Antigenic characterization of the human immunodeficiency viruses

Antigenic characterization of the human immunodeficiency viruses Max Essex, DVM, PhD, Phyllis J. Kanki, DVM, DSc, Richard Marlink, MD, Min-Ji Chou, DSc, and Tun-Hou Lee, DSc Boston, Massachusetts As more is learned about the human immunodeficiency viruses HIV-1 and HIV-2, increasingly sophisticated methods ofacquired immunodeficiency syndrome (AIDS) treatment and prevention may be implemented. Integral to an understanding of these viruses is an analysis of both the viral antigens and the host-immune responses to these antigens, which may differ from HIV-1 to HIV-2. Because levels of both antigen and antibody vary throughout disease development, knowledge of how and why such changes occur will lend insight into viral pathogenic mechanisms and will facilitate the development ofdifferential diagnostic tests for classifying AIDS patients and their disease states. This task becomes very complex when dealing with HIV viruses because they possess an unprecedented number of regulatory genes for members of the retrovirus family. (J AM ACAD DERMATOL 1990;22:1206-10.)

When considering issues of vaccine development, blood screening, and clllical diagnosis and prognosis as they relate to acquired immunodeficiency syndrome (AIDS), it is necessary to look at the relative antigenicity of the proteins of the HIV viruses. A thorough examination of these proteins and their subunit epitopes in both native and nonnative forms lends insight into their relative values for clinical use. The HIV-1 virus is considerably more complex than previously characterized retroviruses in that it contains several more genes. To the extent that their functions can be determined, these additional genes are involved in the regulation of viral replication. 1 The simpler retroviruses, such as the mouse, cat, and chicken leukemia viruses, contain three of the genes found also in the HIV-1 genome: gag, pol, and env. These genes encode for various essential components of the mature viruses. The pol product corresponds to the molecule of reverse transcriptase that must be packaged within every retroviral capsid. It is this unique polymerase that is responsible for the unFrom the Department of Cancer Biology, Harvard School of Public Health. Supported in part by NIH grants CA-39805 (M. E.), HL-33774 (M. E.), and AI-23604, HL-43561 (T. H. L.) and Department of Defense grants DAMD 17-87-C-70n (P. J. K.) and DAMD 17-87C-7031 (T. H. L.). Reprint requests: Max Essex, DVM, PhD, Department of Cancer Biology, Harvard School of Public Health, 665 Huntington Ave., Boston, MA 02115.

16/0/19163

1206

usual replication mechanism used by all retroviruses, a mechanism that allows them to synthesize DNA from RNA templates, reversing the normal order of DNA to RNA synthesis-thus the name "retrovirus." The gag gene product corresponds to various coat proteins and other proteins found in the interior of the virus, while the env gene encodes for the protein components of the viral envelope. 2 These envelope glycoproteins protrude from the viral surface and are usually the major targets for any effective immune response. They are also the most immunogenic portions of the virus, whether the antibodies that react with them are of value for disease prevention or merely for serodiagnosis. 3 In addition to these three standard retroviral genes, HIV-l possesses other genes, including rev, tat, vij, nef vpr, and vpu. All these genes may be involved in the process of viral replication, although the exact functions for several of them are still not completely known. Nonetheless, an understanding of these functions will undoubtedly be of great importance when we try to determine why the virus kills some cells more readily than others and why it remains latent for longer periods in some cells than in others. In fact, it is already clear that HIV-l can be kept in check under some circumstances because it does not cause disease in inoculated chimpanzees. Apparently the immune system of the chimpanzee is well enough equipped to control viral replication that the animal, although infected, does not develop

Volume 22 Number 6, Part 2 June 1990

the clinical symptoms of disease. This phenomenon may be seen in humans as well, in whom various potential pathogens, suchas the Epstein-Barr virus and certain herpesviruses, invoke a similarly effective immune response in the majority of infected persons. The mechanisms responsible for these kinds of regulatory ph~nomena remain to be determined. The antigenic profile of the HIV-1 proteins has been relatively well characterized. Of course, if one considers the proteins of the isolated virus, one gets a different profile from that seen when looking at virus-specific antigens in HIV-infected cells. It was these infected cells that provided the main source of antigen for some of the earliest confirmatory tests performed by fluorescent antibody techniques. 4 Other assays, such as radioimmunoprecipitation, also use whole-cell homogenate from infected cells as antigen. These assays reflect the presence of various precursor polyproteins such as gp160 and p55, which are made as the initial gene products of viral messages and are subsequently cleaved to form smaller proteins such as those found in the extracellular virus particles. Human antisera, or sera COIltaining antibodies, react regularly with the envelope proteins gp120 and gp160, fairly regularly with the gag gene proteins or the pol gene proteins, and only minimally with the protein products of the regulatory genes. BLOOD SCREENING FOR HIV ANTIBODIES

In screening tests such as the enzyme-linked immunosorbent assay (ELISA) or the Western blot confirmatory test, extracellular virus is often used as the antigen source. Here gp160 does not appear in the preparation because it is only an intermediatestage product involved in viral production and therefore does not localize in mature viruses. Instead it is cleaved to gpl20 and gp41. As in assays in which a homogenate from HIVinfected cells is used as the antigen source, those assays in which isolated virus is used as antigen can still detect antibody to p24 and other core proteins, although in certain stages of the disease some individuals may show a high antibody titer to gp120 but not yet to p24. 5 Other proteins, such as gp64, gp53, and gp34, the reverse transcriptase and endonuclease products of the polymerase gene, also appear in preparations of isolated virus and are quite regularly detected with human serum antibodies. They also appear to be highly immunogenic, and they are being more appreciated as an important source of

Antigenic characterization of HIVs 1207 antigens for inclusion in tests designed to look for specific antibodies in human sera. The most frequently used confirmatory test is the immunoblot, or Western blot. This assay is often run with a reduced preparation of the whole virus. Unfortunately, the reduction process destroys the immunogenicity of a good portion of the gp120, and therefore this particular assay may not be as sensitive as others. Nevertheless, this type of assay can often afford some very clear profiles for antibodies directed to both HIV-~ and HIV-2, a second strain of human immunodeficiency virus. At the same time, however some unusual or indistinct profiles can be produced that are often challenging to interpret. In some cases there is evidence ofonly very faint p24 activity, and one wonders whether these profiles represent early seroconversion or merely nonspecific reactivity. Usually such am~iguities are, in fact, nonspecific reactivity. In other cases there is no gp41 antibody, but increased reactivity with the pol antigens. These cases usually indicate exposure to HIV with a specific antibody response. There are also some antibodies in human serum that are directed to the specific gene products ofthe various HIV-1 regulatory proteins, but none ofthese antigens is highly immunogenic. A determination of antibody responses to these proteins may become increasingly important, because it is likely that at least some of them will be useful for such purposes as prognostic monitoring and differential diagnosis. Because not all ofthese regulatory proteins arepart of the virus itself but are involved only in producing the virus, they have to be obtained either by separation from infected cell material or by expression via genetic engineering techniques. Although most of the regulatory proteins are generally of low or modest immunogenicity, all are immunogenic in at least some persons. Based on results obtained by studying persons infected by transfusion, we have been able to d~ter­ mine which types of antibody reactivity occur on a stagewise basis after exposure to the virus. The course of antibody development could be much slower in the case of viral acquisition via some other route-for example, through the mucosal membrane of the reproductive tract. Nonetheless, according to the transfusion studies, there is a 2- to 6week period before antibody seroconversion to any of the HIV-1 antigens occurs. Thus, there is a period of about 1 month during which a person could be excreting virus, hut infected blood, which is capable

1208 Essex et al. of viral transmission, would not be detected or screened out by a blood bank antibody assay. The first antibodies to be detected are usually directed against gp120. In another I to 3 weeks antibody conversion to the p24 gag proteins and to the polymerase proteins occurs. It is only at this point that Western blot test results can be interpreted as positive, and even then the response to gp41 may not be quite apparent. 5 After an additional 1 to 3 weeks, antibodies to gp41 develop, and positive Western blot test results become clear and apparent. Several years of persistent viral infection usually pass before the development of AIDS. With the development of the clinical presentation of AIDS may come a loss of some of these antibodies. For this reason, some investigators test for both p24 antigen and anti-p24 antibodies in their diagnostic procedures. 6 CLINICAL DIAGNOSIS AND PROGNOSIS Some associations have also been made between antibody types and clinical disease progression. For example, antibodies to the reduced gp120 class of epitopes, or R domains, are more prevalent in healthy carriers than in AIDS patients.? There is some hope that this information may be useful for identification of domains of gp120 that might be the most valuable for vaccine development. Evidence for this association comes from several sources. For example, in the Walter Reed staging system, antibodies to most domains of gp120 are as equally likely to be seen in the least advanced stage of disease as they are in the most severe stage of AIDS.8 Antibodies to the R domains are least likely to be seen in AIDS patients, however, and are most likely to be seen in more healthy persons. Of even greater significance is an experiment in which 98 patients were followed for 18 to 24 months to see whether they developed AIDS-related complex (ARC). The results indicated that only about 4 of 85 (5%) of those who had antibodies to the R domains of gp120 developed ARC within 18 to 24 months. Conversely, 4 of 13 (31 %) patients who lacked such antibodies went on. to develop ARC. Another marker that has been and will continue to be used for staging the course of HIV-1 disease is the core protein p24. Levels of both antigen and antibody are significant in the staging process. During the earliest stages of infection, p24 antigen levels appear elevated. A period of latency follows during

Journal of the American Academy of Dermatology

which a person harbors the virus but remains healthy. Throughout this time, which often extends for years, the virus does not replicate. Rather, it remains insidiously integrated in the host cell DNA as a provirus, inactive but potentially pathogenic. Because the virus does not replicate much during this latent period, p24 (core protein) antigen levels remain low. At the time of clinical disease, when the provirus becomes activated and begins a process of replication and reinfection, p24 antigen levels again rise to their previously elevated levels. The titer of anti-p24 antibodies appears to follow a pattern that is roughly reciprocal to that of the p24 antigen. During latency, antibody titer remains high. In contrast to what occurs with the p24 antigen, however, anti-p24 antibody titer is low both at initial infection and again later at the time of disease progression. 9 This phenomenon can be interpreted in two parts: (1) in the early stages of infection, p24 antigen may be detectable in serum well before the antibodies have a chance to develop; (2) in the late stages, it seems likely that viral replication is so extensive that there is enough p24 antigen circulating to saturate the host's supply of anti-p24 antibody, thus soaking up antibodies that would otherwise be detected. There is an exception to this general trend among AIDS patients in Central Africa 10; these patients appear to have significant levels of antibodies to p24 even throughout the most severe stages of disease. In addition to p24, there are other viral antigens to which one can measure antibody responses according to stage of infection. Some of them present even more dramatic results than p24. Yet, none is so valuable that it can be used alone as an indication of disease progression; it is still necessary to consider these data in combination for them to be most useful in staging patients. Perhaps one of the most interesting of these additional markers is the p 19 rev protein. Increased antibodies are most likely to be seen at the stage of ARC just before the development of AIDS, but they then decrease with the onset of clinical AIDS. In the case of nef and viJ, as with p24, antibodies decline as a patient progresses to late-stage clinical disease, whether it be AIDS or ARC. For antigenic characterization ofthe HIV viruses, it is also necessary to distinguish between two types of human immunodeficiency viruses, HIV-1 and HIV-2. HIV-1 is clearly most prevalent in Central

Volume 22 Number 6, Part 2 June 1990

Africa in the region including Zaire, Ruanda, Burundi, Uganda, Congo, Kenya, Zambia, Tanzania, and the Central African Republic. ll HIV-2, on the other hand, is most prevalent in extreme West African nations such as Senegal, Guinea-Bissau, Burkina Faso, and the Ivory Coast,12 The HIV-2 virus shows a Western blot profile that differs from that of HIV-1. In fact, HIV-2 is very closely related to the simian immunodeficiency virus (SIV) family in at least two respects 13: (1) their transmembrane protein is smaller than the gp41 normally seen in HIV_1 14; (2) serums from HIV-2infected persons react far better with the proteins of the monkey virus than they do with the HIV-1 virus, especially when antibody reactivity to the more type-specific gp41 and gp120 antigens is measured. Recent studies conducted in several regions of West Africa indicate that there are considerable variations in HIV-2 prevalence from country to country.12 What is consistent, however, is that sexually active groups regularly have the highest levels of infection. Further, the rates of infection in healthy hospital workers are not much different from the rates of infection in persons with severe infectious diseases that might be indicative of AIDS or AIDSlike immune deficiency-diseases such as tuberculosis or pneumonia. We conducted prospective studies to determine whether healthy seropositive prostitutes developed lymphadenopathy.15 In HIV-2-seropositive prostitutes in Senegal, monitored now for 100 personyellrs of follow-up, no significant progression to lymphadenopathy has been noted. 15 In contrast, similar studies in Central Africa have indicated that healthy persons seropositive for HIV-1 developed significant rates oflymphadenopathy or ARC within the same relative period of follow-up. 16. 17 These and other studies suggest that the HIV-2 virus is not as highly pathogen ic as HIV-1. Although HIV-2 has been isolated from some patients with AIDS or AIDS-like disease, they appear to have a less severe illness than those infected with HIV1. 18,19 In considering these results, it is necessary to understand that the diagnosis of clinical AIDS is often difficult in Africa, where modern equipment is often not available and where the opportunistic pathogens are different from those found in AIDS patients in the West. Diagnosis is particularly difficult in West Africa, where AIDS itself is relatively infrequent. In

Antigenic characterization ofHWs

1209

addition, in such regions, a significant proportion of patients that fit the World Health Organization criteria for the signs and symptoms of AIDS may not have an HIV-related disease at all. However, more advanced diagnostic tests for the AIDS viruses will become available within the next few years. Some of these tests will probably even be useful for monitoring patients, with attention given both to clinical costs and effectiveness of therapy. It is now clear that there is more than one HIV virus, and that there are in fact HIV-related viruses, some of which have different disease-causing abilities from those of the first AIDS virus, HIV-1. We can only hope that the lessons learned from these other viruses will ultimately help us in the development of effective vaccines. REFERENCES

1. Haseltine WA, Wong-Staal F. The molecular biology of the AIDS virus. Sci Am 1988;259:52-62. 2. Allan JS, Coligan JE, Lee TH, et al. A new HTLVIII/LAV encoded antigen detected by antibodies from AIDS patients. Science 1985;230:810-3. 3. Kitchen LW, Barin F, Sullivan JL, et al. Aetiology of AIDS--antibodies to human T-cell leukemia virus (type III) in hemophiliacs. Nature 1984;312:367-9. 4. Sandstrom EG, Schooley RT, Ho DD, et al. DetectIOn of human anti-HTLV-I1I antibodies by indirect immunofluorescence using HTLV-III infected H9 cells. Transfusion 1985;25:308-12. 5. Chou MJ, Lee TH, Hatzakis A, et al. Antibody responses in early human immunodeficiency virus type I infection in hemophiliacs. J Infect Dis 1988;157:805-11. 6. Allain JP, Laurian Y, Paul DA, et al. Serological markers in early stages of human immunodeficiency virus infection in hemophiliacs. Lancet 1986;2:1233-6. 7. Lee TH, Redfield RF, Chou MJ, et al. Association between antibody to envelope glycoprotein gp120 and the outcome ofHIV infection. In: Ginsberg H, Brown F, Lerner R, et al., eds. Vaccines 1988: New chemical and genetic approaches to vaccination. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory, 1988:373-7. 8. Redfield RR, Wright DC, Tramont Be. The Walter Reed staging classification for HTLV-IU/LAV infection. N Engl J Med 1986;314:131-2. 9. Orgad S, Malone G, Zaizov R, et al. Antibodies to HIV in Israeli hemophiliacs: relationship between serological profile and disease development. AIDS Res Hum Retroviruses 1987;3:323-32. 10. Barin F. AIDS vaccine predictions [Letter]. Nature 1987;328:21. 11. Mann JM, Chim J, Piot P, et al. The international epidemiology of AIDS. Sci Am 1988;259:829. 12. Kanki PI, M'Boup S, Richard D, et al. Human Tlymphotrophic retrovirus type IV and the human immunodeficiency virus in West Africa. Science 1987;236:827-31. 13. Essex M, Kanki PJ. Origins of the AIDS virus. Sci Am 1988;259:64-71. 14. Harin F, M'Boup S, Denis F, et al. Serological evidence for

Journal of the American Academy of Dermatology

Essex et al. a virus related to simian T-lymphotrophic retrovirus III in residents of West Africa. Lancet 1985;2:1387-90. 15. Marlink RG, Thior I, Siby T, et al. Observations on the natural history of HIV-2. Fifth International Conference 011 AIDS, Montreal, Canada, 1989. Unpublished abstract. 16. Simonsen N, Plummer F, Gakinya M, et al. Longitudinal study of a cohort ofHTLV-III/LAV infected prostitutes in Nairobi. Second International Conference on AIDS, Paris 1986, Abstract S17d, Communication 197, p 105. 17. Ngaly B, RyderRW, Kapita B, et al. Human irnmunode-

ficiency virus infection among employees in an African hospital. N Engl J Med 1988;319:1123-7. 18. Clavel F, Mansinho K, Chamaret S, et al. Human immunodeficiency virus type 2 infection associated with AIDS in West Africa. N Engl J Med 1987;316:1180-5. 19. Marlink RG, Ricard D, M'Boup S, et al. Clinical, hematologic, and immunologic cross-sectional evaluation of individuals exposed to human immunodeficiency virus type 2 (HIV-2). AIDS Res Hum Retroviruses 1988;4:137-48.

Langerhans cells in HIV-l infection Georg Stingl, MD,a Klemens Rappersberger, MD,a Erwin Tschachler, MD,a Suzanne Gartner, PhD,b Veronika Groh, MD,a Dean L. Mann, MD,c Klaus Wolff, MD,a and Mikulas Popovic, MD, PhDb Vienna, Austria, and Bethesda and Frederick, Maryland

The skin-specific immune surveillance system protects against invading microorganisms and transformed cells expressing tumor-specific neoantigens. This system includes antigenpresenting Langerhai1.s cells, dermal and epidermal T lymphocytes, cytokine-producing keratinocytes, and draining peripheral lymph nodes. In patients infected with human immunodeficiency virus-I (HIV-l), this surveillance system appears to be compromised, as evidenced by a reduction in the epidermal Langerhans cell population. Because human epidermal Langerhans cell express surface-bound CD4 antigens, HLA-DR antigens, and FcIgG receptors, all of which are involved in HIV-l binding to, or entry into, the target cell, the reduction in Langerhans cells in patientS with acquired immunodeficiency syndrome (AIDS) or AIDS~relatedcomplex (ARC) may be a direct consequence of HIV-I infection and subsequent injury to Langerhans cells. Detailed ultrastructural studies have confirmed moderate to severe morphologic damage in some Langerhans cells of such patients and the presence of HIV-I-like particles on Langerhans cell surface membranes and in the extracellular spaces. The biologic consequences of Langerhans cell infection by HIV-l could be either impaired antigen presentation function of viable Langerhans cells or possible transmission of the retrovirus to the T-cell compartment in skin or lymph nodes, with subsequent depletion of CD4+ T cells via widespread syncytia formation between HIV-l-infected and noninfected cells. The facts that herpes simplex virus, specific cytokines, and ultraviolet B radiation can activate signals for HIV-I expression and that epidermal cells can elaborate large amounts of cytokines, particularly with enhanced ultraviolet B exposure, may have important clinical implications for HIV-I-infected patients. (J AM ACAD DERMATOL 1990; 22:1210-7.)

From the Department of Dermatology 1, Division of Cutaneous Immunobiology, University of Vienna Medical School"; the Laboratory of Tumor Cell Biology, National Cancer Institute, Bethesdab; and the Laboratory of Viral Carcinogenesis, National Cancer Institute, Frederick.< Supported in part by a grant from the Medizinisch-Wissenschaftlicher Fonds des Biirgermeisters der Bundeshauptstadt Wien, Vienna, Austria. Reprint requests: Georg Stingl, MD, Head, Division of Cutaneous Tmmunobiology, Department of Dermatology T, University of Vienna Medical School, Alser Strasse 4, A-1090 Vienna, Austria. 16/0/19164

1210

Skin and adjacent mucosal surfaces are major manifestation sites of opportunistic infections and neoplasms that characterize the acquired immunodeficiency syndrome (AIDS).l-S There apparently exists a relationship between the severity of skin infections and the quantitative depletion of the CD4+ T cell subset.} Thus, most infectious and neoplastic processes in the skin of human immunodeficiency virus type 1 (HIV-1 )-infected individuals are a direct consequence of the HIV-l-induced loss of CD4+