Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Dec 1.
Published in final edited form as: Antiviral Res. 2020 Nov 1;184:104966. doi: 10.1016/j.antiviral.2020.104966

BIKE regulates dengue virus infection and is a cellular target for broad-spectrum antivirals

Szuyuan Pu 1,#, Stanford Schor 1,#, Marwah Karim 1, Sirle Saul 1, Makeda Robinson 1, Sathish Kumar 1, Laura I Prugar 2, Danielle E Dorosky 2, Jennifer Brannan 2, John M Dye 2, Shirit Einav 1,#
PMCID: PMC7879702  NIHMSID: NIHMS1666103  PMID: 33137362

Abstract

Global health is threatened by emerging viruses, many of which lack approved therapies and effective vaccines, including dengue, Ebola, and Venezuelan equine encephalitis. We previously reported that AAK1 and GAK, two of the four members of the understudied Numb-associated kinases (NAK) family, control intracellular trafficking of RNA viruses. Nevertheless, the role of BIKE and STK16 in viral infection remained unknown. Here, we reveal a requirement for BIKE, but not STK-16, in dengue virus (DENV) infection. BIKE mediates both early (postinternalization) and late (assembly/egress) stages in the DENV life cycle, and this effect is mediated in part by phosphorylation of a threonine 156 (T156) residue in the μ subunit of the adaptor protein (AP) 2 complex. Pharmacological compounds with potent anti-BIKE activity, including the investigational anticancer drug 5Z-7-oxozeaenol and more selective inhibitors, suppress DENV infection both in vitro and ex vivo. BIKE overexpression reverses the antiviral activity, validating that the mechanism of antiviral action is, at least in part, mediated by BIKE. Lastly, 5Z-7-oxozeaenol exhibits antiviral activity against viruses from three unrelated RNA viral families with a high genetic barrier to resistance. These findings reveal regulation of poorly understood stages of the DENV life cycle via BIKE signaling and establish a proof-of-principle that pharmacological inhibition of BIKE can be potentially used as a broad-spectrum strategy against acute emerging viral infections.

Keywords: Dengue virus, Antivirals, Kinase inhibitors, Drug repurposing, Virus-host interactions

1. Introduction

Dengue virus (DENV) infection is a global health threat, estimated to infect ~390 million people annually (Bhatt et al., 2013; Messina et al., 2019). The geographical range of dengue, including in the developed world, has been expanding due to climate change and rapid urbanization (Messina et al., 2019). DENV is an enveloped, positive single-stranded RNA virus whose 10.7 kb genome encodes a single polyprotein that is proteolytically cleaved into individual proteins (Diamond and Pierson, 2015). The DENV nonstructural (NS) proteins form the viral replication machinery, whereas the capsid, pre-membrane (prM), and envelope (E) proteins form virions. DENV enters cells via clathrin-mediated endocytosis (Diamond and Pierson, 2015). The current model of infectious dengue production suggests that viral particles assemble in ER sites where the E protein is present (Welsch et al., 2009). They become infectious once processed in the trans-Golgi network (TGN) and exit the cell via the secretory pathway (Kudelko et al., 2012). Nevertheless, a comprehensive understanding of the mechanisms that govern viral particle trafficking during DENV entry and assembly/egress is still lacking.

The Numb-Associated Kinases (NAK) family of Ser/Thr kinases is composed of AAK1 (adaptor-associated kinase 1), GAK (G-cyclin associated kinase), BIKE (BMP2-inducible kinase), and STK16 (serine/threonine kinase 16). This is a diverse family of kinases with only limited homology in the kinase domain and low homology in other protein regions (Sorrell et al., 2016). Several NAKs have been shown to regulate intracellular membrane trafficking (Sato et al., 2009; Sorensen and Conner, 2008). AAK1 and GAK phosphorylate the μ subunits of the endocytic adaptor protein (AP) complex 2 (AP2M1) and the secretory AP complex 1 (AP1M1), thereby stimulating their binding to cellular cargo (Conner and Schmid, 2002). STK16, the most distantly related member of the family, regulates secretion in the constitutive secretory pathway at the TGN (In et al., 2014; López-Coral et al., 2018). BIKE, whose structure is closely related to AAK1, was identified as an accessory protein on a subset of clathrin-coated vesicles (CCVs) and was shown to bind the endocytic adaptor Numb and to phosphorylate a synthetic AP2M1 peptide (Kearns et al., 2001; Krieger et al., 2013; Sorrell et al., 2016). However, the biological relevance of AP2M1 phosphorylation by BIKE remained unclear and the functions of BIKE signaling are largely unknown.

We have previously reported that AAK1 and GAK regulate intracellular trafficking of hepatitis C virus (HCV) and DENV during viral entry and assembly/egress in part by phosphorylating AP1M1 and AP2M1 (Bekerman et al., 2017; Neveu et al., 2015, 2012; Xiao et al., 2018). Moreover, we have demonstrated that AAK1 and GAK are required for infections with several other RNA viral families and are the molecular targets underlying the broad-spectrum antiviral effect of various repurposed and novel kinase inhibitors that we have been developing (Bekerman et al., 2017; Pu et al., 2018; Verdonck et al., 2019). The role of the remaining members of the NAK family, BIKE and STK16, in DENV infection remained unknown.

Here, we demonstrate that BIKE, but not STK16, is required for two temporally distinct stages of the DENV life cycle and show that this effect is mediated in part by phosphorylation of a threonine 156 (T156) AP2M1 residue. Moreover, we demonstrate that pharmacological compounds with potent anti-BIKE activity inhibit DENV infection both in vitro and ex vivo, exhibit broad-spectrum antiviral activity with a high genetic barrier to resistance, and their mechanism of antiviral action is, at least in part, mediated by BIKE.

2. Materials and Methods

2.1. Cell lines.

Huh7, Huh7.5 (Apath LLC), T-REx-293, 293T BHK-21 (ATCC), U-87 MG (ATCC) and Vero (ATCC) cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) (Mediatech) supplemented with 10% fetal bovine serum (Omega Scientific), 1% l-glutamine, and 1% penicillin-streptomycin (Gibco), with or without nonessential amino acids, and maintained in a humidified incubator with 5% CO2 at 37°C. Cells were tested negative for mycoplasma by the MycoAlert mycoplasma detection kit (Lonza, Morristown, NJ).

2.2. Plasmids and virus constructs.

DENV2 (New Guinea C strain) TSV01 Renilla reporter plasmid (pACYC NGC FL) was a gift from Pei-Yong Shi (University of Texas Medical Branch, Galveston, Texas, USA) (Zou et al., 2011) and DENV 16681 plasmid (pD2IC-30P-NBX) was a gift from Claire Huang (Centers for Disease Control and Prevention, Public Health Service, US Department of Health and Human Services, Fort Collins, Colorado, USA) (Huang et al., 2010). pCMV-DV2Rep was a gift from Andrew Yueh (Institute of Biotechnology and Pharmaceutical Research, Taipei, Taiwan) (Yang et al., 2013). The plasmid encoding VEEV TC-83 with a nanoluciferase reporter (VEEV TC-83-Cap-NLuc-Tav, hereafter VEEV-TC-83-nLuc) was a gift from Dr. William B. Klimstra (Department of Immunology, University of Pittsburgh) (Sun et al., 2014). The plasmid encoding the kinase domain of BIKE with N-terminal TEV (tobacco etch virus)-cleavable His6 tags was a gift from Dr. Stefan Knapp at the University of Oxford (Sorrell et al., 2016). The plasmids encoding WT and T156A AP2M1 mutant were previously reported (Bekerman et al., 2017).

2.3. Compounds.

5z-7-oxozeaenol was purchased from Cayman Chemical. ML336 was purchased from AOBIOUS. Compound 25A and 34A were synthesized by the Zuercher laboratory (UNC) (Agajanian et al., 2019). Compounds’ structure was defined by NMR and mass spectroscopy. ≥ 95% purity was confirmed by HPLC.

2.4. Western blotting and antibodies.

Cells were lysed in M-Per protein extraction reagent (Thermo Fisher Scientific). For detection of phospho-AP2M1, cells were pretreated with 100 nM calyculin A (Cell Signaling), a PP1 and PP2a phosphatase inhibitor, for 30 minutes prior to lysis. Clarified protein lysates were run on 4%–12% Bis-Tris gels (Invitrogen), transferred onto PVDF membranes (Bio-Rad). Blots were blocked and blotted with anti-BIKE (Santa Cruz biotechnology, catalog sc-134284), anti-AP2M1 (Abcam, catalog ab75995), anti-GLuc (New England BioLabs, catalog E8023S), anti–phospho-AP2M1 (T156) (Cell Signaling, catalog 3843S), and anti–β-actin (Sigma-Aldrich, catalog A3854) antibodies. Signal was detected with HRP-conjugated secondary antibodies. Band intensity was quantified with ImageJ software (NIH).

2.5. RNA interfering.

siGENOME Human BMP2K (55589) siRNA SMART Pool (GAACAUAGACCUGAUAUAU, GGACUGUGCUGUUAAUUCA, GGAACUAUGUACUUUGUGA, CGAUGUGCAUUGAAGCGAA) and siGENOME Non-Targeting siRNA Pool #1 (D-001206-13-05) (UAGCGACUAAACACAUCAA, UAAGGCUAUGAAGAGAUAC, AUGUAUUGGCCUGUAUUAG, AUGAACGUGAAUUGCUCAA) were purchased from Dharmacon. siRNAs (1 pmole) were transfected into Huh7 cells using RNAiMax (ThermoFisher Scientific) 72 hours before infection.

2.6. Generation of BIKE KO cell lines.

CRISPR guide RNA (gRNA) sequences were designed using the CRISPR design tool (http://chopchop.cbu.uib.no/). BIKE (GGTGGCGGCCGACCGCGAAC) sgRNA was synthesized and cloned into the pX458 gRNA plasmid (a gift from Dr. Feng Zhang, Addgene plasmid # 48138), as described (Ran et al., 2013). Single clonal knockout of Huh7 cells were obtained using the PX458 vector that expresses Cas9 and sgRNA against BIKE. Green fluorescent protein (GFP) positive single cells were sorted at 24 hours post-transfection using a BD InFlux Cell Sorter into 96-well plates and screened for knockout via western blot, as described (Ran et al., 2013).

2.7. Protein expression and purification.

Plasmid encoding N-terminally His6-tagged BIKE domain was electroporated into Rosetta strain BL21 E. coli cells using Gene Pulser Xcell™ Electroporation Systems (Bio-Rad) with 1800V. Protein expression was induced with 0.5 mM IPTG at 20 °C overnight. Cells were then harvested and resuspended in lysis buffer composed of 50 mM HEPES (pH 7.5), 500 mM NaCl, 5 mM imidazole, 5% glycerol, and 0.5 mM TCEP [tris-(2-carboxyethyl)phosphine]. Following sonication and spinning at 48,400 xg at 4 °C for 60 minutes the supernatant was harvested. BIKE was purified via Ni-affinity followed by TEV protease digestion to remove the HIS6 tag. BIKE was further purified by size-exclusion chromatography via Superdex S75 10/60 column on AKTA pure (GE Lifesciences). Protein was concentrated via 10 kDa Amicon centrifugal filters (Merck) and suspended in storage buffer composed of 10 mM HEPES (pH 7.5), 300 mM NaCl, 5% glycerol, and 0.5 mM TCEP at −80 °C.

2.8. Virus production.

DENV RNA was transcribed in vitro from pACYC-DENV2-NGC plasmid by mMessage/mMachine (Ambion) kits and electroporated into BHK-21 cells. EBOV (Kikwit isolate) was grown in Vero E6 cells. VEEV-TC-83-nLuc RNA was transcribed in vitro from cDNA plasmid templates linearized with MluI via MegaScript Sp6 kit (Invitrogen #AM1330) and electroporated into BHK-21 cells. Supernatants were collected, clarified and stored at −80°C until further use. Virus stock titers were determined via standard plaque assay on BHK-21 (DENV, VEEV) or Vero E6 (EBOV) cells, and titers were expressed as PFU/ml.

2.9. Infection assays.

Huh7, T-Rex-293, and MDDC cells were infected with DENV in replicates (n = 3-10) at an MOI of 0.01 or 0.05. Overall infection was measured at 48 or 72 hours using a Renilla luciferase substrate or a standard plaque assay. Huh7 cells were infected with EBOV at an MOI of 1 under biosafety level 4 conditions. 48 hours postinfection, supernatants were collected and stored at −80 °C. Cells were formalin-fixed for 24 hours prior to removal from biosafety level 4. Infected cells were detected using an EBOV glycoprotein-specific monoclonal antibody (KZ52) and quantitated by automated fluorescence microscopy using an Operetta High Content Imaging System and the Harmony software package (PerkinElmer). U-87 MG cells were infected with VEEV-TC83-nLuc in 8 replicates at MOI of 0.01. Overall infection was measured at 18 hpi via a nanoluciferase assay using a luciferin solution obtained from the hydrolysis of its O-acetylated precursor, hikarazine-103 (prepared by Dr. Yves Janin, Institut Pasteur, France) as a substrate (Coutant et al., 2019a, 2019b).

2.10. Viability Assays.

Viability was assessed using alamarBlue reagent (Invitrogen) or CellTiter-Glo reagent (Promega) assay according to manufacturer’s protocol. Fluorescence was detected at 560 nm on an InfiniteM1000 plate reader and luminescence on an InfiniteM1000 plate reader (Tecan) or a SpectraMax 340PC.

2.11. Entry assays.

Huh7 cells were infected with luciferase reporter DENV2. 2 hours postinfection, cells were lysed in Trizol, RNA was extracted and the intracellular viral RNA level was measured by qRT-PCR. Alternatively, Renilla luciferase activity was measured at 6 hours postinfection on InfiniteM1000 plate reader (Tecan).

2.12. DENV RNA replication.

As described (Bekerman et al., 2017; Wu et al., 2015; Yang et al., 2013), Huh7 cells were co-transfected with a reporter DNA-launched DENV2 replicon (pCMV-DV2Rep) that contains a minimal cytomegalovirus (CMVmin) promoter regulated by an upstream tetracycline response elements and a TET-ON plasmid (Wu et al., 2015; Yang et al., 2013). Eighteen hours posttransfection, viral RNA transcription was induced by doxycycline and shut down by changing to doxycycline-free medium 24 hours later, which allowed us to shorten the exposure to doxycycline. Replication was monitored by luciferase activity.

2.13. Extra- and intracellular infectivity.

Extracellular infectivity was measured in culture supernatants derived from 24-well plates transfected with 200 ng of DENV RNA for 48 hours and used to infect naive cells for 48 hours. Intracellular DENV infectivity was measured by inoculation of naive cells with lysates of transfected cells subjected to 3 rounds of free-thawing in ethanol-dry ice and clarified at 5,000 g, as previously described (Bekerman et al., 2017). Luciferase or standard plaque assay were used to quantify infectious virus.

2.14. Pharmacological inhibition.

Cells were treated with the inhibitors or DMSO at the time of inoculation. The inhibitors were left for the duration of the assay. Viral infection was measured via luciferase (DENV, VEEV) or plaque (DENV, EBOV) assays.

2.15. Gain-of-function assays.

Doxycycline-inducible cell lines were established to overexpress BIKE via the Flp-In™ recombination system (ThermoFisher). T-REx-293 cells with a pFRT/lacZeo site and pcDNA™6/TR were co-transfected with puromycin-resistant vector encoding Flp-In™ recombination target site and pOG44 plasmid containing Flp recombinase followed by puromycin selection. To overexpress BIKE in the doxycycline-inducible T-REx-293 cell lines, 1 μM of doxycycline was added to the medium and incubated for eight hours. Cells were then infected with DENV (MOI=0.05) and when relevant, treated with the compounds, followed by 72 hours incubation prior to luciferase and viability assays. WT or T156A AP2M1, BIKE, or empty vector controls were transfected into Huh7 cells. 24 hours posttransfection, cells were infected with luciferase reporter DENV (MOI =0.05) and incubated for 48 hours prior to luciferase and viability assays.

2.16. Resistance studies.

VEEV (TC-83) was used to inoculate U-87 MG cells at MOI of 0.1 and passaged every 24 hours by transferring of an equal volume of viral supernatant to naive cells under increasing drug selection (2.5-5 μM, passages 1–3; 5-10 μM, passages 4-7; 10-15 μM, passages 8-10). Upon completion of 10 passages, virus titer from the resulting supernatants was measured by plaque assays. ML336 resistance mutation in NSP2 at passage 10 was confirmed by purification and reverse transcription of viral RNA from cell supernatants, as described in the RNA extraction and quantification section. NSP2 region was amplified with iProof high-fidelity PCR kit (Bio-Rad) using the following primers: (forward: GATGTGGAYYTGAGACCAGCYTCAGCATGGAC; reverse: AGTCAANACTTCACAGAAAGCCCATGTTGTTCTCATCAA (N = any base, Y = C or T)) and sequenced (Sequetech Corp.).

2.17. Quantification and Statistical Analysis

All data were analyzed with GraphPad Prism software. Fifty percent effective concentrations (EC50) were measured by fitting of data to a 3-parameter logistic curve. P values were calculated by 1- or 2-way ANOVA with either Dunnett’s or Tukey’s multiple comparisons tests as specified in each figure legend.

3. Results

3.1. BIKE but not STK16 is required for DENV infection

To probe the functional relevance, we first monitored DENV infection in human hepatoma (Huh7) cells upon depletion of BIKE and STK16 by small interfering RNAs (ON-TARGETplus SMARTpool siRNAs [Dharmacon]) (Fig. 1A). siBIKE, but not siSTK16, reduced infection with DENV2 (NGC strain) expressing a luciferase reporter gene (Zou et al., 2011) by ~65% relative to non-targeting (NT) control, as measured via luciferase assays at 48 hours postinfection, with no apparent cytotoxic effect (Fig. 1B, C). A much greater effect was measured upon BIKE depletion via plaque assays with 2 and 1.5 log reduction of DENV titers in cells infected with an MOI of 0.01 or 0.05, respectively (Fig. 1D), with no cytotoxicity (Fig. 1E). siBIKE similarly suppressed DENV infection in a BIKE-depleted doxycycline-inducible T-REx-293 cell line (Torres et al., 2009). Induction of BIKE expression only partially restored the bulk cellular level of BIKE, yet it completely reversed the effect of siBIKE, largely excluding off-target effects as the cause of the observed phenotype, and suggesting that the local level of active BIKE achieved by overexpression is sufficient to drive its function in DENV infection (Fig. 1F and G). Moreover, induction of BIKE expression in the wild type (WT) T-REx-293 cell line increased DENV infection by ~8 fold relative to non-induced cells, supporting the requirement of BIKE for DENV infection and suggesting that BIKE is rate limiting for DENV infection (Fig. 1H and I). To confirm the role of BIKE in DENV infection, we generated two isogenic BIKE-knockout (BIKEKO) Huh7 cell lines (Fig. 1J). BIKE deletion reduced DENV infection by 3.5-4 fold relative to WT cells (Fig. 1K). Moreover, ectopic expression of BIKE in the BIKEKO cells completely restored or even increased DENV infection, confirming that the observed phenotype resulted from BIKE deletion (Figs. 1J and K). Taken together, these loss-of-function and gain-of-function studies indicate that BIKE is required for DENV infection.

Fig 1. BIKE is required for DENV infection.

Fig 1.

Open in a new tab

(A, F) Confirmation of gene expression knockdown via Western blot in Huh7 (A) and Dox-inducible BIKE-overexpressing T-REx-293 cells (F) transfected with the indicated siRNAs and in these cells upon DOX induction (F). Representative membranes are shown. Numbers represent means +/− SD of protein-to-actin ratio relative to NT control from 3 membranes.

(B-G) Infection (B, D, G) and cellular viability (C, E) measured at 48 hours post-inoculation of the indicated cells with luciferase reporter DENV2 via luciferase (MOI=0.05) (B, G) or plaque assays (MOI=0.01, 0.05) (D) and alamarBlue assays (C, E), respectively. (H, I) BIKE protein (H) and DENV2 infection measured in the stable Dox-inducible BIKE-overexpressing T-REx-293 cell line via luciferase assays 48 hours post-inoculation (MOI=0.05) (I) in the presence or absence of DOX. (J, K) BIKE protein by Western blot (J) and DENV2 infection measured via luciferase assays 48 hours postinoculation (MOI=0.05) (K) of Huh7 cell lines deleted for BIKE via CRISPR/Cas9 and in these cells upon complementation with BIKE-FLAG (BIKE). Samples in the two right panels in H were run on the same gel from which several lanes were cut out. Shown are means +/− SD of results of representative experiments out of two (E, G, I) or three (B, C) conducted each with five replicates. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001 relative to corresponding controls by one-way ANOVA with Dunnett’s (B, C) or Tukey’s (G, K) post-hoc tests or two-tailed unpaired student t-test (D, E, I).

3.2. BIKE is required for both early and late stages of the DENV life cycle.

We next monitored distinct stages in the DENV life cycle in Huh7 cells depleted or deleted of BIKE. We first determined the effect of BIKE depletion on early stages of the DENV life cycle via luciferase assays 6 hours following infection with the luciferase reporter DENV2 (Zou et al., 2011). siBIKE reduced DENV infection by ~50% relative to NT control in this assay (Fig. 1A and 2A). Consistent with these findings, BIKE deletion reduced DENV infection at 6 hours by over 70% in the BIKEKO cell lines (Fig. 1J and 2A). Treatment with a translation inhibitor (cycloheximide), but not a DENV RNA replication inhibitor (SDM25N (van Cleef et al., 2013)) suppressed viral infection at 6 hours postinfection, indicating that this assay monitors DENV entry through translation (data not shown), thereby excluding a role for BIKE in DENV RNA replication. To further pinpoint the specific stage, we thus measured DENV internalization via RT-PCR at 2 hours post DENV inoculation. BIKE deletion had no effect on DENV internalization (Fig. 2B). To probe for a potential role in viral translation, next we monitored translation and RNA replication by luciferase assays at 24 and 72 hours following induction of DENV subgenomic replicon (pCMV-DV2Rep) expression in Huh7 cells (Fig. 2C) (Yang et al., 2013). Neither DENV translation nor DENV RNA replication were affected by BIKE depletion (Fig. 2C). To assess for an effect in later stages of the viral life cycle, we transfected full length DENV RNA into Huh7 cells, obtained clarified cell lysates and culture supernatants at 48 hours, and used these to inoculate naïve cells followed by luciferase or plaque assays at 48 hours. A 2.5-4 fold reduction in both intra- and extracellular infectivity was observed in BIKE depleted versus control cells, suggesting a defect in DENV assembly and egress (Fig. 2D and E). The level of intracellular NS1 and DENV RNA in cell lysates derived from the BIKE-depleted DENV RNA transfected cells was comparable to NT controls, further excluding a role for BIKE in DENV translation and RNA replication (Fig. 2F). Our attempts to monitor RNA replication and infectious virus production in the BIKEKO cell lines were limited by low efficiency of DENV RNA transfection into these cells. These results implicate BIKE in the DENV life cycle via regulation of two temporally distinct steps, an early post-internalization and pre-translation step and infectious virus production.

Fig 2. BIKE is required for an early (postinternalization) and late (assembly/egress) stages of the DENV life cycle.

Fig 2.

Open in a new tab

(A) Early stages monitored by luciferase activity at 6 hours postinoculation of Huh7 cells depleted (left) or deleted (right) for BIKE with luciferase reporter DENV2 (MOI=5). (B) DENV entry measured in BIKE deleted cells via RT-PCR assays 2 hours following viral inoculation. (C) DENV RNA replication monitored by luciferase activity at 24 and 72 hours following co-transfection of Huh7 cells with a Tet-inducible DNA-launched DENV replicon along with TET-ON plasmid and induction by doxycycline for 24 hours (GND is a replication-incompetent DENV). Data are normalized to signal at 24 hours postinduction. (D) DENV infectivity measured via luciferase or plaque assays (E) by inoculating naïve cells with lysates (intracellular) and supernatants (extracellular) derived from Huh7 cells transfected with in vitro transcribed DENV2 RNA 48 hours post-transfection. (F) Intracellular levels of NS1 expression and DENV RNA copy number in BIKE depleted and control cells via Western blot and qRT-PCR assays. Shown are means +/− SD of results of representative experiments out of three conducted each with four replicates. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001 relative to NT or WT controls by (one-way (A, B, D, E) or two-way (C) ANOVA with Dunnett’s post-hoc test or two-tailed unpaired student t-test (F). ns; non-significant, PFU; plaque-forming units.

3.3. BIKE mediates its role in DENV infection in part via AP2M1 phosphorylation at T156

We have previously shown that AP2M1 is required for DENV entry and infectious virus production (Bekerman et al., 2017; Neveu et al., 2015, 2012; Xiao et al., 2018). While recombinant BIKE (rBIKE) was described to phosphorylate a synthetic AP2M1 peptide (Sorrell et al., 2016), the role of AP2M1 as a functional BIKE substrate was not established. To confirm AP2M1 phosphorylation, rBIKE (Fig. 3A) was incubated for 1 hour with ATP and an AP2M1 peptide harboring T156, a residue previously shown to be modified by AAK1 (Ricotta et al., 2002) and GAK (Tabara et al., 2011), followed by Electrospray Ionisation Mass Spectrometry (ESI-MS). A 79.9 Da shift from a molecular weight of 1903 Da, corresponding to the unphosphorylated AP2M1 peptide, to 1982.9 Da was detected, indicating the addition of a single phosphate group by rBIKE (Fig. 3B). To confirm these findings, we measured the level of phospho-AP2M1 upon BIKE depletion. A significant reduction in the phospho-AP2M1, but not total AP2M1 level, was observed in Huh7 cells depleted for BIKE or treated with 10 μM of 5Z-7-oxozeaenol, a pan-NAK inhibitor, relative to NT control (Fig. 3C), confirming that BIKE phosphorylates endogenous, full length AP2M1 in cultured cells. To determine if AP2M1 is a mediator of BIKE activity in DENV infection, we conducted gain-of-function assays. Ectopic expression of WT but not T156A AP2M1 mutant or control vector reversed the antiviral effect of siBIKE (Fig. 3D and 3E). These results indicate that AP2M1 T156 phosphorylation by BIKE is functionally relevant and represents a mechanism through which BIKE regulates DENV infection.

Fig 3. BIKE mediates DENV infection in part through phosphorylation of AP2M1 at T156.

Fig 3.

Open in a new tab

(A) rBIKE by gel filtration following expression in bacterial cells. Representative fractions are shown. Molecular weight markers are indicated on the left (kDa). (B) Mass spectrum of the synthetic AP2M1 peptide following incubation with rBIKE and ATP. (C) Effect of the indicated siRNAs or treatment with 10 μM of 5Z-7-oxozeaenol on AP2M1 phosphorylation in Huh7 cells measured by Western blotting. (D) Infection measured at 48 hours after inoculation (MOI=0.05) of the indicated cells with luciferase reporter DENV2 via luciferase assays. (E) AP2M1-GLuc expression in Huh7 cells transfected with the indicated siRNAs.***, P<0.001 by one-way ANOVA with Tukey’s post-hoc tests.

3.4. A selective AAK1/BIKE inhibitor suppresses DENV infection and AP2M1 phosphorylation in vitro

To determine whether a similar effect on DENV infection to that achieved genetically can be achieved pharmacologically and further validate BIKE as an antiviral target, we treated DENV-infected cells with NAK inhibitors. Compound 25A is a highly selective, potent anti-BIKE and anti-AAK1 kinase inhibitor (dissociation constants (kD) of 68 nM and 71 nM, respectively) that was developed as a chemical probe (Agajanian et al., 2019) (Fig. 4A). We measured a dose-dependent inhibition of DENV2 infection in Huh7 cells following a 2-day treatment with compound 25A with half-maximal effective concentration (EC50) of 1.34 μM by luciferase assays and half-maximal cellular cytotoxicity (CC50) greater than 10 μM by alamarBlue assays in the same samples (Fig. 4B). In contrast, compound 34A, designed as an inactive 25A derivative, demonstrated no antiviral activity (Fig. 4B). To confirm that 25A inhibits phosphorylation of the AAK1 and GAK ligand AP2M1, we measured levels of phospho-AP2M1 upon drug treatment. A ~10-fold reduction in the phospho-AP2M1 to total AP2M1 ratio was measured in Huh7 cells upon treatment with compound 25A but not 34A (Fig. 4C), confirming that the antiviral activity is correlated with functional inhibition of BIKE and AAK1.

Fig 4. A selective BIKE/AAK1 inhibitor inhibits DENV infection and its mechanism of antiviral action is in part mediated by BIKE.

Fig 4.

Open in a new tab

(A) Chemical structures of 25A and 34A. (B) Dose response of DENV infection (blue) and cellular viability (black) to the indicated compounds measured in Huh7 cells. (C) Effect of 1-hour treatment with 25A or 34A at the indicated concentrations on AP2 phosphorylation in Huh7 cells measured by Western blotting. Numbers indicate the ratio of phospho-AP2 (pAP2) to total AP2. Samples were run on the same gel, from which several lanes were cut out. Data in all panels are representative of 2 or more independent experiments. Individual experiments in B had 8–10 biological replicates, means ± SD are shown.

3.5. 5Z-7-oxozeaenol inhibits viral infection in vitro and its mechanism of antiviral action is in part mediated by BIKE

5Z-7-oxozeaenol is a natural product acting as an ATP-competitive irreversible inhibitor of ERK2 (IC50=80 nM), TAK1 (IC50=8 nM) and VEGF-R2 (IC50=52 nM) (Fig. 5A) (Ninomiya-Tsuji et al., 2003; Wu et al., 2013). 5Z-7-oxozeaenol potently binds BIKE, GAK, and AAK1 at 3.8, 10, and 13 % of control, respectively, at 10 μM via a publicly available KINOMEscan (DiscoverX, Harvard Medical School LINCS Dataset (ID:20211)). We measured a dose-dependent inhibition of DENV2 infection in Huh7 cells following a 2-day drug treatment with 5Z-7-oxozeaenol with an EC50 of 2.47 μM and a CC50 of 10 μM and a dramatic reduction in AP2M1 phosphorylation (Fig. 5B and C). Next, we conducted gain-of-function assays, to determine whether inhibition of BIKE is a mechanism underlying the anti-DENV effect of this drug. Following doxycycline-mediated induction of BIKE expression, T-REx 293 cells were infected with a luciferase reporter DENV and treated with 5Z-7-oxozeaenol for 72 hours prior to luciferase and viability assays. DENV infection was partially restored by BIKE expression upon doxycycline induction in 5Z-7-oxozeaenol treated BIKE-inducible T-REx-293 cells (Fig. 5D and 1E). These results pharmacologically validate BIKE as a regulator of DENV infection and as a molecular target that mediates the antiviral effect of these compounds.

Fig 5. 5Z-7-oxozeaenol inhibits DENV infection and its mechanism of antiviral action is in part mediated by BIKE.

Fig 5.

Open in a new tab

(A) Chemical structure of 5Z-7-oxozeaenol. (B) Dose response of DENV infection (blue) and cellular viability (black) to the indicated compounds measured in Huh7 cells. (C) Effect of 1-hour treatment with 5Z-7-oxozeaenol at the indicated concentrations on AP2 phosphorylation in Huh7 cells measured by Western blotting. Numbers indicate the ratio of phospho-AP2 (pAP2) to total AP2. (D) DENV infection normalized to cellular viability in T-REx 293 cells induced with doxycycline to express BIKE or uninduced cells 48 hours postinfection with a luciferase reporter DENV and treatment with 5Z-7-oxozeaenol. ***, P<0.001 relative to uninduced controls at the same concentration by two-way ANOVA with Tukey’s post hoc test. Data in all panels are representative of 2 or more independent experiments. Individual experiments in B and D had 8–10 biological replicates, means ± SD are shown.

3.6. 5Z-7-oxozeaenol inhibits DENV infection ex vivo and demonstrates a broad-spectrum antiviral potential

To further evaluate the therapeutic potential of 5Z-7-oxozeaenol, we studied its antiviral effect in primary human monocyte derived dendritic cells (MDDC); an established ex vivo model system for DENV (Rodriguez-Madoz et al., 2010). We measured a dose-dependent inhibition of DENV infection following a 3-day drug treatment with an EC50 of 0.54 μM by luciferase assays and a CC50 of 5.49 μM by alamarBlue assays (Fig. 6A). Lastly, we investigated whether members of unrelated viral families, Ebola virus (EBOV), a filovirus, and Venezuelan equine encephalitis virus (VEEV), an alphavirus are also susceptible to 5Z-7-oxozeaenol treatment. 5Z-7-oxozeaenol treatment resulted in a dose-dependent reduction in authentic EBOV infection in Huh7 with EC50 value of 4.09 μM and no apparent cellular toxicity within the concentration range tested (Fig. 6B). Similarly, 5Z-7-oxozeaenol treatment inhibited infection of U-87 MG (human astrocytes) with TC-83 (a live-attenuated vaccine VEEV strain) in a dose-dependent manner, as measured by luciferase assays 18 hours postinfection with a nano-luciferase reporter virus, with EC50 value of 2.6 μM and CC50 greater than 20 μM (Fig. 6C).

Fig 6. 5Z-7-oxozeaenol inhibits DENV infection in MDDCs as well as EBOV and VEEV (TC-83).

Fig 6.

Open in a new tab

(A) Dose response of DENV infection (MOI=0.05) (blue) and cellular viability (black) to 5Z-7-oxozeaenol measured in human primary MDDCs via luciferase and alamarBlue assays, respectively. (B) Dose response of EBOV infection (MOI=1) (blue) and cellular viability (black) to 5Z-7-oxozeaenol measured in Huh7 cells under biosafety level 4 containment 48 hours postinfection (MOI=1) via plaque assay and CellTiter-Glo luminescent cell viability assay, respectively. (C) Dose response of VEEV (TC-83) infection (MOI=0.1) (blue) and cellular viability (black) to 5Z-7-oxozeaenol measured in U-87 MG cells via luciferase and alamarBlue assays, respectively. Data in all panels are representative of 2 or more independent experiments. Individual experiments had 5 biological replicates, means ± SD are shown. Shown in A is a representative experiment with cells from a single donor, of two experiments conducted with cells derived from two donors.

3.7. 5Z-7-oxozeaenol has a high genetic barrier to resistance

To determine whether viruses can escape treatment with 5Z-7-oxozeaenol, we chose to focus on VEEV (TC-83) since it has the shortest life cycle among the viruses tested. VEEV was passaged in the presence of 5Z-7-oxozeaenol or the VEEV nonstructural 2p (ns2P) protein inhibitor ML336 (Chung et al. 2020) at increasing concentrations (2.5–15 μM) corresponding to values between EC50 and EC90 in U-87 MG cells. Infectious virus output was quantified over several passages by plaque assays. By passage 3, VEEV overcame inhibition by ML336. In contrast, VEEV remained suppressed for 10 passages under the 5Z-7-oxozeaenol treatment without any phenotypic resistance (Fig. 7A). Moreover, virus from culture supernatants obtained at passage 10 under 5Z-7-oxozeaenol or DMSO treatment remained susceptible to 5Z-7-oxozeaenol . In contrast, virus obtained at passage 10 under ML336 treatment lost its susceptibility to ML336, with the emergence of a previously characterized resistance mutation in NSP2 (Y102C in VEEV TC-83), whereas virus obtained at the same passage under DMSO treatment remained susceptible to ML336 (Fig. 7B, 7C). These results point to 5Z-7-oxozeaenol as a potential broad-spectrum antiviral strategy with a higher relative barrier to resistance than a direct-acting antiviral.

Fig 7. 5Z-7-oxozeaenol has a high genetic barrier to resistance.

Fig 7.

Open in a new tab

(A) VEEV (TC-83) was used to infect U-87 MG cells and passaged every 24 hours by inoculation of naive cells with equal volumes of viral supernatants under DMSO treatment or selection with 5Z-7-oxozeaenol or ML336 (VEEV ns2P inhibitor) increasing from 2.5 to 15 μM over 10 passages. Viral titers were measured by plaque assays. (B and C) Dose response to 5Z-7-oxozeaenol (B) and ML336 (C) of VEEV harvested after 10 passages in U-87 MG cells in the presence of 5Z-7-oxozeaenol (B) and ML336 (C), via luciferase assays. Retained susceptibility to 5Z-7-oxozeaenol and loss of susceptibility to ML336 are shown. Data are representative of at least 2 experiments. Shown are means ± SD. Individual experiments in A and B/C had 2 and 3 biological replicates, respectively.

4. Discussion

We have previously demonstrated regulation of intracellular trafficking of various RNA viruses via AAK1- and GAK-mediated phosphorylation of AP complexes and validated these kinases as targets for broad-spectrum antivirals (Bekerman et al., 2017; Neveu et al., 2015; Pu et al., 2018; Verdonck et al., 2019; Xiao et al., 2018). Yet, the roles of BIKE in healthy and disease states are largely unknown. By integrating molecular virology, genetic, and pharmacological approaches, we reveal a requirement for BIKE in DENV infection and demonstrate that AP2M1 is a substrate that is involved in mediating BIKE’s role. Moreover, we provide a proof of concept that pharmacological inhibition of BIKE can be potentially used as a broad-spectrum strategy against acute emerging RNA viral infections, and we validate BIKE as a molecular target mediating this antiviral activity. These findings provide insights into the virus-host determinants that regulate DENV infection and have potential implications for the design of antiviral strategies.

BIKE was previously implicated in osteoblast differentiation, myopia, and cancer (Buraschi et al., 2012; Kearns et al., 2001; Liu et al., 2009; Mercado-Matos et al., 2018). It was also shown to be required for HIV infection via a genome-wide RNAi screen (Zhou et al., 2008), however, the substrates that mediate this effect, the relevant stage of the viral life cycle, and BIKE’s role in other viral infections have not been studied. We provide evidence that BIKE mediates at least two temporally distinct stages in the DENV life cycle: an early postinternalization but pretranslation stage (such as trafficking from the plasma membrane to early endosomes, uncoating, or viral RNA trafficking to sites of RNA replication) and a later, assembly/egress stage. Like AAK1 and GAK, BIKE therefore represents a regulator of viral infection and a candidate druggable target for antiviral treatment. We predict that a lower dynamic range and partial overlap in substrates of the various NAKs (e.g. AP2M1) account for the relatively modest effect of BIKE depletion or deletion measured on DENV infection via luciferase assays. Indeed, a dramatic effect was measured in BIKE depleted cells via plaque assays and upon treatment with NAK inhibitors. The requirement for BIKE by unrelated RNA viral families suggests that it may be a target for broad-spectrum antivirals.

To date, no functionally proven substrates of BIKE have been discovered. While BIKE was shown to bind Numb and to phosphorylate a synthetic AP2M1 peptide, neither of these endocytic adaptors were shown to be biologically relevant BIKE substrates (Kearns et al., 2001; Krieger et al., 2013; Sorrell et al., 2016). We show that AP2M1 phosphorylation on T156 is biologically relevant via two cell-based assays and that it is involved in mediating BIKE’s role in DENV infection. Since we have previously shown that AP2M1 is required for early and late steps in the DENV life cycle (Bekerman et al., 2017; Neveu et al., 2015, 2012), our current findings suggest that BIKE mediates its role in these steps in part via AP2M1 phosphorylation. We, however, cannot exclude the possibility that additional BIKE substrates exist and are involved in mediating BIKE’s role in viral infections.

We show that 5Z-7-oxozeaenol, a potent, albeit non-selective inhibitor of NAKs, and 25A, a highly selective BIKE and AAK1 inhibitor, restrict DENV infection in vitro. The IC50 or KD values of 5Z-7-oxozeaenol and 25A for BIKE inhibition or binding are 3.8 nM and ~70 nM, respectively, whereas their EC50 values for DENV infection are 0.52-2.47 μM. Such ∼100–500-fold differences between IC50 values measured in vitro and EC50 values measured via cell-based assays are typical for kinase inhibitors and are often attributed to the absence of ATP competition in the in vitro assay, to differences in the fraction of substrate that is consumed under the given assay conditions, to the relative concentration and cellular distribution of the kinase, and to the intracellular bioavailability (determined by the rates of entry to the cells and/or active pumping out of the cells) of the compounds (Bekerman et al., 2017; Knight and Shokat, 2005). We hypothesize that optimization of the compounds via structure-activity relationship analysis will improve their potency as antiviral agents. Dendritic cells represent the primary target of DENV in humans and better model human physiology and disease than immortalized cell lines (Schmid et al., 2014). Our finding that 5Z-7-oxozeaenol treatment exhibits antiviral efficacy in MDDCs with minimal toxicity to host cells therefore further supports the biological relevance of this approach.

We provide evidence that modulation of NAK activity is an important mode of antiviral action of these compounds in DENV infection. Due to lack of affinity of 25A to 5Z-7-oxozeaenol’s cancer targets, the activity of 25A confirms that BIKE and AAK1 are relevant antiviral targets. PDGFR, the only other target significantly bound by 25A via KINOME scan (Agajanian et al., 2019), has no apparent role in DENV infection (Bekerman et al., 2017). Furthermore, we characterized the mechanism by which pharmacological inhibitors of NAKs mediate their anti-DENV effect. We establish that the antiviral activity of 5Z-7-oxozeaenol and 25A correlates with reduced phospho-AP2M1 levels and is mechanistically explained in part by a block in AP2M1 phosphorylation. These findings also present phosphorylated AP2M1 as a useful pharmacodynamic biomarker. Our gain-of-function assays support inhibition of BIKE activity as an important mode of antiviral action of 5Z-7-oxozeaenol. Nevertheless, we cannot exclude the possibility that other targets of 5Z-7-oxozeaenol are involved in mediating this antiviral effect.

Most antivirals inhibit viral enzymes and are thus limited by a narrow-spectrum of coverage and a rapid emergence of drug resistance (Bekerman and Einav, 2015). We predicted that since BIKE is a cellular target and was previously found to be essential for HIV replication, compounds with anti-BIKE activity may have broad-spectrum activity beyond Flaviviridae. Indeed, we reveal activity against the filovirus, EBOV and the alphavirus, VEEV (TC-83). Targeting cellular BIKE subnetworks may thus provide a path for broad-spectrum antivirals. Furthermore, targeting of cellular proteins that are not under the genetic control of viruses is more likely to have a higher barrier to resistance than classical direct-acting antivirals. This is exemplified by our current data and prior data with sunitinib/erlotinib combinations (Bekerman et al., 2017). Although we predict that the genetic barrier to resistance of 5Z-7-oxozeaenol is high, it may be possible to select for resistance over longer-term passage under different conditions or in a different, chronic infection model. The strategies used by viruses to overcome drug-mediated inhibition are similar to those used by cancer cells. Simultaneous inhibition of several kinases or targeting of several pathways by the same drug or drug combination may thus prove attractive in combating viral pathogens, as reported in cancer (Knight et al., 2010). The “polypharmacology” provided by a drug such as 5Z-7-oxozeaenol could therefore increase the effectiveness while minimizing viral resistance.

Covalent kinase inhibitors are widely used in cancer treatment (Barf and Kaptein, 2012), yet their potential in treating viral infections has not been explored. We provide a proof of concept that covalent kinase inhibitors may have utility in combating acute viral infections, an attractive indication requiring a shorter duration of treatment. Notably, E6201, a closely related 5Z-7-oxozeaenol derivative, is undergoing clinical development for cancer, proposing a repurposing potential (Barf and Kaptein, 2012; Tibes et al., 2018).

Taken together, these results validate virus-host interactions required for DENV entry and assembly/release and reveal a novel druggable cellular target with implications for the design of host-targeted, broad-spectrum antiviral strategies with a high barrier to resistance.

Acknowledgements:

This work was supported by award number 1U19AI10966201 (CETR) from the National Institute of Allergy and Infectious Diseases (NIAID), award number W81XWH-16-1-0691 from the Department of Defense (DoD), Congressionally Directed Medical Research Programs (CDMRP), award number HDTRA11810039 from the Defense Threat Reduction Agency (DTRA)/Fundamental Research to Counter Weapons of Mass Destruction and grants from the Stanford SPARK program to S.E.. S.P. was supported by the Maternal and Child Health Research Institute, Lucile Packard Foundation for Children’s Health, as well as the Stanford CTSA (grant number UL1 TR000093). The opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the U.S. Army or the other funders.

References:

  1. Agajanian MJ, Walker MP, Axtman AD, Ruela-de-Sousa RR, Serafin DS, Rabinowitz AD, Graham DM, Ryan MB, Tamir T, Nakamichi Y, Gammons MV, Bennett JM, Couñago RM, Drewry DH, Elkins JM, Gileadi C, Gileadi O, Godoi PH, Kapadia N, Müller S, Santiago AS, Sorrell FJ, Wells CI, Fedorov O, Willson TM, Zuercher WJ, Major MB, 2019. WNT Activates the AAK1 Kinase to Promote Clathrin-Mediated Endocytosis of LRP6 and Establish a Negative Feedback Loop. Cell Rep. 26, 79–93.e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barf T, Kaptein A, 2012. Irreversible Protein Kinase Inhibitors: Balancing the Benefits and Risks. Journal of Medicinal Chemistry. [DOI] [PubMed] [Google Scholar]
  3. Bekerman E, Einav S, 2015. Infectious disease. Combating emerging viral threats. Science 348, 282–283. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bekerman E, Neveu G, Shulla A, Brannan J, Pu S-Y, Wang S, Xiao F, Barouch-Bentov R, Bakken RR, Mateo R, Govero J, Nagamine CM, Diamond MS, De Jonghe S, Herdewijn P, Dye JM, Randall G, Einav S, 2017. Anticancer kinase inhibitors impair intracellular viral trafficking and exert broad-spectrum antiviral effects. J. Clin. Invest 127, 1338–1352. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bhatt S, Gething PW, Brady OJ, Messina JP, Farlow AW, Moyes CL, Drake JM, Brownstein JS, Hoen AG, Sankoh O, Myers MF, George DB, Jaenisch T, Wint GR, Simmons CP, Scott TW, Farrar JJ, Hay SI, 2013. The global distribution and burden of dengue. Nature 496, 504–507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Buraschi S, Neill T, Owens RT, Iniguez LA, Purkins G, Vadigepalli R, Evans B, Schaefer L, Peiper SC, Wang Z-X, Iozzo RV, 2012. Decorin Protein Core Affects the Global Gene Expression Profile of the Tumor Microenvironment in a Triple-Negative Orthotopic Breast Carcinoma Xenograft Model. PLoS ONE. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chung D, Schroeder CE, Sotsky J, Yao T, Roy S, Smith RA, Tower NA, Noah JW, McKellip S, Sosa M, Rasmussen L, White EL, Aube J, Golden JE 2010. ML336: Development of Quinazolinone-Based Inhibitors Against Venezuelan Equine Encephalitis Virus (VEEV) In Probe Reports from the NIH Molecular Libraries Program (National Center for Biotechnology Information (US)). [PubMed] [Google Scholar]
  8. Conner SD, Schmid SL, 2002. Identification of an adaptor-associated kinase, AAK1, as a regulator of clathrin-mediated endocytosis. J. Cell Biol 156, 921–929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Coutant EP, Gagnot G, Hervin V, Baatallah R, Goyard S, Jacob Y, Rose T, Janin Y, 2019a. Bioluminescence Profiling of NanoKAZ/NanoLuc Luciferase Using a Chemical Library of Coelenterazine Analogues. Chemistry – A European Journal. [DOI] [PubMed] [Google Scholar]
  10. Coutant EP, Goyard S, Hervin V, Gagnot G, Baatallah R, Jacob Y, Rose T, Janin YL, 2019b. Gram-scale synthesis of luciferins derived from coelenterazine and original insights into their bioluminescence properties. Org. Biomol. Chem 17, 3709–3713. [DOI] [PubMed] [Google Scholar]
  11. Diamond MS, Pierson TC, 2015. Molecular Insight into Dengue Virus Pathogenesis and Its Implications for Disease Control. Cell 162, 488–492. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Huang CY-H, Butrapet S, Moss KJ, Childers T, Erb SM, Calvert AE, Silengo SJ, Kinney RM, Blair CD, Roehrig JT, 2010. The dengue virus type 2 envelope protein fusion peptide is essential for membrane fusion. Virology 396, 305–315. [DOI] [PubMed] [Google Scholar]
  13. Huttlin EL, Ting L, Bruckner RJ, Gebreab F, Gygi MP, Szpyt J, Tam S, Zarraga G, Colby G, Baltier K, Dong R, Guarani V, Vaites LP, Ordureau A, Rad R, Erickson BK, Wuhr M, Chick J, Zhai B, Kolippakkam D, Mintseris J, Obar RA, Harris T, Artavanis-Tsakonas S, Sowa ME, De Camilli P, Paulo JA, Harper JW, Gygi SP, 2015. The BioPlex Network: A Systematic Exploration of the Human Interactome. Cell 162, 425–440. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. In JG, Striz AC, Bernad A, Tuma PL, 2014. Serine/threonine kinase 16 and MAL2 regulate constitutive secretion of soluble cargo in hepatic cells. Biochemical Journal. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kearns AE, Donohue MM, Sanyal B, Demay MB, 2001. Cloning and characterization of a novel protein kinase that impairs osteoblast differentiation in vitro. J. Biol. Chem 276, 42213–42218. [DOI] [PubMed] [Google Scholar]
  16. Knight ZA, Shokat KM, 2005. Features of Selective Kinase Inhibitors. Chemistry & Biology. [DOI] [PubMed] [Google Scholar]
  17. Knight ZA, Lin H, Shokat KM, 2010. Targeting the cancer kinome through polypharmacology. Nat. Rev. Cancer 10, 130–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Krieger JR, Taylor P, Gajadhar AS, Guha A, Moran MF, McGlade CJ, 2013. Identification and selected reaction monitoring (SRM) quantification of endocytosis factors associated with Numb. Mol. Cell. Proteomics 12, 499–514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kudelko M, Brault J-B, Kwok K, Li MY, Pardigon N, Peiris JSM, Bruzzone R, Desprès P, Nal B, Wang PG, 2012. Class II ADP-ribosylation factors are required for efficient secretion of dengue viruses. J. Biol. Chem 287, 767–777. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Liu H-P, Lin Y-J, Lin W-Y, Wan L, Sheu JJ-C, Lin H-J, Tsai Y, Tsai C-H, Tsai F-J, 2009. A novel genetic variant of BMP2K contributes to high myopia. J. Clin. Lab. Anal 23, 362–367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. López-Coral A, Striz AC, Tuma PL, 2018. A Serine/Threonine Kinase 16-Based Phospho-Proteomics Screen Identifies WD Repeat Protein-1 As A Regulator Of Constitutive Secretion. Sci. Rep 8, 13049. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Mercado-Matos J, Janusis J, Zhu S, Chen SS, Shaw LM, 2018. Identification of a Novel Invasion-Promoting Region in Insulin Receptor Substrate 2. Mol. Cell. Biol 38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Messina JP, Brady OJ, Golding N, Kraemer MUG, Wint GRW, Ray SE, Pigott DM, Shearer FM, Johnson K, Earl L, Marczak LB, Shirude S, Davis Weaver N, Gilbert M, Velayudhan R, Jones P, Jaenisch T, Scott TW, Reiner RC Jr, Hay SI, 2019. The current and future global distribution and population at risk of dengue. Nat Microbiol. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Neveu G, Barouch-Bentov R, Ziv-Av A, Gerber D, Jacob Y, Einav S, 2012. Identification and Targeting of an Interaction between a Tyrosine Motif within Hepatitis C Virus Core Protein and AP2M1 Essential for Viral Assembly. PLoS Pathog. 8, e1002845. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Neveu G, Ziv-Av A, Barouch-Bentov R, Berkerman E, Mulholland J, Einav S, 2015. AP-2-associated protein kinase 1 and cyclin G-associated kinase regulate hepatitis C virus entry and are potential drug targets. J. Virol 89, 4387–4404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Ninomiya-Tsuji J, Kajino T, Ono K, Ohtomo T, Matsumoto M, Shiina M, Mihara M, Tsuchiya M, Matsumoto K, 2003. A resorcylic acid lactone, 5Z-7-oxozeaenol, prevents inflammation by inhibiting the catalytic activity of TAK1 MAPK kinase kinase. J. Biol. Chem 278, 18485–18490. [DOI] [PubMed] [Google Scholar]
  27. Oughtred R, Stark C, Breitkreutz B-J, Rust J, Boucher L, Chang C, Kolas N, O’Donnell L, Leung G, McAdam R, Zhang F, Dolma S, Willems A, Coulombe-Huntington J, Chatr-Aryamontri A, Dolinski K, Tyers M, 2019. The BioGRID interaction database: 2019 update. Nucleic Acids Res. 47, D529–D541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Pu S-Y, Wouters R, Schor S, Rozenski J, Barouch-Bentov R, Prugar LI, O’Brien CM, Brannan JM, Dye JM, Herdewijn P, De Jonghe S, Einav S, 2018. Optimization of Isothiazolo[4,3- b]pyridine-Based Inhibitors of Cyclin G Associated Kinase (GAK) with Broad-Spectrum Antiviral Activity. J. Med. Chem 61, 6178–6192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ricotta D, Conner SD, Schmid SL, von Figura K, Honing S, 2002. Phosphorylation of the AP2 mu subunit by AAK1 mediates high affinity binding to membrane protein sorting signals. J. Cell Biol 156, 791–795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Rodriguez-Madoz JR, Bernal-Rubio D, Kaminski D, Boyd K, Fernandez-Sesma A, 2010. Dengue virus inhibits the production of type I interferon in primary human dendritic cells. J. Virol 84, 4845–4850. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sato J, Shimizu H, Kasama T, Yabuta N, Nojima H, 2009. GAK, a regulator of clathrin-mediated membrane trafficking, localizes not only in the cytoplasm but also in the nucleus. Genes Cells 14, 627–641. [DOI] [PubMed] [Google Scholar]
  32. Schmid MA, Diamond MS, Harris E, 2014. Dendritic cells in dengue virus infection: targets of virus replication and mediators of immunity. Front. Immunol 5, 647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sorensen EB, Conner SD, 2008. AAK1 Regulates Numb Function at an Early Step in Clathrin-Mediated Endocytosis. Traffic 9, 1791–1800. [DOI] [PubMed] [Google Scholar]
  34. Sorrell FJ, Szklarz M, Abdul Azeez KR, Elkins JM, Knapp S, 2016. Family-wide Structural Analysis of Human Numb-Associated Protein Kinases. Structure 24, 401–411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Stark C, Breitkreutz B-J, Reguly T, Boucher L, Breitkreutz A, Tyers M, 2006. BioGRID: a general repository for interaction datasets. Nucleic Acids Res. 34, D535–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Sun C, Gardner CL, Watson AM, Ryman KD, Klimstra WB, 2014. Stable, high-level expression of reporter proteins from improved alphavirus expression vectors to track replication and dissemination during encephalitic and arthritogenic disease. J. Virol 88, 2035–2046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Tabara H, Naito Y, Ito A, Katsuma A, Sakurai MA, Ohno S, Shimizu H, Yabuta N, Nojima H, 2011. Neonatal lethality in knockout mice expressing the kinase-dead form of the gefitinib target GAK is caused by pulmonary dysfunction. PLoS One 6, e26034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Tibes R, Borad MJ, Dutcus CE, Reyderman L, Feit K, Eisen A, Verbel DA, Von Hoff DD, 2018. Safety, pharmacokinetics, and preliminary efficacy of E6201 in patients with advanced solid tumours, including melanoma: results of a phase 1 study. Br. J. Cancer 118, 1580–1585. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Torres JZ, Miller JJ, Jackson PK, 2009. High-throughput generation of tagged stable cell lines for proteomic analysis. Proteomics 9, 2888–2891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Van Cleef KW, Overheul GJ, Thomassen MC, Kaptein SJ, Davidson AD, Jacobs M, Neyts J, van Kuppeveld FJ, van Rij RP, 2013. Identification of a new dengue virus inhibitor that targets the viral NS4B protein and restricts genomic RNA replication. Antiviral Res 99, 165–171. [DOI] [PubMed] [Google Scholar]
  41. Verdonck S, Pu S-Y, Sorrell FJ, Elkins JM, Froeyen M, Gao L-J, Prugar LI, Dorosky DE, Brannan JM, Barouch-Bentov R, Others, 2019. Synthesis and structure-activity relationships of 3, 5-disubstituted-pyrrolo [2, 3-b] pyridines as inhibitors of adaptor associated kinase 1 (AAK1) with antiviral activity. J. Med. Chem [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Welsch S, Miller S, Romero-Brey I, Merz A, Bleck CKE, Walther P, Fuller SD, Antony C, Krijnse-Locker J, Bartenschlager R, 2009. Composition and Three-Dimensional Architecture of the Dengue Virus Replication and Assembly Sites. Cell Host & Microbe. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Wu J, Powell F, Larsen NA, Lai Z, Byth KF, Read J, Gu R-F, Roth M, Toader D, Saeh JC, Chen H, 2013. Mechanism and in vitro pharmacology of TAK1 inhibition by (5Z)-7-Oxozeaenol. ACS Chem. Biol 8, 643–650. [DOI] [PubMed] [Google Scholar]
  44. Wu RH, Tsai MH, Chao DY, Yueh A, 2015. Scanning mutagenesis studies reveal a potential intramolecular interaction within the C-terminal half of dengue virus NS2A involved in viral RNA replication and virus assembly and secretion. J Virol 89, 4281–4295. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Xiao F, Wang S, Barouch-Bentov R, Neveu G, Pu S, Beer M, Schor S, Kumar S, Nicolaescu V, Lindenbach BD, Randall G, Einav S, 2018. Interactions between the Hepatitis C Virus Nonstructural 2 Protein and Host Adaptor Proteins 1 and 4 Orchestrate Virus Release. mBio. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Yang C-C, Tsai M-H, Hu H-S, Pu S-Y, Wu R-H, Wu S-H, Lin H-M, Song J-S, Chao Y-S, Yueh A, 2013. Characterization of an efficient dengue virus replicon for development of assays of discovery of small molecules against dengue virus. Antiviral Res. 98, 228–241. [DOI] [PubMed] [Google Scholar]
  47. Zhou H, Xu M, Huang Q, Gates AT, Zhang XD, Castle JC, Stec E, Ferrer M, Strulovici B, Hazuda DJ, Espeseth AS, 2008. Genome-scale RNAi screen for host factors required for HIV replication. Cell Host Microbe 4, 495–504. [DOI] [PubMed] [Google Scholar]
  48. Zou G, Xu HY, Qing M, Wang QY, Shi PY, 2011. Development and characterization of a stable luciferase dengue virus for high-throughput screening. Antiviral Res. 91, 11–19. [DOI] [PubMed] [Google Scholar]

RESOURCES