Lymphadenopathy-Associated Virus: From Molecular Biology to Pathogenicity

Transcription

1 Lymphadenopathy-Associated Virus: From Molecular Biology to Pathogenicity LUC MONTAGNIER, M.D.; Paris, France Recent data indicate that the lymphadenopathyassociated virus (LAV) is morphologically similar to animal lentiviruses, such as equine infectious anemia and visna viruses. This finding, together with the crossreactivity of the core proteins of LAV with those of the equine infectious anemia virus and a similarity in genome structure and biological properties, allows LAV to be placed in the retroviral subfamily of Lentivirinae. Molecular data indicate a high degree of genetic variation of the virus, especially in the envelope gene, which have important implications for the origin of the virus (the T4 lymphotropism may be a recently acquired property) and for future immunization. Another problem is the role of viral infection in the induction of irreversible immunodeficiency. This syndrome occurs in a minority of infected persons, who generally have in common a past of antigenic stimulation and of immune depression before LAV infection. From the Viral Oncology Unit, Institut Pasteur; Paris, France. Two YEARS of intensive research and international collaboration have resulted in the identification of the primary causative agent of the acquired immunodeficiency syndrome (AIDS). This agent is a type of retrovirus that has not been previously recognized. Major steps in this research are shown in Table 1. The previous experience with retroviruses and the modern techniques in molecular biology have enabled rapid characterization of this retrovirus, and we have today an accurate picture of the anatomic features of the virus and its biologic properties. Although we have not been able to reproduce the disease directly in animals, we have good indirect evidence that infection with this virus in humans can result in the socalled AIDS-related complex or, less frequently, in frank AIDS. However, the exact conditions that determine occurrence of irreversible immunodeficiency, a benign syndrome, or an inapparent infection are not known. To solve these problems, we must address the following questions: How can we place this new group of viruses in the retrovirus family? This question is not only an academic debate between virologists, because we can draw relevant information for pathogenic mechanisms from comparison with animal models. Second, what is the relevance of the cytopathic effects observed in vitro to the in-vivo pathogenicity? Third, what can we learn from seroepidemiologic data, drawn mostly from the cases of transfusion-acquired AIDS, about possible cofactors? Finally, where did the virus come from and did it acquire its current pathogenicity recently? Classification ULTRASTRUCTURE AND VIRION MORPHOGENESIS As reported earlier (1), three characteristic morphologic aspects of the virus can be seen in sections of infected lymphocytes or certain lymphoid cell lines. First, budding of particles occurs at the cell surface: The ribonucleoprotein core forms a dense crescent that is linked to the plasma membrane by structured material. This aspect differs from that seen with type-c particle or human T-cell leukemia virus type I (HTLV-I), in which the core in formation is separated from the plasma membrane by an electron-lucent space. It differs also from the budding type-d particles in which the core is already condensed during morphogenesis (Figure 1). Second, the virus has immature free particles. This stage is an intermediate one in which the released particles still have an uncondensed core. Transition stages of condensation can be seen (Fig- Annals of Internal Medicine. 1985;103: American College of Physicians 689

2 Table 1. Important Steps in the Discovery of the Etiologic Agent of the Acquired Immunodeficiency Syndrome (AIDS) The human T-cell leukemia virus hypothesis January 1983 Isolation of lymphadenopathy-associated virus (LAV1) from a patient with lymphadenopathy 1983 Characterization of the virus (morphology and core proteins) Isolation of immunodeficiency-associated virus (IDAV) from several patients with AIDS Demonstration of the tropism of LAV for T4+ cells and its cytopathic effect Establishment of an enzyme-linked immunosorbent assay with a control cellular antigen, the first seroepidemiologic data 1984 Isolation of similar viruses in the United States (human T-lymphotropic virus type III, LAV, AIDS-related virus) Growth of the virus in continuous cell lines Identification of the viral glycoprotein Molecular cloning and sequencing: evidence for a unique genome structure ure 2). Third, mature virions show a small eccentric core, sometimes bar-shaped. Projections from the cell surface (glycoprotein?) seem to be smaller than those in the immature stage, perhaps due to partial removal of carbohydrates or to configurational changes of the viral glycoprotein. If we compare these ultrastructural characteristics with those of animal or other human retroviruses, we see the closest resemblance with those of the equine infectious anemia virus (2), a virus whose classification within the three subfamilies of Retroviridae is uncertain but which has some similarities with Lentivirinae. not recognized by patient antibodies, because the patient antibodies do not immunoprecipitate the equine infectious anemia virus p25. The only significant homology between LAV and the retroviruses for which sequence data are known lies in the pol gene, which is the most conserved gene in retroviruses, and in the protease region linked to the pol gene. However, the extent of homology in these regions is not greater for HTLV-I than for the avian retrovirus Rous sarcoma virus (1). Envelope Proteins: The external envelope glycoprotein of LAV has been recently identified, through heavy 35 S- cysteine labeling and immunoprecipitation with patients' sera. It has an apparent molecular weight in the range of to on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (3). Recently, the intracellular precursors of gpl 10 have been identified in our laboratory by Frangois Clavel and will be described elsewhere. In brief, pulse chase experiments have shown in continuously infected cells the existence of a first glycosylated precursor of daltons, which is further processed into a second precursor of daltons and finally into the virion gpl 10 and a transmembrane protein of to daltons, possibly through proteolytic cleavage and carbohydrate modifications. In the groups of animal retroviruses, only lentiviruses (visna and caprine encephalitis viruses) and equine infectious anemia virus possess such large glycoproteins. Other Putative Proteins: A unique feature of the LAV genome is its two extra open reading frames, q and f, with coding capacities of to daltons (4). These proteins have not yet been identified, but their deduced amino acid sequence is quite different from that of px proteins of HTLV-I, HTLV-II, and bovine leukosis virus. Half of the open reading frame f is encoded by the 3' VIRAL PROTEINS AND GENOME STRUCTURE gag Proteins: Three structural gag proteins are associated with the internal core, whose apparent molecular weights as determined by polyacrylamide gel electrophoresis under denaturing conditions are, respectively, (pl8), (p25), and (pl3). The p25 protein is methionine rich and readily labeled with 35 S- methionine; pi8 and pi3 are detected by silver staining or labeling with 14 C-labeled amino acids or 35 S-cysteine. These three proteins are synthesized in infected cells under the form of a dalton precursor, which is cleaved into the three proteins by a protease encoded by the viral pol gene. An intermediate cleavage product of daltons can also be detected. The structure of the three gag proteins is also found in HTLV and the lentiviruses. However, no significant sequence homology seems to exist between HTLV core proteins and lymphadenopathy-associated virus (LAV) proteins, confirming earlier data showing no cross-reactivity between patient antibodies. The serum of horses infected with equine infectious anemia virus contains antibodies capable of immunoprecipitation with the LAV p25, suggesting that equine infectious anemia virus p25 and LAV p25 share a common epitope. This epitope is Figure 1. Electron micrograph of a budding lymphadenopathy-associated virus particle at the surface of an infected lymphocyte and of a mature virion. 690 November 1985 Annals of Internal Medicine Volume 103 Number 5

3 Figure 2. Immature virions with condensing core {left) and mature virions {right). long terminal repeat. This finding is reminiscent of the long terminal repeats of mouse mammary tumor virus, which also encode (completely) a new open reading frame. As for mouse mammary tumor virus, the primer transfer-rna is lysine transfer-rna, unlike other mammalian retroviruses that use proline transfer-rna. In conclusion, the retrovirus associated with lymphadenopathy and AIDS is the prototype if a new group of human retroviruses, different from the HTLV group. Its genome structure suggests that it represents the human equivalent of lentiviruses, a subfamily of nononcogenic, pathogenic Retroviridae (Table 2). Origin and Evolution Retrospective seroepidemiologic studies have indicated that patients with AIDS with LAV seropositivity were present in Africa as early as The extent of LAV seropositivity (5% to 10%) in the general population of African equatorial countries (Zaire and Central African Republic) is also suggestive of an earlier diffusion of the virus. Recently, we collaborated with Dr. I. Desmyter (of the Rega Institute, Lou vain) to investigate LAV seropositivity in sera collected from 220 Zairian women in 1970 for hepatitis B studies. The LAV antibodies were detected by indirect immunofluorescence and their presence confirmed by radioimmunoprecipitation. Only 1 woman was found to be positive. ( < 0.5%). This result suggests that the spread of LAV in the Zairian population, although occurring earlier than in the United States and Western Europe, happened rather recently, between 1970 and This finding is in agreement with the observed recent increase in the number of cases of aggressive Kaposi's sarcoma and meningeal crytococcosis linked to immunodeficiency observed in these regions, but is in contrast to data reported by Saxinger and coworkers (5) indicating a high degree of HTLV-III positivity in sera collected in 1970 from Uganda (east equatorial Africa). Molecular studies in progress indicate a high degree of genetic variation of this retrovirus, especially in the protein responsible for its tropism and presumably its pathogenicity. Available sequence data indicate a 20% difference between the glycoprotein of the AIDS-related virus and that of LAV1, whereas the overall sequences differ by 6%. The differences include deletion, point mutations, and insertions, and some of them may suppress or add glycosylation sites. Restriction sites analysis of several isolates, including a Zairian isolate, shows also a great diversity in restriction sites. It should be pointed out, however, that some variation may also occur during invitro propagation, especially in continuous lines in which intense episomal replication of viral DNA is taking place. Despite this genetic variation, antibodies against the glycoprotein of the prototype LAV1 grown on T lymphocytes or continuous lines have been detected in nearly all patients with AIDS, with the exception of two infants (infected at or after birth by their mother) and one adult. However, a greater genetic variation may have occurred in the past and be the cause of the high pathogenicity of the present isolates. Because the molecular basis of the virus tropism for T4+ lymphocytes lies in the glycoprotein, this tropism might be a newly acquired property of the original retrovirus. Such a variation does indeed exist in lentiviruses: Visna and maedi virus are closely related viruses with different cell tropism and pathogenicity encephalitis for visna and progressive pneumonia for maedi. This potential for genetic variation is perhaps the greatest danger in the future of the LAV epidemic. It will make it difficult to design efficient vaccines protective against all strains, and a further change of the virus in its tropism and ways of transmission cannot be excluded. Table 2. Proposed Classification of Human Retroviruses Within Retroviridae Oncovirinae (oncogenic) Avian leukemia virus (birds) Rous sarcoma virus (birds) Human T-cell leukemia virus type I (humans) Human T-cell leukemia virus type II (humans) Lentivirinae (nononcogenic, pathogenic) Visna and maedi viruses (sheep) Caprine encephalitis virus (goats) Equine infectious anemia virus (horses) Lymphadenopathy-associated virus (HTLV-III, AIDSrelated virus)* (humans) Spumavirinae (nonpathogenic) * HTLV-III = human T-lymphotropic virus type III; AIDS = acquired immunodeficiency syndrome. Montagnier Lymphadenopathy-Associated Virus 691

4 In-Vitro and In-Vivo Pathogenicity IN VITRO Soon after the first isolation of LAV, our group was able to show its selective tropism for T4+ lymphocytes purified either from an asymptomatic carrier of the virus or from a blood donor after in-vitro infection. (6, 7). Inhibition of cell multiplication was observed at the onset of virus production, together with a diminution of expression at the cell surface of the marker T4 and, to a lesser extent, T3. Giant cells, formed by fusion between infected cells, were also observed. This fusion activity was shown to be associated with the viral glycoprotein because serum of patients having only antibody against gpl 10 could prevent this process of cell fusion in a continuous cell line derived from the line MOLT4. Recent electron microscopic studies have regularly shown an alteration of mitochondria in infected lymphocytes, including disappearance of cristae and coalescence of mitochondrial membranes (Figure 2). The molecular basis of the virus tropism for T4 cells resides at least partly in the binding of the viral envelope protein to the T4 molecule itself (8, 9). In T8 lymphocytes, which lack the T4 molecule, no productive infection as assessed by the synthesis of proviral DNA can be detected (Klatzmann D. Personal communication). The expression of the T4 receptor is also an important factor for the susceptibility of permanent T-cell lines to LAV infection. The virus can also grow on some B-cell lines obtained by in-vitro transformation with Epstein-Barr virus or derived from Burkitt's lymphoma cell lines (10). The possibility that the susceptibility of these cell lines depends also on the transient expression of the T4 molecule at their surface cannot be excluded (9). Attempts to grow the virus in other types of cell lines, including a human neuroblastoma line and a colon line, have failed so far. IN vivo Primary infection with LAV generally results in an immune response and, less frequently, in clinical signs. The cellular immune response, particularly the formation of a specific cytotoxic reaction, has not been well studied. It is likely that persistent lymphadenopathy is due to a cytotoxic response (T8 cells) against infected T4 cells, which are a minority in the lymph nodes of infected patients. This response is reminiscent of the infectious mononucleosis induced by acute Epstein-Barr virus infection. In contrast, the humoral response occurs as early as 1 month after LAV infection, as has been shown in chimpanzees inoculated with the virus produced in vitro and also in human recipients of blood transfusions from LAV-seropositive donors. Studies of serial samples from patients indicate that seropositivity can persist for years, with some variation in antibody titers and the antigens recognized. All major virion proteins are antigenic, as detected by radioimmunoprecipitation or immunoblotting: by 35 S-labeling of viral proteins, antibodies against the core proteins p25 and pi8 and the viral envelope glycoprotein gpl 10 can be detected. Antibodies against gpl 10 are most constantly found in patients with AIDS or infected persons, and their titers are higher than those raised against the core proteins. Actually, only about 50% of European or North American patients with AIDS show detectable antibodies against one of either core proteins, more frequently p25 (3). However, in a study of 37 Zairian patients with AIDS, 94% had antibodies against the LAV p25. In contrast, asymptomatic white patients or patients with the AIDS-related complex more frequently have antibodies against the core proteins. The reasons for these differences are unclear. One likely interpretation is that antibodies against core proteins are produced when whole virions are expressed by infected lymphocytes and met by the immune system or when the virus-producing lymphocytes are lysed by a cytotoxic response. Perhaps the viral envelope or its cellular precursors are more immunogenic or their expression at the lymphocyte surface is sufficient to raise antibodies. An obvious question is whether such antibodies have neutralizing activity on the viral infectivity. Finding the answer will be technically difficult, because there is no accurate in-vitro infectivity assay of the virus. Through the production of LAV 1-associated reverse transcriptase by infected lymphocytes, we have found a weak neutralizing titer (1/10) in the sera of some patients (11) but not in all. In contrast, the fusion activity associated with LAV-infected MOLT4 cells can be neutralized by sera from most patients with AIDS. Our preliminary conclusion is that glycoprotein epitopes involved in binding to the T4 molecule are poorly immunogenic, perhaps because they are masked by carbohydrates. This event obviously may not be fortuitous, but instead may be an integral part of the virus mechanism to persist and spread in the T4-lymphocyte population. Clinical Signs of Acute Infection: Although the virus grows readily in chimpanzee lymphocytes and can rapidly induce seroconversion in the inoculated animal, no obvious disease has been observed, even after months or years, in the infected animals. In humans, inapparent infection is also the most frequent situation. However, several groups, including ours, have observed in some patients the signs of the AIDS-related complex and, particularly, the generalized lymphadenopathy (12) or mononucleosis-like syndrome, soon after infection. The following case is a typical example. A French man donated blood 20 times between 1979 and He died of AIDS on 1 January Although how he became infected with the virus remains uncertain, he was probably infected in 1982, because all recipients of his blood before 1982 were seronegative and all recipients after 1982 became seropositive. Of the seropositive recipients, all are still asymptomatic except two boys, both thalassemia who once received washed erythrocytes from the donor. Two months after this accidental infection, the boys developed, together with seroconversion, a generalized lymphadenopathy with slight signs of biologic immune depression (12). This syndrome eventually regressed after several months, but the restoration of the 692 November 1985 Annals of Internal Medicine Volume 103 Number 5

5 immune functions is still not complete after 1 year. Why did these two recipients, but not the others, develop this syndrome? Although the differences may just be a matter of time, with the other patients developing this syndrome at a later date, we cannot rule out the possibility that some cofactors played a part: Patients with thalassemia receive erythrocytes monthly, which may act as alloantigens. The pool of activated T4 lymphocytes in these patients may have been abnormally large, increasing the number of target cells for LAV infections. The Problem of Cofactors: For the induction of frank AIDS, we know that cofactors are not obligatory, because patients have developed AIDS after a single blood transfusion. However, cofactors may increase the risk of AIDS in groups at higher risk, such as homosexuals with multiple sexual partners, intravenous drug users, and hemophiliacs. Persons belonging to these groups often have immune dysfunctions, even in the absence of LAV infections, and also are exposed to many antigenic stimuli by repeated infections of various origins or by foreign proteins (in homosexuals and hemophiliacs). This situation may also be the case for Africans living in regions where malaria is endemic. Repeated infection with various LAV strains, like repeated infection with Antiviruses, may also be an AIDSfavoring factor in these high-risk groups. Whatever the importance of such cofactors, the irreversible character of AIDS is striking. Diminution of the viral load by antiviral drugs such as antimoniotungstate (HPA 23) does not restore the immune system (13). Even at late stages of the disease, only a minor fraction of the T4+ population is infected, as shown by in-situ hybridization with DNA probes or by immunofluorescence. It seems as if the virus infection was the trigger of an irreversible process in which host factors play the major part. This mechanism may be also the case in LAV-induced encephalitis, which seems to occur more frequently with prolonged survival in patients with AIDS. Progress in the knowledge of these irreversible mechanisms will come from a close collaboration between virologists, molecular biologists, immunologists, and clinicians. Requests for reprints should be addressed to Luc Montagnier, M.D.; Viral Oncology Unit, Institut Pasteur, 25 rue du Dr. Roux; Paris Cedex 15, France. References 1. WAIN-HOBSON S, ALIZON M, MONTAGNIER L. Relationship of AIDS to other retroviruses [Letter]. Nature. 1985;313: MONTAGNIER L, DAUGUET C, AXLER C, et al. A new type of retrovirus isolated from patients presenting with lymphadenopathy and acquired immune deficiency syndromes: structural and antigenic relatedness with equine infectious anemia virus. Ann Virol (Inst Pasteur). 1984;135E: MONTAGNIER L, KRUST B, CLAVEL F, et al. Identification and antigenicity of the major envelope glycoprotein of lymphadenopathy-associated virus. Virology (In press). 4. WAIN-HOBSON S, SONIGO P, DANOS O, COLE S, ALIZON M. Nucleotide sequence of the AIDS virus, LAV. Cell. 1985;40: SAXINGER WC, LEVINE PH, DEAN AG, et al. Evidence for exposure to HTLV-III in Uganda before Science. 1985;227: MONTAGNIER L, CHERMANN JC, BARRE-SINOUSSI F, et al. A new human T-lymphotropic retrovirus: characterization and possible role in lymphadenopathy and acquired immune deficiency syndromes. In: GAL- LO RC, ESSEX ME, GROSS L, eds. Human T-Cell Leukemia/Lymphoma Virus. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory; 1984: KLATZMANN D, BARRE-SINOUSSI F, NUGEYRE MT, et al. Selective tropism of lymphadenopathy associated virus (LAV) for helper-inducer T lymphocytes. Science. 1984;225: KLATZMANN D, CHAMPAGNE E, CHAMARET S, et al. T-lymphocyte T4 molecule behaves as the receptor for human retrovirus LAV. Nature. 1984;312: DALGLEISH AG, BEVERLY PCL, CLAPHAM PR, CRAWFORD DH, GREAVES MF, WEISS RA. The CD4 (T4) antigen is an essential component of the receptor for the AIDS retrovirus. Nature. 1984;312: MONTAGNIER L, GRUEST J, CHAMARET S, et al. Adaptation of lymphadenopathy associated virus (LAV) to replication in EBV-transformed B lymphoblastoid cell lines. Science. 1984;225: CLAVEL F, KLATZMANN D, MONTAGNIER L. Deficient LAV1 neutralising capacity of sera from patients with AIDS or related syndromes [Letter]. Lancet. 1985;1: BOITEUX F, VILMER E, GIROT R, et al. Lymphadenopathy syndrome in two thalassemic patients after LAV contamination by blood transfusion [Letter]. N Engl J Med. 1985;312: ROZENBAUM W, DORMONT D, SPIRE B, et al. Antimoniotungstate (HPA 23) treatment of three patients with AIDS and one with prodrome [Letter]. Lancet. 1985;1: Montagnier Lymphadenopathy-Associated Virus 693