University of Groningen Virus:host interactions during chikungunya virus infection Bouma, Ellen Marleen

(1)

University of Groningen

Virus:host interactions during chikungunya virus infection

Bouma, Ellen Marleen

DOI:

10.33612/diss.171018969

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2021

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Bouma, E. M. (2021). Virus:host interactions during chikungunya virus infection: Analyzing host cell factors and antiviral strategies. University of Groningen. https://doi.org/10.33612/diss.171018969

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

560817-L-bw-Bouma 560817-L-bw-Bouma 560817-L-bw-Bouma 560817-L-bw-Bouma Processed on: 7-6-2021 Processed on: 7-6-2021 Processed on: 7-6-2021

(3)

Processed on: 7-6-2021 PDF page: 176PDF page: 176PDF page: 176PDF page: 176 176

7

Summarizing Discussion

Chikungunya virus (CHIKV) is a mosquito-borne virus that causes major outbreaks in tropical and subtropical regions of the world. In Chapter 1 of this thesis, I describe that an infection with CHIKV can lead to debilitating, long-lasting disease symptoms and has a substantial impact on human health. Despite this, there are no antiviral therapies or vaccines available. So far, a few vaccine candidates for CHIKV are in Phase 1 or Phase 2 clinical trials1–3_{. A recent study with a live-attenuated}

CHIKV vaccine showed promising results in a Phase 1 clinical trial, with high-titer neutralizing antibodies up to 1 year after a single-dose administration of the vaccine candidate4_{. Currently the vaccine candidate is being tested in a Phase 2 clinical trial}

and preparations are ongoing for the initiation of a Phase 3 clinical trial in the near future5_{. In addition to vaccines, antiviral therapeutics are needed to treat}

CHIKV-infected individuals with a weakened immune system like immunocompromised patients and elderly and non-vaccinated individuals.

CHIKV is a positive-sense single-stranded enveloped RNA virus of ~70nm in diameter6,7_{. To infect a target cell, the virus particle binds to host cell receptors}

after which the virus is internalized via clathrin-mediated endocytosis (CME) and delivered to early endosomes. The acidic environment of the early endosomes subsequently facilitates membrane fusion of the viral envelope with the endosomal membrane via a hemifusion intermediate. After fusion pore formation, the viral nucleocapsid containing the gRNA is released in the cytosol. The process from virus cell binding to gRNA release into the cell cytosol is referred to as virus cell entry and is a prerequisite for infection (for a detailed description of the cell entry pathway of CHIKV I would like to refer to Chapter 1 of this thesis). It also represents an interesting target for antiviral drug development, as you can halt CHIKV infection before the intracellular replication cycle has started. In this way, treatment limits the harmful consequences of viral replication and the immune pathways that are activated due to viral replication.

In this thesis, we investigated the molecular mechanisms involved in the cell entry pathway of CHIKV. Emphasis was on identifying new cellular factors that are important for CHIKV infection and to explore whether these might serve as targets for antiviral therapeutics. In this chapter, I will provide a brief summary of the research described in this thesis (Part 1) and I will discuss my results in the context of recent literature (Part 2) to provide future perspectives on the cellular factors required for CHIKV infection that could be targeted for anti-CHIKV drug development.

(4)

Part 1 - Summary

Cell entry can be studied using a diverse range of methods and technologies including life cell imaging. In this thesis, we visualized fluorescently-labeled CHIKV particles to better understand the virus:host interactions that occur during CHIKV cell entry till the moment of membrane fusion8,9_{. To visualize single CHIKV particles,}

we labeled the membrane of CHIKV particles with the lipophilic fluorescent dye DiD at self-quenching conditions. When the viral outer membrane merges with the endosomal membrane upon membrane fusion, the fluorescent dye dilutes in the endosomal membrane which can be observed as a sudden increase in fluorescent intensity due to dequenching of the probe. Using these DiD-labeld CHIKV particles we can perform single-particle tracking assays, which allow us to follow the entry of single virions in live cells. Additionally, we can use the labeling principle to quantify membrane fusion events using a ‘bulk’ fusion assay where we measure the fluorescent intensity in a larger population of fixed cells. Important to realize for both assays is that dequenching is achieved during the transient hemifusion stage, just before fusion pore formation (See Fig. 5, Chapter 1). Thus, the observed fluorescent signal is not indicative for a complete membrane fusion event but rather indicates the first lipid-mixing step due to hemifusion.

In Chapter 2 we explored the cellular internalization pathway of Mxra8, to identify if Mxra8 is used as an attachment factors or rather as a cell entry receptor. Mxra8 has been identified as a cell surface receptor required for CHIKV infection10_{. We showed}

that Mxra8 is located in acidic compartments, including early endosomes positive for EEA1 (Early Endosome Antigen 1) and lysosomes positive for Lamp1 (Lysosomal-associated membrane protein 1). Importantly, the transmembrane and cytoplasmic domains of Mxra8 are not required for Mxra8 internalization and are not essential for CHIKV infection. This might suggest that Mxra8 internalization is triggered by other interaction partners. In the above outlined bulk membrane fusion assay, we observed a strong reduction in membrane fusion events in Mxra8 knock-out cells compared to control cells, yet this might also be due to impaired binding at the cell surface. Single-particle tracking of CHIKV in Mxra8-GFP expressing cells revealed that half of the CHIKV virions colocalize with Mxra8 at the moment of membrane fusion. This data shows that Mxra8 may indeed function as a cell entry receptor and further research should delineate the potential role of Mxra8 in CHIKV membrane fusion.

(5)

7

Next to Mxra8, microtubules play an important role in the cell entry process of viruses. In Chapter 3 we explored the interaction of CHIKV with the microtubule network during cell entry and infection. Single-particle tracking revealed that the intracellular trafficking behavior of CHIKV particles can be divided into two groups: half of the particles exhibit fast-directed movement prior to membrane fusion, whereas the other half of the particles remain static till the moment of membrane fusion. Subsequent experiments showed that fast-directed movement occurs after CME but before entering Rab5-positive early endosomes. Fast-directed movements are commonly attributed to microtubule-directed transport. Indeed, upon disruption of the microtubule network using the chemical compound nocodazole, we observed a distorted trafficking of CHIKV particles. At these conditions, most particles remained static and fused primarily in the periphery of the cell. While membrane fusion was not inhibited, we did observe a reduced release of gRNA in the cell cytosol and a lower number of productively infected cells. In conclusion, we observed two distinct intracellular trafficking behavior of CHIKV particles, and we found that microtubules are important for infection. Collectively, this indicates that microtubules are required for CHIKV cell entry, which will be further discussed in part 2.

In Chapter 4 we studied the anti-CHIKV effect of serotonergic drugs that bind and/or block serotonin (5-HT) receptors. 5-HT is a neurotransmitter present in the central nervous system and the intestine where it is responsible for various physiological functions including mood, memory, and appetite11,12_{. 5-HT receptors}

are expressed on a large variety of cells and are known to be important for the cell entry of multiple viruses13–15_{. In this chapter we studied the role of serotonergic drugs}

on CHIKV infections in the human bone osteosarcoma cell line U-2 OS. We first analyzed the gene expression of the different 5-HT receptor subtypes within these cells to confirm the possibility of serotonergic engagement by serotonergic drugs. We found that at least 8 different 5-HT receptors are being expressed in U-2 OS cells. Next, we studied the activity of 5-HT receptor agonist, 5-nonyloxytryptamine (5-NT) towards CHIKV infection. Binding of 5-NT to 5-HT receptors leads to complex internalization and activation of the downstream signaling pathways. 5-NT exerted strong antiviral activity against CHIKV: the number of cells infected by CHIKV was reduced by more than 80%. To understand the antiviral mechanism of 5-NT we next determined at which step in the virus replication cycle it exploited its antiviral activity. Intriguingly, while serotonergic drugs target cellular receptors, we did not observe an altered CHIKV cell binding. Moreover, when further dissecting the cell entry pathway, we observed that 5-NT did not influence CHIKV membrane fusion

(6)

activity nor the release of viral gRNA (genomic RNA) into the cell cytosol. Yet, cell entry bypass experiments revealed that 5-NT exhibits its antiviral activity before gRNA entry in the cytoplasm, suggesting that 5-NT inhibits CHIKV cell entry after nucleocapsid release, but before initial RNA translation and/or replication. We additionally assessed CHIKV infection in the presence of a 5-HT receptor antagonist, Methiothepin Mesylate (MM), which in contrast to agonist, binds and blocks the receptors’ underlying signaling pathways. To our surprise, MM did not increase viral infection, but strongly inhibited CHIKV infection. Interestingly, MM specifically reduced the membrane fusion activity of CHIKV particles and consequently also a reduced number of viral RNA copies in the cytosol. In conclusion, we show that 5-NT and MM act differently on the CHIKV replication cycle; MM inhibit CHIKV at, or just before, the membrane fusion step while 5-NT inhibits CHIKV after nucleocapsid release in the cytoplasm. These intriguing findings display the possibilities for serotonergic drugs as antiviral therapies against early steps in CHIKV infection. In Chapter 5 we studied the importance of heat shock proteins 70kDa (Hsp70) by use of Hsp70 inhibitors. The Hsp70 machinery regulates and assists efficient protein folding, protein degradation and protein-protein interactions16–18_{. In stressed}

conditions, including several viral infections, Hsp70 assists in protein folding of key viral components19_{. We show that Hsp70 inhibitors exert strong antiviral activity}

against CHIKV. We observed only a limited effect on CHIKV cell entry; the strongest antiviral effect was seen at post-entry stages of the virus replication cycle. Subsequent analysis of the mode-of-action revealed that the inhibition of the Hsp70 machinery affects the intracellular expression of multiple viral proteins whereas no effect on the number of intracellular viral RNA copies was seen. Furthermore, the transport of the structural protein E1 to the plasma membrane was not compromised. Irrespective of efficient transport of E1 to the plasma membrane, all Hsp70 inhibitors drastically reduced the number of secreted progeny virions. Furthermore, the Hsp70 inhibitor JG-98 strongly affected the infectious properties of newly produced virions. Taken together, although more research is required the results presented in this study clearly indicate an important role for Hsp70 in the late stages of the CHIKV replication cycle and provides new possibilities for intervention in the CHIKV replication cycle. In Chapter 6 we studied the inhibition of CHIKV infection using neutralizing monoclonal antibodies (mAb). The mAb CHK-152 has been isolated from CHIKV-infected mice and was previously shown to protect against CHIKV infection, both in mouse and non-human primate models20,21_{. This antibody binds to the}

(7)

7

fusion20,22_{. The acid-sensitive region is responsible for the destabilization of the}

E2-E1 heterodimer thereby allowing the formation of the fusion-active E2-E1 homotrimer. In this study, we found that CHK-152 interacts with CHIKV under neutral and low pH and shields the virions from binding to lipid bilayers. We also revealed that CHK-152 antibodies prevent E1 homotrimer formation under acidic conditions. Using a single-particle fusion assay we assessed the fusion capacity of single virus particles pre-docked to a lipid-bilayer. We observed that CHIKV fusion is slowed down and blocked in presence of CHK-152. Importantly, CHK-152 dissociated from CHIKV virions at low pH thereby reducing the relative inhibition of membrane fusion. The CHK-152 dissociation enabled us to numerically model the process of cooperative fusion based on CHK-152-free spikes. We show that most likely three to five neighboring trimers are needed for fusion and that membrane fusion can be blocked at sub-stoichiometric antibody binding conditions.

Overall, this thesis aimed to study important virus:host interactions in the CHIKV replication cycle. Using advanced microscopy techniques and antiviral agents, we provided insights in the molecular events required for CHIKV infection. The relevance of these findings for 1) the identification of novel cellular players in the virus replication cycle and 2) the implications for the rational design of antiviral therapeutics will be discussed in the following part of this chapter.

Part 2 - Discussion and Future Perspectives

Chikungunya virus and its interaction with cell surface molecules

Viruses are obligatory parasites and depend on a host cell for the production of progeny virions. The first step in infection involves the interaction of a virus particle with cell surface molecules. These can function as attachment factors thereby playing an important role in the initial attachment of the virus to its target cell or as virus receptors to prime both the virus as well as the target cell to initiate endocytosis and signaling pathways23_{. The viral envelope protein E2 is involved in virus cell binding.}

More specifically, the putative receptor binding domain is located between domain A and domain B of the E2 glycoprotein22_{. While many CHIKV attachment factors have}

been described23,24_{, it was only recently postulated that Mxra8 could function as an}

entry receptor10_{. Mxra8, also known as DICAM, ASP3 or limitrin, is a cell-adhesion}

molecule (CAM) expressed on many different mammalian cell types10_{. It is a type I}

transmembrane protein within the immunoglobulin superfamily and is located in intercellular tight junctions. Soon after the discovery of Mxra8 as a receptor for

(8)

CHIKV, the molecular basis of this interaction was shown by X-ray crystallography and cryo-EM reconstructions25,26_{. The high-resolution structural information showed}

that the extracellular domain of Mxra8 contains two immunoglobin folds which are involved in binding of two E2-E1 heterodimers in one trimeric spike and one heterodimer in a neighboring spike25,26_{. The interactions between Mxra8 and the}

CHIKV glycoproteins suggests that Mxra8 could, besides facilitating attachment, also function as a cell entry receptor, thereby facilitating internalization and potentially also influencing membrane fusion in endosomes10,26–30_.

In Chapter 2, we therefore studied the interaction of the Mxra8 receptor with CHIKV particles in more detail. We showed that Mxra8 is present in early endosomes and lysosomes, indicating that this CAM is targeted towards the same endocytic pathway as CHIKV virions8_{. This endosomal targeting of cell surface molecules is often}

regulated via covalent modifications, like phosphorylation and polyubiquitination, in the cytoplasmic tail of the receptor31_{. In line with this, viruses, including dengue}

virus (DENV) and influenza A virus, often employ covalently modified receptors to be targeted towards the endosomal pathway32,33_{. Here, we verified whether covalent}

modifications of Mxra8 are important for the internalization and endosomal targeting of the receptor by removing the cytoplasmic tail from Mxra8. Notably, Mxra8 was still located in early endosomes and late endosomes/lysosomes (Chapter 2) which suggests that the cytoplasmic domain of the Mxra8 receptor is not required for receptor internalization. This might imply that –next to Mxra8– other cell-surface molecules are involved in the endosomal targeting of CHIKV. Indeed, many receptors are engaged with other receptors and molecules to regulate intracellular signaling pathways and often multiple receptors are required for the cell entry of viruses28,34–36_{. For example, Hepatitis C virus (HCV) requires binding of its structural}

proteins to CD81, which direct the virus via actin-dependent trafficking towards tight junctions where CLDN-1 and OCLN interaction subsequently leads to endocytosis35_.

But also adenoviruses require, in addition to CAR (coxsackievirus and adenovirus receptor), αVβ3 integrin for internalization36_{. Because Mxra8 closely resembles CAR}

and is also associated with αVβ3 integrin26,37_{, it is worthwhile to pursue the possible}

role of these molecules or other molecules localized in tight junctions in CHIKV internalization. Especially because the αVβ1 integrin heterodimer has previously been suggested as potential CHIKV receptor38,39_{. To study their role in CHIKV cell}

entry we could combine a loss-of-function screen of the integrin superfamily with microscopic analysis of receptor colocalization with Mxra8.

(9)

7

In Chapter 4 we show that drugs that target serotonin (5-HT) receptors inhibit CHIKV infection. 5-HT receptors are members of the G-protein coupled receptor family (GPCR) and these often regulate their cell surface expression and signaling via CME30,40_{. Therefore, we here questioned whether the serotonergic antiviral}

drugs alter the CHIKV attachment/binding capacity to cells. However, the drugs that were used in Chapter 4, 5-NT and MM did not alter CHIKV cell binding. Other studies describe that 5-HT receptors or serotonergic drugs can also influence viral entry steps after binding to the host cells. For example, JC polyomavirus requires 5-HT receptors for its internalization via CME as disturbed receptor internalization strongly affected viral internalization as well41,42_{. HCV cell entry is dependent on 5-HT}

receptors via controlling the cell surface expression of Claudin-1 (CLDN-1), the HCV entry receptor that is important for endocytosis15,43,44_{. The results from our study,}

however, suggest that 5-HT receptors control CHIKV infection after virus cell binding. Activated cell surface molecules can also influence viral infection via their underlying signaling pathways28_{. The stimulation of 5-HT receptors by the serotonergic drugs}

used in Chapter 4, could indeed result in the activation or downregulation of protein kinase A (PKA) pathways. 5-NT is a selective agonist of the 5-HT_1Breceptor which is linked to the Gα inhibitor protein (Gα_i). Gα_i can inhibit adenylyl cyclase and cAMP production that subsequently leads to downregulation of PKA activation45_{. MM, a}

non-selective antagonist, can bind to several 5-HT receptors including 5-HT₁, 5-HT₆ and 5-HT₇.5-HT₆ and 5-HT₇ receptors are associated with a Gα stimulatory protein (Gα_s) that activates the PKA pathways. However, the mechanism of intracellular signaling pathways in CHIKV cell entry remain very limited, thus future research related to this topic is warranted.

In conclusion, cell surface molecules play an important role in viral infections by mediating virus attachment/binding to cells, facilitating receptor-mediated endocytosis and by triggering signaling pathways. For CHIKV, future studies are required to identify activated cellular surface receptors that guide CHIKV endocytosis after CHIKV attachment/binding. For example, phosphoproteomic screens could provide valuable information on specific, bona fide kinases responsible for phosphorylation of substrates required for CHIKV cell entry46–49_.

Chikungunya virus and cell surface molecules | Perspectives on antiviral strategies Cell surface molecules are interesting targets for host-directed antivirals because they can be easily targeted with antiviral agents. For example, flavaglines are plant-derived compounds that block binding of CHIKV to the prohibitin receptor50,51_,

(10)

and monoclonal antibodies against the Mxra8 receptor limit CHIKV infection and pathogenesis in mice and rhesus macaques10,52,53_{. However, CHIKV infection was not}

completely blocked in these experiments, as residual infection could be observed10_.

This same holds true for flavaglines where residual CHIKV infection could be observed51_{. These results indicate that blocking CHIKV via its putative receptor is}

challenging. There are multiple cellular attachment factors and cell surface receptors known that the virus could use to infect its target cell. Therefore, it remains to be investigated whether blocking CHIKV-receptor interactions could be used as antiviral therapeutic applications against CHIKV infections.

The intracellular fate of chikungunya virus particles

Upon binding of the virus with its entry receptor, the virus particle is internalized. We and others showed before that CHIKV predominantly infects cells via clathrin-mediated endocytosis (CME)8_{. Life cell imaging in combination with pharmacological}

inhibitors and siRNA-mediated knockdown of CME-related proteins revealed that more than 85% of the particles enter cells via CME8_{. However, clathrin-independent}

cell entry pathways have also been described8,54_{. The cell entry pathway taken by}

the virus is probably dependent on the virus strain and cell type used and might be dictated by the cell receptor the virus interacts with.

Irrespective of the pathway taken, the internalized vesicles are always transported to early endosomes. We have shown before that more than 95% of the CHIKV particles fuse from within Rab5-positive early endosomal vesicles. In Chapter 3 we investigated the intracellular transport behavior of CHIKV particles upon internalization. We revealed two distinct transport behaviors; half of the particles underwent fast-directed movement prior to membrane fusion and the other half of the particles remained static till the moment of membrane fusion. Subsequent analysis revealed that fast-directed movement occurs after CME and prior to entry in Rab5-positive endosomes. Disruption of the microtubule network with the drug nocodazole resulted in a stark reduction in fast-directed movements of CHIKV-containing vesicles demonstrating that microtubules are involved in the intracellular transport of half of the CHIKV particles. These stationary and microtubule-transported vesicles have been described before by Lakadamyali and colleagues55_{. In their report they show that cargo-carrying vesicles are targeted to}

dynamic rapidly maturing Rab5-positive early endosomes or to slowly maturing Rab5-positive early endosomes. Cargo destined for degradation is often targeted to the dynamic rapidly maturing endosomes after microtubule-dependent transport,

(11)

7

while cargo that is recycled back to the plasma membrane, like transferrin receptors, are non-selectively distributed among both the slowly and rapidly maturing early endosomal populations55_{. Based on our observations in Chapter 3 it is likely that}

chikungunya virions are transported in similar ways as the transferrin receptors: non-selectively to both rapidly and slowly maturing endosomes.

Membrane (hemi)fusion was observed for CHIKV particles irrespective of the trafficking behavior. Furthermore, no difference in time to membrane fusion between CHIKV particles that remained stationary or showed fast-directed movements was observed. Intriguingly, however, treatment with nocodazole showed that an intact microtubule network is required for efficient CHIKV infection. Furthermore, we showed that the cytosolic delivery of gRNA is reduced in cells treated with nocodazole. This may indicate that microtubule-dependent transport increases the chance to successfully infect a cell. To test this hypothesis, we are in need for a better understanding of these slowly and rapidly maturing endosomes. Other than the differences in intracellular mobility and maturation kinetics no details are known55_{. There has been new evidence, however, that the classical early endosomal}

sorting station comprises many distinct early endosomal compartments that differ from each other by cellular distribution and effector molecules. For example, the markers SNX15 and APPL1 have been defined as pre-endosomal sorting markers that mark separate endosomal subpopulations that transport clathrin-endocytosed cargo56–59_{. Although most of this research has been done in the context of epidermal}

growth factor receptor (EGFR), it is known that SNX15 and APPL1 distinguish pre-early endocytic intermediate subsets required for receptor recycling, signaling and degradation56–61_{. In our research we have not yet determined the association of}

pre-early endosomal sorting stations with CHIKV infection, though it could be interesting to visualize SNX15 or APPL1-endosomes via high-resolution microscopy assays to determine whether they are associated with the static and fast-directed movements in Rab5-negative vesicles observed in Chapter 3. Because we do observe membrane fusion in Rab5-positive endosomes, it remains to be determined whether these pre-early endosomal sorting stations transport CHIKV virions to different Rab5-positive early endosomal compartments. Future research should focus on characterizing the pre-early endosomal complexes. This can possibly be done via proximity-labeling assays, identifying unique cellular proteins in the initial phase of CHIKV endocytosis. This will enhance our knowledge in cell biology and will uncover whether these host factors are required to productively infect a cell.

(12)

In Chapter 3 we used BS-C-1 cells for our live-cell microscopy assays. As stated above, alternative pathways than CME can be hijacked for virus entry. To demonstrate if the observed trafficking behavior is a general characteristic for CHIKV infection, the trafficking should also be investigated in other cell types. In addition, not all cell lines are equally permissive to infection and it would be interesting the investigate whether CHIKV particles show more microtubule-dependent trafficking in cells highly permissive to infection.

The intracellular fate of chikungunya virus particles | Perspectives on antiviral strategies

A more thorough understanding of the sorting and trafficking of CHIKV particles during cell entry might reveal new avenues for intervention.

In Chapter 4, we dissected the role of 5-NT in CHIKV disassembly during cell entry. 5-NT had a pronounced antiviral effect on CHIKV infection without inhibiting CHIKV cell binding, membrane fusion and gRNA/nucleocapsid release in the cytoplasm. Yet, no major antiviral effect of 5-NT was seen in cell entry bypass studies. Mainou and colleagues reported a distorted distribution of early endosomes upon 5-NT treatment and hypothesized that this aberrant distribution is the consequence of a signaling cascade introduced by 5-NT upon interaction with 5-HT receptor13_.

The authors also showed that reovirus disassembly is disturbed by 5-NT and hypothesized that this might be due to a disturbed transport to acidified organelles. Could the anti-CHIKV activity of 5-NT be also attributed to a disturbed endosomal trafficking? Stimulation of 5-HT receptors is known to induce rapid uptake of receptor-ligand complexes via CME62–64_{. G-protein coupled receptors (GPCRs) are}

often disconnected from their ligand in early endosomes at mildly acidic pH (pH ~6.2) to recycle back to the plasma membrane whereas the ligand is targeted for degradation via the endolysosomal pathway65_{. Sustained receptor stimulation can}

lead to global sequestration of endosomes, altered endosomal trafficking or to receptor downregulation via the ubiquitin-proteasomal or ESCRT pathway65,66_{. It is}

possible that 5-NT alters the endosomal compartment and/or location from where CHIKV particles fuses. If so, this may suggest that the subsequent disassembly step after gRNA/nucleocapsid release is affected by this. To test this hypothesis, we should assess intracellular trafficking of CHIKV particles by single-particle tracking using fluorescent markers for clathrin, Rab5 and preferably the previously mentioned pre-early endosomal sorting markers. Moreover, we could assess the location of membrane fusion like we did in nocodazole-treated cells in Chapter 3. Here, we

(13)

7

observed that CHIKV membrane fusion events occurs both in the periphery, as well as in the perinuclear region. CHIKV particles that remained static fused almost exclusively in the periphery. Future research should delineate if trafficking to the perinuclear region via specific transport is required for CHIKV infection and if 5-NT is responsible for a distorted trafficking of CHIKV particles.

Overall, more research is required to determine whether interference of the intracellular trafficking behavior of CHIKV is an interesting strategy for antiviral drug development.

The requirements for membrane fusion | more than just lipids and pH?

When CHIKV particles are delivered to Rab5-positive early endosomes it is a matter of seconds (on average 37.6s yet 40% fuses within 10s after colocalization with Rab5-positive endosomes) before membrane fusion occurs8_{. It is generally}

accepted that membrane fusion is triggered by exposure of the virus particle to low pH. The low pH in early endosomes destabilizes the E2-E1 heterodimers within the spike which leads to insertion of the E1 fusion loop in the target membrane and the formation of E1 homotrimers. Subsequent re-folding of E1 trimers leads to a transient hemifusion intermediate where the outer leaflets of the opposing membranes merge. Finally, a fusion pore is formed to release the nucleocapsid in the cytoplasm. Membrane fusion is considered a cooperative process, with multiple E1 trimers that form a ring with neighboring spikes67–70_{. In this thesis we modeled}

the cooperativity of E1 proteins in membrane fusion, by simulating the available number of fusion proteins on the viral membrane in presence of CHK-152. Indeed, the sub-stoichiometric neutralization capacity of CHK-152 allowed us to determine that 3-5 neighboring E1 proteins are required to initiate fusion which shows that CHIKV undergoes cooperative membrane fusion.

The basic requirements for alphavirus membrane fusion has been thoroughly described in recent decades8,67,69,71,72_{. We know that membrane fusion of CHIKV is}

promoted, yet not dependent, by the lipids cholesterol and sphingomyelin in the opposing target membrane67,73–75_{. Moreover, inhibition of the vacuolar H}+_{-ATPase with}

Bafilomycin A1 strongly inhibits CHIKV membrane fusion, stressing the importance of this proton pump in CHIKV membrane fusion76_{. Also, CHIKV was shown to fuse}

efficiently with receptor-free liposomes suggesting that proteinaceous receptors are not essential for membrane fusion67_{. Yet, host proteins may promote fusion.}

Indeed, recently, a membrane protein was identified that influences the membrane fusion properties of CHIKV77,78_{. The exact role of this membrane protein, TSPAN9, a}

(14)

member of the tetraspanin protein family with a transmembrane domain with both the N- and C-tail in the cytoplasm, has yet to be unraveled. It is known, however, that TSPAN9 does not control the pH of early endosomes or cellular cholesterol levels77,78_.

Could Mxra8 be involved in CHIKV fusion? In Chapter 2 we observed that half of the particles colocalized with Mxra8 at the moment of membrane fusion. This colocalization during membrane fusion suggests that Mxra8 might function as a cell entry receptor, thereby guiding CHIKV particles towards the acidic environment of early endosomes. High-resolution imaging techniques, like direct Stochastic Optical Reconstruction Microscopy (dSTORM), PhotoActivated Localization Microscopy (PALM) or electron microscopy (EM) should be used to confirm the interaction between Mxra8 and CHIKV E2-E1 heterodimers during membrane fusion, as these technologies have previously been combined to visualize distinct endosomal populations and receptors at nanoscale79_{. Furthermore, it has been postulated by}

Fremont and colleagues that Mxra8 as a receptor could regulate the pH-dependent membrane fusion properties of CHIKV26_{. To address if Mxra8 indeed controls the pH}

threshold for fusion, we could employ the single-particle fusion assay using planar lipid membranes with and without Mxra8, similarly to what has been described in

Chapter 6. Using this system, we have a controlled environment, excluded from

other cellular factors and thereby we can study the specific role of Mxra8 in CHIKV membrane fusion, including the pH requirements for membrane fusion.

The requirements for membrane fusion | Perspectives on antiviral strategies

Blocking membrane fusion is an attractive strategy as antiviral treatment against virus infection. Indeed, Enfuvirtide/T-20 is a fusion inhibitor being used in the treatment against HIV-1. T-20 is a peptide which binds to viral glycoproteins thereby preventing the rearrangements that are required for membrane fusion80_.

CHIKV-specific peptides/small molecules could potentially be used to block CHIKV glycoprotein rearrangements required for membrane fusion. In Chapter 6 we determined that membrane fusion is a cooperative process. Therefore, only a subset of fusion proteins needs to be de-activated, or neutralized, on the viral envelope to prevent membrane fusion. This is beneficial for therapeutic developments. For DENV, small-molecule inhibitors only required ~20-25% occupancy in the virus particle to completely prevent membrane fusion81_{. Thus, the design of fusion inhibitors is an}

interesting direction for future antiviral therapies due to the cooperative mechanism of CHIKV membrane fusion.

(15)

7

Next to peptides/small molecules, antibodies can also be used to ‘freeze’ virus particles thereby preventing membrane fusion and infection. In Chapter 6, we employed a single-particle assay with a planar lipid membrane to study the mechanistic effects of a potent neutralizing antibody CHK-152 on CHIKV membrane fusion. CHK-152 was previously identified as a promising therapeutic antibody as it blocked membrane fusion with the host cell plasma membrane and with liposomes20–22_{. In Chapter 6 we studied the mechanism of neutralization in more}

detail. We observed that CHK-152 interacts with CHIKV under neutral and low pH interaction and shields the virions from binding to lipid bilayers. We also revealed that CHK-152 antibodies prevent E1 homotrimer formation under acidic conditions. Based on the available data we hypothesize that CHK-152 disables CHIKV particles to destabilize its spike heterodimer and thereby preventing stable insertion of the E1 fusion loop and E1 trimer formation.

Fusion inhibitors could block separate stages of the membrane fusion reaction. For example, CHK-152 blocks membrane fusion before E1 homotrimer formation, while the HIV-1 fusion inhibitor T-20 blocks the refolding of the trimer. Alternatively, it has been questioned whether virus particles could be ‘trapped’ in their hemifusion state. In Chapter 3 we show that CHIKV gRNA release was inhibited by nocodazole, while there was an unchanged membrane fusion activity. This may suggest that nocodazole could ‘trap’ the virus particles in the transient state between hemifusion and fusion pore formation. Lipids in the endosomal membrane can promote, or disfavor fusion pore formation82_{. Alternatively, the halt in hemifusion/fusion pore formation could}

be a consequence of unrestricted membrane fusion with intraluminal vesicles83_.

This has been described by Melikyan and colleagues and defined as the fusion decoy model84,85_{. This model explains the observation of membrane fusion events}

without viral nucleocapsid release in the cytoplasm. Could CHIKV be targeted to intraluminal vesicles upon disruption of the microtubule network? In Chapter 3 we observed more CHIKV particles at the peripheral site of the host cell upon nocodazole treatment and intraluminal vesicles are present in endosomes at this location. To test this hypothesis, we should determine the colocalization of CHIKV with intraluminal vesicles in nocodazole-treated cells. This has previously be done using electron microscopy86_.

What about host-directed inhibitors for CHIKV fusion? In Chapter 4, we showed that Methiothepin Mesylate (MM), a 5-HT receptor nonspecific antagonist is able to inhibit CHIKV cell entry at or just before the membrane fusion step. Because MM may inhibit CHIKV entry at several steps between virus binding and membrane fusion,

(16)

we first need to examine which basic key-properties have been changed. First, we should study the internalization properties of CHIKV particles upon MM treatment because the internalization assay used in our research described in Chapter 4 does not discriminate between bound and internalized particles. For example, we could use the live-cell imaging assays to check the delivery of single virus particles to Rab5-positive endosomes. Secondly, the acidification of the endosomal lumen could be altered by MM. This could be assessed using pH-sensitive dyes, in combination with fluorescently labeled Rab5. Finally, alteration in the lipid/protein composition in the early endosomal target membrane could be assessed by staining cellular cholesterol levels using filipin. The information derived from these experiments will provide new insights in the cellular targets and virus:host interactions that are possibly changed by MM and delineate the properties of MM as a CHIKV fusion inhibitor.

Fusion inhibitors are interesting targets for drug development strategies as this process is crucial to infection. A better understanding of the virus:host interactions –in this thesis highlighted with Mxra8 and 5-HT receptors– and the molecular events in membrane fusion –in this thesis highlighted with CHK-152– help to find novel host and viral-directed fusion inhibitors for antiviral strategies in the future.

Nucleocapsid delivery and uncoating | unexplored steps in CHIKV cell entry

After membrane fusion pore formation, the nucleocapsid is released in the cytoplasm. In the particle, the nucleocapsid interacts with the cytoplasmic tail of the viral envelope protein E2 at the C-terminal domain of the capsid protein. The nucleocapsid is composed of 240 copies of capsid proteins that interact with gRNA at the N-terminal domain. Nucleocapsid release and uncoating are required for the initial gRNA translation and replication. In 1984, Wengler and Wengler showed that cellular ribosomes associate with the nucleocapsid of Sindbis virus (SINV)8787_.

One decade later, in 1992, Singh and Helenius proposed that the cellular ribosomes are involved in the uncoating of nucleocapsids of the Semliki Forest virus (SFV)88_.

Around the same time, Wengler and coworkers identified a specific sequence element in the SINV capsid protein89_{which was named the ribosomal binding site}

(RBS) as this region was responsible for ribosomal binding to the capsid protein89_.

However, the RBS is located inside the nucleocapsid and thus it was hypothesized that the alphavirus nucleocapsid requires priming before ribosomes can bind the capsid protein90,91_{. Overall, many details on the mechanism of nucleocapsid release}

(17)

7

One of the host factors that might be involved in CHIKV nucleocapsid release are microtubules (Chapter 3). We used a cellular fractionation assay to differentiate between nucleocapsids/gRNA that is released in the cytosol from those that remain in the endosomal vesicles. This assay has previously been described for other viruses and enabled us to assess the effect of host factors on nucleocapsid/gRNA release in the cytosol92–94_{. In Chapter 3 we show that disruption of the microtubule network}

by nocodazole reduced the number of gRNA copies in the cytosol, independent of membrane fusion; only half of the gRNA copies were released in the cytosolic fraction compared to the control cells. In the previous paragraph we hypothesized that this reduction of gRNA could be related to the fusion decoy model with unrestricted fusion events in ILVs. However, an alternative explanation could be that microtubules are directly involved in nucleocapsid release upon fusion pore formation. Indeed, several viruses were found to be dependent on microtubules for their nucleocapsid release and uncoating within the cytoplasm95–102_{. Influenza A virus requires the}

microtubule-associated protein HDAC6 that links polyubiquitinated proteins – which are encapsidated inside the virion during virus assembly– to the aggresome processing machinery, thereby facilitating the breakdown of the nucleocapsid shell via opposing forces of cytoskeletal motor proteins100_{. The antagonistic forces from}

motor proteins are also crucial for uncoating of HIV-1 in the cytoplasm, via the interaction of the capsid protein with adaptor proteins FEZ1 and BICD297–99,103_{. These}

studies suggest that microtubule-associated proteins directly interact with viral nucleocapsids upon membrane fusion and thereby aid nucleocapsid release and subsequent uncoating.

To understand the role of host factors in CHIKV nucleocapsid release, we are in need for novel live-cell imaging techniques that enable us to track single nucleocapsids upon CHIKV cell entry. For example, fluorescent-tagged nucleocapsids can provide valuable information upon internalization and endosomal trafficking of enveloped viruses. However, unlike the successful labeling of HIV-1 capsid proteins or nucleocapsids from baculoviruses95,104_{, the comprised packaging of CHIKV particles}

restricts the addition of a fluorescent tag to the nucleocapsids of newly produced viruses105,106_{. Alternatives can be found in quantum dots that are smaller and have}

better optical properties107–109_{, or in fluid-markers in between the nucleocapsid and}

the viral envelope, which are released upon membrane fusion fore formation104,110,111_.

However, this fluid-marker rather indicates full fusion pore formation than nucleocapsid release.

(18)

Nucleocapsid delivery and uncoating | Perspectives on antiviral strategies

Is CHIKV nucleocapsid release and/or uncoating an interesting target for intervention? In principle yes, although further research is required to understand these unexplored steps in CHIKV cell entry. Recently a crystal structure of the CHIKV capsid protein in combination with molecular docking of two compounds identified two potent antiviral drugs, MDA and EAB, that bind to the hydrophobic pocket of the CHIKV capsid protein112_{. While these compounds were designed to block}

viral budding, there could also be a potential role of MDA and EAB in blocking nucleocapsid uncoating by stabilizing the icosahedral capsid112_{. Cell-based antiviral}

assays are needed to gain additional information on the potential antiviral efficacy of these compounds in CHIKV cell entry.

In Chapter 4 we demonstrated that serotonergic targeting with the 5-HT receptor agonist 5-NT inhibited CHIKV infection. We showed that in presence of 5-NT direct transfer of in vitro transcribed viral RNA does not have a major influence on CHIKV infection and under normal infection conditions 5-NT did not impair virus cell binding, fusion and gRNA delivery. These puzzling results might indicate that in presence of 5-NT, the nucleocapsid is released but cannot be disassembled. The gRNA –that is still associated with capsid proteins– can be detected in the cytosolic fraction but fails to initiate a productive infection. How could 5-NT inhibit nucleocapsid disassembly? Host proteins could be activated upon 5-NT treatment and stabilize the nucleocapsid within the cytoplasm, something that has been described recently for HIV-1113_{. Death domain-associated protein 6 (Daxx) is activated by interferon}

signaling, and inhibits –besides reverse transcription of the viral RNA– nucleocapsid uncoating of HIV-1113_{. Alternatively, 5-NT could block host factors that are crucial}

for nucleocapsid uncoating. For example, the host protein HDAC6, required for Influenza A virus nucleocapsid release and uncoating can also become inactivated late in the virus replication cycle101,114_{. This blockage is achieved via caspase mediated}

cleavage of the functional domains of HDAC6 and allows nucleocapsid to assemble in progeny virus particles101,114_.

To experimentally assess nucleocapsid uncoating we could detect free gRNA in the cytoplasm by tagging the gRNA with fluorescent markers using quantum dots or MS2-based fluorescence115–117_{. With these techniques one can track incoming single}

mRNA molecules in live cells after the membrane fusion event and subsequent nucleocapsid uncoating, providing crucial information about the dynamics and location of released gRNA in the cytosol. Alternatively, advanced cellular fractionation assays could be used to differentiate intact nucleocapsid cores from

(19)

7

dissociated –soluble– nucleocapsids in the host cell cytoplasm. Recently, HIV-1 nucleocapsid have been separated from soluble capsid proteins in the cellular cytoplasm using ultracentrifugation upon hypotonic lysis and sucrose cushion gradient113_{. A comparable assay could be applied for CHIKV using radioactively}

labeled CHIKV particles or capsid-specific antibodies to localize capsid proteins in the corresponding fractions.

In conclusion, novel live-cell imaging assays that allow tracking of nucleocapsid release and uncoating and biochemical should be designed to study CHIKV infection. When combining these techniques with our established single-particle tracking assay we enable ourselves to study host factors along the complete virus disassembly steps during cell entry, from cell binding until nucleocapsid uncoating.

Protein quality control systems in the chikungunya virus replication cycle

Besides the above mentioned host factors involved in CHIKV cell entry, we were also intrigued by the important role of the heat shock 70kDa proteins (Hsp70) – the molecular chaperones that control protein homeostasis– during cell entry of flaviviruses ZIKV and DENV17,118–120_{. Due to the important role of Hsp70 in CME}

(described in Chapter 1), the resemblance in cell entry between alphaviruses and flaviviruses and the previously described interaction of Hsp70 with CHIKV121_{, we}

studied the antiviral potential of Hsp70 in CHIKV infection in Chapter 5. Despite the strong antiviral activity of the allosteric Hsp70 inhibitors, we observed only a limited effect on CHIKV cell entry based on our time-of-drug-addition experiments. One explanation of these differences could be that flaviviruses are dependent on the ubiquitin-proteasome system during cell entry32,122–124_{. This protein degradation}

system is closely connected with the Hsp70 network125_{. While the}

ubiquitin-proteasome system has been linked to alphavirus replication, it has not yet been linked to CHIKV cell entry126,127_{, which might explain why we observe differences in}

antiviral activity of Hsp70 inhibitors on flavivirus and alphavirus cell entry.

Besides the role in viral entry, Hsp70 proteins can be linked to other steps of virus life cycle including replication, translation, assembly, and release19_{. Viruses, as obligatory}

parasites, can use the protein quality system of Hsp70 proteins in their own benefit, especially when this system is highly upregulated upon stressed conditions, including viral infections19_{. In Chapter 5 we show that Hsp70 inhibitors strongly reduce CHIKV}

protein expression, progeny virus particle secretion and for the Hsp70 inhibitor JG-98 a reduction in the infectivity of newly produced viruses was seen. How can the Hsp70 inhibitor JG-98 reduce viral infectivity? Interestingly, this phenomenon was

(20)

also seen for ZIKV-infected cells treated with 40, a Hsp70 inhibitor similar to JG-98118_{. The authors uncovered an additional, slower migrating, protein band of the}

capsid protein on iodixanol gradients, indicating an alteration in the morphogenesis of capsid proteins upon Hsp70 inhibition128_{. The precise role of these morphological}

changes in the capsid protein is unknown, but it was postulated that it alters the capsid conformation. Ultimately, this could lead to defects in nucleocapsid formation and consequently lower the infectivity of progeny virions. For CHIKV this is unknown although a stark reduction in capsid protein synthesis was observed. Future experiments should unravel whether Hsp70 inhibitors affect nucleocapsid formation. Structural alterations in newly produced CHIKV virions could be assessed via advanced structural studies including X-ray crystallography and electron cryo-microscopy. Other Hsp70 inhibitors, like VER-155008, did not influence the infectivity of progeny virions yet a drastic reduction in virion release was observed. Here, detailed studies into protein translation, stability, virion assembly and/or budding are needed to understand the role of Hsp70 inhibition on the reduction in excreted CHIKV particles.

Protein quality control systems | Perspectives on antiviral strategies

Overall, our research shows that Hsp70 might be a promising target for host-directed antivirals as CHIKV infection is strongly reduced in cells treated with Hsp70 inhibitors. However, the Hsp70 inhibitors used here might not be suitable for antiviral drug development due to reasons that include drug metabolism and cytotoxicity. The used Hsp70 inhibitors block the complete protein quality system and protein homeostasis in cells. To produce specific, non-toxic antivirals for therapeutic applications, it would be helpful to find specificity in the interaction of CHIKV:Hsp70s. There are several ways to find this specificity. First, via the identification of specific proteins in the diverse family of co-chaperones, including the DnaJ family. There are >50 DnaJ (Hsp40) family members present in mammalian cells that provide specificity to the Hsp70 machinery. While all DnaJ proteins contain a highly conserved J domain, the other domains varies majorly16,129_{. The J domain is required}

for binding to Hsp70 proteins, while the C-terminal regions of the DnaJ protein is required for binding with the client protein (often non-native polypeptides)129_.

Identification of the DnaJ protein that specifically interacts with viral proteins might help us to find specificity in the Hsp70 network during CHIKV infection. Alternatively, we can define the specific cellular Hsp70 protein required for CHIKV infections. There are several Hsp70 paralogs present within each cellular compartment; with organelle-specific functions in for example the ER, cytosol or the nucleus130,131_{. For}

(21)

7

example, downregulation of the cytosolic Hsp70 HspA1A, HspA1B and HspA8 (but not HspA6) strongly affected DENV infection, while there was no effect with the ER- and mitochondrial-located Hsp70s119_{. Thus, Hsp70 isoforms might have diverse}

functions in the CHIKV replication cycle and thus more research is warranted. In conclusion, CHIKV-Hsp70 interactions hold promises for antiviral drug development strategies but specificity should be found in the family of co-chaperones and in Hsp70 paralogs.

Concluding remarks

The structural and basic molecular understanding of the alphavirus replication cycle –including CHIKV– is mostly based on studies with the related alphaviruses SFV and SINV. Despite the similarities with CHIKV, there are already multiple examples of host cell responses and molecular mechanisms that differ between these closely related alphaviruses132,133_{. For example, previous publications showed different}

roles of microtubules in the alphavirus replication cycle. We showed in this thesis that microtubules are important for the cell entry of CHIKV. However, previously it was shown that microtubules are involved in the internalization of replication complexes towards the perinuclear location of the cell for SFV134,135_{and microtubule}

are dispensable for SINV infection136,137_{. Besides cellular proteins, the triggered}

intracellular signaling cascades can also differ between alphaviruses. For example, the PI3K-Akt-mTOR pathway was differently activated between SFV and CHIKV138_.

Thus, in order to fully understand the virus:host interactions that occur during CHIKV infection it is important to study CHIKV itself. In this regard I also plead for more structural information and cryo-EM structures of the CHIKV viral particle –– especially of the capsid genome and ribosomal binding site.

In this thesis I proposed future perspectives for elucidating the virus:host interactions during CHIKV infection. First, I advocate investigation into the phosphorylation status of cell surface receptors upon CHIKV binding to the target cell, to better understand receptor internalization and endocytosis of CHIKV particles. Also, we are in need for better understanding of the pre-early endosomal sorting complex and different early endosomal subpopulations. Next, although many details of the membrane fusion reaction are known, more emphasis should be placed on the potential importance of host factors in this process. Finally, cell-based assays are urgently needed to study the virus:host interactions that occur during CHIKV nucleocapsid release and uncoating. Detailed information on the cell entry process

(22)

of CHIKV will further guide the rational design of antiviral drugs that specifically interfere with the early steps in infection.

Thus far, numerous inhibitors have been identified with in vitro efficacy to limit CHIKV cell entry, but fail to enter clinical trials due to several issues, including cytotoxicity or limited pharmacokinetic profiles139,140_{. In my perspective, cellular factors as targets for}

antivirals have several advantages over direct-acting antivirals because they are less prone for the development of antiviral drug resistance141,142_{. Indeed, several potent}

antiviral therapies that target viral proteins required for cell entry quickly led to development of resistant mutants. For example, amantadine and rimantadine inhibit the uncoating step of IAV by targeting and blocking the viral M2 ion channel, but IAV quickly acquired resistance against both drugs143–146_{. Pleconaril, an antiviral drug}

against picornaviruses, binds a specific binding pocket of the capsid protein that inhibits the uncoating process of the virus147_{. After reaching phase 3 clinical trials,}

the drug was not approved by the FDA due to the side effects and drug resistant mutations148_{. Even the currently available HIV-1 fusion inhibitor T-20 has issues}

concerning the development of drug resistant viruses149_{. Host-directed treatments}

that inhibit HIV-1 cell entry are available and/or in development. These inhibitors include maraviroc and vicriviroc, antagonists against CCR5, one of the two HIV-1 receptors150_{. While these antiviral therapies against HIV-1 are continuously under}

development to gain higher effectivity and are still sensitive for the development of drug resistant viruses151,152_{, they show that it is possible to use host-directed antiviral}

against viral infections. These studies provide hopeful perspectives to establish specific, non-toxic cell entry inhibitors against CHIKV infections in the future. Finally, the challenges we face with the current pandemic of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) evidence the need for a better understanding of virus:host interactions to find effective antiviral therapies. At the same time, this pandemic shows the potential in antiviral drug development. Multi-disciplinary research initiatives have been established with virologists, computational and chemical biologists, structural biologists, and clinicians to find and develop new antiviral therapies against SARS-CoV-2153–156_{. The lessons learned during the current}

pandemic, will hopefully aid and speed-up the development of antiviral therapies for other, often neglected, viral diseases, including CHIKV.

(23)

7

References

1 Erasmus JH, Rossi SL, Weaver SC. Devel-opment of vaccines for chikungunya fever. J Infect Dis 2016; 214: S488–S496.

2 Wilder-Smith A, Gubler DJ, Weaver SC, Monath TP, Heymann DL, Scott TW. Epidemic arboviral diseases: priorities for research and public health. Lan-cet Infect Dis 2017; 17: e101–e106.

3 Schrauf S, Tschismarov R, Tauber E, Ram-sauer K. Current Efforts in the Development of Vac-cines for the Prevention of Zika and Chikungunya Virus Infections. Front. Immunol. 2020. doi:10.3389/ fimmu.2020.00592.

4 Wressnigg N, Hochreiter R, Zoihsl O, Fritzer A, Bézay N, Klingler A et al. Single-shot live-attenuat-ed chikungunya vaccine in healthy adults: a phase 1, randomised controlled trial. Lancet Infect Dis 2020. doi:10.1016/S1473-3099(20)30238-3.

5 clinicaltrials.gov. 2020.https://clinicaltrials.gov/ct2/show/NCT04546724?cond=V-LA1553&draw=2&rank=2.

6 Schwartz O, Albert ML. Biology and patho-genesis of chikungunya virus. Nat Rev Microbiol 2010; 8: 491–500.

7 Silva LA, Dermody TS. Chikungunya virus: Epidemiology, replication, disease mechanisms, and prospective intervention strategies. J. Clin. Invest. 2017; 127: 737–749.

8 Hoornweg TE, van Duijl-Richter MKS, Aya-la Nuñez N V., Albulescu IC, van Hemert MJ, Smit JM. Dynamics of Chikungunya Virus Cell Entry Unraveled by Single-Virus Tracking in Living Cells. J Virol 2016; 90: 4745–4756.

9 Ayala-Nuñez N V., Wilschut J, Smit JM. Monitoring virus entry into living cells using DiD-labe-led dengue virus particles. Methods 2011; 55: 137–143.

10 Zhang R, Kim AS, Fox JM, Nair S, Basore K, Klimstra WB et al. Mxra8 is a receptor for multiple arthritogenic alphaviruses. Nature 2018; 557: 570–574.

11 Hoyer D, Hannon JP, Martin GR. Molecular, pharmacological and functional diversity of 5-HT re-ceptors. Pharmacol Biochem Behav 2002; 71: 533–554.

12 Berger M, Gray JA, Roth BL. The Expanded Biology of Serotonin. Annu Rev Med 2009; 60: 355– 366.

13 Mainou B a., Ashbrook AW, Smith EC, Dor-set DC, Denison MR, Dermody TS. Serotonin Receptor Agonist 5-nonyloxytryptamine Alters the Kinetics of Reovirus Cell Entry. J Virol 2015; 89: JVI.00739-15.

14 Assetta B, Maginnis MS, Gracia Ahufinger I, Haley SA, Gee G V., Nelson CDS et al. 5-HT2 Recep-tors Facilitate JC Polyomavirus Entry. J Virol 2013; 87: 13490–13498.

15 Cao L, Chen J, Wang Y, Yang Y, Qing J, Rao Z et al. Identification of serotonin 2A receptor as a novel HCV entry factor by a chemical biology strategy. Protein Cell 2018. doi:10.1007/s13238-018-0521-z.

16 Kampinga HH, Craig EA. The HSP70 chap-erone machinery: J proteins as drivers of functional specificity. Nat. Rev. Mol. Cell Biol. 2010. doi:10.1038/ nrm2941.

17 Bukau B, Horwich AL. The Hsp70 and Hsp60 chaperone machines. Cell. 1998. doi:10.1016/ S0092-8674(00)80928-9.

18 Cyr DM. Swapping Nucleotides, Tuning Hsp70. Cell. 2008. doi:10.1016/j.cell.2008.05.036.

19 Wan Q, Song D, Li H, He M liang. Stress proteins: the biological functions in virus infection, present and challenges for target-based antiviral drug development. Signal Transduct Target Ther 2020; 5. doi:10.1038/s41392-020-00233-4.

20 Pal P, Dowd KA, Brien JD, Edeling MA, Gorlatov S, Johnson S et al. Development of a Highly Protective Combination Monoclonal Antibody Ther-apy against Chikungunya Virus. PLoS Pathog 2013. doi:10.1371/journal.ppat.1003312.

21 Pal P, Fox JM, Hawman DW, Huang Y-JS, Messaoudi I, Kreklywich C et al. Chikungunya Viruses That Escape Monoclonal Antibody Therapy Are Clini-cally Attenuated, Stable, and Not Purified in Mosqui-toes. J Virol 2014. doi:10.1128/jvi.01032-14.

22 Sun S, Xiang Y, Akahata W, Holdaway H, Pal P, Zhang X et al. Structural analyses at pseudo atomic resolution of Chikungunya virus and antibodies show mechanisms of neutralization. Elife 2013. doi:10.7554/ eLife.00435.

23 Schnierle BS. Cellular attachment and entry factors for chikungunya virus. Viruses. 2019. doi:10.3390/v11111078.

(24)

24 van Duijl-Richter MKS, Hoornweg TE, Rod-enhuis-Zybert IA, Smit JM. Early events in chikungunya virus infection—from virus cell binding to membrane fusion. Viruses 2015; 7: 3647–3674.

25 Song H, Zhao Z, Chai Y, Jin X, Li C, Yuan F et al. Molecular Basis of Arthritogenic Alphavirus Re-ceptor MXRA8 Binding to Chikungunya Virus Envelope Protein. Cell 2019. doi:10.1016/j.cell.2019.04.008.

26 Basore K, Kim AS, Nelson CA, Zhang R, Smith BK, Uranga C et al. Cryo-EM Structure of Chikun-gunya Virus in Complex with the Mxra8 Receptor. Cell 2019. doi:10.1016/j.cell.2019.04.006.

27 Bhella D. The role of cellular adhesion mol-ecules in virus attachment and entry. Philos Trans R Soc B Biol Sci 2015. doi:10.1098/rstb.2014.0035.

28 Boulant S, Stanifer M, Lozach PY. Dynamics of virus-receptor interactions in virus binding, signal-ing, and endocytosis. Viruses. 2015; 7: 2794–2815.

29 Sodhi A, Montaner S, Gutkind JS. Viral hi-jacking of G-protein-coupled-receptor signalling net-works. Nat Rev Mol Cell Biol 2004; 5: 998–1012.

30 Pavlos NJ, Friedman PA. GPCR Signaling and Trafficking: The Long and Short of It. Trends Endo-crinol Metab 2017; 28: 213–226.

31 Traub LM, Bonifacino JS. Cargo recognition in clathrin-mediated endocytosis. Cold Spring Harb Perspect Biol 2013. doi:10.1101/cshperspect.a016790.

32 Dejarnac O, Hafirassou ML, Chazal M, Ver-sapuech M, Gaillard J, Perera-Lecoin M et al. TIM-1 Ubiquitination Mediates Dengue Virus Entry. Cell Rep 2018; 23: 1779–1793.

33 Wang G, Jiang L, Wang J, Zhang J, Kong F, Li Q et al. The G Protein-Coupled Receptor FFAR2 Pro-motes Internalization during Influenza A Virus Entry. J Virol 2019. doi:10.1128/jvi.01707-19.

34 Berger EA, Murphy PM, Farber JM. CHEMOKINE RECEPTORS AS HIV-1 CORECEPTORS: Roles in Viral Entry, Tropism, and Disease. Annu Rev Im-munol 1999. doi:10.1146/annurev.imIm-munol.17.1.657.

35 Evans MJ, Von Hahn T, Tscherne DM, Syder AJ, Panis M, Wölk B et al. Claudin-1 is a hepatitis C virus co-receptor required for a late step in entry. Nature 2007. doi:10.1038/nature05654.

36 Salinas S, Zussy C, Loustalot F, Henaff D, Menendez G, Morton PE et al. Disruption of the

cox-sackievirus and adenovirus receptor-homodimeric in-teraction triggers lipid microdomain-and dynamin-de-pendent endocytosis and lysosomal targeting. J Biol Chem 2014. doi:10.1074/jbc.M113.518365.

37 Han SW, Jung YK, Lee EJ, Park HR, Kim GW, Jeong JH et al. DICAM inhibits angiogenesis via sup-pression of AKT and p38 MAP kinase signalling. Cardi-ovasc Res 2013. doi:10.1093/cvr/cvt019.

38 Fraisier C, Koraka P, Belghazi M, Bakli M, Granjeaud S, Pophillat M et al. Kinetic analysis of mouse brain proteome alterations following chikun-gunya virus infection before and after appearance of clinical symptoms. PLoS One 2014. doi:10.1371/jour-nal.pone.0091397.

39 La Linn M, Eble JA, Lübken C, Slade RW, Heino J, Davies J et al. An arthritogenic alphavirus uses the α1β1 integrin collagen receptor. Virology 2005. doi:10.1016/j.virol.2005.03.015.

40 Fessart D, Delom F. Role of phospho-rylation in the control of clathrmediated in-ternalization of GPCR. Int J Cell Biol 2011; 2011. doi:10.1155/2011/246954.

41 Mayberry CL, Soucy AN, Lajoie CR, Du-Shane JK, Maginnis MS. JC Polyomavirus Entry by Clathrin-Mediated Endocytosis Is Driven by β-Arrestin. J Virol 2019; 93. doi:10.1128/jvi.01948-18.

42 Assetta B, Morris-Love J, Gee G V., Atkin-son AL, O’Hara BA, Maginnis MS et al. Genetic and Functional Dissection of the Role of Individual 5-HT 2 Receptors as Entry Receptors for JC Polyomavirus. Cell Rep 2019; 27: 1960–1966.e6.

43 Vandeputte A, Fénéant L, Helle F, Rouillé Y, Song O, Baumert TF et al. Identification of Piperaz-inylbenzenesulfonamides as New Inhibitors of Clau-din-1 Trafficking and Hepatitis C Virus Entry. J Virol 2018; 92: 1–19.

44 Farquhar MJ, Harris HJ, Diskar M, Jones S, Mee CJ, Nielsen SU et al. Protein Kinase A-Dependent Step(s) in Hepatitis C Virus Entry and Infectivity. J Virol 2008. doi:10.1128/jvi.00592-08.

45 Hannon J, Hoyer D. Molecular biology of 5-HT receptors. Behav Brain Res 2008; 195: 198–213.

46 Kalume DE, Molina H, Pandey A. Tack-ling the phosphoproteome: Tools and strategies. Curr. Opin. Chem. Biol. 2003. doi:10.1016/S1367-5931(02)00009-1.