BIOLOGICAL OVERVIEW

Insects resist bacterial infections through the induction of both cellular and humoral immune responses. The cellular response involves the mobilization of hemocytes, whereas the humoralbresponse utilizes antibacterial peptides that are synthesized in the fat bodies and secreted into the circulating hemolymph. To better understand the fly's immune responses, some familiarity with the proteins involved in these responses is useful. In Drosophila, these proteins are coded for by two functionally distinct classes of inducible antimicrobial immunity genes: those that code for antibacterial peptides, namely Cecropins (Kylsten, 1990 and Tryselius, 1992), Diptericin (Wicker, 1990), Drosocin (Bulet, 1993), Attacin (Asling, 1995), and insect Defensin (Dimarcq, 1994), and a gene that codes for an antifungal peptide, Drosomycin (Fehlbaum, 1994)

Recent studies suggest that the induction of the humoral response involves two regulatory proteins, Dif and Dorsal, that are related to mammalian NF-kappa B. Both Dorsal and Dif are expressed in immuno-responsive tissues, and both proteins are translocated into the nuclei after bacterial challenge. These regulatory proteins function as sequence-specific transcription factors that induce the expression of immunity genes (Ip, 1994). It is clear that the cell membrane and intracellular components of the Dorsal group (Toll, Cactus and Dorsal), as well as Dif, control the inducible immune response.

In addition to the dorsoventral pathway, several other pathways are implicated in the immune response. A recessive mutation, immune deficiency (imd), has been described that impairs the inducibility of all genes encoding antibacterial peptides during the immune respone. The antifungal peptide Drosomycin remains fully inducible in imd mutants, pointing to the existence of different pathways leading to expression of the antifungal and antibacterial peptides genes. The imd gene has not yet been cloned (Lemaitre, 1995a).

Another pathway, not yet determined to be distinct from that in which imd participates, involves Jun amino terminal kinase (JNK), coded for by basket. basket is activated by endotoxic lipopolysaccharide (LPS). LPS is a component of bacterial cell walls, and is known to be a stimulant for the immune response in both insects and mammals. Addition of LPS to cultured cell lines causes marked induction of cecropin and diptercin genes. basket/JNK is activated within 5 minutes of LPS addition. The activation of DJun by DJNK in LPS-treated cells may lead to increased AP-1 (a heterodimer of Fos-related antigen and Jun related antigen) transcriptional activity. Targets of Drosophila AP-1 may include the JRA promoter (Sluss, 1996).

Along with the sequences in imd and basket, there is another sequence, in the diptericin promoter that regulates its activity and is homologous to mammalian interferon consensus response element. All these sequences bind a polypeptide that cross-reacts with an antiserum directed against mammalian interferon regulatory factor-I, known to bind to the promoter of interferon-inducible genes (Georgel, 1995).

What is known about the role of Dif in the Drosophia immune response, and how may the relative importance of Dorsal and Dif to one another best be characterized? Dorsal and Dif both rapidly translocate into the nucleus after bacterial challenge. But nuclear localization requires only tissue injury; the simple act of tearing tissue is enough to induce nuclear localization. Nuclear localization is regulated by the transmembrane receptor Toll, which initiates an intracellular regulatory cascade that leads to the dissociation of Cactus (the I-kappaB homolog) from the Dorsal protein. Toll homologs include the mammalian Interleukin 1 gene, but also IL-1 related genes in invertebrates and plants. It is clear that the role of IL-1 in the immune response is conserved in some fashion throughout metazoa. Although interaction of Dif with Cactus has not been studied, it is presumed that like Dorsal, Dif too is regulated by an interaction with Cactus.

Dif binds to promoter motifs whose sequences are evolutionarily conserved. These sequences, called kappaB motifs, are similar to those targeted by NFkappaB, the vertebrate homolog of Dorsal and Dif. The promoters of immunity genes such as cecropin and diptericin have been analyzed for their interaction with both Dorsal and Dif. Dif has a substantially higher binding affiinity for the CecA1 kappaB sequence than does Dorsal. In addition, induction of the immune response gives rise to a nuclear binding activity that recognizes the CeCA1 kappaB motif. This complex is specifically disrupted by pretreating nuclear extracts with anti-Dif antibodies. These results provide strong evidence that the induction of the immune response activates Dif (Ip, 1993).

However, Dif appears to be less efficient as a transactivator than Dorsal when tested using a diptericin reporter gene. The protein complexes produced on kappaB reporter sequences by Dorsal and Dif appear to be different, suggesting that the two proteins do not have identical partners in the activation of promoter sequences by protein complexes. Dif and Dorsal interact differently with different constructs of kappaB sites, suggesting that their DNA binding characteristics are not identical. Mutants containing no copies of dorsal and a single copy of Dif retain their full capacities to express the diptericin and cecropin genes in response to immune challenge. Thus it is clear that Dorsal and Dif are likely to carry out distinct functions, and as yet uncharacterized differences, in target gene activation (Gross, 1996).

It is commonly thought the immune response in Drosophila is related to the acute-phase response in vertebrates. In vertebrates inflammation or injury is accompanied by significant alterations in the serum levels of several plasma proteins, known as acute-phase proteins. Many of the acute-phase proteins act as antiproteinases, opsonins, or blood-clotting and wound-healing factors, which may protect against the generalized tissue destruction associated with inflammation (Baumann, 1994). This analogy only goes so far. Regulation of many of the important accessory proteins of the vertebrate immune response are regulated by the same gene systems involved in the immune response in Drosophila. Making the reasonable assumption that the vertebrate immune response evolved from a non-immunoglobulin based system similar to that in Drosophila, it can be argued that the immune response in Drosophila represents the evolutionary core of the immune response in invertebrates, lacking only an inducible immunoglobulin system. This argument is developed more fully in the rolled gene site.

The Rel protein DIF mediates the antifungal but not the antibacterial host defense in Drosophila

Two Drosophila lines were isolated that carry point mutations in the gene coding for the NF-kappaB-like factor Dif. Like mutants of the Toll pathway, Dif mutant flies are susceptible to fungal but not to bacterial infections. Genetic epistasis experiments demonstrate that Dif mediates the Toll-dependent control of the inducibility of the antifungal peptide gene Drosomycin. Strikingly, Dif alone is required for the antifungal response in adults, but is redundant in larvae with Dorsal, another Rel family member. In Drosophila, Dif appears to be dedicated to the antifungal defense elicited by fungi and gram-positive bacteria. The possibility is discussed that NF-kappaB1/p50 might be required more specifically in the innate immune response against gram-positive bacteria in mammals (Rutschmann, 2000).

Transcription factors of the Rel family have long been suspected to play an important role in the control of the expression of insect antimicrobial peptides. To date, three members of this family have been reported in Drosophila, namely, Dorsal, DIF, and Relish. In this study, two mutations in the gene encoding Dif were generated and show that this Rel protein plays a critical role in the control of the antifungal response in Drosophila (Rutschmann, 2000).

The Dif¹ mutation is a strong mutation, since no significant Drosomycin induction is observed in flies subjected to a natural infection by the fungus B. bassiana in contrast to wild-type flies, where this infection triggers a strong and sustained expression of the antifungal gene. Similar results have been obtained upon injury with either gram-positive bacteria or fungi. Furthermore, in response to a fungal or gram-positive bacterial infection, Dif¹ homozygous and hemizygous flies transcribe the Drosomycin gene at the same low, residual level and show similar survival curves to fungal infections. These data indicate that the Dif¹ mutation is genetically close to a null mutation. It is likely that the Dif² mutation is also a strong mutation, since the induction of Drosomycin by immune challenge and the survival to fungal infections were similar in Dif¹ and Dif² homozygous, hemizygous, or transheterozygous Dif¹/Dif² flies (Rutschmann, 2000).

The two mutations isolated in this study were induced in the Rel homology domain (RHD). Rel proteins bind as dimers to DNA with an unusually high affinity through the conserved RHD. The structure of this domain has been determined by X-ray crystallography in several NF-κB members over the last years. Its main features have been remarkably conserved throughout evolution. The dimer wraps around the DNA, giving the appearance of a butterfly to the complex. Each monomer consists of two immunoglobulin (Ig)-like domains connected by a short linker. The N-terminal Ig-like domain consists of a nine-stranded β barrel and contains a recognition loop (L1) that makes specific contacts with the DNA binding site. This domain also contacts the binding site through a second loop (L2) that clamps the DNA at the central minor groove via basic residues. The C-terminal Ig-like domain consists of a seven-stranded β barrel, which contacts the DNA backbone through two loops (L4 and L5) and mediates homo- or heterodimerization of NF-κB subunits. Together with the nuclear localization signal located right at the C-terminal end of the RHD, the C-terminal Ig-like domain also binds to the six ankyrin repeats of the I-κB inhibitor. Finally, a basic amino acid located in the linker (loop L3) that joins the N-terminal to the C-terminal Ig-like domains makes a specific contact with a nucleotide of the DNA binding site at position ± 2 (the dyad axis of symmetry passes usually through the nucleotide in position 0) (Rutschmann, 2000).

The positions of the two mutations in the Dif structure were visualized by modeling a putative Dif based on the crystallographic data from the mosquito Rel protein Gambif, which is 37% homologous to Dif over the RHD. Glycine 181 is located in the middle of β strand E′ on the inside of the N-terminal domain, close to the DNA clamping loop L2. This glycine has been conserved in all RHDs known to date and presumably plays an important structural role. Its replacement by a bulky, negatively charged amino acid, in close vicinity to the DNA helix, certainly perturbs the local structure and prevents loop L2 from binding to DNA. Thus, it is anticipated that the Dif¹ mutation severely affects the high-affinity DNA binding of Dif. The C-terminal dimerization domain of the RHD that contacts I-κBs is not affected by the mutation and nuclear import of the protein is not altered, as illustrated by nuclear localization of Dif¹ following an immune challenge (Rutschmann, 2000).

The Dif² mutation replaces a serine by a phenylalanine residue. This serine is unlikely to be the target of a kinase, since its alcoholic function is not accessible to solvents. The same holds true for serines at the same position in the other Rel structures that have been determined. It is buried on the inside of the protein where the oxygen of the alcoholic function makes a hydrogen bond with the amide function of leucine 225 in the modeled Dif structure. The minimal hypothesis is favored that the Dif² mutation induces a strong local perturbation of the structure of the N-terminal Ig-like domain. Indeed, serine 245 is part of β strand G that connects β strand F, in particular through its alcoholic function. It is likely that this contact stabilizes the orientation of loop L3, which is highly structured when bound to DNA. It usually makes a specific contact with a central DNA base, probably through lysine 251 in Dif. As for Dif¹, it is expected that the Dif² mutation affects DNA binding, although the overall stability of the protein might also be affected. It is therefore likely that both mutations prevent the recognition of κB binding sites by Dif and that they do not impair protein–protein interactions with putative cofactors (Rutschmann, 2000).

Two lines of evidence indicate that Dif plays a pivotal role in the antifungal response in adult Drosophila. First, Dif flies are more susceptible to fungal infections than wild-type flies. Second, immune induction of Drosomycin, the predominant antifungal peptide gene of Drosophila, is severely reduced, if not abolished in those flies. Furthermore, Defensin induction is significantly decreased in Dif mutants. As regards Cecropin, Dif and dl appear to participate, at least partially, in the regulation of its immune-induced expression. Defensin and Cecropin display some antifungal properties in addition to their antibacterial activities. Even though the decreased expression of antifungal peptides certainly contributes to susceptibility to fungal infection, the possibility that Dif also activates other antifungal defense mechanisms and, namely, cellular responses, cannot be excluded (Rutschmann, 2000).

The Dif phenotype reported in this study is strikingly similar to that of mutants that affect the spz/Tl/tub/pll/cac gene cassette (Lemaitre, 1996). The epistatic relationship between Tl and Dif on one hand and cact and Dif on the other hand provides a clear demonstration that the control of Drosomycin inducibility is mediated through Dif in response to activation of the Tl pathway. Since fungal infections seem to have similar lethal effects on Dif and spz mutants, it is proposed that Dif controls the various aspects of the Tl-mediated antifungal response in adult Drosophila (Rutschmann, 2000).

As regards the induction of Cecropin, Dif and dorsal appear to be functionally redundant in adults. Indeed, immune induction of Cecropin is unaffected in Dif mutants and is significantly reduced in Dif/TW119 hemizygous flies. Since the Dif mutations are strong and are likely to affect DNA binding of Dif rather than putative interactions with cofactors, this effect cannot be ascribed to a hypomorphic effect of the Dif mutations on the Cecropin promoter. Rather, it is likely that the removal of a copy of dl in TW119 combined with the total lack of Dif is responsible for this phenotype in the Dif hemizygous background. It thus appears that the single dose of dl left in this genetic context is not sufficient to mediate the contribution of the Tl pathway to the regulation of Cecropin, and that this reduced but significant expression is regulated via the imd pathway. This inference in the case of Cecropin is further substantiated by the following observations for Drosomycin. In larvae, have shown that Dif and dl are functionally redundant in controlling the inducibility of Drosomycin (Manfruelli, 1999). This inducibility is not markedly affected in a Dif background but is significantly reduced in Dif¹ and Dif² hemizygous larvae where one wild-type copy of dl is left (Rutschmann, 2000).

In contrast to that of Defensin or Cecropin, the immune induction of Attacin does not depend on either Dif or dl, whereas it is strongly decreased in spz, Tl, and pll mutants. These results suggest that a branch point in the Tl pathway exists downstream of pll to regulate an undefined transcription factor, possibly Relish, that controls the inducibility of Attacin in adults (Rutschmann, 2000).

The Dif mutations generated in this study do not fully abolish Drosomycin induction by immune challenge with a mix of gram-positive and gram-negative bacteria. Indeed, some 25% of wild-type levels of induction were consistently observed in these conditions. In contrast, fungi or gram-positive bacteria failed to induce any induction of Drosomycin in Dif¹ mutants. The residual Drosomycin expression is abolished in Dif-kenny double mutants challenged with the mix (kenny is a mutant that shows a phenotype similar to that of Relish). It is proposed that the low level of Drosomycin inducibility triggered by gram-negative bacteria is partially controlled by a Dif-independent pathway, namely, the imd-kenny-Relish pathway. This hypothesis is supported by the result that an infection with E. coli triggers a short-lived induction of Drosomycin, similar to that of fast reactants such as Cecropin and Attacin, that are predominantly controlled by imd-kenny-Relish. Similar observations have been reported in the case of the other genes of the Toll-signaling cassette, where some 25% of wild-type Drosomycin induction was observed after challenge with the mix, whereas no induction was detected when Tl mutants were coated with fungal spores. Furthermore, the low level of Drosomycin expression observed in Tl pathway mutants was totally abolished in Tl-imd double mutants. In conclusion, it is proposed that natural infections with B. bassiana spores trigger Drosomycin expression exclusively by the Tl-Dif-dependent pathway, whereas another pathway, most likely the Relish-kenny-imd pathway, induces, at least partially, Drosomycin expression in response to gram-negative bacteria (Rutschmann, 2000).

It has been shown that Relish is required for the induction of most antimicrobial peptides in response to a gram-negative immune challenge. Specifically as regards Drosomycin, only 20% of wild-type levels were induced in Relish null mutants 6 hr after a challenge with the gram-negative Enterobacter cloacae. The data presented above are compatible with the hypothesis that expression of the Drosomycin gene is controlled by a Relish-Dif heterodimer, which is also in keeping with tissue culture experiments performed with transfected S2 cells. However, the possibility that Relish mediates the gram-negative-induced, Dif-independent component of Drosomycin induction cannot be excluded (Rutschmann, 2000).

It has been demonstrated (Manfruelli, 1999) by a clonal analysis of the TW119 deficiency that Dif and dl are functionally redundant in the control of Drosomycin inducibility in larvae. In contrast, another study (Meng, 1999) has observed that Dif but not dl regulates the expression of Drosomycin in adults. This latter result was obtained by complementation with either a Dif or a dorsal transgene of the small J4 deficiency that removes both genes. The analysis of the Dif point mutants presented in this study resolves this paradox and shows that the regulation of Drosomycin expression in the adult fat body differs from that in the larval fat body. The reasons for this difference are at present unclear. It has to be keep in mind that larval and fat body cells are not fully equivalent and have distinct developmental origins. One explanation for the difference in regulation could be that larval and adult fat body cells produce different levels of either Dif or DL. Another explanation could relate to the presence or absence in larvae versus adults of various cofactors required for full activation of response genes by DL or Dif (Rutschmann, 2000).

Five distinct Rel family members are present in mammals where they play a major role in the host defense by controlling the expression of such diverse immune-response molecules as immunoreceptors, cytokines, adhesion molecules, and acute phase proteins. In addition, they can provide protection against TNFα-mediated apoptosis. The analysis of the respective roles of the five Rel proteins is hampered by partial redundancies and by the complexities or lethality of the mutant phenotypes. In particular, the identity of the Rel proteins that mediate the TLR2-dependent response to peptidoglycans and the TLR4-dependent response to LPS has not been established, since in both cases only the activation of a 'generic' NF-κB binding activity was investigated. Yet, distinct responses are likely to be elicited by different pathogens. In this respect, it is striking that p50 knockout mice are highly susceptible to infections by the pathogenic gram-positive Streptococcus pneumoniae and not by the gram-negative Haemophilus influenzae or E. coli K1, which raises the possibility that p50 is the Rel protein mediating the response to gram-positive bacteria (Rutschmann, 2000).

In Drosophila, mutants for all three individual Rel proteins are now available. Remarkably, the viability of these mutants is not impaired under normal conditions. Dif plays a critical role in mediating the antifungal response that is activated through the Tl pathway in the adult. Strikingly, Dif does not seem to play a role in the humoral immune response against gram-negative bacteria, as can be judged from the lack of effect of Dif mutations on the inducibility of Drosocin, Diptericin, and Attacin, all of which are essentially active on such bacteria. Studies underway should reveal whether the three Rel proteins are indeed complementary in mediating a humoral immune response against diverse pathogens or whether there is some degree of functional redundancy like that observed between Dif and dl in larvae (Rutschmann, 2000).

Calcineurin isoforms are involved in Drosophila Toll immune signaling

Because excessive or inadequate responses can be detrimental, immune responses to infection require appropriate regulation. Networks of signaling pathways establish versatility of immune responses. Drosophila melanogaster is a powerful model organism for dissecting conserved innate immune responses to infection. For example, the Toll pathway, which promotes activation of NF-kappaB transcription factors Dorsal/Dorsal-related immune factor (Dif), was first identified in Drosophila. Together with the IMD pathway, acting upstream of NF-kappaB transcription factor calcineurin A1, acts on Relish during infection. However, it is not known whether there is a role for calcineurin in Dorsal/Dif immune signaling. This article demonstrates involvement of specific calcineurin isoforms, protein phosphatase at 14D (Pp2B-14D)/calcineurin A at 14F (CanA-14F), in Toll-mediated immune signaling. These isoforms do not affect IMD signaling. In cell culture, pharmacological inhibition of calcineurin or RNA interference against homologous calcineurin isoforms Pp2B-14D/CanA-14F, but not against isoform calcineurin A1, decreased Toll-dependent Dorsal/Dif activity. A Pp2B-14D gain-of-function transgene promoted Dorsal nuclear translocation and Dorsal/Dif activity. In vivo, Pp2B-14D/CanA-14F RNA interference attenuated the Dorsal/Dif-dependent response to infection without affecting the Relish-dependent response. Altogether, these data identify a novel input, calcineurin, in Toll immune signaling and demonstrate involvement of specific calcineurin isoforms in Drosophila NF-kappaB signaling (Li, 2014).

GENE STRUCTURE

Dif maps to the 36C region on chromosome 2, which corresponds to the map position of dorsal. The two genes are separated by at least 15 kb; it is probable that they map to between 15 and 88 kb of one another. A 1.8 kb RNA species related in sequence to Dif is specified by a convergently transcribed gene that is tightly linked to Dif; it is related to Dif solely on the basis of a 250 base pair region in the Dif untranslated trailer sequence (Ip, 1993).

Bases in 5' UTR - 562

Bases in 3' UTR - 277

PROTEIN STRUCTURE

Amino Acids - 667

Structural Domains

The Rel domain contains 295 amino acid residues, spanning residues 78 to 372. This Rel sequence is most closely related to the Dorsal sequence; they share 142 identical residues. Among the vertebrate Rel proteins, Dif is most similar to the turkey c-Rel (40% identity in the Rel domain). The Dorsal Rel domain is about equally related to Dif, mouse p65 and Xenopus Rel. The Rel domains of Dif, Dorsal, turkey c-Rel and mouse p65 share 86 invariant residues. Sequences that reside in the Dif Rel domain do not share significant identities with the unique regions of Dorsal or other Rel-containing proteins. The C-terminal region of Dif (amino acid residues 582 to 631) is rich in glutamine, proline, and hydrophobic residues, suggesting that it might function as a transcriptional activation domain (Ip, 1993).

date revised: 1 January 2024