Evidence that hPIV2 paramyxovirus antigenomes are edited during infection

Abstract

Mononegavirus genomes contain cis-acting sequences that direct the formation of mRNA 5' and 3' ends during synthesis (gene start [GS] and gene end [GE]) but are silent during antigenome synthesis from the same genome template. Paramyxo- and filoviruses also carry an additional cis-acting sequence (EDIT) that directs the insertion of specific nucleotide during mRNA synthesis, expanding gene products. This "mRNA editing" is associated with viruses whose replication is governed by the rule of six. One might, therefore, assume that EDIT, like GS and GE, would be restricted to that of mRNA synthesis. However, accurate determination of the ratio of edited to unedited RNAs in different sub-populations of wild-type and mutant human parainfluenza virus type 2 (hPIV2) infections finds evidence that antigenomes are also edited during infection. This alters our view of how the rule of six governs mononegavirus infections and why this rule, RNA editing, and bipartite promoters are linked.

Importance: We have assumed that paramyxovirus editing signals would operate strictly during mRNA synthesis, as it apparently makes no sense to edit antigenomes. Nevertheless, there is evidence here that the opposite is the case. If so, this alters our view of paramyxovirus replication, and we summarize what is known about how its RNA-dependent RNA polymerase carries out its task of expressing alternate open reading frames during mRNA synthesis.

Keywords: RNA editing; paramyxovirus; rule of six; viral replication.

Fig 1

Paramyxovirus bipartite promoter model of the first turn of the NC helix with 13 subunits per turn. The NP subunits are shown as rectangles numbered from the 3′-OH end, each binding precisely 6 nt. The relevant nt sequences of promoter element 1 (PE1) at the 3′ end and the tripartite PE2 in the NP subunits on the next helical turn (and whose bases point toward the solvent) are shown for hPIV2 (black) and SeV (red). Q202 of each subunit can contact the 3′-most base of each hexa-nt, including the genome’s very 3′ uridine, and this last interaction is proposed to prevent NP dissociation from PE1, a prerequisite for its entry into the synthesis chamber in pol’s core for RNA synthesis to initiate. When glutamine is present at position 202, RNA synthesis can initiate, but only if pol can simultaneously interact with a properly aligned tripartite PE2 to reverse the Q202-3′ uridine interaction.

Fig 2

Amplicon sequencing of the editing region of hPIV2 P/V mRNA, viral (+)RNA and (−)RNA. (A) Schematic diagram of the hPIV2 genome and primers for reverse transcription of each viral RNA in wild type (wt) and IQ-infected Vero cells. (B) Scheme for the reverse transcription reaction. PCR amplification was performed for the area containing the P/V gene editing region in mRNA, (+)RNA, and (−)RNA. The sequence of the PCR amplicon is shown, and the RNA editing region is highlighted in yellow. (C) G count in viral RNAs within the infected cells; the number of guanosines in the RNA editing region (left panel) and the magnified graph (right panel). G + 1, 2, and 3 indicate the number of additional Gs over the unedited seven Gs. (D) G count of the viral genome within the purified progeny virions. Data represent means and standard deviations (n = 3). *P < 0.05 and **P < 0.01 (wt vs IQ, a two-tailed unpaired Student’s t-test).

Fig 3

Quantification of NP, P/V, M, F mRNAs and NP-P, P-M, M-F readthrough mRNAs in virus-infected cells. Vero cells were infected with rPIV2^wt or rPIV2^IQ at an MOI of 0.1 for 48 h. Total RNA was extracted from the infected cells, and copy numbers of the NP, P/V, M, F mRNAs and NP-P, P-M, M-F readthrough mRNAs were measured by RT-qPCR. Data represent means and standard deviations (n = 3). **P < 0.01. A comparison between P mRNA and P-M readthrough mRNA in rPIV2^wt and rPIV2^IQ reveals ratios of 1/44 and 1/26, respectively.

Fig 4

Growth of rPIV2^wt and rPIV2^IQ in Vero, HeLa, and A549 cells. Vero, HeLa, and A549 cells were infected with rPIV2^wt or rPIV2^IQ at an MOI of 0.1. The supernatants were then sampled at various times post-infection and titered on Vero cells. Data represent means and standard deviations (n = 3). *P < 0.05 and **P < 0.01 (wt vs IQ, a two-tailed unpaired Student’s t-test).