Nuclear Factor-Kappa B Pathways in Drosophila

S Minakhina and R Steward

Waksman Institute, Department of Molecular Biology and Biochemistry and the Cancer Institute of New Jersey, Rutgers University, Piscataway, NJ, USA

The nuclear factor kappa B (NF-jB) pathways in in a given pathway, developmental process or in disease Drosophila are multi-component pathways, as in verte- (Zhong and Sternberg, 2006). brates, that regulate the expression of many genes There are three nuclear factor kappa B (NF-kB) responsible for the formation of dorsal–ventral polarity homologs in Drosophila: Dorsal, Dif and Relish. All in the early embryo, the innate immune response to three proteins contain the well-conserved N-terminal infection with Gram- negative and positive bacteria and Rel homology (RH) domain. Based on their protein fungi, the cellular immune response and hematopoiesis. structure rather then their sequence similarity, the Overactivation of the fly pathway can result in develop- NF-kB proteins have been subdivided into four sub- mental defects, overproliferation of hemocytes and the families (Huguet et al., 1997). Dorsal and Dif belong to formation of melanotic tumors or nodules. The extra- subfamily I, along with c-Rel, RelA and RelB. The RH cellular events leading to the maturation of the ligand for domain of all these proteins are about 40% identical. initiation of the Drosophila NF-jB pathway is not The Dif and dorsal genes lie within 10 kbp of each other conserved between flies and vertebrates, but the Toll on the second chromosome and likely arose owing to a receptor and downstream events are remarkably similar. gene duplication (Meng et al., 1999). It is interesting to NF-jB proteins have been identified in mollusks, and note that only the Drosophilidae, among insects studied arthropods such as horseshoe crabs and beetles, indicating so far, have two family I proteins and that the RH that this pathway has been established more than 500 domains of Dorsal and other insect family I proteins are million years ago. The fly NF-jB pathways are less more highly conserved (B65% identity) than Dorsal complex than those in vertebrates, with the involvement of and Dif are to one another (45% identical; Bian et al., fewer proteins, but they are, nonetheless, just as important 2005; Tanaka et al., 2005). as their vertebrate counterparts for the life of the fly. The Relish gene lies on chromosome 3 and encodes a Oncogene (2006) 25, 6749–6757. doi:10.1038/sj.onc.1209940 NF-kB protein that contains a C-terminal inhibitory ankyrin repeat domain. It is the only member of Keywords: NF-kappaB; Dorsal; Dif; Relish; immunity; subfamily III and most similar to NF-kB2 or p100, a embryonic polarity member of subfamily II. Drosophila does not have a homolog of the vertebrate subfamily IV, the NF-AT factors. The subfamily II and III proteins are activated by proteolysis that removes the ankyrin repeats, which are also present in IkB proteins. In contrast to Introduction vertebrates, Drosophila contains a single, independent IkB protein, called Cactus, which is most closely related In 1987, two papers reported the conservation in flies to IkBa (Huguet et al., 1997). and vertebrates of oncogenes – genes with a normal The fly NF-kB/IkB proteins function in the formation function in development that, when mutated or mis- of dorsal–ventral polarity, in cellular and humoral expressed, can cause cancer. One of these genes is called innate immunity, in defense against viral and parasitic int in vertebrates and wingless in flies (Rijsewijk et al., infections, in hematopoiesis and in the formation of 1987), and the other c-rel in vertebrates and dorsal in melanotic tumors/nodules. This review summarizes flies (Steward, 1987). In the present day era of genomics, what is currently known about the regulation and the the conservation of genes and entire signaling and function of NF-kB family proteins in Drosophila metabolic pathways is taken for granted and used melanogaster. We discuss new insights into the function computationally to identify additional genes functioning of the NF-kB pathways and NF-kB proteins in several distinct processes. We place special emphasis on the intracellular components of the systems (the actual NF-kB pathways), even though much has also been learned in Correspondence: Dr R Steward, Waksman Institute, Department of the last 7 years about the maturation of the Toll ligand Molecular Biology and Biochemistry and the Cancer Institute of New Jersey, Rutgers University, 190 Frelinghuysen Rd, Piscataway, NJ Spa¨ tzle during oogenesis/embryogenesis. Similarly, 08854, USA. fundamental new insights have been gained in the E-mail: [email protected] extracellular events connecting microbial infection with NF-jBinDrosophila S Minakhina and R Steward 6750 the activation of the Toll and Imd pathways. A more complete summary of the extracellular events can be found in other recent reviews (Roth, 2003; Brennan and Anderson, 2004; Mengin-Lecreulx and Lemaitre, 2005; Moussian and Roth, 2005; Tanji and Ip, 2005).

An overview of the dorsal–ventral signaling pathway in Drosophila melanogaster

Early systematic genetic screens conducted in several laboratories identified maternal-effect mutations that disrupt the dorsal–ventral pattern of the resulting embryos (Anderson and Nu¨ sslein-Volhard, 1984; Schupbach and Wieschaus, 1989). Twelve genes were identified whose products are deposited in the egg by the mother and are required for the proper elaboration of cell-fate along the dorsoventral axis of the embryo. Eleven of these genes comprise the ‘dorsal group.’ Null mutations in each of these genes give rise to embryos in which each cell along the dorsoventral axis assumes a dorsal fate. The remaining gene, cactus, is unique: loss- of-function mutations in this gene result in ventralized embryos. Molecular analyses of these genes and their products, combined with genetic epistasis experiments, have shown that the products are components of a signal- transduction pathway that culminates in the graded nuclear localization of Dorsal protein (Figure 1a) (Roth et al., 1989; Steward, 1989). Threshold levels of Dorsal control the expression of zygotic genes. Thus, dorsal- mediated gene expression represents the transition from maternal to zygotic control of dorsal–ventral patterning in the Drosophila embryo. Embryonic dorsal–ventral asymmetry is set-up in the egg chamber midway through oogenesis and is controlled mainly by epidermal growth factor signaling Figure 1 The Drosophila NF-kB pathways in early embryogenesis (reviewed by Roth et al., 2003; Moussian and Roth, and in the immune response. (a) Cross-section through a syncytial 2005). This asymmetry is transmitted to the embryo blastoderm stage embryo stained with anti-Dorsal antibody. Note through the interaction of two groups of genes. One the graded dorsal–ventral nuclear distribution of the darkly group is expressed specifically on the ventral side of staining Dorsal protein. Superimposed on the cross-section is the fate map. The ventral-most nuclei, which contain the greatest the follicle cells that surround the oocyte and secrete amount of Dorsal protein, will give rise to mesoderm, the adjacent the egg membranes. The other group is mainly made nuclei to mesectoderm and the next threshold level of Dorsal up of the three dorsal group genes gastrulation defective, protein will determine the future neurectoderm . The nuclei on the snake and easter, all three encoding serine proteases. dorsal side of the embryo express certain genes, which are repressed on the ventral side by Dorsal and co-repressors, and these dorsal They are expressed in the oocyte and their products are cells will differentiate into dorsal epidermis and the embryonic secreted into the perivitelline space, between the egg or membrane, called the amnioserosa. The types of enhancers early embryo and the egg membranes. The proteases controlled by Dorsal in different parts of the gradient are also form an activation cascade similar to that observed in indicated. (b) Schematics of Drosophila NF-kB pathways. On the blood clotting (Krem and Di Cera, 2002), that left is the Toll-Dorsal/Dif pathway, and on the right is the Imd- Relish pathway. culminates in the maturation and cleavage of the ligand, Spa¨ tzle, that in turn activates the Toll-Dorsal signaling pathway (Morisato and Anderson, 1994; Weber et al., signaling is ultimately responsible for the formation of 2003). the embryonic Dorsal nuclear gradient (Figure 1b). Spa¨ tzle is thought to form a gradient in the perivitel- Toll encodes a transmembrane protein with homology line fluid. The shape of the gradient is achieved by in the extracellular domain to thrombin receptor and an feedback loops involving Spa¨ tzle itself as well as the intracellular domain with homology to the interleukin three dorsal group serine proteases and a serpin- receptor, the toll/IL-1R (TIR) domain (Hashimoto type serine protease inhibitor Serpin27A (Spn27A) et al., 1988). Several Toll genes exist in Drosophila, (Hashimoto et al., 2003; Ligoxygakis et al., 2003). Toll but only the original Toll functions in dorsal–ventral

Oncogene NF-jBinDrosophila S Minakhina and R Steward 6751 polarity and the immune response. Minor functions activation of the Toll receptor (Bergmann et al., 1996; have been documented for Toll9 and 18 Wheeler, Reach et al., 1996). In addition, Dorsal undergoes another Toll protein (Williams et al., 1997; Ooi et al., signal-induced phosphorylation in the cytoplasm, an 2002; Bilak et al., 2003). essential step for its nuclear localization (Drier et al., Toll protein is present around the entire circum- 1999). Dorsal is then translocated to the nucleus ference of the embryo and is thought to be activated in mediated by its nuclear localization sequence and the a graded manner by its ligand Spa¨ tzle. Dominant, nuclear import machinery (Steward, 1987). Several constitutively active, alleles of Toll cause Dorsal to proteins have been shown to function in this process. translocate into nuclei around the entire circumference Tamo functions as a modulator of Dorsal nuclear of the embryo, even in the absence of the activity of the import, whereas the NTF-2 (nuclear import factor 2), proteases or Spa¨ tzle (Morisato and Anderson, 1994) and Emb/CRM1, Nup214 (nuclear pore protein 214) and lead to severe ventralization of embryos. Binding of the Mbo /Nup88 (Members-only, nuclear pore protein 88) Spa¨ tzle ligand is thought to cause Toll to dimerize/ are components of the general nuclear import/export multimerize, activating the receptor which then recruits machinery (Uv et al., 2000; Bhattacharya and Steward, the intracellular adaptor Tube – Krapfen/Myd88 com- 2002; Minakhina et al., 2003; Roth et al., 2003). plex (Horng and Medzhitov, 2001; Charatsi et al., 2003; Experiments manipulating the genetic dosage of Hu et al., 2004). dorsal and cactus indicate that dynamic mechanisms Tube recruits the Pelle kinase (PII), an IL-1R- exist to achieve stoichiometric equilibrium between associated kinases (IRAK) homolog, to the complex, Dorsal and Cactus (Govind et al., 1993). Both genes thereby increasing the Pelle concentration wherever the exhibit dosage sensitivity resulting in a reduction of complex is formed. Pelle has been shown to auto- embryonic viability and corresponding degrees of phosphorylate, and to phosphorylate Toll and Tube. dorsalization or ventralization of the embryos, depend- Tube has also been found to interact with the RH ing on the dorsal-to-cactus gene copy ratio. The balance domain of Dorsal in the yeast two-hybrid system between Dorsal and Cactus is critical for the proper (Edwards et al., 1997; Yang and Steward, 1997), formation of the gradient of nuclear Dorsal. suggesting that Tube may recruit the Dorsal–Cactus complex to the periphery of the embryo where the two proteins will be subjected to further modifications. Both Dorsal and Cactus are multiply phosphorylated Zygotic Dorsal target genes (Kidd, 1992; Whalen and Steward, 1993; Gillespie and Wasserman, 1994; Drier and Steward, 1997; Drier et al., In the nucleus Dorsal controls the expression of zygotic 1999), but the kinases responsible for these phospho- genes in a concentration-dependent manner and this rylations have yet to be identified. Although Pelle can process results in the patterning of the dorsal–ventral phosphorylate Cactus and Pelle in vitro, IRAK, the embryonic axis. Dorsal generally functions as a tran- vertebrate homolog of Pelle, does not phosphorylate scriptional activator; however, depending on the DNA IkB (Grosshans et al., 1994; Shen and Manley, 2002). elements present in the downstream promoters, co- Casein kinase II (CKII) specifically phosphorylates a set repressors can be recruited to promoters containing of serine residues within the Cactus PEST domain (Liu Dorsal-binding sites to cause repression of Dorsal target et al., 1997), but the in vivo roles of CKII in the Dorsal genes. The expression of specific target genes is and Dif pathways are not clear. dependent on the nuclear concentration of the Dorsal In contrast to the vertebrate inflammatory response, protein and on the composition of the regulatory the fly IkB kinases (IKK) do not function in the Toll elements/enhancers of the target gene. Enhancers con- pathways, but are responsible for the phosphoryla- tain conserved sequence motifs (kB-binding sites) with tion of Relish (see below). Functionally related kinases different affinities for binding by Dorsal and also that phosphorylate Cactus and Dorsal remain to be binding sites for co-regulators. Twist is one of the discovered. earliest target genes controlled by the highest concentra- Dorsal is retained in the cytoplasm by the IkB tion of Dorsal in the mesodermal nuclei (Figure 1a). protein, Cactus. The ankyrin repeats of Cactus are Twist is also a transcriptional activator and cooperates critical for the Dorsal–Cactus interaction (Geisler et al., with Dorsal in activating snail in the mesoderm and 1992; Kidd, 1992). Cactus binds to Dorsal within the mesectoderm. Dorsal and Twist also cooperate to C-terminal third of the RH domain (Tatei and Levine, activate the neurogenic gene, sim (single minded), 1995; Govind et al., 1996). Two specific residues within expressed in the neurectoderm and repressed by Snail Dorsal (C233 and S234) have been implicated in this in the mesoderm and mesectoderm. The three proteins interaction, as mutations in these amino acids render above bind the same enhancer and cooperate to induce Dorsal constitutively nuclear (Lehming et al., 1995; the expression of sim in two lateral stripes (Jiang et al., Drier et al., 2000). 1991; Kosman et al., 1991; Ip et al., 1992; Kasai et al., The nuclear translocation of Dorsal appears to be 1998). Genes that are only expressed in the dorsal regulated at several steps (Drier et al., 2000). The signal- ectoderm and the amnioserosa are repressed by Dorsal transduction system acts to phosphorylate Cactus and and co-repressors in nuclei where Dorsal is present. target it for degradation. The degradation of Cactus Using a computational approach, Levine and co- occurs in a graded manner, reflecting the graded workers have identified 30 genes as direct targets of

Oncogene NF-jBinDrosophila S Minakhina and R Steward 6752 Dorsal. Interestingly, one of these genes, mir-1, encodes 2005). In most cases, Dif is likely to require co- a micro RNA (Biemar et al., 2005). Based on sequence activators, such as TRAP80, to fully activate these and functional analysis of their enhancer regions the immune response genes (Park et al., 2003). Dorsal target genes they can be subdivided into three categories (Figure 1a). Type 1 genes/enhancers are regulated by peak levels of Dorsal protein on the ventral side of the embryo, Type 2 are regulated by intermediate The Imd pathway and the Drosophila immune response Dorsal concentrations, and Type 3 by the lowest concentrations of Dorsal (Stathopoulos and Levine, The Imd pathway responds to infections by Gram- 2002; Stathopoulos et al., 2002; Papatsenko and Levine, negative bacteria and the NF-kB protein in this path- 2005). Examples of specific genes regulated in the three way is Relish (Dushay et al., 1996; Cornwell and domains of the Dorsal gradient are listed in the Table 1. Kirkpatrick, 2001; Khush et al., 2001). In some cases, The enhancers of all these genes contain Dorsal-binding the Imd-Relish pathway can also be activated by certain sites and binding sites for co-factors that are themselves fungi, which do not activate the Dorsal/Dif pathway targets of Dorsal as well as binding sites for ubiqui- (Hedengren-Olcott et al., 2004). As described above, tous transcription factors (Uvell and Engstrom, 2003; Relish is a compound protein with an N-terminal RH Papatsenko and Levine, 2005). domain and a C-terminal ankyrin repeat domain that retains the protein in the cytoplasm. The Imd pathway is activated when proteoglycans in Gram-negative bacteria are recognized by a transmembrane receptor PGRP-LC, The Toll pathway and the Drosophila immune response and by PGRP-LE (Gottar et al., 2002; Takehana et al., 2004; Mengin-Lecreulx and Lemaitre, 2005). How the Natural or experimentally induced infections by fungi or recognition of bacteria by the PGRP proteins is bacteria elicit a specific response in both adult flies and transmitted to the intracellular Imd protein is not larvae. The proteoglycans of Gram-positive and Gram- understood. However, a fly tumor necrosis factor negative bacteria infections are sensed by distinct receptor-associated factor 2 homolog is required for pattern recognition proteins called PGRPs (peptidogly- both Dif and Relish activation (Cha et al., 2003). Other can recognition proteins, highly conserved in verte- downstream events in the Imd pathway are strongly brates) (Royet, 2004). Different PRGPs cooperate to reminiscent of the vertebrate NF-kB pathway (see activate the Toll pathway. The activation of PGRP-SA Figure 1b, Table 1). Genetic epistasis experiments show by Gram-positive bacteria leads to cleavage of Spa¨ tzle that the transforming growth factor beta-activated (Gobert et al., 2003). Fungal infection also leads to the kinase 1 (TAK1) kinase functions upstream from the cleavage of Spa¨ tzle, but the proteolytic cascade in this Drosophila IKK complex (Vidal et al., 2001; Silverman case involves the circulating serine protease Persephone et al., 2003). The fly IKK catalytic subunit is encoded by and a serine protease inhibitor, Necrotic (Ligoxygakis ird5 and NF-kB essential modulator by kenny and et al., 2002a; Pelte et al., 2006). Thus, the maturation of together they control the cleavage of Relish, mediated Spa¨ tzle activates Toll in both early embryo and in the by the caspase Dredd and its partner Fas-associated immune response, but the maturation is controlled by death domain (FADD) (Rutschmann et al., 2000b; different sets of proteases. Georgel et al., 2001; Lu et al., 2001; Leulier et al., 2002; Once Toll is activated, the intracellular events for the Silverman et al., 2003). Surprisingly, Dredd and FADD fly immune response and the proteins involved are are also required for activation of IKK in the Imd almost identical to those in the embryonic dorsoventral pathway, but how this occurs is not known (Zhou et al., pathway (Figure 1b). The TIR domain of Toll interacts 2005). Like many of the steps in activation of the with death domain proteins: Myd88 and Tube and the vertebrate NF-kB pathway, ubiquitination is also kinase Pelle. This receptor adaptor complex relays the involved in Relish activation: that is, the fly homologs signal to the Cactus-NF-kB protein (Dif and/or Dorsal) of the ubiquitin-conjugating enzymes Ubc13 and complex (Ip et al., 1993; Manfruelli et al., 1999; Meng UEV1a are required for activation of the TAK1- et al., 1999). Dif is the predominant transactivator in the >IKK step (Zhou et al., 2005). In addition, the Really antifungal and anti-Gram-positive bacterial infections Interesting New Gene (RING) domain Plenty of SH3s (Rutschmann et al., 2000a; Hedengren-Olcott et al., (POSH) protein controls the induction and termination 2004). Dorsal functions redundantly with Dif in larval of the Relish pathway by acting as a ubiquitinating immune response, but Dorsal does not function in the enzyme that controls the stability of TAK1 (Tsuda et al., adult immune response. In response to bacterial 2005), and similarly a SkpA component of the fly Skp1/ challenge, Dif and Dorsal undergo rapid nuclear cullin/F-box-E3 ubiquitin ligase may also negatively translocation in fat body cells to activate their own regulate the Imd pathway (Khush et al., 2002). In transcription as well as that of the antimicrobial peptide any event, activation of the Imd pathway results in genes. Drosomycin, Mechnikowin and Cecropin are C-terminal partial proteolysis of Relish, much like what expressed in response to fungal infections, and Defensin occurs with p100 in vertebrates, leading to the nuclear and Mechnikovin in response to Gram-positive bacteria translocation of the RH domain part of Relish. Relish (Lemaitre et al., 1995; Levashina et al., 1995; Levashina controls the expression of the antimicrobial peptides et al., 1998; Manfruelli et al., 1999; Imler and Bulet, Attacin, Cecropin and Diptericin (Dushay et al., 1996;

Oncogene NF-jBinDrosophila S Minakhina and R Steward 6753 Table 1 Components and targets of Drosophila NF-kB pathways Gene name Gene product References

Toll pathway Spa¨ tzle (spz) Secreted protein endogenous Toll ligand, NGF-like (Roth, 1994; Weber et al., 2003) Toll (Tl) Transmembrane receptor intracellular: IL-1 receptor; (Hashimoto et al., 1988; Hu et al., 2004) extracellular: thrombin receptor Tube (tub) Adaptor protein (Letsou et al., 1991; Charatsi et al., 2003) Pelle (pll) Serine/threonine kinase IRAK (Shelton and Wasserman, 1993; Muzio et al., 1997) Myd88 Adapter that associates with both Toll receptor and (Horng and Medzhitov, 2001; Charatsi et al., 2003) Pelle; contains Toll/IL-1R homologous region (TIR) Cactus (cact) IkB-protein with ankyrin repeats (Roth et al., 1991; Geisler et al., 1992; Kidd, 1992) Dorsal (dl) NF-kB transcription factor (Steward, 1987) Dorsal-related immunity NF-kB transcription factor (Ip et al., 1993) factor (Dif)

Transcriptional targets of the pathway Type 1 (Stathopoulos and Levine, 2002; Stathopoulos et al., 2002; Papatsenko and Levine, 2005) Twist (twi) Transcription factor Snail (sna) Transcription factor Heartless (htl) FGF-R etc.

Type 2 k Vein (vn) EGF-R binding Rhomboid (rho) brinker Regulation of EGFR (brk) Regulation of EGFR and TGF-b Single-minded (sim) Transcription factor etc.

Type 3 k Decapentaplegic (dpp) TGFb, BMP-signaling Short gastrulation (sog) Vertebrate Chordin (Chd), BMP-signaling Tolloid (tld) Tolloid metalloproteinase, BMP1 Zerknullt (zen) Transcription factor etc. Drosomycin (drs) Antimicrobial peptides (Imler and Bulet, 2005) metchnikowin (mtk) miR-1 MicroRNA (Biemar et al., 2005)

Imd pathway PGRP-LC Peptidoglycan transmembrane receptor (Gottar et al., 2002; Choe et al., 2005) PGRP-LE other PGRPs Peptidoglycan receptors (Royet, 2004; Takehana et al., 2004) Immune deﬁciency (imd) DEATH domain-containing, homologue of (Georgel et al., 2001) TNF-receptor-interacting protein (RIP) TAK1 TGF-b activated kinase 1 homologue of MAP kinase (Vidal et al., 2001) kinase kinase (MAPKKK) Immune response IkB kinase IKK-b (Rutschmann et al., 2000b; Lu et al., 2001) deﬁcient 5 (ird5) Kenny (key) NEMO (Rutschmann et al., 2000b) Relish (Rel) NF-kB transcription factor (N terminus) with (Dushay et al., 1996) IkB-like ankyrin repeats (C terminus)

Transcriptional targets of the pathway Attacin (att) Antimicrobial peptides Reviewed by Imler and Bulet (2005) Cecropin (cec) Diptericin (dpt) etc.

Abbreviations: EGFR, epidermal growth factor receptor; FGF, ﬁbroblast growth factor; NGF, nerve growth factor; IL, interleukin, IKK, IkB kinase; IRAK, IL-1R-associated kinases; NEMO, NF-kB essential modulator; NF-kB, nuclear factor kappa B; PGRP, peptidoglycan recognition proteins; TAK, transforming growth factor beta-activated kinase 1; TGF, transforming growth factor.

Stoven et al., 2003; Hedengren-Olcott et al., 2004; Imler antimicrobial genes by both Relish (and Dif). In an and Bulet, 2005). Exactly how proteolytically cleaved opposite manner, binding of activator protein-1 to Relish activates target genes is not known; for example, several Relish-activated target genes can recruit the whether there is a Bcl-3 homolog in ﬂies that acts histone deacetylase HDAC1 to terminate gene activa- as a co-activator for processed Relish is not known. tion (Kim et al., 2005). However, Yagi and Ip (2005) have recently shown that The ﬁndings that imd mutants are impaired in the the Helicase89B is required for the activation of response to Gram-negative bacterial infections and Toll

Oncogene NF-jBinDrosophila S Minakhina and R Steward 6754 mutants are sensitive to fungal and Gram-positive of the hemocyte population. In these mutants, the bacteria suggest that distinct pathways are used in these overproliferation of blood cells leads to abnormal defenses. However, the level of activation and the activation of the cellular immune response and results expression of the immune peptides depend on the in the formation of melanotic tumors or nodules (Qiu infecting bacteria. Some bacteria apparently can activate et al., 1998; Govind, 1996; Minakhina and Steward, in both pathways. In these situations, Dif and/or Relish press). In addition to the melanotic nodules formed by can activate Cecropin expression, usually thought of as hemocytes, cactus mutants have melanized tumors a Relish target (Hedengren-Olcott et al., 2004). The consisting of diverse tissues, probably caused by Dorsal target enhancers may be bound by homodimers and/or activity in these cells. This phenotype suggests that heterodimers of the three NF-kB proteins. For example, cactus and dorsal are expressed in some tissues that do the enhancer of the immune peptide gene drosomycin not express Toll (Minakhina and Steward, in press). contains kB-elements that can be bound by both Dorsal/ Studies of the lesswright gene, which encodes a small Dif and Relish (Han and Ip, 1999). Thus the transcrip- ubiquitin-like modifier (SUMO)-adding enzyme, indi- tion factors may cooperatively regulate several immune cate that SUMO addition is required for proper peptide genes. regulation of Dorsal and Dif in hemocytes: namely, loss-of-function lesswright mutations result in constitutively nuclear Dorsal and Dif proteins in hemocytes and Additional functions of Drosophila NF-kB proteins the formation of melanotic tumors owing to increased The role of NF-kB pathways in fly immunity is not numbers of hemocytes (Huang et al., 2005). Similarly, restricted to activation of immune peptides. The Imd- the Drosophila Ubc9 SUMO-lating gene also appears to Relish pathway has been linked to a programmed cell repress Dorsal and Dif function in controlling lamelo- death by proteins functioning both in Relish activation cyte proliferation and drosomycin gene expression (Chiu and in apoptosis (Takatsu et al., 2000; Georgel et al., et al., 2005). 2001; Kleino et al., 2005; Tsuda et al., 2005). The Toll Toll-independent functions for Dorsal and Cactus are pathway is a central regulator of multiple aspects of also proposed in the neuromuscular junctions (Bolatto Drosophila immunity, including resistance to viruses, et al., 2003). Mutations in dorsal and cactus cause regulation of the melanization cascade and the cellular neuromuscular defects, but how the proteins are immune response. These aspects of the immune response controlled and function in the junctions is not well are not as well understood as the regulation of immune understood (Bolatto et al., 2003; Beramendi et al., 2005). peptide expression. Few studies of response to viral infection in Droso- phila have been published. Different viruses provoke The NF-jB pathway in other invertebrates specific responses generally not involving the NF-kB pathways (Cherry and Perrimon, 2004; Dostert et al., NF-kB pathways are present in mollusks, arthropods 2005). Toll and Dif have only been implicated in the and vertebrates and generally function in immune response to Drosophila X virus (Zambon et al., 2005). response and development. A Relish homologous gene Toll, along with Jak-signal transducer and activator has been cloned from the Japanese Oyster, Crassostrea of transcription, are key pathways functioning in the gigas. It can activate immune peptides when expressed in cellular immune response. This response involves S2 Drosophila tissue culture cells (Montagnani et al., changes in the hematopoietic system, including blood 2004). Similarly, a Dorsal/Dif homolog has been cell proliferation, differentiation and activation. These isolated from the beetle, Allomyrina dichotoma; this blood cells are the plasmatocytes that function in protein activates the transcription of immune peptide phagocytosis, lamellocytes, that encapsulate foreign genes when transfected into Bombyx mori cells (Sagisaka objects such as parasites, and crystal cells, that control et al., 2004; Tanaka et al., 2005). the melanization cascade (Evans et al., 2003; Meister The ancient origin of NF-kB proteins is underlined by and Lagueux, 2003; Nappi and Christensen, 2005). The their presence in the most ancient arthropod, the melanization cascade is also regulated by serine pro- horseshoe crab, Limulus. This animal has survived teases and the same protease inhibitor, Spn27A, unchanged for B550 million years and has genes that functioning in Spa¨ tzle maturation in the early embryo are homologous to dorsal/Dif and to Relish, as well as an (Castillejo-Lopez and Hacker, 2005; Nappi et al., 2005). IkB homolog. All three function in the regulation of the The Toll pathway is thought to negatively regulate Limulus immune defense (Wang et al., 2006). Spn27A and thereby control melanization (Ligoxygakis The NF-kB proteins and the immune response are et al., 2002b; Castillejo-Lopez and Hacker, 2005; Nappi particularly well studied in mosquitoes (Barillas-Mury et al., 2005). et al., 1996; Cramer et al., 1999; Shin et al., 2005). Mutations in Toll, cactus and tube strongly affect the Generally, mosquitoes have two NF-kB proteins: one encapsulation of wasps in larvae (Qiu et al., 1998; family I NF-kB protein similar to Dif and Dorsal and Sorrentino et al., 2004). The encapsulation capacity is one Relish homolog. The proteins function in both the tightly linked to Drosophila blood cell proliferation cellular and humoral immune response. In Aedes aegypti and activation. Mutations in Toll and cactus that lead the family I NF-kB protein functions in the defense to constitutive activation of the Toll pathway cause against fungal infections, whereas the Relish homolog an overgrowth of hematopoietic organs and expansion functions in defense against Gram-negative bacteria.

Oncogene NF-jBinDrosophila S Minakhina and R Steward 6755 No defense against Gram-positive bacteria has been specific to the NF-kB pathways, but have more general found (Bian et al., 2005). In A. gambiae the Relish functions, such as ubiquination, protein degradation homolog functions in the defense against Gram-positive and nuclear import. and Gram-negative bacteria (Meister et al., 2005). The Drosophila NF-kB pathways are relatively simple Arabidopsis and Caenorhabditis elegans have proteins and involve many fewer genes than their vertebrate weakly related to IkB-related proteins, but they do not counterparts and it is likely that only very few genes have NF-kB proteins, even though they both have a functioning specifically in these pathways remain to be host defense system (Yang et al., 1997; Graef et al., discovered. In the future we will learn the precise 2001; Chisholm et al., 2006). Moreover, C. elegans has function of the pathways in the cellular immune homologs of other genes involved in the immune response and hematopoiesis, and possibly discover yet response in both Drosophila and vertebrates, but they unknown functions. do not function in the nematode defense system (Pujol Because of the impressive conservation of the NF-kB et al., 2001). pathways, Drosophila is used as a model to study defense of the organism to infection with parasites, virus and bacteria. This is done both in the fly and in Drosophila Concluding remarks and future directions tissue culture cells (Brandt et al., 2004). A Drosophila model for studying malaria infection has been esta- When we wrote our review for this journal approxi- blished and it appears that the cellular immune response mately 8 years ago, we predicted that genetic screens is controlling the multiplication of Plasmodia in the fly would identify additional players in the Toll and Imd (Schneider and Shahabuddin, 2000; Vernick et al., pathways. This prediction came only partially true. 2005). No doubt these approaches will identify new, Many new genes functioning in the activation of the Toll conserved genes essential for the defense against specific and Imd pathways were identified and the first steps in pathogens and will hopefully result in the development response to bacterial and fungal infections are much of new treatments to deal with infections. better elucidated now. Our understanding of the events leading to the maturation of the Spa¨ tzle ligand during Acknowledgements oogenesis, embryogenesis and the immune response has also vastly increased. Finally, kinases functioning in the Our work was supported by a grant from the National Imd pathway have been identified. Other genes that Institute of Health and by the W Horace Goldsmith have been isolated in the past several years are not Foundation.

References

Anderson KV, Nu¨ sslein-Volhard C. (1984). Nature 311: Chisholm ST, Coaker G, Day B, Staskawicz BJ. (2006). Cell 223–227. 124: 803–814. Barillas-Mury C, Charlesworth A, Gross I, Richman A, Chiu H, Ring BC, Sorrentino TP, Kalamarz M, Garza D, Hoffmann JA, Kafatos FC. (1996). EMBO J 15: 4691–4701. Govind S. (2005). Dev Biol 288: 60–72. Beramendi A, Peron S, Megighian A, Reggiani C, Cantera R. Choe KM, Lee H, Anderson KV. (2005). Proc Natl Acad Sci (2005). Neuroscience 134: 397–406. USA 102: 1122–1126. Bergmann A, Stein D, Geisler R, Hagenmaier S, Schmid B, Cornwell WD, Kirkpatrick RB. (2001). J Cell Biochem 82: Fernandez N et al. (1996). Mech Dev 60: 109–123. 22–37. Bhattacharya A, Steward R. (2002). EMBO Rep 3: Cramer P, Varrot A, Barillas-Mury C, Kafatos FC, 378–383. Muller CW. (1999). Structure 7: 841–852. Bian G, Shin SW, Cheon HM, Kokoza V, Raikhel AS. (2005). Dostert C, Jouanguy E, Irving P, Troxler L, Galiana- Proc Natl Acad Sci USA 102: 13568–13573. Arnoux D, Hetru C et al. (2005). Nat Immunol 6: 946–953. Biemar F, Zinzen R, Ronshaugen M, Sementchenko V, Drier EA, Govind S, Steward R. (2000). Curr Biol 10: 23–26. Manak JR, Levine MS. (2005). Proc Natl Acad Sci USA Drier EA, Huang LH, Steward R. (1999). Genes Dev 13: 102: 15907–15911. 556–568. Bilak H, Tauszig-Delamasure S, Imler JL. (2003). Biochem Soc Drier EA, Steward R. (1997). Semin Cancer Biol 8: 83–92. Trans 31: 648–651. Dushay MS, Asling B, Hultmark D. (1996). Proc Natl Acad Bolatto C, Chifﬂet S, Megighian A, Cantera R. (2003). Sci USA 93: 10343–10347. J Neurobiol 54: 525–536. Edwards DN, Towb P, Wasserman SA. (1997). Development Brandt SM, Dionne MS, Khush RS, Pham LN, Vigdal TJ, 124: 3855–3864. Schneider DS. (2004). PLoS Biol 2: e418. Evans CJ, Hartenstein V, Banerjee U. (2003). Dev Cell 5: Brennan CA, Anderson KV. (2004). Annu Rev Immunol 22: 673–690. 457–483. Geisler R, Bergmann A, Hiromi Y, Nu¨ sslein-Volhard C. Castillejo-Lopez C, Hacker U. (2005). Biochem Biophys Res (1992). Cell 71: 613–621. Commun 338: 1075–1082. Georgel P, Naitza S, Kappler C, Ferrandon D, Zachary D, Cha GH, Cho KS, Lee JH, Kim M, Kim E, Park J et al. Swimmer C et al. (2001). Dev Cell 1: 503–514. (2003). Mol Cell Biol 23: 7982–7991. Gillespie SK, Wasserman SA. (1994). Mol Cell Biol 14: Charatsi I, Luschnig S, Bartoszewski S, Nu¨ sslein-Volhard C, 3559–3568. Moussian B. (2003). Mech Dev 120: 219–226. Gobert V, Gottar M, Matskevich AA, Rutschmann S, Cherry S, Perrimon N. (2004). Nat Immunol 5: 81–87. Royet J, Belvin M et al. (2003). Science 302: 2126–2130.

Oncogene NF-jBinDrosophila S Minakhina and R Steward 6756 Gottar M, Gobert V, Michel T, Belvin M, Duyk G, Liu ZP, Galindo RL, Wasserman SA. (1997). Genes Dev 11: Hoffmann JA et al. (2002). Nature 416: 640–644. 3413–3422. Govind S. (1996). Biochem Soc Trans 24: 39–44. Lu Y, Wu LP, Anderson KV. (2001). Genes Dev 15: 104–110. Govind S, Brennan L, Steward R. (1993). Development 117: Manfruelli P, Reichhart JM, Steward R, Hoffmann JA, 135–148. Lemaitre B. (1999). EMBO J 18: 3380–3391. Govind S, Drier E, Huang LH, Steward R. (1996). Mol Cell Meister M, Lagueux M. (2003). Cell Microbiol 5: 573–580. Biol 16: 1103–1114. Meister S, Kanzok SM, Zheng XL, Luna C, Li TR, Hoa NT Graef IA, Gastier JM, Francke U, Crabtree GR. (2001). Proc et al. (2005). Proc Natl Acad Sci USA 102: 11420–11425. Natl Acad Sci USA 98: 5740–5745. Meng X, Khanuja BS, Ip YT. (1999). Genes Dev 13: 792–797. Grosshans J, Bergmann A, Haffter P, Nusslein-Volhard C. Mengin-Lecreulx D, Lemaitre B. (2005). J Endotoxin Res 11: (1994). Nature 372: 563–566. 105–111. Han ZS, Ip YT. (1999). J Biol Chem 274: 21355–21361. Minakhina S, Steward R. (2006). Genetics in press. Hashimoto C, Hudson KL, Anderson KV. (1988). Cell 52: Minakhina S, Yang J, Steward R. (2003). Genes Cells 8: 269–279. 299–310. Hashimoto C, Kim DR, Weiss LA, Miller JW, Morisato D. Montagnani C, Kappler C, Reichhart JM, Escoubas JM. (2003). Dev Cell 5: 945–950. (2004). FEBS Lett 561: 75–82. Hedengren-Olcott M, Olcott MC, Mooney DT, Ekengren S, Morisato D, Anderson KV. (1994). Cell 76: 677–688. Geller BL, Taylor BJ. (2004). J Biol Chem 279: Moussian B, Roth S. (2005). Curr Biol 15: R887–R899. 21121–21127. Muzio M, Ni J, Feng P, Dixit VM. (1997). Science 278: Horng T, Medzhitov R. (2001). Proc Natl Acad Sci USA 98: 1612–1615. 12654–12658. Nappi AJ, Christensen BM. (2005). Insect Biochem Mol Biol Hu X, Yagi Y, Tanji T, Zhou S, Ip YT. (2004). Proc Natl Acad 35: 443–459. Sci USA 101: 9369–9374. Nappi AJ, Frey F, Carton Y. (2005). J Insect Physiol 51: Huang L, Ohsako S, Tanda S. (2005). Dev Biol 280: 407–420. 197–205. Huguet C, Crepieux P, Laudet V. (1997). Oncogene 15: Ooi JY, Yagi Y, Hu X, Ip YT. (2002). EMBO Rep 3: 82–87. 2965–2974. Papatsenko D, Levine M. (2005). Proc Natl Acad Sci USA 102: Imler JL, Bulet P. (2005). Chem Immunol Allergy 86: 1–21. 4966–4971. Ip YT, Park RE, Kosman D, Yazdanbakhsh K, Levine M. Park JM, Kim JM, Kim LK, Kim SN, Kim-Ha J, Kim JH (1992). Genes Dev 6: 1518–1530. et al. (2003). Mol Cell Biol 23: 1358–1367. Ip YT, Reach M, Engstrom Y, Kadalayil L, Cai H, Gonzalez- Pelte N, Robertson AS, Zou Z, Belorgey D, Dafforn TR, Crespo S et al. (1993). Cell 75: 753–763. Jiang H et al. (2006). Insect Biochem Mol Biol 36: 37–46. Jiang J, Kosman D, Ip YT, Levine M. (1991). Genes Dev 5: Pujol N, Link EM, Liu LX, Kurz CL, Alloing G, Tan MW 1881–1891. et al. (2001). Curr Biol 11: 809–821. Kasai Y, Stahl S, Crews S. (1998). Gene Expr 7: 171–189. Qiu P, Pan PC, Govind S. (1998). Development 125: Khush RS, Cornwell WD, Uram JN, Lemaitre B. (2002). 1909–1920. Curr Biol 12: 1728–1737. Reach M, Galindo RL, Towb P, Allen JL, Karin M, Khush RS, Leulier F, Lemaitre B. (2001). Trends Immunol 22: Wasserman SA. (1996). Dev Biol 180: 353–364. 260–264. Rijsewijk F, Schuermann M, Wagenaar E, Parren P, Weigel D, Kidd S. (1992). Cell 71: 623–635. Nusse R. (1987). Cell 50: 649–657. Kim T, Yoon J, Cho H, Lee WB, Kim J, Song YH et al. Roth P, Xylourgidis N, Sabri N, Uv A, Fornerod M, (2005). Nat Immunol 6: 211–218. Samakovlis C. (2003). J Cell Biol 163: 701–706. Kleino A, Valanne S, Ulvila J, Kallio J, Myllymaki H, Roth S. (1994). Curr Biol 4: 755–757. Enwald H et al. (2005). EMBO J 24: 3423–3434. Roth S. (2003). Philos Trans R Soc Lond B Biol Sci 358: Kosman D, Ip YT, Levine M, Arora K. (1991). Science 254: 1317–1329. 118–122. Roth S, Hiromi Y, Godt D, Nusslein-Volhard C. (1991). Krem MM, Di Cera E. (2002). Trends Biochem Sci 27: Development 112: 371–388. 67–74. Roth S, Stein D, Nusslein-Volhard C. (1989). Cell 59: Lehming N, McGuire S, Brickman JM, Ptashne M. (1995). 1189–1202. Proc Natl Acad Sci USA 92: 10242–10246. Royet J. (2004). Cell Mol Life Sci 61: 537–546. Lemaitre B, Kromer-Metzger E, Michaut L, Nicolas E, Rutschmann S, Jung AC, Hetru C, Reichhart JM, Meister M, Georgel P et al. (1995). Proc Natl Acad Sci Hoffmann JA, Ferrandon D. (2000a). Immunity 12: 569–580. USA 92: 9465–9569. Rutschmann S, Jung AC, Zhou R, Silverman N, Letsou A, Alexander S, Orth K, Wasserman SA. (1991). Proc Hoffmann JA, Ferrandon D. (2000b). Nat Immunol 1: 342–347. Natl Acad Sci USA 88: 810–814. Sagisaka A, Tanaka H, Furukawa S, Yamakawa M. (2004). Leulier F, Vidal S, Saigo K, Ueda R, Lemaitre B. (2002). Curr Biochim Biophys Acta 1678: 85–93. Biol 12: 996–1000. Schneider D, Shahabuddin M. (2000). Science 288: 2376–2379. Levashina EA, Ohresser S, Bulet P, Reichhart JM, Hetru C, Schupbach T, Wieschaus E. (1989). Genetics 121: 101–117. Hoffmann JA. (1995). Eur J Biochem 233: 694–700. Shelton CA, Wasserman SA. (1993). Cell 72: 515–525. Levashina EA, Ohresser S, Lemaitre B, Imler JL. (1998). J Mol Shen B, Manley JL. (2002). Development 129: 1925–1933. Biol 278: 515–527. Shin SW, Kokoza V, Bian G, Cheon HM, Kim YJ, Ligoxygakis P, Pelte N, Hoffmann JA, Reichhart JM. (2002a). Raikhel AS. (2005). J Biol Chem 280: 16499–16507. Science 297: 114–116. Silverman N, Zhou R, Erlich RL, Hunter M, Bernstein E, Ligoxygakis P, Pelte N, Ji C, Leclerc V, Duvic B, Belvin M Schneider D et al. (2003). J Biol Chem 278: 48928–48934. et al. (2002b). EMBO J 21: 6330–6337. Sorrentino RP, Melk JP, Govind S. (2004). Genetics 166: Ligoxygakis P, Roth S, Reichhart JM. (2003). Curr Biol 13: 1343–1356. 2097–2102. Stathopoulos A, Levine M. (2002). Dev Biol 246: 57–67.

Oncogene NF-jBinDrosophila S Minakhina and R Steward 6757 Stathopoulos A, Van Drenth M, Erives A, Markstein M, Vernick KD, Oduol F, Lazzaro BP, Glazebrook J, Xu J, Levine M. (2002). Cell 111: 687–701. Riehle M et al. (2005). Curr Top Microbiol Immunol 295: Steward R. (1987). Science 238: 692–694. 383–415. Steward R. (1989). Cell 59: 1179–1188. Vidal S, Khush RS, Leulier F, Tzou P, Nakamura M, Stoven S, Silverman N, Junell A, Hedengren-Olcott M, Lemaitre B. (2001). Genes Dev 15: 1900–1912. Erturk D, Engstrom Y et al. (2003). Proc Natl Acad Sci Wang XW, Tan NS, Ho B, Ding JL. (2006). Proc Natl Acad USA 100: 5991–5996. Sci USA 103: 4204–4209. Takatsu Y, Nakamura M, Stapleton M, Danos MC, Weber AN, Tauszig-Delamasure S, Hoffmann JA, Lelievre E, Matsumoto K, O’Connor MB et al. (2000). Mol Cell Biol Gascan H, Ray KP et al. (2003). Nat Immunol 4: 20: 3015–3026. 794–800. Takehana A, Yano T, Mita S, Kotani A, Oshima Y, Kurata S. Whalen AM, Steward R. (1993). J Cell Biol 123: 523–534. (2004). EMBO J 23: 4690–4700. Williams MJ, Rodriguez A, Kimbrell DA, Eldon ED. (1997). Tanaka H, Yamamoto M, Moriyama Y, Yamao M, EMBO J 16: 6120–6130. Furukawa S, Sagisaka A et al. (2005). Biochim Biophys Yagi Y, Ip YT. (2005). EMBO Rep 6: 1088–1094. Acta 1730: 10–21. Yang J, Steward R. (1997). Proc Natl Acad Sci USA 94: Tanji T, Ip YT. (2005). Trends Immunol 26: 193–198. 14524–14529. Tatei K, Levine M. (1995). Mol Cell Biol 15: 3627–3634. Yang Y, Shah J, Klessig DF. (1997). Genes Dev 11: 1621–1639. Tsuda M, Langmann C, Harden N, Aigaki T. (2005). EMBO Zambon RA, Nandakumar M, Vakharia VN, Wu LP. (2005). Rep 6: 1082–1087. Proc Natl Acad Sci USA 102: 7257–7262. Uv AE, Roth P, Xylourgidis N, Wickberg A, Cantera R, Zhong W, Sternberg PW. (2006). Science 311: 1481–1484. Samakovlis C. (2000). Genes Dev 14: 1945–1957. Zhou R, Silverman N, Hong M, Liao DS, Chung Y, Chen ZJ Uvell H, Engstrom Y. (2003). Mol Cell Biol 23: 8272–8281. et al. (2005). J Biol Chem 280: 34048–34055.

Oncogene