Polyprotein-Driven Formation of Two Interdependent Sets of Complexes Supporting Hepatitis C Virus Genome Replication

Abstract

Hepatitis C virus (HCV) requires proteins from the NS3-NS5B polyprotein to create a replicase unit for replication of its genome. The replicase proteins form membranous compartments in cells to facilitate replication, but little is known about their functional organization within these structures. We recently reported on intragenomic replicons, bicistronic viral transcripts expressing an authentic replicase from open reading frame 2 (ORF2) and a second duplicate nonstructural (NS) polyprotein from ORF1. Using these constructs and other methods, we have assessed the polyprotein requirements for rescue of different lethal point mutations across NS3-5B. Mutations readily tractable to rescue broadly fell into two groupings: those requiring expression of a minimum NS3-5A and those requiring expression of a minimum NS3-5B polyprotein. A cis-acting mutation that blocked NS3 helicase activity, T1299A, was tolerated when introduced into either ORF within the intragenomic replicon, but unlike many other mutations required the other ORF to express a functional NS3-5B. Three mutations were identified as more refractile to rescue: one that blocked cleavage of the NS4B5A boundary (S1977P), another in the NS3 helicase (K1240N), and a third in NS4A (V1665G). Introduced into ORF1, these exhibited a dominant negative phenotype, but with K1240N inhibiting replication as a minimum NS3-5A polyprotein whereas V1665G and S1977P only impaired replication as a NS3-5B polyprotein. Furthermore, an S1977P-mutated NS3-5A polyprotein complemented other defects shown to be dependent on NS3-5A for rescue. Overall, our findings suggest the existence of two interdependent sets of protein complexes supporting RNA replication, distinguishable by the minimum polyprotein requirement needed for their formation.

Importance: Positive-strand RNA viruses reshape the intracellular membranes of cells to form a compartment within which to replicate their genome, but little is known about the functional organization of viral proteins within this structure. We have complemented protein-encoded defects in HCV by constructing subgenomic HCV transcripts capable of simultaneously expressing both a mutated and functional polyprotein precursor needed for RNA genome replication (intragenomic replicons). Our results reveal that HCV relies on two interdependent sets of protein complexes to support viral replication. They also show that the intragenomic replicon offers a unique way to study replication complex assembly, as it enables improved composite polyprotein complex formation compared to traditional trans-complementation systems. Finally, the differential behavior of distinct NS3 helicase knockout mutations hints that certain conformations of this enzyme might be particularly deleterious for replication.

FIG 1

Schematic representation of the intragenomic HCV replicons used in this study. The HCV IRES in the 5′ UTR drives translation of ORF1, which encodes the indicated NS polyproteins expressed as part of a Renilla-foot and mouth disease (FMDV) 2A fusion protein. An encephalomyocarditis virus (EMCV) IRES drives translation of the NS3-5B replicase in ORF2. When present, NS5A is expressed from ORF1 as a COOH-terminal V5 epitope-tagged protein. Constructs express NS5A from ORF2 as a FLAG-tagged protein. Coding regions in ORF1 downstream from Renilla 2A represent recoded NS sequences incorporating synonymous sequence alterations compared to the authentic viral counterpart sequence present in ORF2. The table compares the amino acid and nucleotide sequences as well as the CpG and UpA frequencies in ORF1 (synthetic sequence) and ORF2 (authentic viral sequences).

FIG 2

Replication of constructs expressing NS3-5B as a Renilla-FMDV 2A fusion protein. Replication was assessed using both monocistronic (a) and intragenomic (b) replicon constructs. The schematic above each graph indicates the regions of NS3-5B encoded by the recoded and authentic JFH1 coding sequences for each construct. Data represent the means ± standard deviations (SD) from two independent experiments (a) and means ± standard errors of the means (SEM) from three independent experiments (b).

FIG 3

Assessing the capacity of ORF1 polyproteins to support replication of intragenomic replicons with different defects in ORF2 and recue of the same defects by trans-complementation. (a) Replication assay data of intragenomic replicons expressing functional NS3-5A or NS3-5B in ORF1 and expressing NS3-5B carrying a lethal mutation in ORF2 as defined in the schematic above the graph. (b and c) These mutations were introduced into bicistronic replicons expressing firefly luciferase in ORF1, and the capacity of these constructs to replicate was assessed in a stable neomycin-resistant (Neo^r) helper replicon cell line (b) or in naive Huh7.5 cells (c). Data represent the means ± SD from two independent experiments.

FIG 4

Impact of single helicase point mutations K1240N and T1299A on replicon replication. Schematics above the graphs provide details of the constructs used and the positioning of mutations within them. Replication assays either employed a single intragenomic replicon construct per experimental group transfected into Huh7.5 cells (a, c, d) or involved cotransfection of two separate RNAs into naive Huh7.5 cells to assess whether K1240N was dominant negative in trans (b). Results shown from the latter assay also include the firefly signal at 4 and 24 h derived from the replication-defective constructs used. Data represent the means ± SD (a, c, d) or means ± SEM (b) from 2 and 3 independent experiments, respectively. Where n = 3, values significantly different from the control group (mono + tRNA) are highlighted with an asterisk (P < 0.05; 2-tailed t test). Note: the assay performed to obtain the data in panel d was run in parallel with another described in the legend to Fig. 8e, and so they share control groups.

FIG 5

Analysis of mutations using the intragenomic replicon and assessing their dependency on NS3-5B for rescue. The selection of mutations was on the basis that they were refractile to rescue with ORF1 NS3-5A when present in ORF2 but tolerant when ORF1 instead expressed NS3-5B. These mutations were introduced into intragenomic replicons in various combinations, and replication assays were performed. Schematics above each graph (a, c, d, e, f) provide details of the constructs used and the positioning of mutations within them. Also shown are Northern blot data (b) using total cellular RNA taken from a replication assay (a). Data represent the means ± SD from two separate experiments. pol ko, polymerase knockout.

FIG 6

Rescue of lethal mutations in NS3 and NS4A that disrupt processes other than NS3 helicase activity. Genetic complementation of genetic defects was assessed in a variety of ways, including use of intragenomic replicons transfected into Huh7.5 cells (a, c) or using bicistronic replicon expressing firefly luciferase from ORF1 transfected into either a helper replicon cell line (d) or into naive Huh7.5 cells transduced with either baculovirus expressing NS3-5A or β-galactosidase (mock control) (e). Intragenomic replicons were also used to assess whether V1665G, which appeared refractile to rescue with NS3-5B, was dominant negative (c). Schematics above each graph provide specific details of the constructs used and the positioning of mutations within them. Data represent the means ± SD from two separate experiments.

FIG 7

Effects of blocking NS4B5A boundary cleavage on intragenomic replication. Schematics provide details of the constructs used and the positioning of mutations within them. Cleavage of the NS4B5A boundary was blocked by the S1977P mutation. The impact that this had on intragenomic replicons when it was introduced into ORF2 (a) and ORF1 (b) is shown, with cell lysates taken from the latter experiment 72 h posttransfection also analyzed by Western blotting (c). Arrows indicate the position of NS5A and the uncleaved NS4B5A precursor. An asterisk indicates the position of a cross-reacting cellular protein detected by the anti-FLAG antibody. Data from a replication assay comparing the abilities of different mutated NS3-5B ORF1s, including that of S1977P, to suppress replication of intragenomic replicons carrying functional ORF2 (d) are also provided. Data represent the means ± SD from two separate experiments. Note: the assay performed to obtain data in panel b was run in parallel with that described in Fig. 8b, and so the two share control groups.

FIG 8

Impact of blocking NS4B5A cleavage on NS3-5A function. (a) Intragenomic replicons with a functional NS3-5B encoded in ORF2 and expressing either NS5A, NS3-5A, or an S1977P-mutated NS3-5A from ORF1 were transfected into Huh7.5 cells. Confocal microscopy visualized the extent of ORF1 V5-tagged NS5A and ORF2 FLAG-tagged NS5A colocalization 72 h posttransfection. (b and e) Intragenomic replicon replication assays detail the ability of an S1977P-mutated polyprotein expressed from ORF1 to complement either the NS5A S2208I (b) or NS5A G2313A (e) defect in ORF2, the schematics above the graphs indicating further details of the constructs used. (c) Trans-complementation of a firefly luciferase-expressing bicistronic replicon carrying the S2208I mutation is assessed using Huh7.5 cells transduced with baculovirus expressing either β-galactosidase (lacZ) or GFP-tagged versions of NS5A, NS3-5A, and NS3-5A carrying the S1977P mutation. Western blot analysis of cell lysates taken 72 h posttransfection shows levels of baculovirus-transduced expression (d). Graphical data represent the means ± SD from two separate experiments.

FIG 9

Putative model of NS3-5B-dependent complex assembly. (a) The NS3 helicase domain within a recently translated polyprotein precursors binds in a cis-dependent manner to the viral genome (for illustrative purposes, the interaction is shown to occur within the 3′-UTR X region). Subsequent recruitment of additional copies of NS3-5B precursor is not a cis-acting event but does depend on them possessing both NS5B and helicase-dependent RNA binding activity. Our data showing that blocking of NS4B5A cleavage creates a dominant negative polyprotein precursor suggest that while complex activity requires completion of polyprotein proteolytic processing, complex assembly does not. (b to f) The outcome of NS3-5B-dependent complex formation is shown using the same criteria as described above. Replication requires that (i) complex assembly occurs, (ii) any mutated NS protein present is rescued by a functional counterpart, and (iii) a polyprotein carrying a dominant negative mutation is not recruited.