Virus-derived circular RNAs populate hepatitis C virus–infected cells

Figures

• 

  Fig. 1.. Landscape of HCV-derived circRNA (vcircRNA) junctions predicted by RNA-seq analysis. (A) Identification of putative vcircRNA junctions derived from the linear HCV RNA genome. The curved arrow represents a junction-spanning read, which consists of two discontinuous sequences colored in blue and orange and joined by their 5′ and 3′ breakpoints. Aligning with the viral genome also generates the start and the end of read positions. To ensure that the junction originates from vcircRNAs, it is essential to validate the exact order of nucleotide positions, specifically the 5′ breakpoint, end of read, start of read, and finally the 3′ breakpoint. (B) Distribution of predicted vcircRNA junctions on the entire HCV genome. Each junction read is indicated by an arch that connects its 5′ and 3′ breakpoints. The three clusters examined in this study are indicated on the Top. Blue arches on the Bottom demonstrate the distribution of all the other predicted vcircRNA junctions. (C) List of top 20 predicted vcircRNA junctions. (D) Length distribution of the vcircRNAs predicted from junction-spanning reads.

• 

  Fig. 2.. Identification of HCV vcircRNAs in HCV-infected cultured hepatoma cells. (A) Schematic views of vcircRNAs subjected to reverse transcription and subsequent PCR using primers that flank the junction sites of predicted vcircRNAs but diverge on linear RNAs. RT-PCR products of single and multiple repeats of the vcircRNA sequences are denoted as b, c, and d while the smallest amplicon “a” only contains one junction sequence (Right). (B) Agarose gel electrophoresis of RT-PCR products. PCR products representing multiple copies of vcircRNAs from clusters I, II, and III are shown. Primers flanking the variable junctions of cluster I or cluster II vcircRNAs were used. For cluster III 7952/8189, junction-spanning primers were used (listed in Dataset S4). Nested PCR was performed to amplify cluster I vcircRNAs due to their low abundance and a representative image of the second round PCR for cluster I vcircRNAs is shown here in the Upper panel. (C) Representative Sanger sequencing results of RT-PCR products inserted into a TOPO vector. Each sequence trace displays two junction sites denoted as triangles with one full copy of the indicated vcircRNA sequence. (D) Northern blot detection of…

• 

  Fig. 3.. Translation of cluster I IRES-containing vcircRNAs. (A) Translation of the HCV IRES within a circular GFP RNA expressed from a polymerase II promoter-containing split-GFP plasmid that promotes backsplicing to create circular RNAs (see also SI Appendix, Fig. S2E). Plasmids containing the IRES from EMCV, or HCV 5′ UTR, or individual cluster I vcircRNA sequence were transfected into 293FT cells for 48 h. GFP fluorescence in each panel was then analyzed by flow cytometry. (B) Illustrations of two infectious clones of HCV that contain a HiBiT tag of 11 amino acids (21) inserted after the microRNA 122 (miR-122) binding sites in the 5′ noncoding region of HCV, so as not to interfere with miR-122 binding (22). The HCV-c10/405-Hibit mutant has an additional U to G mutation (the G is highlighted in red) to eliminate the upstream UGA codon and allow the expression of the HiBiT tag at the C terminus of vcirc10/405 ORF. In the HCV-c25/1277-Hibit mutant, HiBiT is inserted in-frame with the vcirc25/1277 ORF. (C) HiBiT expression from infectious viral genomes described in Fig. 3B. HiBiT expression from wild-type HCV or mutants were measured as HiBiT NanoLuc activities at indicated time points.…

• 

  Fig. 4.. Translation of the HiBiT-containing linear control IVT RNAs. (A) Schematic view of designing short linear control RNAs of various lengths derived from HCV-c10/405-Hibit. Linear RNA fragments containing HiBiT were generated by in vitro transcription of indicated PCR products amplified from 5′ end of the mutant plasmid of HCV-c10/405-Hibit. (B) Huh7 cells in 12-well plates were transfected with 1 μg of each linear control RNA or in vitro transcribed full-length HCV-WT (lacking HiBiT) or HCV-c10/405-Hibit RNAs. At 12 and 24 h after transfection, HiBiT expression was quantified in the HiBiT Lytic Assay (Promega). At least four replicas were used for each group. Error bars indicate mean ± SEM. Statistical significance was determined by ordinary two-way ANOVA. **P < 0.01 and ***P < 0.001.

• 

  Fig. 5.. Viral circRNAs and circRNA translational product have proviral effects on HCV infection. (A) The orange and blue arms show the corresponding 5′ and 3′ sequences forming the junction of the viral circRNA. JCT siRNAs (siJCT) target vc873/1151 (the most abundant member in cluster II) and vc7952/8189 (cluster III) at junction sequences. Mismatched sequences in the corresponding JCT mutant control siRNA (siJCT mutant) are shown in dashed lines. siHCV was designed to only target HCV linear RNA due to its complete complementarity to the linear genomic sequences. Due to low abundance of individual cluster I circRNAs, a pool of siRNAs targeting the junctions of vc10/405, vc26/369, vc30/372 and vc25/1277 were used. (B and C) Effects of cluster I, II, and III vcircRNA depletion on HCV infection. Huh7.5.1 cells were transfected with each indicated siRNA at a final concentration of 100 nM overnight and were further infected with HCV JFH-1 strain at a multiplicity of infection of 0.5. Three days later, HCV RNA abundance at HCV Core and NS5B regions in cell lysates (B) were quantified by qRT-PCR. Infectious viral titers in cell culture supernatant (C) were measured by focus forming unit (FFU)…