Received 6/18/97, Accepted 7/22/97

Viral vaccines work by imitating natural immune responses to pathogens and subsequently clearing the infectious agents from the host's system. However, retroviruses possesses a much more complex life cycle than any other infectious agent for which we have previously developed vaccines. We must come to better understand the nature of any unique immune mechanism that may exist against retroviruses if we wish to develop an effective vaccine.

First of all, on evolutionary viewpoint, it is difficult to believe that during the course of evolution, higher organisms did not develop some sort of defense mechanism in order to protect themselves from the onslaught of retroviral infections. For example, regions between maize plant genes are packed with retroelements which make up more than 50% of the two billion base pairs that constitute this plant's nuclear DNA (71). Obviously, a mammalian cell can not possibly accommodate so many retroviral genes and must have developed intracellular defenses to counteract insertion of such genes. Retroviruses have the ability to acquire and alter the structure of host-derived sequences, leading to altered genes, pseudogenes, and oncogenes in the host species; they have the ability to insert their own genome into the host's germline, potentially making subsequent generations "transgenic hosts" for this now endogenous virus (72-73); and finally, the seemingly-random insertion of provirus can cause genetic damage to the host, leading to disruptions in the activation or control of specific genes near the site of proviral integration. Therefore, we have hypothesized that higher eukaryotes must possess some sort of intracellular immunity (which we have named molecular immunity) specifically evolved to combat such clear and present genetic dangers. In fact, already there are substantial published data to indicate such immunologic responses exist (74-77).

Even though a majority of individuals exposed to HIV-1 become infected, rare individuals remain uninfected with the virus, despite histories of multiple high-risk sexual exposure to HIV-1 (1, 13-16). In some cases this may simply be the result of defective viruses resulting in abortive or quiescent infection. In other cases, there appears to be a clear evidence of resistance to infection. For example, it has been shown that the CD4+ T-cells of some individuals resist infection with high doses of virus (about 1000-fold higher concentration of virus than what is require to establish infection). While, a small fraction of cells become infected with such a high viral dose, the viral-replication does not take place (13, 78-79). Recently, it has been shown that certain individuals (about 1%) have homozygous defect in one of the HIV-1 coreceptors, CCR5, which makes them resistant to monocyte-tropic strains of HIV-1 (8-9). However, this observation does not explain the fact that why so many of health care workers who got exposed to low doses of HIV-1 did not become infected with the virus? For example, there are over 2084 health care workers in the U.S. who were accidentally exposed to HIV-1 and were monitored by the CDC, (the actual figure may be 10-20 times higher, since most individuals who are accidentally exposed to bodily fluids of HIV-1-seropositive individuals do not inform the CDC). Yet only 4 individuals who have no other source of exposure did became seropositive (18-20) and personal communication: Dr. Denise Cadre, CDC). Considering that many of these individuals had deep percutaneous exposures resulting in visible bleeding from the sites of needle injuries. Recent, more extensive studies indicate that the estimated risk for HIV-1 infection after percutaneous exposure to HIV-1 infected blood is about 0.3% (10-13). This is a very small percentage, considering these sorts of exposure generally involve needle pricks with infected blood. Also, this does not quite fit into the beta-chemokine receptor hypothesis described by Cocchi and others (5-9).

Epidemiological studies indicate that various classifiable subgroups infected with HIV-1 vary considerably in their median incubation period and their susceptibility to HIV-1 infection (80). For example, the frequency of successful transmission of HIV-1 resulting from a single intercourse with an infected partner is relatively low (0.2 to 1%), even though HIV-1 is present in 80-100% of human semen specimens from HIV-1-infected individuals. Some individuals also lack any evidence for infection with HIV-1, despite multiple sexual contacts with HIV-1-infected partners (81-82). Other epidemiological studies suggest that some individuals are truly resistant to HIV-1-infection (13, 78-79). It has been shown that the rate of infection in individuals exposed to a whole unit of infected blood is about 30% (83)—a very high percentage and one that is associated with high-dose exposure.

Furthermore, several investigators have reported isolation of HIV-1 from individuals who remained HIV-1-seronegative and free of disease. Detels et al (13) have observed that some men with many different partners with whom they practiced receptive anal intercourse have remained seronegative, despite repeated exposure. Surprisingly, Bryson et al (84) have documented the clearance of HIV-1 infection in perinatally-infected infants, who subsequently remained without a detectable HIV-1-infection for five years. More recently, Roques et al have also documented 12 cases of perinatally-infected children who cleared the virus (85). Besides these well-documented cases, several other investigators have also reported evidence of individuals who seroreverted, and the prior infection of these individuals with HIV-1 were documented by HIV-positive blood cultures, positive serum HIV-1 p24 antigenemia, and in some cases, positive PCR assays (86).

But perhaps most intriguing of all are the reports of the so called molecular immunity in humans infected with HIV-1 who are long-term nonprogressors (LTNP). The majority of individuals infected with HIV-1 progress to AIDS. The average time from first infection with HIV-1 to death in the progressors is less than 10 years. However, the clinical manifestations of HIV-1-associated illnesses appear much earlier, 4 to 6 years after the infection. In about 5% of individuals infected with HIV-1, the so called "long-term nonprogressors", the natural history of HIV-1-infection is altered. These individuals remain healthy and many of the clinical manifestations of HIV-1-infection are either absent or not as prominent as in the progressors (i.e., low CD4+ cell count, HIV-1 p24 antigenemia, generalized lymphadenopathy and other AIDS-associated infections, lymphomas and Kaposi's sarcoma). In recent years, several reports have emerged, some of which indicate that an attenuated Nef-defective HIV-1 variant may be one of the causes of the LTNP status of individuals, from which these defective HIV-1 variants were isolated (87). Deacon, et al. (83) have described a single index case, an HIV-1-infected blood donor, whose blood or blood products were transfused into six individuals. All of these recipients have remained free of HIV-1-related diseases after 10 to 14 years. The analysis of HIV-1 isolates have shown a Nef-defective gene. In addition, Kestler, et al. (87) have reported that, rhesus monkeys, experimentally infected with nef-negative simian immunodeficiency virus (SIVmac) manifest no signs of immunosuppression. The actual contribution of nef-defective HIV-1 in the LTNP of this retroviral infection is still controversial and several LTNP do not show evidence of nef-defective HIV-1 in their PBMCs (but potential defects at other genetic loci have not been done yet) (88). In addition, the functional analyses of the nef-defective HIV-1 viruses, isolated from ten LTNPs, indicated no significant differences in the replication properties of these isolates (88). Many factors, including multiple HIV-1 variants with different degrees of virulence and replication capabilities, numerous host factors, and environmental influences may play important roles in the ultimate outcome of infection with HIV-1 (89-90). It appears that there is race between the rapid replication of the virus and development of anti-retroviral immunity. If an individual is allowed to develop the proper "molecular immunity" against the virus, then the odds may move in favor of the host. We hypothesize that infection with these relatively-attenuated viruses or low doses of HIV-1 (like in health care workers: 10-12), results in molecular immunity to respective viruses in the infected hosts (38-43, 91-95). A particular point that bolsters this interpretation is the fact that many of the individuals infected with nef-defective HIV-1 almost certainly later came in contact with fully pathogenic strains of HIV-1 but remained nonprogressors due to the induction of molecular immunity by the attenuated strains. For example, one of the documented nonprogressors was a hemophiliac and probably had multiple exposures to virulent strains of HIV-1 (through frequent injection of unscreened Factor VIII). Similarly, several other nonprogressors were homosexual men and also had likely been exposed to various quasispecies of HIV-1, including the fully virulent strains of HIV-1(87-88, 93). Similarly, experimental infection of monkeys with nef-deleted SIVmac has resulted in protection against subsequent infection with the high-dose, full-length, wild-type, virulent strain of the homologous virus (93).