An RNA stem-loop within the bovine coronavirus nsp1 coding region is a cis-acting element in defective interfering RNA replication

Abstract

Higher-order cis-acting RNA replication structures have been identified in the 3'- and 5'-terminal untranslated regions (UTRs) of a bovine coronavirus (BCoV) defective interfering (DI) RNA. The UTRs are identical to those in the viral genome, since the 2.2-kb DI RNA is composed of only the two ends of the genome fused between an internal site within the 738-nucleotide (nt) 5'-most coding region (the nsp1, or p28, coding region) and a site just 4 nt upstream of the 3'-most open reading frame (ORF) (the N gene). The joined ends of the viral genome in the DI RNA create a single continuous 1,635-nt ORF, 288 nt of which come from the 738-nt nsp1 coding region. Here, we have analyzed features of the 5'-terminal 288-nt portion of the nsp1 coding region within the continuous ORF that are required for DI RNA replication. We observed that (i) the 5'-terminal 186 nt of the nsp1 coding region are necessary and sufficient for DI RNA replication, (ii) two Mfold-predicted stem-loops within the 186-nt sequence, named SLV (nt 239 to 310) and SLVI (nt 311 to 340), are supported by RNase structure probing and by nucleotide covariation among closely related group 2 coronaviruses, and (iii) SLVI is a required higher-order structure for DI RNA replication based on mutation analyses. The function of SLV has not been evaluated. We conclude that SLVI within the BCoV nsp1 coding region is a higher-order cis-replication element for DI RNA and postulate that it functions similarly in the viral genome.

FIG. 1.

Location of cis-replication elements within the partial nsp1 coding region in BCoV DI RNA. A. Structure of the naturally occurring BCoV DI RNA relative to the full-length BCoV genome. The DI RNA is composed of the fused ends of the viral genome as described in the text. The 65-nt leader is illustrated by a filled rectangle. The partial nsp1 sequence in parentheses represents the 5′-proximal 288 nt of the 738-nt nsp1 coding region. The cloned, modified BCoV DI RNA is under control of the T7 RNA polymerase promoter, carries a 30-nt in-frame reporter used for Northern probing, and is named pDrep1. Restriction endonuclease sites used for further engineering as described in this report are depicted. The cloned, modified BCoV N mRNA (mRNA 7) is under control of the same promoter as pDrep1, carries the same reporter, and is named pNrep2. The difference in the RNA transcripts from the two plasmids both linearized at the MluI site is a continuous sequence of 421 nt as depicted. At the right are shown the results of Northern analyses depicting the accumulation of pDrep1 RNA at 48 and 96 h posttransfection and at 24 hpi following first virus passage (VP1), and the absence of accumulation of the pNrep2 RNA. RNA, ∼1 ng of transcript used for transfection. B. Structures of modified pDrep1 that carry the 5′ UTR of mRNA 7, named p77Drep, and the 5′ UTR of the genome but without the 288-nt region of nsp1, named p210Nrep. At the right are Northern analyses depicting the accumulation of wt DI RNA and the absence of accumulation for RNAs from p77Drep and p210Nrep. RNA, ∼1 ng of transcript used for transfection. C. Deletion analysis of the 288-nt partial nsp1 coding region in BCoV DI RNA. 3′-terminal deletions of the partial nsp1 coding region were made as indicated, and RNAs from the mutants were tested for accumulation. The results of Northern analyses at the right illustrate that DI RNAs carrying the first 186 nt of nsp1 coding region accumulated, whereas those with less than 186 nt did not. RNA, ∼1 ng of transcript used for transfection.

FIG. 2.

Stem-loops in the 5′-proximal 288 nt of the 738-nt nsp1 coding region as predicted by Mfold. Named nucleotides are those in BCoV. Unless otherwise indicated, they are also the nucleotides in the other named group 2 coronaviruses and are highlighted in gray. Arrows identify bases in MHV-A59, MHV-2, and MHV-JHM unless otherwise identified by superscipts. The compared group 2 coronaviruses and their superscript designations, if any, are BCoV-Mebus, HCoV-OC43 (o), HCoV-4408, HEV-TN11 (h), ECoV-NC99 (e), MHV-A59 (a), MHV-2 (2), and MHV-JHM (j). The free energies noted refer to the stem-loops in the BCoV RNA.

FIG. 3.

Enzyme structure probing of SLV and SLVI in BCoV. A. Schematic summary of structure probing data. B. Electrophoretic analysis of digestion products in the SLV region from extensions of primer 1. Lanes 1 through 4, sequencing ladder generated from primer 1; lanes 5 through 8, RNase CV₁ digestion with 1.0, 0.1, 0.05, and 0.01 U, respectively; lanes 9 through 11, RNase T₂ digestion with 5.0, 1.0, and 0.1 U, respectively; lane 12, undigested RNA. C. Electrophoretic analysis of digestion products in the SLVI region from extensions of primer 2. Lanes 1 through 4, sequencing ladder generated from primer 2; lanes 5 through 7, RNase CV₁ digestion with 0.1, 0.01, and 0.01 U, respectively; lanes 8 and 9, RNase T₂ digestion with 1.0 and 0.1 U, respectively.

FIG. 4.

Effects of SLVI mutations on accumulation of BCoV DI RNA. A. Mutations used to test the cis-requirement of SLVI for DI RNA replication. B. Northern analysis to measure the accumulation of wt and SLVI mutant DI RNAs. The ³²P-labeled oligonucleotide probe specific for the 30-nt TGEV reporter sequence in the DI RNAs was used. RNA, ∼1 ng of transcript used for transfection. C. Translation of wt and mutant DI RNAs. T7 RNA polymerase-generated transcripts of wt and mutant DI RNA were translated in rabbit reticulocyte lysate in the presence of ³⁵S-labeled methionine, and the dried gel was autoradiographed. Lane 1, high-molecular-weight Rainbow marker proteins (Amersham); lane 6, translation product from a reaction mixture to which no RNA was added.