Self-defense: The fruit fly style

The immune system of higher vertebrates consists of innate and adaptive components that differ mainly by the mechanisms of pathogen recognition: the innate immune system uses germ-line encoded receptors to recognize conserved molecular constituents of infectious microorganisms, whereas adaptive immunity is mediated by highly specific antigen receptors that are generated by gene-rearrangement processes during the ontogeny of each individual organism and are distributed clonally on the two types of lymphocytes—T cells and B cells (1, 2). As antigen receptors are generated by random processes, they can potentially be specific for any antigen. One of the essential functions of the innate immune system is believed to be instructing lymphocytes about the nature of the antigen for which they happen to be specific. This choice of effector function is essential for the effective operation of the adaptive immune response, as lymphocytes generate their antigen receptors randomly, and thus can be specific for many different types of antigen. The innate immune system sorts out whether the antigen is derived from a pathogen, in which case a response is induced and used for host defense. In cases in which the antigen is produced by the host organism, an immune response would be harmful to the host and therefore is not induced. Because the receptors of innate immunity recognize molecular structures produced only by microbes and not by the host organism, they activate signaling pathways that inform the lymphocytes of the microbial origin of the antigen. When this activation happens, lymphocytes specific for the antigen in question start to proliferate and differentiate into one of the several possible types of lymphoid effector cells. This differentiation process, however, is not pre-programmed in lymphocytes. Rather, it also depends on signals generated by the innate immune system (3–5). The best known signals of this kind are represented by a subgroup of cytokines that are believed to convey information concerning the type of pathogen invading the host organism. These cytokines thus direct differentiation of the lymphocytes into effector cells, that, in turn, activate the appropriate destructive mechanism to ultimately eliminate or neutralize the microbial pathogens (6). Additionally, some of the lymphocytes derived from the same clone differentiate into long-lived cells that retain the imprint of the instructions they received on initial infection. These cells can rapidly produce the same effector responses to all subsequent infections with the same pathogen. This sophisticated mechanism provides for long-lasting immunological memory to pathogens encountered during the lifetime of the host organism.

Many major unanswered questions concerning the initiation of adaptive immune responses in vertebrates relate to the mechanisms that control the induction of the appropriate cytokines that regulate the initiation of effector functions. CD4 T cells in particular are readily directed in one or other pathway of differentiation by cytokines in their environment in the initial phases of stimulation. The biological significance of these control mechanisms is obvious, because the kind of effector response that needs to be induced depends on the type of pathogen with which the host is infected. Clues to the molecular components involved in these mechanisms may come from studies of immunity in the fruit fly Drosophila melanogaster, where application of genetic analysis allows one to directly correlate a mutant phenotype with a specific gene or genes. Drosophila, like all other invertebrates, relies entirely on innate mechanisms of host defense (7). This fact makes it a particularly attractive and powerful model system, given the remarkable conservation of the signaling pathways that initiate the immune responses in flies and vertebrates (8–10). In this context, the findings reported by Lemaitre and colleagues (18) are particularly important as they demonstrate that even the relatively simple immune system of Drosophila is able to discriminate between different types of pathogens, namely between Gram-positive bacteria, Gram-negative bacteria, and fungi, and to make the effector responses appropriate to the type of infecting pathogen. The relevance of the effector responses in Drosophila to the classes of infecting pathogens is quite straightforward, as the different sets of antimicrobial peptides induced on infection are active only against a particular group of microbes (11, 12). The obvious implication of the present study (18) is that the immune system of Drosophila is able to discriminate between different groups of microbes and that this recognition is mediated, by definition, by innate, germ-line encoded receptors.

The first indication that the Drosophila immune system can distinguish between different types of microorganisms came from the same group in their seminal study demonstrating the role of the spaetzle/Toll/cactus signaling pathway in the antifungal response in adult fruit flies (13). Flies mutant for any of the genes from spaetzle to cactus were unable to induce the expression of the antifungal peptide drosomycin and rapidly succumbed to fungal infection, whereas flies with the same mutations were resistant to bacterial infection. Recent analysis of an immune response in Drosophila larvae carrying mutation in the gene encoding 18 Wheeler, a transmembrane receptor homologous to Toll, has revealed the existence of a distinct pathway of induction of antimicrobial peptides. Interestingly, inactivation of 18 Wheeler results in a defect in antibacterial response, most strongly affecting induction of attacin, a peptide that is active against Gram-negative bacteria (14). The importance of these findings is evident when one considers the homologies between the components involved in immune responses in Drosophila and vertebrates: antibacterial peptides in Drosophila are controlled be the NFκB family of transcription factors (12). Many genes involved in the innate and adaptive immune responses of vertebrates, including the cytokines that control the choice of effector response, likewise are controlled by NFκB (15, 16). In both Drosophila and mammalian species, NFκB and related transcription factors can be activated by Toll receptors, and the similarity of Drosophila and human Toll proteins suggests that the putative and as yet unidentified ligand of human Toll is homologous to the Drosophila Toll ligand spaetzle (17). Most relevant for the present discussion, the molecular components functioning upstream of the Toll ligands are likely to be homologous as well. In Drosophila, the active form of spaetzle is generated by a proteolytic cascade triggered by infection (13). Therefore, the most upstream element in this cascade must be able to detect the presence of the pathogen, most likely by recognizing conserved microbial molecular products, such as lipopolysaccharide in Gram-negative bacteria, teichoic acids in Gram-positive bacteria, or mannans and glucans in fungi. One possibility strongly suggested by the study of Lemaitre et al. (18) is the existence of multiple receptors with specificities for the molecular structures characteristic of different classes of pathogens. These receptors are proposed to act upstream of the protease cascades that generate distinct ligands for different Toll-like receptors. Identification of these receptors in Drosophila and in vertebrates like mouse and humans would allow us to understand in molecular terms the mechanism of induction of effector functions appropriate to the type of infecting pathogens, which could, in turn, be applied to vaccine design.