The influenza virus NP protein also regulates the switch from viral mRNA to full-length (+) strand synthesis (Fig. 6.11). The RdRP for genome replication reads through the polyadenylation and termination signals for mRNA production only if NP is present. This protein is thought to bind nascent (+) strand transcripts and block poly(A) addition by a mechanism analogous to that described for vesicular stomatitis virus N protein. Copying of (+) strand RNAs into (−) strand RNAs also requires NP protein. Intracellular concentrations of NP protein are therefore an important determinant of whether mRNAs or full-length (+) strands are synthesized.
Figure 6.22 Moving-template model for influenza virus mRNA synthesis. During RNA synthesis, the RdRP remains bound to the 5′ end of the genomic RNA, and the 3′ end of the genomic RNA is threaded through the RdRP as the PB1 protein catalyzes each nucleotide addition to the growing mRNA chain. This threading process continues until the mRNA reaches a position on the genomic RNA that is close to the binding site of the polymerase. At this point, the RdRP itself blocks further mRNA synthesis, and reiterative copying of the adjacent U7 tract occurs. After about 150 A residues are added to the 3′ end of the mRNA, mRNA synthesis terminates.
Ambisense RNA
Although arenaviruses are considered (−) strand RNA viruses, their genomic RNA is in fact ambisense: mRNAs are produced both from (−) strand genomic RNA and from complementary full-length (+) strands. The arenavirus genome comprises two RNA segments, S (small) and L (large) (Fig. 6.23). Shortly after infection, RdRP that enters with viral particles synthesizes mRNAs from the 3′ region of both RNA segments. Synthesis of each mRNA terminates at a stem-loop structure. These mRNAs, which are translated to produce the nucleocapsid (NP) protein and RdRP (L), respectively, are the only viral RNAs made during the first several hours of infection. Later in infection, the block imposed by the stem-loop structure is overcome, permitting the synthesis of full-length S and L (+) strand RNAs. It was initially thought that melting of the stem-loop structure by the NP protein allowed the transcription termination signal to be bypassed. It now seems more likely that two different configurations of the RNA polymerases are made in infected cells: one for synthesis of mRNA and a second for synthesis of full-length copies of the genome. The finding that viral mRNAs are capped while genomes are not is consistent with this hypothesis.
Double-Stranded RNA
A distinctive feature of the infectious cycle of double-stranded RNA viruses is the production of mRNAs and genomic RNAs from distinct templates in different viral particles. Because the viral genomes are double stranded, they cannot be translated. Therefore, the first step in infection is the production of mRNAs from each viral RNA segment by the virion-associated RdRP (Fig. 6.24). Reoviral mRNAs carry 5′ cap structures but lack 3′ poly(A) sequences.
Figure 6.23 Arenavirus RNA synthesis. Arenaviruses contain two genomic RNA segments, L (large) and S (small) (top). At early times after infection, only the 3′ region of each of these segments is copied to form mRNA: the N mRNA from the S genomic RNA and the L mRNA from the L genomic RNA. Copying of the remainder of the S and L genomic RNAs may be blocked by a stem-loop structure in the genomic RNAs. After the S and L genomic RNAs are copied into full-length strands, their 3′ regions are copied to produce mRNAs: the glycoprotein precursor (GP) mRNA from S RNA and the Z mRNA (encoding an inhibitor of viral RNA synthesis) from the L RNA. Only RNA synthesis from the S RNA is shown in detail.
Figure 6.24 mRNA synthesis and replication of double-stranded RNA genomes. These processes occur in subviral particles containing the RNA templates and necessary enzymes. During cell entry, the virion passes through the lysosomal compartment, and proteolysis of viral capsid proteins activates the RNA synthetic machinery. Single-stranded (+) viral mRNAs, which are synthesized in parental subviral particles, are extruded into the cytoplasm, where they serve either as mRNAs or as templates for the synthesis of (−) RNA strands. In the latter case, viral mRNAs are first packaged into newly assembled subviral particles in which the synthesis of (−) RNAs to produce double-stranded RNAs occurs. These subviral particles eventually become infectious particles. Only 1 of the 10 to 12 double-stranded RNA segments of the reoviral genome is shown.
In the reovirus core, the λ3 polymerase molecules are attached to the inner shell at each fivefold axis, below an RNA exit pore. Viral mRNAs are synthesized by the polymerase inside the subviral parental core and then extruded into the cytoplasm through this pore. Attachment of the polymerase molecules to the pores ensures that mRNAs are actively threaded out of the particle, without depending upon diffusion, which would be very inefficient. Examination of the structure of an actively transcribing human rotavirus, a member of the Reoviridae, has allowed a three-dimensional visualization of how mRNAs are released from the core particle (Box 6.4). Viral (+) strand RNAs that will serve as templates for (−) strand RNA synthesis are first packaged into newly assembled sub-viral particles (Fig. 6.24). Each (+) strand RNA is then copied just once within this particle to produce double-stranded RNA.
Members of different families of double-stranded RNA viruses carry out RNA synthesis in diverse ways. Replication of the genome of bacteriophage ϕ6 (3 RNA segments) and birnaviruses (2 RNA segments) is semiconservative, whereas that of reoviruses (10 to 12 RNA segments) is conservative: only one of the two strands is copied. During conservative replication, the double-stranded RNA that exits the polymerase must be melted, so that the newly synthesized (+) strand is released and the template (−) strand reanneals with the original (+) strand. In reovirus particles, each double-stranded RNA segment is attached to a polymerase molecule, by interaction of the 5′ cap structure with a cap-binding site on the RdRP. Attachment of the 5′ cap to the polymerase facilitates insertion of the 3′ end of the (−) strand into the template channel. This arrangement allows very efficient reinitiation of RNA synthesis in the crowded core of the particle. The RdRPs of bacteriophage ϕ6 and birnaviruses do not have such a cap-binding site, as would be expected for enzymes that copy both strands of the double-stranded RNA segments. This strategy appears less efficient, but may be sufficient when the genome consists of only two or three double-stranded RNA segments.
Unique Mechanisms of mRNA and Genome Synthesis of Hepatitis Delta Virus
The strategy for synthesis of the (+) strand RNA genome of hepatitis delta virus is very unusual among animal viruses (Fig. 6.25). The genome does not encode an RdRP: viral RNAs are produced by host cell RNA polymerase II, and the hepatitis delta virus RNAs are RNA catalysts, or ribozymes (Box 6.5). The genome of hepatitis delta virus is a 1,700-nucleotide (−) strand circular RNA. As approximately 70% of the nucleotides are base paired, the viral RNA is folded into a rod-like structure.