Innate immune system can be activated under certain circumstances without infection by microorganisms. However, aseptic activation of innate immunity remains poorly understood. Last decade, researchers became enthusiastic to demystify how the sterile inflammation process occurs because if this process is uncontrolled, it can cause serious damage or diseases in humans. Several studies have noted that many pathological conditions or diseases related to inflammatory responses in the absence of pathogens, including myocardial infarction and stroke (1), drug induced liver injury (2), and Systemic Inflammatory Response Syndrome (SIRS) (3).

Sterile inflammation can be triggered by aseptic stimuli, such as mechanical trauma, radiation, ischemia, toxins, minerals, crystals, chemicals, autoimmune conditions, and antigens (4). Based on past findings, sterile inflammation is initiated by two factors: dead (necrotic) cells or a variety of particulates from an intracellular source or extracellularly located molecules (5). Necrotic cells lose their plasma membrane integrity after dying and release or expose their intracellular matrix containing some endogenous materials originated from cytosol, nucleus, and mitochondria, which can be recognized as pro-inflammatory molecules (1). Those factors are collectively called alarmins, damage (danger)-associated molecular patterns (DAMPs) or cell-death associated molecular patterns (CDAMPs). Inflammatory particles are able to stimulate the damaged or diseased tissue in which the particles have accumulated to a sufficient level. Those include inorganic materials (e.g., silica (6), iron oxide (7), calcium pyrophosphate (8)) and organic compounds (e.g., monosodium urate and cholesterol crystal) (8). Those particles are also regarded as alarmins that ultimately contribute to the downstream cellular and vascular manifestations of inflammation as in microorganism infections (9,10).

For the induction of sterile inflammation in mammals, inflammasomes play a central role. Inflammasomes are large multiple cytoplasmic protein complexes, which are particularly structured by Nod-Like Receptors (NLRs) as a sensor, apoptosis-associated speck-like protein containing a caspase recruitment domain or CARD (ASC) as an adaptor, and caspase-1 as a catalyst of pro-IL-1β processing (11-13). The principal function of inflammasome complexes is the cleavage of pro-IL-1β into its active form, i.e., IL-1β, as an important cytokine in inflammatory responses (14,15). Several different NLRs can compose various inflammasome complexes. For instance, the complex of NACHT, LRR, and PYD domains-containing protein 3 (NLRP3), ASC, and caspase-1 are called NLRP3 inflammasome (11-13).

Inflammasome(s) are activated by infectious stimuli known as pathogen-associated molecular patterns (PAMPs), but a number of endogenous and exogenous non-infectious substances or DAMPs can also induce their activation. NLRP3 inflammasome could be activated by various DAMPs, such as ATP, monosodium uric (MSU) crystals, cholesterol crystal, silica, and other particulate matters (14-16). After conditions caused by infection or sterile factors, the release of IL-1β is regulated by two-steps: a priming step and pro-IL-1β processing step by inflammasome(s). In a priming step, PAMPs or DAMPs may activate Toll-like receptors (TLRs) or other innate immune receptors, which induce the nuclear localization of NF-κB that transactivates gene expressions of pro-IL-1β and the components of the inflammasome. Then, inflammasomes are activated by several signaling mechanisms, resulting in the processing of pro-IL-1β into its mature form (11,16). In sterile inflammatory responses, the NLRP3 inflammasome plays a particularly important role for the cleavage of accumulated cytosolic pro-IL-1β. The activation of NLRP3 inflammasome likely collaborates with other surface pattern recognition receptors (PRRs) like TLRs or unknown upstream signaling processes that involve NF-κB induction as a priming step (16) (Figure 1). However, the molecular mechanisms of the priming step are still unclear.

Figure 1.

Model of Sterile Inflammation in Mammals. (1) Priming Step: damage-associated molecular patterns (DAMPs) activate pattern recognition receptors (PRRs), such as Toll-like receptors (TLRs), other immune receptors, and unknown receptors, leading to the production of pro-IL-1β, components of inflammasomes, and other inflammatory cytokines. (2) Processing Step: DAMPs also activate inflammasomes, which result in the cleavage of pro-caspase-1 to its active form caspase-1, which in turn, cleaves pro-IL-1β; into its active form, IL-1β. IL-1β; then binds to its receptor in other immune cells leading to the generation of inflammatory cytokines.

Figure 2.

Model of Sterile Inflammation in Drosophila. Cancer cells, impaired apoptosis, and unknown stimuli are presumed to produce DAMPs, which stimulate signaling pathways including Toll, IMD, and unknown pathways that result in the activation of NF-κB or other unknown transcription factor(s) that lead to the production of AMPs.

To precisely reveal the processes by which inflammation is activated in a sterile milieu, in particular how innate immunity is regulated and activated, many experiments must be established since it seems to be involved in multiple and complex pathways. Therefore, the utilization of animal models, such as insects and mammals, is useful. The fruit fly Drosophila melanogaster has been used as a leading model to study several molecular mechanisms of innate immunity. Drosophila has innate immune receptors to facilitate downstream signaling to activate NF-κB that drives the expression of antimicrobial peptides (AMPs) as their immune responses (17,18). Lacking the NLR gene family and thus inflammasomes, fruit flies are an excellent model to identify mechanisms involved in the priming step, as discussed above. This article will review our current knowledge and our own studies, focusing on how the sterile stimuli induce inflammation, especially the expression of AMPs in Drosophila.

Metazoan animals have many mechanisms that allow them to protect themselves against threats caused by infection, parasitism, and neoplasia (19). The study of D. melanogaster is highly relevant to mammalian innate defense mechanisms particularly in signaling processes and transduction pathways. Without an adaptive immune response, Drosophila provides a powerful model for studying the aspects of the innate immune system (20). There are two main responses in Drosophila innate immunity: the humoral and cellular systems. The humoral responses are induced in an abundant tissue, particularly in larvae, known as the fat body. The fat body can produce and secrete large amounts of AMPs into the hemolymph, while cellular responses involve defense mechanisms, such as phagocytosis and encapsulation of invading pathogens (17).

The humoral innate immunity against bacterial and fungal infection is governed largely by two NF-κB signaling pathways, Toll and IMD (19). TLRs, named after the discovery of the Toll receptor in fly immunity, are evolutionally conserved receptors required for innate immunity as well as developmental processes. Drosophila Toll receptors are composed of a protein family, similar to mammalian TLRs. The Drosophila genome encodes nine Toll receptors, some of which have been functionally characterized, and Toll-1 has an important innate immune role (21). The Toll pathway, the downstream pathway of the Toll receptor, is responsible for resistance to fungal and Gram-positive bacterial infections. Unlike vertebrate TLRs, the fly Toll receptor does not bind directly to pathogens. Instead, it is activated by the endogenous protein Spätzle (Spz), a cysteine-knot protein with structural similarities to nerve growth factor in mammals. Toll activation by the binding of Spz stimulates the recruitment of a protein complex called Myddosome, composed of dMyd88, Tube, and Pelle kinase, via E3 ligase Sherpa (22). The complex around the plasma membrane with Toll leads to destabilization of the IκB protein Cactus with uncharacterized mechanisms, resulting in the activation of the NF-κB proteins Dorsal and Dif, which are responsible for the expression of antimicrobial genes, such as Drosomycin (23).

The IMD pathway is required for defense against infection by Gram-negative bacteria in adult flies. Diaminopimelic acid (DAP)-type peptidoglycans released from Gram-negative bacteria activate peptidoglycan recognition protein LC (PGRP-LC) or PGRP-LE, the transmembrane or intracellular pattern recognition receptors, inducing the recruitment of an adaptor protein IMD. Then, the caspase-like protein Dredd cleaves the IMD protein to form a signaling complex containing the E3 ubiquitin ligase Diap2, which in turn activates TAK1 and then the IκB kinase (IKK) complex. The IKK complex then phosphorylates and activates Relish, a NF-κB to induce the expression of genes that encode antimicrobial proteins including Diptericin (17,19).

The Janus kinase/signal transducer and activator of transcription (JAK/STAT) signaling is one of the evolutionary conserved signaling pathways involved in immunity (24). In Drosophila, the JAK/STAT pathway is important for defense against viral infection, regeneration of the midgut after bacterial infection, hematopoiesis, and lamellocyte differentiation in response to a parasitic infestation (19). The ligand of a receptor for the JAK/STAT pathway consists of three proteins called unpaired (upd), upd2, and upd3. These upd proteins could bind a single receptor, Domeless (Dome), and activate JAK, hopscotch (hop), and STAT transcriptional factor, Stat92E (17,19).

The cellular immune response consists of circulating immune surveillance cells (hemocytes) in the hemolymph. In Drosophila larvae, three classes of hemocytes can be identified: plasmatocytes, crystal cells, and lamellocytes. Plasmatocytes are professional phagocytes most similar to the mammalian monocyte/macrophage lineage and comprise up to 95% of the circulating and sessile population (25). Phagocytosis is a crucial defense mechanism for invading pathogens as well as internally emerged altered-self, such as apoptotic cells. Plasmatocytes utilize several types of receptor proteins for recognizing non-self. These include members of the scavenger receptor family proteins dSR-CI and Croquemort and the EGF-domain proteins Eater and Draper (26,27). Crystal cells secrete components of the phenol oxidase cascade and are involved in the melanization of invading organisms and in wound repair. Lamellocytes are involved in the encapsulation of invading pathogens, such as wasp eggs, that are too large to be phagocytosed (17).

To date, elucidation of numerous signaling pathways responsible in the coordination of innate immune system in Drosophila has shown that conserved mechanisms exist from fruit fly to humans and such beautifully orchestrated events are critical in the maintenance of homeostasis during the life of metazoan species (17, 28). It is not surprising that loss of control in such tightly coordinated processes can cause massive havoc. This is what we are now seeing in the form of sterile inflammation. In the absence of an adaptive immune system, the fruit fly has a layer of innate immune responses as the primary defense mechanism against PAMPs and DAMPs that exist in abnormal circumstances; therefore, they are an excellent model because of the many functional and molecular similarities to mammals (17, 20).

In general, sterile inflammation occurs as a result of unexpected activation of the immune system due to the presence of aseptic triggering signals, either produced intracellularly or of extracellular origin (1, 4, 5, 20). Induction of inflammation by the sterile stimuli, may occur as a result of impairment in tight-coordinated daily routines. One example is the abrupt evolutionally conserved silent elimination of normally present yet potentially harmful apoptotic cells by phagocytes. It is apparent that successful phagocytosis will prevent the activation of inflammatory response (29). However, in the case of patients with impaired phagocytic activities, apoptotic cells are not properly processed thus leading to the activation of a necrotic mechanism where cells die and release their intracellular contents to the extracellular area (20). This in turn abrogates the delicate process of phagocytic elimination of apoptotic cells and creates a chance to activate a sterile inflammatory response. This conceptual framework has been experimentally observed in the case of apoptosis deficiency due to mutation in the caspase genes and a DNase gene.

Experiments by Ming et al. (2014) successfully demonstrated the link between apoptosis deficiency and activation of the Toll signaling pathway. In their study, it was shown that deficiency of initiator caspase Dronc led to the active processing of Spz in a Persephone-dependent manner, which further activated the canonical Toll signaling pathway (30). The physiological role of such an event was then subsequently revealed by Obata et al. (2014). In the later study, the authors demonstrated the relationship of apoptosis deficiency, dFoxO hyperactivation, and Toll activation to the upregulation of glycine N-methyltransferase (Gnmt) that leads to lipolysis and the energy-wasting phenotype in Drosophila (31). In addition to those two studies, an early study by Nagata’s group using a loss-of-function fly mutant that failed to express DNase II demonstrated that the accumulation of endogenous DNA due to DNase II deficiency leads to the expression of Diptericin and Attacin, two well-known AMPs that are under the control of the IMD pathway in Drosophila (32). This suggests that the absence of DNase II resulted in the abruption of apoptotic DNA fragmentation, one observable phenotype in cells undergoing complete apoptosis, and such an event further stimulates the activation of the humoral innate immune system in an IMD-dependent manner. However, it is interesting to note that undigested DNA-mediated sterile inflammation did not occur in a Toll-dependent manner as DNase II lacking Drosophila did not experience an increased production of Toll-dependent AMPs, such as Drosomycin (32). Subsequently, by the same group it was revealed that the activation of the IMD signaling pathway by undigested DNA was associated with the specific role of the threonine phosphatase domain of EYA in Drosophila (33). Collectively, findings reported in those studies strongly suggest that an impaired process of apoptosis could provoke an unexpected sterile inflammatory response in a Toll- or an IMD-dependent manner, supporting the pathophysiological role of DAMPs, such as undigested endogenous DNA, in the induction of sterile inflammation in Drosophila. Undoubtedly, further assessment of the mechanism(s) responsible in the induction of “impaired apoptosis”-mediated sterile inflammation might provide important pharmaceutical insights.

As described, the presence of inflammatory factors is an important trigger of noninfectious sterile inflammation in D. melanogaster. In addition to apoptosis dysfunction, it is important to note that cells with tumor/cancer-like properties play a conserved and central role in the induction of sterile inflammation across metazoan species. Thus, providing readers with a description of such entities will be an important achievement. Innate defense mechanisms are selectively reactive to foreign materials, particularly to imminent threats to host survival, such as pathogenic microorganisms and tumor/cancer cells (17, 20, 28). Like its mammalian counterpart, Drosophila performs routine and timely scans using an array of innate immune sensors and effectors to detect and if possible, eliminate cells with tumor-like properties that might become potential threats. Once such cells are discovered, a tightly controlled disposal procedure is activated (34).

Recent findings in Drosophila confirmed that the presence of long-range, inter-tissue communication between fat bodies and adipocytes plays a role in the hemocytes-mediated activation of the Toll/NF-KB signaling pathway to manage tumor growth via a non-cell autonomous mechanism in Drosophila (35). Tumor cells stimulate the activation of signaling pathways associated with hemocyte proliferation, such as Pvf1/Pvr, Spz/Toll, and Eiger/TNF (35), which have essential roles in triggering systemic immune responses that are important for the initiation of tumor cell death. This occurs in the absence of a pathogen, which recapitulates a hallmark of sterile inflammation. In addition to the stimulation of cellular innate immunity, increased expression of antimicrobial peptide genes, such as Drosomycin, Cecropin A1, and Attacin, have been reported in mutant flies with RAS^v12 (36), suggesting a bipartite link between the induction of Toll- and IMD-mediated activation of humoral immune responses and the rate of tumor progress. Although the precise mechanistic explanation to these events remains elusive, to this end, it is high likely that sterile signals elicited from cells with tumor-like properties can interact with cellular and humoral components of innate immune responses in Drosophila.

Previous studies from several groups have reported that sterile wounding in Drosophila adults induces the expression of AMPs, although the intensity of the response seems to be less than that upon infection. In contrast, Kenmoku et al. (2017) demonstrated that aseptic injury with tungsten needles strongly induced the expression of AMPs in larvae, the level of which is comparable to septic injury with bacteria (37). Intriguingly, they realized that even pinching with forceps, the standard handling method for collecting larvae, can trigger AMP expression powerfully. They noticed that AMP expression was induced not only by stimulus with forceps, but by just touching larvae with almost any material, such as toothpicks or brushes. Notably, induced genes include several AMPs, IM genes, and stress response genes, indicating that pinching by forceps can induce humoral innate immunity. They confirmed that this induction is infection independent by using germ-free larvae. Approximately 15% of pinched larvae showed small melanization spots inside their bodies, suggesting that pinching by forceps causes tissue injury. As with the regular systemic induction of AMPs, pinching-induced humoral immunity occurs at fat bodies. In sum, they revealed that larvae are surprisingly susceptible to mechanical stimuli in terms of the induction of innate immunity. The physiological significance of this phenomenon is unclear, although it may be an innately imprinted protective measure against expected infection, since larvae are frequently attached by a variety of predators, such as parasitoid wasps.

Next, they examined which signaling pathways are involved in pinching-induced Drosomycin (Drs) expression in larvae. As mentioned earlier, Drs induction is totally dependent on the Toll pathway upon systemic infection with Gram-positive bacteria. Thus, they first examined the Toll pathway by using psh, modSP, spz, dMyd88, and Dif mutant animals. Pinching-induced Drs expression is partly dependent, though the dependency is not high, on spz and dMyd88. Remarkably, the induction is totally independent from psh, modSP, and Dif. These results indicate that the canonical Toll pathway is partially involved in the mechanically induced Drs, and the protease cascade that is required for the activation of Spz upon adult infection is not engaged in larvae. Note that their preliminary experiment implies that the processing of Spz is SPE independent (data not shown). They next assessed the involvement of the IMD pathway using imd and relish mutant larvae and indicated that there was no requirement of the IMD pathway for the induction of Drs upon pinching. Furthermore, the double mutant larvae of the Toll and IMD pathways still exhibited the substantial induction of Drs, indicating that large part of the pinching-induced Drs utilizes conventional innate immune pathways in adult flies. They additionally tested the involvement of the JNK, p38, PPO, and insulin signaling by using a mutant strain of the corresponding pathways, but no mutants showed decreased expression of Drs upon pinching, suggesting the requirement of the unknown signaling mechanism for the sterile inflammation is induced by mechanical stimuli.

Though pinching by forceps is performed only on part of a larval body, Drs expression is induced in the whole fat body. This suggests that some signaling factors would be released and circulated to activate the fat body. They first wondered whether sensory neurons under the cuticle could recognize and send signals to induce Drs in the fat body. However, apparently no involvement of sensory neurons was demonstrated by their experiments. First, the induction of Drs was not affected by the removal or dysfunction of sensory neurons, using the overexpression of apoptosis-inducing genes reaper and hid, or of the dominant negative shibire. Second, artificial activation of sensory neurons by expressing the dTrpA1 ion channel without pinching did not activate the expression of Drs. They next investigated the role of hemocytes for pinching-induced Drs expression. By using hemocyte-deficient larvae, which were generated again with a hemocyte-specific overexpression of reaper and hid, they showed that no difference could be detected between wild-type and hemocyte-ablated larvae with respect to the level of Drs induction. So far, they have not succeeded in the determination of responsible cells or tissues that signal from pinching stimuli to the fat body, which is a key challenge to be tackled in future studies.

To analyze sterile inflammation of Drosophila, Kanoh et al. attempted to identify DAMPs that activate innate immune signaling. Since it is generally believed in vertebrates that tissue damage could activate an inflammatory signal immediately after the damage, they reasoned that DAMPs may already exist inside the cells in Drosophila. To test this idea, they prepared tissue extracts of Drosophila by several extraction methods and found that the larval tissue extract prepared using an organic solvent strongly potentiated the expression of Drs in DL1 cultured cells (38). They did not detect such activity in cell lines or adult flies (data not shown).

As a next step, the authors examined whether the tissue extract activity that induces Drs is already a known Toll ligand, Spz. They prepared a tissue extract from the strong hypomorphic spz mutant spz^rm7 as well as from the null mutant spz^Δ8 (data not shown), and tested Drs inducing activity. They found that those extracts possessed equal Drs-inducing activity, suggesting that the activity from the larval tissue extract is not derived from the Toll ligand Spz. In addition, they prepared the tissue extract from germ-free larvae to exclude the possibility that the Drs-inducing activity is derived from bacterial components or pathogen-associated molecular patterns. This extract also contained comparable activity to the normal tissue extract. These results collectively indicate that larval tissue extract contains novel Drs-inducing activity (38).

Next, they performed RNAi experiments to investigate whether the Toll pathway is involved in Drs induction by larval tissue extract. Inhibition of expression for all the canonical Toll pathway components, Toll, dMyd88, Tube, Pelle, and Dif by RNAi resulted in the reduction of the expression level of Drs stimulated with the extract, indicating that the Toll pathway is required for this Drs expression. Of note, knockdown of the Toll receptor also exhibited the effect, suggesting that a novel Drs-inducing activity in the extract is an unknown ligand for the Toll receptor.

The novel ligand activity for Toll is sensitive to protease treatment, suggesting that the molecular nature of the activity is composed of protein. In their following study, Nonaka et al. used the candidate gene approach to identify the novel ligand (39). They expressed six Spz family of proteins, from Spz to Spz6, in S2 cells by the standard transfection method and prepared the cell extract to test Drs-inducing activity. This experiment revealed that the extract from Spz5-expressing cells contained strong activity. RNAi experiments showed that Drs induction by Spz5 is Toll- and dMyd88-dependent. These results suggest that Spz5 could be a ligand for the Toll receptor. To prove this possibility, they generated spz5-deficient flies using the CRISPR/Cas9 technique. They successfully obtained a null mutant for spz5 and prepared the larval extracts from the mutant larvae. The Drs reporter assay with DL1 cells clearly revealed that the spz5-deficient larval extract lost the activity almost completely, indicating that Spz5 is the responsible protein of the ligand activity for Toll receptor. Then, they examined whether Spz5 could function as a DAMP in terms of sterile induction of Drs in larvae with pinching by forceps. Unfortunately, Drs induction upon pinching of larvae with the spz5 null mutant indicated that Spz5 is dispensable for pinching-induced sterile immunity. Future studies are necessary to reveal in vivo roles of Spz5, particularly its role, if any, as a DAMP.

The Spz family of proteins is known to become active by proteolytic cleavage. For example, Spz is cleaved by Easter or SPE (4041), Spz3 by Easter (42), and Spz5 by Furin proteases (43). The cleaved form of Spz functions as a Toll ligand, and cleaved Spz5 works as a neurotrophin through Toll-6 (44). Furin proteases are evolutionally conserved proteases that process pro-neurotrophins, such as brain-derived neurotrophic factor (45,46). Foldi et al. (43) showed that the Furin cleavage site at R284 in Spz5 is a functional target site and is the predominant cleavage site of Spz5. Therefore, the authors introduced mutations into Spz5 at the R284 to make a Furin-resistant form, and then, these constructs were transfected into S2 cells and cell extracts were prepared from the cells. They confirmed that wild-type Spz5 were cleaved and the mutated Spz5R284G were not. By using these forms of Spz5, they performed the Drs reporter assay and showed that the R284G mutant form of Spz5 can activate Toll signaling. These results suggest that protease cleavage of Spz5 is unnecessary for its activity as a Toll receptor ligand. This implies that Spz5 might function as a DAMP that is released by cell damage or necrotic cell death since proteolytic processing is required for secretion of Spz5 from cells. But again, this is just a possibility that should be pursued in future studies.