By Nick Puso, Biochemistry & Molecular Biology ’23

Author’s Note: Nick Puso is a biochemistry & molecular biology graduate of the class of 2023. He wrote this review because, in his own words, he “really loves vesicles”. Nick found this topic particularly exciting to write on because it combines biochemistry, structural biology, genetics, drug design, and lipidomics. He is currently applying to medical school and hopes to become an ER doctor so he can put his love of the biological sciences to good use. Academics aside, Nick loves weightlifting, cycling, backpacking, and pushing wheelchairs at the VA where he volunteers. If you’re looking for him, your best bet is to comb the Sierra Nevadas in a helicopter.

INTRODUCTION 

Flaviviruses, including ZIKA, Dengue virus, and others infect over 400 million people annually [1], causing serious neurological symptoms, hemorrhaging, and birth defects. This issue is exacerbated by the failure of past research to produce any antivirals that prove safe and effective in early clinical trials [2]. An effective antiviral would significantly reduce the global mortality and morbidity of flaviviruses, but development is long and difficult. The process of producing an antiviral relies on knowledge of a critical mechanism in a virus’s infectious biology. In the case that this knowledge applies to more than one virus, the potential exists to develop an antiviral with broad efficacy- saving essential time and money. Flaviviruses as a genus share many fundamental attributes in their infectious biology, particularly concerning their use of vesicles. These conserved features could be exploited to create antivirals with broad efficacy within the genus. The flaviviruses referred to here include Zika (ZIKV), Dengue virus (DENV), Langat virus (LGTV), West Nile virus (WNV), and Powassan virus (POWV). This review will discuss recent advancements in our understanding of the role vesicles play in the infectious biology of flaviviruses. This will include release from endosomes after entry into the cell, viral replication within the unique “vesicle packet” organelle, and a novel exosome-based mechanism of infection. Furthermore, we will explore how this knowledge could inform drug development. 

Background 

Flaviviruses are a genus of enveloped viruses, meaning that their genetic material is contained within a lipid bilayer. Embedded in this bilayer are the Envelope (E) protein (sometimes called E-glycoprotein) and Membrane (M) protein. E-Protein is responsible for attaching naked viral particles (virions) to target cell membranes, upon which they enter the cell via clathrin-mediated endocytosis [3]. The clathrin coated proteins mold a small bubble of host membrane material–an endosomal vesicle–which carries the virions into the cell.

For the viral genome to escape the endosome, and enter the cytoplasm, the viral envelope must fuse with the membrane of the endosome. Fusion is initiated by the acidic pH of endosomes, which drives a conformational shift in viral E-protein that merges the viral envelope with the endosome [3]. This mechanism is conserved in all flaviviruses. 

Flaviviruses also utilize a unique replication organelle, derived from the membrane of the endoplasmic reticulum (ER), called the vesicle packet. The vesicle packet is an invagination into the ER membrane in which the machinery of viral replication is contained, and likely protected from host cell innate-immune factors. A combination of host factors and viral proteins, including non-structural protein 1 (NS1) conserved in flaviviruses, work together to construct the vesicle. [4]. However, the specific host factors involved in vesicle packet formation were largely unknown until recently. 

Once translated and assembled, virions exit the cell via the secretory pathway, including the trans-Golgi-network (TGN). Like the endosome earlier in the life cycle, the TGN has an acidic pH. M-protein requires this acidic pH to mature. Specifically, prM, the precursor to M-protein, undergoes an acid-driven conformational shift which exposes cleavage sites for Furin-protease, a host protease [5]. Cleavage releases mature M-protein from ‘prM’. The ‘pr’ element itself is responsible for preventing premature fusion of the virion envelope with host membranes, so it remains associated with the virion until secretion [5]. The dissociation of ‘pr’ after cleavage allows the mature virion to later undergo acid-driven fusion with endosomes in a new cell, completing the cycle [5]. 

Endosomal release and niclosamide’s antiviral potential 

A compound that safely inhibits endosomal and secretory pathway acidification could represent a pan-flavivirus antiviral. This is because attenuation of acidic pH would prevent the conformational shift in E-protein necessary for the fusion of the virion with the endosome. Recently, researchers determined that niclosamide–an FDA-approved antiparasitic drug previously known to inhibit ZIKV infection–did exactly this in DENV-infected cells [6]. Niclosamide is a protonophore;a compound that can exchange protons across membranes, bypassing the proton pumps responsible for endosomal acidification. Treatment with niclosamide significantly reduced viral load in baby-hamster-kidney (BHK21) cultures infected with DENV. Using pH-sensitive dye A0, which turns from clear to pink at pH 6, the mechanism was confirmed to be attenuation of endosomal acidification. Additionally, a western blot of DENV protein-E showed that the acid-driven confirmation shift in protein-E necessary for endosomal fusion did not take place. This western blot relied on the fact that E-protein’s conformational shift results in its degradation. An antibody could visualize E-protein in the SDS-page (electrophoresis which separates proteins by mass) of the treatment group, meaning the degradation did not take place. This indicates that the conformational shift was blocked by niclosamide as hypothesized. Because no effects were noted on viral genome replication or endocytosis independently of endosomal release, the authors do not put forward any additional mechanisms of action. 

However, another study found that niclosamide also prevents the maturation of DENV in human (Huh7) cells via deacidification of the TGN [7]. Researchers determined that deacidification inhibited cleavage of DENV/ZIKV prM protein by preventing the exposure of furan-protease cleavage sites. Compared to a control, Western blot of the treatment grouprevealed uncleaved prM when niclosamide was present. Virions with immature prM protein due to niclosamide treatment would be unable to fuse with endosomes upon infection of a new cell and are thus dead. The authors suggest discrepancies from Kao et al [6] arrose because they used different time points in their analysis of niclosamide’s effect on genome replication and virion maturation. 

Niclosamide’s ability to de-acidify endosomes and the TGNdemonstrates its potential as a pan-flavivirus antiviral, given that E-protein and prM are conserved and essential in all flaviviruses. Both authors argue that niclosamide’s safety as an antiparasitic drug and its effect on ZIKA and DENV warrant investigation into antiviral applications. Additionally, both studies put forward inhibition of endosomal release as a viable strategy in anti-flavivirus drug development overall. Jung et al also recommend inhibition of prM cleavage for therapeutic investigation.

Targeting host factors in ‘Vesicle packet’ formation 

Replication inside the vesicle packet is another potential target for pan-flavivirus antiviral development because it is conserved in the genus. Knowledge of the involved host factors–proteins in the host that the virus manipulates to its own ends–could allow for antivirals that knockout vesicle packet formation, inhibiting viral replication. Recently, a class of host ER-shaping proteins called Atlastins were found to play a role in the formation of the vesicle packet [8]. Researchers used small-hairpin RNAs (shRNAs), which are processed into silencing siRNA to knock out various atlastins in A529 cells exposed to ZIVK, DENV, and WNV. Transmission electron microscopy, which uses electrons to visualize microscopic structures, revealed that Atlastin-2 (ATL2) knockout shrunk both the size and number of vesicle packets, and disrupted their localization. Immunofluorescence, which visualizes the location of small molecules using a fluorescent antibody, then showed that the localization of viral double-stranded RNA (dsRNA) was changed. This indicates that the coordination of vesicle packets with the replicating viral genome was disrupted. Also, the overall viral load of ZIKV, DENV, and WNV was reduced by Atlastin-2 knockout. In the case of ZIKV, this reduction was potentially as much as 16-fold. While the authors make no specific recommendations, future research should explore whether these factors inhibit flavivirus replication outside of DENV, ZIKV, and WNV. Given that all flaviviruses form a vesicle packet for replication, the host factors they manipulate to do so may be the same. If this is the case, disruption of flavivirus/Atlastin interaction may be a powerful therapeutic approach. 

A host protein called Receptor of Activated C Kinase (RACK1)has also been identified as essential to vesicle packet formation for flaviviruses WNV, DENV, POWV, and LGTV [9]. Researchers used a similar approach to Neufeldt et al, employing small ‘interfering’ RNAs (siRNAs) to silence host factors and measure the effect on viral load. Specifically, a CRISPR-based genome-wide knockout screen using LentiCRISPRv2-GecKO (an siRNA library) provided siRNAs to human Huh7.5 cells. RACK1 knockout produced the strongest reduction in viral load out of all targets. To determine the mechanism, researchers analyzed the time dependency of viral load reduction on RACK1 knockout. The results indicated that RACK1 did not act on viral entry or translation, but rather on replication, which brought attention to the vesicle packet. Transmission electron microscopy then demonstrated a significant reduction in vesicle packets after RACK1 knockout. Immunofluorescence found that viral NS1, a known critical factor in vesicle packet formation, failed to localize to the ER. The failure of NS1 to localize to the ER during RACK1 knockout would suggest that an interaction between NS1 and RACK1 is necessary for vesicle packet formation. Co-expression of RACK1 and NS1 gave strong evidence for their co-localization. To confirm the interaction, a pulldown assay- which measures binding affinity between two molecules- was performed and RACK1 was proven to bind NS1. The authors thus posit that RACK1 is responsible for localizing NS1 and that RACK1 knockout disrupts vesicle packet formation by the failure of NS1 to localize. 

The discovery of ATL2 and RACK1 as host factors necessary to the formation of the vesicle packet offers new targets for disrupting flavivirus replication. The respective authors argue that dependence on RACK1 and Atlastin-2 for vesicle packet formation is likely conserved in all flaviviruses. Shue et al thus argue that RACK1 is a promising target for antiviral development, and recommend further research into RACK1 knockout as a therapeutic strategy. ATL2’s importance to vesicle packet formation warrants exploration into its potential as a pan-flavivirus antiviral target. 

Exosome-based infection: SMase and Tsp29Fb as drug targets 

Recently, three studies have identified a novel exosome-based mechanism of flavivirus infection and identified potential targets for antiviral development. Zhou et al, (2018) found that LGTV uses extracellular vesicles (exosomes) from its tick host to infect mammalian cells, and that infected mammalian brain-endothelial cells, which make up the blood-brain barrier, produce exosomes capable of infecting neuronal cells [10]. In a follow-up to their 2018 study, Zhou et al (2019) found that ZIKA-infected, neuronal-cell-derived exosomes readily infected neurons of the cortex [11]. A third study, Vora et al, demonstrated that exosomes derived from DENV-infected mosquitoes were infectious to human blood-endothelial cells [12]. The ability of flaviviruses to use exosomes, derived from host cells, to infect other cells was previously unknown and represents a significant advancement in our understanding of their infectious biology as a genus. 

Additionally, GW4869, an exosome release inhibitor, proved to reduce viral load in cell cultures infected with their respective flavivirus in all cases. GW4869 works by inhibiting sphingomyelinase (SMase), an enzyme 

that produces lipids necessary for exosome formation. The effect of GW4869 on viral load confirms that exosomes provide significant infectious potential, in a novel mechanism, for 3 prominent flaviviruses: LGTV, ZIKV, and DENV. GW4869’s ability to inhibit this mechanism in all three suggests that an antiviral targeting exosome release may have efficacy across the genus. All three studies thus cite GW4869 as a promising target for drug development to disrupt flavivirus infection via exosomes. GW4869’s efficacy also suggests the anti-flaviviral potential of SMase inhibition more broadly. 

All three studies used immunoblotting and quantitative real-time PCR to show the enrichment of viral RNA and proteins in infectious exosomes.

All three also confirmed that these viral components are fully inside the exosomes. To do this, they exposed these exosomes to RNase and flavivirus envelope protein antibodies, which neutralize the infectious potential of naked virions by degrading exposed viral RNA and protein. The infectious potential of exosomes was unaffected by RNase or antibody treatment. All three studies thus argue that the exosome protects viral RNA and protein from antibodies and RNase. 

Vora et al also found that Tsp29Fb–an ortholog of high similarity to human exosome marker CD63–was enriched on infected extracellular vesicles [12]. Specifically, qRT-PCR, which tracks the expression of RNA over time, found that Tsp29Fb was overexpressed during infection. Additionally, the CD63 antibody (which is highly cross-reactive with Tsp29Fb) was used in immunoblotting to show that Tsp29Fb is enriched on the exosomes. Researchers then used immunoprecipitation, which removes a target from solution using an antibody, to find that Tsp29Fb binds DENV envelope protein. This binding interaction indicates Tsp29Fb may play a role in DENV’s ability to use exosomes for infection. To test this, researchers silenced Tsp29Fb with siRNA and found drastically reduced viral protein in the resulting exosomes. The researchers posit that Tsp29Fb thus plays a role in loading protein into secretory vesicles post-translation, which then exit the cell as infectious exosomes. For that reason, the authors put Tsp29Fb silencing forward as an antiviral development tactic.

These three studies demonstrate that extracellular vesicles provide infectious potential to three flaviviruses. Additionally, two promising candidates for antiviral research have been put forward to disrupt viral exosomes. Namely, GW4869 targeting SMase, and Tsp29Fb as a target itself. 

CONCLUSION

Flaviviruses as a genus have great similarities in their infectious biology concerning their manipulation of host membranes and vesicles. Post-entry endosomal release and the formation of the vesicle packet are shared features of the flavivirus infection cycle. In a completely new infectious mechanism, LGTV, ZIKV, and DENV were all found to invade new cells using exosomes. All three discussed uses of vesicles have been proven essential to infection and are present in multiple flaviviruses. Furthermore, workable targets for the inhibition of these faculties have been identified, which subsequently reduce viral load in vitro. Multiple promising candidates for drug development have thus emerged from these targets. Future research should focus on whether an exosome-based infection is present in other flaviviruses and establish the relevance of SMase and Tsp29Fb to the broader genus. If exosome-based infection proves critical to flaviviruses more broadly, SMase may be a viable target, and GW4869 might have potential as a pan-flavivirus antiviral. With respect to targeting replication in the vesicle packet, RACK1 and ATL2 knockout should be investigated for safety and efficacy. Niclosamide’s ability to inhibit post-entry edosomal release must also be established in the genus as a whole, and the therapeutic dose already approved by the FDA for other conditions should be tested for efficacy against flavivirus.

REFERENCES

1. Pierson, T.C., Diamond, M.S. The continued threat of emerging flaviviruses. Nat Microbiol 5, 796–812 (2020). https://doi.org/10.1038/s41564-020-0714-0
2. Qian, Qi. 2022. Mosquito-Borne Flaviviruses and Current Therapeutic Advances. Viruses. 2022, 14(6), 1226; https://doi.org/10.3390/v14061226
3. Modis Y, Nayak V. Molecular Mechanisms of Flaviviral Membrane Fusion. J Virol. 2009:265–86. doi: https://doi.org/10.1007%2F978-0-387-79840-0_12 
4. Chatel-Chait, Bartesenschlager. 2014. Dengue Virus- and Hepatitis C Virus-Induced Replication and Assembly Compartments: the Enemy Inside—Caught in the Web. Virology. 2014; 88(11). doi: https://doi.org/10.1128/jvi.03404-13
5. Yu et al. 2009. Association of the pr Peptides with Dengue Virus at Acidic pH Blocks Membrane Fusion. Virology. 2009 Dec; 83(23): 12101–12107. doi: 10.1128/JVI.01637-09 
6. Kao JC et al. 2018. The antiparasitic drug niclosamide inhibits dengue virus infection by interfering with endosomal acidification independent of mTOR. PLoS Negl Trop Dis. 2018 Aug 20;12(8):e0006715. doi: https://doi.org/10.1371%2Fjournal.pntd.0006715 
7. Jung, E. et al. 2019. Neutralization of Acidic Intracellular Vesicles by Niclosamide Inhibits Multiple Steps of the Dengue Virus Life Cycle In Vitro. Sci Rep 9, 8682 (2019). https://doi.org/10.1038/s41598-019-45095-1
8. Neufeldt et al. 2019. ER-Shaping Atlastin proteins act as central hubs to promote flavivirus replication and virion assembly. Nat Microbiol. 2019 Dec; 4(12): 2416–2429. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6881184/
9. Shue et al. 2021. Genome-wide CRISPR screen identifies RACK1 as a critical host factor for flavivirus replication. Virology. 2021; 95(24). doi: https://doi.org/10.1128/JVI.00596-21
10. Zhou et al. 2018. Exosomes serve as novel modes of tick-borne flavivirus transmission from arthropod to human cells and facilitates dissemination of viral RNA and proteins to the vertebrate neuronal cells. PLoS Pathog. 2018 Jan; 14(1): e1006764. doi: 10.1371/journal.ppat.1006764 
11. Zhou et al. 2019. Exosomes mediate ZIKA virus transmission through SMPD3 neutral sphingomyelinase in cortical neurons. Emerging Microbes and Infections. PLoS Pathog.2019; 8(1): 307–326. doi:1080/22221751.2019.1578188
12. Vora et al. 2018. Arthropod EVs mediate dengue virus transmission through interaction with a tetraspanin domain containing glycoprotein Tsp29Fb. Proc Natl Acad Sci USA. 2018 Jul 10; 115(28): E6604–E6613.https://doi.org/10.1073%2Fpnas.1720125115