Using the fruit fly to take on Zika virus

Zika virus (ZIKV), a member of the Flaviviridae family, preferentially targets and damages neural stem cells and progenitor cells. Therefore, ZIKV is particularly dangerous during fetal development as infection can lead to congenital brain abnormalities.

In adults Zika infection is generally asymptomatic; however recent reports have identified few cases of severe complications associated with the central nervous system, including encephalitis. Whether there are intracellular host pathways that control Zika infections in neuronal cells is still largely unknown. In the News and Views section of Nature Microbiology, Carolyn B. Coyne summarizes how recent work carried out in Drosophila has proposed autophagy as an innate antiviral defense mechanism against ZIKV in neuronal cells.

Drosophila as a model for the study of innate immunity and viral infections

Drosophila melanogaster has been used as a model system to elucidate the molecular mechanisms involved in innate immunity and infections. Also, an interesting point is that a stage of ZIKV life cycle, as well as of other neurotropic and non-neurotropic flaviviruses, includes a stage of replication in mosquitos. As in higher organisms, infection of Drosophila with various microbes and viruses leads to the activation of the transcription factor NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells), a pivotal mediator of multiple innate and adaptive immune functions, including the expression of pro-inflammatory genes (cytokines, chemokines, and adhesion molecules). However, the downstream targets and signaling pathways responsible for ultimately clearing infections are often unknown.

dSTRING protects the fly brain against Zika

In their recent publication, Liu et al. made use of the genetically tractable Drosophila system to investigate the signaling pathways that control ZIKV infection in the brain. These authors found that NF-kB activates the expression of Drosophila stimulator of interferon genes (dSTING) to induce antiviral autophagy, thereby controlling ZIKV infection in the brain. Below we take a close look at this study.

Zika targets the fly brain

In establishing their fly model of Zika infection, Liu et al. use a combination of molecular biology (qRT-PCR), virology (TCID50 in Vero cells), and microscopy (confocal) approaches to show that Zika targets and replicates in the Drosophila brain, more than in any other tissue.

NF-kB signaling (but not RNAi) fights Zika in the fly brain

Next, Liu et al. set out to identify the innate immune responses active against Zika infection in the fly brain. RNAi has been identified as an antiviral pathway in the Drosophila brain and seemed a likely candidate for a fly anti-Zika response. However, when the team measured ZIKV levels in mutant flies missing key genes of the RNAi pathway (Dcr-2 and Ago2), they detected no change in viral replication. This finding suggests that ZIKV infection of Drosophila is not inhibited by RNAi, either systemically or in the brain. Does ZIKV encode a suppressor of the antiviral RNAi pathway? This would be interesting to follow-up as it may reveal a mechanism of ZIKV evasion of immunity in arthropods.

Liu et al. next tested the inflammatory NF-kB pathways TOLL and IMD as potential anti-Zika mechanisms in the fly brain. Indeed, infection of Drosophila with ZIKV did induce the IMD NF-kB pathway (but not the TOLL pathway), which was dependent on the transcription factor Relish; ZIKV ran riot in Relish mutant flies (more virus and more lethality), also when the gene was selectively knocked down in neurons and glia.

What happens downstream of Relish?

To determine the downstream Relish-dependent genes that control viral infection, Liu et al. performed transcriptional profiling of flies that were either uninfected or infected with the model virus Drosophila C (DCV), a picornavirus that naturally infects Drosophila.

The authors identified 145 genes induced by the viral infection, many of which – unsurprisingly – were canonical effectors associated with NF-κB signaling. Perhaps more interestingly, Liu et al. also detected induction of the of the Drosophila homolog of STING (dSTING) in the virus-infected flies. STING is a highly evolutionarily conserved molecule that responds to both exogenous and endogenous ligands (cyclic dinucleotides) to trigger the innate immune system, and has been previously identified as virally induced. Using Relish mutant flies, the team also showed that ZIKV induction of dSTRING is Relish-dependent as dSTRING is not induced in the absence of Relish.

What does dSTRING do to protect neurons?

Lui et al. next reasoned that dSTRING could be inducing autophagy (a natural, regulated mechanism that disassembles unnecessary or dysfunctional components) as an antiviral effector mechanism. There were good reasons for this thinking: STING has been shown to induce autophagy to resist intracellular pathogens; autophagy protects neurons from diverse stressors (including viral infection); and autophagy can clear pathogens without causing cell death (useful for mature neurons!).

Their hunch turned out to be good. Using elevated Atg8-II levels as a marker for autophagy, Lui et al. found that ZIKV infection activates autophagy in the fly brain. Using a series of mutant flies, the authors go on to demonstrate that autophagy is essential for the control of ZIKV and reduces lethality.

Moreover, inducing autophagy (by feeding flies rapamycin) before infection protected the flies from ZIKV. If this pathway was confirmed in mammals, could pharmacological activation of autophagy be part of a therapeutic strategy against ZIKV? Perhaps, although the authors caution that this may not be a tenable approach given that Zika infection of non-brain tissues might be exacerbated by the drug.

Outlook

This study reveals some interesting new concepts relative to the innate immune response that arthropods mount against pathogens, and particularly the role of dSTING in activating autophagy to restrict viral replication in the brain. However, its relevance to higher species remains to be assessed.

While STING is highly conserved across genera, its activation and its downstrean effectors are different between different species. In particular, in mammalian cells STING is activated directly by pathogen associated molecular patterns (PAMPs), or indirectly by cyclic dinucleotides, and its primary dowstream effect is the activation of an interferon response. In the fly, the main activator of STING seems to be the NF-kB pathway, and this work (and others) suggest that the main downstream effect is activation of authophagy. While activation of autophagy has been described downstream of STING also in mammalian cells, its relevance in innate immunity and pathogen defence is not entirely clear, but likely not to be the major antiviral component.

Another outstanding question is the impact of STING in the fly antiviral defences, as viral replication seems to occur even when the pathways is active, although to a lower level.

Infection of the fly brain is also an interesting point, as flaviviruses tend to primarily infect and replicate in the gut and salivary glands. Whether other neurotropic flaviviruses infect primarily the fly brain is unclear. Finally, while ZIKV is known to infect neural progenitor cells, infection of mature neurons is controversial and the very limited cases of encephalitis in humans seems to confirm this notion.

This work provides important information on the antiviral activity of STING, a highly evolutionary conserved component of innate immunity, in arthropods (which are the vehicle of many flaviviruses). It also suggests a possible ancestral mechanism of STING-mediated activation of autophagy, needed for antiviral defence, particularly in tissues that do not regenerate. The importance of this pathway in mammalin systems and particuarly in neurons deserves investigation: it would be interesting to know whether this mechanism is what makes differentiated neurons resistant to ZIKV, as well as its impact in other neurotropic flaviviruses.