Dengue Virus Genome Uncoating Requires Ubiquitination

Abstract

The process of genome release or uncoating after viral entry is one of the least-studied steps in the flavivirus life cycle. Flaviviruses are mainly arthropod-borne viruses, including emerging and reemerging pathogens such as dengue, Zika, and West Nile viruses. Currently, dengue virus is one of the most significant human viral pathogens transmitted by mosquitoes and is responsible for about 390 million infections every year around the world. Here, we examined for the first time molecular aspects of dengue virus genome uncoating. We followed the fate of the capsid protein and RNA genome early during infection and found that capsid is degraded after viral internalization by the host ubiquitin-proteasome system. However, proteasome activity and capsid degradation were not necessary to free the genome for initial viral translation. Unexpectedly, genome uncoating was blocked by inhibiting ubiquitination. Using different assays to bypass entry and evaluate the first rounds of viral translation, a narrow window of time during infection that requires ubiquitination but not proteasome activity was identified. In this regard, ubiquitin E1-activating enzyme inhibition was sufficient to stabilize the incoming viral genome in the cytoplasm of infected cells, causing its retention in either endosomes or nucleocapsids. Our data support a model in which dengue virus genome uncoating requires a nondegradative ubiquitination step, providing new insights into this crucial but understudied viral process.

Importance: Dengue is the most significant arthropod-borne viral infection in humans. Although the number of cases increases every year, there are no approved therapeutics available for the treatment of dengue infection, and many basic aspects of the viral biology remain elusive. After entry, the viral membrane must fuse with the endosomal membrane to deliver the viral genome into the cytoplasm for translation and replication. A great deal of information has been obtained in the last decade regarding molecular aspects of the fusion step, but little is known about the events that follow this process, which leads to viral RNA release from the nucleocapsid. Here, we investigated the fate of nucleocapsid components (capsid protein and viral genome) during the infection process and found that capsid is degraded by the ubiquitin-proteasome system. However, in contrast to that observed for other RNA and DNA viruses, dengue virus capsid degradation was not responsible for genome uncoating. Interestingly, we found that dengue virus genome release requires a nondegradative ubiquitination step. These results provide the first insights into dengue virus uncoating and present new opportunities for antiviral intervention.

FIG 1

Fate of the capsid protein during DENV infection. (A) Schematic representation of the events during viral entry and uncoating. (B) (Top) Experimental design showing the times of infection (above time line) and times of sample collection (below time line). (Bottom) Western blots using DENV anticapsid antibodies are shown. Adsorption lanes indicate the capsid protein present in adsorbed virions in the absence of proteinase K (PK) treatment. In the presence of PK, complete removal of adsorbed virions was observed. The presence of the capsid protein during viral infection at different times from 1 to 20 h postinfection after PK treatment is shown. (C) Levels of capsid protein inside the host cell as a function of time postinfection in the presence (orange bars) and absence (white bars) of cycloheximide (CHX). AU, arbitrary units.

FIG 2

Capsid protein decay is proteasome dependent, but its degradation is not required for initial viral translation. (A) A representative Western blot showing viral capsid protein decay during infection in the presence of CHX and DMSO (top) or CHX and the MG132 proteasome inhibitor (bottom). On the right, a plot shows quantified capsid levels as a function of time in the presence of DMSO or MG132. (B) (Top panel) Schematic representation of a reporter dengue virus, DV-R, showing the duplication of the cis-acting elements (CAE) and the location of luciferase and viral protein-coding regions. FMDV, foot-and-mouth desease virus 2A site; C, capsid. (Bottom panel) Measurement of the first rounds of viral translation by analysis of luciferase activity in the presence (black line) or absence (green line) of puromycin (Puro). (C) Luciferase activity measurement in cells infected with DV-R in the presence (gray bars) or absence (green bars) of MG132 at 4 or 6 h postinfection. Data correspond to the averages of results of three experiments. Error bars indicate standard deviations. (D) Cartoon representing the results obtained where capsid degradation was not necessary for the first rounds of viral translation.

FIG 3

Capsid degradation and viral translation depend on ubiquitination. (A) A representative Western blot showing viral capsid protein decay during infection in the presence of CHX and DMSO (top) or CHX and the ubiquitination inhibitor UBEI 41 (bottom). (B) Luciferase activity measurement in cells infected with DV-R in the presence (gray bars) or absence (green bars) of UBEI 41 at 3 and 5 h postinfection. (C) Luciferase activity measurement in cells transfected with DENV reporter RNA to bypass infection, in the presence or absence of UBEI 41. (D) (Left panel) Schematic representation of the experimental design: after 2 h of infection with DV-R, UBEI 41 (or DMSO [control]) was added; luciferase activity was measured 4 h after addition of the compound. (Right panel) Luciferase levels in the presence or absence of UBEI 41 as indicated. (E) (Left panel) Experimental design showing the times of infection, UBEI 41 addition, and sample collection. (Right panel) Luciferase activity measurement in cells infected with DENV reporter in the presence or absence of UBEI 41 at 6 h postinfection. Data correspond to the averages of results of three experiments. (F) DENV reporter infections of cells silenced with siRNAs directed to UBA1 (ubiquitin-like modifier activating enzyme 1; required for ubiquitination), UBA3 (ubiquitin-like modifier activating enzyme 3; required for neddylation), Renilla (positive control), or a nonrelated siRNA (NR [negative control]). AU X 103, AU × 10³. (Inset panel) Quantification of mRNA levels of UBA1 or UBA3 in cells silenced for one or the other gene by real-time RT-PCR. (G) Luciferase activity measurement in cells infected with reporter DENV in the presence of UBEI 41, DMSO, or the MLN4924 neddylation inhibitor as indicated at 6 h postinfection.

FIG 4

Infectivity of recombinant DENVs with substitution of lysine (K) by arginine (R) in the capsid protein. (A) (Top panel) Location of the capsid K residues that were substituted in the mutant viruses (Mut 1, Mut 2, and Mut 3). Combined mutations (Mut 1 plus Mut 2 [Mut 1+2], Mut 1+3, Mut 2+3, and K-less) were also designed. (Bottom panel) Results of immunofluorescence (IF) assays of cells transfected with WT or mutated viral RNAs at different times as indicated for each panel. (B) Expression and processing of capsid was evaluated by Western blot analysis after RNA transfection of the WT and each mutant: Mut 1, Mut 2, Mut 3, Mut 1+2, Mut 1+3, Mut 2+3, or K-less. (C) (Top panel) Schematic representation of a DENV reporter system, DV-RD, indicating the locations where the K substitutions were incorporated. (Bottom panel) Translation and replication of viral RNAs in reporter-transfected cells followed by luciferase activity measurements. A WT virus, a K-less mutant, and a replication impaired control virus (Mut NS5 [polymerase mutant]) are shown. hpt, hours posttransfection. (D) Schematic representation of the reporter dengue virus, DV-R, showing the locations of the mutations inserted in the capsid protein. (E) Translation and replication of viral RNAs corresponding to WT DV-R, a capsid-deleted control (ΔC), and the lysine mutants. (F) Evaluation of production of infectious particles. Levels of luciferase activity measured postinfection with supernatants obtained from the transfections presented in panel E.

FIG 5

Ubiquitination inhibition blocks viral genome release during infection. (A) Levels of viral RNA measured by real-time RT-PCR as a function of time postinfection of cells treated with DMSO (left panel [Control]) or UBEI 41 (right panel). Adsorbed viral particles were removed with PK treatment at all time points. Data correspond to the averages of results of triplicates, and error bars indicate standard deviations. (B) Cartoon representing the steps of the viral life cycle that are susceptible to ubiquitination inhibition.