Effects of mutations of the initiation nucleotides on hepatitis C virus RNA replication in the cell

Abstract

Replication of nearly all RNA viruses depends on a virus-encoded RNA-dependent RNA polymerase (RdRp). Our earlier work found that purified recombinant hepatitis C virus (HCV) RdRp (NS5B) was able to initiate RNA synthesis de novo by using purine (A and G) but not pyrimidine (C and U) nucleotides (G. Luo et al., J. Virol. 74:851-863, 2000). For most human RNA viruses, the initiation nucleotides of both positive- and negative-strand RNAs were found to be either an adenylate (A) or guanylate (G). To determine the nucleotide used for initiation and control of HCV RNA replication, a genetic mutagenesis analysis of the nucleotides at the very 5' and 3' ends of HCV RNAs was performed by using a cell-based HCV replicon replication system. Either a G or an A at the 5' end of HCV genomic RNA was able to efficiently induce cell colony formation, whereas a nucleotide C at the 5' end dramatically reduced the efficiency of cell colony formation. Likewise, the 3'-end nucleotide U-to-C mutation did not significantly affect the efficiency of cell colony formation. In contrast, a U-to-G mutation at the 3' end caused a remarkable decrease in cell colony formation, and a U-to-A mutation resulted in a complete abolition of cell colony formation. Sequence analysis of the HCV replicon RNAs recovered from G418-resistant Huh7 cells revealed several interesting findings. First, the 5'-end nucleotide G of the replicon RNA was changed to an A upon multiple rounds of replication. Second, the nucleotide A at the 5' end was stably maintained among all replicon RNAs isolated from Huh7 cells transfected with an RNA with a 5'-end A. Third, initiation of HCV RNA replication with a CTP resulted in a >10-fold reduction in the levels of HCV RNAs, suggesting that initiation of RNA replication with CTP was very inefficient. Fourth, the 3'-end nucleotide U-to-C and -G mutations were all reverted back to a wild-type nucleotide U. In addition, extra U and UU residues were identified at the 3' ends of revertants recovered from Huh7 cells transfected with an RNA with a nucleotide G at the 3' end. We also determined the 5'-end nucleotide of positive-strand RNA of some clinical HCV isolates. Either G or A was identified at the 5' end of HCV RNA genome depending on the specific HCV isolate. Collectively, these findings demonstrate that replication of positive-strand HCV RNA was preferentially initiated with purine nucleotides (ATP and GTP), whereas the negative-strand HCV RNA replication is invariably initiated with an ATP.

FIG. 1.

(A) Diagram of the in vitro transcription and replication cycle of HCV replicon RNA. The HCV replicon RNAs were transcribed by a T7 RNA polymerase from DNA vectors digested with restriction enzymes as indicated on the right side of panel B. The in vitro-transcribed RNA was transfected into Huh7 cells. Upon replication, the positive-strand HCV RNA is converted to the complementary negative strand, which in turn serves as a template for synthesis of more positive-strand RNA. The 5′- and 3′-end nucleotides of both positive- and negative-strand RNAs are highlighted in boldface letters. (B) Schematic presentation of the in vitro-transcribed HCV replicon RNAs. The nucleotides at the 5′ and 3′ ends of the replicon RNAs are highlighted in boldface letters. The T7 promoter is indicated by an open box. The restriction enzyme used to linearize each cDNA clone of the replicon RNA is shown on the right side. The replicon was named after the 5′- or 3′-end nucleotide, as shown on the left side. The original HCV replicon RNA contains a G at the 5′ end and a U at the 3′ end (37).

FIG. 2.

Effects of the 5′-end nucleotide mutations on cell colony formation. Then, 2 μg each of the in vitro-transcribed RNAs was transfected into 8 × 10⁶ Huh7 cells by electroporation. The RNA-transfected Huh7 cells were incubated with DMEM containing 10% FBS. After 24 h of incubation at 37°C and 5% CO₂, cell culture medium was replaced by DMEM containing 10% FBS and 0.5 mg of G418 sulfate/ml. The medium was changed twice a week. After an ∼4-week selection with G418, cell clones were fixed, stained by a solution containing 0.01% crystal violet and 19% methanol, and photographed.

FIG. 3.

RPA determination of the positive-strand and negative-strand RNAs of HCV replicons isolated from different Huh7 cell lines. (A) Detection of both positive- and negative-strand RNAs of HCV replicons isolated from Huh7 cell colonies resulted from replication of the 5′UTR-A RNA. Total cellular RNA was extracted with TRIzol reagent from Huh7 cells. A total of 15 μg of total RNA was used in an RPA for hybridization with 5 × 10⁴ cpm of [α-³²P]UTP-labeled β-actin RNA probe (Ambion) and either 10⁵ cpm of [α-³²P]UTP-labeled (−)3′UTR (for the detection of positive-strand RNA) or (+)5′UTR RNA probe (for the detection of negative-strand RNA). After RNase A/T₁ digestion, RNA products were analyzed in a 6% polyacrylamide-7.7 M urea gel. The RNA levels were determined by quantitation with a PhosphorImager (Molecular Dynamics). The sizes of the RNA markers are indicated on the left, and arrows on the right highlight the RNA products. Numbers on the top indicate different cell colonies that resulted from replication of the 5′UTR-A RNA. Huh7, RNA extracted from regular Huh7 cells as a negative control; 5′UTR-G, RNA extracted from a Huh7 cell line resulted from transfection with the 5′UTR-G RNA (wild type). (B) Detection of both positive-strand and negative-strand RNAs of HCV replicons recovered from Huh7 cells transfected with the 5′UTR-C RNA. Total RNA was extracted from Huh7 cell colonies that resulted from transfection with the 5′UTR-C RNA. Otherwise, detection was the same as described for panel A. Positive- and negative-strand RNAs are indicated at the bottom.

FIG. 4.

Effects of the 3′-end nucleotide mutations on cell colony formation. The in vitro-transcribed replicon RNAs contain nucleotide mutations from a U (3′UTR-U) to a C (3′UTR-C), to a G (3′UTR-G), and to an A (3′UTR-A) at the 3′ end, respectively. Otherwise, transfection and selection were done in the same way as described as for Fig. 2. The 3′UTR-U is a wild-type replicon RNA.

FIG. 5.

Determination of the positive- and negative-strand RNAs of the subgenomic HCV replicons isolated from Huh7 cells transfected with the 3′UTR-C and 3′UTR-G RNAs. (A) Quantitation of replicon RNAs derived from the 3′UTR-C RNA-transfected Huh7 cells. RPA was essentially the same as described for Fig. 3 except that total RNAs were extracted from different Huh7 cell lines as a result of replication of the 3′UTR-C RNA and selection with G418 sulfate. W.t. is the 5′UTR-G RNA (Fig. 1). (B) Quantitation of replicon RNAs isolated from different Huh7 cell lines that resulted from replication of the 3′UTR-G RNA by RPA. The positive- and negative-strand RNAs are indicated at the bottom. Numbers indicate replicon RNAs derived from different cell lines.

FIG. 6.

Determination of the positive- and negative-strand RNAs of HCV replicon RNAs isolated from Huh7 cells transfected with the 5′UTR-A/3′UTR-C RNA. Total RNAs were extracted from six different Huh7 cell lines resulted from replication of the transfected 5′UTR-A/3′UTR-C RNA. RPA was done in the same way as described in the Fig. 3 legend. 5′-3′/w.t., RNA extracted from an Huh7 cell line resulting from replication of the transfected wild-type replicon RNA (5′UTR-G RNA, see Fig. 1B). Replicon RNAs derived from different cell lines are indicated by numbers on the top. The positive- and negative-strand RNAs are indicated at the bottom.

FIG. 7.

Determination of nucleotides at the 5′ and 3′ ends of HCV replicon RNAs recovered from G418-resistant Huh7 cells. (A) Schematic of the RNA ligase-mediated rapid amplification of cDNA ends (RLM-RACE) (Ambion). An adapter RNA (open bar) was ligated to the 5′ end of either the positive-strand (for determination of the 5′-end nucleotide) or the negative-strand (for determination of the 3′-end nucleotide) HCV replicon RNA isolated from different cell lines. The cDNAs of both positive and negative strands of HCV RNA were reverse transcribed by using HCV-specific primers and amplified by PCR with 5′RACE outer/inner primers (gray arrow) and HCV-specific primers (open arrow). The 5′-end nucleotides of both positive and negative strands of HCV RNA were determined by DNA sequence analysis. (B) 5′- and 3′-end nucleotides of HCV RNA determined by RLM-RACE. Letters indicate nucleotides at the 5′ and 3′ ends as shown on the top. The number indicates the number of cell clones from which HCV replicon RNAs were analyzed. Numbers in parentheses indicate the frequency of the nucleotides determined. The HCV replicon RNAs given on the left column were the ones transcribed in vitro by a T7 RNA polymerase.

FIG. 8.

Determination of the 5′-end nucleotide of T7 RNA transcripts and HCV genomic RNAs derived from clinical HCV isolates. The 5′UTR-A and 5′UTR-C RNAs were in vitro transcribed by a T7 RNA polymerase. HCV RNAs were extracted with TRIzol reagent from serum samples of patients infected with different genotypes of HCV. The 5′-end nucleotide of T7 transcripts and HCV genomic RNAs were determined by RLM-RACE as described in Fig. 7A legend. Nucleotide sequences of each chromogram are complementary to HCV genomic RNA (in black and red) and adapter RNA sequence (in blue), as highlighted underneath each chromogram. Nucleotides in red are the 5′-end nucleotides of HCV RNAs. The highlighted sequence underneath each chromogram is in the order from 3′ to 5′ ends. HCV genotypes are indicated by 1a, 2b, 3a, and 4. P1 and P2 stand for HCV RNAs derived from patient 1 and patient 2, who were infected with the same genotype. A mutation from C to U at position 3 from the 5′ end was identified in one clinical isolate (1a-P1) and is highlighted by a purple letter U.