protein that binds to T-cell epitopes. This means that only the T lymphocytes of some individuals (those with suitable MHC proteins) can recognize a given T-cell epitope. Genetic restriction is most easily studied in inbred mouse strains that have MHC proteins that are identical for all individuals of that strain but different from the MHC proteins of other strains. Most T-cell epitopes on malaria antigens appear to be genetically restricted (Del Giudice et al., 1986; Good et al., 1986, 1987, 1988a,b,c; Good, 1988; Hoffman et al., 1989a). That is, only mice of a few strains can recognize these epitopes. The human cell-mediated immune response to malaria antigens also appears to be genetically restricted (Good, 1988; Good et al., 1988d; Hoffman et al., 1989c; de Groot et al., 1989).
If only a minority of individuals can mount effective immune responses against protective T-cell epitopes after immunization, a vaccine would have to include multiple T epitopes, perhaps from different target antigens, to increase the probability that at least one T-cell epitope was recognized by almost every individual within a population. Indeed, the apparently universal ability of live attenuated sporozoite vaccines to induce protective immunity may result in part from the presence of multiple T-cell epitopes on multiple different target antigens. Although knowledge in this area is limited, it is encouraging that three of four volunteers immunized with irradiated sporozoites produced cytotoxic T lymphocytes against a defined T-cell epitope in the P. falciparum CS protein (Malik et al., 1991). The importance of genetic restriction to malaria antigens in humans remains to be determined.
Assays Predictive of Protection One of the major obstacles to malaria vaccine development is the lack of laboratory assays that correlate with protective immunity (Hoffman et al., 1987). The only certain way to test vaccine efficacy is to immunize volunteers and determine whether they are protected after exposure to malaria-infected mosquitoes. Development of laboratory assays that predict protective immunity would be a major advancement.
Length of Protection Nonimmune visitors to endemic areas may need a vaccine that protects them for only several weeks or months. In contrast, long-term residents of endemic regions require a vaccine that provides protection for years.
Immunization with live attenuated sporozoites induces long-lasting protective immunity in mice, but in the few humans tested protection disappears after a few months (Clyde et al., 1973b, 1975; Rieckmann et al., 1979). The early experience with malaria subunit vaccines based on the P. falciparum CS protein and tested in humans indicates that antibody levels generally decline quickly. Approaches that may overcome this problem