The fruit fly Drosophila melanogaster Toll signaling pathway has an evolutionarily conserved role in controlling immune responses. Whereas the microbial recognition mechanisms and the core signaling pathway leading to activation of the humoral immune response via the NF-κB transcription factors have been well established for many years, the mechanistic understanding of the effector functions at the molecular level is currently rapidly evolving. In this review, we discuss the current developments in elucidating the role of the Drosophila Toll signaling pathway in immunity. We discuss the emerging role of Toll in viral infections and sex-specific differences in immunity. Mainly, we focus on Toll pathway regulation, the effector molecules, and cellular immunity.
In 2011, the importance of the Drosophila model in innate immunity studies was recognized by awarding the Nobel Prize in Physiology or Medicine to the researchers who discovered the fundamental basis of innate immune responses and their role in activating adaptive immunity. One half of the prize was awarded jointly to Bruce A. Beutler and Jules A. Hoffmann “for their discoveries concerning the activation of innate immunity,” and the other half was awarded to Ralph M. Steinman “for his discovery of the dendritic cell and its role in adaptive immunity.” In the work of Prof. Hoffmann’s group, the Drosophila melanogaster Toll receptor was identified as being essential in the defense against fungal infections (1). This finding was soon followed by the discovery of TLRs in mammals, opening new horizons for a deeper understanding of how mammalian immune responses are regulated (2, 3). In our Brief Reviews article from January 2011 in The Journal of Immunology, we reviewed the literature leading to the understanding of the Drosophila Toll pathway function in both embryonic development and immunity (4). In the present review, we revisit the topic of the D. melanogaster Toll signaling pathway and describe, in particular, the immune-related developments in Toll pathway research during the past decade, including findings concerning both humoral and cell-mediated arms of innate immunity.
Developments in microbe recognition at the receptor level and in the core Toll pathway
The Drosophila Toll receptor differs from the mammalian TLRs in that the Drosophila Toll receptor functions as a cytokine receptor (reviewed in Ref. 5), whereas the mammalian TLRs recognize foreign structures directly and thus are pattern recognition receptors (PRRs). In Drosophila, there are nine genes encoding Toll receptors (Toll-1 and Toll-9), out of which Toll-1 (Toll) has the main role in mediating innate immune signaling (4). Other Toll receptors may have tissue- and/or infection type-specific roles (described below).
Events upstream of the Drosophila Toll receptor to activate the Toll pathway in different contexts have been thoroughly dissected earlier and are reviewed in Valanne et al. (as shown in figure 1 in Ref. 4). Recent developments include clarifying the structure of the Spatzle (Spz)/Toll receptor complex; in two independent studies it was shown that a single Spz dimer binds one Toll receptor ectodomain in 1:1 complex (6, 7). The stoichiometry of Spz binding to Toll is similar to some mammalian neurotrophins, where one cystine-knot dimer binds one receptor chain (7). Furthermore, Kellenberger et al. (8) have resolved the crystal structure of Grass, the clip serine protease involved in Toll pathway activation upstream of Sphinx1/2/Spirit/Spheroide (8). In addition, the role of thioester-containing proteins (TEPs) in immune response has been studied, with the secreted TEPs (TEP1, 2, 3, and 4) shown to play a role in Toll pathway activation, likely by taking part in the recognition of certain Gram-positive bacteria and fungi (9).
The activation of pathogen recognition receptors by microbial molecules has also been thoroughly studied (e.g., as shown in figure 1 of Ref. 4; see also Ref. 10). In the current model of Toll pathway activation, bacterial and fungal structures are recognized by specific PRRs, leading to the activation of downstream cascades and, ultimately, the cleavage and activation of the Toll receptor ligand Spz. Recently, Gyc76C, a receptor guanylate cyclase, was shown to function as a parallel immune receptor to Toll, modulating NF-κB signaling downstream of MyD88 (11). Furthermore, it was shown that Gyc76C mediates both humoral responses (e.g., antimicrobial peptide (AMP) induction) and cellular responses (hemocyte (fly blood cell) proliferation), but with distinct mechanisms: for the humoral response, Gyc76C-mediated AMP induction requires production of the secondary messenger cGMP, whereas hemocyte proliferation is cGMP-independent (12).
Another proteolytic cascade leading to Spz activation is initiated by proteases secreted by microbes, which can be considered as danger signals (i.e., damage-associated molecular patterns or danger-associated molecular patterns [DAMPs]) (13–15). DAMPs can also be endogenous molecules generated upon injury or cellular damage, but in this review we discuss the danger signals coming from microbes upon infection. The mechanism behind the function of Persephone (Psh) in recognizing DAMPs upstream of Spz processing enzyme (SPE) was recently further studied (16). It was shown that certain fungal or bacterial proteases, which are important virulence factors for host colonization, prime Psh for the cleavage and activation by the endogenous cysteine cathepsin 26-29-p. Specifically, the microbial proteases act as danger signals to the host before tissue damage occurs, and the prodomain of Psh functions as a bait for a broad range of these proteases. Subsequent action of the cysteine cathepsin 26-29-p on the primed Psh leads to the activation of the Toll pathway. This highlights the potential importance of cysteine cathepsins also in mammalian inflammatory diseases, a factor that has recently been discussed (e.g., Ref. 17). Of note, it was recently discovered that psh is likely to be a relatively recent duplication of the serine protease gene Hayan, and that these two proteins redundantly activate the Toll pathway downstream of PRRs (18). It is evident that this system of proteolytic activation by danger signals can sense a plethora of microbes, regardless of their origin, type, or specificity. Therefore, this finding leads to a conceptually novel immune system function in animals, although similar guard mechanisms have been known to play a role in plants (19). Recently, a parallel immune mechanism has been identified also in mammals; it was shown that the NLRP1 inflammasome is proteolytically activated by diverse microbial enzymes (20).
The core Toll pathway was extensively mapped already by 2011 (21, 22), but one important question remained—what is the kinase phosphorylating the Drosophila IκB homolog Cactus (Cact)? Cact needs to be phosphorylated for its degradation and the subsequent activation of the pathway. After years of speculation, Daigneault et al. (23) showed that Pelle phosphorylates Cact at the serines required for signal transduction and thus acts as the Cact kinase. Pelle can also phosphorylate the required sites of IκBα (23). Whereas the understanding of the core pathway has not changed much during the past 10 y, much more insight has been gained relating to regulation and fine-tuning of the Toll pathway.
Regulation of the Toll pathway
As Toll signaling is central in inflammatory and immune responses, it needs to be tightly controlled. Many aspects of the regulation of the Toll pathway have been investigated in detail (Fig. 1). At the level of modifying the structure of chromatin, the Osa-containing Brahma complex (BAP) was shown to negatively regulate Toll pathway-mediated immune reactions both in vitro and in vivo in Drosophila (24). In a transcriptome study, Osa was also shown to regulate the expression of metabolic genes, highlighting the importance of the interplay between immunity and metabolism (24, 25). Another identified negative regulator of the Toll pathway is the retromer complex, shown to function upstream of the Toll receptor but downstream of SPE. Retromer is a protein complex originally identified in yeast (26). The complex is associated with the cytosolic side of the cell membrane and regulates the trafficking of protein cargo from endosomes to the trans-Golgi network (26, 27). Retromer is composed of five components: sorting nexin 1/2 (SNX1/2), SNX5/6, vacuolar protein sorting 29 (Vps29), Vps26, and Vps35. Zhou et al. (28) speculated that retromer is involved in an as yet unclear mechanism of Spz maturation. Besides general Toll pathway regulation, tissue-specific regulation mechanisms of the immune response have been studied in Drosophila respiratory epithelium, that is, trachea, where Tollo (Toll-8) was shown to negatively regulate the immune response signals coming via the Imd pathway. The ligand (or one of the ligands) to activate Tollo is a Spz homolog Spz2/DNT1, but the exact mechanism between Imd pathway and Tollo interplay in the tracheal tissue remains elusive (29).
Regulatory mechanisms studied in greater detail in recent years also include posttranslational modifications such as ubiquitination and sumoylation. Both ubiquitination (reviewed in Ref. 30) and sumoylation (reviewed in Ref. 31) are mechanisms that can regulate immune pathway proteins, either by activating them, repressing them, or targeting them for degradation (e.g., Refs. 32, 33). In the Drosophila Toll pathway, Pellino (Pli) has been identified as a Pelle-interacting factor (34–36). First, Pli was suggested to positively regulate Toll pathway activity, as ubiquitous overexpression of Pli resulted in enhanced Toll pathway target gene Drosomycin (Drs) expression (35). Somewhat controversially, it was later demonstrated that knockdown or overexpression of Pli in the fat body, or in D. melanogaster Schneider 2 (S2) cells, has effects that suggest that Pli acts as a negative regulator of the Toll pathway in these contexts (36). The authors demonstrated that at the plasma membrane, Pli interacts with the adaptor protein MyD88, regulating its ubiquitination and targeting it for degradation (36). In mammals there are several Pli homologs that have opposing roles in different cells/tissues, indicating that the regulation mediated by Pli family members is complex and appears to be context-dependent (37). Looking further into MyD88-related regulatory mechanisms, a detailed study on MyD88 function showed that Drosophila MyD88 binds to the phosphatidylinositol 4,5-bisphosphate (PIP2)-rich regions on the plasma membrane. PIP2-guided localization of MyD88 on the membrane was shown to be essential for its function as a Toll pathway signaling adaptor and the subsequent activation of immune reactions. The authors concluded that Drosophila MyD88 serves as a sorting adaptor, and functionally it is the equivalent of the mammalian sorting adaptor TIRAP (38, 39).
Anjum et al. (40) showed that the concerted action of Drosophila β-arrestin Kurtz (Krz) and a sumo protease Ulp1 is needed to keep Toll signaling at bay in the fat body via desumoylation of Dorsal (Dl). Silencing of Krz and Ulp1 led to activation of Toll signaling and was lethal to the larvae (40). Hegde and et al. (41) show that in a sumoylation-resistant Dl mutant (DlK382R), Dl transcriptional activation is increased. This somewhat contradicts the earlier finding (40); however, Anjum et al. (40) speculated that in their study there are perturbations in the general sumoylation machinery, which may affect also other sumoylation targets besides Dl.
At the level of translational regulation of Toll pathway proteins, Wang et al. (42) provide evidence that Dicer-2, part of the RNA interference (RNAi) machinery, is involved in translation of the Toll protein by binding to the Toll mRNA 5′ untranslated region. Through this mechanism, Dicer-2 is involved in regulation of Toll pathway -mediated immune reactions.
Noncoding RNAs, including long noncoding RNAs (lncRNAs), microRNAs (miRNAs), and small interfering RNAs, have emerged as important regulatory mechanisms across a wide range of biological contexts. A limited number of recent studies have identified examples of modulation of Drosophila Toll pathway activity by both miRNAs and lncRNAs. The miRNA miR8, related to the miR200 family of miRNAs conserved in mammals, appears to downregulate the Toll pathway by interacting with mRNAs of multiple Toll pathway genes, including Toll and Dl (43, 44), with this occurring specifically in the fat body tissue of the fly (43). Other miRNAs suggested to downregulate Toll signaling by targeting various genes in the pathway are miR958 (45), miR964 (46), miR317 (47), as well as members of the miR959–962 cluster of RNAs (48). The lncRNA CR11538 (49) has been shown to bind to Dif/Dl proteins to prevent transcription of immune effector genes, and CR46018 (50) and CR33942 (51) upregulate Toll signaling through a similar mechanism. Finally, Zhang et al. (52) identified the lncRNA VINR as being involved in the immune response against both Drosophila C virus (DCV) and bacterial infections through a noncanonical activation of Toll signaling involving Cactin. However, the complete picture of how miRNAs, small interfering RNAs, and lncRNAs regulate the Toll pathway in different tissues and during immune challenge is yet to develop.
Toll pathway effector molecules
In Drosophila, the immune response against Gram-negative bacteria is primarily orchestrated by another NF-κB signaling pathway called the Imd pathway, whereas the Toll pathway has a more important role in the defense against Gram-positive bacteria and fungi (53). These responses are mediated through effector molecules. Marked progress in the past decade has been made in analyzing the Toll pathway effectors and their function. Although many of the recently characterized effector molecules had already been identified nearly 25 y ago in a mass spectrometric analysis of immune-induced molecules (IMs) (54), the modern CRISPR/Cas9 gene editing technology has now facilitated the dissection of their roles in the Drosophila immune defense. Such molecules include the Daisho peptides Daisho1 and Daisho2 (previously called IM4 and IM14), which are related peptides with partially shared functions. Daishos are needed in the defense against a group of pathogenic, filamentous fungi (55). Another recently characterized gene based on the original IM findings is Baramicin A (BaraA) (56). The BaraA gene encodes a polypeptide precursor that is cleaved into multiple peptides that correspond to one third of the originally described IMs (IMs 5, 6, 8, 10, 12, 13, 22, and 24) (54, 56). The most abundant products of the BaraA gene are IM10 and IM10-like peptides, cleavage products from the other produced IMs. These have a synergistic antifungal effect with an antifungal agent, pimaricin. Moreover, BaraA mutant flies are highly susceptible to Beauveria bassiana fungal infection, indicating that BaraA is required in the defense against fungi (56). Recently, it was shown that a Baramicin paralog, encoded from the IM24 Baramicin domain, also has nonimmune functions in the nervous system (57).
Bomanins (Boms) make up another gene family that is induced upon the Toll pathway activation (58). Some of the Boms were found in the mass spectrometric analysis and previously designated as IMs (54), whereas others were found through bioinformatics analysis (58). Ten out of the 12 Boms are found in a cluster on chromosome 2 at cytogenetic position 55C, whereas the remaining two are located on chromosome 3. Deletion of the Bom55C cluster shows that it is specifically required for the Toll-mediated response against certain bacteria and fungi (59). In another study, it was shown that Boms are the main contributors to Gram-positive bacterial and fungal resistance; the BomΔ55C mutant flies are as susceptible to infections as Toll pathway mutants, whereas mutant flies lacking 14 AMP genes that are induced upon systemic infection show a much milder phenotype (60). Bom peptides form three distinct groups, that is, short, tailed, and two-headed (or bicipital), and were renamed accordingly a few years after initial characterization (61). Furthermore, a key factor called Bombardier (Bbd), controlling expression of short-form Boms and therefore Toll pathway -mediated humoral immunity, was recently identified (62).
Furthermore, two novel peptide-encoding genes, namely Induced by Infection (IBIN) and IBIN-like, were recently identified as induced by Gram-positive bacteria Micrococcus luteus infection in Drosophila (63). It was previously thought that IBIN and IBIN-like are noncoding RNA molecules (CR44404 and CR45045, respectively), but they have been reannotated as peptide-encoding genes with strong homology to each other (63). The M. luteus -mediated induction of IBIN expression is dependent on the Toll pathway; however, IBIN can be also induced by Gram-negative bacteria, in which case the Relish/Imd pathway is required. IBIN overexpression has effects on the expression of metabolic genes, but the exact effector role of IBIN molecules is not known (63). Other studies have recently shown that IBIN is induced upon sight of parasitoid wasps (64) and social isolation (65), indicating an additional role for IBIN peptides in other stress-related situations besides infection.
In addition to Osa (24) and IBIN, the connection between the Toll pathway and metabolism was established in the gut: peptidoglycan recognition protein SA (PGRP-SA) recognizes intestinal bacteria on the surface of enterocytes, activates the intracellular Toll pathway, and thus increases the phosphorylation of 4E-BP/Thor transcription enabling fat catabolism and maintenance of the gut microbiota (66).
Sex differences in Toll pathway responses
Female and male flies differ in their response to infection, and this variation has been noted to be pathogen-specific (reviewed in Ref. 67). The Toll pathway has been shown to mediate sex-specific differences in response to both bacterial and fungal infections. Besides involvement in immunity, the Toll pathway also participates in the female-specific process that occurs in the eggs, the dorsoventral embryonic patterning. As the transmembrane receptor Toll is shared between the two processes, females have higher overall Toll expression levels due to expression in the ovaries (68). However, in various infection models, males have better survival rates and resistance compared to females (67). Duneau et al. (69) showed that in the absence of Toll signaling males were less resistant than females when challenged with Enterococcus faecalis. They also showed that males exhibit higher expression of Toll pathway effectors at the basal level and when infected with Providencia rettgeri, and that the loss of function of the psh gene abolished the sex differences. Gene expression levels of Drs and Metchnikowin during the first 24 h of infection (70), Toll-5 upon infection, and Toll-7 at the basal level (69) have been shown to be higher in males than in females. Loss of Toll has been additionally shown to affect the expression of Attacins and Diptericins in Enterobacter cloacae -infected males more than in females (71). Males also seem to have better survival rates when exposed to certain fungal infections. Shahrestani et al. (72) showed that females were more susceptible to fungal entomopathogen B. bassiana, with loss-of-function mutations of Toll pathway genes removing the sex differences in survival. Resistance to Candida albicans was also altered more strongly in males in loss-of-function mutants of Toll and Toll-7 (68). Belmonte et al. (67) speculated that the involvement of the Toll pathway in sex-specific differences in immunity may be due to the dual role of the Toll pathway in females, as the Toll pathway immune responses in females are somewhat restricted by potential consequences on egg development, a limitation that is absent in males. Although it is clear that the Toll pathway mediates sex-specific differences, the reasons for this are as yet unresolved.
Toll pathway in viral immunity
Immunity against viral infection in Drosophila appears to be largely dependent on RNAi, as well as the JAK/STAT and Imd pathways (73–76). Recent studies have presented evidence against a general major role for the Toll pathway in defense against viral infections in Drosophila in vivo, although suggesting that the pathway is involved in certain situations. In addition, several transcriptional profiling studies have shown no, or only very limited, upregulation of Toll pathway genes during viral infection of flies and/or S2 cells (77, 78). For example, Liu et al. (79) showed upregulation of the Imd pathway target gene Diptericin, but not the Toll pathway effector gene Drs, in the brains of Zika-infected Drosophila.
Limited evidence for a role for the Toll pathway has been published. Kallithea virus has been shown to suppress Toll pathway activity in the fly (80), suggesting a potential role for Toll in response to viral infection. In a separate study, invertebrate iridescent virus 6 (IIV-6) suppressed both Imd and Toll pathways (81). Separately, the Toll pathway in planthoppers was shown to be activated upon infection with a plant pathogen virus (82). In S2 cells, Flock House virus and vesicular stomatitis virus (VSV) have both been shown to trigger Drs expression (42), and while describing a transcriptional pausing mechanism for the control of virus response genes, Xu et al. (83) showed that expression of Toll, Toll-2, Toll-7, and Tollo are all upregulated during infection with VSV and Sindbis virus. The gut has been suggested as a tissue in which Toll plays a role in responses to specific viruses, for example, DCV (84), despite apparent lack of upregulation in systemic infection with this virus.
Beyond Toll itself, other Toll family members may have roles in viral immunity. Toll-7 was suggested to be involved in antiviral autophagy in two articles with somewhat contradictory results as to the role of Toll-7 and the downstream signaling pathway. Nakamoto et al. (85) found higher VSV replication in Toll-2 and Toll-7 knockdown S2 cells, and in flies with Toll-7 knockdown. Toll-7 was suggested to act as a PRR, not dependent on canonical Toll signaling through MyD88. A second study (86) supports the role of Toll-7 in an autophagy reaction against specific viral infections; however, in this case, Toll-7 signaling was suggested to use the canonical Toll signaling pathway. Lamiable et al. (87) have since shown that in their experiments, Toll-7 was not needed for resistance to VSV infection, and that autophagy only plays a limited role in this reaction. Apart from the Toll family of receptors, an lncRNA, VINR, has been shown to act as a PRR, recognizing viral suppressors of the RNAi pathway, and triggering the expression of Toll and Imd target genes (52). VINR was shown to be relevant to limiting viral replication of DCV (but not of other viruses) in S2 cells, providing further evidence for the role of Toll and Imd effectors in response, in particular, to DCV infection.
Toll pathway in blood cell homeostasis and cell-mediated immune response
The Drosophila blood cells, called hemocytes, can be classified into three main types: the macrophage-like plasmatocytes; crystal cells, central for melanization responses at wound sites and against microbes; and lamellocytes, an immune-inducible hemocyte type needed for the encapsulation and melanization response against parasitoids. Many thorough reviews on the Drosophila blood cell system and its similarities to its mammalian counterpart exist for an interested reader (for instance, Refs. 88–90). Despite the first findings on the role of Toll signaling in the formation of melanized masses via the action of lamellocytes having been made over 30 y ago (91–93), the intricacies of Toll signaling in the cellular innate immune response have been much less studied than in the humoral response. Besides lamellocyte differentiation, the Toll-induced hemocyte phenotype includes the release of hemocytes from their sessile reservoirs, as well as hemocyte hyperproliferation. Multiple studies have further elaborated on the roles of Toll signaling in the control of immune cells and on cell-mediated immune responses in the larval hematopoietic organ (the lymph gland) and in the mature hemocytes, or via signaling from other tissues, such as the fat body. Because lamellocyte differentiation occurs at the larval stages, the studies discussed below were conducted on larvae unless otherwise stated. (Fig. 2 gives a schematic summary of the findings discussed below, concentrating on the role of Toll signaling in differentiation of lamellocytes, which can be considered as a hallmark of hemocyte activation in D. melanogaster.
Qiu et al. (94) were the first to show that the Toll/Cact signaling axis is involved in the control of hematopoiesis in the lymph gland. Several papers have elaborated on the roles of Toll signaling in hematopoietic homeostasis in different compartments of the lymph gland. Gueguen et al. (95) showed that Dif and Dl are nuclear, and hence active, specifically in the posterior signaling center (PSC), which acts as a niche maintaining hemocyte progenitor cells located in the medullary zone of the lymph gland. Not only PSC-specific overexpression of either Dif or dl, but also infection by parasitoid wasp Leptopilina boulardi eliciting the cell-mediated immune response including lamellocyte formation, increased the nuclear localization of the NF-κB factors in the niche, and resulted in lamellocyte differentiation in the lymph gland (95). Louradour et al. (96) showed that larvae mutant for various Toll pathway components exhibited delayed disruption of the lymph gland, and subsequently delayed release of lamellocytes as a response to L. boulardi parasitization, leading to a reduced immune response against the parasitoids. They also showed that Toll signaling is activated in the PSC upon wasp infection via increased reactive oxygen species (ROS) production in a Psh-dependent manner, and that this activation requires Dif, but not Dl. In contrast, Dl, but not Dif, in the prohemocytes was shown to regulate the prohemocyte pool in the lymph gland medullary zone during steady-state conditions, and overexpression of dl or knockdown of cact in prohemocytes initiated their differentiation into lamellocytes (97).
Several studies have looked at the role of Toll signaling on hemocyte activation outside of the lymph gland. Schmid et al. (98) showed that although expressing the Toll gain-of-function mutant (Toll10b) in the fat body, midgut, or in mature hemocytes was sufficient to induce lamellocyte formation, Toll activation in the fat body was required for the full spectrum of the Toll-induced hemocyte phenotypes (98). They also showed that parasitization suppressed Toll activation in the fat body, but that the response against L. boulardi does not seem to require Toll, neither in the fat body nor in the hemocytes. Similarly, Yang and Hultmark (99) reported that silencing of the Toll receptor in the fat body or in hemocytes does not affect the killing of L. boulardi. However, Toll signaling has been shown to be suppressed by parasitoid wasp infection also in other insects (for example, Ref. 100), suggesting a role for Toll signaling in the cell-mediated immune response against parasitoids. To that end, Yang et al. (101) observed that pupal ectoparasitoid Pachycrepoideus vindemmiae infection induces Toll signaling as measured by Drs induction.
Schmid et al. (102) focused on the molecular underpinnings of the Toll-induced hemocyte mobilization. In their deletion screen they identified the gene immune response deficient 1 (ird1) mutant as a suppressor of this phenotype. Interestingly, other Toll-induced hemocyte traits, such as melanotic nodules and increased number of circulating hemocytes, were not suppressed, but rather enhanced in ird1 mutants. The authors showed that Toll signaling was induced in ird1 mutant larvae in the fat body, but not in hemocytes. Ird1 encodes a serine/threonine kinase important in several vesicle trafficking pathways, but it remained unclear how its loss may activate Toll signaling in the fat body. The authors suggest that the observed relocalization of the Toll receptor in ird1 mutant larvae might contribute to Toll activation. Also, when Yu and others (103) knocked down the Ras-like GTPases Rab5 and Rab11 with important roles in vesicle transport in hemocytes, Dif and Dl were localized into the nucleus and lamellocytes were formed, requiring Dif but not Dl.
The complex tissue-specific functions of Toll signaling are further highlighted in several papers discussing the link between a winged helix/forkhead transcription factor, Jumeau (Jumu), and Toll signaling. First, Zhang et al. (104) showed that simultaneous overexpression of jumu in the fat body and in hemocytes, but not in either tissue individually, led to activation of Toll signaling and formation of melanotic nodules and lamellocytes. Second, Hao and Jin (105) showed that loss of jumu throughout the lymph gland induced lamellocyte differentiation in a Dif-dependent manner. The authors noted that Jumu might regulate Toll indirectly, via the transcription factor Knot/Collier (105). In a third study, Hao et al. (106) showed Toll activation in transheterozygous jumu mutants, both in the fat body and in hemocytes. Nuclear Dif and Dl localization was accompanied by lamellocyte formation only in hemocytes, as shown by silencing of jumu tissue specifically.
As Toll signaling is responsive not only to pathogen-associated molecules, but also to various DAMPs, it has been shown to alter the hemocyte response also via these signals. Ming et al. (107) discovered that apoptosis-deficient Drosophila larvae systemically activate Toll signaling as a response to DAMPs in the hemolymph. This activation led to classical Toll-dependent effects on hemocytes: hyperproliferation and the formation of melanotic nodules, and Spz secretion from the hemocytes into the hemolymph. The systemic Toll activation as a response to DAMPs was dependent on the action of the serine protease Psh in the hemolymph (107). Arefin et al. (108) showed that apoptosis induction in non-lamellocyte hemocytes induced melanotic masses and lamellocyte differentiation, which was correlated with increased activity of Toll signaling measured as increased expression of Drs. Incidentally, Shields et al. (109) showed that in apoptosis-induced proliferation of epithelial cells, Toll-9 interacts with Toll, leading to the activation of the core Toll pathway. This results in nuclear translocation of Dl and induced expression of proapoptotic genes reaper and hid, recruitment of hemocytes, and JNK pathway activation (109). Evans et al. (110) looked at Toll signaling in the lymph gland and in circulating hemocytes, in the context of sterile wounding. They showed that injury alone was able to activate Spz in an SPE- and Grass-dependent manner in the hemolymph. Spz, in turn, activated Toll signaling in hemocytes, initiating lamellocyte differentiation via Toll-activated JNK signaling. Rather than microbe sensors, activation of Toll signaling in hemocytes required hydrogen peroxide production at the wound site (110). Chakrabarti and Visweswariah (111) similarly showed that in adult flies, a burst of ROS at the wound led to hydrogen peroxide production in hemocytes, as well as activation of Toll signaling in those hemocytes. Toll activity was required for the survival of the flies after wounding (111).
These studies emphasize the various roles of Toll signaling in the cell-mediated immune response and especially in the control of hemocyte differentiation in the lymph gland, in hemocytes, and via signals from the fat body. Recently, research on the Drosophila blood cell system has moved into the single-cell RNA sequencing era, enabling more detailed analysis of hemocytes under various conditions. The data so far have already indicated enriched expression of Toll pathway components in certain subtypes of plasmatocytes (112, 113). Further experiments focusing on transcriptomics and proteomics at the single-cell level will aid in dissecting the role of Toll signaling in detail in different hemocyte subtypes.
Toll pathway regulators and responses have been extensively studied with Drosophila, especially upon systemic bacterial and fungal infection. However, the roles of Toll in viral and parasitoid infections, as well as tissue-specific Toll pathway responses and the effect of the sex of the animal on Toll pathway activation and resulting outcomes, require further investigation. Open questions for future research include, for example, what the signals from a Toll-activated fat body to hemocytes are that result in hemocyte activation, what downstream events are affected by Toll signaling in different tissues, and how do different effectors affect immunity at the molecular level.
This work was supported by a Tampere Tuberculosis Foundation grant to S.V., by Academy of Finland Grant 32273 and Sigrid Jusélius Foundation Grant 3122800849 to T.S.S., and by Sigrid Jusélius Foundation Grant 220161 and a Tampere Tuberculosis Foundation grant to M.R. The Drosophila work was carried out in the Tampere Drosophila Facility, which is partly funded by Biocenter Finland.
Abbreviations used in this article:
damage-associated molecular pattern
Drosophila C virus
Induced by Infection
immune response deficient 1
long noncoding RNA
pattern recognition receptor
posterior signaling center
reactive oxygen species
Spz processing enzyme
vesicular stomatitis virus
The authors have no financial conflicts of interest.