UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Alloimmunity, autoimmunity, and AIDS Grant, Michael David 1989

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1989_A6_7 G76.pdf [ 6.06MB ]
JSON: 831-1.0097452.json
JSON-LD: 831-1.0097452-ld.json
RDF/XML (Pretty): 831-1.0097452-rdf.xml
RDF/JSON: 831-1.0097452-rdf.json
Turtle: 831-1.0097452-turtle.txt
N-Triples: 831-1.0097452-rdf-ntriples.txt
Original Record: 831-1.0097452-source.json
Full Text

Full Text

ALLOIMMUNITY, AUTOIMMUNITY, AND AIDS By MICHAEL DAVID GRANT B.Sc, The University of British Columbia, 1978 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Microbiology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August 1989 © Michael David Grant, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract The sensitivity of the human immune system to the retrovirus HIV is difficult to explain on the basis of viral cytopathicity. AIDS develops often long after initial HIV infection in spite of a vigorous and sustained immune response against the virus which effectively contains viral replication. Although more virulent strains of HIV can be isolated from immunodeficient persons, there is no evidence that these strains predominate in vivo or that a large increase in production of infectious virus accompanies progression to disease. There is as yet no satisfactory mechanism to explain the immunosuppression, T4 cell depletion, autoimmunity, and immunodeficiency associated with HIV infection. An hypothesis that AIDS results from immune responses to HIV gp 120 and to allogeneic cells stems from recent developments in the symmetrical network theory. It is proposed that T-cell selection processes produce a T-cell idiotypic network "centrepole," which bears an internal image of self MHC class II within T-cell receptors. This network centrepole interacts with the anti-class II helper cell repertoire to stabilize both its own expression and expression of the helper T-cell repertoire. The idiotypes involved in this mutual stabilization are related to certain idiotypes present on allogeneic lymphocytes and to the envelope glycoprotein of HIV gpl20. Through its relationship to CD4 protein, and hence class II MHC, HTV gpl20 is thought to mimic the T-cell network centrepole. Anti-host class II receptors on allogeneic lymphocytes induce antibodies which also mimic the proposed centrepole. The immune response to gpl20 is directed against the centrepole and against the antibodies (anti-anti-class II) induced by the receptors of allogeneic lymphocytes. The hypothesis evaluated in this thesis project is that the immune responses described above synergize in an attack on both the centrepole and the helper cell repertoire, resulting in aberrant immune regulation, autoimmunity, and eventually, AIDS. Sera from persons with AIDS or at risk of AIDS were examined for antibodies implicated in the above scheme. Anti-anti-CD4/anti-gpl20 antibodies (putative anti-centrepole) were found in only a small minority of subjects and did not correlate with disease, while anti-anti-class II antibodies were almost never detected. A n t i - M H C class I antibodies, reflecting alloimmunity, were associated with H I V infection and to some extent with disease progression. Autoantibodies against denatured collagen, reflecting autoimmunity, were found in almost all AIDS patients. The prevalence of these autoantibodies increases in HIV infection and with disease expression. Antibodies against denatured collagen show an interesting distribution which suggests they are related to the idiotypic determinants involved in the pathogenesis of AIDS and other similar immune disorders. The specificity of these antibodies suggests they arise through immunoregulatory defects induced through idiotypic network interactions. Table of Contents Abstract ii. Table of Contents iv. List of Tables v. List of Figures vi. List of Abbreviations viii. Acknowledgement ix. Introduction 1. Materials and Methods 16. Results 21. Discussion 54. Bibliography 69. List of Tables Table Title I. Serum IgG Levels and Correlation Coeffecients. LI. Mean ELISA OD values of group ± sem. LTI. Significance Values. VI List of Figures Figure Title Page 1. Anti-anti-CD4 ELISA values of homosexual, hemophiliac, and 22. control sera. 2. Inhibition of anti-anti-CD4 ELISA with HIV gp 120. 23. 3. Anti-anti-CD8 ELISA values of homosexual, hemophiliac, and 25. control sera. 4. Anti-anti-CD4 and anti-anti-CD8 ELISA values of rheumatoid 26. arthritis sera. 5. Anti-anti-gp 120 ELISA values of homosexual, hemophiliac, 27. and control sera. 6. Anti-gelatin ELISA values of homosexual, hemophiliac, and 29. control sera. 7. Inhibition of anti-gp 120 ELISA activity of AIDS sera with sera 36. from autoimmune HIV- homosexuals and controls. 8. Titration of anti-gelatin ELISA activity. 38. 9. Molecular sizing and immunoblot analysis of gelatin-agarose 39. affinity purified material. 10. Inhibition of anti-gelatin ELISA with human placental 40. collagens. 11. Inhibition of anti-gelatin ELISA with fibronectin. 42. 12. Immunoblot analysis of anti-collagen reactivity. 44. v i i Figure Title Page 13. Correlation of anti-gelatin ELISA values with serum IgG levels. 45. 14. Anti-collagen reactivity profiles of homosexual anti-gelatin 46. positive sera and control sera. 15. Inhibition of anti-gelatin ELISA with C l q . 48. 16. Immunoblot reactivity of purified anti-gelatin antibodies. 49. 17. Non-specific binding of antibodies to C l q immobilized on 50. nitrocellulose. 18. Non-reactivity of purified anti-gelatin antibodies with the 51. isolated collagen-like sequences of C l q . 19. Augmentation of anti-Clq reactivity by IgG capture of C l q . 53. List of Abbreviations Abbreviations Meaning A c h acetylcholinesterase AIDS acquired immune deficiency syndrome A Z T azidothymidine B S A bovine serum albumin CDC Centre for Disease Control E D T A ethylene diamine tetraacetic acid ELISA enzyme linked immunosorbent assay G V H D graft versus host disease H I V human immunodeficiency virus Ig immunoglobul in kd kilodaltons M H C major histocompatibility complex OD optical density P A G E polyacrylamide gel electrophoresis PBS phosphate buffered saline R A rheumatoid arthritis SDS sodium dodecyl sulphate SLE systemic lupus erythematosus SRBC sheep red blood cells Tris tris hydroxymethyl aminomethane Acknowledgement I gratefully acknowledge the support, encouragement, and guidance of Dr. Geoffrey Hoffmann, whose determination and insight made this project possible. Suggestions and evaluation provided by committee members Dr. F. Tufaro and Dr. J. Chantler have been invaluable towards completion of this project as has the cooperation and assistance of the management and staff of the B. C . Provincial Health Lab. I thank Dr. Michael Weaver, Dr. Christos Tsoukas, and the B. C . Provincial Health Lab for providing serum samples and thank the N H R D P for financial support during the second year of my Master's program. Introduction Shortly after recognition of a novel acquired immunodeficiency syndrome (AIDS), a new human retrovirus (human immunodeficiency virus-1, ELTV-l) was isolated from lymphocytes of affected individuals (1, 2). With the optimization of culture conditions and development of viral antibody and nucleic acid detection tests, HIV infection can now be demonstrated in essentially all AIDS patients. A clear association between HIV and AIDS is established, as is HTV's tropism for the T4 lymphocytes which disappear during progression to AIDS. H I V uses the C D 4 protein, expressed on the surface of lymphocytes, monocytes, dendritic cells and glial cells, as a receptor, and is cytopathic for some CD4+ cell lines under some in vitro conditions (3). Cel l death in vitro occurs through expropriation of cellular metabolic machinery, cytolysis associated with the release of viral particles, or by fusion of cellular membranes (syncytia formation) (4-6). Infection and syncytia formation are inhibited in vitro by antibodies against epitopes of CD4 or HIV gpl20 involved in the CD4/gpl20 interaction (7, 8). A number of regulatory genes found in H I V imply a complex mechanism of life cycle regulation. Transcriptional and translational enhancer and inhibitor elements complement the usual retroviral genes encoding structural, envelope, and catalytic proteins (9). Depending on the presence of proteins capable of interacting with HIV enhancer sequences, HIV may grow slowly, rapidly, or remain latent (10). Although investigators generally agree on HIV as the cause of AIDS, the pathogenesis of AIDS is not simply related to the biology of H I V infection. Biochemical evidence does not support a direct role for HIV in the T4 lymphocyte depletion which characterizes AIDS. In situ hybridization and polymerase chain reaction studies reveal only 1/1000 to 1/10,000 T-lymphocytes infected with HIV at any stage of disease (11, 12). There is also little evidence of in vivo cell fusion, especially of T-lymphocytes (13). A number of indirect mechanisms for HIV induced T4 depletion have been proposed, but no firm basis for this depletion and resultant immunodeficiency has been clinically or experimentally established. Some immunodeficiency arises even before depletion of T4 cells (14). .Infection with HIV occurs primarily via anal sex or parenteral exposure. In order to replicate and establish infection, H I V must contact permissive CD4+ cells found mainly in the blood and lymph. During acute infection, lymphadenopathy and other symptoms similar to those of mononucleosis often develop and there can be considerable production of virus (15, 16). At this stage, no significant depletion of T4 lymphocytes occurs, but the T4 /T8 ratio decreases due to increased numbers of CD8+ lymphocytes (17). Symptoms accompanying primary infection usually resolve as a specific immune response develops, but HIV establishes a persistent, mainly latent infection. Despite the apparent success of the immune system in combatting HIV, some insidious form of damage to the immune system actually occurs from this point onwards. Coincident with the development of H I V neutralizing antibodies, the absolute number of T4 lymphocytes begins to fall in HTV infected individuals. As the T4 number continues to fall, autoimmunity, neurological symptoms, immunodeficiency, malignancies, and opportunistic infections develop (18). There is no clear evidence that a rise in infectious virus titres precipitates these events, but increased serum levels of p24, the HIV core protein, tend to foreshadow or parallel clinical deterioration. Serum levels of fJ-2 microglobulin and curiously, neopterin, a marker of T4 lymphocyte activation, also tend to foreshadow AIDS onset (19). Paradoxically, infectious HIV becomes most difficult to detect in the final stages of AIDS (20). 3 The protracted course from infection with H I V to AIDS suggests slow persistent growth such as occurs with other animal lentiviruses. Success for these viruses depends on evasion of the immune system by antigenic variation, latency and reactivation, and or immunosuppression induced by a variety of mechanisms. H I V exhibits all these characteristics, but perhaps the most important in the pathogenesis of AIDS is the mechanism of immunosuppression. In relation to other human immunosuppressive viruses (eg. measles, cytomegalovirus, Epstein-Barr virus) H I V is present at extremely low levels (21). Since some immunodeficiency is seen prior to T4 cell depletion, an indirect mechanism of immunosuppression appears to operate at least early in infection. The nature of this mechanism is unknown, but may involve secreted viral proteins, induced or reduced secretion of cellular regulatory proteins, interference with antigen presentation, disruption of substituent T4 lymphocyte subsets, idiotypic network interactions, or some combination of the above. The immunodeficiency of human H I V infection is especially mysterious in light of present animal models of infection. Chimpanzees are the only animal so far experimentally infected with HTV and although infectious virus is produced, no evidence of immunodeficiency is observed (22). Failure of H I V to induce disease in chimpanzees and the indirect mechanism of immunosuppression and T4 cell depletion in human HIV infection raises the possibility that HIV infection alone is insufficient to cause AIDS. This possibility is also supported by the predominance of AIDS within particular risk groups. A n important question is whether this predominance reflects only risk of exposure or whether it reflects risk group specific cofactors active in disease progression. In North America, AIDS is largely confined to blood product recipients, intravenous drug abusers, and male homosexuals (23). In Africa, the presumed continent of origin of HIV, infection and AIDS occur equally 4 in males and females. The primary means of H I V transmission appear to be exposure to a contaminated blood supply and heterosexual intercourse between carriers of other sexually transmitted diseases (24). Epidemiological studies on cohorts of HIV infected homosexuals in North America reveal no correlation between rate of progression to AIDS and exposure to other sexually transmitted diseases (25). Therefore, investigators acknowledge risk factors for H I V infection, but assume a similar prognosis for everyone once infection with HTV occurs. This assumption is, however, yet to be substantiated by epidemiological studies of cohorts of H I V infected persons not within the known risk groups. Despite agreement on overall average rates of progression to AIDS, there is tremendous individual variation in the clinical course of disease. A number of proposed cofactors in AIDS pathogenesis relate to enhanced viral replication. These possible cofactors include immune stimulation, immunosuppression, virulence increasing mutations, and coinfection with pathogens capable of activating HTV replication, or with other strains or variants of HIV. The significance of any of these cofactors rests on the demonstration that increased viral replication causes the T4 depletion which characterizes AIDS. The predominance of homosexual AIDS cases in North America is believed to reflect the route of introduction of H I V to the continent and extremely rapid early spread of the virus via anal sex and promiscuity. A corollary of this belief is that once the virus enters the general population, over time the proportion of heterosexual AIDS cases will gradually increase. So far, this is happening only within the intravenous drug abusing population, and it is still unclear to what degree HIV is distributed outside of the recognized risk groups. A n important question remains whether the lack of AIDS cases outside the primary risk groups reflects low risk of infection or whether factors usually responsible for exposure to HIV also entail increased susceptibility to the effects of HIV. H I V infection usually occurs concurrently with exposure to allogeneic proteins, including M H C antigens, lymphocyte receptors, and in some cases viable cells. Exposure to these agents induces an immune response which is theoretically directed against the immune response to the envelope glycoprotein of HIV, gpl20 (26). It has been proposed that these immune responses include a response directed against immune system regulatory elements and that synergy between anti-HIV gpl20 immunity and complementary immunity induced by allogeneic cells induces destabilization of the immune system idiotypic network. Some homosexuals, hemophiliacs, drug addicts, and transplant recipients show evidence of disturbed immune regulation in the absence of infection with H I V (27). Conceivably, this evolves from exposure to allogeneic cells and the immune response to allogeneic cells impels the pathogenicity of HIV through complementarity to anti-gpl20 idiotypes. There is considerable indirect evidence implicating the immune response in the pathogenesis of AIDS. AIDS develops despite humoral and cell mediated responses against HIV. Although profound immunodeficiency is the hallmark of AIDS , most patients show signs of prolonged immune system activation. Hypertrophy of lymphoid organs and lymphocyte hyperplasia occur early in HTV disease while hypoplasia and atrophy are late features (28). Despite the poor response of B and T-lymphocytes from H I V infected persons to mitogenic stimulation in vitro, polyclonal activation of B-lymphocytes is an early and sustained feature of HIV infection (29). Soluble factors which suppress the proliferation of T-lymphocytes are found in the serum of AIDS patients, yet activation of T-lymphocytes is necessary for HTV replication in these cells (30, 31). Skin test anergy, first to new, and later to recall antigens, demonstrates the deficiency in cell mediated immunity associated with progressive HIV disease (32). Humoral responses to new antigens also decrease following HIV infection (33). Infected T-cells display a selective defect in signal transduction through the CD3 protein and reduced production of, and response to interleukin-2. (34,35) Even uninfected T-cells from HIV infected persons respond subnormally in T-cell colony formation assays (36). H I V infected monocytes are defective in antigen presentation and release less interleukin-1 upon stimulation than uninfected monocytes (37, 38). Natural killer cell function and antibody dependent cell mediated cytotoxicity may also decrease in HTV infection (39, 40). The initial fall in T4 lymphocytes was reported to occur specifically within the 2H4+ suppressor inducer subset (41, 42). There is a dramatic fall in this subset in severe burn victims who suffer immunodeficiency and this subset is also deficient in persons with systemic lupus erythematosus (SLE) (43, 44). A fall in the activity of this subset would likely amplify any direct effect of H I V on B-cell activation since, in vitro, 2H4+ T4 cells act via CD8+ T-cells to lessen immunoglobulin production by mitogen stimulated B-cells (45). Conceivably, this subset of T4 cells operates in vivo to suppress the production of autoantibodies, and depletion of this subset could precipitate autoimmune phenomena and immune dysregulation associated with HTV infection. Hence the loss of T4 cells could result in both a poor response to foreign antigens and increased autoimmunity. Although no biochemical basis for the selective depletion of this T4 subset has been defined, it may be that the migratory characteristics conferred by the cell surface proteins defining the specific T-cell subsets are involved. Despite the progressive immunodeficiency of H I V infection, autoimmunity is a common and constant feature of H I V disease (46). Shortly after infection, serum immunoglobulin levels increase and circulating immune complexes and autoantibodies arise (47). These features tend to persist throughout HIV infection. Decreased hemolytic complement activity and depletion of specific complement components consistent with chronic complement activation also occur (48). Similar serological and clinical features present in systemic lupus erythematosus, lepromatous leprosy, and graft versus host disease (GVHD) . This suggests that these immune disorders reflect pathogenic features of H I V infection, specifically aberrant immune system regulation (49-51). IgM and IgG autoantibodies against red blood cells, platelets, T and B lymphocytes, nuclear bodies, cardiolipin, and collagen occur in AIDS patients (52-56). Thrombocytopenia develops due to excessive clearance of platelets coated with immune complexes or antibodies against a 25 kilodalton protein in the platelet membrane (53, 57). Immune complexes eluted from platelet membranes of HIV infected persons are composed of anti-viral antibodies and anti-idiotypic antibodies, rather than anti-viral antibodies and viral antigens (58). The antibody nature of the immune complexes and the huge amounts of anti-gpl20 purified from the serum of H I V infected thrombocytopenics (200-400 u.g/mL) (58) strongly suggest the involvement of id iotypic network interactions in the induct ion of thrombocytopenia. Other clinical signs of autoimmunity include vasculitis, nephritis, encephalitis, peripheral neuropathy, and arthritis. Cutaneous and gastrointestinal manifestations also may result from autoimmunity (59). Autoimmunity associated with H I V infection is treated with prednisone or even by splenectomy, and several independent groups have noted overall biological improvement in patients after splenectomy (60, 61). Nearly half of HTV infected persons treated with cyclosporin A responded with large increases in numbers of circulating T4 cells and all showed resolution of lymphadenopathy (62). Immunoglobulin therapy, used to treat thrombocytopenia and Rh factor incompatibility, was an early treatment for AIDS which produced some benefit (63). More recently, infusion of AIDS patients with serum from healthy H I V infected persons or with autologous plasma and lysed lymphocytes has produced clinical improvements such as weight gain, increased T4 count, and decreased incidence of 8 opportunistic infections (64, 65). Alpha-interferon has been effective in the treatment of patients with kaposi's sarcoma (66, 67). Although some treatments result in lower circulating levels of HIV antigens, the effects cannot be attributed solely to inhibition of viral replication. Clinical improvements may reflect direct effects on the immune system idiotypic network, and though so far temporary, suggest that damage is preventable or reversible if appropriate treatment is initiated at an early enough stage of disease. Treatment of AIDS with azidothymidine (AZT) , an inhibitor of reverse transcriptase, is now a common procedure. A Z T prevents new infection by blocking provirus production and may to some extent also inhibit virus reactivation. A Z T decreases the level of circulating viral antigen, reduces the incidence of opportunistic infections, and often dramatically relieves neurological symptoms. However the benefits of A Z T are short-lived (68). There is no prolonged elevation of T4 levels or restoration of immune function and the toxic effects of A Z T often become overwhelming within six months. Recently, A Z T resistant strains of HIV have been observed after A Z T treatment (69). Trials are beginning which employ A Z T at earlier stages of HIV infection, but early results have produced little cause for optimism. There is definitely a pressing need for more effective, less toxic therapy. A number of investigators have speculated on autoimmune pathogenic mechanisms of AIDS. It was first suggested that T-cells binding viral antigens could become bystander Jtargets of the antiviral immune response (70). Other theories consider aberrant or anti-idiotypic immune responses resulting in immunity against M H C class II proteins or C D 4 protein (71, 72). Immunological cross-reactivity has been demonstrated between H I V envelope glycoprotein and class II proteins (73, 74) and the concept of an anti-idiotypic response against the CD4 protein is consistent with observed immune responses against other virus receptor proteins (75). Genetic predisposition to immune hyperactivity and autoimmunity hastened progression to AIDS in one cohort of H I V infected hemophiliacs (76). This also suggests the action of an (auto)immune component in AIDS pathogenesis. The most detailed autoimmune theory of AIDS takes into account the distribution of AIDS within specific groups and postulates that two immune responses are necessary to cause AIDS (77). This theory assumes a role for risk group specific cofactors and provides an explanation for the failure of HIV to induce AIDS in chimpanzees. In this theory of AIDS pathogenesis, exposure to allogeneic cells is a critical cofactor for progression to disease. The theory is based on current knowledge of T-cell selection mechanisms and on complementarity observed between HIV gpl20 and the CD4 protein. Class II restricted T-cells are selected in the thymus on the basis of weak complementarity between the T-cell receptor and self class II antigens. According to the network hypothesis, T-cells which regulate the activity of class II restricted T-cells will have receptors complementary to the anti-class II receptors. Selective pressure should impel convergence of regulatory T-cell receptors to shapes which can interact with the maximum number of anti-class II receptors. The dominant shape selected would be an internal image of class LI, which mimicks class II proteins but is unique since it is encoded by different genes and not necessarily constrained to bind short peptides. A n internally focussing T-cell idiotypic network results, providing mutual stabilization between diverse anti-class II T-cells and stringently selected regulatory T-cells with receptors mimicking class LI proteins. The important specificities in this scheme are self M H C class II, anti-class II T-cell receptors, and anti-anti-class II T-cell receptors. Normal regulation of the T-cell network is presumably maintained by the interaction of T-cells bearing the relevant specificities. In a murine model of autoimmune disease, anti-anti-self class II antibodies precede the appearance of a number of specific autoantibodies (78). These anti-anti-class II antibodies could infringe on the postulated T-cell regulatory network by virtue of their specificity and induce disregulation and the production of autoantibodies. Anti-anti-self class II antibodies can be induced in mice by immunization with allogeneic lymphocytes (79). Multiple alloimmunizations occur in blood product recipients, intravenous drug abusers, and homosexuals, therefore it is likely that people within these groups at high risk for infection with H I V also produce anti-anti-self antibodies. There is evidence for immunoregulatory disruption in persons within these groups even in the absence of H I V infection. Conceivably, disregulation occurs by the same process as in the autoimmune mouse model, and relates to anti-anti-self antibodies. HTV gpl20 induces antibodies which may be complementary to anti-anti-self class II antibodies. Gpl20 is complementary to the CD4 protein which has class II proteins as natural ligands (80, 81). Regulatory T-cell receptors, HTV gpl20, and the antibodies induced by allogeneic cells are all anti-anti-class II and M H C class II mimicking. Antibodies against gpl20 could react with the regulatory T-cell receptors and the anti-anti-self antibodies. The anti-anti-self antibodies are potentially against receptors on CD4+ T-cells. The response to gpl20 and the response to allogeneic cells are both potentially disruptive and are symmetrically directed against one another. Instead of mutual stabilization of the T-cell regulatory network by ant i -MHC receptors and anti-anti-MHC receptors, mutual destruction by two complementary sets of antibodies could result. Therefore, synergy between two symmetrically opposed immune responses is proposed to provoke network destabilization, immunological abnormalities, and eventually, AIDS. Several paradoxical aspects of AIDS are explained by this theory. Nearly all persons infected with HTV" produce antibodies against gpl20 which neutralize HTV in vitro. In vivo, however, these antibodies provide no lasting protection and may actually enhance the infectivity of HTV" (82). The failure of HTV" to induce AIDS in chimpanzees is expected in the absence of their exposure to allogeneic lymphocytes, and the extremely low viral titres observed in human AIDS patients are not inconsistent with this theory of pathogenesis. A consequence of this theory, if it is correct, is that people exposed to allogeneic lymphocytes who are vaccinated with H I V envelope glycoprotein alone, are at risk to develop AIDS. So far, in North America, AIDS is largely confined to persons exposed either to allogeneic cells or alloantibodies which cross the placenta. The corollary of this hypothesis is that HTV infected persons not exposed to allogeneic lymphocytes are at less risk to develop AIDS. A t the heart of this autoimmune theory of AIDS is the prediction of two symmetrically opposed immune responses; one a consequence of HIV gpi20 and the other a consequence of allogeneic lymphocytes. Idiotypic complementarity between these two immune responses is consistent with current knowledge of T-cell selection and restriction mechanisms and of the HTV gp l20 /CD4 interaction. A number of hypotheses arise from this theory which can be evaluated serologically. Implications of this theory extend beyond AIDS to autoimmunity and immune system regulation in general. Evaluation of immune phenomena in AIDS, in the context of this theory and current concepts in autoimmunity, may shed new light on immune system regulatory processes. 12 Objective The overall objective of this project was to identify particular idiotypes present in AIDS patients, persons at risk for AIDS and persons progressing to AIDS, which correlate with disease progression. Recent progress in the symmetrical network theory of immune regulation suggests that idiotypic links exist between alloimmune and autoimmune responses. Alloimmunity and autoimmunity are consistent features of persons within AIDS risk groups and it is suggested that these features amplify susceptibility to the effects of HIV. In the context of the symmetrical network theory, this amplification results from complementarity between anti-HIV idiotypes and the idiotypes l inking alloimmunity and autoimmunity. We wished to identify autoantibodies specific to individuals within AIDS risk groups and probe for an idiotypic relationship between these autoantibodies, anti-HIV gpl20 antibodies, and antibodies associated with alloimmunity. Cross sectional analysis of serum samples, was employed to identify antibodies associated with the development of AIDS. If antibodies associated with AIDS are linked to alloimmunity this supports the contention that idiotypically mediated immune system deregulation occurring in experimental models of autoimmunity occurs in persons progressing to AIDS, and functions as an important cofactor in AIDS pathogenesis. If particular anti-HIV gpl20 antibodies are associated with AIDS, the immune response to H I V will be implicated in immune system destruction. If idiotypic links between al loimmunity, autoimmunity, and anti-gpl20 immunity can be demonstrated, this will support the hypothesis that complementary immune responses synergize to provoke AIDS through idiotypic interactions. The experiments done to evaluate the hypotheses regarding network interactions in the pathogenesis of AIDS can be grouped into four sections. 13 1. Anti-anti-CD4 Antibodies In the network theory of AIDS, the immune response to gpl20 is implicated in the context of opposition to the anti-anti-class II immune response to allogeneic cells. The relevant determinant of gpl20 is therefore that part of the molecule that binds to the anti-class II protein, CD4. Previous work has shown that the gpl20 /CD4 interaction is prevented by certain antibodies against CD4 and that these antibodies bear an internal image of gp 120 (83, 84). If the immune response to this region is critical to the development of AIDS, then at some point in infection, all persons progressing to AIDS should make anti-gpl20 antibodies which bind to the internal image of gpl20 present on the monoclonal antibodies which prevent the gp!20 /CD4 interaction. By ELISA, sera were screened for antibodies against an appropriate anti-CD4 monoclonal and the specificity of the detected antibodies confirmed by inhibition with viral gpl20. This was done to establish whether such antibodies are made and if they are associated with AIDS. Sera were examined for reactivity with the anti-CD4 monoclonal Leu 3a and the anti-CD8 monoclonal Leu 2a. Leu 2a is related to class I M H C by virtue of complementary to the natural ligands of class I and can thus be used to probe for antibodies against internal images of class I M H C antigens. 2. Anti-anti-class LI Antibodies The important complementary immune response to anti-gpl20 is predicted to be anti-anti-class LI and should be present in the serum of HLV infected persons progressing to AIDS and in non-HIV infected persons exposed to allogeneic lymphocytes. Sera were examined for antibodies reacting with a monoclonal antibody against the non-polymorphic region of the class II H L A D R antigen and for antibodies against an anti-gpl20 monoclonal which prevents the CD4/gp l20 interaction (85, 86). This monoclonal bears an internal image of the anti-class II region of CD4 so it can also be considered anti-class LI. 3. Complementary Immune Responses In H I V infected persons progressing to AIDS, it is postulated that two immune response symmetrically directed against one another are involved. During progression to AIDS, these immune responses act to destabilize the immune system network at a fundamental level. This suggests that there will be a strong convergence within the anti-gpl20 immune response to antibodies bearing the relevant idiotype. Therefore, late in disease most of the anti-gpl20 antibodies are expected to be anti-anti-anti-class II antibodies. Serum from AIDS patients was mixed with serum from non-HIV infected persons, who had been exposed to allogeneic lymphocytes and were expressing autoantibodies, in an attempt to inhibit the anti-gpl20 ELISA activity of the AIDS serum and thus demonstrate the hypothesized complementary immune responses. 4. Anti-collagen Antibodies From early experiments in this study, it was observed that antibodies prevalent in AIDS sera reacted with gelatin in commercial monoclonal antibody diluent. Since gelatin is denatured collagen, this probably represented antibodies against denatured collagen, which occur in other immune disorders similar to AIDS. Common antibodies in these disorders may reflect a common underlying idiotypic etiology. The prevalence and levels of this activity were determined in serum samples from distinct clinical subgroups to evaluate the relationship to 15 AIDS risk and AIDS development. Specificity of the antibodies for collagen was evaluated by anti-gelatin ELISA inhibition with purified human collagens and by immunoblotting. Cross-reactivity with the collagen like regions of the complement component C l q was similarly evaluated. Anti-collagen positive sera and affinity purified anti-collagen antibodies were examined for idiotypic relationship to anti-anti-class II and anti-gpl20 antibodies implicated in the pathogenesis of autoimmunity and immunodeficiency. Materials and Methods Serum. Serum samples from 16 homosexual AIDS patients and 71 homosexuals not meeting clinical criteria for AIDS were obtained from the Vancouver Lymphadenopathy AIDS Study. Serum samples from 36 asymptomatic hemophiliacs were obtained from the B. C . Center for Disease Control. The sera from B. C. hemophiliacs had previously been tested by ELISA and western blot for antibodies against HTV. Serum samples from 19 HTV" and 12 HTV"1" hemophiliacs at various C D C stages of disease were obtained from Dr. C . Tsoukas of Montreal General Hospital. Serum samples from 12 patients with clinically defined rheumatoid arthritis were obtained from the British Columbia Arthritis Center and control sera were obtained from 11 healthy laboratory personnel. A l l sera were heat inactivated at 60°C for 30 min. Sera not previously tested for anti-HIV antibodies was tested by commercial ELISA (Dupont HTLV-III ELISA) and by recombinant gpl20 (Genentech) ELISA for antibodies against HIV. ELISA Assays. Replicate wells of plastic flat bottomed 96 well microtitre ELISA plates (Dynatech Immulon) were coated overnight at 4 ° C with 50 LiL/well of a 1 ng/LiL solution of the desired antigen in carbonate buffer. Following overnight incubation, wells were individually rinsed 4 times with PBS plus 0.5% Tween 20. Residual binding sites were then blocked by a 90 min incubation with 200 [iL of 5% fat free casein (BDH) in PBS. Wells were again rinsed 4 times with PBS Tween 20 and 100 uL of each serum sample diluted 1/50 in 1% casein was added to duplicate wells for 90 min. Samples tested against C l q were diluted in both 1% casein and 1% casein with 20 m M E D T A . Wells were then rinsed 6 times and 100 uL of alkaline phosphatase conjugated goat anti-human IgG (Dupont) diluted 1/400 in 1% casein was added to all wells for 60 min. Wells were rinsed a final 6 times and 100 uL of 1 m g / m L p-dinitrophenyl phosphate (Sigma) in 10% diethanolamine buffer was added to each well. After 45 min incubation O D 405 was read on a Dupont microplate reader. A l l steps other than overnight coating were carried out at room temperature. Specific ODs were calculated by subtracting the O D of each sample obtained in BSA (fraction IV, Sigma) coated wells from the O D obtained in the well of the test antigen. Samples were considered positive if their specific O D reading was 3 or more standard deviations above the mean specific O D of the control samples. Antigens used in this assay include gpl20 (Genentech), leu 3a and leu 2a (Becton-Dickinson), L234 an IgGl class anti-human class II mouse monoclonal antibody, obtained from the A T C C , anti-gpl20 and HTLV-Ln (Dupont), BSA, gelatin (J.T. Baker Chemical Company), human IgG, human placental collagens type I, III, IV, and V , fibronectin and complement component C l q (Sigma) and anti-human IgG (Technical Research Associates). Modifications to the basic ELISA procedure employed to measure serum IgG levels and binding to antibody bound C l q are described in later sections. Inhibition Assays. Inhibition assays were carried out to assess the ability of human collagens, fibronectin, C l q , IgG, and D N A to inhibit the anti-gelatin ELISA and the ability of gp!20 to inhibit the anti-leu 3a ELISA. Usually, various amounts of the potential inhibitors were added to different serum dilutions or dilutions of purified antibodies and incubated overnight at 4°C. ELISAs were then carried out as previously described except that the serum samples were incubated on the ELISA plates for only 30 min. Inhibition of the anti-gelatin ELISA by fibronectin was assessed by adding fibronectin to diluted serum immediately before standard ELISA testing. Collagens were obtained as lyophilized protein and were solubilized by 18 incubation overnight in 0.1 M acetic acid. Solubilized collagens were heat denatured at 60 °C for 30 min. D N A was denatured by heating to 100°C for 10 min. Antibody Purification and Analysis. Serum antibodies were precipitated with 50% ( N H 4 ) 2 S C » 4 , redissolved in PBS, precipitated with 40% ( N H 4 ) 2 S C » 4 , redissolved in PBS and exhaustively dialyzed against PBS. Dialyzed material was applied three times to gelatin-agarose (Sigma) affinity columns (equilibrated with PBS) and columns were washed with 3 volumes of PBS. Bound material was eluted with 3 volumes of 4 M urea p H 7.3 in PBS and immediately dialyzed into PBS. Columns were washed with 3 volumes 8 M urea PBS and 5 volumes PBS before reuse. Antibodies were purified from individual and pooled serum samples positive for anti-gelatin antibodies and from equivalent volumes of control serum. Purified material from the equivalent of 1 \iL serum was dissolved in non-reducing 2% SDS, 50% sucrose sample buffer, heated to 9 0 ° C for 3 min and electrophoresed on 6% SDS polyacrylamide gels. A continuous buffer system with 192 m M glycine and 25 m M Tris ph 8.3 was used for all electrophoresis and electroblotting. Replicate lanes from 6% gels were then either silver stained as in (87) or electroblotted onto nitrocellulose (BioRad) as in (88). Nitrocellulose strips were rocked for 60 min in 5% skim milk in Tris buffered saline p H 7.3 and then for 120 min with alkaline phosphatase conjugated goat anti-human IgG antibodies diluted 1/1000 in 5% skim milk. Strips were washed 3 times for 5 min with PBS 0.5% Tween and then incubated for 10 m i n with alkaline phosphatase immunoblotting substrate (89). H u m a n collagens were dissolved in sample buffer with 5% 2-mercaptoethanol, heated to 1 0 0 ° C for 5 min, electrophoresed on 6% SDS polyacrylamide gels and electroblotted onto nitrocellulose. Nitrocellulose strips were incubated for 60 min in 5% skim milk and overnight with serum or purified antibodies diluted 1/500 in 5% skim milk. Development of strips was carried out as described above. Collagens were loaded onto gels at approximately 300 ng/lane. Measurement of Serum IgG. Serum IgG was measured by ELISA. Plates were coated with polyclonal goat anti-human IgG antibodies and ELISA was carried out as stated except that serum samples were diluted 1 / 1 0 ° in 1% casein. Known concentrations of human IgG (Cappel) were run in parallel with serum samples to construct a standard curve from which serum IgG concentration was estimated. Reactivity Against Collagen-Like Sequences of Clq. C l q was electrophoresed under reducing and non-reducing conditions on 6% and 12% polyacrylamide gels in the presence and absence of 5% 2-mercaptoethanol. The collagen-like fragments of C l q were obtained by pepsin digest (90), and electrophoresed on 12% SDS polyacrylamide gels under the same conditions. Transfer to nitrocellulose and immunoblotting with purified anti-gelatin antibodies was performed as described for collagens. Reduced collagen type I or III was electrophoresed and used as a positive control. Reactivity Against Antibody Bound Clq. In order to determine if collagen-like antigenic epitopes of C l q were exposed upon binding to antibody, a modified ELISA was developed in which IgG antibodies were coated directly onto microtitre plates. C l q was then allowed to react with the bound IgG in the presence of purified anti-gelatin antibodies and control antibodies. Plates were coated as previously described with 500 ng human IgG (Sigma). C l q was added for 90 min at 500 ng/well in 1% casein in the presence of 0.5 | i g / m L purified biotinylated anti-gelatin antibodies, or 5 u.g/ m L biotinylated normal human IgG. Biotinylation was carried out as in (91). Biotinylated material purified by gelatin-affinity chromatography r 20 from normal sera was also used as a control in this ELISA. The ELISA was developed as previously described except that a 1/400 dilution of avidin conjugated alkaline phosphatase (Sigma) was used to quantitate the binding of biotinylated antibodies to ELISA plates Samples were also added to replicate wells coated with 500 ng of C l q or IgG alone. Complement Measurements. Hemolytic complement activity was measured in fresh sera from healthy volunteers by the rise in O D 415 following 60 min incubation with antibody sensitized sheep red blood cells (SRBC) (Sigma) at 20°C. From 5 to 40 jig purified anti-gelatin antibodies were added to 10 uL of test sera diluted 1/15 in PBS and 150 uL sensitized SRBC to assess any effect on complement activity. Manufacturer's instructions were followed for testing except that smaller aliquots of SRBC were incubated with sera in eppendorf tubes. After incubation, tubes were centrifuged for 10 min in an eppendorf centrifuge and 100 [iL of each supernatant transferred to individual wells of an ELISA plate. O D 415 was read immediately after transfer on a microplate reader. Statistical Analysis. Significant differences were assessed by the Student's t test and correlation between different measures assessed by linear regression analysis using the Statworks software program. 21 Results Section I. Antibodies, H I V Infection, and Disease Progression Anti-anti-CD4 Antibodies. One hundred and seventy-three serum samples were tested by E L I S A for antibodies reactive with the anti-CD4 monoclonal antibody leu 3a. The samples included 10 controls, 38 HTV- homosexuals, 33 HIV+ homosexuals, 16 homosexual AIDS patients, 21 HIV+ hemophiliacs, 42 HIV" hemophiliacs, and 12 R A patients. Results of ELISA testing are shown in figures 1 and 4. Few homosexuals in any category express anti-anti-CD4 antibodies and there appears to be no correlation with disease progression. This suggests that anti-anti-CD4 antibodies are not associated with the development of AIDS. The finding that some HIV" homosexuals and hemophiliacs express anti-anti-CD4 antibodies suggests that some of these antibodies do not arise against HIV gpl20, but more likely are related to exposure to foreign M H C class LI antigens. Soluble purified HTV gpl20 prevents only 30 to 40 per cent of the binding to leu 3a (figure 2) in either H I V + or HTV" homosexual serum samples. Inhibition of anti-anti-CD4 activity of HTV- serum with gpl20 suggests that antibodies against M H C class II antigens cross-react with HIV gpl20. Thus, there is the potential for interaction between the immune response to M H C antigens and the immune response to gpl20. Anti-anti-CD8 Antibodies. The same serum samples as above were tested for reactivity with the anti-CD8 monoclonal antibody leu 2a. Results are shown in figures 3 and 4. In this case, there does appear to be a relationship between infection with H I V , disease progression, and expression of anti-anti-CD8 antibodies. The prevalence of these antibodies varies from 24% (HIV") to 40% (HTV+) to 50% (AIDS) in serum from homosexuals (figure 3A). Anti-anti-CD8 antibodies probably result 22 1.0 Anti-anti-CD4 OD 405 nm 0.5-0.0 1 • — 3 SD cutoff • • a Controls (n=8) HIV-(n=38) HIV+ (n=33) AIDS (n=15) B 1.0 Anti-anti- CD4 OD 405 nm 0.5-0.0 Controls (n=10) - 3 SD cutoff HIV-(n=43) HIV+ (n=21) Figure 1. Anti-anti-CD4 FXKA values obtained with individual serum samples within homosexual (A) and hemophiliac (B) subgroups, together with normal control sera. A number of hemophiliac sample values are not visible due to a net OD of zero. 1.0-0.8-0.6-OD (0gp120) 02 0.0 0.0 .05 0.5 5.0 ng gp120/mL 50 05 0.5 5.0 hg gp120 50 1.0 0.8 0.6-OD (+qp120l OD (0gp120) 0.2-0.0-0.0 .05 i 1 r 0.5 5.0 50 ng gp120/mL Figure 2. Inhibition of anti-anti-CD4 i ELISA reactivity with recombinant gpl20. Gpl20 inhibits some of the anti-anti-CD4 activity of serum from two HIV infected individuals diluted 1/1000 (A) and 1/250 (B) and of serum from a non-infected homosexual diluted 1/250 (C). from exposure to foreign M H C class I antigens and our data suggests that such exposure increases the risk of infection with HTV and the risk of HTV related disease progression in homosexuals. A similar, but stronger correlation between HIV infection and anti-anti-CD8 antibodies was seen in hemophiliacs; prevalence in HIV" hemophiliacs (38%), prevalence in HTV+ hemophiliacs (90%) (figure 3B). The sera of hemophiliacs also expressed much higher levels of anti-anti-CD8 activity, presumably as a result of repeated exposure to concentrated preparations containing al lo-MHC class I antigens. Anti-anti-CD8 antibodies were found also in 25% of serum samples from rheumatoid arthritis patients (figure 4). The patient with the highest level of anti-anti-CD8 antibodies had serum drawn during a clinically active disease flare. The presence of anti-anti-CD8 antibodies in persons with autoimmune disease suggests a link between production of these antibodies and aberrant immune system regulation. One of three samples from persons with SLE also contained anti-anti-CD8 antibodies (not shown). These antibodies may be involved in, or signify, disease progression in HTV infected persons. This hypothesis is complicated by the significant correlation of anti-anti-CD8 antibody levels with serum IgG in homosexual AIDS patients and H I V + hemophiliacs, although these factors are independent in other groups (table 1). Anti-anti-class II Antibodies. Several monoclonal antibodies were used to test by ELISA for anti-anti-class II M H C antibodies in controls, homosexuals, hemophiliacs, and R A patients. Results using an anti-gpl20 monoclonal which bears an internal image of CD4 are shown in figure 5. Low antibody levels, not significantly different from control levels, were expressed in homosexual samples of any category (tables II and HI) and no correlation with disease progression was apparent (figure 5A). Hemophiliacs expressed higher levels of anti-anti-class II 25 0.50 Anti-anti-CD8 OD 405 nm 0.25-0.00' Controls HIV-(n=8) (n=38) 3 SD cutoff HIV+ (n=33) AIDS (n=16) 2.0 Antl-antl-CD8 OD 405 nm 1.0-0.0 • a Controls (n=10) B a a 3 SD cutoff HIV-(n=43) HIV+ (n=21) Figure 3. Anti-anti-CD8 ELISA values obtained with individual serum samples within homosexual (A) and hemophiliac (B) subgroups, together with normal control sera. A number of hemophiliac sample values are not visible due to a net OD of zero. 0.50 O D 4 0 5 n m 3 S D cutof f 0.25 0.00 OD 405 nm -T»-3SD cutoff Anti-ant i-CD4 Ant i-ant i-CD8 Figure 4. Anti-anti-CD4 and anti-anti-CD8 ELISA values obtained with individual serum samples from rheumatoid arthritis patients. Cutoff lines indicate 3 standard deviations above the mean ELISA OD obtained with normal control sera analyzed at the same time, but not shown. A 0.230 -An t i - an t i - gp 120 O D 405 n m 0.115 •3 SD cutoff 0.000-C o n t r o l s (n=8) HIV-(n=22) HIV+ (n=33) A I D S (n=14) 1.0-An t i - an t i - gp 120 O D 405 n m 0.5 0.0 C o n t r o l s (n=10) B HIV-(n=43) 3 SD cutoff HIV+ (n=21) Figure 5. Anti-anti-gpl20 (anti-anti-class II M H C ) ELISA values obtained with individual serum samples within homosexual (A) and hemophiliac (B) subgroups, together with normal control sera. A number of hemophiliac sample values are not visible due to a net O D of zero. Serum samples from 12 rheumatoid arthritis patients were negative for anti-anti-gpl20 activity (not shown). antibodies (figure 5B), but the levels of these antibodies was significantly correlated with serum IgG (table 1). ELISA results using a monoclonal antibody (L234) against a non-polymorphic region of human M H C class LI ( H L A DR) were consistently negative (not shown). These results demonstrate difficulty in detecting anti-anti-class II activity in homosexual serum samples using monoclonal antibodies as antigens. Anti-gelatin Antibodies. Anti-gelatin antibody levels were measured by ELISA in 178 serum samples. Samples included 8 controls, 12 rheumatoid arthritis patients, 44 HTV- hemophiliacs, 24 HIV+ hemophiliacs, 38 H I V - homosexuals, 33 H I V + homosexuals, and 16 homosexual AIDS patients. Results are shown in figures 6A and 6B. The prevalence of autoantibodies against gelatin increased in homosexuals with infection with H I V (32% HIV", 66% H I V + ) and with progression to AIDS (100%). Prevalence also increased with HIV infection in hemophiliacs, but to a much lesser degree than in homosexuals (0% HIV", 13% HTV+). Although prevalence increased with HIV infection and disease progression in homosexuals, anti-gelatin antibody levels reached titres as high in non-infected homosexuals as in AIDS patients (figure 6A). The mean anti-gelatin activity of the H I V -homosexuals d id not differ significantly from controls (table III), even though individuals within the H I V - homosexual population had high levels of anti-gelatin antibodies. Serum IgG Levels. Serum IgG levels were measured to assess correlations between total serum IgG and particular anti-idiotype or autoantibody levels. The mean serum IgG levels of the different serum categories and correlation coefficients with different antibodies are shown in table I. Mean serum IgG was elevated in the 29 2.0-Anti-gelatin OD 405 nm 1.0 0.0 ."« n r- 3 SD cutoff Controls HIV-(n=8) (n=38) HIV+ AIDS (n=33) (n=16) Anti-gelatin OD 405 nm 0 .8 -0.4-0.0 Controls (n=8) B HIV-(n=44) 3 SD cutoff HIV+ (n=24) Figure 6. Anti-gelatin ELISA values obtained with individual serum samples within homosexual (A) and hemophiliac (B) subgroups. A number of sample values are not visible due to a net O D of zero. Serum samples from 12 rheumatoid arthritis patients were negative for anti-gelatin activity (not shown). Table I. Serum IgG Levels and Correlation Coeffecients IgG/aocCD8 IgG/aaCD4 IgG/aagp 120 IgG/agel AIDS .681 .345 .166 .085 HlV+a .136 .380 .261 .353 HIV-fl .105 .322 .179 .359 HIV+b .730 .708 .775 .490 HIV-b .159 .000 .206 .320 R A .086 .120 .178 .036 Controls .170 .153 .302 .714 mg IgG/mL (mean+sem) 16.50+2.11 17.29±0.81 11.63±0.36 30.73±3.52 19.61±1.48 13.71+1.20 9.9410.65 Footnotes. a= homosexuals b= hemophiliacs 31 Table LT. Mean ELISA O D values of group ± sem AIDS aaCD8 .2841.030 (n=15) aaCD4 .2651.023 (n=15) aagp 120 cxGelatin .092±.010 1.091±.104 (n=14) (n=16) HIV+fl .223±.013 (n=33) .252±.038 (n=33) .096±.004 .6431.078 (n=33) (n=33) HIV-a .2081.012 (n=38) .2511.020 (n=38) .0861.004 .4201.080 (n=22) (n=38) HIV+b .9831.095 (n=21) .3551.064 (n=21) .5281.086 .1711.033 (n=21) (n=24) HTV-b .3531.054 (n=43) .1991.041 (n=43) .3021.060 .0951.011 (n=43) (n=44) R A .3011.031 (n=12) .2181.031 (n=12) .0321.01 .0771.010 (n=12) (n=12) Controls .1791.009 (n=8) .1861.009 (n=8) .0801.004 .1261.024 (n=8) (n=8) Footnotes. a= homosexuals b= hemophiliacs group of H I V infected homosexuals, all hemophiliacs, and persons with rheumatoid arthritis (tables I and HI). Significance Levels. Significance levels were determined by Student's t test to see which parameters distinguished relevant populations (table III) and which parameters differed independently of differences in serum IgG. Mean ELISA O D levels of different antibodies in each different group are shown in table LI. Results of Student's t testing for significant differences in population means are shown in table III. Anti-anti-CD4 levels do not distinguish AIDS patients from HIV+ homosexuals, but do distinguish AIDS patients from normal controls. Anti-anti-C D 8 levels distinguish AIDS patients from H I V infected homosexuals and other groups but don't distinguish H I V infected homosexuals from non-infected homosexuals. Anti-anti-CD8 levels also distinguish H I V infected from non infected hemophiliacs. These parameters may relate more to risk factors for infection ie. exposure to alloantigens than to the progression of HTV related disease. Anti-gelatin antibody levels show the best correlation with disease progression, as the population means discriminate between H I V + homosexuals and homosexual AIDS patients and between H L V + and H I V - homosexuals and hemophiliacs, but do not distinguish H I V - homosexuals or hemophiliacs from controls. Section II. Complementary Immune Responses Serum from HIV- homosexuals expressing autoantibodies and serum from normal controls partially inhibited the binding of antibodies from a particular AIDS patients to gp 120 (figure 7A). Repetition of this inhibition experiment with 4 control sera, 4 autoimmune sera, and 4 different AIDS patients produced highly variable results from individual to individual (figure 7B). There was no consistent Table ILL Significance Values A. Serum IgG levels A B C D E F G A . AIDS ns .001 .001 ns ns .038 B. HIV+homosexuals .001 .001 ns .023 .001 G HTV- homosexuals .001 .001 .029 ns D. HTV+hemophiliacs .002 .001 .001 E. HIV- hemophiliacs .011 .001 F. R A patients .027 G Controls B. Anti-anti-CD4 levels A B c D E F G A . AIDS ns ns ns ns ns .050 B. HIV+homosexuals ns ns ns ns ns C HTV- homosexuals ns ns ns ns D. HTV+hemophiliacs .050 ns .002 E. HIV- hemophiliacs ns ns F. R A patients ns G Controls Table LTI. continued C. Anti-anti-CD8 levels A B C D E F G A. AIDS .050 .005 .001 ns .040 .020 B. HIV+homosexuals ns .001 ns ns ns C. HTV- homosexuals .001 .025 ns ns D. HTV+hemophiliacs .001 .001 .001 E. HIV- hemophiliacs ns ns F. RA patients ns G. Controls D. Anti-anti-gp 120 levels A B C D E F G A. AIDS ns ns .001 ns .001 ns B. HIV+homosexuals ns .001 .004 .001 ns C HTV- homosexuals .001 .012 .001 ns D. HTV+hemophiliacs .034 .001 .001 E. HIV- hemophiliacs .021 ns F. RA patients .011 G Controls Table III. continued E. Anti-gelatin levels A A. AIDS B. HIV+homosexuals C. HIV- homosexuals D. HIV+hemophiliacs E. HIV- hemophiliacs F. RA patients G. Controls B C D E F G .004 .001 .001 .001 .001 .001 .050 .001 .001 .001 .002 .020 .001 .020 ns .009 ns ns ns ns .011 Footnotes. Numbers in the tables represent the probability that the means of the various populations compared by Student's t test are the same. If the probability that the population means were the same was greater than .05 in any test, the difference between those populations was considered non-significant and ns entered into the table. Populations in the table columns are denoted by letters which correspond to the denotations of the populations named in the table rows. % Inhibition of anti-gp120 1/100 1/10 Inhibitor serum dilution % Inhibition of anti-gp 120 B • Control AB • Control J • Control H Control R • HIV- 26 • HIV-7 m HIV- 35 m P AIDS AIDS AIDS AIDS 18 19 46 53 Figure 7. Serum from a normal control (R) and non-FTIV infected anti-gelatin positive homosexual (12) slightly inhibited the anti-gpl20 ELISA activity of an individual AIDS patients' serum diluted 1/500 (A). There was no consistently greater inhibition by homosexual (26, 7, 35) or other alloimmune (P) serum samples (1/20 dilution) expressing anti-gelatin activity, than by normal control sera (AB, J, H, R) of the anti-gpl20 activity of any of four AIDS sera (18, 19, 46, 53) diluted 1/500 (B). Samples shown in legend that do not appear on graph gave zero inhibition. relationship between supposed exposure to allogeneic cells and subsequent anti-anti-class II antibody response and the ability to inhibit anti-gpl20 antibodies from AIDS patients' serum. The relevant immune responses, anti-anti-class II in homosexuals and anti-anti-anti-class II in AIDS patients either exhibit too much individual variation for specific complementarity to be demonstrated by this method or have already formed immune complexes in the AIDS patients' sera. The ability of serum from normal controls to bind anti-gpl20 antibodies of AIDS serum was unexpected and is of unknown significance. Section III. Characterization of Anti-gelatin Antibodies Titration of Anti-Gelatin Activity. Titration of anti-gelatin antibody activity in 3 homosexual serum samples showed titres 100 times those of control sera. Anti-gelatin antibodies were routinely detectable at a 1/10,000 dilution of sera (figure 8). Antibody Nature of Anti-gelatin Activity. Material purified by gelatin agarose affinity chromatography from control, H I V + , and AIDS serum was shown by SDS P A G E sizing to have a molecular weight of 150,000 kd and was shown by immunoblot analysis to be human IgG (figure 9). Much higher levels of antibody were purified from AIDS and H I V + serum relative to control serum, but some IgG was purified from control serum by gelatin-agarose affinity chromatography. Anti-collagen Specificity of Anti-gelatin Antibodies. Four types of native and heat denatured human collagens were used to inhibit the anti-gelatin ELISA. Figure 10 shows that all collagens inhibited the anti-gelatin ELISA more effectively after heat denaturation. Almost complete inhibition occurred with 50 ng of heat 38 3 Anti-gelatin OD 405 nm 2 -1 -Control _ HIV-HIV+ AIDS log (serum dilution) Figure 8. Titration of anti-gelatin activity in representative homosexual and control serum samples shows that the level of anti-gelatin activity is roughly 100 times higher in the homosexual samples. Difference between these samples and control samples was easily detectable at dilutions up to 1/10,000. 39 1 2 IgG A B c C B A Figure 9. Gelatin-agarose affinity purified material from control (A), HIV+ homosexual (B), and AIDS (C) sera was subjected to 6% non-reducing SDS PAGE and transferred to nitrocellulose for immunoblot visualization with alkaline-phosphatase conjugated goat anti-human IgG antibodies (1) or silver stained along with molecular weight standards (2). Figure 10. 40 100 -% inhibition 5 10 15 jig/mL denatured collagen B 100 -% Inhibition jig/mL collagen (non denatured) Figure 10 continued. 41 100 -% Inhibition jig/mL collagen (non-denatured) Figure 10. A l l human collagens tested (types I, III, IV, and V) inhibited the anti-gelatin ELISA activity of a representative anti-gelatin positive homosexual serum sample diluted 1/100 (A&B) or 1/200 (C). Denatured collagens (A) inhibit the ELISA more effectively than non-denatured collagens (B&C). Native and denatured collagens at 20 Lig / m L did not inhibit the anti-gpl20 ELISA activity of HIV+ serum (not shown). 100 80 % Inhibition of Anti-gelatin 60 -OD 405 nm 40 -20 -5 10 jig/mL fibronectin 15 —T -20 Figure 11. Purified fibronectin inhibits the anti-gelatin ELISA activity of representative anti-gelatin positive sera diluted 1/100. Fibronectin was added to diluted serum immediately pror to the ELISA and the incubation times were as for a standard ELISA. Anti-gelatin positive sera does not react with fibronectin directly in standard ELISA testing (not shown). Inhibition of the anti-gelatin ELISA by fibronectin is through direct reaction of fibronectin with gelatin and masking of the antibody reactive determinants. Fibronectin at up to 20 Lig /mL did not inhibit the anti-gpl20 ELISA activity of FTIV+ serum (not shown). denatured types I and II collagen added to 100 uL of a 1/100 dilution of serum (figure 10A). Roughly 80 times as much native collagen type I or LLI was required to elicit the same degree of inhibition (figure 10A, B). Figure 11 shows the ability of fibronectin to inhibit antibody binding to gelatin. This suggests that the anti-gelatin antibodies bind to the same determinants of collagen as does fibronectin. Figure 12 demonstrates the strong immunoblot reactivity of an anti-collagen positive serum and of purified anti-collagen antibodies with the alpha chains of reduced collagens type I and HI, confirming the preferential reactivity demonstrated by inhibition studies. Correlation With Serum IgG Levels. Figure 13 and table I summarize the relationship by group of anti-gelatin activity with serum IgG levels. Only control sera shows significant correlation of these measures. This suggests anti-gelatin antibodies are independent of polyclonal B cell activation. Anti-Collagen Profiles. In order to see whether overall differences in anti-collagen activity existed between homosexual H I V - , HIV+, and AIDS patients, ELISA activity was measured in six samples from each group against all four collagens in both native and heat denatured conformations. Relative reactivity was very consistent within and between these groups as shown by the similar profiles and small standard error bars, but also very different from the control group (figure 14). Reactivity in the "anti-gelatin positive" sera from homosexuals rose selectively against denatured collagens, especially against denatured types I and IH, compared to control sera. This also argues against polyclonal activation as a source of autoantibodies against gelatin. A B 116k 97k 66k 45k 29k I m IV V rn i v v I T r Figure 12. Immunoblot analysis of reduced electrophoresed collagens (6% SDS PAGE) with anti-gelatin positive serum (A) and affinity purified anti-gelatin antibodies (B) shows preferential reactivity with the alpha chains of collagens type I and III. Only slight reactivity is seen with collagen type IV and reactivity with type V collagen is barely detectable. 2.0 Anti-gelatin OD 405 nm LO-CO-+ + • . • + - % T 10 20 30 40 Serum IgG Level (mg/mL) I 50 Figure legend o Control M RA + AIDS • HIV- homosexuals a HIV+ homosexuals * HIV- hemophiliacs • HIV+ hemophiliacs Figure 13. Serum IgG levels and anti-gelatin ELISA activity plotted against each other for individual serum samples from homosexuals (FLTV+, FLTV', and AIDS), hemophiliacs ( H I V + , HIV"), rheumatoid arthritis patients, and normal controls. No correlation between these variables is evident, which suggests that anti-gelatin activity is not due to polyclonal activation. Figure 14. The average relative reactivity (± sem) of six sera from each of four groups against native and denatured forms of collagen types I, III, TV, and V. Relative reactivity is defined as the ELISA OD against a given collagen divided by the highest ELISA OD obtained under identical conditions against a native or denatured collagen with the same serum sample. The different specificity profile for control sera compared to the "anti-gelatin positive" sera from the homosexual AIDS, HIV+, and HIV" groups argues against anti-gelatin activity arising through polyclonal activation. Section IV. Reactivity of Anti-gelatin Antibodies With C l q Direct ELISA Reactivity. Complement component C l q is known to express collagen-like sequences and triple helical conformation. Since the anti-gelatin antibodies reacted to some degree with all collagens tested, it seemed reasonable to expect some degree of reactivity with C l q as well. Reactivity with native C l q bound to ELISA plates, however, did not correlate with anti-gelatin activity and was unaffected by the addition of 20 m M E D T A to the serum diluent (not shown). This concentration of E D T A abrogates binding of C l r and C i s to C l q and prevents any inhibition of antibody binding to the collagen like region of C l q . Antibody preparations used in this study reacted non-specifically with immobilized C l q , presumably through the Fc binding regions of C l q . Since C l q is used to measure serum immune complexes in a similar ELISA format, this form of ELISA is inappropriate for the measurement of specific anti-Clq antibodies. Although some anti-gelatin positive sera had greater ELISA reactivity against C l q than control sera, it is not clear what the C l q ELISA measures; total serum IgG, serum immune complexes, or anti-Clq antibodies. Soluble C l q inhibited the anti-gelatin ELISA to a measurable, but minor extent, suggesting some weak cross-reactivity of the anti-gelatin antibodies with soluble C l q (figure 15). It is possible that this small amount of reactivity is obscured in a solid phase binding assay such as an ELISA, by the non-specific binding of IgG via Fc portions to C l q . Immunoblot Reactivity. Purified anti-gelatin antibodies did not react with the primary chains of C l q produced upon reduction of intact C l q (figure 16). Similar to the C l q ELISA, non-specific antibody binding to intact C l q and to C l q constituent chain oligomers ocurred in immunoblot testing (figure 17). After digestion of C l q with pepsin to prepare collagen-like fragments, there was no 100 80-% Inhibition of Anti-gelatin 60 -OD 405 nm 40 -20 -1 0 jig/mLC1q 1 5 —I 20 Figure 15. Purified complement component C l q inhibits the anti-gelatin activity of a representative anti-gelatin positive sera (diluted 1/100) to a minor extent. C l q was added to diluted serum in the presence of 20 m M E D T A and incubated overnight at 4°C before anti-gelatin ELISA testing. C l q at up to 20 u g / m L did not inhibit the anti-gp!20 ELISA activity of H T V + serum (not shown). 49 1 2 3 4 5 6 7 1 2 3 4 5 6 M M M & 2 r J Figure 16. Molecular weight standards (1), fibronectin (2), gpl20 (3), Clq in 8M urea (4), Clq in sample buffer (5), type III collagen (6), and reduced Clq (7) were subjected to 6% SDS PAGE and silver stained (A). Reduced Clq (1), type III collagen (2), Clq in sample buffer (3), Clq in 8M urea (4), gp!20 (5) and fibronectin (6) were transferred from the same gel to nitrocellulose for immunoblotting with purified anti-gelatin antibodies (B). Proteins were loaded at approximately 250 ng/lane. Anti-gelatin antibodies reacted with type III collagen (B2) and non-reduced Clq (B3,4)/ but not with recombinant gpl20 (B5), fibronectin (B6), or reduced Clq (Bl). Fig. 17 A B C Figure 17. Binding of purified anti-gelatin antibodies to non-reduced Clq (Al) and type I collagen (A4), but not to reduced Clq (A3) was seen by immunoblotting. Samples were subjected to 12% SDS PAGE and transferred to nitrocellulose. Gelatin at 0.5% inhibited binding to reduced type I collagen (B3), but not to Clq (B5). Immunoblot strips (C) show that purified anti-gelatin antibodies (1), purified normal human IgG (2), and goat anti-human antibodies (3) all bind to non-reduced Clq transferred to nitrocellulose following 6% SDS PAGE. This suggests that non-specific binding of antibodies to immobilized Clq occurs through the Fc binding regions of Clq. Figure 18. Demonstration that anti-gelatin antibodies do not react on immunoblots with the collagen-like regions of Clq or acetylcholinesterase. Non-reduced pepsin alone (lane 1), non-reduced Clq digest (lane 2), non-reduced acetylcholinesterase digest (lane 3), (standards lanes 4 and 5), reduced pepsin alone (lane 6), reduced Clq (lane 7), reduced Clq digest (lane 8) reduced acetylcholinesterase digest (lane 9), and reduced type I collagen (lane 10) were subjected to 12% SDS PAGE and silver stained (A) or transferred to nitrocellulose for immunoblot analysis with purified anti-collagen antibodies (B). Immunoblot shows reactivity of purified antibodies only with type I collagen (BIO) and residual intact Clq left after pepsin digest (B2). 52 immunoblot reactivity of the anti-gelatin antibodies or anti-gelatin positive sera with reduced or non-reduced fragments (figure 18). IgG Enhancement of Anti-Clq Reactivity. When C l q and purified anti-gelatin antibodies were allowed to react in the presence of solid phase bound human IgG, there was a significant enhancement of ant i -Clq activity. This enhancement was not seen with normal IgG or with material purified by gelatin-agarose affinity chromatography from normal sera (figure 19). This suggests that a conformationally altered from of C l q present upon binding to the Fc region of IgG reacts specifically with anti-gelatin antibodies. Hemolytic Complement Testing. Anti-gelatin antibodies added to diluted fresh control serum at up to an equivalent amount by weight of the C l q present in the test serum, d id not inhibit complement hemolytic activity (not shown). Although this represents a large excess of anti-gelatin compared to the relative amounts of C l q and anti-gelatin antibodies found in AIDS sera, the in vitro complement assay may create artificial conditions under which hemolysis is heavily favoured over normal conditions in vivo. This possibility need be considered, although these results suggest that anti-gelatin antibodies binding to activated C l q do not inhibit complement function and are not resposible for the depressed hemolytic complement activity seen in F n v + homosexuals. OD 405 nm 0.4-0.0 • Normal serum • Purified IgG 11 Anti-gelatin Figure 19. The binding of purified anti-gelatin antibodies to C l q is enhanced when C l q interacts with immobilized IgG antibodies (IgG + Clq ) , compared to the binding to C l q or IgG alone (end groupings). Enhanced binding was not seen with material affinity purified from normal serum with gelatin-agarose or with 10 times the quantity of purified IgG. Columns represent the mean O D obtained from three separate experiments ± sem. 54 Discussion Infection with H I V can be asymptomatic for long periods or rapidly devastating (92). Severe immunodeficiency follows disappearance of the T4 cells necessary for specific immune responses (93). Before clinical signs of immune deficiency develop, immunological signs of hyperactivity are pronounced. B cell hyperplasia, elevated serum immunoglobulins and circulating immune complexes arise soon after infection (47). These symptoms can reflect direct activation of B cells by HIV or other agents, but they also occur in the absence of infection (94). Immune activation in autoimmune diseases, for example, produces similar serological and pathological signs. Increased growth factor production, direct B cell activation or effects on T cell regulation could explain the early signs of HIV infection. H I V induces more immunoglobulin synthesis in lymphocyte cultures of B and T cells than in cultures of B cells alone, which suggests, but doesn't prove, that H I V exerts some effects on B cells through an effect on T cells (95). The maintenance of immune activation, even in periods of limited viral replication, suggests that an immunoregulatory disorder develops early in the course of HIV infection. This immunoregulatory disorder may pertain to the later development of overt AIDS. Network theories of immune system regulation maintain that idiotypes recognizing foreign antigens such as HIV are themselves recognized by idiotypes present on the cell surface of T and B lymphocytes (96). The character of the network topology in the immune system is postulated to be largely determined by M H C proteins. Interaction between T cell idiotypes depends upon complementarity of antigen responsive idiotypes to self M H C and mimicry of M H C by internal idiotypes. Selection criteria for internal idiotypes impels recognition of M H C restriction elements, while antigen responsive idiotypes are selected for 55 recognition of short foreign peptides (77). The regulatory idiotypes must resemble M H C but differ in that there is no constraint to bind short peptides directly. Diversity among antigen responsive idiotypes is favoured, while convergence to shapes most closely resembling the M H C sites of T cell interaction is favoured among regulatory T cells. T cell regulation through this type of internal recognition is postulated to involve a network focussing topology (77). Focussing to highly conserved internal regulatory idiotypes could occur if similarities between M H C molecules are more recognizable than differences. In this case the regulatory structures of immune systems would be highly conserved. Alternatively, there may be many regulatory idiotypes, each corresponding to a different M H C protein involved in presenting conserved sets of peptides. Regulation of multiple antigen reactive cells restricted to particular M H C proteins could occur through one regulatory T cell in either case. Although the role of suppression in the ontogeny and maintenance of self tolerance is sometimes questioned, self reactive T and B lymphocytes do exist (97). In some circumstances, polyclonal activation, autoantibody production, and immune complex production seem to occur through defective suppression of self reactive cells. In human lupus, autoantibodies with different antigenic specificity express a common idiotype defined by a monoclonal antibody anti-16.6 (98). These autoantibodies may share an idiotype which is related to the idiotype of regulatory cells which normally suppress autoantibody production. Constraints of the symmetrical network theory dictate that suppressor cells have high network connectivity and interact idiotypically with a large number of lymphocytes (99). Thus point idiotypic defects in suppression can have widespread effects on the immune system network. Polyclonal activation may actually result from idiotypically specific effects on regulatory cells and thus produce antibodies with common idiotypes. 56 H I V confronts the central axis of the immune system network with its external envelope glycoprotein, gpl20. Gpl20 is complementary to the cell surface protein CD4, which is found most commonly on T lymphocytes and monocytes (80). CD4 normally functions in signal transduction by interacting with M H C class LI proteins and this interaction is inhibited by gpl20 (100). H I V thus relates to class LI M H C proteins in manner analogous to the internal regulatory idiotypes. HTV gpl20 recognizes M H C restriction elements on the CD4 protein and the regulatory idiotypes recognize M H C restriction elements on T cell receptors. This relationship of gpl20 to CD4 and class LT M H C and potential mimicry of regulatory idiotypes is the basis of the network theory of AIDS. In the context of this theory, gpl20 expresses an internal image of class II. Theoretically, prohibition of immune responses against internal images of class II M H C is not as strictly enforced as against class LT proteins themselves. Subtle differences between class TI proteins and internal image of class II proteins, elicited through different selection criteria, may render internal images of class II M H C on regulatory idiotypes targets for the immune response against HTV gpl20. Since it is the anti-CD4 portion of gpl20 which intersects with the immune system network, an hypothesis of the network theory of AIDS is that the immune response against this portion of gpl20 is involved in AIDS pathogenesis. Detailed mapping of the gp l20 /CD4 interaction with monoclonal antibodies has identified anti-CD4 antibodies which bind to the same region of CD4 as does gpl20 (83). These antibodies mimick gpl20 and can be used to detect anti-gpl20 antibodies of the relevant specificity produced in HIV infection (84). The network theory of AIDS also maintains that the pathogenesis of AIDS depends on coincident immune responses against HTV gpl20 and against allogeneic cells (77). Most people who develop AIDS are exposed to allogeneic cells in blood products or semen. Lymphocytes present in these fluids will transiently proliferate in response to host M H C proteins and expose the host immune system to anti-host M H C receptors. These anti-host M H C receptors induce an anti-anti-class LI antibody response in experimental alloimmunizations (79). Anti-anti-class II M H C antibodies are idiotypically related to class LI proteins in an analogous manner to both gpl20 and the internal regulatory idiotypes of the immune system. Potential complementarity between the immune response to gpl20 and the anti-anti-class II immune response exists such that the anti-anti-class II response will stimulate the anti-gp!20 response and vice versa. Two symmetrically opposed immune responses along the central axis of the regulatory network may synergize to effect the gradual erosion of network stability. This potential synergy implies that antibodies complementary to the anti-anti-CD4 antibodies may be involved in AIDS pathogenesis and that the specificity of these antibodies would be anti-anti-class E M H C . Disruption of the internal regulation of the immune system network by HTV is a possible cause of polyclonal activation and eventually of T4 cell depletion and immunodeficiency. Autoreactive antibodies and T cells could result if the regulatory functions of T cells were undermined by inappropriate immune responses. If AIDS results from specific, idiotypically related immune responses to conserved antigens, as suggested by this theory, this should be reflected by the emergence of common idiotypes in the sera of AIDS patients. This study focussed on the identification of antibodies which, according to the network theory of AIDS, reflect common targets of a pathogenic immune response associated with HIV infection and are involved in the induction of AIDS. A primary objective was to assess the prevalence of anti-anti-class II antibodies and antibodies against that portion of gp!20 which binds to CD4. Association between the appearance or level of these antibodies and disease progression should occur if they are involved in destabilizing the immune system network. A secondary objective was to identify and characterize antibodies associated with disease progression, as these antibodies should be idiotypically related to network destabilizing immune responses if such immune responses are occurring and leading to AIDS. Direct E L I S A testing in this study revealed very few H I V infected homosexuals expressing antibodies against the CD4 binding region of gpl20. A similar low prevalence was observed by Lundin et al. using an alternative method of detection (101). Their data also corroborates our observation of high levels of anti-anti-CD4 antibodies in sera from a small minority healthy H I V infected persons. The sensitivity of our ELISA to the anti-anti-CD4/anti-gpl20 antibodies of interest was confirmed using soluble recombinant gpl20 to inhibit the anti-leu 3a ELISA. Surprisingly, inhibition studies indicated that only about half of the anti-anti-CD4 activity was related to H I V gpl20. Although this could be due to variability in HTV gpl20, the CD4 binding region is expected to be highly conserved. It is more likely that the remaining 50% of the anti-anti-CD4 antibodies, and 100% of the anti-anti-CD4 antibodies seen in H I V negative persons are actually anti-foreign class II antibodies, which potentially include anti-anti-CD4 activity. If exposure to alloantigens induces the anti-anti-CD4 antibodies, anti-anti-CD8 antibodies should be more prevalent. The CD8 protein has a relationship to class I proteins analogous to that of class LI proteins and CD4. In alloresponses, class I antigens induce the predominant alloantibody response and in our homosexual and hemophiliac serum samples, as expected, there are higher levels and a greater prevalence and of anti-anti-CD8 antibodies than anti-anti-CD4 antibodies. There also appears to be some correlation between this anti-foreign response and HIV infection and progression to AIDS. This is expected in that exposure to HTV often occurs coincident with exposure to alloantigens and allogeneic cells have been implicated in the network model as a cofactor in AIDS pathogenesis. There were also much higher levels and a greater prevalence of anti-anti-CD8 antibodies in hemophiliac sera than homosexual sera, and the presence of these antibodies correlated well with FLTV infection. Presumably this reflects that the more blood products a hemophiliac is exposed to the greater the a l l o -MHC response and the greater the chance of HTV infection. Since nearly all HTV infected hemophiliacs are also alloimmune it is not possible to associate the anti-anti-CD8 response with AIDS in this group. Our results and those of Lundin et al. (101) suggest that anti-anti-CD4 antibodies are not involved in the pathogenesis of AIDS. Anti-anti-CD4 antibodies were present in the sera of some healthy H I V infected homosexuals and did not increase in level or prevalence in AIDS patients. It is possible that the regulatory T-cell idiotypes are not conserved to the degree that an immune response against them is detectable in a varied population with a single idiotype. A relationship between anti-HIV immunity and ant i -MHC immunity is apparent, however, from the ability of H I V gpl20 to inhibit the anti-anti-CD4 antibodies in sera from HTV-persons. Cellular immunity of the corresponding specificity may be more important in AIDS pathogenesis. Recent work showed that individuals not infected with H I V mount an in vitro cytotoxic T cell response against gpl20 characteristic of previous priming to gpl20 determinants (102). This may reflect, as proposed in the network focussing topology, prior T cell selection for recognition of M H C mimicking idiotypes present on regulatory T cells, which are themselves mimicked by gpl20. Possibly, the T cell cytotoxic response stimulated by HTV cross-reacts with regulatory T cell idiotypes. We would then expect, as has been observed, that despite the unusually high anti-gpl20 C T L response induced in vitro, that there will be little anti-gpl20 C T L activity detectable in unstimulated peripheral blood lymphocytes from H I V infected persons (102). T cells in H I V infected persons which are cytotoxic towards normal T cells have been reported, but specificity for T cell receptor idiotypic determinants has not been determined (22). The interaction between M H C class II and CD4 has not been mapped in the same detail as the gpl20/CD4 interaction, so the selection of appropriate anti-class LI monoclonals with complementarity to the anti-anti-class LT antibodies implicated in the network theory of AIDS is problematic. Direct ELISA testing with a monoclonal antibody (L234), against a non-polymorphic region of H L A DR, revealed no anti-anti-class LI antibodies in AIDS patients or H I V infected persons. ELISA testing using an anti-gpl20 monoclonal which bears an internal image of CD4, and is thus anti-class II, revealed very low levels of anti-anti-class II antibodies in sera from some AIDS patients and H I V infected persons. The low levels and relative prevalence of these antibodies do not support a role for the antibodies in AIDS pathogenesis. Complementarity between anti-gpl20 antibodies in AIDS sera and putative anti-anti-class II antibodies in serum from non HIV infected homosexuals was also not apparent by inhibition of anti-gpl20 ELISA. This could reflect absence of the antibodies of the predicted idiotypic specificity or prior formation of antibody antibody immune complexes in the AIDS patients' sera. It is also possible that the complementary idiotypes on the anti-gpl20 antibodies are expressed independently of the gpl20 binding site so that the formation of complexes does not preclude binding of anti-gpl20 antibodies to gpl20. This phenomenom was observed with anti-gpl20 and anti-anti-gpl20 antibodies in immune complexes eluted from the platelets of HTV infected persons (58). Overall, these results don't support the hypothesis that complementary anti-anti-CD4 and anti-anti-class II antibodies cause AIDS. If such conserved immune responses cause AIDS, they may occur at the cellular level. Alternatively, the idiotypic mechanism involved is considerably more complicated and less conserved than the model predicts. Evaluation of cellular immunity pertinent to the network theory of AIDS will require development of new techniques for the 61 generation and analysis of cytotoxic T cells. Whether an idiotypic mechanism operates at the cellular or humoral level and whether or not it involves exactly the idiotypes predicted, a common idiotypic destructive mechanism should be reflected in the emergence of common idiotypes in affected persons. Some of the focus on aberrant immune regulation in AIDS stems from the the association between autoimmunity and HTV infection. Autoantibodies in AIDS patients are often attributed to polyclonal B cell activation, induced by HTV directly, other pathogens, or by H I V induced depletion of T4+ suppressor inducer cells. Autoantibodies generated by polyclonal activation are generally of low affinity and reflect the composition of the B cell repetoire (103). Levels of these autoantibodies correlate with serum immunoglobulin levels and usually don't reach the levels of antibodies produced in specific immune responses. If A I D S is an immunoregulatory problem induced by specific idiotypes, common specific autoantibodies should appear in AIDS patients and these antibodies should not exhibit the characteristics of autoantibodies produced by polyclonal activation. Although many different autoantibodies have been described in AIDS patients, most occur with low frequency and probably reflect individual differences in immunological or genetic background, rather than an idiotypic mechanism of AIDS pathogenesis. We have detected common specific autoantibodies in AIDS patients which are independent of polyclonal activation and which occur in several immune disorders with features similar to AIDS. The distribution of these autoantibodies suggests they are related to an idiotypic mechanism involved in the induction of severe immune disorder and immunodeficiency. Characterization of these antibodies and identification of factors associated with their presence may therefore reveal important factors in the development of AIDS. These autoantibodies were detected by reaction with the diluent of the anti-C D 4 commercial monoclonal leu 3a, which contains mostly gelatin. Sera from seven different groups were therefore systematically assayed for ELISA reactivity with gelatin. A clear association between anti-gelatin antibodies and HIV infection and progression to AIDS was observed with homosexual serum samples. Association of anti-gelatin antibodies with H I V infection was observed in hemophiliac sera. N o anti-gelatin antibodies were seen in control sera, sera from non-HIV infected hemophiliacs, or sera from rheumatoid arthritis patients. The prevalence of these antibodies in different groups suggests they are useful indicators of the risk of developing AIDS. If a specific idiotypic mechanism is involved in AIDS pathogenesis, it is possible that these anti-gelatin antibodies are related to this mechanism. Several experiments demonstrated that anti-gelatin antibodies are not due to polyclonal activation. The level of anti-gelatin antibodies in many serum samples is over 100 times the level in normal sera and anti-gelatin reactivity is easily detectable at 1/10,000 dilutions of sera. Linear regression analysis showed that serum IgG levels do not correlate significantly with anti-gelatin levels except for in control sera. The correlation of anti-gelatin activity with total serum IgG in controls probably reflects a minor increase in background dependent upon serum IgG. The collagen type specificity profile of antibodies in control sera differs from the anti-gelatin positive sera in that in the positive sera there is specific expansion of antibody reactivity with denatured collagens type I and III. Inhibition and immunoblot studies showed that determinants most commonly expressed on denatured type III collagen are targets for anti-gelatin antibodies present in AIDS sera. These studies suggest that anti-gelatin antibodies result from a specific immune response occurring in association with progression to AIDS. Serum fibronectin shows a reactivity profile with collagen similar to that of the autoantibodies seen in this study (104), and in inhibition experiments, prevented most of the interaction of anti-gelatin antibodies with gelatin (104). Anti-gelatin reactivity was shown by ELISA, SDS P A G E sizing, and immunoblot to be IgG in composition and unrelated to fibronectin. Fibronectin levels have been measured in AIDS sera and are not different from levels of normal sera (105). It is difficult to attribute the anti-gelatin antibodies to the presentation of new determinants to the immune system since the determinants involved are likely to be masked by fibronectin. Although immune responses against collagen occur in disorders characterized by polyclonal activation, the antibodies we describe are unrelated to polyclonal activation. The conditions under which tolerance to collagen is broken during, or before, progression to AIDS suggest an immunoregulatory problem exists. The key questions are how this immunoregulatory problem develops and affects the production of specific anti-gelatin antibodies. The broad reactivity of the anti-gelatin antibodies with all collagens tested implies that they react with other proteins containing collagen-like sequences. Two such proteins are the enzyme acetylcholinesterase (Ach) and the complement component C l q , both of which also interact with serum fibronectin (106, 107). Autoantibodies against A c h have been described in association with neuromuscular disorders and autoantibodies against C l q occur in human lupus in association with elevated levels of immune complexes and decreased hemolytic complement activity (108, 109). This is relevant in the context of AIDS in light of the elevated levels of immune complexes and decreased complement activity observed in this syndrome (47, 48). Experimentally produced anti-collagen antibodies cross-react with C l q and bind to the same region of C l q as do C l r and C i s (110). C l q is composed of three 64 different chains , A , B, and, C , which are disulphide linked as C C homodimers and A B heterodimers. Six A B dimers and three C C dimers associate into the intact molecule bearing six globular heads and a stalk region which exhibits collagen-like triple helical structure. The primary A , B, and C chains of C l q are each composed of approximately one third collagen like sequence (106). When intact C l q interacts via its globular heads with the Fc region of bound immunoglobulin molecules, a conformational change in the collagen like region of C l q results which enables activation of C l r and C i s (111). One theory proposed to explain the production of a n t i - C l q antibodies in human lupus is the presence of high levels of conformationally altered C l q in circulating immune complexes (109). This might result in the presentation of neoantigens to the immune system and the induction of antibodies reactive with collagen like sequences. Previously described monoclonal antibodies which cross-react with C l q and collagen are specific for denatured collagen (112). Since the proposed mechanism of autoantibody production against C l q is consistent with the symptoms of FLTV" infection, reactivity of the anti-gelatin antibodies with C l q was investigated. Immunoblot analysis showed no reactivity between purified anti-gelatin antibodies and the reduced primary chains of C l q . This is somewhat surprising in that the anti-gelatin antibodies reacted with primary determinants of each of the four collagens tested. Apparently, the repeating sequence XYgly expressed in the A , B, and C chains of C l q does not itself confer cross-reactivity with the primary determinants against which the anti-gelatin antibodies are directed. Solid phase bound C l q used in this study reacted with all antibody preparations when intact or when individual dimers were associated into oligomers larger than 100 kd. Reduction of C l q abrogated all antibody binding to C l q . ELISA measures of serum reactivity showed no difference in the presence or absence of E D T A suggesting there was no competition between antibody binding to C l q and the C a + + dependent binding of O r and Cis to Clq. Soluble Clq inhibits the anti-gelatin ELISA only to a slight extent compared to inhibition seen with denatured collagens. Purified anti-gelatin antibodies did not inhibit complement function even when added in equivalent amounts (by weight) to the Clq concentration used in the complement assay. However, the same preparation of anti-gelatin antibodies reacted strongly with purified Clq bound to immobilized IgG. When Clq is digested with pepsin so that collagen like sequences alone are left intact, reactivity with anti-gelatin antibodies is still absent. These results suggest that only under select circumstances is Clq a specific target of the anti-gelatin antibodies described in this study. These circumstances could involve depletion of other complement components or the entrapment of Clq in circulating immune complexes. Such circumstances exist in HIV infection, but it is still speculative to attribute the production of anti-gelatin antibodies to immune complex bound Clq. Conceivably, anti-gelatin antibodies are caused by conformationally altered Clq and this would provide a simple explanation for the loss of self tolerance to collagen. Cross reactivity of the anti-gelatin antibodies with Clq may, however, simply be a biochemical phenomenom without physiological relevance. Reactivity with select conformations of Clq present in immune complexes is an attractive explanation for production of anti-gelatin antibodies, but more biological evidence is needed to evaluate this possibility. Antibodies with the idiotypes predicted to cause AIDS were found, in this study not to be linked to the development of AIDS. Exposure to alloantigens and a subsequent immune response against MHC class I antigens was linked to HIV infection and AIDS in this study, but the data does not distinguish whether exposure to allogeneic proteins or cells increases just the risk of infection or also the risk for developing AIDS. It is possible that immune disorder caused by HIV itself induces an anti-class I like antibody response. Similar antibodies, anti-anti-CD8 were detected in a number of lupus and rheumatoid arthritis patients not exposed to alloantigens. Whether these antibodies are a cause or consequence of immune disorder is not clear, but it is interesting that they are associated with immune disorder outside of H I V infection and exposure to allogeneic cells. In the homosexual AIDS patients and HIV infected hemophiliacs, anti-anti-CD8 levels correlate significantly with serum IgG levels. The reason for this relationship is also obscure, but perhaps some form of selection occurs during progression to AIDS which expands the relative numbers of cells involved in this immune response. Alternatively, alloantigens inducing the anti-anti-CD8 antibodies or the anti-anti-CD8 antibodies themselves cause polyclonal activation in certain circumstances. In the context of idiotypic network regulation, polyclonal activation does not necessarily imply non-specific activation. Antibodies against gelatin were clearly linked in this study to progression to AIDS. Yet high levels of these antibodies also occur without H I V infection in homosexuals and persons with lupus, leprosy, or graft versus host disease (113). This suggests there is a common element in the immunopathogenesis of these disorders. The anti-gelatin antibodies are not related to polyclonal activation, but do seem to correlate with exposure to alloantigens or their equivalent. Homosexuals are exposed to allogeneic lymphocytes and often show immunological priming against alloantigens (114). Graft versus host disease involves B and T cell mediated immunity against host M H C antigens. A n immunodominant cell surface protein of mycobacterium leprae, the cause of leprosy, shows immunological cross-reactivity with H L A DR proteins, and with HIV proteins (115, 116). A mouse monoclonal antibody against HIV gpl20 reacts with a monocyte cell surface antigen involved in antigen presentation and antibodies against HTV gp41 react with human M H C class II antigens (73, 74). Lupus is not associated with alloimmunity, but in a murine model of lupus, the M R L l p r / l p r strain, anti-anti-self M H C antibodies precede the usual autoantibodies characteristic of lupus (78). Interestingly, this strain of mouse also expresses high levels of anti-gelatin activity. Though alloimmunity correlates with anti-gelatin antibodies, it is not clear if, and how, these immune responses are biochemically or idiotypically linked. Purified anti-gelatin antibodies don't bind lymphocyte cell surface antigens or inhibit or compete with anti-anti-CD4 antibodies, anti-anti-CD8 antibodies, or anti-anti-class II antibodies. Anti-gelatin antibodies don't appear to be ant i -MHC themselves, complementary to ant i -MHC antibodies, or to share idiotypes with ant i -MHC or anti-anti-MHC antibodies. It is possible though, as is seen with other antibodies, that the idiotype anti-gelatin antibodies have in common with the M H C related antibodies is expressed independently of the antigen binding region and will require anti-idiotypes for its demonstration. A possible clue to the origin of the anti-gelatin antibodies is their relative prevalence in H I V infected homosexuals and hemophiliacs. Homosexuals, in common with persons with graft versus host disease and leprosy, are exposed to M H C or M H C - l i k e antigens on viable cells or organisms, whereas hemophiliacs are exposed to alloantigens in lyophilized protein preparations. Viable cells or organisms capable of proliferation or active infection are more likely to induce cytotoxic T cell mediated immunity. Perhaps it is this form of alloimmunity that is specifically linked to the production of anti-gelatin antibodies. The capacity of HTV to mimic class II antigens in the induction of cytotoxic T cell responses is consistent with a role for HTV in anti-gelatin antibody production and is also consistent with a synergystic effect of HIV infection and exposure to allogeneic cells. In lupus, perhaps T cell subset abnormalities favour the induction of T cell immunity against M H C or M H C mimicking proteins. T-cells isolated from the synovium of severely affected rhematoid arthritis patients often proliferate in response to autologous lymphocytes. The correlation between T cell alloimmunity and anti-gelatin antibodies does not explain how the anti-gelatin antibodies arise or if and how AIDS is related. The T cell alloimmune response may include a response against the M H C mimicking, regulatory idiotypes of the network focussing topology and may help to provoke the network instability envisioned in the network theory of AIDS. If this hypothesis is correct, then what we see in autoimmune diseases such as lupus may be simply a less virulent form of AIDS. HTV alone, or in concert with exposure to allogeneic lymphocytes, would amplify T cell alloimmunity and, as a result of retroviral persistence, eventually result in AIDS. This may explain the chronically elevated levels of cytotoxic T8 cells seen in HIV infection and the failure of these cells to protect against AIDS. t In conclusion, the etiology of AIDS remains a mystery. The susceptibility of the human immune system to a poorly replicating, but persistent retrovirus, in the minds of a number of immunologists is best explained by the unique relationship of H I V antigens to the regulatory structures of the immune system. The similarities to other immunoregulatory disorders and the negative effects of anti-HIV immunity speak for an immune mechanism of AIDS pathogenesis that will not be counteracted by anti-virals, nor without a better understanding of the internal regulation of the immune system. 69 Bibliography 1. Barre-Sinoussi, F., J. C. Chermann, F. Rey, M. T. Nugeyre, S. Chamaret, J. Gruest, C. Dauguet, C. Axler-Blin, F. Vesinet-Brun, C. Rouzioux, W. Rozenbaum, and L. Montagnier. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS). Science 1983; 220: 868. 2. Vilmer, E., C. Rouzioux, F. Brun-Vezinet, A. Fischer, J. C. Chermann, F. Barre-Sinoussi, C. Gazengel, C. Dauguet, P. Manigne, C. Griscelli, and L. Montagnier. Isolation of a new lymphotropic retrovirus from two siblings with haemophilia B, one with AIDS. Lancet 1984 i; 753. 3. Dalgleish, A., P. Beverley, P. Clapham, D. Crawford, M. Greaves, and R. Weiss. The CD4 (T4) antigen is an essential component of the receptor for the AIDS retrovirus. Nature 1984; 312: 763. 4. Somasundaran, M. and H. L. Robinson. Unexpectedly high levels of HTV-1 RNA and protein synthesis in a cytocidal infection. Science 1988; 242:1544. 5. Leonard, R., D. Zagury, I. Desportes, J. Bernard, J. Zagury, and R. C. Gallo. Cytopathic effect of human immunodeficiency virus in T4 cells is linked to the last stage of virus infection. Proc. Natl. Acad. Sci. U.S.A. 1988; 85: 3570. 6. Sodroski, J., W. C. Goh, C. Rosen, K. Campbell and W. A. Haseltine. Role of the HTLV-III/LAV envelopes in sncytium formation and cytopathicity. Nature 1986; 322: 470. 7. Maddon, P. J., A. G. Dalgleish, J. A. McDougal, P. R. Clapham, R. A. Weiss, and R. Axel. The T4 gene encodes the AIDS virus receptor and is expressed in the immune system and the brain. Cell 1986; 47: 333. 70 8. Lifson, J. D., M. B. Feinberg, G. R. Reyes, L. Rabins, B. Banapour, S. Chakrabarti, B. Moss, F. Wong-Staal, K. S. Steimer, and E. G. Engleman. Induction of CD4-dependent cell fusion by the HTLV-LII/LAV envelope glycoprotein. Nature 1986; 323: 725. 9. Wong-Staal, F. Human immunodeficiency virus: genetic structure and function. Seminars in Hematology 1988; 25: 189. 10. Nabel, G. and D. Baltimore. An inducible transcription factor activates expression of human immunodeficiency virus in T cells. Nature 1986; 326:711. 11. Harper, M. E., L. M. Marselle, R. C. Gallo, and F. Wong-Staal. Detection of lymphocytes expressing human lymphotropic-T virus type UL in lymph nodes and peripheral blood from infected individuals by in situ hybridization. Proc. Natl. Acad. Sci. U.S.A. 1986; 83: 772. 12. Rickman, D., J. McCutchan, and S. Spector. Detecting human immunodeficiency virus RNA in peripheral blood mononuclear cells by nucleic acid hybridization. J. Infect. Dis. 1987; 156: 823. 13. Cohen, M. B and J. Beckstead. in "AIDS: pathogenesis and treatment" Levy, J. A. ed. Dekker, New York, in the press. 14. Gurley, R. J., K. Ikeuchi, R. A. Byrn, K. Anderson, and J. E. Groopman. CD4+ lymphocyte function with early human immunodeficiency virus infection. Proc. Natl. Acad. Sci. U.S.A. 1989; 86: 1993. 15. Tindall, B., S. Barker, B. Donavan. Characterization of the acute clinical illness associated with human immunodeficiency virus infection. Arch. Intern. Med. 1988; 148: 945. 16. Levy, J. A., L. S. Kaminski, W. J. W. Morrow, K. Steimer, P. Luciew, D. Dina, J. Hoxie, and L. Oshiro. Infection by the retrovirus associated with the human immunodeficiency syndrome. Ann. Int. Med. 1985; 103:694. 71 17. Boyko, W. J., M . T. Schechter, K. J. P. Craib, B. Willoughby, B. Douglas, P. Sestak, W. A. McLeod,and M. O'Shaughnessy. The Vancouver lymphadenopathy-AIDS Study: 7. Clinical and laboratory features of 87 cases of primary HIV infection. Can. Med. Assoc. J. 1987; 137:109. 18. Groopman, J. E.Clinical symptomatology of the acquired immunodeficiency syndrome (AIDS) and related disorders. Prog. Allergy. 1986; 37:182. 19. Moss. A. R. and P. Bacchetti. Natural history of HTV infection. AIDS 1989; 3: 55 20. Gallo, R. C, S. Z. Salahuddin, M. Popovic, G. M. Shearer, M. Kaplan, B. F. Haynes, T. J. Palker, R. Redfield, J. Oleske, B. Safai, G. White, P. Foster, and P. D. Markham. Frequent detection and isolation of cytopathic retroviruses from patients with AIDS and at risk for AIDS. Science 1984; 224: 500. 21. Rouse, B. T. and D. W. Horohov. Immunosuppression in viral infections. Rev. Infect. Dis. 1986; 8; 850. 22. Zarling, J. Primate models for evaluation of AIDS vaccines. AIDS 1988; 2 (suppl): 113. 23. Statistics from the World Health Organization and the Centres for Disease Control. AIDS any issue. 24. Biggar, R. J. The AIDS problem in Africa. Lancet 1986 i; 79. 25. Moss, A. R., P. Bacchetti, D. Osmond, W. Krampf, R. E. Chaisson, D. Stites, J. Wilber, J. Allain, and J. Carlsin. Seropositivity for HTV and the development of AIDS or AIDS related conditions: three year follow up of the San Francisco General Hospital cohort. Br. Med. Journal 1988; 296: 745. 26. Hoffmann, G. W. On I-J, a network centre pole and AIDS. In: "The Semiotics of Cellular Communication in the Immune System." Sercarz, E., Celada, F., Mitchison, N. A., and Tada, T. eds. 1988 Springer Verlag, New York. 72 27. Madhok, R., A. Grade, G. D. O. Lowe, A. Burnett, K. Froebel, E. Follett, and C. D. Forbes. Impaired cell mediated immunity in hemophilia in the absence of infection with human immunodeficiency virus. Br. Med J. 1986; 293: 978. 28. Wood, G S., C. F. Garcia, R. F. Dorfman, and R. A. Warnke. The immunohistology of follicle lysis in lymph node biopsies from homosexual men. Blood 1985; 66: 1092. 29. Friedman, A. E., L. H. Laubenstein, P. Rubinstein, E. Buimovici-Klein, M. Marmor, R. E. Stahl, K. S. Kim, I. Spigland, and S. Zolla-Pazner. Kaposi's sarcoma in homosexual men. Ann. Int. Med. 1982; 96: 693. 30. Cunningham-Rundles, S., M. A. Michelis, and H. Masur. Serum suppression of lymphocyte activation in vitro in acquired immunodeficiency disease. J. Clin. Immunol. 1983; 3: 156. 31. Folks, T., J. Kelly, S. Benn., A. Kinter, J. Justement, J. Gold, R. Redfield, K. W. Sell, and A. S. Fauci. Susceptibility of normal human lymphocytes to infection with HTLV-LLI/LAV. J. Immunol. 1986; 136: 4049. 32. Seligman, M., A. J. Pinching, F. S. Rosen, J. L. Fahey, R. M. Khaitov, D. Klatzmann, S. Koenig, N. Luo, J. Ngu, G. Riethmuller, and T. J. Spira. Immunology of human immunodeficiency virus infection and the acquired immunodeficiency syndrome. Ann. Int. Med. 1987; 107: 234. 33. Ballet, J., L. C. Couderc, C. Rabian-Herzog, C. Duval-Roy, M. Janier, F. Danon, J. Clauvel, and M. Seligmann. Impaired T-lymphocyte dependent immune responses to microbial antigens in patients with FnV-1 associated persistent generalized lymphadenopathy. AIDS 1988; 2: 291. 34. Linette, G. P., R. J. Hartzman, J. Ledbetter, and C. H. June HTV-1-infected T cells show a selective signaling defect after perturbation of CD3/antigen receptor. Science 1988; 241: 573. 73 35. Murray, J. L., E. M. Hersh, J. M. Reuben, C. G. Munn, and P. W. A. Mansell. Abnormal lymphocyte response to exogenous interleukin-2 in homosexuals with acquired immune deficiency syndrome (AIDS) and AIDS related complex (ARC). Clin. expl. Immunol. 1985; 60: 25. 36. Winkelstein, A., L. A. Kingsley, R. S. Klein, D. W. Lyter, T. L. Evans, C. R. Rinaldo, L. D. Weaver, L. L. Machen, and R. C. Schadle. Defective T-cell colony formation and IL-2 receptor expression at all stages of HTV infection. Clin. expl. Immunol. 1988; 71: 417. 37. Prince, H. E., D. J. Moody, B. J. Shubin, and J. L. Fahey. Defective monocyte function in acquired immunodeficiency syndrome (AIDS): evidence from a monocyte- dependent T-cell proliferative system. J. Clin. Immunol. 1985; 5: 21. 38. Roy, S., L. Fitzgibbon, L. Poulin, and M. A. Wainberg. Infection of human monocytes/macrophages by HIV-1: effect on secretion of IL-1 activity. Immunology 1988; 64: 223. 39. Rook, A. H., H. Masur, H. C. Lane, W. Frederick, T. Kasahara, A. M. Macher, J. Y. Djeu, J. F. Manischewitz, L. Jackson, A. S. Fauci, and G. V. Quinnan. Interleukin-2 enhances the depressed natural killer and cytomegalovirus-specific cytotoxic activities of lymphocytes from patients with the acquired immunodeficiency syndrome. J. Clin. Invest. 1983; 72: 398. 40. Bender, B. S., F. A. Auger, T. C. Quinn, R. Redfield, J. Gold, and T. M. Folks. Impaired antibody-dependent cell-mediated cytotoxic activity in patients with the acquired immunodeficiency syndrome. Clin. Exp. Immunol. 1986; 64: 166. 41. Nicholson, J. K. A., J. S. McDougal, T. J. Spira, B. M. Jones, and E. L. Reinherz. Immunoregulatory subsets of the T helper and T suppressor cell populations in homosexual men with chronic unexplained lymphadenopathy. J. Clin. Invest. 1984; 73: 191. 74 42. Nicholson, J. K. A., J. S. McDougal, and T. J. Spira. Alterations of functional subsets of T helper and T suppressor cell populations in acquired immune deficiency syndrome and chronic unexplained lymphadenopathy. J. Clin. Immunol. 1985; 5: 269. 43. Calvano, S. E., H. F. DeRiesthal, M. A. Marano, and A. C. Antonacci. The decrease in peripheral blood CD4+ T cells following thermal injury in humans can be accounted for by a concomitant decrease in suppressor-inducer CD4+ T cells as assessed using anti-CD45R. Clin. Immunol. Immunopathol. 1988; 47: 164. 44. Morimoto, C, P. L. Romain, D. A. Fox, P. Anderson, M. DiMaggio, H. Levine, and S. F. Schlossman. Abnormalities in CD4+ T lymphocyte subsets in inflammatory rheumatic diseases. Am. J. Med. 1988; 84: 817. 45. Schlossman, S. F. Address; BRC Opening Symposium, Vancouver, Canada. 1988. 46. Calabrese, L. H. Autoimmune manifestations of human immunodeficiency virus infection. Clin. Lab. Med. 1988; 8: 269. 47. Zolla-Pazner, S. Laboratory studies in AIDS: serologic studies, in "Basics on AIDS: a guide to clinicians, ed. M. Melbye Munksgaard, Copenhagen 1984; 151. 48. Nara, P. L., W. G. Robey, M. A. Gonda, S. G. Carter, and P. J. Fischinger. Absence of cytotoxic antibody to human immunodeficiency virus-infected cells in humans and its induction in animals after infection or immunization with purified envelope glycoprotein gp 120. Proc. Natl. Acad. Sci. U.S.A. 1987; 84: 3797. 49. Kopelman, R. G. and S. Zolla-Pazner. Association of human immunodeficiency virus infection and autoimmune phenomena. Am. J. Med. 1988; 84: 82. 50. Gill, H. K. and T. Godal. Deficiency of cell mediated immunity in leprosy. Prog. Allergy 1986; 37: 377. 75 51. Klatzman, D. and J. C. Gluckman. HTV infection: facts and hypotheses. Immunol. Today. 1986; 7: 291. 52. Toy, P. T., M. E. Reid, and M. Burns. Positive direct antiglobulin test associated with hyperglobulinemia in acquired immunodeficiency syndrome (AIDS). Am. J. Hematol. 1985; 19: 145. 53. Walsh, C. M., M. A. Nardi, and S. Karpatkin. On the mechanism of thrombocytopenic purpura in sexually active homosexual men. N. Engl. J. Med. 1984; 311: 635. 54. Williams, R. C, H. Masur, and T. J. Spira. Lymphocyte reactive antibodies in acquired immune deficiency syndrome. J. Clin. Immunol. 1984; 4: 118. 55. Canoso, R. T., L. I. Zon, and J. E. Groopman. Anticardiolipin antibodies associated with HTLV-m infection. Br. J. Haematol. 1987 65, 495. 56. Grant, M. D., Weaver, M. S., Tsoukas, C, and Hoffmann, G. W. Distribution of antibodies against denatured collagen in AIDS risk groups and homosexual AIDS patients suggests a link between autoimmunity and the immunopathogenesis of AIDS. 1989 submitted for publication. 57. Strieker, R. B., D. I. Abrams, L. Corash, and M. A. Shuman. Target platelet antigen in homosexual men with immune thrombocytopenia. N. Engl. J. Med. 1985; 313: 1375. 58. Karpatkin, S., M. Nardi, E. T. Lennette, B. Byrne, and B. Poiesz. Anti-human immunodeficiency virus type 1 antibody complexes on platelets of seropositive thrombocytopenic homosexuals and narcotic addicts. Proc. Natl. Acad. Sci. U.S.A. 1988; 85: 9763. 59. Berman, A., L. R. Espinoza, J. D. Diaz, J. L. Aguilar, T. Rolando, F. B. Vasey, B. F. Germain, and R. F. Lockey. Rheumatic manifestations of human immunodeficiency virus infection. Am. J. Med. 1988; 85: 59. 76 60. Sanderson, F., N. Gratecos, J. G. Fuzibet, Y. Birtwisle, A. Pesce, H. Vinti, J. P. Casuto, and P. Dejardin. Splenectomy for thrombocytopenia in HIV infected patients: outcome of the disease compared to medically treated patients. Abstract 7647, IV International AIDS Conference, Stockholm, 1988. 61. Beldjord, K., J. M. Tourani, C. Audroin, J. M. Siksik, P. Even, and J. M. Andrieu. Evolution of lymphocyte subsets in peripheral blood of splenectomized HTV patients. Abstract 7645, IV International AIDS Conference, Stockholm, 1988. 62. Andrieu, J. M., P. Even, A. Venet, J. M. Tourani, M. Stern, W. Lowenstein, C. Audroin, D. Erne, D. Masson, H. Sors, D. Israel-Biet, and K. Beldjord. Effects of cyclosporin on T-cell subsets in human immunodeficiency virus disease. Clin. Immunol. Immunopathol. 1988; 47:181. 63. Dwyer, J. M., J. D. McNamara, L. H. Sigal, and C. C. Wood. Immunological abnormalities in patients with the acquired immune deficiency syndrome (AIDS). Clin. Immun. Rev. 1984; 3: 25. 64. Jackson, G. G., M. Rubenis., M. Knigge, J. T. Perkins, D. A. Paul, J. C. Despotes, and P. Spencer. Passive immunoneutralization of human immunodeficiency virus in patients with advanced AIDS. Lancet 1988 i; 647. 65. Bruster, H. T., B. M. E. Kuntz, and J. W. Scheja. Autovaccination plus heat inactivated autologous plasma in AIDS patients. Lancet 1988 i; 1284. 66. Lane, H. C, J. Feinberg, V. Davey, L. Deyton, M. Baseler, J. Manischewitz, H. Masur, J. A. Kovacs, B. Herpin, R. Walker, J. A. Metcalf, N. Salzman, G. Quinnan, and A. S. Fauci. Anti-retroviral effects of interferon-alpha in AIDS-associated kaposi's sarcoma. Lancet 1988 i; 1218. 77 67. De Wit, R., C. A. B. Boucher, K. H. N. Veenhof, J. K. M. E. Scattenkerk, P. J. M. Bakker, and S. A. Danner. Clinical and virological effects of high-dose recombinant interferon-alpha in disseminated AIDS-related kaposi's sarcoma. Lancet 1988 i; 1214. 68. Dournon, E., W. Rozenbaum, C. Michon, C. Perrone, P. DeTruchis, E. Bouvet, M. Levacher, S. Matheron, S. Gharakhanian, P. M. Girard, D. Salmon, C. Leport, M. C. Dazza, B. Regnier. Effects of zidovudine in 365 consecutive patients with AIDS or AIDS-related complex. Lancet 1988 i; 1297. 69. Larder, B. A., G. Darby, and D. D. Richman. HIV with reduced sensitivity to zidovudine (AZT) isolated during prolonged therapy. Science 1989; 243: 1731. 70. Klatzmann, D. and L. Montagnier. Approaches to AIDS therapy. Nature 1986; 319: 10. 71. Andrieu, J. M., P. Even, and A. Venet. AIDS and related syndromes as a viral induced autoimmune disease of the immune system: an anti-MHC LI disorder. Therapeutic implications. AIDS Research 1986; 2: 163. 72. Ziegler, J. L. and D. P. Stites. Hypothesis: AIDS is an autoimmune disease directed at the immune system and triggered by a lymphotropic retrovirus. Clin. Immunol. Immunopathol. 1986; 41: 305. 73. Golding, H., F. A. Robey, F. T. Gates, W. Linder, P. R. Beining, T. Hoffman, and B. Golding. Identification of homologous regions in HTV-1 and human class II p 1 domain. J. Exp. Med. 1988,167: 904. 74. Beretta, A., F. Grassi, M. Pelagi, A. Clivio, C. Parravicini, G. Giovinazzo, F. Andronico, L. Lopalco, P. Verani, S. Butto, F. Titti, G. Battista Rossi, E. Ginelli, and A. G. Siccardi. HIV env glycoprotein shares a cross-reacting epitope with a surface protein present on activated human monocytes and involved in antigen presentation. Eur. J. Immunol. 1987; 17: 1793. 78 75. Krah, D. L. and P. W. Choppin. Mice immunized with measles virus develop antibodies to a cell surface receptor for binding virus. J. Virol. 1988; 62:1565. 76. Steel, C. M., D. Beatson, R. J. G. Cuthbert, H. Morrison, C. A. Ludlam, J. F. Peutherer, P. Simmonds, and M. Jones. HLA haplotype A1B8DR3 as a risk factor for HIV related disease. Lancet 1988 i; 1185. 77. Hoffmann, G. W., T. A. Kion, R. B. Forsyth, K. G. Soga, and A. Cooper-Willis. The N-dimensional network, in "Theoretical Immunology, Part 2. A. S. Perelson, ed. Addison Wesley, p. 291. 78. Kion, T. A. unpublished. 79. Hoffmann, G. W., A. Cooper-Willis, and M. Chow. A new symmetry: A anti-B is anti-(B anti-A), and reverse enhancement. J. Immunol. 1986; 137: 61. 80. McDougal, J. S., M. S. Kennedy, J. M. Sligh, S. P. Cort, A. Mawle, and J. K. A. Nicholson. Binding of HTLV-ILT/LAV to T4+ T cells by a complex of the 110k viral protein and the T4 molecule. Science 1986; 231: 382. 81. Gay, D., P. Maddon, R. Sekaly, M. A. Talle, M. Godfrey, E. Long, G. Goldstein, L. Chess, R. Axel, J. Kappler, and P. Marrack. Functional interaction between human T cell protein CD4 and the major histocompatibility complex HLA-DR antigen. Nature 1987; 328: 626. 82. Robinson, W. E., D. C. Montefiori, and W. M. Mitchell. Antibody-dependent enhancement of human immunodeficiency virus type 1 infection. Lancet 1988 i: 790. 83. Sattentau, Q. J., A. G. Dalgleish, R. A. Weiss, and P. C. L. Beverley. Epitopes of the CD4 antigen and HIV infection. Science 1986; 234: 1120. 84. Chanh, T. C, G. R. Dreesman, and R. C. Kennedy. Monoclonal anti-idiotypic antibody mimics the CD4 receptor and binds human immunodeficiency virus. Proc. Natl. Acad. Sci. U.S.A. 1987; 84: 3891. 79 85. Savage, S. M., R. P. Searles, G. M. Troup, and C. M. Brozek. Anti-idiotypic antibodies to Anti-DR in patients with rheumatoid arthritis. Clin. Immunol. Immunopathol. 1987; 42: 183. 86. Fields, A. P., D. P. Bednarik, A. Hess, and W. S. May. Human immunodeficiency virus induces phosphorylation of its cell surface receptor. Nature 1986; 333; 278. 87. Blum, H., H. Beier, and H. J. Gross. Improved silver staining of plant proteins, RNA, and DNA in polyacrylamide gels. Electrophoresis 1987; 8: 93. 88. Towbin, H., T. Staehelin, and J. Gordon. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U.S.A. 1979; 76: 4350. 89. Blake, M. S., K. H. Johnston, G. J. Russel-Jones, and E. C. Gotschlich. A rapid, sensitive method for detection of alkaline phosphatase conjugated anti-antibody on western blots. Anal. Biochem. 1984; 136: 175 90. Reid, K. B. M. Isolation, by partial pepsin digestion, of the three collagen-like regions present in subcomponent Clq of the first component of human complement. Biochem. J. 1976; 155: 5. 91. Bayer, E. A. and M. Wilchek. in "Methods of biochemical analysis", D. Glick ed. John Wiley and Sons New York 1980; 26: 1. 92. Giesecke, J., G. Scalia-Tomba, O. Berglund, E. Berntorp, S. Schulman, and L. Stigendal. Incidence of symptoms of AIDS in 146 Swedish hemophiliacs and blood transfusion recipients infected with human immunodeficiency virus. Br. Med. Journal 1988; 297: 99. 93. R. A. Kaslow, J. P. Phair, IT. B. Friedman, D. Lyter, R. E. Solomon, J. Dudley, B. Frank, and W. Blackwelder. Infection with human immunodeficiency virus: clinical manifestations and their relationship to immune deficiency. Ann. Int. Med. 1987; 107: 474. 80 94. Lambert, P. H., M. Goldman, J. C. Renversez, P. Morel, S. Luzuy, and J. Louis. Immune complex diseases resulting from idiotypic or allogeneic interactions, in "Recent advances in SLE" Academic Press, London 1984; p. 69. 95. Reuben, J. M., L. Liang, A. Rios, S. Li, G Brewton, P. Sarin, and P. W. Mansell. Polyclonal activation of B cells by HIV isolates is T-cell dependent. Abstract 2514 IV International AIDS Conference, Stockholm, 1988. 96. Jerne, N. K. Towards a network theory of the immune system. Ann. Immunol. 1974; 125: 373. 97. Londei, M., C. M. Savill, A. Verhoef, F. Brennan, Z. A. Leech, V. Duance, R. Maini, and M. Feldmann. Persistence of collagen type II-specific T-cell clones in the synovial membrane of a patient with rheumatoid arthritis. Proc. Natl. Acad. Sci. U.S.A. 1989; 86: 636. 98. Schwartz, R. S. Address, 19 International Leukocyte Culture Conference, Banff, Canada, 1988. 99. Hoffmann, G. W. On network theory and H-2 Restriction, in "Contemporary topics in immunology volume LT" ed. N. Warner Plenum publishing, New York 1980. 100. Diamond, D. C, B. P. Sleckman, T. Gregory, L. A. Lasky, J. L. Greenstein, and S. J. Burakoff. Inhibition of CD4+ T cell function by the HIV envelope glycoprotein, gp 120. J. Immunol. 1988; 141: 3715. 101. Lundin, K., A. Karlsson, A. Nygren, A. Lofstrom, D. Gigliotti, and H. Wigzell. Certain human gp 120-HIV antibodies react with anti-CD4 antibodies. Scand. J. Immunol. 1988; 27: 113. 102. Hoffenbach, A., P. Langlade-Demoyen, G. Dadaglio, E. Vilmer, F. Michel, C. Mayaud, B. Autran, and F. Plata. Unusually high frequencies of HIV-specific cytotoxic T lymphocytes in humans. J. Immunol. 1989; 142: 452. 81 103. Souroujon, M., M. E. White-Scharf, J. Andre-Schwartz, M. L. Gefter, and R. S. Schwartz. Preferential autoantibody reactivity of the preimmune B cell repertoire in normal mice. J. Immunol. 1988; 140: 4173. 104. Engvall, E. and E. Ruoslahti. Affinity of fibronectin to collagens of different genetic types and to fibrinogen. J. Exp. Med. 1978; 147: 1584. 105. Ogden, T. M., H. E. Fischer, F. J. Lui, A. Rios, and B. Lichtiger. Plasma fibronectin values in patients with acquired immunodeficiency syndrome (AIDS) and AIDS-related complex. Am. J. Clin. Path. 1988; 90: 293. 106. Reid, K. B. M. Proteins containing collagen sequences. in"Collagen in health and disease" eds. J. B. Weiss and M. I. V. Jayson Churchill Livingstone, Edinburg 1982; p. 18. 107. Kleinman, H. K. and C. M. Wilkes. Interaction of fibronectin with collagen. in"Collagen in health and disease" eds. J. B. Weiss and M. I. V. Jayson Churchill Livingstone, Edinburg 1982; p. 198. 108. Livneh, A., I. Sarova, D. Michaeli, M. Pras, K. Wagner, H. Zakut, and H. Soreq. Antibodies against acetylcholinesterase and low levels of cholinesterases in a patient with an atypical neuromuscular disorder. Clin Immunol. Immunopathol. 1988; 48: 119. 109. Antes, A., H. Heinz, and M. Loos. Evidence for the presence of autoantibodies to the collagen-like portion of Clq in systemic lupus erythematosus. Arthritis Rheum. 1988; 31: 457. 110. Reid, K. B. M., R. B. Sim, and A. P. Faiers Inhibition of the reconstitution of the haemolytic activity of the first component of human collagen by a pepsin-derived fragment of subcomponent Clq. Biochem. J. 1977; 161: 239. Ul . Golan, M. D., R. Burger, and M. Loos. Conformational changes in Clq after binding to immune complexes: detection of neoantigens with monoclonal antibodies. J. Immunol. 1982; 129: 445. 82 112. Heinz, H. P., K. Rubin, A. La well, and M. Loos. Common epitopes in Clq and collagen type II. Mol. Immunol. 1989; 26: 163. 113. Choi, E. K. K., P. A. Gatenby, N. W. McGill, J. F. Bateman, W. G. Cole, and J. R. York. Autoantibodies to type II collagen: occurrence in rheumatoid arthritis, other arthritides, autoimmune connective tissue diseases, and chronic inflammatory syndromes. Ann. Rheum. Dis. 1988; 47: 313. 114. Tung, K. S., K. F. Foster, D. C. Bernstein, P. W. Kriebel, S. M. Payne, and G. M. Shearer. Elevated allogeneic cytotoxic T-lymphocyte activity in peripheral blood leukocytes of homosexual men. J. Immunol. 1985; 135: 3163. 115. Anderson, D. C, W. C. A. van Schooten, M. E. Barry, A. A. M. Janson, T. M. Buchanan, and R. R. P. deVries. A mycobacterium leprae-specific human T-cell epitope cross reacts with an HLA-DR2 peptide. Science 1988; 242: 259. 116. Raj, L. S., S. A. Patel, and A. Girdhar. Antigenic relatedness between human immunodeficiency virus (HIV) and M. leprae. Abstract 5549, IV International AIDS Conference, Stockholm, 1988. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items