AIDS Viruses of Animals and Man ow does a host usually develop nately, appear to be resistant to destruc- The Search a state of protection against an tion by complement. H invading virus? Three major Studies performed in my laboratory host responses to invading viruses in- in 1987 showed no evidence of ACC clude activation of complement, produc- activity in humans at various stages tion of neutralizing and complement- of AIDS despite the presence of large Protective fixing antibodies, and cell-mediated im- amounts of antibody directed against mune responses. Traditionally, when both HIV and HIV-infected cells. The Host a new viral disease is recognized in a absence of ACC is also documented in species, efforts to understand the pro- visna-maedi. No ACC activity has been Responses tective immune states are derived from reported for the other animal-lentivirus its surviving members. These individ- systems mentioned here. Both my lab uals serve as immune benchmarks, and and others have shown that human com- subsequent studies often reveal impor- plement is incapable of inactivating HIV tant clues to the eventual production of either directly or in the presence of neu- a vaccine. Here we will review studies tralizing antibody. Recently, we have of the major antiviral immune responses discovered a heat-sensitive serum factor to HIV and see that none of them are in various laboratory and wild animal completely effective, although some av- species that does inactivate the human enues of developing traditional vaccines AIDS virus in vitro. Further studies are for AIDS are still open. underway to characterize this factor, or factors, and to understand how to recruit Complement. One possible response its activity in humans and why human to HIV would be the activation of the complement does not work against HIV. complement system, known to be a powerful, continuous, ever-present, Neutralizing Antibody. Neutralizing microbe-eliminating system (see Fig. 11 antibodies have been shown to be one in the main article). Complement is a of the major lines of defense in viral group of serum proteins circulating in diseases of human and other animals. the bloodstream that bind to, become Following the infection of the host by activated, and destroy invading mi- the AIDS virus, plasma cells produce crobes by creating holes in their surface antibodies directed against various parts membranes. Complement proteins are of the virus. The antibodies are of two synthesized by activated macrophages, major types, functional (when they bind liver cells, and epithelial cells. Com- to the virus they inactivate or destroy plement inactivates some Type C on- it) and nonfunctional. A nonfunctional coviruses directly due to the presence antibody recognizes various parts of the of as yet undefined receptors on the vi- virus; however, they do not mediate any ral envelopes. Complement can also antiviral effects in vitro or in vivo. Also work in conjunction with antibodies. the nonfunctional antibodies can coat The antibodies produced in response the virus and thereby block or inter- to a viral infection may bind both to fere with otherwise effective antibodies, complement and to the virus or virally such as neutralizing or complement- infected cells, resulting in destruction fixing ones. An antibody that is coat- of the intruder. The destruction of vi- ing a virus can also, as previously de- rally infected cells through this mech- scribed, facilitate entry of the virus into anism is called antibody-dependent, monocytes and macrophages, thus in- complement-mediated cytolysis (ACC). fecting these cells. Some evidence for The lentiviruses as a group, unfortu- this type of antibody-facilitated infectiv- Los Alamos Science Fall 1989 85 AIDS Viruses of Animals and Man ity has been reported in the visna-maedi izes the virus. Subsequent infectivity system. studies with HIV I demonstrated that Functional neutralizing antibodies the virus can be neutralized at various are produced by plasma cells derived times, even after it has attached via the from B lymphocytes that have been CD4 receptor to the host cell (Fig. 1). specifically activated against a partic- It appears that the virus binds to sus- ular antigen. We have already indicated ceptible lymphocytes at the diffusion- -9 that the ability of the neutralizing anti- limited rate of 4.0 x 10 M (see “The (b) Binding of Antibody to HIV body to inactivate lentiviruses is highly Kinetics of Viral Infectivity”). After variable. Horses infected with the EIA binding, however, the virus only slowly virus make antibody that is capable of enters the cell by the fusion process. neutralizing the initial infecting viral Thus, neutralizing antibody is capable of strain. However, antigenic drift even- neutralizing the virus during the 30 to tually produces viruses that can avoid 60 minutes between binding and entry those neutralizing antibodies. into the lymphocyte. This is a singu- Our information on visna-maedi is larly encouraging finding for vaccine more detailed. In vitro, antibody against development. However, only the sera visna-maedi is capable of neutralizing from HIV-infected humans or HIV- the virus so that it cannot infect a cell. infected chimpanzees were capable of However, neutralization of the virus oc- neutralizing more than one HIV I strain. curs only if the virus and antibody Moreover, these strains may only be a allowed to interact for 15 minutes c subpopulation of the virus present in longer before being introduced to the any one infected individual. Studies target cells. It appears that after this of the role of neutralizing antibody in preincubation the antibody prevents the preventing infection of monocytes and virus from attaching to the sheep’s cells. macrophages will have to await the de- However, when the virus and antibody velopment of new assay methods. are added to the target cells simulta- Our studies also show that neutraliz- neously, no neutralization of the virus ing antibody derived from sera of goats Mechanism Unknown occurs. These studies suggest that the infected with the purified envelope of antibody produced during the infection one HIV strain is also capable of neu- is not biologically functional in vivo. In tralizing the virus either before or after the host the virus probably encounters it has bound to a target cell. The ma- and infects target cells before neutral- jor limiting feature however was the izing antibody has sufficient time to narrow specificity of the neutralizing an- neutralize it. The virus’s escape from tibody produced. We, in collaboration antibodies appears to be related to the with Jaap Goudsmit, Scott Putney, and high sugar content of the viral envelope others, have discovered that neutralizing proteins, which conceal neutralization antibody reacts only with the immun- OF HIV epitopes (protein shapes that serve as odominant neutralizing epitope of gp 120 antibody binding sites). shown in Fig. 2. Further, this portion of Fig. 1. In vitro studies suggest that neutral- Fortunately, the neutralizing anti- gp120, which is about 30 amino acids in izing antibody against HIV can neutralize the body present in HIV-infected humans, length, appears to be changing its amino virus even after it has bound to a target-cell HIV-infected chimpanzees, and animals acid content rapidly in infected humans membrane. The figure shows neutralizing an- that have been vaccinated with the vi- and more slowly in chimpanzees. In tibodies attaching to the viral envelope after ral envelope protein gp 120 are more particular, even the first viruses isolated the virus has bound to and begun to fuse with effective. Recent detailed kinetic stud- from chimps infected with a specific and the cell membrane. The antibodies somehow ies in my laboratory revealed that the well-characterized human AIDS virus prevent infection, but the details of the neu- serum from these hosts rapidly neutral- were resistant to a typing sera made tralization mechanism are unknown. 86 Los Alamos Science Fall 1989 AIDS Viruses of Animals and Man HIV ENVELOPE GLYCOPROTEINS CD4 Binding Site lmmunodominant Loop Amino Acid Variability Constant Occasionally Altered q Highly Variable Fig. 2. The structure of the HIV envelope glycoproteins gp120 and gp41, showing constant regions Ci, including the binding site for the CD4 receptor, and variable regions V i including the immunodominant loop, a region that binds most of the known neutralizing antibodies. Both the CD4 binding site (c3) and the immunodominant loop V 3, have very few sugar molecules compared to most other regions of gp120 and are therefore exposed to the immune system. The immunodominant loop consists of about 30 or so amino acids, many of which are highly variable from one viral strain to another. The loop mediates all known neutralization phenomena in infected humans, chimpanzees, and viral envelope-based vaccine constructs. However, there are reports of other binding sites for neutralizing antibody on gp120, which are not yet as well characterized. Studies show that the ability of the immunodominant loop to bind neutralizing antibody is altered by the change of a single amino acid in either the fusion domain of gp41 (see figure) or in the immunodominant loop. Such changes may alter the conformation of the loop (indicated by the arrows). Also gp41 may contain an as yet uncharacterized neutralization binding site. from goats immunized with gp 120 of that other sites in the viral envelope and variations that cover the spectrum the original innoculated virus as well as also must contribute to the interaction of gp120 variations made during viral sera from other chimps infected with the between the neutralizing antibody and replication. We might then be able to original virus. We are currently studying the immunodominant loop (see Fig. 2). manufacture an anti-gpl20 vaccine that the amino acid sequence of the relevant When completed, the molecular study of would be effective against all these vari- pieces of the viral envelope protein in the viral-envelopes from chimp isolates ations.