lymphocyte CD4 or HLA antigens or by viral "molecular mimicry" of cellular proteins. See, e.g., Mann et al. (1983) J. Immunol. 131:2021-2024; Klatzmann et al. (1986) Nature 319:10-11. The latter mechanism has been postulated in other autoimmune diseases. Billings et al. (1983) Proc. Natl. Acad. Sci. USA 8.0:7104-7108; Fujinami et al. (1983) Proc. Natl. Acad. Sci. USA 80.:2346-2350; Walker et al. (1986) Lancet ϋ:605-607. Once the autoantibody is pre¬ sent, progressive destruction of CD4+ T-cells may occur on the basis of concomitant T-cell activation and B-cell deregulation. Lane et al. (1983), supra; Wachter et al. (1986) Lancet i:97; Yarchoan et al. (1986) J. Clin. Invest. 78_: 39-447. Thus an autoimmune response against stimulated T-cells may play an important role in destruction of the immune system in AIDS. Third, recog¬ nition of autoantibody production against stimulated T-cells may lead to new directions in AIDS therapy aimed either at the autoantibody response itself or at inhib¬ iting expression of the target antigen on T-cells. For example, cyclosporin A inhibits the expression of at least two T-cell surface antigens. This inhibition probably constitutes an important immunomodulatory effect of the drug, although its mechanism of action remains unidentified. See, e.g., Groenewegen et al. (1985) Transplantation 4j0:21-25; Gaucht et al. (1986) Proc. Natl. Acad. Sci. USA 83:6430-6434. Experimental Methods

Detection of anti-T-cell antibody by immunoblotting. Human T-cells were obtained from con¬ trol blood by Ficoll-Hypaque separation followed by sheep red cell rosetting. Anderson et al. (1978) Blood 5^,:57-67. The cells were grown in the presence or absence of phytohemagglutinin (PHA), 4 μg/ml, for 5 days and then harvested and solubilized in 2% sodium dodecyl sulfate (SDS) sample buffer at 5 x 10⁷ cells/ml (5 mg/ml). B-cells not exposed to PHA were harvested after T-cell separation and solubilized at the same concentra¬ tion in 2% SDS. Monocytes were separated by adherence to plastic plates, and granulocytes were prepared by sucrose gradient separation. Jtd. These cells were solubilized in 2% SDS at 5 mg/ml. Platelets were pre¬ pared as previously described in Strieker et al. (1985), supra, and solubilized at 5 x 10^ cells/ml (5 mg/ml). Immunoblotting was performed as previously described, lά. ; Strieker et al. (1985) J. Clin. Invest. 26:1274-1278. Briefly, 50 microliters of the target cell preparations were subjected to SDS-polyacrylamide gel electrophoresis (SDS-PAGE) using 12% gels under non-reducing conditions. Pre-stained molecular weight standards (BRL, Gaithersburg, MD) were electrophoresed with the samples. After transfer to nitrocellulose and blocking with 5% gelatin in Tris-buffered saline (TBS), the nitrocellulose was incubated with serum, F(ab)2 fragments (3 mg/ml) or T-cell eluates (1 mg/ml) diluted 1:50 in TBS containing 1% gelatin. Antibody binding was detected using a biotin-avidin-peroxidase system.

Biotinylated F(ab)2 goat anti-rabbit IgG was used to detect rabbit antibody binding, while biotinylated F(ab)2 goat anti-human κ/λ light chain was used to detect patient F(ab)2 binding on the nitrocellulose.

In a representative immunoblot using serum from a homosexual man with ITP, serum antibody did not bind to proteins from unstimulated peripheral blood

T-cells. In contrast, serum antibody bound to an

18,000-dalton protein in PHA-stimulated uninfected T-cells. This protein was not detected on B-cells, monocytes, platelets, or granulocytes. The 18,000-dalton antigen was also found on unstimulated control T-cells that were infected with human immunodeficiency virus (HIV) after isolation from peripheral blood. The patient's own serum also con¬ tained antibody against a previously described

25,000-dalton platelet antigen.

Specificity of antibody binding to CD4+ cells. Freshley isolated peripheral-blood T-cells were sepa¬ rated by panning into Leu 3-positive (CD4+) and Leu 2-positive (CD8+) populations, as defined by monoclonal antibodies (Becton-Dickinson, Mountain View, CA) . Engelman et al. (1981) J. Immunol. 127:2124-2128. The purity of each population was >90%, as determined by flow cytometry using these antibodies. The cells were grown in PHA for 5 days and then solubilized in SDS at 5 7 x 10'd cells/ml. Immunoblotting was performed as described above. Serum antibody from a homosexual AIDS patient bound to the 18,000-dalton antigen on stimulated CD4+ cells, but not on stimulated CD8+ cells. When T-cell proteins were subjected to disulfide bond reduc¬ tion using dithiothreitol prior to SDS-PAGE and immunoblotting, antibody binding to the 18,000-dalton protein on CD4+ cells was no longer detected. TABLE 1

SUBJECTS SERUM ANTIBODIES Anti-HIV Anti-18 KD

Results of serum antibody testing in control and patient groups. The number positive/number tested is presented in Table 1. The control group consisted of 10 heterosexual subjects with ITP and 31 asymptomatic homosexual men and hemophiliacs. Seventeen of the asymptomatic controls were HIV antibody-negative (10 homosexual, 7 hemophiliac) while 14 were HIV antibody-positive (7 homosexual, 7 hemophiliac). The patient group consisted of six patients with lymphadenopathy (4 homosexual, 2 hemophiliac), 10 patients with ITP (6 homosexual, 4 hemophiliac), 16 patients with Pneumocystis carinii pneumonia (10 homo¬ sexual, 6 hemophiliac) and seven patients with Kaposi's sarcoma (all homosexual). Serum antibody against HIV was detected by IFA and immunoblotting [Levy et al. (1984), supra: Carlson et al. (1985), supra] , while serum antibody against the 18,000-dalton T-cell protein (18 KD) was detected as described above.

Characterization of anti-T-cell antibody.

Binding of F(ab)2 fragments and T-cell eluates to the human helper/inducer T-cell line, HUT-78. Levy et al. (1984), supra. Uninfected HUT-78 cells were grown in the presence or absence of PHA (4 μg/ml) for 5 days and then solubilized at the same concentration as fresh T-cells. In addition, non-PHA-stimulated HUT-78 cells infected with HIVgF2 (formerly AIDS-Associated Retrovirus-2, or ARV2) were grown under the same condi¬ tions. Viral infection was demonstrated by a positive reverse transcriptase assay. Id. Purified IgG and F(ab)2 fragments were prepared from the serum of homo¬ sexual ITP and AIDS patients as previously described. Strieker et al. (1985) N. Engl. J. Med., supra. The F(ab)2 fragments at a concentration of 3 mg/ml were shown to be free of residual IgG by SDS-PAGE.

Serum incubation with lectin-stimulated HUT-78 cells and subsequent antibody elution was performed by the acid elution technique described by Falus et al. (1982) Clin. Exp. Immunol. £7:103-109. PHA-stimulated HUT-78 cells (1 x 10') were incubated with one millili- ter of control or AIDS serum for one hour at room tem¬ perature. The cells were washed in phosphate-buffered saline (PBS), and antibody was eluted using 0.1 M glycine, pH 2.5. The eluate was neutralized with 1 M Tris and dialyzed against PBS. Analysis of the eluates by SDS-PAGE revealed two bands corresponding to IgG and albumin. The IgG concentration in the eluates was determined by a standard rate nephelometric technique. See, e.g., Basic and Clinical Immunology (Stites et al. eds. 1987).

Immunoblotting was performed as described above. Serum antibody from a homosexual ITP patient bound to an 18,000-dalton protein on PHA-stimulated HUT-78 cells and on HIV-infected HUT-78 cells. Unstimulated, uninfected HUT-78 cells failed to express the 18,000-dalton antigen. In binding F(ab)2 fragment (3 mg/ml) from a homosexual ITP patient to proteins from PHA-stimulated, HIV-infected HUT-78 cells, the F(ab)2 fragments bound to the same 18,000-dalton antigen. The T-cell antibody eluate (1 mg/ml) from a hemophiliac AIDS patient failed to bind to unstimulated HUT-78 cells. In contrast, the eluate bound to the 18,000-dalton antigen on PHA-stimulated HUT-78 cells and on PHA-stimulated control CD4+ cells.

Characterization of the target T-cell protein in relation to other cellular antigens. Serum and autologous T-cells were obtained from a hemophiliac AIDS patient. The T-cells were shown to be infected with HIV, as determined by a reverse transcriptase assay and IFA. Levy et al. (1984), supra. These cells were grown in the absence of PHA for 5 days and then solubilized in SDS. Rabbit polyclonal antisera against HLA Class I and HLA Class II framework antigens were obtained from Pel-Freez Biologicals, Rogers, AR. Rabbit antiserum reactive with 82~^mic^ro9lobulin was obtained from Dako Corp., Santa Barbara, CA. In addition, rabbit antiserum was raised against sucrose-purified HIV. .Id. By immunoblotting, this antiserum was shown to recognize six viral proteins separated by SDS-PAGE under condi¬ tions of disulfide bond reduction: the core (p24-25), envelope (gp41, gpl20), and polymerase (p31, p51, p66) proteins. Carlson et al. (1985), supra: Levy et al. (1985) Ann. Int. Med. 103:694-699.

Immunoblotting was performed as described above. A serum antibody from the hemophiliac AIDS patient bound to the 18,000-dalton protein on autologous HIV-infected T-cells. Rabbit antiserum was incubated with proteins from PHA-stimulated, HIV-infected T-cells. The 18,000 T-cell antigen was distinct from the 40,000 dalton HLA Class I (ABC) framework antigen and from the 28,000-dalton HLA Class II (DR) framework antigen. Fur¬ thermore, the 18,000-dalton protein was not recognized by antibody specific for -microglobulin (M_r=12,000). Binding of rabbit anti-HIV antiserum to sucrose-purified HIV proteins (core, envelope, and polymerase) immobi¬ lized on nitrocellulose was also shown.

Effect of purified IgG on proliferation of PHA-stimulated T-cells isolated from peripheral blood. T-cells were incubated in the presence of purified IgG according to a modification of the method of Kiprov et al. supra. Panned Leu 3-positive (CD4+) or Leu 2-positive (CD8+) T-cells (2 x lOVml) were grown in the presence of PHA (4 μg/ml) and purified IgG (200 μg/ml final concentration) for five days at 37°C. Incu¬ bations were performed in triplicate using T-cells from two normal donors in separate experiments. Cell prolif¬ eration was assessed after five days by ³H-thymidine incorporation. .Id. Purified IgG from two homosexual AIDS patients inhibited CD4+ cell uptake of ^H-thymidine by a mean of 62% compared to control IgG (range 48-77% inhibition). When the concentration of IgG in the cul¬ ture medium was decreased, the inhibitory effect of the AIDS samples on cell proliferation was abolished. Puri¬ fied IgG from two other homosexual AIDS patients inhibited CD4+ cell proliferation by a mean of 42% com¬ pared to proliferation of CD8+ cells (range 35-49% inhibition). Uptake of ³H-thymidine by CD8+ cells in the presence of purified IgG from the AIDS patients

(mean, 5843 cpm) was comparable to the uptake of "i

^H-thymidine by CD4+ cells in the presence of purified control IgG (mean, 6100 cpm).

Cytotoxicity of HUT-78 antibody eluates. Eluates were prepared as described above. The eluates were adjusted for IgG concentration (200 μg/ml) and added to cultures of PHA-stimulated CD4+ or CD8+ peripheral-blood T-cells in the presence or absence of a 1:5 concentration of rabbit complement (Pel-Freez, Rog¬ ers, AR) . Cytotoxicity was assessed by propidium iodide uptake after 24 hours using a flow cytometric technique. Stites et al. (1986), supra. Two separate experiments were performed in triplicate using different T-cell donors, and eluates from two AIDS patients were studied. Seventy percent of the PHA-stimulated T-cells were acti¬ vated, as assessed by binding of monoclonal anti-HLA-DR antibody to the cultured cells (range, 68-73%). Incuba¬ tion of CD4+ or CD8+ cells with complement alone or with a control antibody eluate (200 μg/ml) in the presence of complement failed to induce cytotoxicity. Incubation with the AIDS patient antibody eluates in the absence of complement also failed to induce cytotoxicity. However, incubation of CD4+ cells with and AIDS patient antibody eluates in the presence of complement induced cytotoxicity in 65% of these cells (range, 60-70%). In contrast, no eytotoxic effect was seen in the CD8+ cells. As a positive control, monoclonal antibody (200 μg/ml) against pan-T Leu 1 marker was shown to be eytotoxic for 90% or more of both CD4+ and CD8+ cells in the presence of complement. The AIDS patient antibody eluates had no eytotoxic effect on CD4+ cells grown in the absence of PHA.

It was then determined that the 18,000-dalton antigen described above is the histone protein H2B. This was discovered by sequencing the 18,000-dalton histone, purified by standard techniques, and determin¬ ing a partial amino acid sequence. It was also deter¬ mined by showing cross-reactivity between the 18,000-dalton histone and anti-H2B antibodies. See, e.g., Mathieu et al. (1981) J. Cell Biol. 9.1:135 (rabbit anti-H2B antibodies). It was then shown that commer¬ cially available H2B preparations bound the autoantibody in immunoblots. While the commercial preparation of H2B showed a molecular weight of 16,000 daltons on a gel, this is probably due to the- differences in source. The 18,000-dalton antigen is isolated from cells and may have additional proteins or carbohydrates associated with it, such as the other histones to which H2B com¬ plexes. It is also known that the presence of a large number of proteins in the gel (as is obtained from solubilized cells) can affect the rate of migration of particular proteins on the gel. Either of these reasons would explain the slightly higher M.W. observed above.

While the present invention has been described above by way of illustration, it is to be understood that the scope of the present invention is not limited to the specific embodiments described above, and should be defined by the following claims.

Claims

1. A method for diagnosing the clinical stage of a human immunodeficiency virus (HIV) infection by determining the presence of anti-H2B antibodies in a patient comprising:

(a) providing an antibody-containing sample from human patient infected by HIV; and

(b) determining whether said sample contains ant bodies to histone H2B.

2. A method according to claim 1 wherein said sample is contacted with histone H2B epitopes and the presence of anti-H2B antibodies is determined in an immunoassay.

3. A method according to claim 2 wherein said immunoassay is a direct assay.

4. A method according to claim 2 wherein said immunoassay is a competition assay.

5. A method according to claim 2 wherein said immunoassay is an ELISA.

6. A method according to claim 2 wherein: (i) said sample is serum;

(ii) said histone H2B epitopes are immobilized a solid support; and

(iii) the presence of anti-H2B antibodies is determined after said contacting by reaction wit labeled anti-human Ig antibodies.

7. An immunoassay for diagnosing the clinical stage of an HIV infection by determining the presence of anti-H2B antibodies in a patient comprising:

(a) providing serum from a human patient infecte by HIV;

(b) contacting said serum with a sample of immob lized histone H2B epitopes to provide a serum-contacted sample;

(c) contacting said serum-contacted sample with labeled anti-human Ig antibodies to provide label-contacted sample; and

(d) determining the presence of the label on sai label-contacted sample.

8. The method of claim 7 further comprising the steps of washing and said serum-contacted sample prior to step (c) and washing said label-contacted sam¬ ple prior to step (d) .

9. The method of claim 7 wherein step (d) comprises an enzyme reaction producing a color change.

10. The method of claim 7 wherein step (d) comprises detecting a fluorescent label.

11. The method of claim 2 wherein said H2B is non-human H2B.

12. The method of claim 7 wherein said H2B is on a human T-cell.

13. A kit for an immunoassay to diagnose the clinical stage of an HIV infection by determining the presence of anti-H2B antibodies in a patient comprising: (a) a solid immunoassay support having immobiliz thereon histone H2B epitopes; and

(b) a container comprising labeled anti-human Ig antibody.

14. The kit of claim 11 wherein said solid immunoassay support is a plastic plate containing recep¬ tacles for a serum sample coated with isolated histone H2B, said labeled anti-human Ig antibody is labeled directly or indirectly by an .enzyme, and said kit fur¬ ther comprises both a container comprising a substrate for said enzyme and a container comprising a wash fluid.

15. The kit of claim 11 wherein said H2B epitopes are non-human H2B.

16. The kit of claim 14 wherein said anti-human Ig antibody is biotinylated and said enzyme is conjugated to avidin.

17. A kit for an immunoassay to diagnose the clinical stage of an HIV infection by determining the presence of anti-H2B antibodies in a patient comprising:

(a) a predetermined amount of histone H2B epitope; and

(b) a predetermined amount of labeled anti¬ body to said histone H2B epitope.

18. A kit according to claim 17 wherein said histone H2B epitope is immobilized on a solid immunoassay support.

19. The kit of claim 17 wherein said H2B epitopes are non-human H2B.