Positive strand RNA viruses differ in the constraints they place on the folding of their negative strand

Abstract

Genome replication of positive strand RNA viruses requires the production of a complementary negative strand RNA that serves as a template for synthesis of more positive strand progeny. Structural RNA elements are important for genome replication, but while they are readily observed in the positive strand, evidence of their existence in the negative strand is more limited. We hypothesized that this was due to viruses differing in their capacity to allow this latter RNA to adopt structural folds. To investigate this, ribozymes were introduced into the negative strand of different viral constructs; the expectation being that if RNA folding occurred, negative strand cleavage and suppression of replication would be seen. Indeed, this was what happened with hepatitis C virus (HCV) and feline calicivirus (FCV) constructs. However, little or no impact was observed for chikungunya virus (CHIKV), human rhinovirus (HRV), hepatitis E virus (HEV), and yellow fever virus (YFV) constructs. Reduced cleavage in the negative strand proved to be due to duplex formation with the positive strand. Interestingly, ribozyme-containing RNAs also remained intact when produced in vitro by the HCV polymerase, again due to duplex formation. Overall, our results show that there are important differences in the conformational constraints imposed on the folding of the negative strand between different positive strand RNA viruses.

Keywords: double-stranded RNA; positive strand RNA virus; replication; replication intermediate; ribozyme.

FIGURE 1.

Placing the HdV Rbz in the [−] RNA of HCV cleaves this RNA and suppresses replication. (A) Schematic depiction of the HCV gt2a replicon encoding the HdV Rbz in its [−] RNA. Image includes the region of both [−] and [+] strands recognized by the strand-specific probes used for northern blotting. (B) Replication of constructs carrying either an active (wt) or inactive (ko) HdV Rbz in their [−] RNA. Included are replication-competent and replication-defective controls lacking the inserted Rbz sequence. Significant differences between HdV(ko) and (wt) constructs are highlighted ([*] P < 0.05; paired t-test; n = 6). (C) Northern blot of RNA from transfected cells. The arrow highlights the position of full length transcripts and the arrow heads the position of products produced as a result of Rbz activity.

FIGURE 2.

Hammerhead Rbzs are highly effective at cleaving the [−] RNA of HCV replicon constructs. (A) Schematic depiction of the HCV gt2a replicons used in this experiment. (B) Replication of constructs carrying either an active (wt) or inactive (ko) sTRSV or N79 hammerhead Rbz in their [−] RNA. Included are replication-competent and replication-defective controls lacking the inserted Rbz sequence. Significant differences between sTRSV(ko) and (wt) constructs, and between N79(ko) and (wt) constructs, are highlighted ([*] P < 0.05; paired t-test; n = 3). (C) Northern blot of RNA from cells transfected with replicon constructs 48 h earlier. The arrow highlights the position of full length transcripts and the arrowheads the position of products produced as a result of Rbz activity. Other bands on the gel (*) coinicident with the position of ribosomal RNAs, represent background artifacts.

FIGURE 3.

Rbzs embedded in the [−] RNA of HCV suppress replication irrespective of the position they are located at. (A) Schematic depiction of HCV gt2a replicons carrying two copies of the N79 Rbz positioned at two different locations within their [−] RNA. (B) Replication data from the constructs depicted in A as well as from replication-competent and replication-defective controls lacking the inserted Rbz sequence. Significant differences between the (ko/ko) construct and other 2× N79 containing constructs are highlighted ([*] P < 0.05; one-way ANOVA; n = 3). No significant differences were observed between other 2× N79 experimental groups.

FIGURE 4.

Embedding Rbzs in the [−] RNA of a diverse set of positive strand RNA virus constructs results in different replicative outcomes. A schematic depiction showing the positioning of the reverse complemented N79 Rbz or sTRSV Rbz sequence is provided for an (A) HCV gt1b replicon, (B) FCV virus, (C) YFV replicon, (D) CHIKV replicon, (E) HRV replicon, and (F) HEV replicon. The regions ΔE and ΔVP1 in the YFV and HRV schematics represent the carboxy-terminal ends of the E protein and VP1 protein, needed for proteolytic processing at the amino terminus of the NS1 and P2 boundaries, respectively. Also shown are luciferase replication data from those same isolates, and in the case of the FCV virus experiments representative images illustrating differences in cpe development over time. The number of experimental repeats for each luciferase assay is indicated next to the respective graph. Significant differences between inactive (ko) and active (wt) Rbz constructs are highlighted ([*] P < 0.05; paired t-test).

FIGURE 5.

The N79 Rbz is not restricted in its ability to cleave the [−] RNA of positive strand RNA virus constructs when this RNA is in a single-stranded state. (A) Single-stranded [−] RNAs from replicons containing a reverse complemented N79 sequence were generated by in vitro transcription and their cleavage was assessed by gel electrophoresis and image capture. Significant differences between experimental groups are highlighted ([*] P < 0.05; one-way ANOVA; n = 3). (B) RNAs from cells transfected 24 h earlier with YFV or HRV replicons carrying a reverse complemented N79 sequence were collected. They were subsequently subject to treatment designed to release the [−] RNA from its double-stranded state and enable it to fold before Mg²⁺ was added to activate Rbz activity. Experimental groups lacking Mg²⁺ addition were included as controls. Cleavage was assessed by northern blot analysis . The arrows represent the position of full length transcripts and arrowheads the position of N79-cleaved product.

FIGURE 6.

Rbz-containing RNAs are not cleaved when synthesized in vitro by the HCV RNA polymerase NS5B due to extensive base-pairing with the template strand. (A) Schematic depicting how NS5B and T7 polymerase transcription reactions were used to produce an N79 Rbz containing RNA. Templates were such that both sets of reactions were expected to transcribe identical RNAs, either encoding an active N79 Rbz sequence or one with a single nucleotide substitution in the active site [N79(wt) and N79(ko) respectively]. The expected sizes of RNAs produced, whether cleaved by Rbz activity or not, are shown. (B) All four [α-³²P] labeled RNAs produced from the transcription reactions outlined in A were subjected to a series of treatments as detailed in the provided diagram before being run on a denaturing polyacrylamide gel. The high-salt RNaseA treatment step was used to selectively degrade single-stranded RNAs. The heat denaturation/renaturation step was used to melt dsRNA and allow single-stranded RNA folding. The asterisks indicate RNAs produced as a result of Rbz cleavage. Additional smaller bands in the NS5B transcription reactions likely arise from internal initiation. Results shown are representative of one of two experiments where initial transcription reactions were carried out at room temperature.

Morgan Herod

Joseph Ward