Identification of determinants involved in initiation of hepatitis C virus RNA synthesis by using intergenotypic replicase chimeras

Abstract

The 5' nontranslated region (NTR) and the X tail in the 3' NTR are the least variable parts of the hepatitis C virus (HCV) genome and play an important role in the initiation of RNA synthesis. By using subgenomic replicons of the HCV isolates Con1 (genotype 1) and JFH1 (genotype 2), we characterized the genotype specificities of the replication signals contained in the NTRs. The replacement of the JFH1 5' NTR and X tail with the corresponding Con1 sequence resulted in a significant decrease in replication efficiency. Exchange of the X tail specifically reduced negative-strand synthesis, whereas substitution of the 5' NTR impaired the generation of progeny positive strands. In search for the proteins involved in the recognition of genotype-specific initiation signals, we analyzed recombinant nonstructural protein 5B (NS5B) RNA polymerases of both isolates and found some genotype-specific template preference for the 3' end of positive-strand RNA in vitro. To further address genotype specificity, we constructed a series of intergenotypic replicon chimeras. When combining NS3 to NS5A of Con1 with NS5B of JFH1, we observed more-efficient replication with the genotype 2a X tail, indicating that NS5B recognizes genotype-specific signals in this region. In contrast, a combination of the NS3 helicase with NS5A and NS5B was required to confer genotype specificity to the 5' NTR. These results present the first genetic evidence for an interaction between helicase, NS5A, and NS5B required for the initiation of RNA synthesis and provide a system for the specific analysis of HCV positive- and negative-strand syntheses.

FIG. 1.

Structure and replication kinetics of genotype 1b and 2a replicons. (A) Structures of the replicons Luc Con (genotype 1b) and Luc JFH (genotype 2a) used for transient replication assays. luc, firefly luciferase; EI, encephalomyocarditis virus IRES; VR, variable region and poly(U/UC) tract of the HCV 3′ NTR; X, X tail. Coding sequences are indicated by rectangles, and noncoding regions are indicated by oval forms. Con1 sequences are indicated in white with black letters, and JFH1-derived sequences are indicated by black, filled forms with white lettering. This code was kept throughout the whole study. Adaptive mutations E1202G, T1280I, and K1846T in Con1 are indicated by black dots. (B) Transient replication of Luc Con and Luc JFH. Huh7-Lunet cells were transfected with 5 μg of in vitro transcripts corresponding to Luc Con, Luc JFH, ConΔGDD, or JFHΔGDD. Luciferase activity was determined in cell lysates prepared at 2, 4, 8, 12, 18, 24, 48, and 72 h posttransfection. Data were normalized for transfection efficiency among identical replicons as determined by measurement of the luciferase activity 2 h after transfection. Electroporations and luciferase assays were performed in duplicate. Note the logarithmic scale of the ordinate and the abscissa.

FIG. 2.

Impact of heterologous NTR sequences on replication efficiency of Con1 and JFH1 replicons. (A) Effect of the heterologous 5′ NTR and 3′ X tail on replication efficiency of a Luc JFH replicon. Luc JFH replicon RNAs with homologous NTRs (5′ X JFH), 5′ NTR, or 3′ X-tail sequences derived from Con1 (5′ Con/X JFH, 5′ JFH/X Con, or 5′ X Con, respectively) or a replication-deficient control replicon (ΔGDD) were transfected into Huh7 cells; transfected cells were seeded and harvested at 4, 24, 48, and 72 h after transfection, and cell lysates were analyzed for luciferase activity. Luciferase activity is given as the change in RLU (n-fold) relative to the value obtained 4 h after transfection. VR, variable region and poly(U/UC) tract of the HCV 3′ NTR; X, X tail. (B) Genotype specificity of replication signals in the 5′ NTR. Luc JFH replicons with 5′ NTRs derived from isolates JFH1 and J6 (both genotype 2a), H77 (genotype 1a), and Con1 (genotype 1b) were transfected into Huh7 cells; luciferase activity was determined as described above. JFH1 and J6 5′ NTRs differed at positions 4, 78, and 302 and Con1 and H77 at positions 11, 12, 13, 34, 35, 204, and 243. (C) Heterologous NTR sequences affect the replication efficiency of Luc Con replicons. Luc Con replicon RNAs with homologous NTRs (5′ X Con), 5′ NTR, or 3′ X-tail sequences derived from JFH1 (5′ JFH/X Con, 5′ Con/X JFH, or 5′ X JFH, respectively) were transfected into Huh7 cells and analyzed for replication efficiency as described above. Schematic drawings of replicons are shown at the top of panels A and C. JFH1- or Con1-derived sequences are given in black or white, respectively; portions given in gray are variable within one panel. For a detailed description of the luciferase replicons, refer to the legend to Fig. 1A.

FIG. 3.

Impact of 3′ X-tail and 5′ NTR chimeras on negative- and positive-strand syntheses of replicon Luc JFH. Analysis of transient replication of Luc JFH replicons with different NTR sequences by Northern hybridization. (A) Huh7 cells were transfected by electroporation with Luc JFH replicons harboring homologous NTR (5′ X JFH), 5′ NTR, or 3′ X-tail sequences derived from Con1 (5′ Con/X JFH, 5′ JFH/X Con, or 5′ XCon, respectively) or with a replication-deficient control replicon (ΔGDD) and analyzed for positive- and negative-strand RNA syntheses at 0 to 72 h after transfection. β-Actin RNA levels were used to normalize for loading in each lane. (B) Huh7-Lunet cells were transfected with the same replicons as those used for panel A and analyzed 0 to 12 h after electroporation. Replicon Luc JFH 5′ X Con was excluded from this analysis due to inefficient replication. Cells were harvested at the time points postelectroporation indicated above lanes 2 to 21. Ten micrograms of total RNA was analyzed by Northern hybridization with a ³²P-labeled negative- or positive-sense riboprobe to detect HCV positive-strand RNA [(+)RNA] or negative-strand RNA [(−)RNA], respectively, or with a negative-sense riboprobe specific for β-actin as given on the right. Four micrograms of total RNA from naïve Huh7 cells was used as a negative control (Huh7). Luc JFH in vitro transcripts (10⁷ or 10⁸) of positive-sense [(+)RNA] or negative-sense [(−)RNA] orientation were loaded to quantify HCV RNA, as shown above the two outer left or right lanes, respectively.

FIG. 4.

Mapping of genotype-specific signals in the 3′ X tail and 5′ NTR. (A) Secondary structure of SLI of the X tail (8). The consensus sequences of genotypes 1 and 3 to 6 are given, and deviations found in the consensus sequence of genotype 2 are indicated by arrows. (B) Schematic representation of the secondary structures determined for the 3′-terminal 156 nucleotides of HCV negative-strand RNA (59), complementary to the 5′ NTR of the positive strand. Note that numbering starts at the 3′ end. Differences between the Con1 and JFH1 sequences are indicated by filled circles. The open circle indicates an isolate-specific deviation of the JFH1 sequence at position 78 which was excluded from the analysis. (C) Impact of single-nucleotide replacements in the 3′ X-tail sequence on HCV replication efficiency. Luc/ubiquitin JFH1 replicon RNAs with a homologous 3′ X tail (X JFH), with individual nucleotide exchanges in the 3′ X tail (C72U, UA74/75AU, and U81C, as indicated in panel A), or with the entire 3′ X-tail sequence derived from Con1 (X Con) were transfected into Huh7 cells, and 4, 24, 48, and 72 h after transfection, cell lysates were analyzed for luciferase activity. Luciferase activity is given as the change in RLU (n-fold) relative to the value obtained 4 h after transfection. The structure of the replicon Luc/ubi JFH used in this experiment is given at the top. luc, firefly luciferase; ubi, ubiquitin; VR, variable region and poly(U/UC) tract of the HCV 3′ NTR; X, X tail (for a detailed explanation, refer to the legend to Fig. 1A). (D) Analysis of genotype-specific signals in the 5′ NTR. Luc JFH replicons with authentic 5′ NTRs (JFH1), 5′ NTRs derived from Con1 (5′ Con), or chimeric JFH 5′ NTRs containing parts of Con1 clustered according to the secondary structure prediction of the 3′ end of the negative strand (B) were electroporated into Huh7 cells. Transfected cells were analyzed for luciferase activity at 4 h and 24 h after transfection. The ratio of luciferase activity at 24 h to that at 4 h after transfection is shown. For further explanation, refer to the text.

FIG. 5.

Purification of JFH1 NS5BΔC21 and analysis of genotype-specific template recognition in vitro. (A) Expression of NS5B from isolates JFH1 (lanes 1 to 5) and Con1 (lanes 6 to 10) in E. coli and purification by differential solubilization and affinity chromatography. Both proteins lack 21 C-terminal amino acids and are fused to a His₆ tag. T, total bacterial lysate after induction; S1, supernatant 1 after treatment of bacterial cells with LBI and centrifugation (note that NS5B was not soluble under these conditions); S2, supernatant 2 obtained from solubilization of the pellet remaining from S1; S3, supernatant 3 after incubation of S2 with Ni-NTA agarose; E, eluted protein. Numbers on the left refer to the sizes (in kDa) of reference proteins run on the same gel. For a detailed explanation, refer to Materials and Methods. (B) Schematic representation of different template RNAs used for in vitro RdRp assays. Portions corresponding to Con1 sequences are given in white with black letters, and JFH1-derived sequences are indicated by black, filled forms with white lettering. The positions of cis-acting RNA elements in the coding sequence of NS5B are indicated by schematic stem-loop drawings. Sequences corresponding to the 3′ terminus of the negative strand are given upside down and in 3′-to-5′ orientation. VR, variable region and poly(U/UC) tract of the HCV 3′ NTR; X, X tail. (C) Analysis of genotype-specific template recognition in vitro by NS5B of isolates JFH1 and Con1. Different in vitro-transcribed RNAs as indicated above each lane and as shown schematically in panel B were incubated with 200 ng purified NS5BΔC21 from isolate JFH1 (5B/JFH) or 1 μg Con1 (5B/Con) in the presence of [α-³²P]CTP. Reaction products were separated by denaturing agarose-glyoxal gel electrophoresis; the gel was dried and subjected to autoradiography. mock, elution buffer instead of purified polymerase was added to the reaction; M 3′ (+), in vitro-transcribed, radiolabeled RNA corresponding to nucleotides 9259 to 9678 of the HCV JFH1 positive strand, identical in size to template 3′ (+) X JFH; M 3′ (−) JFH, in vitro-transcribed, radiolabeled RNA corresponding to the 3′-terminal 340 nucleotides of the negative strand of HCV JFH1, identical in size to template 3′ (−) JFH. Note that 3′ (−)C JFH and 3′ (−) Con encompass 341 nucleotides in length. The sizes of the template RNAs are indicated by arrows on the left [3′ (+)] and right [3′ (−)]. For a detailed explanation of the different template RNAs, refer to the text.

FIG. 6.

Structure and replication competence of intergenotypic replicase chimeras. (A) Schematic structures of replicons harboring a chimeric replicase. White areas with black lettering represent Con1-derived sequences, and black areas with white lettering refer to JFH1. Adaptive mutations E1202G, T1280I, and K1846T in the Con1 sequence are indicated by black dots. luc, firefly luciferase; EI, encephalomyocarditis virus IRES; VR, variable region and poly(U/UC) tract of the HCV 3′ NTR; X, X tail. For further details, refer to the legend to Fig. 1A. (B) Replication competence of intergenotypic replicase chimeras. Luciferase replicon RNAs derived from either JFH1 (Luc JFH) or Con1 (Luc Con) or replicon RNAs harboring chimeric NS3-to-NS5B coding regions or a replication-deficient control replicon (ΔGDD) were transfected into Huh7-Lunet cells by electroporation; transfected cells were seeded and harvested at 4, 24, 48, and 72 h after transfection, and cell lysates were analyzed for luciferase activity. Luciferase activity is given as the change in RLU (n-fold) relative to the value obtained 4 h after transfection. Electroporations and luciferase assays were performed in duplicate; data represent the averages and standard deviations of at least four single values. Note the logarithmic scale of the ordinate.

FIG. 7.

NTR preferences of intergenotypic replicase chimeras. Genotype Con1-based luciferase replicons harboring the sequences of different nonstructural proteins from isolate JFH1 were fused to the 5′ NTR and 3′ X tail from JFH1 (5′ X JFH), Con1 (5′ XCon), or mixed NTRs (5′ Con/X JFH and 5′ JFH/X Con) and tested for replication efficiency as described in the legend to Fig. 6B. Schematic drawings of replicons are shown at the top of panels A to D. JFH1- or Con1-derived sequences are given in black or white, respectively; portions given in gray are variable within one panel. luc, firefly luciferase; EI, encephalomyocarditis virus IRES; VR, variable region and poly(U/UC) tract of the HCV 3′ NTR; X, X tail. For a detailed description of the luciferase replicons, refer to the legend to Fig. 1A.